SEAFCO is committed to continue its research-minded initiation to link

theory and practice in deep foundations. We hope that researched works

and findings presented in this book will be useful and serve as a source
of reference to all in the field of geotechnical and foundation engineering,
particularly those who are involved in the construction industry of Thailand.

40th Anniversary Celebration of Seafco Public Company Limited

In Thailand, according to the authors experience as a contractor, development

in both construction and design aspects of wet-processed deep-seated
bored piles, barrette foundations and diaphragm walls in past four decades
are significant. With recognition of technical and economic advantages of
using these high capacity cast-in-place foundations by local practitioners,
it is expected that they will be more popular in the future construction industry
of Thailand and the region. However, in the authors opinion, there is much
work to be done with particular focus on constructability issues, concrete
technology for wet-processed bored piles and barrette, reliable but cost-effective
quality control testing and application of value-engineering in design and
construction. Starting from the planning stage, site investigation, design,
construction and inspection should be integrated to enable designers,
contractors and construction inspectors to participate as a team with
a common goal. Appropriate and practical specifications should be established
jointly by these parties for local soil conditions and construction methods.
Continuing education should be promoted for designers, inspection engineers,
and contractors.

40th Anniversary Celebration of Seafco Public Company Limited

19 December 2014

Current Practices in Deep Foundations and Diaphragm Wall Construction in Thailand

Current Practices in Deep Foundations and Diaphragm Wall Construction

in Thailand is published to mark the auspicious occasion of SEAFCO
40th Anniversary. It is a compilation of 48 selected published papers in the
international conferences. The case studies and researched works presented
in these papers are mainly from the projects carried out by SEAFCO in
Bangkok and other parts of Thailand. Published papers authored by SEAFCOs
geotechnical engineers in other major projects and research studies overseas
are also included.

Seafco Public Company Limited

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510

TEL : (662) 919-0090 - 7, FAX : (662) 919-0098
www.seafco.co.th

WE BUILD MODERN FOUNDATIONS

2014

Current Practices in Deep Foundations

and Diaphragm Wall Construction in Thailand
Editors:
Zaw Zaw Aye
Aung Win Maung
Thayanan Boonyarak
Chalermpol Taechakumthorn
Nutthachai Prongmanee
Nutthapon Thasnanipan

40th Anniversary Celebration of Seafco Public Company Limited

19 December 2014

Current Practices in Deep Foundations

and Diaphragm Wall Construction in Thailand
Editors:
Zaw Zaw Aye
Aung Win Maung
Thayanan Boonyarak
Chalermpol Taechakumthorn
Nutthachai Prongmanee
Nutthapon Thasnanipan
Member of

PREFACE
Seafco Public Company Limited is a deep foundation contractor established in 1974. The company
has been involved in numerous foundation and deep excavation projects in Thailand, including
large-scale foundation works, infrastructure projects, and more recently the underground structures
of Bangkoks second subway project (MRT Blue Line Extension Project). SEAFCO expanded its
operation overseas and successfully took part in the construction of diaphragm walls for Marina
Bay Sands in Singapore in 2007. SEAFCO has also established Seafco (Myanmar) Company
Limited in late 2013 and completed two projects to date.
Consolidating state-of-the-art technology, sophisticated equipment and engineering professionals,
SEAFCO initiated the innovative techniques and solutions in geotechnical and foundation
engineering in Thailand and Myanmar. To improve the currently used methods, material, system,
design concepts and construction techniques, SEAFCO has spent considerable time and effort
on Research and Development activities.
SEAFCOs R&D team continuously conducts research on bored piles, barrettes, diaphragm
walls, associated testing methods, deep excavation and other geotechnical engineering works.
New findings and research results have been presented in a number of conferences and
seminars organized by various institutions both domestically and internationally. In forty years,
R & D Division of SEAFCO has produced over 100 technical papers (both in English and Thai)
which were presented and published in international conferences held in various countries. The
purpose of this publication is to mark the auspicious occasion of SEAFCO 40th Anniversary. It is a
compilation of selected published papers in the international conferences. The case studies and
research works presented in these papers are mainly from the projects carried out by SEAFCO
in Bangkok and other parts of Thailand. Published papers authored by SEAFCOs employees in
other major projects and research studies overseas are also included.
I would like express my appreciation to the clients, consultants, designers, universities, institutions,
the Engineering Institute of Thailand under H.M. the Kings Patronage (EIT) and partners who have
been continuously supporting SEAFCO for 40 years. I also would like to acknowledge the effort of
SEAFCOs R&D team for their commitment and hard work in the research works.
I hope that researched works and findings contained in these publications will be useful and serve
as a source of reference to all in the field of geotechnical and foundation engineering, particularly
those who are involved in the construction industry of Thailand.

Narong Thasnanipan
President & CEO
Seafco Public Company Limited

FOREWORD
First of all, I would like to extend my congratulations to SEAFCO for 40 years of successful
establishment, not only as a business entity but also as a great contributor for construction industry
of Thailand.
SEAFCO is well known for research minded initiatives in foundation engineering. SEAFCO shares
outcomes of its R&D works to the engineering society by publishing technical papers, organizing
conferences, and delivering lectures in various institutions in both Thailand and overseas. I
proudly acknowledge the contributions SEAFCO has been making to strengthen the foundation of
civil engineering society in Thailand and beyond.
Since the time I was working on my doctoral research, I have found SEAFCOs publications
in various international conferences very useful to understand practical aspects of deep
foundations. SEAFCO puts the innovative theory into practice by equipping itself with knowledge,
experience and professionalism. SEAFCO R&D team has put in commendable effort to compile
all of their published papers and produced the book Current Practices in Deep Foundations
and Diaphragm Wall Construction in Thailand. Geotechnical specialists, foundation engineers,
structural engineers and practitioners in construction industry will find this book beneficial.
I, as the President of the Engineering Institute of Thailand under H.M. The Kings Patronage (EIT),
would like to sincerely thank Seafco Public Company Limited for endless support they have been
providing to the various activities of EIT under the leadership of CEO Khun Narong Thasnanipan. In
particular, SEAFCOs great support in a leading role to Thailand Underground & Tunneling Group
(TUTG) in winning the bid for Thailand to host the World Tunnel Congress 2012 and successfully
organizing the event at Queen Sirikit National Convention Center in Bangkok from 18 to 23 May
2012.
I believe SEAFCO will continue to grow in business with trust and confidence of its stakeholders.
I am also confident that SEAFCO will maintain its position as the leading innovators of foundation
and geotechnical engineering.
I wish SEAFCO all the success.

Prof. Dr. Suchatvee Suwansawat

President
The Engineering Institute of Thailand

Table of Contents

Bored Pile and Barrette: Construction Methods, Design,

Analysis and Performance
Barrettes Founded in Bangkok Subsoils, Construction and Performance
Thasnanipan N., Maung A.W. & Tanseng P. .................................................................. 1
Design, Construction and Behavior of Bored Cast In-Situ Concrete Piles
in Bangkok Subsoil
Thasnanipan N., Teparaksa W., Maung A.W. & Ganeshan B. ........................................ 9
Large Diameter Bored Piles in Multi-Layered Soils of Bangkok
Thasnanipan N., Tanseng P. & Anwar M.A. .................................................................. 19
Effect of Construction Time and Bentonite Viscosity on Shaft Capacity
of Bored Piles
Thasnanipan N., Ganeshan B. & Anwar M.A. .............................................................. 29
Performance Comparison of Bored and Excavated Piles in the Layered Soils
of Bangkok
Thasnanipan N., Anwar M. A., Maung A.W. & Tanseng P. .......................................... 39
Base grouting of wet process bored piles in Bangkok subsoils
Teparaksa W., Thasnanipan N. & Anwar M. A. ............................................................. 51
Review of the Shaft Capacity Degradation of Bored Piles Constructed
with Bentonite Slurry
Thasnanipan N., Anwar, M. A. & Maung A. W. ............................................................. 57
Failure Mechanism of Long Bored Piles in Layered Soils of Bangkok
Thasnanipan N., Anwar, M. A. & Maung A. W. ............................................................ 69
Barrettes : A Versatile Foundation for Transmission Line Towers
Thasnanipan N., Tanseng P., Maung A.W. & Anwar M. A. ............................................ 77
P ractical Installation of Stanchions for Top-Down Construction
in Bangkok Subsoil
Thasnanipan N., Maung A. W. & Z. Z. Aye ................................................................... 85
Concrete for Wet Process Bored Piles
Thasnanipan N., Maung A. W., Tanseng P. & Navaneethan T. .................................... 97

Record Load Test on a Large Barrette and its Performance in the layered
soils of Bangkok
Thasnanipan N., Maung A. W. & Z. Z. Aye ................................................................. 107
Performance of Wet-Process Bored Piles Constructed with
Polymer-Based Slurry in Bangkok Subsoil
Thasnanipan N., Z. Z. Aye, Submaneewong C. & Teparaksa W. ............................... 117
Barrette of Over 50,000 kN Ultimate Capacity Constructed in
the Multi-Layered Soil of Bangkok
Thasnanipan N., Z. Z. Aye & Teparaksa W. ................................................................ 135
Performance of Toe-grouted Large Bored Piles in Sand-Gravel
Formation of Chiang Mai City, Northern Thailand
Thasnanipan N., Z. Z. Aye & Teparaksa W. ................................................................ 145
Application of Polymer-based Slurry for Wet-process Bored Piles
Construction in Mylti-layered Soil of Bangkok
Z. Z. Aye, Singtogaw K. & Submaneewong C. ........................................................... 155
Development in Construction of Bored pile and Barrette for
Infrastructure Projects in Thailand, A Country Report
Thasnanipan N., Singtogaw K., Z. Z. Aye & Pravesvararat S. .................................... 163
Behaviour of Polymer-based Slurry for Deep-seated Bored Piles
in Multi-layered Soil of Bangkok
Thasnanipan N., Z. Z. Aye & Boonyarak T. ................................................................. 175
Effectiveness of Toe-Grouting for Deep-Seated Bored Piles
in Bangkok Subsoil
Thasnanipan N., Z. Z. Aye & Submaneewong C. ....................................................... 185
C urrent Practice and Future Trends of Cast-in-place Deep Foundation
in Thailand
Thasnanipan N., Singtogaw K., Z. Z. Aye & Pravesvararat S. ................................... 199
Development and Achievement of Deep-seated Bored Piles and
Barrettes Construction in Thailand in the Past Forty Years,
A Country Report
Thasnanipan N., Z. Z. Aye, Submaneewong C. & Boonyarak T. ................................. 211
Barrette for High-rise Buildings and Heavy Structures in Bangkok
Sub-soil Thasnanipan N., Z. Z. Aye, Boonyarak T., Kitpayuck P.
& Thasnanipan N. ...................................................................................................... 223

Diaphragm Wall and Deep Excavation

Performance of a Braced Excavation in Bangkok Clay, Diaphragm Wall
Subject to Unbalanced Loading Conditions
Thasnanipan N., Maung A.W., Tanseng P. & Wei S.H. ............................................... 245
Prediction and Performances of Short Embedded Cast In-Situ Diaphragm
Wall for Deep Excavation in Bangkok
Thasnanipan N., Teparaksa W., Maung A.W. & Wei S.H. ............................................ 253
Lessons from the collapse during construction of an inlet pumping station:
Geotechnical instrumentation aspect
Teparaksa W., Thasnanipan N., Maung A. W. & Tanseng P. ....................................... 263
Behavior and performance of diaphragm walls under unbalanced
lateral loading along the Chao Phraya River,
Thasnanipan N., Maung A. W. and Tanseng P. & Teparaksa W. ............................... 271
Analysis of Lateral Wall Movement for Deep Braced Excavation of
Bangkok Subsoils
Teparaksa W., Thasananipan N., & Tanseng P. ....................................................... 279
Diaphragm Wall and Barrette Construction for Thiam Ruam Mit Station Box,
MRT Chaloem Ratchamongkhon Line, Bangkok
Thasnanipan N., Maung A.W. & Ganeshan B. ............................................................. 291
Monitoring of Diaphragm Wall Displacement and Associated Ground
Movement, Braced Excavation Adjacent to Historical Building at the Bank
of Chao Phraya River
Tanseng P., Z. Z. Aye & Submaneewong C. ............................................................... 299
Performance of Buttress-Support Thin Diaphragm Wall for Underground
Car Park in Bangkok
Thasnanipan N., Z. Z. Aye & Submaneewong C. ........................................................ 309
Stability of a Trial Trench Excavated under Polymer Slurry in Bangkok
Soft Clay
Thasnanipan N., Maung A. W., Boonyarak T. & Aye, Z. Z. .......................................... 315
Application of Top-down Construction Method for Deep Excavations
in Urban Area of Bangkok
Thasnanipan N., Z. Z. Aye, Submaneewong C. & Boonyarak T. ................................. 321

Construction of diaphragm wall for basement excavation adjacent to

tunnels in Bangkok subsoil
Thansnanipan N., Maung A. W., Z. Z. Aye, Submaneewong C.
& Boonyarak T. ........................................................................................................... 337
Application of Observational Method in Diaphragm Wall Support
Basement Excavation
Z. Z. Aye, Submaneewong C., Boonyarak T. & Chanchad C. ...................................... 351
Diaphragm Wall Support Deep Excavation in Bangkok
Thasnanipan N., Z. Z. Aye, Maung A. W. Boonyarak T., Kitpayuck P.&
Thasnanipan N. ........................................................................................................... 369
Underpass Construction in Congestive Area of Bangkok
Z. Z. Aye, Boonyarak T., Kitpayuck P., Mahasantipiya S. & Thasnanipan N. ...................... 387

Quality Control Tests on Deep Foundations

Sonic Integrity Test on Piles Founded in Bangkok Subsoil-Signal
Characteristics and Their Interpretations
Thasnanipan N., Maung A.W. & Ganeshan B. ........................................................... 405
Sonic Integrity Test of Piles-integrity Effected by Basement Excavation
in Bangkok Soft Clay
Thasnanipan N., Maung A. W., Tanseng P. & Z. Z. Aye ............................................. 415
Non-Destructive Integrity Testing on Piles Founded in Bangkok Subsoil
Thasnanipan N., Maung A. W. & Navaneethan T. & Z. Z. Aye ..................................... 425
Damages to Piles associated with Excavation Works in Bangkok Soft Clay
Thasnanipan T., Maung A. W. & Tanseng P. ............................................................... 435
Quantifying Pile Head Condition Before Basement Excavation by
Cross-hole Sonic Logging Tests
Thasnanipan N., Z. Z. Aye, Boonyarak T. & Kampananon N. ...................................... 445

Advantages and Limitations of Integrity Tests for Large Diameter Bored

Piles in Bangkok
Thasnanipan N., Z. Z. Aye, Boonyarak T. & Kampananon N. ...................................... 453
Application of Low-strain Integrity Tests for Quality Assurance of
Deep Foundations in Thailand
Maung A. W., Z. Z. Aye & Boonyarak T. ...................................................................... 463

Geotechnical Engineering : Research and Practice

Ground Movement Prediction and Building Damage Risk-Assessment
for the Deep Excavations and Tunneling Works in Bangkok Subsoil
Z. Z. Aye, Karki D. & Schulz C. ................................................................................... 473
Key Factors in Application of Observational Method Bangkok Experience
Phienwej N. & Z. Z. Aye ........................................... ................................................... 495
Ground Movements Associated with the Underground Excavations
of the First Bangkok MRT Line
Phienwej N., Sirivachiraporn A., Suwansawat S. & Z. Z. Aye ...................................... 525
Geotechnical Risk Assessment and Management for Tunnelling Project
in Congestive Urban Area Bangkok Experience
Z. Z. Aye, Kitpayuck P. & Phienwej N. ......................................................................... 535
Tunneling Effect on Pile Group Response in Bangkok
Boonyarak T. & Ng C. W. W. ....................................................................................... 547

SEAFCO PUBLIC CO LTD

SEAFCO

D EEP FOU N D ATI ON S

Barrettes
Strip Piles

When single pile safe

working load in excess of
2000 tons is required or
conventional bored piling
is impractical, barrettes or
strip piles are the best alternative. Various dimensions, cross-sections and
orientation can be formed
and arranged to suit design requirements.
SEAFCO barrettes are
used in foundations for expressways, MRTA stations
and high-rise buildings.

MEMBER OF

A - 13

A - 14

A - 15

A - 16

A - 17

Cast-in-situ bored piles with base

grouting are in common use in
Bangkok for foundations of heavy
structures. Base grouting is usually
employed in bored piles to improve
pile capacity and performance. The
unavoidable problems arising from
loosening of soil and sedimentation
at the base caused by boring can
be minimized by base grouting.
The first base grouted bored pile in
Bangkok was used in 1985 at the
Rama IX cable-stayed bridge
across the Chao Phraya River.
Teparaksa, W. (1994) reported that
base grouted piles in the second
sand layer of Bangkok subsoils
have an increment in their performance as below;

Conditions

Increment %

In Failure Load
In Skin Friction
In End Bearing

12-24
9-27
11-21

The above increments were determined at fully mobilized skin friction

1.59-1.86% and 1.04% of pile diameter for base grouted and nonbase grouted piles respectively.
Mechanism of improvement by
base grouting involves;
Replacement of sediments with
grout under the pile base.
Squeezing water out of the
sediments.
Improving the sand stratum
around the pile base.
Prestressing of the pile.
Increasing skin friction in sand
layers.

GROUTING ASSEMBLY
Current base grouting practice
uses two U-circuits consisting of
12mm diameter PE tubes attached at the bottom with 12mm
GI pipes with perforations
wrapped with rubber sleeves
(manchettes). They are attached
with the reinforcement cage to
place the manchettes at the base
of the pile.

GROUTING CRITERIA
Grout Pressures:
Maximum
pressure developed is a single
most important parameter.
It
should be gradually achieved by
low injection rates to avoid hydraulic fracturing.
Grout Volumes:
Usually
total grout take is not a limiting
criteria and is used to decide
about the staged grouting. Generally, the higher the required
maximum pressure to be
achieved, the higher will be the
grout volume under similar condi-

Injection Rates:
It is not a
direct controlling parameter but
instead used to achieve maximum pressure in a gradual
manner. Injection rates of 510litres/min are proved to be
satisfactory.
Residual Pressures: Normall y
residual pressures of the order
of 20-30% of target maximum
pressure are desirable.

PERFORMANCE OF
GROUTED PILES

BASE

Base grouting with tube-amanchettes technique has been

used by SEAFCO in many projects in Bangkok.
Improvements achieved have been
proved by load tests on both
non-grouted and base-grouted
bored piles on the same project
sites.

N on-G route d v s. B ase -G route d P ile s (S ilom)

L o ad (to n )
0

500

1000

1500

2000

2500

3000

3500

4000

-10

0
-20

10

Settlement (mm)

PILE BASE GROUTING IN

BANGKOK SUBSOILS

N
G

B
G

B
G

Soft Clay
Stiff Clay
Silty Clay

-30

20

Dense Sand

30

-40

40

Hard Clay

-50

50

Dense Sand

60
N o n -G ro u te d P ile - D ia . 1 .5 x6 0 .0 0 m
B a s e -G ro u te d P ile A - D ia .1 .5 x5 5 .1 6 m (3 S ta g e s - to ta l 2 0 0 0 ltr)
B a s e -G ro u te d P ile B - D ia .1 .5 x6 0 .0 0 m (3 S ta g e s - to ta l 2 0 0 0 ltr)

-60

SEAFCO

26/10 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

A - 18

Tel. (662) 919 0090-4, Fax. (662) 919 0098

Clayey Sand

Silty Clay

-30
Sand
Silty Clay
Silty Clay

-50

1000

1500

2000

2500

Sand

3000

Age: 61 days

30

-30

Bas e-Gro u ted Pile - Dia. 1.5x59.72m (2 Stag es - to tal 1500 ltr)

500

Silty Clay

3000

3500

20

30
Non-Grouted Pile - Dia. 1.5x60.2m
Base-Grouted Pile - Dia. 1.5x60.3m (2 Stages - total 1500 ltr)

Non-Grouted vs. Base-Grouted (Silom) - JTC

Load (ton)
0

1500
-20

1000

1500

2000

-30

2500

3000

3500

Age: 115 days

10

Dense Sand

10

500

Stiff Clay

20
30

20

-40
Desne Sand

30

-50

40

-60

40
Age: 60 days

50
Silty Clay

Non-Grouted Pile - Dia. 1.0x53.0m

Bas e-Grouted Pile - Dia. 1.0x55.0m (500 ltr 1 Stage)

SEAFCO

2500

Age: 82 days

Sandy Clay

Dense Sand

-60

2000

Age: 62 days

-10

1000

Settlement (mm)

-50

1500

10

Soft Clay

Load (ton)
0

Hard Clay

Hard Clay

Sand

-20

-40

1000

-60

No n -Gro u ted Pile - Dia. 1.5x60.2m

Non-Groute d vs. Ba se -Groute d Pile s (Ra m a III)

-30

Load (ton)
500

-50

70

Stiff Clay

Base-Grouted Pile - Dia. 1.2x55.90m (Virgin Soil)

Age: 57 days

60

B
G

Base-Grouted Pile - Dia. 1.2x55.87m (Grouted Soil)

Stiff Clay

-40

50

Soft Clay

N
G

40
Silty Clay

-10

40

Sandy Clay

-10

30

Dense Sand

-20

20

80

-60

20

Settlement (mm)

-50

Settlement (mm)

Hard Clay

2000

Non-Grouted vs. Base-Grouted (V-Rangsit) -CNW

10

-40

1500

load settlement curves of base-grouted piles, before

and after all adjacent piles have been base-grouted.

-20

1000

10

Settlement (mm)

Stiff Clay

500

500

Soft Clay

Lo ad (ton)

-10

-30

-60

N on-G route d v s. B ase -G route d (Prathu-N am)

Soft Clay

Stiff Clay

-20

-40

All rights reserved. No part of this publication may be reproduced in any

form or by any means, without the prior written permission of SEAFCO CO.,
LTD.

Base-Grouted Piles - CP
Load (ton)

Soft Clay

-10

Settlement (mm)

Generally, total volume of grout at pile base

ranged from 500 liters to 4000 liters at single or
multiple stage grouting. Grouting pressure ranges
from 25 bars to 60 bars depending on the depth of
pile base level and relative density of sand layer in
the project area.

60
Dense Sand

Non-Grouted Pile - Dia. 1.5x63.72m

Base-Grouted Pile - Dia. 1.5x63.72m (2 Stages - total 1000 ltr)

References:
Teparaksa, W., Thasnanipan, N. and Anwar M. A. (1999), Base Grouting of Wet Process Bored Piles in Bangkok
Subsoils, Eleventh Asian Regional Conference, International Society for Soil Mechanics and Geotechnical Engineering, Seoul, Korea, August 16-20, 1999. Pp. 269-272.
Lui, J. and Gao W. S. (1999), The Behavior of Post-grouting Pile Groups under Vertical Load, Eleventh Asian Regional Conference, International Society for Soil Mechanics and Geotechnical Engineering, Seoul, Korea, August 1620, 1999. Pp. 209-212.
Thasnanipan, N. (1997), Current Development of Bored Piling Techniques in Bangkok (in Thai), Technical Seminar
on Foundation 1997 organized by the Engineering Institute of Thailand, Bangkok, Thailand, February 4, 1997. Pp. 126.
Teparaksa, W. (1992), Behavior of Base-Grouted Bored Piles in Bangkok Subsoils, Piling: European Practice and
Worldwide Trends. Conference organized by the Institution of Civil Engineers, London, April 7-9, 1992, Thomas Telford. Pp. 296-301.

26/10 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

Tel. (662) 919 0090-4, Fax. (662) 919 0098

A - 19

Object
Contiguous bored pile wall and bored piles for construction of basement and foundation of the Rama
Thibodi Hospital.

Project Description
Bored pile wall with temporary braced excavation
method for construction of 3 level basements for car
parking.

Type of Work
Foundation Piling, Contiguous bored pile wall.

Owner
Layout of contiguous and foundation bored piles.

Rama VI Hospital.

Designer
Arun Chaiseri Consulting Engineers Co., Ltd.

Main Contractor
Ch. Karnchang Public Company Limited.

Project Schedule
2005-2006

Construction Method

Excavation with contiguous pile wall using bottom up construction method to 11.5m.

Construction Details
Foundation bored pile:

Contiguous Pile Wall:

7 (0.8mx55m)
121 (1.0mx55m)
163 (1.2mx55m)
32 (1.5mx55m)
364 (1.0mx20.0m)

Excavation with braced pile wall for construction of footings and basements.

Subsoil Conditions
Soft clay:
Medium clay:
Stiff clay:
Dense sand:
Hard clay:
Very dense sand:

0.0-11.0m
11.0-14.0m
14.0-28.0m
28.0-30.0m
30.0-47.0m
47.0-55.0m

Exposed contiguous bored pile wall.

No. 144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: [email protected],
Homepage: www.seafco.co.th.

A - 20

MahaNakhon

Deep Foundation Works

Project Description

Barrettes, bored piles and diaphragm walls for a 75

storey high-end Hotel and Residence Tower with 2
level basements, and a retail complex with 4 level
basements, located on Narathiwas Road.

Type of Work

Foundation piling and diaphragm walling.

Owner

Pace Project One Co., Ltd., Pace Project Two Co.,

Ltd. and Pace Project Three Co., Ltd.

Project Schedule
2011-2012

Construction Details
Foundation bored pile:

Diaphragm Walls:

Layout of the project.

0.80mx57-65m = 59 +11
1.00mx57-65m = 13+14
1.20mx65.0m = 4
0.80mx3.0mx65.0m = 1
1.20mx3.0mx65.0m = 114
9,264sqm (0.6m to 0.8m
thick x 16.0m to 22.0m)

Subsoil Conditions
Soft clay:
Medium clay:
Stiff clay:
Dense sand:
Hard clay:
Very dense sand:
Hard Clay:
Very Dense Sand

2.0-8.0m
8.0-13.0m
13.0-24.0m
24.0-38.8m
38.8-43.0m
43.0-49.1m
49.1-55.0m
55.0-67.0m

Diaphragm Wall Construction in Progress.

General View of the Site.

Rebar Cage Installation for Diaphragm Wall.

No. 144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: [email protected],
Homepage: www.seafco.co.th.

A - 21

Object
Construction of basement structures of a 42 storey
office tower on 9 storey retail podium.

Project Description
Construction of bored pile foundations, diaphragm
walls and deep excavation works for 2-3 level basements construction.

Type of Work
Foundation piling, diaphragm walling and substructure works.
Layout of the project.

Owner
Land and House Property Company Limited.

Designer
Palmer &Turner (Thailand) Limited.

Project Schedule
2008-2009

Construction Method

Bored piling and diaphragm walling

Deep excavation to 13.0m with diaphragm
walls using bottom up construction method.

Construction Details
Foundation bored pile:

Diaphragm Walls;
Voume of Excavation:
Temporary Bracing:
Temporary Platform:
Concrete Volume:
Steel Reinforcement:
Basement Floor:

0.80mx57m = 22
1.20mx57m = 65
1.50mx57m = 183
1.80mx57m = 25
8,172 sq.m. (0.80m thick
x 18m deep)
95,000 cu.m.
1,987 tons
3,807 sq.m.
24,295 cu.m. (substructure)
4,350 tons (substructure)
30,909 sq.m.

Excavation with braced diaphragm wall for construction of basements.

Subsoil Conditions
Soft clay:
Medium clay:
Stiff to very stiff clay:
Hard clay:
Dense Sand:

2.5-9.0m
9.0-15.3m
15.3-40.0m
40.0-42.8m
42.8-60.0m
Completed substructure.

No. 144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: [email protected],
Homepage: www.seafco.co.th.

A - 22

ASOKE COMPLEX - INTERCHANGE TOWER

Substructure Works
Object

Construction of foundation mats, footings, deep basements for car parking and wastewater treatment
tanks.

Project Description

Diaphragm wall with 2-3 levels temporary bracing for

deep excavation up to 15.90m to construct 3.5 level
basements for car parking.

Type of Work

Deep excavation with temporary bracing, Civil Construction.

Layout of temporary bracing.

Owner

B&G Park Company Limited.

Designer

Palmer and Turner (Thailand) Ltd.

Project Schedule

March 2005 November 2006

Construction Method

Braced excavation to 15.90m with diaphragm

wall using bottom up construction method.

Construction Details
Temporary bracing:
Earth excavation:
Basement floor area:

1,478 tons
54,000cu.m
13,000sq.m.

Rebar fixing for foundation mat in progress.

Subsoil Conditions
Soft clay:
Medium clay:
Stiff clay:
Hard clay:
Very dense sand:

0.0-14.0m
14.0-15.0m
15.0-24.0m
24.0-40.0m
40.0m+

Concrete casting of 3.0m thick foundation mat in progress.

144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: [email protected],
Homepage: www.seafco.co.th.

A - 23

UBC III AND EM2

Deep Foundation Works
Project Description

Bored piles and diaphragm walls for a 32 storey

office tower and shopping complex with 2 to 4 level
basements, located on Sukhumvit Road.

Type of Work

Foundation piling and diaphragm walling.

Owner

Bhiraj Buri Co., Ltd (UBC III) &

Sukhumvit City Mall Co., Ltd. (EM2)

Project Schedule
2011

Construction Details
Foundation bored pile:
Bored Pile Wall
Bored Pile Wall
Foundation bored pile:

Diaphragm Walls:

Layout of the project.

0.60mx18.5m = 126
0.60mx22-23.0m = 433
0.80mx18.5m = 393
0.80mx60.0m = 158
1.00mx60.0m = 30
1.20mx60.0m = 27
1.50mx60.0m = 308
1.80mx60.0m = 140
0.80mx3.00mx60.0m = 4
1.00mx3.00mx60.0m = 22
7,057.5 sqm (0.8-1.0m thick
x18.0-22.0m)

Subsoil Conditions
Soft clay:
Medium clay:
Stiff clay:
Dense sand:
Hard clay:
Dense Sand:

1.5-13.5m
13.5-16.3m
16.3-32.5m
32.5-34.0m
34.0-53.9m
53.9-58.0m

Diaphragm Wall Construction in Progress.

General View of the Site.

Drilling for Large Diameter Bored Piles.

No. 144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: [email protected],
Homepage: www.seafco.co.th.

A - 24

Object
Construction of a 2 level underground car park under
a football field.

Project Description
Construction of bored pile foundations, diaphragm
walls and deep excavation works for 2 level basements and football field on top.

Owner
Srinakrin Wiroj University

Architects & Engineers

Krunthep Tanakom Co., Ltd. (Designer)
Plan Consultants Co., Ltd.
Kumjornkit Construction Co., Ltd. (MC)

Layout of the project (Courtesy of the Designer)

Project Schedule
2008-2010

Construction Method

Bored piling and diaphragm walling

Deep excavation to 7.3m with diaphragm
walls using top-down construction method

Construction Details
Foundation bored pile:
Diaphragm walls;
Excavation volume:
Basement floor:
Concrete volume:
Reinforcement:

0.60mx18.5m = 508
7,664 sq.m. (0.60m thick
x 16m deep)
102,000 cu.m
36,000 sqm
39,624 cu.m
4,218 tons

Excavation with diaphragm wall using top-down construction method for basement construction.

Subsoil Conditions
Soft clay:
Medium clay:
Stiff clay:
Hard clay:
Sand:
Hard clay:
Dense sand:

2.0-14.5m
14.5-17.0m
17.0-22.0m
22.0-41.5m
41.5-46.0m
46.0-49.0m
49.0-60.0m

Substructure works in progress.

No. 144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: [email protected],
Homepage: www.seafco.co.th.

A - 25

Object
Deep foundation work for wastewater treatment plant.

Project Description
Bored piles and diaphragm walls for construction of
deep basements for a wastewater treatment plant in
Bangsu, Jatujak area.

Type of Work
Foundation piling, diaphragm walling and soil improvement.

Owner
The Bangkok Metropolitan Administration.

Layout of pile foundation and diaphragm walls.

Designer & Engineers

The Bangkok Metropolitan Administration.
CH. Karnchang Public Company Limited (MC)

Project Schedule
2008-2009

Construction Details
Foundation bored pile:

0.80mx57.0m= 971
1.00mx57.0m = 7
1.00mx57.0m = 380
1.50mx57.0m = 46

Diaphragm Walls:

2,460 sq.m (1.5m thick x

24.7m deep )
10,225 sq.m (1.2m thick x
24.7m deep )

Soil mixing:

Birds eye view of the project site.

1,734 cu.m

Subsoil Conditions
Soft clay:
Medium clay:
Stiff to very stiff clay:
Dense sand:

2.0-10.0m
10.0-13.5m
13.5-24.0m
24.0-60.0m

Diaphragm wall and bored pile construction in progress.

No. 144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: [email protected],
Homepage: www.seafco.co.th.

A - 26

Object
Underground car park with disable access, natural
light and ventilation for the Central Hospital of Bangkok Metropolitan Authority. Large open space and
garden on the roof top.

Project Description
Diaphragm wall with top down excavation method for
construction of 4 level basements for car parking and
roof top garden.

Type of Work
Foundation Piling, Diaphragm Walling, Deep Excavation, Civil Construction.
Diaphragm wall and bored pile construction.

Owner & Designer

The Bangkok Metropolitan Administration

Project Schedule
2005-2007

Construction Method

Excavation with diaphragm Wall using topdown construction method to 13.0m.

Construction Details
Foundation bored pile:
Diaphragm Wall:
Basement floor area:
Landscaping:

69 (1.8mx22m)
4,900sq.m. (0.8m thick,
20.0m deep)
11,200sq.m.
2,800sq.m.

Excavation with permanent RC beams lateral support in progress.

Subsoil Conditions
Soft clay:
Medium clay:
Stiff clay:
Dense sand:
Hard clay:
Very dense sand:

0.0-11.0m
11.0-14.0m
14.0-30.0m
30.0-37.0m
37.0-42.0m
42.0-50.0m

Birds eye view of the completed underground car park.

Cross section of underground car park.
No. 144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: [email protected],
Homepage: www.seafco.co.th.

A - 27

Object
Deep foundation works for a 47 storey luxury condominium.

Project Description
Bored piles, contiguous pile wall, diaphragm walls
and soil improvement for construction of deep basement.

Owner
Sathorn Park Co., Ltd. & Grace Ivory Co., Ltd.

Designer

Layout of pile foundation and diaphragm walls.

Meinhardt (Thailand) Ltd. (Structural Engineers)

Project Schedule
2007-2008

Construction Details

Foundation bored piles: 1.00mx55.0m = 192

1.20mx55.0m = 238
Contiguous bored pile: 1.00mx22.0m = 83
Diaphragm Walls:
6,352 sq.m. (1.0m thick x
20-22.0m deep)
1,377 sq.m. (1.2m thick x
20-22.0m deep)
Soil mixing:
275 cu.m.

Subsoil Conditions
Soft clay:
Medium clay:
Stiff clay:
Dense sand:

2.5-10.0m
10.0-16.0m
16.0-26.0m
26.0-60.0m

Exposed diaphragm wall and contiguous piles.

Diaphragm wall construction in progress

Bored piling in progress.

No. 144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: [email protected],
Homepage: www.seafco.co.th.

A - 28

SRINAKRIN UNDERPASS
Civil Works
Project Description

Construction of an underpass for the traffics along

Srinakrin Road.

Type of Work

Foundation piling, diaphragm walling and civil works.

Owner

The Bangkok Metropolitan Administration (BMA).

Project Schedule
2009-2011

Construction Details
Foundation bored pile:

1.00mx29.0m =
73
0.80mx3.0x25.0m = 3

Diaphragm Walls:

13,672sqm
(0.8m thick x 25.0m)

Structural Concrete:
Reinforcement Steel:

42,230cu.m
5,266tons

Construction of the Underpass in Progress.

Subsoil Conditions
Soft clay:
Medium clay:
Stiff clay:
Hard clay:

3.0-15.0m
15.0-18.0m
18.0-19.5m
19.5-25.0m

Construction of Diaphragm Wall under the Existing Highway.

Layout of the Project.

No. 144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: [email protected],
Homepage: www.seafco.co.th.

A - 29

Diaphragm walls are embedded

32-35m deep in sand layer for
24.5m deep excavation work. 63
barrettes (1.2mx3.0m) embedded
44.5-55.0m deep in conjunction
with pre-placed steel stanchions
at the top are used to facilitate
top down construction and to
support the station box.
Glass Fibre Polymer Reinforcement was used in diaphragm wall
panels at the tunnel opening locations to allow direct break
through for TBM.

RI
VE
R

BANG SU STATION

AY
A

Thiam Ruam Mit (Thailand Cultural Center) Station (S12) is one

of 18 underground stations of the
Metropolitan Rapid Transit Authority (MRTA) Initial System Project. The station is located on
Ratchadapisek Road. The initial
system project or Blue Line is the
first link in the network and 20km.
Long. The perimeter walls of the
station box are made of cast insitu concrete diaphragm walls
1.0m and 1.2m thick. The south
wall of the station is a 1.2m thick
diaphragm wall and has openings
for three tunnels linking the south
and north underground stations
to the Depot. A Tunnel Boring
Machine (TBM) driving shaft for
twin tunnels is also included in
the station box on the northern
end.

LEGEND :
MRTA ISP (Underground Structures-North)
MRTA ISP (Underground Structures-South)
MRTA Interchange Station
MRTA Station
SRT System (Hopewell)
BMA System (BTS)

CH
AO
PH
R

GENERAL

THIAM RUAM MIT STATION

HUA LAMPHONG STATION

Location of Thiam Ruam Mit Station.

WORK UNDERTAKEN
BORED PILES:

Dia. 0.6mx19.5-23.5m
Dia. 1.0mx36.0m
Dia. 1.2mx36.0m
Dia. 1.5mx36.0m
1.2mX3.0mX44.5-55.0m
859 tons
24,338.9 sq.m. (1.0m Thick)
1,612.8 sq.m. (1.2m Thick)

BARRETTES:
STEEL STANCHION:
DIAPHRAGM WALL:
20.25

6.03

6.03

12 nos.
95 nos.
9 nos.
5 nos.
63 nos.

20.25
Foot Path

+101.000

TOP SLAB

+98.900

CONCOURSE SLAB

Soft Clay
ORANGE LINE PLATFORM
ORANGE LINE SLAB

Stiff Clay

DIAPHRAGM WALL

BASE SLAB

+80.450
+78.450

Silty Sand

BARRETTE
C
L TRACK

C
L TRACK

D04/99

Cross section of station box (Schematic).

TYPE OF WORK:
OWNER:
MAIN CONTRACTOR:
DESIGNER:
PERIOD:

Foundation Piles, Diaphragm Wall and Barrettes

The Metropolitan Rapid Transit Authority
Nishimatsu Construction Co., Ltd. (ION Joint Venture)
Ove Arup and Partners International Limited
1998-1999

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

A - 30

SEAFCO

Tel. (662) 919 0090-7, Fax. (662) 919 0098

2-SC

2-SC

2-SC

Mr. P FRESH MILK

ENTRANCE
12.1

OPEN
AREA

4-CF
4-CF

ENTRANCE
12.2

CL

CL

CL

LINE

TO RAMA IX
ROAD

2-SC

THAI
INSURANCE

R.S.TOWER

2-SC

VEN
T.
SHAFT
1

TBM
DRIVE
SHAFT

RATCHADAPHISEK RD.

ENTRANCE
12.4

CL

27.000

TO LATPHRAO
ROAD

CL

ORANGE

CL

ENTRANCE
12.3

VEN
T.
SHAFT
2

359.500

Layout of station, diaphragm wall and barrettes.

U n d ra in e d S h e a r S tre n g th
0

S u (t/m 2 )

10

15

SPT
N (B lo w s /ft)
40
60

B H -3
-1 0

-2 0

-2 0

-3 0

-3 0

Depth (m)

-1 0

Depth (m)

SO FT C LAY

-4 0

80

B H -2

B H -3

S T IF F C L A Y

M E D IU M
DENSE SAND

20

F IL L

B H -3 7

-4 0

HARD CLAY

General view of the site (diaphragm wall

and barrette work).

-5 0

-5 0

-6 0

-6 0

-7 0

-7 0

Subsoil conditions at the site.

Panel excavation with grab.

A large volume of bentonite slurry supply, storage and treatment facilities for
an excavation up to 300 cu.m.

Concrete pouring using double tremie

sets for T-Shaped panel.

Placing a strip of water-stop

in the stop-end plate.

Lowering a reinforcement cage with glass

fibre polymer reinforcement soft eye for
tunnel opening.

A wall panel being cast with a water-stop between panel joints.

A - 31

Exposed diaphragm wall inside (TBM) driving shaft.

A scene of barrette construction activity.

Barrette construction in progress (SEAFCO Guide

Frame with extended section to control verticality of
stanchion in the foreground)

Left Lowering rebar cage attached with sonic logging access

tubes.
Centre Setting up SEAFCO
guide frame for steel stanchion
to be installed with great accuracy.
Right Placing a steel stanchion
in position.

Concreting a barrette with pre-placed steel stanchion, using double tremie sets .
Steel stanchions on
ready to be installed.

site,

Verticality of stanchion can also be

checked against tall buildings nearby.

A - 32

Checking trench verticality and dimension with Koden drilling monitoring

equipment..

Drilling monitoring record

showing good verticality of
trench and dimensions.

Attaching a vibrating wire strain gauge

in the reinforcement cage of diaphragm
wall.

All rights reserved. No part of this publication may be reproduced in any form or by
any means, without the prior written permission of SEAFCO PUBLIC CO., LTD.

A jack-out pressure cell to be installed

in the reinforcement cage of diaphragm
wall.

Monitoring and Checking Bentonite

Slurry Quality.

Checking concrete quality of barrette

with sonic logging method.

Testing on barrette capacity with fully

instrumented load test.
Sonic logging test record showing good quality of concrete .

SEAFCO

References

Teparaksa, W., Thasnanipan, N., Maung, A. W. and Wei, S. H. (1998), Prediction and Performances of Short Embedded Cast In-situ Diaphragm Wall for Deep Excavation in Bangkok Subsoil, Fourth International Conference on
Case Histories in Geotechnical Engineering, St. Louis, Missouri, March 9-12, 1998. Pp. 686-692.
Thasnanipan, N., Anwar, M., A., Maung, A. W. and Tanseng, P. (1999), Performance Comparison of Bored and
Excavated Piles in the Layered Soils of Bangkok, Symposium on Innovative Solutions in Structural and Geotechnical
Engineering in Honor of Professor Seng-Lip Lee, AIT, Bangkok, May 14-15, 1999. Pp. 345-353.

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

Tel. (662) 919 0090-7, Fax. (662) 919 0098

A - 33

GENERAL
Rangsan Silom Precious (present name is Royal Charoen Krung)
building is located at the corner of Silom and Charoen Krung Roads,
Bang Rak, Bangkok. It is a 63-storey tower with six basement levels
constructed by using top-down method. Foundation bored piles of
1.5m in diameter founded in sand layer at depths up to 60m, support
the building. The piles were base-grouted. A 1.02m thick cast in-situ
diaphragm wall (toe depth 36.0m below ground level) was used for
basement excavation and construction. The maximum excavation
depth was about 20.0m. 95 built-up steel stanchions pre-founded in
bored piles were used for supporting the basement slabs and parts of
superstructure during simultaneous construction of basements and
the super structure. Three basement slabs, B1, B3 and B5 with
openings were used as lateral supports to the diaphragm wall to allow
excavation reach to the required depths.
WORK UNDERTAKEN
BORED PILES:

Section of building.

Dia. 1.0mx36.0m
5 nos.
Dia. 1.5mx60.0m
529 nos.
534 piles
Built-up Sections
95 nos.
13,748 sq.m. (1.02m Thick)
5 Inclinometer Tubes in the wall and 1
Tube in the soil behind the wall.

BASE GROUTING:
STEEL STANCHIONS:
DIAPHRAGM WALLS:
INSTRUMENTATION:

I-1

I-2

I-6

I-3
LEGEND
BORED PILE
BORED PILE WITH STANCHION

I-5

DIAPHRAGM WALL PANEL

INCLINOMETER TUBE
0

15

25

50m

I-4

View of basement construction with top-down method.

TYPE OF WORK:
OWNER:
SUPERSTRUCTURE
CONTRACTOR:
DESIGNER:
PERIOD:

Diaphragm wall and pile layout.

Foundation Piles and Diaphragm Walls

Rangsan Silom Precious Co., Ltd.
Samsung (Thailand) Development Co., Ltd.
Siam Engineering Consultants Co., Ltd.
1991-1992

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

SEAFCO

Tel. (662) 919 0090-7, Fax. (662) 919 0098

A - 34

SPT

Undrained Shear Strength

0

Top Fill

10

Su (t/m2)
20

30

N (Blows/ft)
0

40

-10

-10

-20

-20

20

40

60

80

100

Soft to Medium Clay

Depth (m)

Dense Sand

Depth (m)

Stiff to Hard Clay

-30

-30

-40

-40

Hard Silty Clay

-50

-50
BH-1

BH-1

BH-2

Very Dense Sand

BH-2
-60

-60

Subsoil conditions.
All rights reserved. No part of this publication may
be reproduced in any form or by any means, without
the prior written permission of SEAFCO PUBLIC
CO., LTD.

Diaphragm wall panel excavation.

Lowering a rebar cage section

with box-outs for basement slab
connections.

N on-G route d v s. B ase -G route d P ile s

L o ad (to n )
0

500

1000

1500

2000

2500

3000

3500

4th FL. +15.30

3rd FL. +10.80

4000

B1 -2.70

Settlement (mm)

10

B3 -7.90

20

MAT FOUNDATION

30

D-WALL
-36.0m.

40

B5 -13.10
B6 -15.70

D-WALL
-36.0m.

50
60
N o n -Gro u te d P ile - D ia . 1 .5 x6 0 .0 0 m
B a s e -Gro u te d P ile A - D ia .1 .5 x5 5 .1 6 m (3 S ta g e s - to ta l 2 0 0 0 ltr)
B a s e -Gro u te d P ile B - D ia .1 .5 x6 0 .0 0 m (3 S ta g e s - to ta l 2 0 0 0 ltr)

Floors and basements constructed

Pile load test results.

Excavation below the basement slab.

Installing a 24.0m-long built-up

steel stanchion in the borehole.

Preparation for connections between basement floor and diaphragm wall.

References:
Teparaksa, W., Thasnanipan, N. and Tanseng, P. (1999), Analysis of Lateral Wall Movement for Deep Braced Excavation in Bangkok Subsoils, Civil and Environmental Engineering Conference New Frontiers & Challenges, AIT,
Bangkok, November 8-12, 1999.
Lee, S. L. and Yong, K. Y. (1992), Base Grouting of Bored Piles, 2nd International Conference on Deep Foundation
Practice incorporating PILETALK 1992, Singapore, 4-6 November 1992. Pp. 147-153.

SEAFCO
144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

Tel. (662) 919 0090-7, Fax. (662) 919 0098

A - 35

GENERAL

DON MUANG

N
FROM

IC
THE V

TORY

M EN T
M ONU

AT
ED
EX THE
PR
ES FIRS
SW T S
AY TA
GE

TO
5 15

35

AS
OK
EDIN

DA
EN
G

EL
EV

The Din Daeng Underpass is the

first of its kind in Bangkok Metropolis to ease the traffic congestion. It is an alternative to flyovers
and overpasses. The underpass
route extends for 415m with two
lanes with one way traffic from the
Victory monument. It runs along
the middle of the Asoke-Din
Daeng Road, and passes under
the elevated First Stage Expressway as a cut and cover tunnel.
82m long approach sections at
both ends of the underpass were
constructed with 0.45m thick reinforced concrete walls. The 252m
long middle section of the underpass, which includes portals and
cut and cover tunnel, is constructed with 1.0m thick diaphragm
walls having toe level
21.0m
deep. Cantilever walls and reinforced concrete strut-supported
walls are used in the portal sections. Excavation depths for base
slab construction varied from 4.9m
to 6.5m below the pavement level.

Layout of Din Daeng Underpass.

WORK UNDERTAKEN
DIAPHRAGM WALL:
INSTRUMENTATION:

Thickness 1.0m
Toe Depth 21.0m
10,545sq.m.
8 Inclinometer Tubes in the wall
2 Inclinometer Tubes in the ground
behind the wall, 8 Sets of VWSG. and
2 Vibrating Wire Piezometers

0.00m
Back Fill
Soft Clay

-6.50m

Drain Pipe
-21.00m
Stiff Clay

Diaphragm Wall

D02/99

Schematic section of cut and cover tunnel.

TYPE OF WORK:
OWNER:
MAIN CONTRACTOR:
DESIGNER:
PERIOD:

Diaphragm wall panel excavation in progress.

Diaphragm Wall
Bangkok Metropolitan Administration.
Unique Engineering and Construction Co., Ltd.
Epsilon Co., Ltd.
1996

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

A - 36

SEAFCO

Tel. (662) 919 0090-7, Fax. (662) 919 0098

U n d ra in e d S h e a r S tre n g th

SPT

N (B lo w s /ft)

S u (t/m )
0

10

15

20

25

30

F IL L

10

20

30

40

-5

B H -2

B H -3

-5

B H -4

Depth (m)

Depth (m)

M E D IU M
DENSE SAND

B H -4

-1 0

-1 5

-2 0

-2 0

-2 5

-2 5

-3 0

-3 0

S T IF F C L A Y

B H -3
B H -5

-1 0

-1 5

60

B H -1

B H -5

SOFT CLAY

50

0
B H -1
B H -2

Subsoil conditions at the Site.

Diaphragm wall panel excavation in the middle of

traffics.

All rights reserved. No part of this publication may be reproduced in any form or by
any means, without the prior written permission of SEAFCO PUBLIC CO., LTD.

Diaphragm wall panel excavation with

modified grab crane under the existing
expressway.

SEAFCO

Portal section.

Tunnel section.

References:
Thasnanipan, N., Maung, A. W. and Singtogaw, K. (1996). Experience in Construction of an Underpass Tunnel using
Diaphragm Wall Bangkok (in Thai), Technical Seminar on Construction of Basements 1996 organized by Engineering
Institute of Thailand, Bangkok, 18 September 1996. Pp. 55-70.
Teparaksa, W., Thasnanipan, N. and Tanseng, P. (1999), Analysis of Lateral Wall Movement for Deep Braced Excavation in Bangkok Subsoils, Civil and Environmental Engineering Conference New Frontiers & Challenges, AIT, Bangkok, November 8-12, 1999.

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

Tel. (662) 919 0090-7, Fax. (662) 919 0098

A - 37

D02/99

General view of Din Daeng Underpass.

Lowering a rebar cage section.

GENERAL

+14.80

The Thammasat Administration

Building project site is located on
Tha Prachan Campus, Bangkok
nearby the Chao Phraya River.
The site is surrounded by a historical building and other existing
structures. The construction of
this building commenced about
one year after completion of library building with three basements being separated by an existing building. Bored piles with
pile tip in dense sand layer are
used for supporting the building.
A diaphragm wall of 0.8m in thickness was designed for excavation
9.7m deep with two levels of temporary bracing. The diaphragm
wall toe was embedded down to
28.0m to achieve the overall stability of the excavation on the river
bank. Various types of instrumentation were installed in the wall
and existing buildings to observe
the ground movements and response of the buildings during
excavation.

10

+11.30

20M

+7.80

+0.70
0

-3.20
-5.80
-8.40

Soft to
Medium
CLAY

SPT-N (Blows/ft)

20 40 60 80 0

50

100

Phase 1
Phase 2

5
10
15

Bored Pile

0.80m Thick D-Wall

Stiff to Very
Stiff CLAY

20
25

-28.00 (D-Wall Tip)

Medium to
Very Dense
SAND

30
35

-48.00 (Pile Tip)

Basement section and soil profile.

WORK UNDERTAKEN
BORED PILES:

DIAPHRAGM WALL:
EARTH WORK:
TEMPORARY BRACING:
INSTRUMENTATION:

Excavation and basement construction.

TYPE OF WORK:
OWNER:
MAIN CONTRACTOR:
DESIGNER:
PERIOD:

Vane Shear
Strength (kPa)

+4.30
+0.80

Dia. 0.8mx48.0m
17 nos.
Dia. 1.0mx48.0m
33 nos.
Dia. 1.2mx48.0m
5 nos.
Dia. 1.5mx48.0m
23 nos.
7,742sq.m. (0.8m Thick)
59,592cu.m.
586.50 tons
6 Inclinometer Tubes, 10 Tiltmeters,
5 Vertical Beam Sensors, VWSGs in
one Panel, 4 Earth Pressure Gauges,
20 Surface and Deep Settlement
Plates.

Diaphragm wall panel excavation.

Foundation Piles, Diaphragm Wall and Excavation

The Thammasat University
Ch. Karnchang Public Co., Ltd.
SJD-3D Co., Ltd.
1999

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

A - 38

SEAFCO

Tel. (662) 919 0090-7, Fax. (662) 919 0098

Chao Phraya River

D4

D10
D9

S4
I-1

S10
S9 I-2

S3
D3

5 Sto
rey B

EL1

S8
D8

S2
D2

T2

T10
EL5

I-3

T9

uildin
g

D7

EP-1

T3
S5

I-5

S6

T4
5 Storey Building

Legend

D5

I-4 D6

T8
EL4

S7
T7

T5

EL2

T6

EL3
5 Storey Building
0

Inclinometer in D-Wall
Tilt plate
Vertical beam sensor
Surface settlement plate (1.0m)

5 Storey
Building

D1
I-6
S1

T1

6 10

20m

Deep settlement plate (5.0m)

Earth pressure gauge
Pre-load location

Temporary bracing and working platform.

Layout of instrumentation and the project site.

Bored Pile 0.8x5.98m

Load (Tons)
0

200

400

600

800

0.0

Settlement (mm)

All rights reserved.

No part of this
publication may be
reproduced in any
form or by any
means, without the
prior written permission
of
SEAFCO PUBLIC
CO., LTD.

5.0

10.0

15.0

20.0

Excavation completed to the final depth.

Pre-loading the struts.

SEAFCO

Pile load test results.

Earth pressure gauges for checking

the strut force.

Preparation for base slab casting.

References:
Thasnanipan, N., Maung A. W. and Tanseng, P. (1999), Behavior and Performance of Diaphragm Walls under Unbalanced Lateral Loading along the Chao Phraya River, 5th International Symposium on Field Measurements in Geomechanics FMGM99, 1-3 December, 1999, Singapore. Pp 267-272.
Teparaksa, W., Thasnanipan, N. and Tanseng, P. (1999), Analysis of Lateral Wall Movement for Deep Braced Excavation in Bangkok Subsoils, Civil and Environmental Engineering Conference New Frontiers & Challenges, AIT,
Bangkok, November 8-12, 1999.
Thasnanipan, N., Maung A. W., Tanseng, P. and Wei, S. H. (1998), Performance of a Braced Excavation in Bangkok
Clay, Diaphragm Wall Subject to Unbalanced Loading Conditions, 13th Southeast Asian Geotechnical Conference,
16-20 November, 1998, Taipei, Taiwan, ROC. Pp 655-660.

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

Tel. (662) 919 0090-7, Fax. (662) 919 0098

A - 39

GENERAL

I-3

The Sathorn Complex project is

planned for a multipurpose office
and shopping complex. It is located at the Sathorn and Rama
IV Road intersection. The foundation bored piles and barrettes
are embedded 60m deep in sand
layer to support the building. A
0.8m thick cast in-situ diaphragm
wall 18.0m in depth, with threelevel temporary bracing was initially planned for construction of a
four-level basement. After completion of diaphragm construction,
basement excavation was later
modified with four level-bracing to
allow excavation down to 15.5m
below ground level for construction of five basement levels. For
the foundation in the area of the
main tower, barrettes were employed together with bored piles
to increase load bearing capacity.
Bored piles and barrettes were
also incorporated as legs in the
diaphragm wall to carry the load
of building.

Excavation -15.5m
Bored Pile
Barrette
Inclinometer Tube

I-2
89.35

Excavation -14.0m

Foundation and diaphragm wall layout.

WORK UNDERTAKEN
BORED PILES:
BARRETTES:
DIAPHRAGM WALL:
INSTRUMENTATION:

-14.00

-6.90
10

-9.60

-12.30
-13.30

-18.00

-15.50
-18.00

525

700

CONSTRUCTION STAGE

1050

1225

1400

1575

1750

-10
-15
-20
-25
-30

-40

TEMPORARY BRACING

LAOD (Tons.)

Basement excavation and construction sequence (Schematic).

TYPE OF WORK:
OWNER:
MAIN CONTRACTOR:
DESIGNER:
PERIOD:

875

-35

LEGEND
2

350

-5

12

SETTLEMENT (mm.)

175

- 4.20

11

-11.80

14

-9.00

BORED PILE 1.2m. TOE DEPTH 60.17m.

-1.50

13

-5.10

Dia. 1.2mx60.0m
443 nos.
0.8mX2.7mX60.0m
50 nos.
7,002sq.m. (0.8m Thick)
4 Inclinometer Tubes in the wall

+1.70

15
1

I-1

120.25

16

-2.40

I-4

Pile load test result.

Foundation Piles, Diaphragm Wall and Barrettes

Quality Houses Public Company Limited.
Kay-Thai Co., Ltd.
K.C.S. Associates Co., Ltd.
1995-1996

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

A - 40

SEAFCO

Tel. (662) 919 0090-7, Fax. (662) 919 0098

1925

Su (t/m2)
0

10

15

SPT N

20

25

0.0

Top Fill

-3.0

BH-1

Medium to Very
Stiff Clay

Depth (m)

Depth (m)

40

-9.0

-15.0

-18.0

-18.0

-21.0

-21.0

-24.0

-24.0

-27.0

30

-12.0

-15.0

Dense Sand

20

-6.0

-9.0
-12.0

(Blows/ft)

-3.0

BH-2

-6.0

Soft Clay

10

0.0

-27.0

Subsoil conditions at the site.

Lowering a reinforcement cage attached with inclinometer tube void former.

All rights reserved. No part of this publication may be reproduced in
any form or by any means, without the prior written permission of
SEAFCO PUBLIC CO., LTD.

Work in progress.

Exposed barrette top sections.

Exposed diaphragm wall.

SEAFCO

References:
Thasnanipan, N. and Singtogaw, K. (1996), Construction of Diaphragm Wall and Bored Piling for Sathorn Complex
Project (in Thai), Technical Seminar and Site Visit on Construction of Deep Basements using Short Embedded Diaphragm Wall in Bangkok, organized by Engineering Institute of Thailand. Bangkok, 19 December 1996. Pp. 129-143.
Teparaksa, W., Thasnanipan, N., Maung, A. W. and Wei, S. H. (1998), Prediction and Performances of Short Embedded Cast In-situ Diaphragm Wall for Deep Excavation in Bangkok Subsoil, Fourth International Conference on Case
Histories in Geotechnical Engineering, St. Louis, Missouri, March 9-12, 1998. Pp. 686-692.
Teparaksa, W., Thasnanipan, N. and Tanseng, P. (1999), Analysis of Lateral Wall Movement for Deep Braced Excavation in Bangkok Subsoils, Civil and Environmental Engineering Conference New Frontiers & Challenges, AIT,
Bangkok, November 8-12, 1999.

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

Tel. (662) 919 0090-7, Fax. (662) 919 0098

A - 41

GENERAL
I-6

The Rajavej Hospital is located

on Phayathai Road, Rajatheewi,
Bangkok. It has five basement
levels constructed with top-down
method for car parking and
rooms for utilities. Foundation
bored piles of 0.8m, 1.2m and
1.5m in diameter founded in sand
layer at depths of 58m, support
the building. A 1.0m thick cast
in-situ diaphragm wall (toe depth
21.0m below ground level) was
used for basement excavation
and construction. The maximum
excavation depth was about
14.5m. 80 steel stanchions prefounded in bored piles were used
for supporting the basement
slabs and parts of superstructure
during simultaneous construction
of basements and the super
structure. Two basement slabs,
B1 and B3 with openings were
used as lateral supports for the
diaphragm wall to allow excavation to the required depths.

I-7

I-5

I-4
INDEX
BORED PILE 0.80m.
BORED PILE 1.20m.

I-3

I-1

BORED PILE 1.50m.

STEEL STANCHION
INCLINOMETER TUBE

10

20m

I-2

Diaphragm wall and foundation layout.

WORK UNDERTAKEN
BORED PILES:
STEEL STANCHIONS:
DIAPHRAGM WALL:
INSTRUMENTATION:

Dia. 1.0mx58.0m
5 nos.
Dia. 1.2mx58.0m
114 nos.
Dia. 1.5mx58.0m
56 nos.
414x405-498x432mm 80 nos.
6,058 sq.m. (1.0m Thick)
7 Inclinometer Tubes in the wall.

0.0m
B1

B3

-6.7m

D-WALL
-11.8m
-14.5m

BORED
PILE
-21.0m

Lowering a steel stanchion into

borehole before casting the pile.

TYPE OF WORK:
OWNER:
SUPERSTRUCTURE
CONTRACTOR:
DESIGNER:
PERIOD:

Basement section.

Foundation Piles and Diaphragm Wall

Rajavej Hospital Co., Ltd.
SAE Thailand Co., Ltd.
P & CYGNA Consultants Co., Ltd.
1994-1995

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

SEAFCO
Tel. (662) 919 0090-7, Fax. (662) 919 0098

A - 42

SPT

Undrained Shear Strength

0

Top Fill

10

Su (t/m2)
20

30

N (Blows/ft)
0

40

-10

-10

-20

-20

20

40

60

80

100

Soft to Medium Clay

Depth (m)

Dense Sand

Depth (m)

Stiff to Hard Clay

-30

-30

-40

-40

Hard Silty Clay

-50

-50
BH-1

BH-1

BH-2

Very Dense Sand

BH-2
-60

-60

Subsoil conditions.
All rights reserved. No part of this publication may be
reproduced in any form or by any means, without the
prior written permission of SEAFCO PUBLIC CO.,
LTD.

Diaphragm wall panel excavation.

21m-long reinforcement cage

with box-outs for basement slab
connections.
BORED PILE DIA. 1.2mX58.16m
Applied Load (Tons)
0

500

1000

1500

2000

Pile Head Movement (mm)

OPENING

OPENING

B3 SLAB

10

20m

10

20

30

40

50

Pile load test result.

Basement slab with temporary openings for excavation work.

Installing a 18.0m-long steel

stanchion in the borehole.
Excavation at final depth of 14.5m
below ground level.

Excavation below the basement slab.

SEAFCO

References:
Thasnanipan, N. and Singtogaw, K. (1996), Bored Piling and Diaphragm Wall Work for Excavation of Five-level Basement of Rajavej Hospital (in Thai), Technical Seminar and Site Visit on Experience in Construction of Deep Basements in Bangkok Soft Clay, organized by Engineering Institute of Thailand. Bangkok, 26 March 1996. Pp. 104-140.
Teparaksa, W., Thasnanipan, N. and Tanseng, P. (1999), Analysis of Lateral Wall Movement for Deep Braced Excavation in Bangkok Subsoils, Civil and Environmental Engineering Conference New Frontiers & Challenges, AIT,
Bangkok, November 8-12, 1999.

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510.

Tel. (662) 919 0090-7, Fax. (662) 919 0098

A - 43

KABAR AYE EXECUTIVE RESIDENCE

Bored Piling & Diaphragm Wall Works
-

A - 44

PROCEEDINGS OF THE THIRTEENTH

SOUTHEAST ASIAN GEOTECHNICAL
CONFERENCE
November 16-20 1998, Taipei, ROC

Barrettes Founded in Bangkok Subsoils, Construction and Performance

N. Thasnanipan, A. W. Maung and P. Tanseng
Seafco Co., Ltd., Bangkok, Thailand

Sponsored by;
Chinese Institute of Civil and Hydraulic Engineering
Southeast Asian Geotechnical Society

Thirteenth Southeast Asian Geotechnical Conference, 16-20 November, 1998, Taipei, Taiwan, ROC

Barrettes Founded in Bangkok Subsoils, Construction and

Performance
N. THASNANIPAN

SEAFCO Co., Ltd., Bangkok, Thailand

A. W. MAUNG

SEAFCO Co., Ltd., Bangkok, Thailand

P. TANSENG

SEAFCO Co., Ltd., Bangkok, Thailand

SYNOPSIS
Barrettes become not an uncommon foundation element used to support multi-storey buildings, elevated
expressways and underground station boxes in Bangkok, Thailand. The demand for barrettes is necessitated mainly by
constraints of site conditions, construction method, equipment and extensive bearing capacity requirement, etc. This paper
presents construction practice and the performance of barrette constructed in Bangkok metropolis. Trial trenching near a canal
for assessment of trench stability and soil deformation are presented. Choice of barrette and common defects found in barrette
are also discussed.

INTRODUCTION

slurry used for trenching in Bangkok subsoil usually comply

with DFCP-4 of the Oil Companies Materials Association
and the Institution of Civil Engineers (ICE) Specification
for Piling and Embedded Retaining Walls. The usual
bentonite slurry properties specified are 1.10g/ml for density,
30-50sec for viscosity (Marsh cone), 8-11.5 for pH and not
more than 4% for sand content. Usually 3m to 6m long
sections of steel tremie pipe of 25cm in diameter are joined
together to form a continuous pipe with required length for
concrete pouring under the slurry.

Recently published papers indicate that barrettes become a

common foundation type used in various soil conditions to
carry very high load for sky scrapers. Development of
properties and infrastructures such as construction of multistoried buildings, elevated expressways, subway railway
stations, etc. demand deep foundations with large bearing
capacity in Bangkok metropolis, Thailand. Bangkok is well
known for its subsoils constraints especially, by the presence
of thick soft clay on the top, followed by a series of
alternating stiff clay or hard clay and dense or medium dense
sand layers. Due to not only the subsoil conditions, but also
non-availability of adequate construction area and
overlapping developments above and underground, barrettes
become common to replace the usual large diameter bored
cast in-situ piles in circular shape. The earliest barrettes of
0.82mx2.98m-3.0m with toe depth of 50m were constructed
in Bangkok around 1985 for International Trade Centre
building in conjunction with diaphragm wall for basement
construction. The first static load test on a barrette
(0.82x2.7x61.8m) in Bangkok was carried out in 1992 for a
55 storey building on Silom Road, Bangkok.

Cast-in-situ reinforced concrete, pre-cast concrete and

temporary steel box guide walls appropriate for site
conditions are used for trenching. In normal site conditions,
guide walls with a minimum depth of 1.2m are commonly
used. Base grouting is also used to improve the capacity and
performance of barrettes if necessary.
Generally an average rate of excavation about 10-15m2/hr is
achieved with mechanical grab in Bangkok subsoil. Usually,
excavation rate decreases with increment in excavation
depth, having non-linear relationship between excavation
time and depth. Figure 1 illustrates excavation time and rate
vs. excavation depth for barrettes of 1.50x3.00x55.00m. with
no site constraint and idling time.

CONSTRUCTION METHOD
Normally a cable hung grab mounted on crawler crane is
used for trenching under supporting fluid which is
commonly bentonite slurry. Concrete pouring is carried out
under bentonite slurry using tremie pipes. Properties of

573

TRIAL TRENCH PERFORMANCE

24.00
B15
B20
B3
B10
B19
Best Fit

20.00

Time (Hours)

16.00

A trial trench (0.8mx2.7mx14m), with above mentioned soil

reinforcement close to the canal and high slurry head, was
carried out to study the feasibility in construction of barrettes
to support the elevated expressway across the Canal. The
major concern was stability of 11.0m thick soft clay layer
(Fig. 2) which has a very low shear strength less than 15 kPa
and natural water content of up to 80%. The planned trench
was about 1.5m away from the canal water and four
reinforced sheet piles were installed between the trench and
the canal prior to excavation. In addition to the reinforced
sheet piles, a temporary sheet pile wall was also constructed
for protection of slurry spillage into the canal as the local
authorities strictly required.

12.00
8.00
4.00
0.00
0.00

10.00

20.00

30.00
Depth (m)

40.00

50.00

60.00

30.00
B15
B20
B3
B10
B19
Best Fit

Excavation Rate (m2/hr)

25.00
20.00
15.00

Creep (mm)

10.00
0.0

100

Slurry
Level

Guide Wall
Creep (mm)

-100

0.0

Canal

TOP SOIL

5.00
0.00
0.00

10.00

20.00

30.00
Depth (m)

40.00

50.00

60.00

2.0

2.0

Steel
Sheet Pile

Soft clay

0.1

Stiff clay

0.2

0.3

Med. sand

0.4

0.5

V. stiff clay

0.6

0.7

Dense sand

0.8

0.9

Usually 1.0m or more over casting of concrete above

designed cutoff level is practiced. Observation of 3
completed projects indicates that concrete over break of
barrettes was found to be in a range of 6% to 14% for
barrettes with about 2.0m over casting.

4.0

4.0

6.0

6.0

8.0

8.0

10.0

12.0

10.0

12.0

MEDIUM
CLAY

TRENCH STABILITY

Depth (m)

SOFT
CLAY

Figure 1. Excavation time, rate vs. depth for barrettes

1.5mx3.0m.

Depth (m)

After 24 Hours
After 48 Hours

In general site conditions, Bangkok soft clay layer with

undrained shear strength value of 15- 20 kPa does not pose a
particular problem for trench stability during barrette
excavation under bentonite slurry. However trenching in the
close vicinity of the river or canal/khlong required careful
assessment of soil properties, especially sensitivity and
natural water content of the soft clay and presence of
permeable soil layers connecting the river bed with high pore
pressure. Normally, higher guide wall with adequate slurry
pressure and reinforcement of the surrounding soft clay by
sheet piles placed perpendicular to the alignment of the
planned barrette and driven to a stiffer soil layer, overcome
the instability of the trench.

14.0

14.0

Figure 2. Soil movements on opposite faces of trial trench

The trial trench filled with bentonite slurry was open for 48
hour after excavation down to 14m. Lateral movements of
trench face along the depth were studied with graphs of the
trench profiles plotted by a Koden drilling monitoring
equipment stationed at a fix position on the guide wall at
every 6 hour interval.
Trench profiles plotted were
compared and changes were manually measured using the
reference scales shown on the graph. The recording
accuracy of the equipment is in a range of +/- 0.2% (in this
case the accuracy would be within 2-3mm). 20mm to 30mm
lateral soil movement inward the trench was observed on the
monitoring records on non-reinforced soil side after 24 hours

574

and 48 hours respectively in the soft clay layer. For trench

face close to the canal , observed movements of the soil
being reinforced with sheet piles are generally less than those
of non reinforced soil and some movements are considered
induced by the sheet pile wall installed for canal water
protection. A profile of trench showing soil movement
plotted from the monitoring records is presented in Figure 2.

capacity with safety factor of 2.0 to 2.5. The toe depth of

constructed barrettes ranges from 45.0m to 60.0m. Concrete
cube strength of 30MPa to 40MPa are commonly used for
barrettes. Minimum reinforcement of 0.5% of the cross
sectional area of barrette is usually provided for either full
length or partially as per load requirements.
Barrettes can be constructed with flexible layout plan for
both vertical and lateral loads. The layout pattern of
barrettes can be arranged in a continuous row or column,
radial, alternating long and short axis of barrette and a
combination of two or more of such patterns. One type of
equipment can be used for constructing both barrettes and
diaphragm walls in particular projects thus reduces the
mobilization cost. In addition to the large bearing capacity
requirements, on site difficulties such as limited head room,
also demands the barrettes. Piling rigs cannot be utilized
under such situations like presence of high voltage power
cables overhead, existing overpasses or structures for
elevated expressways and planned subway stations.
A summary of observation on barrettes of 19 projects
completed, with respect to selection criteria is presented in
Table 1.

There is no particular problem associated with trench

stability in the sand layers 20 to 45m of typical Bangkok
subsoil below the ground level. The drilling monitoring
records show no collapse of trench surface along the depth
section (Fig. 3).

Table 1. Summary of Barrette Selection

Criterion

No. of

Remarks

Project

Figure 3. Trench profile of barrette (1.5mx3.0mx55m)

plotted by Koden drilling monitoring equipment
LOAD BEARING CAPACITY OF BARRETTES

High Load Capacity

Minimise Construction
Equipment
Limited Head Room
for Excavation

Combination
with
Diaphragm Wall
Foundation as well as
portion of column

10

incorporated with bored

piles
Alternative for bored
piles
Elevated
Highways
under or above existing
structures/power lines
As diaphragm wall legs
provision for the future
requirement

Since basic shape of the cross section of barrette is

rectangular, certain dimensions of barrette can provide a
greater perimeter for skin friction than that of the
conventional bored pile (in circular shape) with equivalent
cross sectional area. Hence, for friction piles, more load
carrying capacity per unit volume of concrete can be
achieved by a barrette than that of a circular bored pile.

Common type of barrettes constructed in Bangkok area is

rectangular shape with dimensions ranging from 0.6mx2.7m
to 1.0mx2.7m. These barrettes have been mostly constructed
incorporated into diaphragm walls for deep basements of
high rise buildings. However, individual larger barrettes
(1.2mx2.7m and 1.5mx3.0m) are also currently in use to
support large column loads as the grabs of large dimensions
are currently available in the market. A number of load test
on such large barrettes will be carried out in the near future.
Maximum bearing capacity of barrettes is usually designed
up to allowable concrete stress of 5MPa. Toe depth of
barrettes are limited by subsoil conditions and required load

Figure 4 illustrates that typical barrette of 2.7m with various

thickness can give 12 to 30% higher friction area than that of
circular piles with corresponding equivalent sectional area.
575

PERFORMANCE OF BARRETTES

11
10

Available load test results, 4 nos of static load test and one
dynamic load test, are summarized below in Table 3.

9
Perimeter (m)

8
7
6

Table 3. Summary of Barrette Load Test

Perimeter of
Square
Perimeter of
Circle
Barrette 2.7x X

5
4
3

Project

2
1

0
0

Equivalent Gross Cross-sectional Area (m )

Figure 4. Comparison of shape effect between barrette and

circular pile

D
QUALITY CONTROL

Construction tolerance allowed for barrettes are generally

identical to that for bored piles. Normally verticality of
1:100 is allowed in barrette construction in Thailand.
Drilling monitoring is employed to check the verticality of
the trench on random barrettes or on all barrettes.
After construction, integrity testing such as sonic integrity
(low strain dynamic test) and sonic logging tests and
concrete coring are sometimes applied. Common defects
found in barrettes constructed in Bangkok subsoils are
presented in Table 2.

Note:

Barrette
Dimension
(m)
0.82x2.70x
61.80*

Test
Load
(kN)
14000
28000
35000
12000
24000
12900
25800
30140
11000
27500
20555

0.82x2.70x
44.00
0.80x2.70x
55.00

Total
Settlement
(mm)
4.83
12.56
17.63
5.65
34.15
4.46
11.01
12.35
3.07
7.97
14.00

0.80x2.70x
50.00
1.00x2.70x
48.94**
* with base grouting
** load test by dynamic load

Permanent
Settlement
(mm)
0
0
3.47
0.15
24.72
0.24
2.10
2.17
0.21
0.15
3.0

All the barrettes tested, except the barrette of project B were

founded in dense sand layer with embeddment more than 5m
into it. The barrette of project B was embedded into about
0.5m in dense sand layer and failed under the maximum test
load of 24MN. Base grouting was employed for the barrette
of project A. A soil profile and load vs. settlement graph of
test barrette of project A are presented in Figures 5 and 6
respectively.

Table 2. Common Defects Found in Barrettes

SPT-N (BLOWS/30cm)
0

Defect found

Possible Causes

Concrete cover

improper cage position, provision of

inadequate spacers
improper tremie concrete pouring,
improper desanding and recycling of
slurry, inadequate concrete slump
inadequate supervision

Bentonite slurry
inclusion or poor
quality concrete
Reinforcement
cage deviation
Uplift of
reinforcement cage

50

100

-4

-1 0

-1 6

SOFT
CLAY

STIFF
CLAY

-2 2

-2 8

DENSE
SAND

-3 4

-4 0

improper tremie concrete pouring

(with very long tremie pipe
embedment in concrete poured) in
partially reinforced barrettes

-4 6

SILTY
CLAY
DENSE
SAND

-5 2

-5 8

-6 4

-7 0

Figure 5. Soil profile at project A site

576

SILTY
CLAY

DENSE
SAND

Load (MN)
0

10

20

30

40

Settlement (mm)

5
10
15
20
25
30
35

Cycle - 1
Cycle - 2
Cycle - 3

40

Figure 8. Project E -Simulated load settlement curves (PDA)

Figure 6. Load vs. settlement graph of barrette load test Project A.

DEPTH
(m)

Barrettes of project E were constructed on both side of water

supply canal and one barrette was tested with high strain
dynamic load test as no anchor piles were available. A
200kN steel hammer was dropped at a height of 2.6m above
the pile cap. A graph of dynamic load test signal analysis
and simulated load settlement curve are shown in Figures 7
and 8 respectively.

SPT-N (BLOWS/30cm)
0

50

100

FILL

-3
-6
-9

SOFT
CLAY

-1 2
-1 5
-1 8
-2 1
-2 4
-2 7
-3 0
-3 3
-3 6

SILTY
CLAY

-3 9
-4 2
-4 5
-4 8
-5 1
-5 4
-5 7

SILTY SAND

-6 0
-6 3
-6 6

Figure 9. A soil profile of Project E site

Load test results listed in Table 3 indicate that all barrettes
tested are able to carry the design load with Factor of Safety
of 2 or higher. From the available load test results, the end
bearing capacity of test barrettes cannot be estimated.
However, all barrettes are considered as friction piles, except
for project A in which base grouting was used to improve the
end bearing of barrette. For test barrette of project A, no
permanent settlement was observed under test load of 28MN
which is 200% of the design load and base grouting is
considered contributed in good performance of barrette.

Figure 7. High strain dynamic load test signal (force and

velocity) and PDA analysis - Project E

Barrette load tests with full instrumentation devices such as

strain gauges and magnetic extensometers are planned to be
carried out for subway stations in the near futures. The
577

performance of barrettes in Bangkok subsoil can then be

evaluated with a great accuracy.
CONCLUSION
Construction practice adopted in the foundation industry for
barrette installation in Bangkok subsoil has been discussed in
this paper. There is no indication of trench instability under
bentonite slurry in Bangkok subsoil for barrette construction
with proper site preparation. This is evidenced by trial
trenching and profiles plotted by drilling monitoring
equipment. To achieve a good end bearing capacity for
barrette founded in sand, base grouting is recommended.
Performance of barrettes has been discussed with available
load test results, and found to be satisfactory with load
bearing performance.
ACKNOWLEDGEMENT
The authors express their appreciation to the colleagues,
especially to Mr. Ganeshan Baskaran and Mr. Muhammad
Ashfaq Anwar for their invaluable suggestion and assistance
in preparation of this paper.
REFERENCES
Schlosser F., Simon B. and Morey J., (1989) High
Capacity Barrette in a Region of Old Underground
Quarries, Proceedings of the International Conference on
Piling and Deep Foundations. London, 15-18 May 1989.
Pp 119-130.
Woo S. M., Lee K.H. and Hsieh K.J. (1993) The Causes of
Cast-In-Situ Diaphragm Wall Defects, Eleventh Southeast
Asian Geotechnical Conference, 4-8 May, 1993,
Singapore, Pp 793-798.
Ho C. E. (1993) Deep Barrette Foundation in Sigapore
Weathered
Granite
Eleventh
Southeast
Asian
Geotechnical Conference, 4-8 May, Singapore, Pp 529534.
Ho C. E. and Lim C. H. (1998) Barrettes Designed as
Friction Foundations: A Case History, Fourth
International Conference on Case Histories in Geotechnical
Engineering, St. Louis, Missouri, USA. March 9-12, 236241.
Dibiagio E and Myrvoll F. (1983) Full Scale Field Test of a
Slurry Trench Excavation in Soft Clay, Proceedings of the
Conference on Instrumentation in Civil and Geotechnical
Engineering, 4-5 July 1983 Singapore. Pp 49-58.
578

Fourth International Conference on

Case Histories in Geotechnical

Engineering
March 8 12, 1998, St. Louis, Missouri, USA

Editor : Shamsher Prakash

Design, construction and behavior of bored

cast-in-situ concrete piles in Bangkok subsoils
Thasnanipan N. , Maung A. W. and Ganeshan B.
SEAFCO Co., Ltd., Bangkok, Thailand

Teparaksa W.

Chulalongkhon University, Bangkok, Thailand

University of Missouri-Rolla
Rolla, Missouri

Proceedings: Fourth International Conference on Case Histories in Geotechnical Engineering,

St. Louis, Missouri, March 9-12, 1998

DESIGN, CONSTRUCTION AND BEHAVIOUR OF BORED CAST IN-SITU CONCRETE PILES IN BANGKOK SUB SOIL

Narong Thasnanipan
Seafco Co., Ltd.
Bangkok 10320
Thailand

Wanchai Teparaksa
Civil Engineering Dept.
Chulalongkorn University
Bangkok 10330
Thailand

Aung Win Maung

Seafco Co., Ltd.
Bangkok 10320
Thailand

Ganeshan Baskaran
Seafco Co., Ltd.
Bangkok 10320
Thailand

Paper No.1.20

ABSTRACT
This paper briefs design considerations, construction methods and materials, static load test results and pile integrity test results of the
cast in-situ bored piles constructed as foundation elements for a wastewater treatment plant. The treatment plant is located in
Bangkok, capital of the Kingdom of Thailand and is to serve about 195,000 inhabitants of the city. A total of 402 piles, 296 nos. of
1500 mm diameter and 96 nos. of 1000 mm diameter, were designed to carry the whole structural load of the treatment plant.
Maximum working load on individual piles was designed to be 1000 tons and 500 tons for 1500 mm and 1000 mm diameter pile
respectively with a safety factor of 2.5. Initially the pile toe depth was designed to be 60m below the ground level, and embedded into
hard clay for 1500 mm diameter piles while the smaller diameter pile embedded into second sand layer at a depth of 53m and base
was not decided to be grouted. Pile load test results on fully instrumental pilot piles showed the failure well before reaching the
designed test load. Piles were re-designed and pile toe depth was set to be 55m, embedded into dense sand layer for both sizes and
base grouting was decided to be done for all piles. Proof load test results performed on four working piles(contract piles), two for
each size, produced well acceptable results conforming to the design and structural specifications and requirements. However,
settlements observed were higher than that observed for other projects in the vicinity. Base Grouting was done using tube - a
manchette method. A maximum volume of 500 liter of cement grout (w/c ratio ~ 0.5) or maximum grouting pressure of 40 bars was
used as limiting criteria. However, maximum pressure of 40 bars was hardly achieved. From the proof load test results ultimate load
capacity was estimated and compared using different methods suggested by previous researchers. Integrity test was decided to be
performed on each pile and results available currently are summarized.
KEYWORDS
Bored Piles, Base Grouting, Static Pile Load Test, Integrity Test, Pile Capacity

281

11

INTRODUCTION

sensitive and has an-isotropic and time dependent stress strain

properties. Below is stiff to very stiff clay layer underlain by
hard clay layers. These clay layers extend down to about 42m
in depth. However, the subsoil condition below 36m from the
ground level in the project area is variable - a sand layer with
a thickness of about 13m is present sporadically. In some
locations within the project site, a stiff to hard dark gray clay
is also found at the depth between 40m and 49m. Generally a
thick dense sand layer is located below 50m from the existing
ground level. However, at this area alternated layers of dense
sand and hard clay of 5m thick also was encountered. Fig. 2
shows the general bore log data observed at the site.

The city of Bangkok has been emerged as one of heavily

populated urban cities in the world. Due to its rapid economic
growth, the construction industry has been challenged to
develop the infrastructure to cope with the booming economy.
As a result, while multi storied private owned buildings
become most attractive sky scrapers in this city of Angels,
huge portion of the Government budget is swallowed by new
projects involving construction of highways and railways,
both elevated and at-grade, sub ways, large span bridges and
water and wastewater treatment facilities. Construction
industry is further challenged by the citys subsoil condition
as it is plagued with a lot of constraints such as, presence of
thick soft clay as top strata, land subsidence and high risk of
flooding. These constraints compel most structures be relied
on pile foundation usually, on large diameter bored cast in-situ
piles, embedded into first or second sand layers. Though, the
first large bored pile was installed nearly two decades ago in
this city and hundreds of thousands of piles have already been
constructed, design, construction methods, performances and
effectiveness of these piles are seldom studied and rarely
reported.
The project site is located at the Rama III road, with east
boundary along the Chao Phraya river which is the largest of
its kind in the kingdom. The piles are usually of 1500mm and
1000 mm diameters. The maximum design working load on
piles are 500 tons and 1000 tons for 1000 mm and 1500 mm
piles respectively. Pile toe was designed to be embedded into
second sand layer at depth 55m from the ground level. Fig. 1
shows the general layout of the piles. All the piles are cast in
situ bored piles under wet process and base grouted.

Data from 3 bore holes within the project site

Fig. 2 Soil profile at the project site

PILE CAPACITY AND DESIGN CONSIDERATIONS
The load capacity of the piles were estimated for the soil
failure using the available soil data. Safety Factor of 2.5 was
used for both end bearing and shaft bearing capacities
considering the negative skin friction as well. The estimated
safe working load in tons, using individual three bore hole
data are 480, 508 and 583 for 1000 mm piles and 839, 880
and 993 for 1500 mm piles.
The estimated load capacity, with required safety factor, from
the soil bore log results is, in many cases, below the designed
safe working load. Hence, means of improving pile bearing
capacity was considered. The pile bearing capacity can be
improved either by extending the pile length further or by
grouting the pile or by enlarging the pile base. Usually,
enlarging the base for long piles is hardly practiced in
Bangkok. Piling contractors rather prefer to extend the length
of the shaft and additional required capacity is achieved by the
shaft friction. For the bored piles designed at this project site,
if an extra bearing capacity of 250 t/m2 could be achieved by

Fig. 1 General lay-out of bored piles

SUBSOIL CONDITIONS
The geotechinical investigation at the site indicates that a
strata of 14 -15m thick soft clay layer occurs in the project
area as common in Bangkok sub soil. This soft clay is

282

12

base grouting, the grouted piles would be capable of carrying

the designed safe working loads with adequate safety factor.

Slurry properties
The limiting values for the bentonite slurry according to the
contract specification were 1.04 - 1.20 for density; 30-40s for
viscosity (Marshs cone); not less than 7 for pH and not more
than 3% for sand content.

Based on the results from a project just near by the site and
past experiences in the vicinity it was proposed that the above
mentioned required capacity could be achieved by base
grouting the piles with tip founded at about 55 m from the
ground.

On average 3.5 % of bentonite concentration was used.

Bentonite properties were monitored before concreting for all
piles. Arbitrary checking was done during drilling. From the
bentonite results, no significant changes were observed. On
average, the viscosity observed was 32.8s (Marsh cone); with
maximum of 40s and a minimum of 29s. The average values
of 1.11g/cc, 8.9, 1.62% were noted for density, pH and sand
content respectively. The minimum values are 1.02g/cc, 7,
0.1% and maximum values are 1.11g/cc, 9 and 1.55
respectively. Eventhough, the sand content was set to be
within 3%, at times it was difficult to achieve this. In this
cases complete recycling of the slurry was done.

The option of extending the pile length was given up as the

clay layer is found to be at about depth of 60 m and thus
possibly higher initial settlement was expected if the piles are
embedded into the clay layer. Moreover, the end bearing
capacity would be reduced and this reduced end bearing
capacity might influence the shaft bearing capacity too. This
conclusion was reinforced by the previous pile load test
results on a pilot pile, founded at 60m and non base grouted.
This non grouted and longer pile of diameter 1500 mm,
constructed by some other piling contractor, failed at about
1750 tons, while the designed maximum test load was 2500
tons. The 1000mm diameter pilot pile was, though not
plunged to failure, settlement observed was too excessive for
the structural requirements. Load settlement characteristics of
these pilot test piles have been shown in Fig. 4(a) and Fig.
4(b). Finally the previous piling contractor was disqualified
and the new contractor was called upon.

Concreting and Reinforcement

Concrete cube strength of 350 ksc (35 Mpa) was used with
minimum cement content of 400 kg/m3. Concreting was done
by usual tremie method of 250mm tremie pipe. Concreting
was requested to be done with in twenty four hours after final
excavation. Since a concrete batching plant was established
within the project area, concrete supply was usually regular
and timely supplied otherwise it would have been a major
concern in Bangkok because of its ill fated traffic congestion.
Average concreting rate was about 25 - 30 cu.m per hour.
After pouring the concrete to the desired level, the temporary
casing was withdrawn.

CONSTRUCTION METHODS, MATERIALS AND

PROBLEMS
Wet process
The usual wet process using the bentonite slurry was adopted.
Temporary casing of length 15m was used to maintain the
hole stability in top soft layer. The holes were drilled using
augers and then continued using the bucket method. The
slurry was introduced upon reaching the first sand layer and
maintained all the time higher enough to support the bore. The
slurry was monitored continuously for the properties of
density, viscosity, pH and sand content.

Reinforcement was provided to full length of the pile

considering the possible tensile forces during construction
period of the structure. Maximum of 0.5% of cross sectional
area was used for both sizes of the piles.
All the main bars used for reinforcement are of high strength
deformed bars of 4000 ksc (400 Mpa) and round bars of 2400
ksc (240 MPa) were used for spirals. Reinforcement was
provided by fabricating in the form of cages as usual and the
maximum length of a cage was set to be 12 m. A minimum lap
length of forty times the diameter of the bars was used at the
joints of the cages.

Construction of bored piles started during the rainy season and

therefore the site needed additional requirements to improve
the working condition. As the water level was high in the
Chaophraya river flood protection was needed. Because of the
nature of the Bangkok sub soil and the project site was
previously used as urban waste dumping area, the site was
very swampy during the rainy season gave a lot of
inconveniences in solving problems for drainage of water
accumulated. The situation was further aggravated as the
excavated soil was not able to be removed on time. All these
reduced the productivity rate drastically. However, all the
piles were cast within the time frame set according to the
contract.

Inclusions in pile top and concrete level below trimming level

Few cases were observed where the concrete level was below
the designed cut off level and/or inclusions in pile top portion.
Mainly, this has been contributed by the high un- predicted
slumping down of concrete upon extracting the temporary
casing. As Fleming et al. (1985) pointed out concrete would
normally be retained by cohesive soils with shear strength in
excess of 15 kN/m2 and this limit is rarely achieved by the top
soft clay of Bangkok sub soil.

283

13

Final sound concrete level has been a talking point between

the main contractor and piling contractor for both cases where
the final concrete level is above and below the designed trim
level. Usually for the first case the main contractor suffered to
trim the long pile heads but in the latter cases the piling
contractor was called upon for remedial works. However, it
should not be neglected that the possibility of failure to over
flush enough concrete at the pile head to ensure that all
concrete present below the trim level is of full strength and
thus having the inclusions at the pile top.

the grout volume was the limiting factor. Usually the grouting
pressure was within the range of 20 - 30 bars.
Injection rate of grout was 10 - 15 liters/ min. Experiments on
grout samples collected from three of the bored piles
(1500mm dia.) in this site was performed by Anwar (1997).
Comparison of base grouted and soft base (non base grouted)
piles were studied and summarized in Table 1. It was
recommended that low injection rate about 1-2 liters/min
would improve the effectiveness of base grouting. However,
practicality and sustainability of grout pump, especially for
large volume of grout, can hardly justify this requirement.
Though the direct comparison of effect of base grouting is not
possible as the compared piles are founded in different soil
strata other parameters in Table 1 are worth enough to be
noted.

Verticality and Position Errors

Verticality tolerance was set to be 1 in 100 and no piles have
been found to be out of this tolerance. However, four piles
(about 1% of the total) have been found to be out of position
deviating the specified tolerance of 7.5cm plus the allowance
for verticality. However, many of them were found to be
under the mat on which the columns were to be erected.
Hence, dealing with these erroneous locations was not
dramatic and modification of reinforcement for the mat have
been done according to the redistributed load.

Table 1 Load distribution and settlement behavior of base

grouted and soft base piles.

GROUTING

Diameter(mm)
Base Depth (m)
Design Load Qd (ton)
Failure Load (ton)
Load at 20mm total settlement (QL)
% increase in QL
Portion of QL carried by shaft (QS)
% increase in QS
Portion of QL carried by base (Qb)
% increase in Qb
Total settlement at (Qd) (mm)
Total settlement at 1.5 Qd (mm)
* - Estimated

The effectiveness of base grouted piles in Bangkok sub soil

was studied by many previous researches. Teparaksa (1994)
suggests base grouting mainly aims to increase soil stiffness
beneath the pile base which was affected by the boring
process. Also he noted that for these piles the displacement at
fully mobilized skin friction is in the order of 1.59 to 1.86
percent of the diameter. Test results in this project too agrees
the range suggested.
The PE pipes (0.5) and tube-a manchette were in-corporated
into the reinforcement cages and lowered into the bore to
facilitate the base grouting. The grouting was done after the
concrete is set, usually the base was cracked using a high
water pressure not later than twenty four hours, and grout was
injected.

Base
Soft
Grouted
Base
1500
1500
55
60
1000
1000
2725 *
1750
2290
1520
50.66
1720
1450
18.6
570
70
714
5
8.13
7.5
15.08

PILE LOAD TEST AND RESULTS

Four piles, two from each size, were selected to perform the
static pile load test. All the test piles and anchor piles are
contract piles. The maintained load test was performed in two
cycles for each pile with the first cycle having maximum test
load of safe design working load. The maximum test load for
second cycle was 150 % of the design safe working load. The
maximum test load for these both cycles were maintained for
24 hours.
The third cycle was performed as a quick test and each
incremental load was maintained for 10 minutes. The
maximum test load on this third cycle was 250 % of the safe
design working load. The results were satisfactory and
complying the design requirements (Max. allowed settlement
of 10mm under working load)

Fig.3 Tube a manchette used for grouting

The grouting process was controlled by either maximum grout
volume of 500 litters or a maximum pressure of 40 bars. But,
the pressure of 40 bars was rarely achieved and thus usually

284

14

Ultimate load interpretation

Table 3 (b) CYCLE - 2

Predicted ultimate loads were based on graphical methods

proposed by Mazurkiewicz (1972), Fuller and Hoy (1970) and
Butler and Hoy (1977). It may be appropriate to mention here
that ultimate capacity determined from the above mentioned
methods generally not yield a load corresponding to a
plunging failure. The ultimate capacity is defined by the
load at which the pile moves rapidly (plunging) without any
further load increment. Most of the existing methods for
determining the ultimate capacity from pile load test, however,
are mostly based on some arbitrary displacement criteria and
there is no consensus on the best method of interpreting the
failure load. Fellenius (1980) pointed out that preferred
method of interpretaion often depend on an individuals
experience and he further recommended that several methods
be applied and that a preferred method be chosen on the basis
of the users needs and special conditions of the job.

Pile
No.

Dia.
(m)

Base
depth
(m)

TP - 1
TP - 2
TP - 3
TP - 4

1000
1000
1500
1500

55
55
55.3
55.5

Cycle 2 (Maintained Load Test)

Max.
Head
Elastic
Load
Movement Recover
(ton)
(mm)
y
(mm)
750
8.55
7.66
1250
17.40
12.49
1500
7.72
6.59
2500
27.19
16.00

Table 3 (c) CYCLE - 3

For base grouted piles in Bangkok sub soil, the predicted

failure load based on the above methods well agrees each
other (Teparaksa, 1994). For this particular project the
variation of predicted loads by these different methods are in
the range of 2.5% to 19%. It should be noted that the tests
were not performed to failure of piles. As the piles were not
tested for failure it was difficult to compare the results with
actual failure load.

Pile
No.

Dia.
(m)

Base
depth
(m)

TP - 2
TP - 4

1000
1500

55
55.5

Cycle 3 (Quick Test)

Max.
Head
Elastic
Load
Movement Recover
(ton)
(mm)
y
(mm)
1250
17.05
12.43
2500
29.26
17.52

Load test results on soft base pilot piles

Table 2 Ultimate load predicted by different methods

Max.
Ultimate Load (tons)
Test. Mazurk
Fuller
Butler and
Pile
Dia.
Load -iewicz and Hoy
Hoy
No.
(mm) (ton)
(1972)
(1970)
(1977)
TP - 1 1000
750
1625
1585
1575
TP - 2 1000
1250
1700
1480
1460
TP - 3 1500
1500
3600
3200
3185
TP - 4 1500
2500
3250
2775
2725
(Note: Piles were not tested to failure)

Fig. 4(a) Load test on pilot pile with non base grouted pile
(Diam. 1500mm, Toe embeddment at depth 60m
below ground)

Summary of pile load test results

Table 3 (a), (b) and (c ) summarize the pile load test results
and Fig. 5 and Fig. 6 show the corresponding load settlement
characteristics.
Table 3 (a) CYCLE - 1

Pile
No.

Dia.
(mm)

Base
depth
(m)

TP - 1
TP - 2
TP - 3
P-4

1000
1000
1500
1500

55
55
55.3
55.5

Cycle 1 (Maintained Load Test)

Max.
Head
Elastic
Load
Movement Recover
(ton)
(mm)
y
(mm)
500
4.69
4.52
500
4.66
3.98
1000
4.55
4.47
1000
5.47
5.43

Fig 4 (b) Load test on pilot pile with non base grouted pile
(Diam. 1000mm, Toe embeddment at depth 53m
below ground)

285

15

Load test results on base grouted contract piles

Almost all the cracks were indicated on piles adjacent to the

sheet pile cofferdam which was used for deep excavation.
These piles in many cases were repaired by coring and cement
grouted. Physical damages to piles located adjacent to the
excavation boundary were a major source of trouble as the
sheet piles used for temporary retaining purpose cannot
completely retain the soil movement and local soil strains
induced could be sufficient enough to cause cracking.
Possibility of providing more main reinforcement bars for
piles adjacent to the excavation zone to prevent these worse
effects would have been worth to have considered.

Fig. 5 (a) (proof) Load test on contract pile (TP - 1)

Dia. 1000mm, Toe depth 55m below Ground Level

Table 4 Interpretations by sonic integrity test

Available test results

Sectional increase (small)
Sectional increase
(prominent)
Sectional Decrease (small)
Intermittent sectional
variations (1)
Cracks (small) (1)
Cracks (prominent) (2)

Fig. 5 (b) (proof) Load test on contract pile (TP - 2)

Dia. 1000mm, Toe depth 55m below Ground Level

Diam.
1500 mm
nos.
%
300
100
25
8.3
04
1.3

Diam.
1000mm
nos.
%
59
100
10
16.9
02
3.4

04
10

1.3
3.3

04

6.8

10
03

3.3
1.0

08
-

13.6
-

Notes:
1. These variations were attributed to the sub soil conditions
2. All the cracks have been found on piles adjacent to the
excavation supported by steel sheet piles
CONCLUSION
Fig. 6 (a) (proof) Load test on contract pile (TP - 3)
Dia. 1500mm, Toe depth 55.3 m below Ground Level

A case history of cast in-situ concrete bored piles in Bangkok

sub soil has been summarized. A total of 402 piles of 1000mm
(96 nos.) and 1500mm (306 nos.) have been constructed as
foundation elements of a wastewater treatment plant.
Base grouting has been considered to improve the bearing
capacity and settlement behavior of bored piles. Maximum
pressure of 40 bars or 500 liter of grout volume has been used
as limiting criteria. Base grouting with tube -a manchette
method has been adopted and found to be effective.
Static pile load test results on contract piles and pilot test piles
have been summarized. Pilot piles with base non grouted,
have shown either high degree of settlement or failed well
before the designed maximum test load. Ultimate load
capacity has been estimated by different methods and
variation of the ultimate capacity found to be within 2.5 to 19
percent.

Fig. 6 (b) (proof) Load test on contract pile (TP - 3)

Dia. 1500mm, Toe depth 55.5 m below Ground Level
INTEGRITY TEST AND SIGNAL INTERPRETATIONS
It was requested to perform the integrity test on all the piles
cast. Commonly interpreted defects by the integrity testing
equipment developed by IBBC-TNO are shown in Table 4.

Pile integrity test was decided to be performed on all piles and

commonly indicated signal characteristics have been tabulated
for available data.

The integrity testing results showed clearly the physical

damages in many piles where the ill considered means were
practiced to cut the piles to trimming level.
286

16

ACKNOWLEDGEMENT
The authors are much thankful for all the staff and employees
of SEAFCO CO., LTD. especially for Mr. Kamol
Yooyuenpattana and Mr. Chetsadar Plongkrathok.
REFERENCES
Anwar M. A [1997], Base grouting to improve the
performance of wet process bored piles in Bangkok sub soils,
AIT Master Thesis, Bangkok, Thailand.
Fleming, W. G. K., F. Reiding and P. Middendorp. [1985]
Faults in cast in place piles and their detection, Proc.
Second Intern. Conf. On Structural Faults and Repair., ICE,
Westminister, London, pp. 301 - 310.
Joshi, R. C., and H. D. Sharma [1987] Prediction of ultimate
pile capacity from load tests on bored and belled, expanded
base compacted and driven piles Proc. Intern.. Symp. On
Prediction and Performance in Geotechnical Engineering,
Galgary, pp. 135 - 144.
Ng Kim Cheng [1983], The construction problems and
performance of large bored piles in second sand layer, AIT
Master Thesis, Bangkok, Thailand
Teparaksa, W. [1994], Newly developed toe-grouted bored
pile in soft Bangkok clay; Performance and Behaviour,
Intern. Conf. On Design and Construction of Deep
Foundations, FHWA, Florida, pp. 1337 - 1351..
Woo Siu Mun and C. H. Juang [1994], Analysis of pile load
test results, Conf. On Deep Foundations and Ground
Improvement Schemes, Bangkok, Thailand.

287

17

PROCEEDINGS OF THE 3RD INTERNATIONAL GEOTECHNICAL SEMINAR ON DEEP

FOUNDATIONS ON BORED AND AUGER PILES/ GHENT/ BELGIUM/19-21 OCTOBER 1998

Deep Foundations on Bored

and Auger Piles
BAP III
Edited by

W.F.VAN IMPE & W.HAEGEMAN

Laboratory of Soil Mechanics, Ghent University, Belgium

Large diameter bored piles in multi-layered soils of Bangkok

N. Thasnanipan, P. Tangseng & M. A. Anwar
Seafco Company Limited Bangkok, Thailand

A.A. BALKEMA / ROTTERDAM / BROOKFIELD / 1998

19

Deep Foundations on Bored and Auger Piles, Vam Impe & Haegeman (eds) 1998 Balkema, Rotterdam, ISBN 90 5809 022 1

Large diameter bored piles in multi-layered soils of Bangkok

N. Thasnanipan, P. Tanseng, M. A. Anwar
Seafco Co., Ltd., Bangkok, Thailand

ABSTRACT: Static pile load test results of ten large diameter bored piles founded in the multi-layered soils
of Bangkok are discussed in this paper. All piles were instrumented with vibrating wire strain gauges
(VWSG) and telltale extensometer rods. Range of shaft friction and end bearing values mobilized in different
soil horizons are reported. Progressive type of failure mechanism which is peculiar with the long piles in
multi-layered soils, and found to be initiated from the soil layers present at the middle reach of pile shaft and
then extending towards the top and bottom layers, is also presented. Brittle type of failure observed in some
sand and very stiff to hard clay layers is also discussed. Settlements at design loads calculated from global
factor of safety concept and limit load concept are also compared. Additionally, back calculated
and values are also compared with the previously recommended values for similar soils.
1 INTRODUCTION
Bored cast in situ piles are extensively used as a
foundation element in Bangkok due to the
requirements of huge loads to be transferred and
limitations of using driven piles like limited
capacities, associated soil movements and some
unwanted environmental effects. Diameters of these
piles normally fall in the range of 0.80 to 1.50 m and
toe depths down to 60m from ground level are quite
common. Soil profiles of plain of Bangkok are
always evident of the presence of marine Bangkok
Soft Clay (BSC) at the top changing to medium at
about 15 to 18m. First sand layer is usually 5 to 10m
thick and found at 25m to 30m depth, below is a
series of stiff to hard clay and medium to very dense
silty sand layers (Ref. to the Figs. 5 to 14 for typical
soil profiles). Actual pore pressure conditions in the
upper BSC are hydrostatic from circa 1m below
ground level. Then the hydrostatic conditions
changes to piezometric draw down near the bottom
level of BSC. Piezometric draw down resulted in
increased effective overburden pressure of about 20
ton/m2 in the first sand layer and below as shown in
Figure 1. The under drainage in the Bangkok soils is
attributed to the deep well pumping in the area.
According to the latest reports, piezometric draw
down conditions have started recovering since last
few years due to the control over deep well pumping
and shifting of industrial areas away from Bangkok
metropolitan.

Figure 1. Variation of effective overburden pressure

with depth, (after KERDSUWAN, 1984)
2 METHOD OF CONSTRUCTION
Bentonite slurry in conjunction with rotary bucket is
the normal drilling procedure of pile construction
when piles are to be founded in the first sand or
below. Top 15 to 18m depth of soft to very soft clay
is almost always temporarily cased to assure the
stability of borehole. Firstly, auger is used to drill

511

21

with unconfined compression test while for deeper

stiff to hard clay layers Cu is indirectly estimated
from Standard Penetration Test (SPT-N) data as Cu
= C1.SPT-N (ton/m2), with C1 equal to 0.674 and
0.507 for high and low plasticity clays respectively.
C1 values mentioned here are based on the statistical
analysis carried out by Pitupakorn, 1985. Skin
friction capacities for the sand layers are calculated
as fs = v.Ks.tan, with coeficient of horizontal
earth pressure Ks equal to 0.7 and equal to 0.75.
Angle of internal friction is also estimated from
SPT-N values by first correcting the N values for
overburden correction (Bowels, 1988). Some
designers also use an equivalent term of instead of
Ks.tan. It must be noted that the empirical
parameters discussed here are based on the effective
overburden pressure v which is calculated with the
assumption of hydrostatic conditions from the
ground level, ignoring the piezometric draw down
conditions below BSC shown in Figure 1.

within the temporary casing followed by rotary

bucket with bentonite slurry down to final depth of
excavation. Special cleaning bucket or air lift
technique is normally applied to clean the borehole
base of any congregated sediments before lowering
the reinforcement cage which is followed by tremie
concreting. Bentonite slurry viscosity is normally
maintained with in the range of 30 to 50 seconds
(Marsh cone viscosity) and construction time,
starting from the casing driving till the completion
of concreting, usually fall in the range of 10 to 20
hours excluding some accidental delays due to
equipment break down, unavailability of ready
mixed concrete or similar reasons.
3 DESIGN PARAMETERS
A combination of total stress method for clay layers
and effective stress method for sand layers is
normally adopted due to its simplicity and
limitations regarding unavailability of reliable
effective stress parameters. There is no standard
criteria of acceptable settlement for these piles, but a
limit of 5mm at working loads is normally accepted
for high rise buildings.

3.2 End bearing

Ultimate end bearing capacity for the sand layers is
calculated as Nq.v with bearing capacity factor Nq
in the range of 5 to 20 depending upon the relative
density of the sand layers. A limit of 500 ton/m2 for
maximum end bearing is normally used for piles
seated in the second sand layer. For clay layers
ultimate end bearing is estimated as Nc.Cu with
bearing capacity factor Nc equal to 9. Global factor
of safety of 2.5 on the accumulated ultimate skin
friction and end bearing is normally used to
calculate the safe working loads.

3.1 Skin friction

Skin friction capacity of clay layers is estimated as fs
= .Cu, with adhesion factor varying from 0.9 to
0.3 depending upon the undrained shear strength of
clay layers. Undrained shear strength Cu of shallow
clay layers is usually determined in the laboratory

Table 1. Principal data of test piles with interpreted failure loads.

14

2.0xQd1

13

15

16

17

QL
Limit load
at a
settlement of
1% of Dia.
(ton)

Gross
Qd3
settlement Difference of
at Qd3
Qd3 as
Design
(mm)
compared to
load
'=QL/1.5
original Qd1
(%)
(ton)

6=(4/5)

18

19

20

21

TP-1 1.2x57.10 650

1582

1230

1.29

1435 1715

1575

630 0.97 6.8 13.2 35.5

975

650

6.8

0.0

TP-2 1.2x46.25 650

1929

930

2.07

2815 2017

2416

966 1.49 5.5

9.3

15.0

1140

760

5.5

16.9

TP-3 1.0x46.51 500

1118

800

1.40

1477 1651 1544 1557

623 1.25 4.5

8.2

12.3

920

610

5.5

22.0

TP-4 1.0x49.47 450

970

930

1.04

1172 1270 1138 1193

477 1.06 6.3 12.8 40.5

650

430

5.5

-4.4

TP-5 1.0x43.00 450

1106

690

1.60

1399 1642 1347 1463

585 1.30 4.5 11.1 34.5

675

450

4.5

0.0

TP-6 1.0x41.00 360

641

620

1.03

759 997

878

351 0.98 3.2

5.0

7.2

830

550

5.0

52.8

TP-7 1.2x43.6

450

1499

850

1.76

891 1390

1141

456 1.01 2.5

4.2

5.4

1150

770

4.5

71.1

TP-8 1.2x43.5

450

519

640

0.81

1012 1180 978

1057

423 0.94 3.3

5.9

48.5

850

570

4.7

26.7

TP-9 1.0x43.5

400

701

530

1.32

867 946

975 929.3 372 0.93 4.0

5.2

36.5

750

500

4.9

25.0

TP-10 1.2x54.00 500

799

1100

0.73

3.5

6.4

16.3

1000

670

5.0

34.0

512

22

12

Qd1

11

Qd2/Qd1

10

Qd2=
Qu ave/2.5

Average Failure
Load
Qu ave

Chin-Konder's

Mazurkiewicz's

Mobil/Estim
Skin
Friction

Brinch-Hansen's

Friction
(ton)

De'Beer

Dimensions Design Skin

(Dia. x
Load Friction
Depth)m (ton)
(ton)

Davisson

Pile
No.

SETTLEMENT(mm) AT

Failure Load (ton)

1.5xQd1

Total Estimated
Skin
Qd1 Mobilized

than the estimated skin friction values with the

exception of piles TP8 and TP10 who have 19% and
27% less capacity respectively. It must be noted that
TP10 was not tested to failure so ultimate shaft
capacity was not yet mobilized.
Unit skin friction values mobilized in different
soil layers are given in Table 2. It must be noted that
the unit skin friction values reported here are the
actual maximum mobilized capacities only. Since
piles were not instrumented to measure the
corresponding displacements at the interface of each
soil horizon, minimum values are not reflecting the
ultimate unit friction values.

4 TEST PILES
Test piles discussed in this paper were all
constructed following the procedure discussed in
section 2. All piles were instrumented with VWSG
at five to seven different levels to estimate the shaft
friction transferred to different soil horizons. One set
of telltale extensometer rods was also used near the
tip of each test pile to measure the elastic shortening
of the pile shaft and finally pile base movement at
different stages of load testing. Static maintained
load testing method was used for all piles. Test loads
were applied using a system of hydraulic jacks
against the reaction frame of steel girders fixed
against anchored reaction piles. Normally three
cycles of loading are applied with first cycle up to
the design load Qd and maintained for 12 hours
followed by second cycle of loading up to 2 times of
Qd and maintained for 24 hours. Third cycle of
loading is applied up to 2.5 times Qd or maximum
pre-decided test load. Some times a fourth quick
loading cycle up to the maximum test load is also
applied and maintained for two hours. All test piles
except TP10 were tested to well above the expected
failure loads with the concept of sufficiently
mobilizing the end bearing.

Table 2. Skin friction values mobilized in different

soil horizons and corresponding values.
Soil Type

SPT-N

Ave.

Skin Friction

Cu

Mobilized (ton/m2)

(Ave.)

Average

Maximum

Soft to medium Clay

2-12

4.7

6.0

13.9

1.3

Stiff to very stiff

20-35

14.0

7.5

18.8

0.54

Hard Clay

40-46

21.5

7.3

24.6

0.34

First silty Sand

18-68

11.2

19.9

Second silty Sand

20->100

15.6

25.1

Clay

5 TEST LOAD RESULTS

Average skin friction values mobilized in the soft to

medium clay layers seems to be even higher than
average Cu of these layers, most likely reasons for
this are: 1) actual diameter in the soft to medium
clay layer is slightly higher than the nominal
diameter of the pile due to the use of temporary
casing; 2) actual Cu of the top few meters of
weathered crust is higher than the Cu used in the
analysis. Values of reported in Table 2 are quite
conforming to the previous recommendations like
Ng, (1983) and Tomlinson (1995). Skin friction
values mobilized in the very stiff to hard clay layers
are relatively low, since at failure loads considerable
portion of the pile head settlement is absorbed by the
long pile shaft itself which reduces the relative
pile/soil interface movement near the deeper soil
layers which results in partial mobilization of skin
friction capacity of these soil horizon. In other
words, in case of long piles the geotechnical
capacity of deeper soil layers may not be fully
exploited before they are declared to be failed due to
the excessive pile head movement, same is the case
for end bearing of these piles. Average values of unit
skin friction mobilized in the first sand and second
sand layers are 11.2 ton/m2 and 15.6 ton/m2
respectively which corresponds to a value of 0.45
to 0.55, without considering piezometric draw down
discussed in section 1. But if we consider
piezometric draw down, corresponding values will
be 0.23 to 0.28. This confirms the suggestions made

Test load results with pile dimensions are

summarized in Table 1 and Figures 5 to 14. It is
quite evident that the safe design loads (Table 1,
col.13) calculated from the average of failure loads,
interpreted by different available methods (cols.711), are in reasonable agreement with the estimated
design loads (col.3). But the settlements at the
design loads (col.15) have a wide scatter between
2.5mm and 6.8mm. So instead of computing failure
loads from methods mentioned in Cols. 7-11, if we
use the limit load concept at settlement of 1% of the
diameter (col. 18) and then use a partial FOS of 1.5
to calculate design loads (col.19), resulted
settlements at this design load (col. 20) will be quite
uniform and within the limits of allowable
settlement at design loads for high rise buildings.
Global FOS against overall ultimate failure will still
be in excess of 2.5 against the ultimate failure of the
pile which still have not been reached even at large
pile head movements (Buttling, 1992). This can also
be confirmed from the maximum end bearing values
mobilized under maximum test loads i.e. only 25%
of the ultimate end bearing is mobilized (ref. Section
5.2)
5.1 Skin friction
Actually mobilized shaft friction capacities (ref.
Table 1, col.4) are on the average by 50% higher

513

23

considerable portion of the total pile head settlement

is absorbed by the elastic shortening of the pile
shaft. So even at 3 to 3.5 times working loads
maximum pile tip movement is less than 7% of the
diameter of the pile. Ultimate end bearing values
which need a pile base settlement of 20% to 30% of
the pile base diameter (Tomlinson, 1995) are still
only 25% mobilized.
Maximum mobilized end bearing for piles with
tips in the sand layers are plotted in Figure 4. It is
clear that the estimated end bearing value of

by Meyerhof (1976) who concluded that value is

also dependent on the length of the pile and can be
as low as 0.15 for very long piles. Maximum unit
skin friction mobilized in sand layers is also
comparable to test results by Reese and ONeill
(1988) who measured a maximum value of 19.15
ton/m2 for sand layers.
Dense sand and very stiff to hard clay layers
found to exhibit a brittle type of failure mechanism
in majority of the cases, a typical family of curves
representing such type of behavior are shown in
Figure 2. Residual shaft friction in dense sand layers
with increasing butt loads found to be dropped to
50% of the maximum mobilized values in some
cases. For very stiff to hard clay layers, such
reduction is found to be as high as 60%.

Dense Sand
Hard Clay

V.Stiff Clay

Hard Clay

12

Soft Clay
V. Stiff Clay

Dense Sand

9
Skin Friction Mobilized (ton/m
)

Mobilized Skin Friction (ton/m )

10

14

Dense Sand

10

Dense Sand
Hard Clay
Hard Clay

Hard Clay
Hard Clay
Hard Clay

6
4
2

V.Stiff Clay

200

400

600

800

1000

Butt Load (ton)

3
2

Figure 3: Skin friction mobilization in different soil

horizons for TP10

1
0
0

200

400

600

800

1000

1200

Butt Load
(ton)

Table 3. Maximum end bearing mobilized and

corresponding pile base settlement from telltale data.

Figure 2: Family of curves showing typical brittle

failure in dense sand and very stiff to hard clay
layers.

Pile
No.

TP-1
TP-2
TP-3
TP-4
TP-5
TP-6
TP-7
TP-8
TP-9
TP-10

Failure mechanism of different soil horizons with

increasing butt loads exhibit a progressive type of
behavior for some test piles. A typical, unit skin
friction development with increasing butt load
(TP10) is shown in Figure 3. In number of cases
dense sand and very stiff to hard clay layers found to
reach their ultimate unit friction values well before
the maximum test loads which shows that pile head
displacements required to mobilize ultimate shaft
friction in these layers are quite small. An other
point to be noted here is that the progressive failure
mechanism found to be started from the layers
present in the middle reach of the pile shaft and then
it extended towards top and bottom. Once the sand
and stiff to hard clay layers started yielding most of
the loads were shifted to the medium and soft clay
layers at the top.

Max. End Estimate Average

Max. Load
Elastic
Total
Net
% of
Bearing d End SPT-N Soil Type near
Transfered to Compressi Settlement Settlement
Pile
Mobilized Bearing Near
the Toe
the Base
on of Pile of Pile Head of Pile Base
Dia.
2
2
VWSG (ton)
(mm)
(mm)
(mm)
(ton/m ) (ton/m ) the Toe

418
71
382
230
194
199
101
481
199
201

25.7
16.2
18.3
16.3
13.0
9.1
15.8
9.1
14.1
8.0

94.1
34.7
81.6
90.7
60.5
13.8
58.4
90.9
69.0
16.3

68.4
18.4
63.3
74.4
47.5
4.7
42.7
81.9
55.0
8.3

5.7
1.5
6.3
7.4
4.7
0.5
3.6
6.8
5.5
0.7

370
63
487
293
247
254
89
426
253
178

500
500
500
206
430
500
183
500
500
228

100
100
70
45
35
50
40
60
60
50

Clayey Sand (SC)

SAND (SP-SM)
Silty Sand (SM)
Silty CLAY (CH)
Silty Sand (SM)
Silty Sand (SM)
Silty CLAY (CL)
Silty Sand (SM)
Silty Sand (SM)
Silty CLAY (CL)

500 ton/m2 in most of the cases correspond to a base

settlement of approximately 7% of the pile diameter
and this level of settlement have rarely been
achieved even at such a high pile top displacements.
End bearing mobilized in the sand layers for
different values of base settlement (ref. Figure 4)
match well with the recommendations of Reese and
ONeill (1988) who recommended a value of critical
end bearing equal to 5.76*SPT-N (ton/m2)
corresponding to a base settlement of 5% of the pile
base diameter for sand layers of similar relative
density. Since the average value of SPT-N for test
piles founded in sand layers is approximately 60,
critical end bearing comes out to be 5.76*60 = 350
ton/m2 and from the test piles plotted data Figure 4,

5.2 End bearing

End bearing values mobilized under maximum test
loads are given in Table 3. It is quite evident that

514

24

500

TP3

SPT-N (Blows/ft)
0 20 40 60 80 100 0

400

TP1

Load (Tons)
1000 1500 2000

2500

Soft to Medium
CLAY (CH)

10

300

20

TP5

250
200

TP9

30

TP2 (1.2m x 46.25m)

350

500

Fill Material

TP8

Depth (m)

Mobilized End Bearing (t/m

450

Stiff to Very Stiff

CLAY

Dense SAND
Very Stiff CLAY
Dense SAND
Very Stiff CLAY

150
40

TP2

100
50

Dense to Very Dense

SAND

TP-2

R = 0.889

50

0
0

3
4
% (Base Sett./Dia. of
Pile)

Applied Load (Tons)

0

500

2000

Fill Material

10

1500

2000

20
Pile Head Movement (mm)

Load (Tons)
1000
1500

1000

Figure 4: Development of end bearing with base

settlement as % of pile diameter.
SPT-N (Blows/ft)
0 20 40 60 80 100

500

Soft to Medium
CLAY (CH)

40
60
80
100

TP-2

30

40

TP1 (1.2m x 57.1m)

Depth (m)

20

120
Medium to Dense
SAND (SM or SC)

Figure 6: Test pile TP-2

Stiff CLAY
(CH)

50

SPT-N (Blows/ft)
Very Dense SAND
(SC)

TP-1

60

Load (Tons)

20 40 60 80 100 0

400

800

1200

1600

Top Soil

Soft Clay

10

Applied Load (Tons)

500

1000

1500

2000
Depth (m)

Pile Head Movement (mm)

20

20

30

40

TP3 (1.0m x 46.51m)

Medium to Stiff Clay

Medium to Dense
Sand
Stiff Clay
Medium Dense Sand

Very Stiff to Hard

Clay

60

40
Dense to Very Dense
Sand

80

TP-3

50

100
120

TP-1

Applied Load (Tons)

0

500

1000

1500

Figure 5: Test pile TP-1

Pile Head Movement (mm)

20

critical end bearing corresponding to 5% base

settlement is 270 ton/m2, a difference between the
predicted value of end bearing is due to the reason
that end bearing values predicted by Reese and
ONeill (1988) are based on the SPT data with
energy ratio (ER) of 0.55, while SPT-N used for the
test piles are not corrected for ER.
Back calculated average value of Nc for clay

40
60
80
100
120

TP-3

Figure 7: Test pile TP-3

515

25

2000

SPT-N (Blows/ft)
0

20

40

60

400

800

Load (Tons)

SPT-N (Blows/ft)

Load (Tons)
80 0

1200

1600

Medium Stiff Clay

20

40

60

80 0

200

400

600

800

1000

Fill Material

Very Soft to Soft

Clay

Medium Dense Sand

Depth (m)

Depth (m)

30

TP4 (1.0m x 49.47m)

Very Stiff Clay

20

Soft to Medium
CLAY (CH)

10

Loose Sand

Very Stiff Clay

20

TP6 (1.0m x 41.0m)

10

Very Stiff CLAY

(CL)

30

Dense to Very Dense

Sand

40

Dense to Very Dense

(SM)

40
Hard Clay

50

Dense to Very Dense

Sand

TP-6

50

TP- 4

60

500

1000

1500

Pile Head Movement (mm)

Applied Load (Tons)

2000

Pile Head Movement (mm)

20
40
60
80
100

40

60

80 0

500

Load (Tons)

SPT-N (Blows/ft)

1000

1500

Fill Material

Soft to Medium
CLAY

20

40

60

80 0

400

800

1200

Fill Material

Soft to Medium
CLAY (CH)

10

30

40

(1.
0m
x
43.
0m
) .

20

Stiff CLAY

Depth (m)

De 20
pth
(m)

TP
5

Dense to Very Dense

SAND

30

TP7 (1.2m x 43.6m)

10

TP- 6

Load (Tons)

SPT-N (Blows/ft)
20

20

Figure 10: Test pile TP-6

Figure 8: Test pile TP-4

0

40

TP-4

120

Applied Load (Tons)

500
1000

Very Stiff CLAY

(CH)
Very Stiff CLAY
(CL)

Dense to Very Dense

SAND (SM)

40
Hard CLAY (CL)

TP-5

50

50

TP-7

Dense to Very Dense

SAND (SM)

Applied Load (Tons)

0

500

1000

1500

2000

Pile Head Movement (mm)

Pile Head Movement (mm)

20
40

60
80

100
120

Figure 9: Test pile TP-5

Applied Load (Tons)

500
1000 1500 2000

20
40
60
80
100
120

TP- 5

TP- 7

Figure 11: Test pile TP-7

26

Load (Tons)

SPT-N (Blows/ft)
0 20 40 60 80 100

400

1200

0 20 40 60 80 100 0

Very Soft to Medium

CLAY (CH)

500

1000

Fill Material

10

Very Soft to Soft

CLAY

30
Hard CLAY (CH, CL)

40

30

20

Medium Dense
SAND (SM)

TP10 (1.2m x 48.9m)

Stiff CLAY (CH-CL)

Depth (m)

TP8 (1.2m x 43.5m)

20

Stiff to Very Stiff

CLAY

Dense to Very Dense

SAND

40
Dense to Very Dense
SAND (SM)

TP-8

Very Stiff CLAY (CH)

50

50

Very Stiff to Hard

Clay

Applied Load (Tons)

0

500

1000

1500

60
Applie d Loa d (Tons)

20

60
80
100

TP- 8

0 20 40 60 80 100 0

20

40

Load (Tons)

SPT-N (Blows/ft)

500

1000

TP- 10

Figure 14: Test pile TP-10

Fill Material

layers is 8 with 4.5, 6.0 and 12.8 for the three piles
seated in clay layers. From Table 3 it can be
confirmed that the end bearing values in clay layer
also depend upon the base settlement and values of
Nc more than 9 can also be mobilized at high base
settlement ratios. An other possible reason for the
value of Nc more than 9 may be the underestimation
of undrained shear strength of clay from correlation
with SPT-N discussed in section 3.1.

Very Soft to Medium

CLAY (CH)

10

Stiff CLAY (CH-CL)

Medium Dense
SAND (SM)

30
Hard CLAY (CH, CL)

40
Dense to Very Dense
SAND (SM)

TP-9

Very Stiff CLAY (CH)

6 CONCLUSIONS

Applied Load (Tons)

0

500

1000

1500

1. Skin friction and end bearing values of wet

process bored piles mobilized in the subsoils of
Bangkok are reported and compared with the
previous researches made for similar cases.
2. Dense sand and very stiff to hard clay layers
often exhibit a brittle type of failure with
residual friction capacities as low as 50 to 60%
of the peak mobilized values.
3. Failure of long piles in multi-layered soil is
dominantly progressive in nature and some soil
horizons reach their ultimate friction capacities
well before the over all ultimate failure loads.
4. Calculation of ultimate failure loads from the
normally available methods and then design

2000

20
Pile Head Movement (mm)

50

1000

40

Figure 12: Test pile TP-8

20

500

120

TP-10

2000

Pile Head Movement (mm)

Pile Head Movement (mm)

TP9 (1.0m x 43.5m)

Depth (m)

10

Load (Tons)

SPT-N (Blows/ft)

800

Fill Material

40

60
80

100
120

TP- 9

Figure 13: Test pile TP-9

516

27

loads by applying a global FOS give widely

scattered settlements at design loads and if we
calculate design loads from limit load concept,
quite uniform settlement at design loads can be
achieved.
AKNOWLEDGEMENTS
Writers gratefully acknowledge the continuous
support and invaluable suggestions made by their
colleagues, especially Mr. Aung Win Maung and
Mr. Ganeshan Baskaran from analysis to final draft
of this paper.
REFRENCES
Bengt B.Broms and M.F.Chang (1988), Bored Piles
in Residual Soil and Weathered Rocks in
Singapore, Deep Foundations on Bored and
Auger Piles, Van Impe (ed.), Balkema,
Rotterdam.
Bowles, J.E.(1988), Foundation Analysis and
Design, McGraw-Hill.
Buttling, S. (1992), Prediction of the Performance
of Some Instrumented Pile Tests in Hong
Kong, Proc. of Symposium on Prediction Vs
Performance in Geotechnical Engineering,
SEAGS, Bangkok.
Meyerhof, G.G. (1976), Bearing Capacity and
Settlement of Pile Foundation, Jour. Of
Geotch. Engg. Div., ASCE Vol102, No GT3,
pp195-228.
Ng, K.C. (1983), The Construction Problems and
Performance of Large Diameter Bored Piles in
Second Sand Layer, M. Engg. Thesis no. GT82-26, AIT, Bangkok.
Pitupakorn, W. (1985), Predictions of prestressed
concrete pile capacity in Bangkok stiff clay and
clayey sand, 8th Southeast Asian Geotechnical
conf, Kuala Lumpur.
Reese and O Neil (1988), Drilled Shafts: Design
Procedures and Design Methods, US Dept. of
Transportation
F.
H.A.
Office
of
implementation, McLean, Va.
Tomlinson, M.J. (1995), Pile Design and
Construction Practice, 4th Ed. Chapman & Hall.

517

28

PROCEEDINGS OF THE 3RD INTERNATIONAL GEOTECHNICAL SEMINAR ON DEEP

FOUNDATIONS ON BORED AND AUGER PILES/ GHENT/ BELGIUM/19-21 OCTOBER 1998

Deep Foundations on Bored

and Auger Piles
BAP III
Edited by

W.F.VAN IMPE & W.HAEGEMAN

Laboratory of Soil Mechanics, Ghent University, Belgium

Effect of construction time and bentonite viscosity on shaft capacity of

bored piles
N. Thasnanipan, G. BAskaran & M. A. Anwar
Seafco Company Limited Bangkok, Thailand

A.A. BALKEMA / ROTTERDAM / BROOKFIELD / 1998

29

Deep Foundation on Bored and Auger Piles, Avan Impe & Haegeman (eds) 1998 Balkema, Rotterdam, ISBN 90 5809 022 1

Effect of construction time and bentonite viscosity on shaft capacity of bored

piles
N. Thasnanipan, G. Baskaran & M. A. Anwar
Seafco Co., Ltd., Bangkok, Thailand

ABSTRACT: Effect of construction time and slurry viscosity on the shaft friction capacity of cast in situ
bored piles have been studied by various investigators in the past but the extent to which these parameters
effect the shaft capacity is still not clear. Major obstacle in this regard is that it is hard to normalize the shaft
capacity degradation against other numerous parameters which are influencing simultaneously. Results from
eleven pile load tests, constructed with different slurry viscosities and construction times, in the layered
alluvial strata of Bangkok are presented in this paper.
It has been concluded that slurry viscosity, do not have significant effect on the shaft load transfer of these
piles but considerably reduced with increase in construction time. Trend of reduction in capacity seems to
follow an exponential decrease with increase in construction time with major part of degradation within first
24 hours of construction time.
60 m from ground level are quite usual. Soil profile
of plain of Bangkok generally consists of Quaternary
alluvial deposits of alternating sand and clay layers
as shown in Figure 1, down to the rock face which is
reported to be at least 550 m deep in the area
(Balasubramanium, 1991).

1 INTRODUCTION

Average Engineering
Parameters

P o re P re s sure (t/m )
0

20

30

Bangkok Soft Clay (BSC)

(CH)

40

Su = 10 - 20 KPa
t = 14-16 KN/m3
-1 0

Medium to Very Stiff Clay Su = 40 - 140 KPa

-2 0
(CH)
t = 17-21.0 KN/m3

-3 0

c
tati
ros

SPT-N = 20 - 50
t = 20.0 KN/m3

Su = 250 KPa -4 0

Lin
wn
Do

t = 21.2 KN/m3

w
Dra

Hard Clay (CL)

etric

40

Lin

om

iez
al P

Dense to Very Dense

Sand (SM)

Actu

Depth (m)

10

20

Weathered Crust

Hy

Shaft friction capacity of bored piles, especially, in

granular soils is greatly affected by the installation
procedures, but the sole effect of bentonite slurry
(slurry) properties and its exposure time is still
unclear. Results from the previous researches on
deleterious effects of slurry properties and its contact
time on the shaft load transfer of wet process bored
piles are a bit anecdotal and to some extent
contradictory. Major problem in this regard is that,
the effect of either parameter can not be totally
isolated from other influencing variables. Due to
this reason researches have been found to report
almost converse conclusions for the similar type of
soils.
In order to assess the effect of slurry viscosity and
exposure time on the shaft load transfer, data from
eleven instrumented bored pile load tests is
presented in this paper. All test piles were
constructed using same type of equipment and
procedures, in a similar type of soil conditions. So it
is assumed that the influence of other variables
would be minimized and the effect of slurry
viscosity and exposure time, to some extent, could
be estimated.

50

60

-5 0
Dense to Very Dense
Sand (SM)

SPT-N = 50 - 100
t = 20.0 KN/m3

-6 0

2 BORED PILES AND SOIL CONDITIONS

Figure 1. Typical soil profile of Bangkok plain with
piezometric draw down conditions.

Use of bored cast in situ piles in Bangkok, is very

common. Diameters of these piles normally fall in
the range of 0.80 to 1.50 m and toe depths down to

171

31

60

Below the top few meters of weathered crust, a thick

layer of well known Bangkok Soft Clay (BSC)
changing to medium at about 15 to 18 m, is present.
First sand layer is usually 5 to 10 m thick and found
at 25 m to 30 m depth, below is a series of stiff to
hard clay and medium to very dense silty sand
layers. Actual pore pressure conditions in the upper
BSC are hydrostatic from approximately 1 m below
ground level. Then the hydrostatic conditions
changes to piezometric draw down near the bottom
level of BSC, Figure 1. Piezometric draw down
resulted in increased effective overburden pressure
of about 20 ton/m2 in the first sand layer and below.
The depressurization of sand layers is attributed to
the deep well pumping in the area, Nutalaya (1981).

noted that major part of the reduction i. e. 43 percent

took place in the first 8 hours of contact time and
only 13 percent (56-43) reduction took place in the
rest of 89 hours (97-8), for the second barrette.
These findings support the concept that major part of
the shaft friction capacity reduction with increase in
construction time, took place in first few hours of
construction time and further increase in
construction time have minor contribution towards
the reduction in shaft friction capacity.
Corbette (1975), presented the results of load testing
of a diaphragm wall panel excavated in conjunction
with bentonite suspension, and construction time of
approximately 48 hours. He inferred that the
presence of bentonite suspension even for such a
long time has not adversely affected the
development of skin friction.

3 REVIEW OF PREVIOUS WORK

Degradation of perimeter load transfer in permeable
soil layers is attributed to the formation of filter cake
which is left in place and get sandwiched at soil-pile
interface and is not scrapped of by fluid concrete.
Scouring capability of fluid concrete depends upon
its shear strength and that of the filter cake and it is
argued that if shear strength of filter cake is more
than the fluid concrete, it can not be scoured and left
in place resulting in the degradation of load transfer
capacity of the shaft. In case of impermeable soils
reduction in the shaft load transfer is mainly due to
the softening of the soil at the perimeter of the
borehole, because the formation of thick filter cake
is not possible as slurry can not permeate in to the
soil, but some researchers e. g. Veder, C. (1963)
observed a thin cake of few millimeters even in clay
formations and argued that it may be the result of
electrical forces or chemical reaction of bentonite
suspension on the wall of the borehole.

Littlechild and Plumbridge (1998), concluded from

the pile load tests of BERTS project Bangkok, that
shaft friction capacity of bored piles constructed
under slurry tends to decrease as the construction
time and slurry viscosity increases.
Majano and ONeil (1993) attempted to model the
formation of filter cake in the laboratory with
different slurry dosages, differential pressures and
exposure times. They stated that perimeter load
transfer is a complex function of the physical and
chemical characteristics of the slurry and the
geomaterials, the roughness of the borehole, the
fluid pressures exerted by concrete, the shearing
properties of the soil, and possibly the chemistry of
the fluid concrete. They attempted to correlate the
potential degradation of the soil-pile interface to the
thickness and shear strength of the cake formed
against the walls, and argued that the shear strength
and thickness can not be measured with only one or
two parameters like slurry dosage and differential
pressure. With a bentonite concentration of 72
kg/m3 and differential pressure of 0.5 psi they
achieved a filter cake thickness of 3.14 and 4.5 mm
after a contact time of 4 and 24 hours. Authors also
compared the laboratory tests with the actual field
conditions and proved that due to the presence of
enormously high differential pressures in the field in
some cases, thicknesses of filter cakes which would
be considered detrimental for shaft load transfer,
require few hours to develop (They proved that a
filter cake of 6 mm thickness would require only 2.3
hours to develop in the field under a differential
pressure of 10 psi).

ONeil et al. (1992), reported the effect of slurry

properties on two drilled shafts constructed in
similar ground conditions (alternating layers of very
stiff clay and dense sand). Shafts 1 and 2 were
constructed with Marsh cone viscosity of 37 and 49
sec.
respectively, time between opening the
borehole and completion of concreting for both piles
was 5 to 7 hours. Minor difference in load transfer
of both shafts was found and authors concluded that
the difference in slurry properties used for the
construction for both shafts had a little effect on
shaft load transfer.

Cernak (1976), performed full-scale load tests on

three barrettes in sandy gravels, two barrettes were
constructed using slurry with different exposure
times, 8 and 97 hours. Load test results indicated a
4 METHOD OF CONSTRUCTION
decrease of skin friction capacity of 43 percent and
56 percent, respectively, as compared to the barrette
All the eleven piles reported in this paper were
excavated dry and concreted immediately. It must be
constructed, following the internationally accepted
172

32

guidelines and specifications. Slurry in conjunction

with rotary bucket is used for all piles.
Temporary casing, having length depending upon
the thickness of the BSC at the site is driven using a
high frequency vibro-hammer. Drilling operation is
commenced with an auger down to the base of the
temporary casing. Further drilling below the bottom
of temporary casing is continued using a drilling
bucket by first filling the borehole with slurry.
Bentonite, classified as Activated Sodium Bentonite
is usually used and slurry properties are controlled
within the limits given in Table 1.

must also be noted that the actual contact time for

slurry starts, once it is fed to the borehole till the up
flowing concrete reaches the bottom level of
temporary casing. Since the accumulated time
required for concreting above the bottom level of
temporary casing and auger drilling inside the
temporary casing roughly required one hour,
construction times given in Table 2 are supposed to
be appropriate for comparison.

Table 1. Slurry specifications followed.

Slurry Property
Control Limits

A commonly accepted design approach which

consists of a combination of total stress method for
clay layers and effective stress method for sand
layers is adopted to estimate the shaft friction
capacity reported in Table 2, col. F. Skin friction
capacity of clay layers is estimated as fs = .Cu,
with adhesion factor . taken from the curve
proposed by Suchada (1989), Figure 2, for Bangkok
subsoils. It can be noted that value of used is quite
comparable to the previous recommendations by
Tomlinson (1957) and Kulhawy (1984).

30 - 60
1.04 - 1.15

pH

7 - 11

Sand Content

<4%

Outer diameter of the drilling bucket is slightly less

than the nominal size of the pile to provide an
annular by-pass for the slurry during lowering and
hoisting operations. Required nominal size of the
pile is achieved using a cutting edge with
sidecutters. Slurry level in the borehole is
maintained within 1 to 1.5 m from top of temporary
casing. After drilling down to the required tip level
of pile, special cleaning bucket and/or air lift
technique is normally applied to clean the borehole
base of any congregated sediments before lowering
the reinforcement cage. Concreting is done using a
tremie pipe at the center of the borehole. Total time
to carry out this operation, starting from the casing
driving till the completion of concreting, usually fall
in the range of 10 to 20 hours excluding some
accidental delays due to equipment break down,
delay in ready mixed concrete due to traffic
congestion or similar reasons. In most of the cases
whole operation of pile construction is accomplished
within 24 hours, but in some cases, due to the local
authority regulations which do not permit to operate
during night time hours, borehole has to be left
exposed for 10 to 12 hours unagitated. In such
cases, drilling is intentionally curtailed above the
final excavation level and resumes next day. We
assume that in such cases, during drilling after a
delayed period of 10 to 12 hours, any possible filter
cake on the borehole walls is automatically scraped
of. Construction times more than 24 hours reported
in Table 2 are examples of such cases. Construction
time reported in Table 2, col. D is the time elapsed
between the start of auger drilling to the finishing of
concreting. Longer construction times for the test
pile reported are usually due to the time consumed in
instrumentation installation with the rebar cage. It

1.20
Tomlinson,1957

1.00

Density (g/cc)

Adhesion Factor ,

Viscosity (Marsh cone)

5 SHAFT CAPACITY ESTIMATION

Bored Cast Insitu

Piles

Stas and Kulhawy, 1984

0.80
Suchada,1989
for bored piles in Bangkok soils

0.60

0.40

0.20

0.00
0.00

5.00

10.00

15.00

20.00

25.00

30.00

Undrained Shear Strength (ton/m )

Figure 2. Comparison of Adhesion factor used for bored piles

in Bangkok Subsoils.

Undrained shear strength Cu of shallow clay layers

is normally determined in the laboratory with
unconfined compression test while for deeper stiff to
hard clay layers, Cu is indirectly estimated from
Standard Penetration Test (SPT-N), as Cu =
C1.SPT-N (ton/m2), with C1 equal to 0.674 and
0.507 for high and low plasticity clays respectively.
C1 values mentioned here are based on the statistical
correlations suggested by Pitupakorn, 1985. Skin
friction capacities for the sand layers are calculated
as fs = v. Ks.tan. with coefficient of horizontal
earth pressure Ks equal to 0.7 and equal to 0.75 .
173

33

Table 2. Construction parameters and shaft capacities of test piles.

Pile No.

Pile
Dimensions
(m)

Bentonite
Slurry
Viscosity
(sec.)

Construction
Time
(hrs)

Pile Age at
the Time of
Load Testing
(days)

TP-1

1.2x57.1

38

43.0

32

TP-2

1.2x46.3

38

13.1

29

TP-3

1.0x46.5

37

9.8

25

TP-4

1.0x49.5

37

38.7

TP-5

1.0x43.0

38

26.0

TP-6

1.0x41.0

45

TP-7

1.2x43.6

55

TP-8

1.2x43.5

TP-9

1.0x43.5

Actual Shaft
Capacity
(ton)

Actual/Estimated
Shaft Capacity

H=(G/F)

1230

1200

0.98

930

1300

1.40

800

1250

1.56

15

930

750

0.81

19

690

700

1.01

11.8

32

620

800

1.29

16.8

32

850

1100

1.29

38

12.3

24

640

750

1.17

37

11.3

18

530

700

1.32

Estimated
Shaft
Capacity
(ton)

TP-10

1.2x62.0

41

32.4

27

1534

2000

1.30

TP-11

1.2x54.2

38

30.0

39

1050

1300

1.24

for 12 hours followed by second cycle of loading up

to 2 times of Qd and maintained for 24 hours. Third
cycle of loading is applied up to 2.5 times Qd or
maximum pre-decided test load. Occasionally, a
fourth quick loading cycle up to the maximum test
load is also applied and maintained for two hours.
All test piles, except TP-2 and TP-10, were tested to
well above the expected failure loads with the
concept of sufficiently mobilizing the end bearing.
Load displacement curves of the test piles are shown
in Figure 3 for comparison.

Value of Ks used is in good agreement with the

recommended values by different researchers like
Fleming (1977), who suggested a value not more
than 0.75 to account for the possible degradation due
to the formation of filter cake. Angle of internal
friction is also estimated from SPT-N values by
first correcting the N values for overburden
correction (Bowels, 1988). Some designers also use
an equivalent term of instead of Ks.tan. It must
be noted that the empirical parameters discussed
here are based on the effective overburden pressure
v which is calculated with the assumption of
hydrostatic conditions from the ground level,
ignoring the piezometric draw down conditions
below BSC shown in Figure 1.

7 ACTUAL SHAFT FRICTION CAPACITIES

Actually mobilized shaft friction capacities are
calculated from the VWSG load transfer results at a
level where load displacement curves show a
maximum curvature and are given in Table 2, col. G.

6 TEST PILES
Test piles discussed in this paper were all constructed
following the procedure discussed in section 4. All
piles were instrumented with Vibrating Wire Strain
Gauges (VWSG) at five to seven different levels to
estimate the shaft friction transferred to different soil
horizons. One set of telltale extensometer rods was
also used near the tip of each test pile to measure the
elastic shortening of the pile shaft and finally pile
base movement at different stages of load testing.
Static maintained load testing method was used for
all piles. Test loads were applied using a system of
hydraulic jacks against the reaction frame of steel
girders fixed against anchored reaction piles.
Normally three cycles of loading are applied with
first cycle up to the design load Qd and maintained

8 EFFECT OF SLURRY VISCOSITY

Effect of slurry viscosity on the shaft load transfer
can best be seen from the plot between the ratio of
Actual/Estimated shaft capacity versus slurry
viscosity, Figure 4. Over all scatter of data do not
suggest any trend. Since the effect of slurry viscosity
is difficult to separate, without normalizing the
effect of construction time. Test piles TP-6, 7 and 9
have a slight difference in construction time and pile
age at load testing but slurry viscosity is
significantly different, 45, 55, and 37 sec.
respectively but the ratio of Actu./Esti. shaft
capacities for these piles are necessarily the same
174

34

5, plot between Actu./Esti. shaft capacity versus

construction time. Though the data is widely
scattered, best fit curve shows the trend of
exponential decrease of shaft load transfer with
increase in construction time. Best fit curve for the
available data also reveals that the available shaft
capacity is more than the estimated capacity for a
construction time as long as 40 hours. This confirms
the validity of recommendations made by some
researchers like Fleming (1977), who suggest to
complete the concreting within 24 hours after
finishing the drilling, and seems to be more
reasonable because slurry is kept continuously
agitated during drilling operation and actual
exposure time for slurry starts, once the drilling is
completed. Test pile TP-4 which exhibited the
lowest capacity and has the minimum slurry
viscosity (37 sec.), second longest construction time
(38.7 hours) and the minimum age (15 days) at the
time of load testing, in the group reveals that the
shaft capacity reduction is contributed by increased
construction time as well as reduced pile age while
slurry viscosity have minor effect.
It can also be noted that the major part of the shaft
capacity reduction is contributed by the first 24
hours of construction time which normally
considered feasible and agreeable by most of the
piling contractors under normal circumstances, and
further delays beyond 24 hours have a minor
addition to the shaft load capacity degradation.
These findings are in good agreement with the
previous observations discussed in section 3.

Applied Load (Tons)

0

500

1000

1500

2000

2500

0
20

TP-1

Movement (mm

TP-2
TP-3

40

TP-4
TP-5
TP-6

60

TP-7
TP-8
TP-9

80

TP-10
TP-11

100
120

Figure 3. Load settlement curves of the test piles.

which suggest that the sole effect of slurry viscosity

on shaft capacity reduction is insignificant.
Additionally, TP-1, 4 and 5 which exhibit a
maximum reduction in shaft capacity have almost
same slurry viscosity but different construction time
which confirms that slurry viscosity within the data
set available do not directly have any effect on the
shaft load transfer of the piles.
9 EFFECT OF CONSTRUCTION TIME
Since ten out of eleven bored piles were constructed
with slurry viscosity varying within narrow range
effect of construction time can best be seen in Figure
1.60

25

Actual / Estimated Shaft Capacity

TP-1
1.40

29(Pile age at load testing, days)

18

27

TP-2
TP-3

32

32

39

1.20

TP-4

24

TP-5
TP-6

19
32

1.00

TP-7
TP-8
TP-9

15

0.80

TP-10
TP-11

0.60
35

40

45
50
Bentonite Viscosity (sec)

55

Figure 4. Effect of slurry viscosity on shaft capacity of bored piles.

175

35

60

Actual / Estimated Shaft Capacity

1.60

25

TP-1(Tip in SAND)
TP-2(Tip in SAND)

29 (pile age at load testing, days)

1.40
18
1.20

32

TP-3(Tip in SAND)

27

32

TP-4(Tip in CLAY)

39

TP-5(Tip in SAND)

24

TP-6(Tip in SAND)
TP-7(Tip in CLAY)

19

1.00

32

TP-8(Tip in SAND)
TP-9(Tip in SAND)

15

0.80

TP-10(Tip in
SAND,Base Grouted)
TP-11(Tip in
SAND,Base Grouted)

0.60
5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

50.0

Construction Time (hours)

Figure 5. Effect of construction time on shaft capacity of bored piles.

10 DISCUSSION

or Maintained Load (ML). Reduction in construction

time, generally, lead to better performance of pile
shafts but this reduced construction time could
never exclude the accidental occurrence of delays.
Any potential degradation in load carrying capacity
of such piles caused by uncontrollable delays must
be considered along with the conceivable trend in
reduction under these conditions.
It must also be noted that because of high
differential pressures imposed due to depressurized
sand layers in Bangkok sub soils (differential
pressures of the order of 30 psi will be imposed if
slurry level is maintained near the ground level due
to piezometric draw down of 20 m in the sand
layers), formation of filter cakes which are
considered to degrade the shaft load transfer would
theoretically take fraction of seconds to develop, if
we calculate on the basis of procedure recommended
by Majano (1993). This means that the degradation
of pile shaft interface due to the formation of filter
cake in permeable soil layers is unavoidable within
the practical construction time limits.

From Figure 5 it can be noted that actually

mobilized shaft capacity is equal to the estimated
shaft capacity for a construction time of
approximately 35 - 40 hours or in other words, the
empirical design parameters commonly used to
estimate the shaft friction capacities discussed in
section 5, are justifiable for a total construction time
of longer than 24 hours, especially, to account for
the unexpected delays during construction in the
downtown areas of Bangkok which are notorious for
its traffic congestion and other constraints imposed
by local authorities. This also confirms that the
allowable time of 24 hours after finishing the
drilling operation (Fleming, 1977) is also justifiable.
However, with no dispute, shorter the construction
time better the shaft load transfer response.
It must therefore be emphasized that the selection
of empirical parameters like , , and Ks, to
estimate shaft load capacities at the design stages
must also consider the practically possible
construction times by suitably incorporating the
effects of any possible delays specific to the site and
working conditions. Empirical factors mentioned
above are not the fundamental properties of the soilpile system, and need to be accounted for the
installation procedures or even the method of pile
load testing like Constant Rate of Penetration (CRP)

11 CONCLUSIONS
1.

176

36

Bentonite slurry viscosity within the data set

available do not have significant effect on the
shaft load transfer of bored cast in situ piles.

2.

3.

4.

Longer construction time reduces the shaft

carrying capacity of bored piles constructed
under slurry. However, the major portion of
degradation occurs within the first 24 hours
period, which is unavoidable under normal
construction time.
Degradation of shaft capacity with increase in
construction time beyond 24 hours is not
excessive and can be adjusted within the
empirical design parameters or factor of safeties
chosen at the design stages.
Piles constructed within 24 hours showed
generally higher shaft capacities than estimated
as per normal design practice adopting
empirical parameters suggested by different
researchers for Bangkok subsoils and can safely
be used for slurry viscosity up to 55 sec. and
total construction time up to 40 hours.

Suchada, (1989), Bored Piles in Bangkok Subsoils, , M. Engg.

Thesis, AIT, Bangkok.
Tomlinson, M. J. (1957) The adhesion of Clay Piles Driven in
Clay Soils, Proc. 4th Intl. Conf. on Soil Mechanics and
Foundation Engineering, London.
O Neil et al.(1992), Behaviour of Axially Loaded Drilled
Shafts in Beaumont Clay, The University of Texas at
Austin.
Veder, C. (1963), An Investigation on The Electrical
Phenomena At The Area of Contact Between Bentonite
Mud And Cohesionless Material, 5th Intl. Conf. on Soil
Mechanics, Paris.

AKNOWLEDGEMENTS
Writers wish to acknowledge the co-operation of
Mr. Aung Win Maung and Mr. Pornpot Tanseng, in
the preparation of this paper.
REFERENCES
Baker, Jr. C. N. et al. (1970), Load Carrying Capacity
Characteristics of Drilled Shafts Constructed with the Aids
of Drilling Fluids, Research Report 89-6, Center for
Highway Research, The University of Texas at Austin, TX.
Balasubramanium, A. S. (1991), Inaugural Lecture on
Contributions in Geotechnical Engineering Soil Mechanics
and Foundation Engineering, AIT, Bangkok, March 14,
1991.
Bowles, J.E.(1988), Foundation Analysis and Design,
McGraw-Hill.
Cernak, B. (1976), The Time Effect of Suspension on the
Behaviour of Piers, 6th European Conf. On Soil Mechanics
and Foundation Engineering, Vienna.
Corbett, B. O. et al. (1975), A Load Bearing Diaphragm Wall
at Keinsington and Chelsa Town Hall, London, Conf. on
Diaphragm Walls and Anchorages, ICE, London.
Fleming, W. K. (1977), The Use and Influence of Bentonite in
Bored Pile Construction, CIRIA Report PG3.
Littlechild, B. and Plumbridge, G. (1998), Effects of
Construction Technique on the Behaviour of Plain Bored
Cast In Situ Piles Constructed Under Drilling Slurry, 7th
Intl. Conf. and Exhibition on Piling and Deep Foundations,
Vienna.
Majano, R. E. and ONeil (1993), Effect of Slurry Dosage and
Exposure Time on Perimeter Load Transfer of Drilled
Shafts, ADSC Publication.
Nutalaya, P. (1981), Subsidence Studies of the Bangkok Plain,
AIT Research Reports.
Pitupakorn, W. (1985), Predictions of Prestressed concrete pile
capacity in Bangkok stiff clay and clayey sand, 8th
Southeast Asian Geotechnical conf., Kuala Lumpur.
Stas, C. V. and Kulhawy, F. H. (1984), Critical Evaluation of
Design Methods for Foundations Under Axial Uplift and
Compression Loading, Report EL-3771, EPRI, Palo Alto,
Calif.

177

37

PROCEEDINGS OF THE SYMPOSIUM IN HONOR OF

PROFESSOR SENG-LIP LEE
May 14-15, 1999
Asian Institute of Technology, Pathumthani, Thailand

Innovative Solutions in Structural and

Geotechnical Engineering
Edited by

Pisidhi Karasudhi and Anil C. Wijeyewickrema

Asian Institute of Technology, Thailand

K. Y. Yong and T. Balendra

The National Institute of Singapore

Performance comparison of bored and excavated piles

in the layered soils of Bangkok
N. Thasnanipan, M.A. Anwar, Maung A. W. and P. Tangseng
Seafco Company Limited Bangkok, Thailand

Asian Institute of Technology

National University of Singapore

39

Symposium on Innovative Solutions in Structural and Geotechnical Engineering

in Honor of Professor Seng-Lip Lee, May 14-15, 1999

PERFORMANCE COMPARISON OF BORED AND EXCAVATED PILES, IN

THE LAYERED SOILS OF BANGKOK
Narong Thasnanipan, Muhammad Ashfaq Anwar, Aung Win Maung, Pornpot Tanseng
SEAFCO Co., Ltd., Bangkok, Thailand.
ABSTRACT: Installation techniques are considered to effect the performance of cast in situ piles.
This paper compares the load transfer characteristics of a fully instrumented bored pile (Dia. 1.5 m)
and barrette (1.5 x 3.0 m) having same length equal to 57 m, embedded in identical ground
conditions (on same site, 30m apart). Total time consumed to construct the pile and barrette differ
considerably, 27 and 75 hours, respectively. No significant difference in load transfer has been
observed between the pile and the barrette. It has been concluded that, by observing good engineering
practices, any conjectured degradation effects on load capacity of such piles can be eliminated.
Additionally, the load applied on barrette was one of the highest test load in the region with a
reaction frame tower capacity of 6000 tons. Reaction frame and loading system arrangement, to
achieve such a high load capacity is also discussed.
1. INTRODUCTION
Barrettes become a common foundation element in combination with bored piles and diaphragm
walls in Bangkok (Refer: Thasnanipan et al, 1998-b). Foundations for the BECM tower project
located at Rama IX road Bangkok, Thailand, required 560 bored piles of 1.2 m and 1.5 m diameter
and 24 number of barrettes having cross section size of 1.5 m x 3.0 m. Foundation plan of the tower
is shown in Figure 1. Barrettes were necessary to transfer the very high loads coming from the
central 52 story towers lift shafts. Piles were approximately 57.0 m long with tips embedded into
the second sand layer. Depending upon the load requirements piles were base grouted in the central
tower area and at the locations of high column loads. Load testing on working piles was required to
verify the design parameters, and it was proposed to test one bored pile of 1.5 m diameter and one
barrette with full instrumentation. Since, pile and barrette were constructed using two different
techniques and the parameters like construction time, bentonite properties etc, were considerably
different, results from the load tests have provided a unique chance to assess the extent of difference
in behavior, which have been speculated to effect considerably in the past by some theoreticians.
2. SOIL PROFILE
Soil profile at the site is no different than the typical soil profiles of plain of Bangkok and is shown in
Figure 2. Below the top 2 m of weathered crust 11.5 m thick Bangkok Soft Clay is present. First stiff
clay starts from 13.5 m depth and extends to 26 m, below is the First sand layer and is about 10 m
thick. Below the First sand layer a thick layer of Hard clay which extends down to 54 m, is present,
and then starts the second sand layer which extends down to about 80 m where the next hard clay
layer starts. Undrained shear strength, Su obtained from unconfined compression test and Standard
Penetration Test (SPT) are shown in Figure 2. Ground water table depth at the site generally varies
from 1 to 1.5 m, piezometeric conditions of the site are also quite similar to the other parts of the
Bangkok with piezometric draw down of about 22 m near the bottom of the first stiff clay layer.

41

346
132.10

96.20

Test-Barrette

Test
Pile

132.68

94.00

Figure 1. Foundation footprint of BECM tower project, showing the location of test pile and barrette.
3. CONSTRUCTION PROCEDURE
3.1 Bored Pile
Wet process method with rotary drilling bucket was employed. Bentonite slurry conforming to the
widely accepted specifications were used as supporting fluid. Top 15 m soft clay was temporarily
cased to assure the stability of the borehole. Firstly, auger was used to drill within the temporary
casing, followed by rotary bucket with bentonite slurry down to final depth of excavation. Airlift
technique was applied to clean the borehole base of any congregated sediments. Before lowering the
reinforcement cage special cleaning bucket was used to scrap of the borehole walls and the base.
Reinforcement cages were lowered inside the borehole by simultaneously attaching the
instrumentation at specified locations. Soon after lowering the rebar cage tremie concreting was
started. Properties of the bentonite slurry used are given in Table 1. Polystyrene grains plug was
used before the first charge of concrete to avoid the mixing of bentonite with concrete. Time
consumed in different construction activities is plotted in Figure 3.
3.2 Barrette
Mechanical rope-grab in conjunction with bentonite slurry was used to excavate the trench. A guide
wall cast with inside clear dimensions slightly more than the nominal size of the barrette was used to
guide the grab during initial bites. Since time consumed in the preparation of instrumentation was
quite long, refer Figure 3, desanding was continuously done to keep the bentonite slurry agitated,
which also helps to minimize the growth of filter cake by actually reducing the exposure time.

42

347
0

DEPTH (m)

Max. Mobilized Skin Friction

(ton/m 2)

BUTT LOAD (TON)

Su (UC) t/m2
5.0
10.0
1000

2000 3000

1000

2000

3000

4000

5000

0.0

10.0

-20

First Stiff
Clay

-30

First
Sand

-40

30.0

Bored Pile

Soft
Clay

-10

20.0

Weathered
Crust

Barrette

Hard
Clay

-50
Second
Sand

-60

Tip Level
-57.5 m

Tip Level
-57.5 m

60
1.50 m
0

50

3.00 m

100

60

SPT-N (blows/30cm)

VWSG

Bored Pile

1.50 m

Barrette

Mechanical Extensometer

Figure 2. Comparison of bored pile and barrette, with soil profile at the site.
As another measure, trench was once again occupied by grab to scour the trench walls and to
remove, if any, filter cake formed on the walls (Refer: Drilled Shaft Inspectors Manual, 1989). It is
authors opinion that, if due to some unforeseen reasons, reinforcement cage lowering have to be
delayed for considerable period of time or for overnight. It is a good practice to use the grab again to
scrap the trench walls which eliminates any foreseeable bad effects due to such unexpected delays.
After lowering the rebar cage, tremie concreting was done.
Table 1. Comparison of bentonite slurry properties.
Pile

Barrette

Before feeding to the borehole

After Recycling & Before

concreting
(near
borehole base)

Before feeding to the

trench

After Recycling & Before concreting

(near trench base)

Viscosity (sec)

33

36

36

49

Density (g/cc)

1.08

1.10

1.10

1.17

pH value

Sand Content (%)

0.1

0.8

1.0

1.1

43

348
B arrette
D rilling
D eSanding
C age Low ering
Trem ie Pipe Preparation
C oncreting

B ored P ile

10

20

30

40

50

60

70

80

T im e Co n su m e d (h o u rs)

Figure 3. Comparison of time consumed in different construction activities.

4. INSTRUMENTATION
Both, pile and barrette were instrumented with Vibrating Wire Strain Gauges (VWSGs) and
Mechanical Extensometers (ME) at five levels along the pile shaft at the known interface boundaries
of different soil layers. Pile has four sets of VWSGs and one set of ME at each level. While the
Barrette has six sets of VWSGs and two sets of ME at each level, as depicted in Figure 2.
5. LOAD TEST ARRANGEMENT
Reaction frame anchored against the four working piles was used to load the piles. Reaction
frame used to load the barrette was one of the biggest of its kind used in the region with overall
height equal to about 8 m as shown in Figures. 4a and 4b. Five numbers of built-up steel girders
supported on each side by two 1st level cross girders were used to achieve the required maximum
frame capacity of 6000 tons. 1st level crossbeams were supported against the 2nd level cross girders.
2nd level cross girders were then anchored against the surrounding barrettes using anchor blocks at
the top. Especially fabricated rigid transfer girders as shown in Figure 5 were used to distribute the
tension force coming from the Tie-bars to the dowel bars above the anchor barrette heads. Sixteen
number of hydraulic jacks each having 500 ton capacity, were installed between the barrette cap and
main girder of the reaction frame. Jacks were arranged in a symmetrical fashion as shown in Figure
4b, to avoid any possible eccentricity of loading. Ball bearings were used between the main girders
and hydraulic jacks to keep the line of loading in the true vertical direction.
6. EFFECT OF SHAPE ON SHAFT LOAD TRANSFER
Assessment of shape effect on the shaft load transfer has been attempted by various researchers
in the past. Hosoi et al (1994), concluded that the earth pressure acting on the flat surface of a
rectangular diaphragm wall panel is larger than that of circular bored piles. Numerical analysis
performed by them showed that the earth pressure on the flat surface of a trench with L/B = 1000
(plain strain conditions), where, L and B are the length and width of the trench in plan respectively, is
higher than that of circular borehole. In order to assess the effect of different aspect ratios (L/B) on
the earth pressure developed around the trench, attempt has been made to model the behavior of
trench using finite element computer program. Figure 7 shows the orientation of principal stresses
and stress contours developed around the trenches with different L/B ratios. It has been observed that
sufficient hoop compression stresses develop around the trench and if L/B <3. But if L/B is
increased above 5, longer sides of the trench start yielding and it can be concluded that under such
conditions hoop stresses around the trench become less effective which give rise to the increased

44

349

Figure 4a. View of the test frame having maximum capacity of 6000 tons.
5.00 m

5.00 m

2nd Cross Beam

I 350 x 2000 x 8000 mm.

B20

1st Cross Beam (R)

2 x I 600 x 2700 x 10000 mm.

5.00 m

Main Beam
5 x I 400 x 2000 x 9000 mm.
B13
B14

B15

5.00 m

Hydraulic Jacks

Reference Beam

B
B9

2nd Cross Beam

I 350 x 2000 x 8000 mm.

1st Cross Beam

2 x I 600 x 2700 x 10000 mm.

2nd Cross Beam

I 350 x 400 x 7500 mm.

Figure 4b. Barrette load test plan.

45

350

Figure 5. Stiffening girder used to distribute force to dowel bars.

Figure 6. A view of the instrumented barrette cap with jacks.

46

351

L/B = 1

L/B = 3

L/B = 5
Figure 7. Orientation of principal stresses and stress contours around the excavated trenches with
different aspect ratios.

47

352
earth pressure on flat surface of trenches with large aspect ratios. Although the exact determination
of earth pressure developed on the surface of trench depends on the soil properties and the over
burden pressure at the level of consideration, it can be concluded that no significant difference in
earth pressure is observed between the bore hole (L/B=1) and the barrette (L/B =2).
7. RESULTS OF LOAD TEST
Values of unit skin friction developed at the interface of different soil horizons along the pile
and barrette shafts are given in Table 2 and are also plotted in Figure 2. Comparison of unit skin
friction values prove that there is no significant difference in shaft load transfer between pile and
barrette, even though the construction methodology adopted is different. These findings are in line
with the conclusions made by Thasnanipan et al (1998-a). Load settlement curves of the pile and
barrette are shown in Figure 8.
Table 2. Comparison of unit skin friction and end bearing mobilized for the test pile and barrette.
Mobilized Skin Friction (ton/m2)
Depth

Soil Type

Pile

(m)

Barrette

At Design Load Max. Mobilized

0 - 13.5
13.5 - 25.0
25.0 - 35.0
35.0 - 50.0
50.0 - 55.0

Bangkok Soft Clay (CH)

Stiff - V. Stiff Clay (CH)
Med.-Dense Silty Sand (SM-SP)
Hard Silty Clay (CH)
3.5m Hard Clay + 1.5m Clayey Sand (SC)
Maximum Mobilized End Bearing (Tom/m2)

0.20
5.90
11.46
3.09
3.43

2.63
7.01
22.76
11.11
14.84

At Design Load

Max. Mobilized

0.26
3.58
13.25
3.55
3.19

2.24
8.86
25.28
9.61
7.49

270

101

A p p l i e d L o a d (T o n s)
0

500

1000

1500

2000

2500

3000

3500

4000

P ile H e a d M o ve m e n t (m m

20
B ar r ette
Pile

40

60

80

Figure 8. Comparison of Load Settlement Curves of Barrette and Pile.

48

4500

5000

5500

353
8. CONCLUSIONS
1) No significant difference in shaft load transfer has been observed between the wet process bored
pile and barrette, in spite of the presence of considerable difference in construction techniques
followed and other relevant parameters like bentonite slurry viscosity and construction time.
2) All foreseeable negative effects attributed to construction parameters can be eliminated by
observing good engineering practices.
3) Reaction frame can be designed and erected to apply very high compression loads of the order of
6000 tons for pile load testing.
ACKNOWLEDGEMENTS
The authors wish to thank Dr. Wanchai Teparaksa and their colleague Mr. Kamol Singtogaw for
their invaluable suggestions. They also wish to thank Mr. Natamon of STS Engineering Consultants
Co., Ltd. for providing the necessary data.
REFERENCES
Drilled Shaft Inspectors Manual (1989), Prepared by The Joint caisson Drilled Shaft Committee of
the ADSC: The International Association of Foundation Drilling and DFI: Deep Foundation
Institute, NJ. USA.
Hosoi, T, Yagi N & Enoki, M. (1994), Consideration to the Skin Friction of Diaphragm Wall
Foundation, 3rd Intl. Conf. on Deep Foundation Practice incorporating PILETALK, Singapore.
Static Load Testing Report of Barrette Pile for BECM Tower Project, Rama IX Road, Prepared by
EDE-STS Engineering Joint Venture, Bangkok, Thailand.
Static Load Testing Report of Bored Pile for BECM Tower Project, Rama IX Road, Prepared by
EDE-STS Engineering Joint Venture, Bangkok, Thailand.
Thasnanipan, N, Baskaran G. & Anwar, M.A. (1998-a), Effect of Construction Time and Slurry
Viscosity on Shaft Capacity of Bored Piles, Proc. of the 3rd Intl. Geotechnical Seminar on Deep
Foundations on Bored and Auger Piles, Ghent, Belgium, Balkema.
Thasnanipan, N., Maung, A.W. & Tanseng, P (1998-b), Barrettes Founded in Bangkok Subsoils,
Construction and Performance, Proc. of the 13th Southeast Asian Geotechnical Conf. Taipie,
Taiwan, R.O.C.

49

PROCEEDINGS OF THE ELEVENTH ASIAN REGIONAL CONFERENCE ON SOIL MECHANICS

AND GEOTECHNICAL ENGINEERING : SEOUL/ KOREA/ 16-20 AUGUST 1999

Soil Mechanics and

Geotechnical Engineering

Eleventh Asian Regional Conference

Edited by
S.W.Hong, M.M.Kim, G.S.Yang, S.R.Lee, S.S.Chung, C.C.Ihm, H.T.Kim
J.B.Park & B.S.Lee

Base grouting of wet process bored piles in Bangkok subsoil

Teparaksa W.

Chulalongkorn University, Thailand

Thasnanipan N., Anwar M. A.

Seafco Company Limited Bangkok, Thailand

A.A. BALKEMA / ROTTERDAM / BROOKFIELD / 1999

51

Eleventh Asian Regional Conference on Soil Mechanics and Geotechnical Engineering, Hong et al. (eds)
1999 Balkema, Rotterdam, ISBN 90 5809 053 1

Base grouting of wet process bored piles in Bangkok subsoils.

W. Teparaksa

Chulalongkorn University, Bangkok, Thailand.

N. Thasnanipan & M.A. Anwar

Seafco Company Limited. Bangkok, Thailand.

ABSTRACT: Base grouting techniques are widely used to rectify the soft toe problem of bored piles in Bangkok. Two different
techniques, commonly known as flat jack and tube-a-manchette are used. This research study is mainly focused on the tube-amanchette technique. Six piles were constructed with the provision of PVC-casing inside the shaft to allow the drilling through the
pile bases in order to collect the soil samples below the pile tip. Piles were base grouted by varying the controlling parameters such as
grout volumes, pressures and injection rates. Soil samples were collected from beneath and at some distance away from the pile tip,
and analyzed. It is observed that grout does not penetrate into the surrounding sand of the pile toe even at high pressures. It partially
replace and/or precompress the sediments present under the pile tip and most of the grout just rises up the soil/pile interface by
replacing the bentonite cake. Recommendations are made to eliminate the soft toe problem more effectively.
0

2 BASE GROUTING IN BANGKOK

Teparaksa (1991-1992) reported the construction and
performance of base-grouted bored piles in Bangkok subsoils. It
was found that base grouting not only helps to improve the end
bearing capacity but skin friction as well (ref: Table 1).
Table 1. Improvements made by base grouting (Teparaksa, 1994).
Layer of
Pile tip

Ist Sand
2nd Clay
2nd Sand

Increase
in

failure
load
(%)
26-66
27
12-24

Increase

Increase

in skin
friction
(%)

in end
bearing
(%)

24-66
51
9-27

28-61
1
11-21

NGB: Non Grouted Base, GB: Grouted Base

Displacement at fully
mobilized skin friction

-10

X- Section through pile base

1
2
Manchettes
Grouting
circuit
(U-loop)

FIRST STIFF
CLAY

-20
DEPTH
(m)

FIRST
SAND

-30

-50

-60
-65

PVCCASING

SECOND STIFF
CLAY

-40

SECOND SAND

The method of constructing bored piles under bentonite

suspension has become well established in Thailand. It is
particularly used for large diameter piles, where powerful rigs
can be used with rotary tools to bore through otherwise unstable
water bearing strata. It is almost always used with a temporary
casing to support the top most Bangkok Soft Clay (BSC),
resulting in a hybrid of the casing and bentonite methods.
Construction problems are affecting integrity and
performance of such piles. A pile without a sound toe and
without a proper seat into the virgin ground should prove
disastrous particularly if required to act in point bearing. Even
otherwise, such piles could lead to higher orders of detrimental
settlements. The known causes of ending up with a defective
pile toe, are the loosened materials during boring operation and
accumulation of sediments from the bentonite column at the base
of the borehole. The technology of flowable concrete and tremie
placing are well known to avoid this problem up to certain extent
and it is already a standard procedure to use a cleaning bucket or
an airlift or both to clear the borehole base of any soft or loose
material.
Another solution was found in pressure grouting the base
after the completion of the pile. This improves the condition of
the soil and pre-stresses the pile. Base grouting is widely used in
Thailand to rectify the soft toe problem.

FILL

SOFT
CLAY

1 INTRODUCTION

30 cm

Pile
tip

Concrete
coring

Manchettes

0.5-1.0m

SPT

2-3m

Position of PVC-casing
with test pile no.
Washboring & soil
sampling location

Figure 1. Schematic diagram of the positions of PVC casings in test piles

with typical soil profile at the site.

Pile base grouting was started in Bangkok in 1985, during the

construction of Rama IX cable stayed bridge (Morrison, 1987).
Here 172 bored piles, 2.0 m dia. and 35 m long seated in the 1st
sand were base grouted. Tube-a-manchette technique was used
with grout volumes up to 4000 liters and maximum pressures of
60 bar. With the passage of time engineers started applying this
technique for the foundations of expressways, high-rise
buildings and other heavy structures. In the absence of standard
guidelines, the criteria and the controlling parameters changed
with time. In current practice, pressures ranging from 20 to 60
bar and grout volumes 500 to 1000 liters are normally applied
with Injection Rate (IR) varying from 5 to 36 liters/min.

(% of Pile Dia.)
NGB
0.49 0.51
0.37
1.04

GB
1.34-1.35
0.74
1.59-1.86

3 GROUTING PROCEDURE
Two grouting circuits in the form of U-loops with manchettes
placed at a level of 5 to 10 cm above the pile tip arranged in a

53

269

Portland cement Type I grout with w/c ratio equal to 0.55 was
used. Volume of grout was measured from the number of pump
strokes and also crosschecked against the volume consumed
from the agitator tank. Nominal grout volume used for each pile
was 540 liters but a slight difference shown in Table 2 in actual
volume of grout measured might be due to the difference in IR
used during grouting and at the time of calibration.
All the three piles were grouted with different IR varying
from 3 to 36 liters/min. In each case, after injecting target grout
volume, grout circuit valve was closed to lock the maximum
pressure achieved. Residual pressures two minutes after closing
the pump were also measured and are given in Table 2.
After 15 to 20 days of base grouting, diamond bit rotary
drilling with double tube sampler was used to core through
concrete present below the base of the PVC-casing. Right after
reaching the pile tip, Standard Penetration Test (SPT) was
started. Steel liners were used inside the split spoon sampler to
collect soil samples. All the concrete and soil samples retrieved,
carefully examined on site to determine the grout content
present. Additionally, in order to determine the lateral spreading
of grout into the surrounding soil of pile tip, additional boreholes
were drilled 30 cm away from the pile periphery. Washboring
was used down to approximately 1 m above the pile tip level and
then SPT was employed to collect the samples.

A: Base of PVC-casing, B: Manchettes

Figure 2. PVC casing attached to the rebar-cage, before lowering in to

the borehole.

symmetrical fashion are used (Figure 1). Normal practice is to

use PE tubes above the manchettes running along the rebar cage
to the top, but if specification require cross-hole sonic logging,
same sonic logging tubes are used to connect the manchettes.
Normal practice is to start grouting 24 hours after concreting
when high pressure criteria is not to be strictly followed. This is
especially suitable for congested piling plans where subsequent
maneuvering of equipment can damage the circuit tubes. If high
pressure is to be maintained and/or sonic logging is required,
grouting is started after seven to ten days. First the pile toe is
cracked with high-pressure water flow to flush and open the
grouting circuit to make way for forth coming grout and then
cement grouting is started. Grouting is stopped when the target
grout volume or maximum required pressure is achieved.

5 CONCRETE CORES RECOVERED

From the concrete cores recovered no trace of grout was found
except in case of test pile TP-1 where a 6.5 cm thick grout layer
was recovered right below the pile tip as shown in Figure 3.

4 TESTING METHODOLOGY

In order to investigate the mechanism of grout spreading beneath

and in the surrounding soil of pile tip and the effect of different
grouting parameters, an attempt was made to collect the soil
samples from below the tip of the base grouted piles (Anwar,
1997). PVC-casings were placed in the prototype bored piles as
shown in Figure 1 and 2, to allow for the passage of drilling
tools down to the pile base. Three piles, 1.0 m in diameter and
length in the range 50 to 55 m were constructed with the
provision of 7.5 cm diameter PVC-casings at a site in Bangkok
metropolitan area. Unfortunately, only one pile out of this group
could be drilled because of the failure to lower the drilling tools
inside the PVC-casings. For the second time another group of
three piles having 1.5 m diameter and length 56 - 57 m were
constructed. This time diameter of PVC-casing was increased to
10 cm. Positions of the PVC-casings were staggered with
respect to the manchettes as depicted in Figure 1. All the three
piles were base grouted following the parameters shown in Table
2.
Table 2. Grouting parameters of the test piles.
Test
Pile

Grout
Volume
(Liters)

Injection
Rate
(Liters/min)

Maximum
Pressure
Achieved
(bar)

Residual
Pressure 2 min
after closing the
pump
(bar)

TP-1
TP-2
TP-3

500
545
520

32 3*
36
8

26 16*
36
24

14
3
8

A: Grout core, B: PVC-casing bottom cap.

Figure 3. Grout layer of 6.5 cm thickness recovered from TP-1.

Below the grout core recovered no traces of grout were found in

the sand samples recovered by SPT. Similarly for the rest of two
test piles no traces of grout were seen in the soil samples
recovered through SPT.
6 DEVELOPMENT OF GROUTING PRESSURE
Development of grout pressure with the increase in grout volume
is shown in Figure 4. TP-1 was grouted with a high IR of 32
liters/min for the first 140 liters and then IR reduced to 3
liters/min for the rest of the grouting operation. It must be noted
that as the IR is reduced, pressure dropped from 26 to 6 bar and
then increased gradually to 16 bar at the end of the grouting.
Two minutes after closing the pump, pressure dropped to 14
bars. TP-2 was grouted at a constant and very high IR of 36
liters/min, which corresponds to the maximum available speed of
the grout pump. It must also be noted that pressure development

* in two phases, refer Figure 4 for clarification.

54
270

in this case was very quick as it jumped to 15 bars within first

few strokes of grout and then rose to 36 bar at about 200 liters.
No further increase in pressure was observed till the target
volume of grout had been pumped. Pressure abruptly dropped to
about 3 bar two minutes after closing the pump. TP-3 was
grouted with a constant IR of 8 liters/min. Development of
pressure in this case was quite gradual and it approached
maximum pressure of 24 bar and dropped to 8 bar 2 min after
closing the pump.

SAND

SILT

Type III

CLAY

100
90

MC-300
MC-500

80

Percent Finer (WT)

70
Portland Type I

60
50
40

D15 - Range of
Sand Samples

30
20

Injection Rate
TP-1
TP-2
TP-3

10

100

200

300

400

500

20
10

600

0.1

Grain Size ( mm)

0.01

0.001

Figure 5. Grain size distribution curves of samples collected from

Second Sand Layer around the pile tip.

In order to assess the possibility of permeation grouting of the

2nd sand layer present around the pile tip, Groutability Ratio
(GR) of the sand layers was determined. GR is a useful
parameter for checking the applicability of grout for use in sand
and is given by the formula as follows (Mitchell, 1970):

Volume of Grout (Liters)

Figure 4. Curves showing gradual increase in grout pressure.

GR = D15/D95

(1.1)

Where D15 = The particle diameter of the soil to be grouted 15%

of which is finer by weight; and D95 = the particle diameter of
the grout 95% of which is finer by weight.
Weaver (1991) summarizes the possibility of grouting a soil
for GR ranges as:
GR > 24, usually;
GR < 19, not likely; and
GR < 11, not possible.
GR range for the 2nd sand layer is found to be 1.2 to 4 for the
Portland cement type I which means that its not possible for the
cement grout to permeate into the sand layers. GR using
microfine cements MC-300 (Henn, 1996) shown in Figure 5
comes out to be 12 and 30, so the silty sand layer encountered
can be permeation grouted using microfine cements.

7 GROUT INJECTION RATE

7.1 High Injection Rate

It has been observed that IR is a key factor in the process of
base grouting. If IR is kept very high, grout pressure rises
abruptly but can not be smoothly maintained. A clear fluctuation
in pressure is observed during grouting with high IR. Although
the apparent maximum pressure indicated by the dial gauge is
high at the initial stages but can not be increased further. It
seems that at high IR, hydraulic fracturing of the surrounding
soil took place which is also evidenced by sharp fluctuations in
pressure during grouting Similar phenomenon have been
observed by Yeats, 1989. It must also be noted that maximum
pressure achieved in this case dissipates very quickly as the
grouting operation is terminated. This means that pressure
developed in this case mainly consists of the circuit resistance
and is not the true pressure induced at the pile/soil interface.
Since pressure ceases to increase beyond 400 liters it seems that
no further increase in pressure is possible even if large volumes
of grout have been injected.

9 INSTALLATION EFFECTS
SPT-N values obtained right below the pile tips and 30 cm away
from the pile are compared with the values from nearest
borehole (10 20 m) performed at the time of soil investigation
in Figure 6. It is quite clear that the density of the sand
surrounding the tip have been increased appreciably, especially,
below the tip of the piles.

7.2 Low Injection Rates

SPT-N (Blows/30cm)
25

If IR is kept low, development in pressure took place quite

gradually and the pressure fluctuation is minimal. It must also be
noted that the pressures developement by keeping IR low is quite
steady and can also be maintained for certain duration of time, as
shown in Figure 4 for TP-1 where a residual pressure of 14 bars
has been observed two minutes after closing the pump with a
total drop of only 2 bar. It must also be noted that pressure keeps
on increasing gradually and higher pressures than that of TP-2
can be achieved with more grout volumes.

50

52

75

100 25

50

TP-1

75

100 25

TP-2

50

75

100

TP-3

53

56

8 PENETRATION OF GROUT INTO THE SAND.

Pile Tip

55
Pile Tip

Depth (m)

54
Pile Tip

Pressure

20

30

Injection Rate (Liters/min)

Pressure (bar)

30

10

40
Residual pressure after 2 min.

40

57

From the samples collected, no traces of grout have been

observed in the sand present below or surrounding the pile tip.
Grain size distribution curves of sand samples collected are
shown in Figure 5. It can be seen that the second sand layer falls
in the range of medium to fine sand range. Grain size
distribution for different types of available cements is also
shown in Figure 5.

58
Under the pile tip

30cm away from pile

Nearest borehole

Figure 6. Comparison of SPT-N values around the grouted pile tip and
the nearest borehole from soil investigation.

55

271

10 MANCHETTE BREAKING MODELS

Thickness of concrete cover present below the manchette is very
important and determines the flow path of the grout injected. If
concrete cover present below the manchette is thicker than the
side cover to the outlet points on manchette, it can be suspected
that grout directly percolates toward the pile shaft by bursting
the side cover. In order to monitor the bursting mechanism of
manchettes, 9 model piles 1.2 m diameter and 2.0 m deep were
constructed by varying manchette levels above the pile tip
(Thasnanipan et al, 1998). Model piles were water burst in a
similar way comparable to the normal bored piles. Colored
grout was injected to differentiate it from normal pile concrete.
Suitable time was given to allow for the grout setting before
these piles were exhumed to determine the flow pattern of the
injected grout. Some of the unearthed model piles are shown in
Figure 7 and 8.

A: colored grout zone

Figure 8. Model pile showing the grouted zone below the tip.

3.

4.

Figure 7. Typical model pile used to determine manchette breaking

mechanism.

Low IR of the order of 35 liters/min are recommended for

base grouted piles in silty sand layers of Bangkok.
Significant improvement in the density of sand around the
pile shaft near the base and especially, below the grouted
pile tip has been observed.
Although this is also
contributed by increased overburden pressure due to
concrete, base grouting helps to improve the soil around the
pile tip, disturbed during the drilling process.
Position of manchettes above the pile base is very important
to effectively grout the base of the pile. Care must be
exercised to place the manchettes as close to the pile tip as
possible but if due to some unforeseen reasons manchettes
are lifted up grout can be conjectured to percolate towards
the pile base if the concrete cover thickness below the
manchettes does not exceed 30 cm.

REFERENCES
From the analysis of the model piles, it was revealed that: a) full
pile base area coverage by grout can be achieved if manchette
level is less than 20 cm above the pile tip; b) a partial base area
coverage was found up to 30 cm and if; c) manchette level is
more than 40 cm above the tip, grout directly percolates towards
the pile shaft by bursting the side concrete cover. Presence of
grout on the vertical pile/soil interface at the shaft was also
confirmed in all cases. Although a direct comparison of grout
flow in model piles and prototype piles can not be made due to
the difference in overburden pressures but the upward flow of
grout along the pile shaft can be the most likely phenomenon in
base grouting. Which, subsequently act as a rock socket and
significantly increases the pile stiffness (Francescon, 1992).

Anwar, M.A. 1997, base grouting to improve the performance of wet

process bored piles in Bangkok sub-soils, M. Engg. Thesis, AIT,
Bangkok, Thailand.
Francescon, M. & Solera, S.A. 1992, Base grouting to improve
performance of piles bored under bentonite in the Thanet Sands, Proc.
of the Conf. on Grouting in the ground, ICE, London.
Henn, R.W. 1996, Practical guide to grouting of underground structures,
ASCE publication, NY.
Mitchell, J.K. 1970, In-place treatment of foundation soils, Journal of the
Soil Mechanics and Foundation Division, ASCE, 96(1), 73-110.
Morrison, I.M. 1987, Bored piled foundation for Chao Phya river
th
crossing at Wat Sai, Bangkok, 9 SEAGS Conf. Bangkok, pp. 6-207
to 6-218.
Teparaksa, W. 1991, Performance of base-grouted bored piles in
Bangkok subsoils, First Young Asian Geotechnical Engineers
Conference, AIT, Bangkok, Thailand, pp.117-128.
Teparaksa, W. 1992, Behavior of base-grouted bored piles in Bangkok
subsoils, Piling Conference, European Practice and Worldwide
Trends, Institution of Civil Engineers, London, pp.296-301.
Teparaksa, W. 1994, Newly developed toe-grouted bored piles in soft
Bangkok clay: performance and behavior, Proc. of the Intl. Conf. on
Design and Construction of Deep Foundation, FHWA, Orlando,
USA, pp.1337-1351.
Thasnanipan, N. et al, 1998, Design, Construction and behavior of bored
th
cast in-situ concrete piles in Bangkok subsoil, Proc. 4 Intl. Conf. on
Case Histories in Geotechnical Engineering, St. Louis, Missouri,
USA, paper no.1.20.
Thasnanipan, N. & Umesh, S. 1998, Pile base cracking models with
different manchette levels from the pile tip, SEAFCO internal report.
Weaver, K.D. 1991, Dam Foundation Grouting, American Society of
Civil Engineers, NY.
Yeats, J.A. & ORiordan, N.J. 1989, The design and construction of large
diameter base grouted piles in Thanet Sand at Blackwall Yard,
London, Proc. Intl. Conf. on Piling and Deep Foundations, London,
Vol. 1, pp. 455-461, Balkema.

11 CONCLUSIONS
From the study carried out following conclusions can be drawn.
1.

2.

Most of the grout injected during base grouting rises up

along the pile/soil interface, and does not permeate in to the
surrounding sand layer to any degree. So the improvement
made by base grouting is mainly contributed by increase in
skin friction in the form of rock-socket effect which
increases the pile stiffness significantly. Since the increase
in skin friction depends only upon the volume of grout
injected, pressures achieved during grouting are of
negligible importance in this regard. Partial replacement
and/or precompression of the sediments below the pile tip is
possible by following low IR hence, smoothly increasing
and maintaining the pressures achieved, if it is desired so.
IR is the key factor in the base grouting process, especially
if, higher pressures are to be achieved and maintained.

56
272

Under the Royal Patronage of His Majesty King Bhumibol Adulyadej

Sixth Cycle Celebration of

His Majesty the King of Thailand and
th
40 Anniversary of the Asian Institute of Technology
Civil and Environmental Engineering Conference
New Frontiers and Challenges
8-12 November 1999
Bangkok, Thailand

Conference Proceedings
Volume 2 (Part II)

Editors:
D. G. Lin
D. T. Bergado
N. Phien-wej
P. Nutalaya
A. S. Balasubramaniam

Review of the shaft capacity degradation of bored piles

constructed with bentonite slurry
N. Thasnanipan, M.A. Anwar and Maung A.W.
Seafco Company Limited Bangkok, Thailand

57

Civil and Environmental Engineering Conference

New Frontiers and Challenges, 8-12 November 1999, Bangkok, Thailand

REVIEW OF THE SHAFT CAPACITY DEGRADATION OF BORED PILES

CONSTRUCTED WITH BENTONITE SLURRY
Narong Thasnanipan, Muhammad Ashfaq Anwar, Aung Win Maung
SEAFCO Co., Ltd., Bangkok, Thailand.
ABSTRACT
Over the last few decades, different researchers have studied the influence of different
construction parameters on the shaft load transfer of bored piles constructed with slurries.
Still the issue is not fully resolved, different parties involved in the bored piling industry
believe on the degree to which these parameters effect the perimeter load transfer. This paper
attempts to summarize all the significant works previously done on this subject. It is
concluded that the realistic formulation of specifications and selection of design parameters
with the consideration of practical constraints at site is very important. Performance of the
bored piles is effected by installation procedures, all the ill effects attributed to the
construction parameters can be eliminated by following good engineering practices.
INTRODUCTION
It is commonly accepted that the construction parameters like slurry properties,
construction time, equipment used, etc., effect the performance of wet process bored cast insitu piles. But the extent to which these parameters effect the pile capacity is not clearly
understood due to the wide variation in the soil conditions and other influencing factors which
are simultaneously acting. It has also been recommended in some codes of practice that
observing good engineering practices can eliminate all the foreseeable ill effects attributed to
these parameters.
CONSTRUCTION PARAMETERS
Among the other various parameters, construction time and slurry properties are
particularly considered responsible for the degradation of the pile shaft capacity. In fact the
total construction time is not directly related to the pile performance. Critical time, which
might contribute to the degradation of the pile, starts after the borehole excavation is
completed till the completion of the concrete pouring. Time consumed in the excavation
operation is not generally considered to contribute any degradation since slurry is
continuously kept agitated during drilling operation. A maximum elapsed time of 24 hours
after the completion of the borehole excavation is generally recommended under normal
conditions. Slurry properties especially the viscosity is believed to be another important
factor related to the quality of the constructed pile, because it determines the degree of
replacement by concrete during tremie concreting.
Previous published researches on this issue along with the brief conclusions made are
given in the following in a chronological order.
Chaidesson (1961) observed a better load-settlement behavior for one pile drilled with
slurry than the other pile drilled by a casing method.
V ~ 59

59

Corbett et al. (1970) presented the results of load testing of test panel excavated under
bentonite slurry, excavation took 2 days and on completion the panel was left open overnight,
but bentonite was renewed prior to concrete. The authors inferred that the presence of
bentonite suspension even for such a long time was not adversely affected the development of
skin friction.
Farmer, Buckley & Sliwinski (1970) conducted the laboratory tests on model concrete
pile. The results indicated that presence of a bentonite filter cake at the concrete/sand
interface has little effect on load transfer. Full scale tests on three deep bored piles
constructed under bentonite, indicated that the load transfer was extremely higher than the
expected. They concluded that high load transfer level may be obtained from piles
constructed under bentonite, this may be due to uneven side-wall configuration.
Reese & Tauma (1972) found that the values of load transfer developed in clayey soils
on the sides of the shaft constructed by the slurry displacement method are comparable to
those in the shafts constructed in the dry.
Wates & Knight (1975) concluded from their investigation that a bentonite filtercake of
a significant thickness would develop in 24 hours. In order to prevent a build up of filtercake,
the slurry should be left in place much less than 24 hours.
Cernak (1976) performed full scale load test on three barrettes in sandy gravels. Two of
the barrettes were constructed using slurry with different exposure times (8 and 97 hours) and
their load test results indicated a decrease of skin friction capacity of 43% and 56%
respectively, as compared to the other barrette excavated dry and concreted immediately. It
must be noted that the major part of the reduction i.e. 43% took place in the first 8 hours of
construction time while only 13% (56-43) reduction took place in the rest of 89 hours (97-8)
for the second barrette. These findings support the conclusion of Wates & Knight (1975).
Sliwinski (1977) reported that displacement of suspension of slurry from the sides of the
bored hole does not constitute a major problem. The rising column of concrete will displace
the fluid suspension because of the considerable difference in unit weight and shear strength
of the materials. The author stated that the field and laboratory tests seem to indicate that the
influence of some bentonite in the parent soil has an insignificant influence on load support
providing that the properties of the slurry are within reasonable limit and the concreting is
done within a reasonable short time after the excavation is completed.
Fleming & Sliwinski (1977) reported on 49 filed tests from several countries. The test
results suggest that the use of bentonite suspension has no detrimental effect on shaft friction
while in granular soil, there is an indication that skin friction at high displacement may be
slightly be reduced, though there may be reasons unrelated to the bentonite process. They
suggested that concreting should be completed within 24 hours of the completion of boring.
Reese & Tucker (1985) concluded that the capacity of bored piles constructed under
slurry could be substantial with proper construction technique. The slurry will be ejected from
the excavation and substantial bond will develop at the interface of the concrete and the
supporting soil. However, the authors also suggested that the slurry must not be left more
than few hours in the excavation without operation of drilling tool; otherwise, a thick filter
cake will develop on the side of the excavation.
Reese & O Neil (1988) suggested that a solution to the problem of reduced skin
friction due to excessive filter cake is to maintain the properties of the slurry within tolerable
limits and to place the concrete the same day that the excavation is completed. If for some
reason, it is impossible to place the concrete without undue delay, the drilling machine must
be re-occupied the excavation to rework the borehole. If the slurry remains for a period of
time without agitation, the filter cake can become thick.
V ~ 60

60

Fleming et al. (1992) concluded that the process of forming a pile under bentonite
suspension does not materially reduce the shaft friction in both granular soil and cohesive
soil. The rising column of concrete with a slump in excess of 175 mm from tremie process
will largely remove the filter cake layer on the wall surface of shaft. However, if the hole is
left filled with slurry for long period (no agitation action), this will give rise to a significant
reduction in shaft friction on completed piles.
ONeil & Reese (1992) reported the effect of slurry properties on two drilled shafts
constructed in similar ground conditions (alternating layer of very stiff clay and dense sand).
Shafts 1 and 2 were constructed with Marsh cone viscosity of 37 and 49 sec respectively.
Time between opening the borehole and completion of concreting for both piles was 5 to 7
hours. Minor difference in load transfer of both shafts was found and the authors concluded
that the difference in slurry properties used for the construction for both shafts had a little
effect on shaft load transfer.
Wardle et al. (1992) concluded from four piles tested in London that the axial capacity
of driven and jacked piles after one or two months were 14 to 28% greater than those
measured during or immediately following installation. CRP tests performed after about 3
years recorded further increase of between 14 to 20% (total became 28 to 48%). Although, it
was not possible to record the capacity of the bored pile immediately after installation, the
capacity of this pile increases by 47% between 2 months and three years.
Majano & ONeil (1993) attempted to model the formation of filter cake in the
laboratory with different slurry dosages, differential pressures and exposure times. They
stated that perimeter load transfer is a complex function of the physical and chemical
characteristics of the slurry and the geomaterials, the roughness of the borehole, the fluid
pressures exerted by concrete, the shearing properties of the soil, and possibly the chemistry
of the fluid concrete. They attempted to correlate the potential degradation of the soil-pile
interface to the thickness and shear strength of the cake formed against the walls, and argued
that the shear strength and thickness can not be measured with only one or two parameters
like slurry dosage and differential pressure. With a bentonite concentration of 72 kg./m3 and
differential pressure of 0.5 psi, they achieved a filter cake thickness of 3.14 and 4.5 mm after
a contact time of 4 and 24 hours. Authors also compared the laboratory tests with the actual
field conditions and proved that due to the presence of enormously high differential pressures
in the field in some cases, thickness of filter cakes which would be considered detrimental for
shaft load transfer, require few hours to develop (Authors proved that a filter cake of 6 mm.
thickness would require only 2.3 hours to develop in the field under a differential pressure of
10 psi).
Hosoi et al. (1994) reported that when the diaphragm wall foundation is cast in slurry,
mud cake could not be fully scoured by concrete. So the evaluation of the shaft friction
mobilized on the wall foundation cannot overlook the effect of mud cake. They concluded
from laboratory tests that the friction resistance between concrete and the sample soil
(decomposed granite) through bentonite slurry cake is smaller than the shear strength of the
sample soil.
Ho & Lim (1998) reported that the barrette sized 2.8 m x 0.60 m x 47.4 m excavated
under bentonite slurry support at Singapore Post Center project with a total construction time
of about 110 hours has been tested. The test result showed that substantial shaft friction could
be mobilized to carry load satisfactory with a reasonable factor of safety and within a
tolerable displacement of 12 mm at working load.
Littlechild & Plumbridge (1998) presented that the values of actual shaft resistances
mobilized for in the five pile load test carried out at BERTS project and compared to the
calculated shaft resistances, where calculated shaft resistance is based on the design approach
adopted on the project. For cohesive soil fs = Cu, taken as 0.5 for stiff and hard clay

61
V ~ 61

layers and for sand layers shaft resistance of 120 KPa has been determined for all sand layers.
Construction time and slurry viscosity on this project was controlled to be less tan 24 hours
and 35 sec respectively. They concluded that shaft friction capacity of bored piles constructed
under bentonite slurry tends to decrease as the construction time and slurry viscosity increases.
Thasnanipan, Baskaran & Anwar (1998) presented the shaft friction values mobilized
from eleven instrumented bored piles constructed under slurry with different construction
times and bentonite viscosity from various sites in Bangkok area. They concluded that the
slurry viscosity up to 55 sec./Qt do not have significant effect on the shaft load and
degradation of shaft capacity with increase in construction time beyond 24 hours is not
excessive.
DESIGN CONSIDERATIONS AND PARAMETERS
Selection of design parameters, which suits the local construction practices and
procedures, is very important. Design parameters are normally selected based on the previous
researches of local institutions/organizations along with some special considerations specific
to the sites. Following are the recommendations and suggestions that need to be observed for
the estimation of shaft load capacities in Bangkok subsoils.
Sliwinski and Fleeming (1974) suggested that in clay soil, the bentonite has no effect on
friction and no reduction of the normal factors for deriving adhesion is justified. There may
indeed be a slight increase of adhesion. In granular soils, the bentonite has some effect. The
reduction of friction calculated from effective lateral pressure and coefficient of friction
should be considered in the region 10-30%.
Oonchittikul (1990) back analyzed a parameter = (Ks.tan ) from instrumented bored
pile load tests conducted with bentonite slurry in Bangkok subsoils by using the effective
stresses with both the hydrostatic pore pressure and the declined pore pressure. The resulted
value in the first and second sand layers from effective overburden pressure based on
hydrostatic pore pressure falls in the range of 0.4 to 0.65 and 0.2 to 0.4 respectively while
value in the first and second sand layers from effective overburden pressure based on declined
pore pressure falls in the range of 0.1 to 0.48 and 0.1 to 0.2 respectively. In both cases
value in second sand layer is less than the value in the first sand layer. The author inferred
that the overburden pressure might not fully develop at great depth in sand layer as it has been
assumed in the calculations.
Hooley & Brooks (1992) have suggested that if the traditional pile design approach is
used, the proper interpretation must be given to the accuracy of the data employed by soil
investigation as the potential error in the calculation of ultimate base resistance increases
steadily with depth and shaft resistance error increases gradually with depth. The authors
reported that it is noticeable that the amount of scatter in the test result from soil investigation
samples increases with depth and thus the average line is more difficult to define at greater
depths. If the scatter is analyzed statistically, upper and lower bound of average line can be
identified on the basis of chance of error 0.01% probability.
Eide & Bellis (1992) extracted data from 13 numbers of loading tests taken to failure
from the test results of over 109 tests in Bangkok which are presented by Oonchittikul (1990).
The authors found average shaft friction in stiff clay and dense sand in the range 78 to 130
KPa and adopted 100 KPa as design value, which is commonly used for design in Bangkok
Subsoil.
Teparaksa (1992) reported the problem from bentonite sedimentation causes a soft base.
Unexpectedly, low skin friction can be mobilized at failure (design at yield point) in soft base
condition. The author also reported that the static load test results on 1.50 m dia. pile at Pok

62

V ~ 62

Klao bridge project in Bangkok indicates an excessive settlement at 4000 KN load which is
caused by deposition of bentonite at the base of the pile. Analysis of load transfer indicates
that the skin friction of first stiff clay, first sand, and second stiff clay are equal and about 40
KPa at 4000 KN load. After the base grouting with cement-mortar, the analysis shows the
skin friction of this soil is between 40 to 90 KPa at about 5 mm. movement. They are in the
expected range for each soil layer, which is not affected by grouting.
Meiner et al. (1993) studied the influence on bearing behavior of bored piles when the
pile base is approaching a layer of soft soil layer and concluded that a significant punching
mechanism is obtained for value of t/d =2, where t is the distant from pile base and the clay
layer and, d is the pile diameter. This punching mechanism would decrease significantly both
shaft and base resistance.
Tomlinson (1996) suggested that if pile is excavated by grab under water, there is
considerable loosening of the soil. This causes a marked reduction in both end bearing and
skin friction. Then the design calculation will be calculated on the basis of a low relative
density of the granular soil.
Littlechild et al. (1998) reported that the shaft resistance of both the plain and shaft
grouted bored piles in sand appears to be mostly independent of the vertical effective stress
over the stress range encountered. From the test result of plain bored piles constructed in
accordance with a given specification of BERTS project in Bangkok i.e. bentonite viscosity
<35 sec and construction time of less than 24 hours, the shaft resistances fall in the range 65
to 170 KPa. In addition, they have determined a design line for shaft resistance of plain bored
piles in Bangkok sand of 120 KPa for all ranges of vertical effective stresses.
Thasnanipan et al. (1998) revealed that the end bearing capacity of bored pile where toe
is embedded in to the clay layer would be reduced and this end bearing capacity reduction
might influence the shaft resistance capacity too. This conclusion was reinforced by the pilot
pile load test result in the project located on Rama III road in Bangkok having diameter 1500
mm, toe founded at 60m in clay layer, failed at only about 17500 KN with total settlement of
over 90 mm. While the designed maximum test load was 25000 KN compared to the pile load
test from contract pile of same diameter toe depth founded in sand layer at depth 55 m and
base grouted. Result of test up to 2500 tons with total settlement of only 29.26 mm and elastic
recovery was 17.52 mm recorded.
CODES AND SPECIFICATIONS
Specifications are prepared for the good of the project. Specs drafted by ignoring the
practical site constraints and specific project needs can cause a job to become problem for all
parties. For example, some specs have been found to stressing on maximum allowable
construction time of 24 hours, which sometimes become impossible to adhere to because of
the prevailing traffic congestion in the city centers and local authority regulations in Bangkok,
and ultimately instigate the contractor to breach the law in order to comply with such
unpractical specs. A summary of the slurry specifications used worldwide is given in Table 1.
In the following are given some suggestions related to the excavation under bentonite slurry.
Fleming et al. (1975) suggested that the specifications should set out a cause of action to
perform the work, which is reasonable from the contractors point of view and gives a
reasonable expectation of achieving the desired end product. A specification loses
creditability on site if its requirements are outside the range of practical achievement and such
a specification can only be detrimental to the whole job. They further suggested that obscure
requirements will naturally be the first to be forgotten. Thus, for example, it is better to say
after mixing concrete shall be placed before it lose the specific workability than to say
concrete shall be placed within X minute (s) from mixing.
V 63
~ 63

Lttlechild & Plumbridge (1998) reported that the majority of the piling contractors state
that a maximum construction time for bored pile constructed to depth 50m under bentonite
slurry for BERTS project in Bangkok of less than 24 hours per pile would not be realistic. A
maximum construction time of 24 hours and slurry viscosity of less than 35 seconds (use of
Bentosund bentonite only) were therefore adopted for the project.
Chodorowski & Duffy (1998) presented the typical control and progress charts for 24
hours limit on construction time of bored piles at BERTS project, Bangkok. Charts prepared
by the authors show that during the 24 month period of bored pile construction, there was
noticeable percentage of pile constructed longer than 24 hours of approximately 15% which
in some months as much as 30%. They reported that longer construction time of more than
24 hours per pile was due to difficulties in securing continuous concrete supply as a result of
traffic congestion in the area of different piling sites.
Table 1: Summary of recommended values of slurry properties used for bored piles.

Authors
Parameter
Slurry Type
Density
(gm/
cc)

pH
Sand Content
Marsh Funnel
Viscosity
(sec/
Qt)

Plastic (cP
Viscosity

)
Yield(Pa
Point
)
10 min. Gel Strength
(Pa
)
Differential Head
(m
)
Fluid Loss
(30 minute test)
* Maximum Contact
Time
(hr.)

FP
S
1975,
1977

FDOT
Hutchin Hodgso
Holden
gso
,
n
1987,
n
1984
1977
et al.,
1988
1975

Bentonite Calcium
Bentonite

<1.10

1.024 to
1.218

9.5 to 12

<11.7

<6%
(by weight)

Bentonite

1.024 to
1.135
10.8 to 11.7

<14%
<35%
by volume
(by weight) (5% max.
by weight)

Bentonite
Sodium
or
Bentonite
Attapulgite
1.03 to
1.20
8 to 11
10% max.
(by volume)

Rees
e&

O'Neil
1988

Bentonite
or
Polymer

Bentonite
or
Bentonite
Attapulgite
<1.10

<1.36
Friction
Shaft

1.03 to 1.10
1.03 to 1.20

1.02 to 1.07
1.02 to 1.13

<1.10
<1.15

8 to 11

9.5 to 12

8 to 12

8 to 11
8 to 11

8 to 10
8 to 10

9.5 to 10.8
9.5 to 11.7

(by volume)

<25%
(by volume
Friction
Shaft)

<4%
<4%
(by volume)

1.03 to 1.20

1.03 to 1.20

8 to 12
<4%

28 to 40

28 to 45

<20

<20

3 to 20

<20

<20

<20

4.2 to 41.8

2 to 10

1.9 to 10

4 to 40

1 to 1.5

>1

Bentonite

1989

30 to 40

10 to 40

Bentonite

1988

3.6 to 20

Bentonite
or
Attapulgite

AASHTO

4 to 40

ICE
FPS
ARUP
1996

ACI

30 to 90

1992

Majano
&
O'
Neil
1993

ICE

30 to 90

26 to 50

28 to 45
28 to 45

<4 %
<10%

<2%
<2%

32 to 60
32 to 60

30 to 70
< 90

6 to 8.5
6 to 10.0

2 to 6.0
2 to 6.5

4 to 40
4 to 40

1.5

< 40
<60

Note: For two rows in each cell (last three cloumns)

First row: fresh or recycled as supplied to borehole
Second row: prior to concrete pouring

* Without agitation and side-wall cleaning

ADSC and DFI drilled shaft inspectors manual clause 4.5.1, advised that where
bentonite slurry remains in the shaft unagitated for 3 or 4 hours, or more than 24 hours even
though agitated, a mud cake may develop on the shaft walls which can reduce friction.
Unless, the designer has accounted for caking and appropriately reduced design friction
value or the specification gives other limitation, portion of the shaft or socket designed for
side friction should either be excavated and concreted within a 24 hours period (and no more
than 3 hours when the slurry is not agitated) or re-reamed or re-scraped to remove the mud

64
V ~ 64

cake prior to concreting, or some method which is acceptable to the geotechnical engineer and
the inspector used to demonstrate that detrimental mud caking has not occurred.
ADSC Specification Clause 2.3.5.2 (g) specified that the drilled pier constructed under
bentonite slurry should be concreted and completed the same day that of excavation. If this is
not possible, the excavation shall be re-drilled, cleaned and slurry tested before concreting.
Shaft friction could be mobilized to carry load satisfactorily with a reasonable factor of safety
and within a tolerable displacement of 12 mm at working load.
FHWA guide for drilled shaft specification, stipulates that, that drilling slurry is an
effective method of stabilizing drilled shaft excavation until either a casing has been installed
or concrete has been placed. Primary concerns connected to slurry use are: the shape of the
borehole would be maintained during the excavation and concrete placement; the slurry does
not weaken the bond between the concrete and both the natural soil and rebar; all of the slurry
is displaced from the borehole by the rising column of fresh concrete; and any sediment
carried by the slurry is not deposited in the borehole. The engineers concerns regarding the
behavior and effectiveness of slurry project, can be satisfied by appropriate specification
requirements. These requirements include: specifying a suitable range of slurry properties
both prior to and during excavation and prior to concreting; performing slurry inspection test;
and construction of pre-production trial shaft by the slurry method.
FULL SCALE TEST TO ASSESS THE EFFECT OF CONSTRUCTION PARAMETERS
Since the degree of effect of different parameters greatly depends upon the soil
properties encountered, conclusions made for one type of soil are not directly applicable to
other locations with different geology. Static, instrumented load test program on bored and
barrette pile conducted at BECM tower project Rama IX road, Bangkok, provided the unique
chance to determine the effects of construction parameters on the shaft capacity (Thasnanipan
et al., 1999). Bored pile 1.5m diameter and a barrette 1.5x3.0m in cross-section had same
lengths of 57.5m founded in the second sand layer and located 30m apart having similar soil
embedment conditions. Properties of the bentonite slurry used are given in Table 2.
Table 2. Comparison of bentonite slurry properties for test pile and barrette.
Pile
Properties
Viscosity (sec)
Density (g/cc)
pH value
Sand Content (%)

Before feeding
to the borehole

33
1.08
8
0.1

Barrette

After Recycling &

Before concreting
(near borehole base)

36
1.10
8
0.8

Before feeding
to the trench

36
1.10
8
1.0

After Recycling &

Before concreting
(near trench base)

49
1.17
9
1.1

Total construction time for the bored pile and barrette was 27 and 75 hours,
respectively. However, the total elapsed time between the final excavation (re-drilling &
cleaning) and concreting for the pile and barrette was 16.25 and 20 hours, respectively. In
spite of considerable difference in the total construction time and slurry viscosity, values of
unit shaft friction mobilized in different soil layers did not show any difference in the shaft
load transfer characteristics. This proves that the total construction time is not a crucial
parameter, which effects the shaft load capacity if good engineering practices like re-drilling
the borehole is applied.
V ~ 65 65

CONCLUSIONS
Conclusions of various researchers can be summarized as follows:

Bentonite viscosity up to the normally recommended range i.e. 60 sec., given in Table 1
does not have any significant effect on the skin friction capacity of bored piles. In
authors point of view, filter loss is the more relevant property of the slurry, which
directly effects the formation of the filter cake and finally results in the shaft capacity
reduction rather than the viscosity.
Selection of pile design parameters with due consideration of site constraints is equally
important. Adequate strength reduction factors based on the local experiences suggested
by local research institutions need to be considered.
Concreting should be completed within 24 hours after the finishing excavation of
borehole. If due to some unavoidable reasons borehole is left filled with slurry beyond 24
hours with out agitation, re-drilling must be carried out to scarp off any filter-cake present
on the walls of the borehole.
Performance of bored piles, to a large extent, depends upon the installation procedures
and, the skill of the operating crew, if proper construction procedures have been followed

all the negative effects can satisfactorily be eliminated.

It must be emphasized that the specifications for each project must be tailored to suit the
actual site conditions keeping in view the constraints like possible delays in concrete
supply due traffic congestion, local authority regulations etc. Among the available control
limits for bentonite slurry shown in Table 1, this is authors opinion and experience that
AASHTO specs are practically more suitable for pile excavation works in Bangkok
subsoils.

REFERENCES
ADSC and DFI (1989), Drilled Shaft Inspectors Manual prepared by The Joint CaissonDrilled Shaft Committee of the ADSC: The International Association of Foundation
Drilling and DFI: Deep Foundations Institute.
ADSC Standards and Specifications, Foundation Drilling Industry. (1995). Incorporations.
ACI 336.1 94.
American Concrete Institute (1989), Standard Specifications for the Construction of Drilled
Piers, ACI Structural Journal 89, 1, pp. 94-103.
Cernak, B. (1976), The Time Effect of Suspension on the Behavior of Piers, 6th European
Conf. on Soil Mechanics and Foundation Engineer, Vienna.
Chaidesson, R. (1961), Influence of the Boring Method on the Behavior of Cast-In-Place
Bored Piles (in French with English Summary). Proc. 5th Int. Conf. Soil Mech. Paris.
Chodorowsi A.R. & Duffy, M.R. (1998), Quality Assurance and Testing on a contract for
16,000 Large Diameter Bored Cast In-Situ Piles. 7th Int. Conf. and Exhibition on Piling
and Deep Foundations, Vienna, Austria.
Corbett, B.O., Davies, R.V. & Langford, A.D. (1970), A Load Bearing Diaphragm Wall at
Kensington and Chelsea Town Hall, London. Proc. of the Conf. on Diaphragm Wall &
Anchorage organized by ICE, London.
Eide, O. & Bellis, J.A. (1992), Foundation of Bridges in Bangkok. Int. Conf. on Deep
Foundation. London.
V ~ 66

66

Farmer, I.W., Buckley, P.J.C. & Sliwinski, Z. (1970), The Effect of Bentonite on the Skin
Friction of Cast-in-Place Piles. Proc. of the Conference on Behavior of Piles. organized
by ICE London.
Federation of Pilling Specialists (1975), Specifications for Cast-In-Place Piles Formed under
Bentonite Suspension, Ground Engineering, 8, No. 2, 50 pp.
Fleming, W.G.K. & Sliwinski, Z.J. (1977), The Use and Influence of Bentonite in Bored Pile
Construction, CIRIA REPORT PG3.
Fleming, W.G.K., Fuchsberger, M., Kipps, O. & Sliwinski, Z. (1975), Diaphragm Wall
Specification, Proc. of the Conference on Diaphragm Wall & Anchorage organized by
ICE, London.
Fleming, W.G.K., Weltman, A.J., Randolph, M.F. & Elson, W.K. (1992), Piling Engineering
(2nd edition). John Willy & Sons.
Florida Department of Transportation (1988), FDOT Standard Specification for Road and
Bridge Construction, Section 455: Foundations of Structures, Tallahassee, FL.
Guide Drilled Shaft Specification, U.S. Department of Transportation, Federal. Highway
Administration. Geo-Technical Engineering Note Book. Geotechnical Guideline no. 14.
Distributed by ADSC, Texas, USA.
Ho, C.E. & Lim, C.H. (1998), Barrettes Designed as Friction Foundations, Proc. of Fourth Int.
Conf. on Case Histories in Geotechnical Engineering. St. Louise, Missouri.
Hodgson, F.T. (1977), Design and Control of Bentonite/Clay Suspensions and Concrete in
Diaphragm Wall Construction, Seminar on A Review of Diaphragm Walls, ICE,
London.
Holden, J.C. (1984), Construction of Bored Piles in Weathered Rocks, Part 4: Bentonite
Construction Procedures, Technical Report No. 69, Road Construction Authority of
Victoria, Australia.
Hooley, P. & Brooks, S.R. (1992), Site Investigation of Bored Piling in London Clay. Piling
European Practice and Worldwide trends. Proc. of a Conf. organized by the ICE,
London.
Hosoi, T., Yaki, N., & Enoki, M. (1994), Consideration to the Skin Friction of Diaphragm
Wall Foundation. 3rd Int. Conf. on Deep Foundation Practice incorporating Piletalk
1994, Singapore.
Hutchinson, M.T., Daw, G.P., Shotton, P.G., & James, A.N. (1974), The Properties of
Bentonite Slurries Used in Diaphragm Walling and Their Control, Proc. of Conf. on
Diaphragm Walls and Anchorages, ICE, London.
Littlechild, B. & Plumbridge G. (1998), Effects of Construction Technique on the Behavior of
Plain Bored Cast in Situ Piles Constructed under Drilling Slurry, 7th International Conf.
and Exhibition on Piling and Deep Foundations Vienna, Austria.
Littlechild, B.D., Plumbridge , G.D. & Free, M.W. (1998), Shaft Grouted pile in sand and
clay in Bangkok .7th Int. Conf. and Exhibition on Piling and Deep Foundations, Vienna,
Austria.
Majano, R.E. & ONeil, M.W. (1993), Effect of Mineral and Polymer Slurry on Parameter
Load Transfer in Drilled Shaft. A Report to ADSC; by University of Houston, USA.

V ~ 67

67

Meibner, H., Shen, Y.L., Van Impe, W.F & Vogt, C. (1993), Punching Effect for Bored Piles.
Proc. of the 2nd Int. Geotechnical Seminar on Deep Foundation on Bored and Auger
Piles; Ghent, Belgium.
ONeil, M.W. & Reese, L.C. (1992), Behavior of Axially Loaded Drilled Shafts in Beaumont
Clay. Report No. 89-8, The University of Texas at Austin. U.S.A.
Oonchittikul, S. (1990), Performance of Bored Piles in Bangkok Subsoils. M. Engg Thesis,
V ~ Bangkok,
67
Asian Institute of Technology (AIT),
Thailand.
Reese, L.C. & Tauma, F.T. (1972), Load Test of Instrumented Drilled Shafts Constructed by
the Slurry Displacement Method. Center for Highway Research. The University of
Texas at Austin. U.S.A.
Reese, L.C. & Tucker, K.L. (1985), Bentonite Slurry in Construction Drill Piers. Proc. of a
Session sponsored by the Geotechnical Engineering Division of the American Society
of Civil Engineers in Conjunction with the ASCE Convention, Denver, Colorado.
Reese, L.C. & ONeil, M.W. (1988), Drilled Shafts: Construction Procedure and Design
Methods prepared for U.S. Department of Transportation, Federal Highway
Administration office of Implementation, Mc.Lean, Virginia. in cooperation with
ADSC: The International of Foundation Drilling. Dallas, Texas.
Sliwinski, Z. & Fleming, W.G.K. (1974), Practical Consideration Affecting the Performance
of Diaphragm Walls, Proc. of the Conf. on Diaphragm Wall and Anchorages, ICE,
London.
Sliwinski, Z.J. (1977), The Effect of the Bentonite Displacement Method of Construction on
the bearing Capacity of piles and Diaphragm Wall Elements. Notes for the meeting of
the concrete Society. Croydon, October 12, 1977, (Unpublished.).
Teparaksa, W. (1992), Behavior of Base Grouted bored piles in Bangkok Subsoils. Piling
European Practice and Worldwide Trends. Proc. of the conf. organized by the ICE,
London.
Thasnanipan, N., Baskaran G. & Anwar, M.A. (1998), Effect of Construction Time and
Bentonite Viscosity on Shaft Capacity of Bored Piles, Proc. of the 3rd Int. Geotechnical
Seminar on Deep Foundations on Bored and Auger Piles, Ghent, Belgium.
Thasnanipan, N., Maung, A.W. & Tanseng, P. (1998), Barrettes Foundation in Bangkok
Subsoils, Construction and Performance. Thirteenth Southeast Asian Geotechnical
Conference, Taipei, Taiwan.
Thasnanipan, N., Teparaksa, W., Maung, A.W. & Baskaran, G. (1998), Design, Construction
and Behavior of Bored Cast In-Situ Concrete Piles in Bangkok Sub Soil. Proc. of
Fourth Int. Conf. on Case Histories in Geotechnic Engineering, St. Louise , Missouri.
Tomlinson, M.J. (1996). Pile Design and Construction Practices. (4th Edition).
Wardle, I.F., Price, G., & Freeman J., (1992), Effect of Time and Maintained Load on the
Ultimate Capacity of Pile in Stiff Clay. Piling European Practice and Worldwide trends.
Proc. of the Conf. Organized by the ICE, London.
Wates, J.A. & Knight, K. (1975), The Effect of Bentonite in the Skin Friction in Cast in-Place
Piles and Diaphragm Walls. Proc. of Sixth Regional Conf. for Africa on Soil Mechanics
and Foundation Engineering, Dorhan, South Africa.

V ~ 68

68

Under the Royal Patronage of His Majesty King Bhumibol Adulyadej

Sixth Cycle Celebration of

His Majesty the King of Thailand and
th
40 Anniversary of the Asian Institute of Technology
Civil and Environmental Engineering Conference
New Frontiers and Challenges
8-12 November 1999
Bangkok, Thailand
Editors:
W. Kanok-Nukulchai
C. Polprasert
P. Nutalaya
A. S. Balasubramanian

Failure mechanism of long bored piles

in layered soils of Bangkok
N. Thasnanipan, M.A. Anwar and Maung A.W.
Seafco Company Limited Bangkok, Thailand

69

Civil and Environmental Engineering Conference

New Frontiers and Challenges, 8-12 November 1999, Bangkok, Thailand

FAILURE MECHANISM OF LONG BORED PILES IN LAYERED SOILS OF

BANGKOK
Narong Thasnanipan, Muhammad Ashfaq Anwar, Aung Win Maung
SEAFCO Co., Ltd., Bangkok, Thailand.
ABSTRACT
Simple superposition principle for the estimation of ultimate shaft friction capacity of
bored piles sometimes found to be not directly applicable for the piles embedded in the
multilayered soils of Bangkok. Because of the early yielding of the stiff clay layers, a
localized plunging of the load-settlement curve is often observed. For these layers, although
the maximum mobilized skin friction corresponds to the adhesion factor equal to around
0.35, suggested by different researchers. But due to the movement softening behavior of
these layers, residual friction values near the ultimate failure of the piles drops to a
corresponding values of the order of 0.15 or even smaller. It is recommended that due to
the progressive nature of failure, lower values for stiff to hard clay layers should be
considered for the estimation of ultimate skin friction capacity under such conditions.
INTRODUCTION
Ultimate shaft friction capacity of piles embedded in the layered soils is calculated as
summation of the ultimate shear resistances offered by different soil horizons present along
the pile shaft. Load movement curves obtained from instrumented pile load tests have
revealed that the very stiff to hard clay layers present within the Bangkok aquifer sands
frequently exhibit a well defined movement softening behavior. Peak friction values in these
layers have been found to be mobilized at very low pile head movements and well before the
ultimate friction values of other layers present along the pile shaft. When rest of the layers
reach their ultimate values these layers have already reached their residual friction values,
which are as low as 50 percent of the peak friction values. So the simple superposition of
ultimate friction values give rise to overestimation of the total shaft friction capacity under
these conditions.
SOIL PROFILE
Typical soil profile of plain of Bangkok is shown in Figure 1. Thickness of Bangkok
Soft Clay (BSC) at the top varies generally between 15 to 18m. BSC changes to medium stiff
consistency before the first stiff clay, which is generally present at 20 to 25 m depth. First
sand layer is usually 5 to 10 m thick and found at 25 m to 30 m depth, below is a series of stiff
to hard clay and medium to very dense silty sand layers.
PILES EMBEDMENT CONDITIONS
Pile embedment conditions are designed depending upon the actual soil profiles
encountered at the sites, as depicted by three locations P-1, P-2 and P-3 in Figure 1.
Thasnanipan et al. (1998), presented the instrumented pile load test results of wet process
bored piles tested in Bangkok subsoils. The authors have revealed that in case of pile
V ~ 69

71

P-1

P-2

Depth (m)

P-3

Figure 1: Typical soil profile along a 4.5 km long stretch of an expressway in the plain of
Bangkok, with different pile embedment conditions.
embedment conditions like P-1 and P-3, 2nd stiff clay often exhibit a brittle type of failure and
peak skin friction values are mobilized well before the sand layers.
A family of t-z curves obtained from the instrumented pile load tests showing this type
of failure is presented in Figure 2. The peak friction values in these layers are mobilized at
very low pile head movements (10 15 mm depending upon the depth of stiff clay layer) and
well before the ultimate friction values of other soil layers present along the pile shaft. When
rest of the soil layers reach their peak friction values stiff clay layers have reached the residual
friction values.
10
8

Skin Friction Mobilized (ton/m )

9
7
6
5
4
3
2
1
0
0

200

400

600

800

1000

1200

Butt Load (ton)

Figure 2: Family of curves obtained from instrumented pile load tests showing typical brittle
failure in stiff clay layers.

72

V ~ 70

A typical load test showing this type of behavior is shown in Figure 3. It can be seen
that the stiff clay layers reach their ultimate friction values well before the sand layers and at
the maximum test load stiff clay layers reach their residual friction values. Design parameters
like and suggested by different researchers like Suchada (1995) for the estimation of
ultimate skin friction in Bangkok subsoils give a reasonable estimate for the embedment
conditions like at P-2 in Figure 1 or where the thickness of 2nd stiff clay layer is relatively
low. Exceptions have been observed in case of embedment conditions especially like P-1,
where a thick 2nd stiff clay layer is present. Owing to the difference in shaft friction
development with pile head movement in the stiff clay layers as shown in Figure 3, direct
summation of (simple superposition) ultimate shaft friction capacities in different soil
horizons leads to over estimation of shaft capacity of these piles. Due to the brittle nature of
failure mechanism in very stiff to hard clay layers peak friction resistance is mobilized at
comparatively very little movement at the pile/soil interface. Residual values of friction
resistance in these layers drop to 50 percent or even smaller of the peak friction values
mobilized.

Butt Load
0.00
14

Soft Clay

Su = 1.5 t/m 2

10.00

40.00

1st Sand

2nd Stiff
Clay

Su = 13.5 t/m 2

SPT-N = 35

Su = 25.5 t/m 2

50.00

Medium Clay
1st stiff Clay

Mobilized Skin Friction (ton/m )

Depth (m)

30.00

1st Stiff
Clay

Su = 4.5 t/m 2

Bored Pile

Medium
Clay
20.00

Soft Clay

12

10

1st Sand
2nd Stiff Clay
2nd Sand

2nd Sand

SPT-N = 50

400

800

1200

1600

2000

Butt Load (ton)

60.00

Figure 3: Skin friction development with the increase in butt load which clearly shows a
typical early yielding of stiff clay layers.
On the extreme, from some instrumented pile load test results it has been observed that
stiff clay layers reach their peak friction values near the design service loads. Rest of the soil
layers need further pile/soil interface movement for the full mobilization of skin friction, and
at this stage, stiff clay layers have already arrived at their residual friction values. Now
depending upon the relative thickness of stiff clay layers, overall shaft friction capacity
mobilization is affected differently. If the thickness of very stiff to hard clay layers is such
that their contribution in the total shaft friction capacity is proportionally high, when stiff clay
layer yields, load settlement curve shows a localized plunging behavior. The movement of
such plunging consists of few millimeters and when the other soil layers shares the deficit in

73

V ~ 71

Table 1: Comparison of maximum skin friction values mobilized and the values at maximum
test load.

Soft Clay

1.5

Mobilized Unit Skin Friction Values

(ton/m2)
Max. Mobilized
At Max. Test Load
fs
fs

1.7
1.13
1.6
1.07

Medium Clay

4.5

2.2

0.49

1.9

0.42

1st Stiff Clay

13.5

5.8

0.42

5.1

0.38

2nd Stiff Clay

25.5

8.8

0.34

4.2

0.16

Undrained
Shear
Strength
Su (ton/m2)

Clay Layer

peak and residual friction capacity of these layers, once again pile starts taking load followed
by this local plunging, as shown in figure 4, till the final plunge exhibiting the failure of all
soil layers along the pile shaft reaches. It must be noted that this localized plunging was
earlier conjectured to be because of the soft toe of the piles. On the other hand if the
contribution of skin friction derived from very stiff to hard clay layers is proportionally low,
localized plunging due to earlier failure of very stiff to hard clay layer is not visible. Normal
practice to estimate the contribution of skin friction from clay layers is by using method,
with a value of equal to around 0.3 for very stiff to hard clay layers. Instrumented pile load
test results have shown that skin friction values calculated using value equal to 0.3
corresponds to the peak skin friction in these layers. Owing to the brittle type of failure
mechanism in these layers, values at full mobilization of total shaft friction capacities drop
to as low as 0.15 in some cases and the ultimate shaft capacity calculated gives over estimates
as compared to the actually mobilized capacities. Skin friction values mobilized in the pile
load test shown in Figure 3 are given in Table 1.
0

500

Butt Load (Tons)

1000

1500

2000

Pile Head Movement (mm)

0
20

TP-1
TP-2

40
60
80
100
120

Figure 4: Typical load-settlement curves showing localized plunging before the overall failure
of the pile.

74

V ~ 72

Peak and residual values mobilized in the stiff clay layers from Table 1 are plotted in
figure 5 along with the suggested curves by different researchers. It can be noted that the
peak value mobilized in 2nd stiff clay layer corresponds well to the value suggested by Stas
& Kulhawy (1984). But the residual value at the maximum test load drops below the curve
suggested by Suchada (1989). So the values proposed in figure 5 for the stiff to hard clay
layers gives overestimate of the ultimate shaft friction under theses conditions.

1.20
Bored Cast Insitu Piles
1.00

Adhesion Factor ,

Tomlinson,1957
0.80
Suchada,1989
for bored piles in Bangkok soils

0.60

0.40

Peak
Stas and Kulhawy, 1984

0.20

Residual

0.00
0.00

5.00

10.00

15.00

20.00

25.00

30.00

Undrained Shear Strength (ton/m )

Figure 5: Comparison of adhesion factor , suggested by different researchers with the actual
mobilized in the stiff clay layers.
EFFECT OF LOWER END BEARING VALUES ON SKIN FRICTION
Meibner & Van Impe (1993) firstly reported the influence of lower end bearing values
on the shaft friction capacity of piles. Laboratory model tests as well as finite element
analysis performed by the authors showed that the presence of soft soil within a depth of two
to four times pile diameter below the pile tip, not only decrease the end bearing but also the
skin friction capacity. Thasnanipan et al. (1998) also revealed that the end bearing capacity of
bored pile where toe is embedded in to the clay layer would be reduced and this end bearing
capacity reduction might influence the shaft resistance capacity too. This conclusion was
reinforced by the pilot pile load test result in the project located on Rama III road in Bangkok
where bore pile having diameter 1500 mm, toe founded at 60m in clay layer, failed at only
about 1750 tons with total settlement of over 90 mm. While the designed maximum test load
was 2500 tons. An other contract pile of same diameter and founded in sand layer at depth 55
V ~75
73

m and base grouted was successfully tested up to 2500 tons with total settlement of only
29.26 mm and elastic recovery was 17.52 mm recorded. Even though the second pile was
shorter in length than the first, more skin friction was mobilized due to the improved end
bearing conditions. A similar decrease in the shaft friction capacity in the sand layers has
been observed from the instrumented pile load tests at another project in Bangkok.
CONCLUSIONS
For the piles embedded in the multi layered soils of Bangkok, estimation of ultimate shaft
friction capacity needs to consider the brittle type of failure mechanism of stiff to hard clay
layers and the values selected need to be adjusted accordingly. Where the pile tips are
embedded in the clay layer, effect of lower values of end bearing on shaft capacity should also
be considered.
ACKNOWLEDGEMENTS
The authors are grateful to their colleagues, especially Mr. Pornpot Tanseng for their
assistance in the preparation of this paper.
REFERENCES
Meibner, H., Shen, Y.L., Van Impe, W.F. & Vogt, C. (1993), Punching Effect for Bored
Piles. Proc. of the 2nd Int. Geotechnical Seminar on Deep Foundation on Bored and
Auger Piles; Ghent, Belgium.
Stas, C.V. & Kulhawy, F.H. (1984), Critical Evaluation of Design Methods for Foundations
Under Axial Uplift and Compression Loading, Report EL-3771, EPRI, Palo Alto, Calif.
Suchada (1989), Bored Piles in Bangkok Subsoils, M. Engg. Thesis, AIT, Bangkok.
Thasnanipan, N., Baskaran G. & Anwar, M.A. (1998), Large diameter bored piles in
multilayered soils of Bangkok, Proc. of the 3rd Intl. Geotechnical Seminar on Deep
Foundations on Bored and Auger Piles, Ghent, Belgium, Balkema.
Thasnanipan, N., Teparaksa, W., Maung, A.W. & Baskaran, G. (1998), Design, Construction
and Behavior of Bored Cast In-Situ Concrete Piles in Bangkok Sub Soil. Proc. of Fourth
Intl. Conf. on Case Histories in Geotechnic Engineering, St. Louise, Missouri.
Tomlinson, M. J. (1957), The adhesion of Clay Piles Driven in Clay Soils, Proc. 4th Intl.
Conf. on Soil Mechanics and Foundation Engineering, London.

76
V ~ 74

GeoEng2000

An International Conference on
Geotechnical & Geological Engineering
19-24 November 2000
Melbourne Exhibition and Convention Centre
Melbourne, Australia

Barrettes : A versatile foundation for transmission line towers

Narong Thasnanipan, Pornpot Tanseng, Aung W. Maung and Muhammad A. Anwar
SEAFCO Co., Ltd., Bangkok, Thailand

ISBN No. 1-58716-068-4

77

GeoEng2000 An International Conference on Geotechnical and Geological Engineering 19-24 November 2000, Melbourne, Australia

BARRETTES; A VERSATILE FOUNDATION FOR

TRANSMISSION LINE TOWERS
Narong Thasnanipan1, Pornpot Tanseng2, Aung W. Maung3 and Muhammad A.
Anwar4

ABSTRACT
An efficient design of foundations for transmission line towers has always been a challenge for the
engineers due to the variety and cyclic nature of the loads. Foundations, especially for the four-legged
towers, are subjected to all types of loads (compression, tension, torsion and shear) in different combinations.
The cyclic nature of the loads further complicates the situation. Available design parameters proposed by
different researchers are also mostly based on the monotonic loading conditions and are not directly
applicable for tower foundations. This paper presents the analysis, design and construction practice of the
barrettes used for the transmission line towers in Thailand. Adaptability of the barrettes under different site
constraints like sensitive underground pipelines is also described.
INTRODUCTION
The 230kV-transmission line project presented in this paper involved construction of 40 transmission
towers, 400m apart. The project site was located about 200km, east of Bangkok, Thailand. The transmission
line was planned to convey electricity from the power plant to a sub-station. The alignment of 15.6km-long
transmission line partially runs along the boundary of an Industrial Estate and thus the foundations of these
towers are in close vicinity of existing underground gas pipelines and utilities above ground (Fig.1).
Foundation layouts, construction method and the work area were thus restricted by presence of these utilities.
GULF OF THA
ILAND

LEGEND :
LATTICE TYPE TOWER
MONOPOLE TYPE TOWER
TRANMISSION LINE

POWER
GENERATION
PLANT

GAS PIPELINE

CHEMICAL
PLANT

CHEMICAL
PLANT

CHEMICAL
PLANT

1.5km

GAS
SEPARATION
PLANT

CHEMICAL
PLANT

CHEMICAL
PLANT

0km

CHEMICAL
PLANT

2.5km

5km

SUB STATION

Figure 1. Layout of project site

1

Managing Director, Seafco Co., Ltd.. Bangkok, Thailand

Geotechnical Engineer, Seafco Co., Ltd.. Bangkok, Thailand
3
Project Manager, Seafco Co., Ltd.. Bangkok, Thailand
4
Project Manager, Middle East Foundations Group, Dubai, UAE
2

79

Initially, the structural designer proposed

large diameter bored piles of 2.0m in
diameter
for
tower
foundations.
Construction of such large bored piles
would require heavy equipment and plant,
and would induce unacceptable vibration
level to the existing underground gas pipe
lines located within and along the
construction area.
As an alternative
construction method and foundation type
to suit the site conditions, barrette
construction that utilizes relatively less
equipment and induces less vibration than
conventional bored pile construction was
adopted.

Construction of barrettes in a site with limited headroom and constraints and performance of barettes in
Bangkok soil has been reported by Thasnanipan et. al. (1998) and (1999). The past experience indicated the
feasibility of this project.
SOIL CONDITIONS

40

39

38

37

36

35

34

33

32

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

15

14

13

13/A

12

11

10

BH

No.

Initially, only a few boreholes were made to investigate subsoil conditions. However as the great distance
between towers meant that substantial variations in soil conditions could occur. An extensive soil
investigation was later carried out at all tower locations by drilling to verify subsoil condition to optimize
barrette design and for excavation feasibility. The general subsoil profile revealed fine to coarse sand at the
top, overlaying silty clay (lateritic soil) with varying thickness. The lateritic soil was not present at some
locations. Below the sand and lateritic soil was completely decomposed granite at varying depths. Figure 2
shows cross sections of subsoil at tower locations along the transmission line. The groundwater level found
also varied from about 0.6m to 9.2m below ground surface.

0.00

5.00

DEPTH (M)

10.00

15.00

20.00

25.00

30.00

LEGEND
CLAYEY SAND
SANDY CLAY (LATERITIC SOIL)
SANDY CLAY (COMPLETELY DECOMPOSED GRANITE)
SILTY SAND (COMPLETELY DECOMPOSED GRANITE)

Figure 2. Soil profile along the transmission line

0

BH-3

BH-11

SPT-N

SPT-N

50

100

50

BH-20

BH-32

SPT-N
100

50

BH-39
SPT-N

SPT-N
100

50

100

50

100

0.00

5.00

DEPTH (M)

10.00

15.00

20.00

25.00

30.00

Figure 3. SPT test results at some tower locations.

SPT N values of soil layers from different boreholes indicated that the depth of competent layer varied
significantly (Fig. 3). This meant that penetration depth of barrettes would vary with soil conditions and
barrettes need to be designed individually.

80

BARRETTE DESIGN
The design of barrettes mainly involved (1) layout arrangement and (2) structural and geotechnical
capacity of barrettes in connection with load, site and soil conditions and construction practicality.
Regarding the layout, a four-barrette group (Fig. 5) was selected to support 38 lattice towers (up to 61.4m
high) individually where a large area was available. Otherwise, a cruciform barrette (Fig. 6) was used for
monopole towers (51.4m high). Compression, uplift and lateral forces were considered the significant loads
to barrettes and they depended on the position of the tower in the line, the natural forces (wind) exerted on
towers and cables and the weight of tower and cables. The towers were generally classified according to
their positions on the line; end towers, edge towers and towers on a straight line between two other towers.
The end towers were longitudinally loaded from one side by the cables and the overturning moment on such
towers was large. The edge towers were loaded both vertically and transversally while the towers on straight
line mainly carried a vertical load. A summary of estimated loads on the individual barrettes of the towers is
presented in Table 1. To achieve load carrying and uplift capacity based on the imposed load and soil
conditions, barrette sizes of 1.0x2.7m and 0.8x2.7m with depths from 11.0m to 22.0m were selected.

0.8x2.7

0.8x2.7

0.8x2.7

0.8x2.7

0.8x2.7

0.8x2.7

0.8x2.7

0.8x2.7

0.8x2.7

1.0x2.7 Edge tower

1.0x2.7 End tower

32

33

34

35

36

37

38

39

40

0.8x2.7

31

0.8x2.7

29

30

0.8x2.7

0.8x2.7

28

0.8x2.7

18

1.0x2.7 Edge tower

0.8x2.7

17

27

0.8x2.7

16

0.8x2.7

1.0x2.7 Edge tower

0.8x2.7

15

Horizontal (kN)
1087
898-1087
309-505

26

1.0x2.7 Edge tower

23

0.8x2.7

14

1.0x2.7 Edge tower

1.0x2.7 Edge tower

22

0.8x2.7

13/A

25

1.0x2.7 Edge tower

21

0.8x2.7

13

0.8x2.7

0.8x2.7

12

11

10

0.8x2.7

Tower on a straight line

24

0.8x2.7

0.8x2.7

20

0.8x2.7

Uplift (kN)
3078
2804-3078
1032-1565

19

1.0x2.7 Edge tower

1.0x2.7 Monopole

1.0x2.7 Monopole

1.0x2.7 Monopole

1.0x2.7 Edge tower

2

1.0x2.7 End tower

1

Size
No.

Tower Barrette

Remark

Table 1. Summary of Loads on individual barrettes

Tower Classification
Compression (kN)
End Tower
3792
Edge Tower
3479-3792
Tower on straight line
1370-2039

Tower on a straight line

0.00

5.00

DEPTH (M)

10.00

15.00

20.00

25.00

30.00

Figure 4. Dimensions and type of barrettes for the towers

BARRETTE
0.80X2.70m

2.47

15.90

GAS PIPE
LINE
34"

2.67

15.90

3.48

4.47
GAS PIPE LINE
34"

A
-

0.

80

70

2.

15.90

SECTION A-A

Figure 5. Typical barrette layout for a transmission tower (Lattice Type)

81

Generally, a safety factor higher than 2.3 and 1.2 was adopted for compression and uplift load capacity
respectively. For uplift capacity, reduced friction capacity (up to 50%) in the cohesionless soil above fixity
level was used in calculation since barrettes were subjected to horizontal cyclic loading. Alexander J.
Verstraeten. (1987) has described decrease of friction by dynamic loading patterns in details. Finite element
method (FEM) was used for calculating bending stesses in the barrettes and prediction of lateral movements.
A total of 6 pile types was classified according to imposed loads, pile dimensions and soil conditions, and
then analysed by FEM. Reinforcement of barrettes was designed using strength design method. A concrete
cover 75mm was used for all barrette types.
The lattice towers were anchored in the barrettes by stubs and cleats which were cast into the extended
barrette section after trimming the barrette top. For monopole towers, anchor bolts and steel flanges were
used.
Three monopole towers were used where the lattice type tower could not be allocated. Due to the
presence of underground gas pile lines and the narrow area available, it was difficult to allocate barrette
groups with adequate alignment to resist the maximum bending moment and horizontal force up to
25,840kN-m and 635kN respectively. Cruciform barrettes were then designed to resist such high bending
moments and forces acting on the base of the monopole.
MONOPOLE
TOWER

2.00

1.00 2.70 3.40

PILE CAP

CRUCIFORM
BARRETTE
1.0X2.70m

1.00
2.70
3.40

SECTION A-A

Figure 6. Cruciform barrette for a monopole type tower

BARRETTE CONSTRUCTION
A mechanical rope grab was used to excavate the trench under bentonite slurry. The same crane was used
for excavation and lifting reinforcement cages as well as construction plant and facilities. A guide wall with
inside clear dimensions slightly more than the nominal size of the barrette was used to guide the grab.
Circulation of slurry was continuously done to keep the bentonite slurry agitated, to minimize the building up
of filter cake on trench wall surfaces. Properties of bentonite slurry were maintained within the specified
ranges in wide use. Construction sequences are shown schematically in Figure 7.

GUIDE WALL
TREMIE PIPE

REINFORCEMENT
CAGE
BENTONITE
SLURRY

EXCAVATION

INSTALLATION OF
REINFORCEMENT

Figure 7. Barrette construction sequences (schematic)

82

CONCRETING

Figure 8. Guide wall for a cruciform barrette

Figure 9. Reinforcement being installed in the trench

Sediment or loose materials at the bottom of the trench were removed and any built-up filter cakes were
scraped off by the grab before reinforcement cage installation. Entire barrette length was fully reinforced
and up to 24m-long reinforcement cages were fabricated in one complete section, for installation into the
trench. The reinforcement cage of the cruciform barrette was as the same shape as the barrette. For the
purpose of lifting and handling the cage, temporary stiffeners, lacing and tie bars were necessary. These
temporary bars were removed section by section while the cage was lowered into the trench.
After excavation, the trench profile was checked with Koden drilling monitoring equipment. Tremie
concreting method was used for casting the barrettes. Since the cutoff level of barrettes was generally at
ground level, ready-mixed concrete was poured until all slurry and slime in the trench was completely
displaced and fresh concrete could be seen.
Up to 172kg/m3 and 381kg/m3 of SD40 steel bars with maximum steel percentage at barrette top section
(about 12.0m) were used to reinforce the group and cruciform barrettes respectively. Ready-mixed concrete
with cube strength of 38MPa at 28days was used for casting barrettes.
CONSTRUCTION PROBLEM AND CORRECTION
A trench collapse was reported during desanding operation after trenching was completed for a barrette
of lattice tower. A sudden fall of bentonite slurry level by a few meters in the trench was observed and then
the trench was found filled with soil up to 10m depth. Mode of failure of the trench was found to be in the
form of a localized cavity failure in the loose sand present at 2.0m to 6.0m depth. Failure surface did not
extend up to the ground surface and the guide wall was still intact without any distortion or damage. The
most probable cause of the trench collapse was due to heavy rainfall during construction. Surface water was
seen in the vicinity of the construction area. The rain had raised the groundwater level to near the ground
surface and caused a groundwater flow reversal towards the trench, triggering cavity formation in the loose
sand stratum. The cavity progressively increased in size, leading to the collapse of the trench during the
desanding operation. The collapse trench was backfilled with cement-bentonite mix, and wooden piles were
driven into the collapse area outside of the barrette outline. The guide wall was then raised up to 1.5m above
ground level. The trench was re-excavated under the slurry head 1.5m above the expected groundwater level
and the barrette was completed.
QUALITY CONTROLS
All trenches were checked for verticality and dimensions prior to reinforcement cage installation to avoid
any obstruction associated with trench inclination. In particular, all sides of the cruciform trench were
checked. If necessary, verticality of trench was improved by careful chiselling with the grab. Profiles of the
trench for a cruciform barrette is shown in Figure 10.
As the subsoil conditions varied from one location to another, they were observed during trenching
process and verified with the soil conditions assumed in design, especially in the lower section of the
barrette. Slurry quality was regularly tested and maintained within the specified ranges. Since barrettes

83

were highly reinforced, in order to achieve a good flow of concrete, the concrete mix was checked for
appropriateness of slump and cohesiveness prior to casting.

Figure 10. Trench profile of a cruciform barrette

recorded with Koden equipment - Tower no. 5

Figure 11. Test signal showing sound integrity and

toe reflection of a barrette - Tower no. 2

Sonic integrity/seismic test was carried out on all barrettes about 5 days after casting for checking barrette
integrity. The test indicated that integrity of all barrettes were sound. Figure 11 shows a signal acquired by
sonic integrity test on the cruciform barrette.
DISCUSSION AND CONCLUSION
For barrette piling in loose cohesionless soil, chances of rapid groundwater level rise due to the rain, tide,
etc., need to be considered during trenching. If necessary, guide wall with appropriate height above the
possible highest groundwater level must be used.
Barrettes are versatile and can be used efficiently in the areas where conventional bored piling cannot be
done. Moreover, barrette construction uses less quantity of equipment than bored piling and thus it is
suitable for the piling work requiring shifting from one location to another.
Verticality of trench is critical for barrettes with complex cross-sectional shape, such as cruciform, T and
H for reinforcement cage installation. A thick concrete cover up to 100mm is thus recommended, in
particular for cruciform barrettes for ease of reinforcement cage installation and to maintain the adequate
concrete cover, considering trench verticality against high stiffness of rebar cage.
REFERENCES
Thasananipan N., Singtowkaew K. and Tanseng P. (1999), "Experiences in Construction of Barrette (in
Thai)", National Convention on Civil Engineering, Pattaya, Thailand, May 24-26, 1999. pp. GTE-144149.
Thasnanipan, N. Maung, A. W. and Tanseng, P. (1998), "Barrettes Founded in Bangkok Subsoils,
Construction and Performance", Thirteenth Southeast Asian Geotechnical Conference, Taipei, Taiwan,
ROC, November 16-20, 1998. pp. 573-578.
Alexander J. Verstraeten. (1987), "Steel Pile Foundations for Transmission Line Towers, as Used in Western
Europe", Proceeding of the 2nd International Conference of Deep Foundation Institute, Luxembourg, May
4-7, 1987. pp. 297-325.

84

Proceedings of the

GEOTECH YEAR 2000

Development in Geotechnical Engineering
Editors:
A. S. Balasubramaniam
Dennes T. Bergado
Lin Der-Guey
Tian Ho Seah
Kinya Miura
Noppadol Phien-wej
Prinya Nutalaya

Practical installation of stanchions for top-down

construction in Bangkok subsoil
Narong Thasnanipan, Aung Win Maung & Zaw Zaw Aye
SEAFCO Co., Ltd., Bangkok, Thailand

27-30 November 2000

Bangkok, Thailand

85

GEOTECH-YEAR 2000
Developments in Geotechnical Engineering
27-30 November 2000, Bangkok, Thailand

PRACTICAL INSTALLATION OF STANCHIONS FOR TOP-DOWN

CONSTRUCTION IN BANGKOK SUBSOIL
Narong Thasnanipan, Aung Win Maung & Zaw Zaw Aye
SEAFCO Co., Ltd., Bangkok, Thailand
ABSTRACT
Top-down construction method has been used for construction of deep basement at
major projects in Bangkok, Thailand. To be able to apply top-down construction method,
structure elements are required to sustain the construction load and to utilize as a part of
bracing system. Prefabricated steel columns known as stanchions embedded in bored piles or
barrettes are commonly used for this requirement. This paper presents some experiences in
the practical installation of the different stanchion types in large diameter deep-seated bored
piles and barrettes in Bangkok. The problems encountered during the actual installation of the
stanchions and solutions to them are also discussed.
INTRODUCTION
Conventionally, buildings with underground basements are built by bottom-up method
where sub-structure and super-structure floors are constructed sequentially from the bottom of
the sub-structure or lowest level of basement to the top of the super-structure. Though this
conventional method, also called as bottom-up method, is simple in both design and
construction, it is not feasible for the gigantic projects with limited construction time and/or
with site constraints. Top-down construction method which provides the significant saving of
the overall construction time has been adopted for some major projects where time factor is of
primary importance. To be able to apply top-down construction method, rigid permanent
retaining wall and pre-founded structural columns are required. Diaphragm wall is normally
used as basement retaining wall whereas prefabricated steel stanchions embedded in bored or
barrette piles are utilized to sustain the construction load as well as to form as a permanent
column at later stage. The common length of the stanchions used in Bangkok ranges from 15
to 30m for excavation depth of 12 to 25m. Installation of heavy and lengthy stanchion with
stringent positional tolerance in the boreholes or excavated trenches filled with drilling fluid
in prevailing subsoil condition of Bangkok is a challenging job for the foundation contractors.
SUBSOIL CONDITION AND PILE CONSTRUCTION METHOD
Bangkok subsoil is featured by alternating clay and sand layers of thick Quaternary
deposits. As characteristics and properties of Bangkok subsoil layers have been well
documented in many literatures, only a general description of Bangkok subsoil is presented in
this article as follows.
Made-Ground consists predominantly of Fill-Materials, Clayey Sand or Silty Clay with
some cement block rubble and rock fragments, is commonly found up to 4m depth. Soft to
very soft, highly compressible dark gray marine clay lies beneath Made-Ground and in some
337

87

areas it lies under weathered crust layers of 2m thick. Depending on the location, this layer is
extended up to 12-18m. About 2m thick Medium Clay layer can be observed between Soft
Clay and underlying Stiff Clay. Generally Stiff Clay layer occurs directly underneath
Medium Clay and its depth goes up to 22m. Below Stiff Clay layer, First Sand layer 5-8m in
thickness can be found. This First Sand layer, however, is absent in some areas. Stiff to Hard
Clay layer underlies First Sand and it is found to be about 5m thick. Second Sand layer
generally occurs at depths between 45 to 65m.
Stanchions are mainly installed in the large diameter deep-seated bored piles and
barrettes founded in first or second sand layer. Bored piles of diameter 1.50m to 1.8m and
rectangular-shape barrettes of 0.80x2.8m to 1.5x3.0m (thickness x width) founding at depth
between 50m and 60m are commonly used to accommodate the stanchions. Wet process
method is always used for construction of these large bored piles and barrettes in Bangkok.
For bored piles, temporary casing of 14-15m in length is used as a support in soft clay layer.
Soil inside the casing is normally excavated by auger applying rotary drilling method and
drilling is continued with the bucket under the slurry from the top of sand layer to the final
depth. A cable-hung-grab mounted on crawler crane is used for excavation of trench with
slurry support in barrette construction. Guide walls having a minimum depth of 1.2m are
commonly used for initial guiding of grab excavation at the top of the trench. Tremie
concreting is necessary for casting both bored piles and barrettes.
APPLICATION OF TOP-DOWN CONSTRUCTION METHOD
Top-down construction method as the name implies, is a construction method, which
builds the permanent structure members of the basement along with the excavation from the
top to the bottom. Top-down method is mainly used for two types of urban structures, tall
buildings with deep basements and underground structures such as car parks, underpasses and
subway stations. For tall buildings with deep basements, a full application of top-down
method enables superstructure to build concurrently with excavation and construction of the
basement giving a significant advantage in reducing the overall construction time. Top-down
method is the most appropriate solution in minimizing the duration and disruption to surface
traffic and other urban activities, for the construction of the underground car parks,
underpasses and subway stations where the structure is located directly underneath the
existing roads.
Application of top-down method offers three main advantages as outlined below.

It allows early commencement of super-structure construction without the need to

wait for excavation reaching the bottom level. Overall construction duration can be
significantly less than that of conventional method thereby saving the cost.

It does not require temporary bracing system so that time requirement and cost for
temporary works (both material and labor costs) are eliminated, in turn significantly
saving cost.

Minimize soil movement induced by excavation works which is an important factor

for some sensitive locations.
STRUCTURAL MEMBERS REQUIRED FOR TOP-DOWN CONSTRUCTION
Design and construction principles for top-down method primarily call for two major
structural elements.

Columns with sufficient capacity must be pre-founded in bored piles or barrettes to

sustain the construction load and to utilize as part of bracing system.

88
338

Excavation for basement must be carried out with the support of permanent
retaining wall so that basement floor slabs can be utilized as lateral bracing.
Diaphragm wall of 0.8m to 1.2m in thickness with sufficient embedment in firm soil
layers is commonly used as a retaining wall whereas prefabricated steel columns known as
stanchions embedded in either large diameter deep-seated bored piles or barrettes are utilized
as structural columns. Figure 1 illustrates the top-down construction method with utilization
of stanchions and diaphragm wall.

F3
Top

F2

F1

GF
B1

Stanchion

B2
Down

B3
B4

Bored Pile

D-wall

Figure 1. Top-down construction with stanchion and diaphragm wall

TYPES OF STANCHION AND THEIR APPLICATION IN BANGKOK
Pre-fabricated H section steel columns and steel built-up-section columns are mainly
used as pre-founded structural columns or stanchions. Pre-cast reinforce concrete columns are
very seldom used (Manoharn S. & Aye Z. Z. , 1994). For the purpose of convenient reference
throughout the paper, the type of stanchion is categorized by size and capacity as presented in
the table below.
Table 1. Technical data, application and limitation of three types of stanchion
Type of
Stanchion
Light
stanchion

Material & Example Size

Steel H-beams
350x350x137kg/m

General Information
Application
For semi top-down
construction
For temporary decking

Mediumsized
stanchion

Steel H-beams
350x350x390 kg/m

For semi and full top-down

construction of shallow to
medium deep excavation

Heavy
stanchion

Steel

Full top-down construction in

deep excavation

H-beams
508x457x738kg/m
Composite steel columns
built up by 2 or more small
to medium size H-beams
Large section pre-cast RC
column (seldom use)

89

Limitation
Limited capacity of light
stanchion does not allow for
construction
of
super
structure until completion of
basement construction
Limited number of super
structure floors construction
Depending on the loading
condition, numbers of
superstructure floors can be
constructed before
completing basement
excavation

339

Embedded length of stanchion usually ranges from1.5 to 5m depending on the loading

condition applied in design and construction.
ALLOWABLE TOLERANCE FOR POSITION OF STANCHIONS
Allowable tolerance for stanchions is usually called by the designer and it is mainly
governed by the structural tolerance of the steel as well as required position of the finished
columns. In most projects, allowable vertical and horizontal deviation of stanchion are
specified as 25 to 50mm whereas verticality is required between 1:200 and 1:400. There are a
number of constraints to get highly accurate position of heavy and lengthy stanchions which
have to be installed in deep-seated foundation piles constructed by wet process in Bangkok
subsoil as described in the following sections.
STANCHION INSTALLATION METHODS
Stanchion installation method is usually selected by the piling contractor who take into
consideration three main factors such as installation depth, size of stanchion and size of bored
or barrette piles. Though installation details may be different from one contractor to another,
stanchion installation can be categorized under two main methods, post-concreting or
plunging installation and pre-concreting installation or placing stanchion prior to concreting.
Post-concreting installation or plunging method
In this method, stanchion is installed immediately after completion of bored pile
concreting process. General construction sequence involved in this method is demonstrated in
Figure 2. Guide frame is used to install the stanchion at the correct position.
Concrete
Guide frame

Tremie pipe

Stanchion

Stanchion

Slurry

Slurry

Slurry

Backfill

Temporary
casing

Temporary
casing

Temporary
casing

Pile cut-off
level

Pile cut-off
level

Pile cut-off
level

Concrete

Complete drilling and

reinforcement installation

Concreting by tremie
method

Installing stanchion by
plunging into concreted pile

Concrete

Complete casing extraction

and backfilling

Figure 2. General construction sequence of pre-concreting installation method

Pre-concreting or pre-placing installation method
In this method, stanchion is installed immediately after completion of drilling and
reinforcement lowering prior to concreting process. In some projects stanchion is attached to
the last section of reinforcement and installed together with the reinforcement. General
construction steps involved in this method are demonstrated in Figure 3.

90
340

Concrete

Guide frame

Stanchion

Stanchion

Stanchion

Slurry

Slurry

Tremie pipe

Temporary
casing

Temporary
casing

Temporary
casing

Pile cut-off
level

Pile cut-off
level

Pile cut-off
level

Pile cut-off
level

concrete

concrete

Complete drilling and

reinforcement installation

Stanchion installation

Concreting
(1 tremie for bored pile and
2 tremie for barrette)

Backfill

Complete casing extraction

and backfilling

Figure 3. General construction sequence of pre-concreting installation method

Figure 4 shows the installation of a 24m long built-up steel stanchion into the pile
borehole prior to concreting.

(a)

(b)

Figure 4. (a,b) Installation of 24m long stanchion into borehole prior to concreting
Guide Frame for stanchion positioning
Guide frame is used for positioning of stanchion in both installation methods.
Effectiveness of guide frame plays one of the important roles in achieving positional accuracy
of stanchion. Figure 5 (a,b) shows the guide frame used for pre-concreting installation method
in one major project constructed by Seafco Co., Ltd.

91

341

(a)
(b)
Figure 5 (a,b) Guide frame used by Seafco for pre-concreting installation in one major project
REVIEW OF THE STANCHION INSTALLATION METHODS USED IN OTHER PARTS
OF THE WORLD
The stanchion installation methods used in other parts of the world have been presented
in some published papers. Among these, Findlay (1989) reviewed a number of stanchion
installation methods used in construction of large diameter bored piles for top-down
construction particularly in UK. The author reported that steel columns can be placed with
better accuracy by dry process bored piling method than wet process method (with support
fluid).
Hollingsworth (1991) demonstrated the installation of stanchions by plunging into
concrete with the aid of an adjustable guiding frame called Cemloc (European Patent No.
0302717) for seven levels basement of 25m deep constructed by top-down technique in
London.
Ressi di Cervia and Tamaro (1991) briefly explained the stanchion installation method
used for construction of an underground garage with top-down technique in Boston, USA.
Steel columns were installed prior to concreting in load bearing elements (LBE) or barrette
excavated by cable-hung mechanical clamshell buckets. The authors cited that using the same
equipment for construction of diaphragm wall and load bearing elements minimized the
congestion on site and streamlined the schedule.
Crawley and Stones (1996) presented the installation of 30m long composite steel
columns for top-down construction applied for Westminster Station in Central London.
Columns were placed with the aid of guide frame followed by concrete pouring. Accuracy of
25mm at the top of column with verticality of 1:200 was achieved as stated by the authors.
Arz (1989) presented an installation of heavy composite columns fabricated from six
steel H-beams encased by concrete (cast during fabrication) for construction of six storey
building with five basement using top-down method in Germany. Due to the high accuracy
requirement of 25mm, adjustable guide frame equipped with three hydraulically control arms
were used. Stanchions were installed prior to concreting as reported by the author.
From the available literatures it is noted that pre-concreting method is more common in
use than post-concreting method.
PROBLEMS ENCOUNTERED IN POST-CONCRETING INSTALLATION METHOD
The major problem commonly encountered in this installation method is inability to
install the stanchion at the design position due to one or combination of the following causes.

Inclination of the borehole / trench

92

342

Stanchion can not be inserted up to the required depth as concrete becomes hard due
to the premature setting

During lowering, stanchion is stuck by reinforcement cage and due to the hardening
of the concrete, extraction of stanchion becomes impossible for reinstallation
As the installation of the stanchion is often associated with many unforeseen problems,
it is likely in many cases that concrete becomes stiff or prematurely set during the installation.
In Bangkok, premature setting of concrete is usually found to be attributed by;

long delivery time due to traffic conjunction

inappropriate mix

severe weather

disruption in concreting / equipment breakdown

PROBLEMS ENCOUNTERED IN PRE-CONCRETING INSTALLATION METHOD

Using this method can minimize some problems encountered during installation
provided that adequate clearance is available between pile reinforcement and the stanchion for
tremie pipes lowering. Inclination of borehole is also important for this installation method.
Proper planning in concrete ordering is required to avoid unnecessary waiting of concrete
trucks for stanchion placing which may affect the total batching time limitation. As stanchion
has to be hung in the position until completion of concrete placement, appropriate and
sufficient capacity of the hanging and positioning device should be provided to avoid falling
and deviation of stanchion in the borehole. Temporary casing should not be used as a hanging
system without additional support for the large stanchion.
FACTORS EFFECTING THE POSITIONAL ACCURACY OF THE STANCHION
It is hardly possible to install a stanchion so that its position and verticality is always as
designed. As-built position of stanchion is found to be generally influenced by the installation
method. The factors affecting the positional accuracy for two different methods are
summarized in the Table 2 and 3.
Table 2. Post-concreting installation method factors effecting positional accuracy

Factors effecting
accuracy
Verticality of piles

Concrete stiffening
or hardening
during installation
Inappropriate
backfill material
Improper back
filling
Improper
temporary casing
extraction

Discussion
Due to the use of fixed guide arm for guiding
stanchion against borehole wall, its verticality
is solely influenced by the verticality of
borehole.

Due to the stiffening or hardening of concrete,

stanchion may not be able to place at the right
elevation. Using vibro-hammer to force down
stanchion can cause stanchion to deviate from
the position.
Large aggregates backfill can cause locking
up casing with stanchion upon casing
extraction which create stanchion to deviate
Back filling from one side of stanchion can
cause deviation of stanchion.
Stanchion deviation in both horizontal and
vertical can be caused by improper temporary
casing extraction.

93

Recommended measures to improve

accuracy
Use adjustable guide
Maintain good horizontal position of
drilling rig and verticality of
temporary casing
Perform drilling monitor test after
completion to check verticality prior
to stanchion installation
Use concrete with appropriate
admixture to achieve longer setting
time
Use

appropriate backfill such as sand

/ fine aggregates.

Maintain

equal distribution of back

fill around stanchion
Apply proper equipment and method
in casing extraction to suit the site
condition and design requirement

343

Table 3. Pre-concreting installation method factors effecting positional accuracy

Factors effecting
accuracy
Verticality of piles

Concrete flows
pushing at
stanchion during
tremie concreting
Improper back
filling
Improper temporary
casing extraction

Recommended measures to improve

accuracy
Maintain good horizontal position of
In this method stanchion position may be
independent from borehole verticality prior to drilling rig and verticality of
concreting. However due to poorer borehole
temporary casing
Perform drilling monitor test after
verticality in comparison with that of
completion of drilling to check
stanchion, clearance between stanchion and
verticality prior to stanchion
borehole wall becomes smaller with depth at
installation and adjust the tremie pipe
one side of stanchion. Thus during the tremie
position according to borehole
pipe lowering and extraction, stanchion is
verticality
pushed by tremie causing stanchion to move
horizontally.
During pouring of concrete from tremie pipe, Use appropriate fixing (lock)at top
force induced by concrete flow tends to move Use 2 sets of tremie pipes placing at
other side of stanchion if possible
stanchion in both horizontal and vertical
direction
Same as post-concreting installation method
Same as post-concreting installation
presented in above Table 2
method presented in above Table 2
Same as post-concreting installation method
Same as post-concreting installation
presented in above Table 2
method presented in above Table 2
Discussion

Skill and experience of the contractor plays a major role in achieving positional
accuracy of the stanchion provided that the installation is practical with the design elements
such as size of foundation piles in relation to size of stanchion.

Figure 6. View of stanchions embedded in bored piles at final excavation level

STANCHION POSITION EFFECTED BY EXCAVATION
In Bangkok, post-installation movement (horizontal movement) of the stanchion is
frequently encountered and it is mainly caused by excavation induced soil displacement
particularly at initial stage of excavation where diaphragm wall deflection is characterized by
large cantilever rotation within Soft Clay layer. In some projects, deviated stanchions had to
be pushed or jacked back to the original position and restrained by means of temporary
support until permanent slab was cast.
REVIEW OF THE COMPLETED PROJECTS
The authors were involved in a number of projects constructed by Top-down method in
Thailand particularly in Bangkok.

94
344

Table 4 shows the summarized technical information of some major projects with
different stanchion type, size and installation depth.
Table 4. Summarized technical information of stanchion application in some major projects in
Thailand
Project
No.

Stanchion Type and Size

Pile Type and

Size

Installation
Method

Built-up steel column

850x850mm, L=24m
built-up by 4 H-beams
270x248x157kg/m
H-section steel column
508x457x 738 kg/m, L=30m*
H-section steel column
419x407x390kg/m, L=24.7m*
H-section steel column
414x405x232 kg/m, L=18m
Pre-cast RC column
500x1000mm, L=15m
500x1500mm. L=15m

Bored pile
1.5x 60m

Pre-concreting

Bored pile
1.8x65m
Barrette pile
1.2x3.0x47m
Bored pile
1.8x65m
Bored pile
1.5x35m
1.8x35m

Postconcreting
Pre-concreting

-29.5m

Pre-concreting

-14.5m

Pre-concreting

-11.5m

2
3
4
5

Basement
Excavation
Depth (m)
-20.0m

-24m

Location

Silom Road
Bangkok
Rama IV Road
Bangkok
Ratchadapisek
Road Bangkok
Payathai Road
Bangkok
Haadyai

Note: * Top levels of stanchion in Project 2 and 3 are 3m and 1.8m below ground level respectively.

A view of heavy stanchions (built-up sections) exposed after excavating to the first
basement level is presented in Figure 7. Pre-concreting installation was used to install the
stanchions in this project.

Figure 7. View of built-up stanchions exposed after excavation to the first level of basement
COMPARISON OF AS-BUILT POSITIONAL TOLERANCE ACHIEVED BY TWO
INSTALLATION METHODS
Although the as-built position of stanchions installed by two different methods have yet
to be statistically analyzed, field survey data of Projects 2 and 3 indicated in above Table 4
suggested that stanchions could be more accurately installed by pre-concreting method.
Stanchions with as-built horizontal deviation greater than 100mm installed by two methods
(expressed by percentage) are presented in the table below.

95
345

Table 5. Comparison of as-built position of stanchions by two installation methods

Project

Installation Method

Depth of
Measurement

As-built Position of Stanchions with

Horizontal Deviation >100mm

Post-concreting Method

15.5m B.G.L

20 %

Pre-concreting Method

13.5m B.G.L

4%

CONCLUSION
Two common stanchion installation methods have been presented with particular
emphasis on the common problems encountered by each method and recommended measures
to alleviate them. According to the authors experience, pre-concreting installation method
provided fewer problems in practical installation and achieved better positional accuracy. A
comparison of as-built position of stanchion between two projects also suggested that
stanchions could be more accurately installed by pre-concreting installation method.
The designers such as structural engineers and architects should be aware of the
accuracy achievable by practical installation method and take it into consideration at the
design stage.
Stanchion installation contractor should select the appropriate method, equipment as
well as experienced personnel plus a well-formed plan with the consideration of all potential
problems to achieve the successful construction of foundation structure, which is of primary
importance for top-down technique.
REFERENCES
Arz P. 1989. Foundations of the Ko-Galerie in Dusseldorf Using Top-down Building
Technologies, Proceedings of the International Conference on Piling and Deep
Foundations, London, 15-18 May 1989 : 185-192. Rotterdam: Balkema.
Crawley J.D. & Stones C. S. 1996. Westminister Station Deep Foundation and Top-Down
Construction in Central London, Geotechnical Aspects of Underground Construction in
Soft Ground, London : 93-97. Rotterdam : Balkema.
Findlay J. D., Wren G.E. 1989. Review of the Methods Used to Construct Large Diameter
Bored Piles for Top Down Construction. Proceedings of the International Conference
on Piling and Deep Foundations, London, 15-18 May 1989 : 199-210. Rotterdam:
Balkema.
Finsted J. A. 1991. Royal Christiana Hotel : Basement with Permanent Sheet Pile Wall, Upand-down Method, 4th International DFI Conference : 387-392. Rotterdam: Balkema.
Hollingsworth J.R. 1991.
Diaphragm Walls, Load Bearing Piles and Piled Soil
Reinforcement for a Deep Top-down Basement Construction, 4th International DFI
Conference : 79-84. Rotterdam : Balkema.
Manoharn S. & Aye Z. Z. (1994) , Bored Pile with King-post and Secant Pile Wall
Construction for Top-Down Method Central Sukhothai Hotel Haddyai, Latest
Technology in Construction Technology Conference, Bangkok, Thailand
Yong K.S., Tsen C.N. & Tang S.K. 1998. Performance of a Deep Excavation with Top-down
Construction, 13th Southeast Asian Geotechnical Conference, 16-20 November 1998.
Taipei, Taiwan ROC : 587-593

96
346

Proceedings of the

GEOTECH YEAR 2000

Development in Geotechnical Engineering
Editors:
A. S. Balasubramaniam
Dennes T. Bergado
Lin Der-Guey
Tian Ho Seah
Kinya Miura
Noppadol Phien-wej
Prinya Nutalaya

Concrete for wet processed bored piles

Thasnanipan N., Maung A. W., Tanseng P. & Navaneethan T.
SEAFCO Co., Ltd., Bangkok, Thailand

27-30 November 2000

Bangkok, Thailand

97

GEOTECH-YEAR 2000
Developments and Geotechnical Engineering
27-30 November 2000, Bangkok, Thailand

CONCRETE FOR WET PROCESS BORED PILES

Thasnanipan Narong, Maung Aung Win, Tanseng Pornpot & Navaneethan Thiruchelvam
SEAFCO Co., Ltd., Bangkok, Thailand.
ABSTRACT
The final quality of wet process bored piles is not only depending on the good
construction practice of piling contractors but also on the characteristics of concrete supplied.
The concrete characteristics are of the major items that influence the performance of piles
after casting. Due to the nature and environment of wet process bored piling, both proper
drilling and pouring of concrete would not produce the good quality piles, if the quality of
concrete did not meet the basic characteristics required for bored piling process. This paper
reviews the basic characteristics of concrete required in wet process bored pile and presents
some example problems caused by improper concrete mix. Additionally, the problems caused
by improper concrete pouring process and preventive measures are also described.
INTRODUCTION
The quality of piles depends on a good construction process including drilling,
reinforcement installation and concrete pouring as well as on good quality of concrete.
Quality of concrete has some influence on the workmanship with interrelated performances.
For wet-process bored piles, concrete is cast under drilling slurry using tremie pipes. Good
quality concrete in bored piling sense means that the properties and characteristics of the
concrete are suitable for the process of work and subsequently meet requirements of the
finished product. Continuous concrete pouring which is mandatory in piling and it is
sometime disrupted by blockage of segregated or prematurely set concrete mix in the tremie
pipe. Early setting of concrete after pouring in bored hole can also cause discontinuities in
pile by accidental lifting of set concrete during extraction of the temporary casing. Dampness
is sometimes found in top section of piles constructed in water-bearing permeable soil layer.
The dampness was found to be caused by capillary action of ground water through
interconnecting voids formed in the improperly mixed concrete.
Ready mixed concrete for bored piles usually specified as self-compacting concrete and
its compacting factor is very important for achieving required strength, especially at the top
section of pile which usually carries the maximum portion of transferred load. However, the
compacting factor is seldom mentioned in piling practice. Besides, no vibration is allowed for
tremie concrete while structural concrete is compacted by vibrator during casting. The
concrete strength test is usually carried out on the compacted test samples and thus the actual
strength of pile can be different from sample strength according to the compacting factor.
WET PROCESS BORED PILING METHOD
In wet process bored piling, bentonite or other type of suitable drilling slurry is used as
drilling fluid to support the borehole during construction. A steel temporary casing is usually

347

99

used to case the top weak soil subjected to heavy construction loads. Drilling and
reinforcement cage installation and concrete placing are successively executed under drilling
slurry. The concret is poured with tremie pipes, displacing the slurry well above the cutoff
level. The temporary casing is then extracted immediately after concreting.
CHARACTERISTICS OF CONCRETE AND CONCRETE MIX
Concrete mix for bored piles is designed according to concrete pouring process and
mechanical properties required. Concrete for wet process piles needs to be specially mixed
having cohesiveness with high workability (high slump/excellent fluidity) which is not prone
to segregation and retain it workability as far as possible throughout the tremie placing
operation for the complete pour. Addition to those characteristics, compaction under selfweight, resistance to harsh environment, resistance to leaching, and appropriate strength are
essential.
Workability
Excellent fluidity is essential that the concrete has the ability to flow readily through the
tremie pipe, to flow laterally through a reinforcement cage and a high lateral stress against the
sides of borehole. High workability is best achieved with rounded natural aggregates and
natural sand in the mix.
Self Compaction
Compaction under self-weight is essential as vibration of concrete is impractical, except
near the surface. The degree of compaction achieved is determined by the density ratio (the
ratio of density actually achieved to the density of the same concrete fully compacted). The
recommended compacting factor for the required workability of tremie concrete is 0.95 to
0.96 (Xanthakos 1994). Fresh concrete is usually placed through tremie pipes and displaces
the slurry by gravity action only. In some cases, lack of self-compaction in the concrete will
lead to defects, such as reversed hanging up, and whirls in the completed pile. If the
initial shear of the concrete is very high, the flow is likely to restrain, resulting in bentonite
trapped in areas not reached by the concrete (Xanthakos 1994).
Resistance to Segregation
The concrete mix should be cohesive and resistant to segregation, as improperly
designed mixes will segregate during placement, resulting in inferior concrete containing
honeycombs and high permeable zones within the pile shaft. Concrete that bleeds or
disintegrates under the pressures of its own weight can also block the tremie pipe or accept
bentonite.
Controlled Setting
The concrete must retain it fluidity thorough the depth of borehole during complete
placement of the concrete in the borehole and attain an appropriate strength within a
reasonable time after placement. Retarders are used to prevent premature stiffening of some
cements or to delay stiffening under difficult placing conditions. The setting time must be
checked against the time necessary to complete the placement. The retarders should be used
under competent technical advice and after adequate testing.

100

348

Resistance to Aggressive conditions

The concrete should have high density and low permeability to resist the possible
(chemical and physical) attack of an aggressive sub-subsurface condition. In some instances
there is an underground flow of water that can cause a weakening of the concrete after it is
placed, and a properly designed mix should be resistant to such flow. However, if the rate of
ground water flow is substantial, a permanent casing will be necessary.
Good Mechanical Performance
The mechanical properties of hardened concrete can be satisfied in most instances.
However, appropriate tensile strength for the concrete without reinforcement in piles and high
level of bending and axial stress must be considered in some cases.
Reese and ONeil (1988) emphasis that the design of the concrete mix must be given
appropriate attention and the design of the mix is dependent strongly on the particular job, and
the cement will be selected to be consistent with the design requirement. They observed that
bleeding is not a problem for concrete mixes that are properly designed. The trial mix method
is usually used in the laboratory. It is necessary to follow-up to see that the materials and
proportions used by the batch plant are those of that are recommended. Inspection at the
batch plant should include checking the nature, quantity and temperature of the components
of the mix, the aggregates, cement, water, admixtures and of the completed mix for
conformance with the specifications. For testing at the job site, the organization of the job
must be such that time required to perform tests at the job site is kept to a minimum.
Excessive job-site testing can lead to harmful effects. No delay in pouring should occur due
to field test.
Adding of water to the concrete with very low slump on site to increase the workability
can have detrimental effect of reducing the strength, compactability and impermeability of the
concrete. The results of adding water could be a significant change in the characteristics of
mix and the possibility of segregation as the pour is made. Segregation of concrete during
pouring can also lead to increase in permeability of concrete, especially at the top section of
piles due to upward migration of water in the concrete mix. Adding of water to the concrete
on site must not be allowed unless specified.
Suggested concrete mix by Fleming & Sliwinski (1977) for bored piles cast under
Bentonite was shown in Table 1 while range of cement content and water-cement ratio in
general use for concrete mixes by Bartholomew (1979) is shown in Table 2.
Table 1. Suggested Concrete Mix for Bored Piles Cast under Bentonite (After Fleming &
Sliwinski 1977)
Slump
Water/Cement

>175 mm.
Below 0.6

Aggregate Type

Natural round stone if possible, 20 mm. max. size

Sand Type

Natural and complying with zone 2 or 3 grading

Sand Content

30% to 45% of total aggregate weight

Cement Content

Not less than 400 kg./m3

Admixture

The use suitable admixture which will improve the workability and
extend the period during which such workability is maintained are to be
advocated.

101
349

Table 2. Range of Cement Content in kg/m3 and Water Cement Ratio in General use for
Concrete Mixes (After Bartholomew 1979)
Conditions
Normal
Moderately
Highly and Very
Pile Type
Aggressive
Highly
Aggressive
450-475
450-475
450-475
Precast
0.4 0.5
0.4
0.4
300-450
350-450
380-500
Bored Piles Dry Process
0.5 0.55
0.475 - 0.5
0.45 - 0.5
Bored Piles with Tremie Process
350-450
350-450
400-500

Driven Cast-In-Situ Piles

0.5 0.6
280-370

0.475 - 0.5
330-450

0.43 - 0.45
370-500

0.25 0.6

0.3 0.55

0.3 - 0.45

Concrete for bored piles compared to that for pre-cast drive piles is the least dense
concrete due to pouring and casting process. Bored pile concrete was cast in the aggressive
subsurface conditions such as, high salinity, ground water fluctuation or in the vicinity of sea
and river. In such environment, the concrete can easily be leached out by ground water. In
this case cement content and water cement ratio given in the Table 2 needs to be reviewed
considering the site conditions.
Fleming et al (1977) pointed out that the high cement content favored for in-situ pile
construction enable the necessary workability mixes to be used with adequate margins of
safety against the inevitable variations in strength and workability. It also compensates for
some reduction in strength, which may occur on interfaces during displacement.
Concrete is permeable to water to the extent that it has interconnecting void spaces
through which water can move. Calcium hydroxide liberated by hydrating cement is watersoluble and may leach out of harden concrete, leaving voids for the ingress of water.
Permeability of concrete is governed by amount of cementitious material, water content,
aggregate grading, consolidation and curing efficiency.
MEHTOD OF CONCRETE POURING
The quality of piles depends on a good pouring procedure as well as on good quality of
concrete. In wet process, the concrete is usually placed by a steel tremie pipe of 20-25cm in
diameter (minimum 6 times of coarse aggregate size).
Prior to charging the tremie pipe with concrete, the bottom of tremie needs to be sealed
by a plug of some descriptions may be inserted at the top of tremie before or after the tremie
is placed in the bored hole as appropriate. There are two potential problems that are associate
with the initial charging of tremie with concrete; the concrete can segregate during placement
and the air in tremie will prevent the complete filling of the tremie. These problems can be
avoided if the tremie is filled slowly. Faulty initial charging of tremie during concreting can
cause entrapment of mud within the concrete.
Excessive initial lifting of tremie can result in possible distribution of leached concrete
caused by concrete falling through the slurry. The bottom of tremie must stay well below the
top of the column of fresh concrete all the time. Moreover, the tremie must not be lifted and

102

350

lowered rapidly to avoid the cause of contamination of concrete with slurry. It is suggested
that the tremie pipe must not be lift and lowered rapidly to start or restart the flow of the
concrete (Reese & Neil 1988). However Xanthakos (1994) suggested that if the concrete is
not deposited easily the triemie pipes may be moved up and down with movement not
exceeding 30cm. Moreover, the tremie pipes should not be moved horizontally.
In the area of high ground water level, the concrete must be deposited above the
external water table before the casing is withdrawn. The hydrostatic pressure in the concrete
column should be greater at all time than the pressure in any column of fluid outside the
casing.
COMMON DEFECTS
The common defects of piles are cold joints, zone of segregated or contaminated
concrete, trapping of bentonite mud and cavities. The first two types of defect result from
interruptions in the concrete placement or premature extracting of the tremie pipe either
partially or completely above the concrete-slurry interface. Mud trappings are caused by
concrete of low workability and impediment to the flow of concrete due to closely spaced
bars. Discontinuities or partial separation in the piles at the bottom edge of temporary casing
can be caused by accidental lifting of low workability concrete or concrete without controlled
setting during casing extraction. If the concrete in the casing is too stiff and has considerable
frictional resistance against the casing, a column of concrete can be pulled up with the casing.
Permeability of concrete depends on the capillary porosity, water-cement ratio and
degree of hydration. High permeability of concrete can be contributed by presence of
capillary pores that are interconnecting voids in the concrete (Figs 1 & 2). The
interconnecting voids are caused by bleeding of concrete due to excessive water used in the
mix. Bleeding raises the water and air bubbles to top surface of fresh concrete and raises the
water-cement ratio of concrete upper part of the forms, thereby reducing the strength and
increasing porosity of concrete. Concrete mixed with a water-cement ratio higher than 0.6
can be more permeable. Function of concrete self-compaction can also reduce the
permeability and increase the density. In piles, usually the fresh concrete compacts under its
own weight, resulting in an increasing density with depth. Besides, concrete in a great depth
of pile is generally cured in stable temperature and moisture.

Figure 1. High permeability of concrete

caused by interconnecting voids

Figure 2. Low permeability of concrete

caused by less interconnecting voids

Usually one type of concrete mix is used for piling in a particular project in most cases.
However, the project in the vicinity of the river, some bored piles are to be installed closer to

103

351

the river than the remaining piles and thus the concrete mixes need to be designed
accordingly. For the piles installed in the deposition side of the river where sand deposits
occur, these piles are usually subjected to be effected by the ground water flow (Fig. 3). If
the cement content is not high enough in the concrete mix and bleeding or segregation occurs,
dampness or wet patches caused by capillary suction to ground water through the previous
tremie location and vertical steels can be found on the pile head (Figs 4 & 5). Figure 6 is a
photo of core sample obtained from the pile that exhibits the dampness on top, showing
segregated concrete caused by bleeding.

Crust Soil
Capillary Suction
Thru Previous
Tremie Location

Hydraulic pressure
and Capillary suction

Permeable
Soil
Water Flow into
High Permeable
Concrete

Soft to Stiff
Clay

Figure 3. Capillary flow of

ground water in bored pile

Figure 4. Capillary suction and hydraulic pressure

in concrete

Figure 5. Wet pile head caused by capillary

flow of ground water through segregated
concrete in the pile

Figure 6. Core samples showing

segregated concrete in the pile

Figure (7) show the some segregated concrete extracted from the tremie pipe, which
was blocked with such concrete. Settlement of solid particles or aggregates in the concrete
mix cause bleeding by migration of the water to the top surface of fresh concrete, reducing
water-cement ratio of the lower part of the mix. As a result, lower part of the concrete mix
stiffens rapidly with aggregates. In such case stiffening concrete can also block the tremie

104
352

pipe during concrete pouring (Fig. 8) causing disruption in casting process and thus effects
the quality of pile. Equipment for concrete pouring must also be adequate and reliable to
avoid any interruption due to a breakdown in continuous tremie concrete pouring operation.

Figure 7. Tremie blocked by segregated

concrete.

Figure 8. Tremie blocked by stiffening

concrete

Due to a congested bar arrangement, concrete cannot flow through the bars and as a
result concrete cover can be lost. Good arrangement of reinforcing bars is thus necessary.
The horizontal spacing of main bars should be at least 10cm and 15cm (minimum 4 times of
the maximum size of aggregate) for small and large diameter bars respectively.
RECOMMENDATIONS AND CONCLUSIONS
Foundation designers and concrete suppliers should pay more attention to workability
and many other important factors required for cast in-situ tremie concrete than strength alone.
It is essential to design the better quality mix than the concrete for other structural works in
some aspects, considering the process of work.
For the projects, such as elevated highways, viaducts and bridges, variable soil and
ground water conditions can be encountered along the project area. In such cases, concrete
mixes need to be designed to suit these conditions and used accordingly in bored pile
construction.
An adequate cement content and water cement ratio is necessary to have good
impermeability of concrete which is one of factors influencing the durability of concrete,
especially for bored piles in water bearing subsoil.
It is concluded that appropriate concrete mix and casting practice is essential in bored
piling work to achieve good quality piles.
REFERENCES
Bartholomew, R. F. (1979), The Protection of Concrete Piles in Aggressive Ground
Conditions: An International Appreciation, Proceedings of the Conference on Recent
Developments in the design and Construction of Piles, ICE London, 21-22 March 1979,
pp. 131-141
Fleming, W. K. and Sliwinski, Z. J. (1977), The Use and Influence of Bentonite in Bored
Pile Construction, CIRIA Report . PG. 3.

105
353

Reese, L. C. & O Neil, M. W. (1988), Drilled Shafts: Construction Procedures and Design
Methods, U.S. Department of Transportation, Federal Highway Adminstration, Virginia
incorporation with ADSC., Dallas, Texas, USA.
Xanthakos, P. P (1994), Slurry Walls as Structural systems, 2nd Edition, McGraw-Hill Inc.

106

354

5th International Conference on

DEEP FOUNDATION PRACTICE

incorporating

PILETALK International 2001

4 6 April 2001, Singapore

RECORD LOAD TEST ON A LARGE BARRETTE AND ITS

PERFORMANCE IN THE LAYERED SOILS OF BANGKOK
Narong Thasnanipan, Aung Win Maung, Zaw Zaw Aye
SEAFCO Co., Ltd., Bangkok, Thailand

Pornpot Tanseng

Suranaree University of Technology, Nakornratchasima, Thailand

Organized by
CI-PREMIER CONFERENCE ORGANIZATION

ISBN: 981-04-2512-0

107

th

5 International Conference on Deep Foundation Practice incorporating Piletalk : 4 6 April 2001, Singapore

Record load test on a large barrette and its performance

in the layered soils of Bangkok
Narong Thasnanipan, Aung Win Maung, Zaw Zaw Aye
SEAFCO Co., Ltd., Bangkok, Thailand
Pornpot Tanseng
Suranaree University of Technology, Nakornratchasima, Thailand

ABSTRACT
This paper presents the load transfer characteristics of fully instrumented barrette of 1.5x3.0m in size
seated about 57m below the ground level. The test results were compared with those from the
instrumented load test on bored pile of diameter 1.5m with the same length, located 30m away. No
significant difference in load transfer has been observed between barrette and pile despite the
considerable difference in construction method applied and time consumed.
Keywords : Barrette, static load test, load transfer characteristics

1. INTRODUCTION
Barrettes have been used as deep foundations for various structures in Bangkok for a number of
years. The large load carrying capacity achievable by flexible dimension and length of barrettes
provides a major advantage in prevailing subsoil condition of Bangkok. In addition to extensive bearing
capacity requirement, the demand for barrettes is necessitated mainly by site constraints, applicable
construction method and equipment. Barrettes with dimension ranging from 0.80mx2.7m to
1.5mx3.0m for safe working load capacity from 1100 to 2300 ton have been used in some major
projects. To assess the performance of large-capacity foundation element in layered subsoil of
Bangkok, instrumented load testing is compulsory. This paper presents the static load test results of
instrumented barrette tested up to 5290 ton for foundation of a fifty-storey building in Bangkok. The
test results, particularly load transfer characteristics, and shaft friction capacity were compared with
those of bored pile diameter 1.50m with the same length located 30m away.
2. OVERVIEW OF THE PROJECT
Foundation of the fifty-storey tower called for 560 bored piles of 1.2m and 1.5m diameter and 24
number of barrettes having cross section size of 1.5m x 3.0m. Bored piles and barrettes were seated
at approximate depth of 57m in the second sand layer. Barrettes were designed to support the large
load for towers lift shafts as bored piles were not feasible to utilize in such case. Base grouting was
applied for barrettes and bored piles at the locations of high column load mainly in the central tower
area. Instrumented static pile load test was proposed for one barrette and one bored pile of diameter
1.5m. The design safe working loads for the base grouted barrettes and bored piles are 2,300 ton and
1,175 ton respectively. Foundation plan of the project is shown in Figure 1.
363

109

132.10

96.20

Test-Barrette

Test
Pile

132.68

94.00

Figure 1 Layout of foundation showing location of test barrette and bored pile
3. SUBSOIL CONDITION
Soil investigation from five boreholes at different location reveals that subsoil layers along the site
are relatively consistent. Similar to other localities in Bangkok a typical subsoil profile at the site is
characterized by the alternating layers of clay and sand deposits as soil succession shown in Figure 7.
Soft, highly compressible dark gray marine clay lies beneath weathered crust layers of 2m thick and
extends up to 13.5m. Stiff Clay layer occurs directly underneath Soft Clay and its depth goes up to
26m. Below Stiff Clay layer, First Sand layer of 10m in thickness can be found. Hard Clay layer
underlies First Sand and it is found to be about 12m thick. Second Sand layer occurs at depths
between 50 to 72m. Undrained shear strength (Su) obtained from unconfined compression test and
Standard Penetration Test (SPT) are shown in Figure 7.
4. CONSTRUCTION METHODS
Both barrette and bored pile were constructed by wet process under bentonite slurry. Bentonite
slurry conforming to the widely accepted specification was used. Properties of the bentonite slurry
used are given in Table 1. Single stage base grouting was applied 24 hours after concreting for both
barrette and bored pile. The detailed procedures utilized in the construction of two different foundation
structures are outlined below.
Table 1. Comparison of bentonite slurry properties
Properties

Before feeding
to the borehole

Barrette
After Recycling &
Before Concreting
(near trench base)
1.17

Before feeding to
the borehole
1.08

Bored Pile
After Recycling & Before
concreting
(near borehole base)
1.10

Density (g/cc)

1.10

Viscosity (sec)

36

49

33

36

Sand Content (%)

1.0

1.1

0.1

0.8

pH value

Mechanical rope-grab was used to excavate the trench. A guide wall cast with inside clear
dimensions slightly larger than the nominal size of the barrette was used to guide the grab during
364

110

initial bites. Since time consumed in the preparation of instrumentation was relatively long, desanding
was continuously done to keep the bentonite slurry agitated, which also helps to alleviate the growth of
filter cake by minimizing the actual exposure time. As another measure, trench was once again
occupied by grab to scrap the trench walls to remove, if any, filter cake formed on the walls. This
attempt is in line with the recommendation made by Reese and ONeill (1988) [1]. It is authors opinion
that, if due to some unforeseen reasons, reinforcement cage lowering have to be delayed for
considerable period of time, it is a good practice to use the grab again to scrap the trench walls. This
measure eliminates any foreseeable negative impacts caused by unexpected delays. After lowering
the rebar cage, tremie concreting was done.
Rotary drilling was employed for bored pile excavation. Different from barrette excavation,
temporary casing of 15 m length was used as a support in Soft Clay layer for bored pile drilling to
assure the stability of the borehole. Firstly, auger was used to drill within the temporary casing,
followed by rotary bucket with bentonite slurry down to final depth of excavation. The base of the
borehole was cleaned by recycling technique to minimize any congregated sediments. Before lowering
the reinforcement cage special cleaning bucket was used to scrap of the borehole walls and the base.
Reinforcement cages were then lowered inside the borehole while attaching the instrumentation
simultaneously at specified locations. Soon after lowering the rebar cage tremie concreting was
commenced. Polystyrene grains plug was used before the first charge of concrete to avoid the mixing
of bentonite with concrete. Time consumed in different construction activities is plotted in Figure 2.

B arrette
D rilling
D eSanding
C age Low ering
Trem ie Pipe Preparation
C oncreting

B ored P ile

10

20

30

40

50

60

70

80

Tim e Co nsum e d (ho urs)

Figure 2 Time consumed in different construction activities of barrette and bored pile
( after Thasnanipan, 1999 [2] )
5. LOAD TEST PROGRAM AND INSTRUMENTATION
5.1 Test Pile Layout
Test pile layout of barrette is presented in Figure 3. With overall height of 8m, reaction frame
utilized for static load testing on barrette in this project was claimed to be one of the biggest of its kind
in the region. Four barrettes were used as anchoring system. Five numbers of built-up steel girders
st
supported on each side by two 1 level cross beams were used as main beams to achieve the
maximum capacity of 6000 tons. First level beams were supported against the second level cross
beams. Second level cross beams were anchored against surrounding barrettes using anchor blocks
at the top. Specially fabricated rigid transfer girders were used to distribute the tension force coming
from the tie-bars to dowel bars above the anchor barrette heads. Sixteen numbers of hydraulic jacks
each having 500 ton capacity were placed between the test barrette cap and the main beams of the
reaction frame. General view of the barrette load test set up is presented in Figure 4.
Test pile layout of bored pile was similar to those of other static bored pile load tests in Thailand.
Steel test frame anchored against four bored piles was used in bored pile load test. Different from the
test frame of barrette, only 1 layer of cross beams was required for that of bored pile.

111365

5.00

5.00

2n d C ro ss
B eam

1st C ro ss
B eam

5.00

AN C H O R
B AR R E TTE

M ain B eam

H ydrau lic
Jacks

5.00

TE S T
B AR R E TTE

2n d C ro ss
B eam

1st C ro ss B eam
2n d C ro ss
B eam

Figure 3

Barrette test pile layout showing the configuration of 6000 ton capacity reaction frame

5.2 Monitoring System and Instrumentation

Direct measurement from four dial gauges placed in diametrically opposite positions having
equidistance from the test pile axis was used as a main monitoring system of test pile head
movement. Precise leveling and piano wire were also utilized as backup for pile head movement
measurement. Additional two dial gauges were also used to monitor the lateral movement of the pile.
Eight number of SINCO load cells of 500 ton capacity each were installed on top of the hydraulic jacks
to evaluate the actual applied load of the first load cycle . Vibrating wire strain gauges (VWSGs) and
Mechanical Extensometers (ME) were fixed at five levels along the shafts at the known interface
boundaries of different soil layers. At each level, four sets of VWSGs and one set of ME were installed
for the pile whereas six sets of VWSGs and two sets of ME were installed for the barrette.

Figure 4 General view of the reaction frame for static load test on barrette

112

366

5.3 Load Test

Pile load tests were carried out in accordance with ASTM D 1143-81. Pile and barrette were
tested under three and four cycles of loading and unloading respectively.
6. LOAD TEST RESULTS AND INTERPRETATIONS
6.1 Pile Head Movement at Applied Load
Load vs pile head movement graph of barrette and bored pile is illustrated in Figure 5. Measured
pile head movements at design load, double of design load and maximum test load are presented in
Table 2. As can be seen in the table, both barrette and bored pile experienced only negligible pile
head movement at relevant design load.
Table 2 Predicted and measured pile head movement at specific applied load
Max. Test Load

Measured gross pile head movement at applied load

At DL
At 2 x DL
At Max. Test Load

Test Pile

Design Load (DL)

Barrette

2300 ton

5290 ton

-5 mm

-12 mm

-61 mm

Bored Pile

1175 ton

2700 ton

-6 mm

-19 mm

-67 mm

Table 3 shows the estimated ultimate load capacity of barrette and bored pile from the load vs pile
head movement using different method.
Table 3 Estimated ultimate capacity of barrette and bored pile
Test Pile
Barrette

Bored Pile

Ultimate capacity (Qult) and pile head movement

at Qult estimated by different method
Davissons
Mazurkiewiczs
Butler and Hoys

Parameter
Qult (ton)

5180 ton

5156 ton

5100 ton

Pile head movement at Qult

18 mm

18 mm

17 mm

Qult (ton)

2675 ton

2734 ton

2620 ton

Pile head movement at Qult

30 mm

38 mm

22 mm

Applied Load (Tons)

0

1000

2000

3000

4000

5000

6000

Pile Head Movement (mm)

20
Barrette
Pile

40

60

80

Figure 5 Pile head movement at the last load cycle (maximum loading) of barrette and bored pile
367

113

6.2 Mobilized Skin Friction vs Pile Head Movement

The graphs showing the ratio of mobilized skin friction to maximum mobilized skin friction against
ratio of pile head movement to pile diameter (s/D) are presented in Figure 6a and 6b respectively.
Symbol D shown in Figure 6a represents the barrette dimension in equivalent diameter. According to
the figures, in general, skin friction mobilized to the maximum values at pile head movement of 0.6%
and 1.6% of equivalent shaft diameter for barrette and diameter of bored pile respectively.

1 .2

(fs As) m obiliz e d / (fs As) m a x im um

(fs As)m obiliz e d / (fs As) m a x im um

1 .2

1 .0

0 .8

0 .6

0 .4

0 .2

0 .8

0 .6

0 .4

0 .2

0 .0

0
0 .0

0 .5

1 .0

1 .5

2 .0

2 .5

0 .0

1 .0

2 .0

s / D (% )

3 .0

4 .0

5 .0

s / D (% )

6b (Bored Pile)

6a (Barrette)

Figure 6 Ratio of mobilized skin friction to mobilized maximum skin friction against ratio of
pile head movement to equivalent pile diameter

6.3 Load Transfer Characteristics of Barrette and Bored Pile

Load transfer curves along the shaft of barrette and bored pile at various applied load are shown
in Figure 7. The values of unit skin friction developed at the different soil layers along the shaft of
barrette and bored pile in comparison with those of calculated ultimate unit skin friction are
demonstrated in Figure 8. This comparison suggests that mobilized skin frictions of barrette and bored
pile are higher than those of calculated values using empirical formulas. It proves the findings of
various researchers on shaft friction improvement of base grouted piles ( eg. Teparaksa et al 1999,
[3] ). It is also evident that overall shaft resistance of both pile was not fully developed at design load.
S u ( U C ) t/ m2
DEPTH (m)

BUTT LOAD (TON)

10

0
Weathered
Crust

1000

2000

3000

1000

2000

3000

4000

5000

Soft
Clay

-10

-20

First Stiff
Clay

-30

Fi r st
Sand

-40
Hard
Clay

-50
Second
Sand

50

100

S P T - N ( b l ows / 3 0 c m)

1.50 m
B ored
P ile

3.00 m

1.50 m

T i p Level
-57.5 m

T i p Level
-57.5 m

-60

B arrette

60

VW SG
M echanical
Extensom eter

Figure 7 Load transfer curves of test barrette and bored pile with typical soil profile at the site

114 368

10

0
Weathered
Crust

-20

First Stiff
Clay

-30

F i r st
Sand

-40

20

30

0
0

Bored Pile
Barrette
Calculated

Soft
Clay

-10

10

10

10

20

20

Depth (m)

Depth (m)

DEPTH (m)

Max. Mobilized Unit

Skin Friction (ton/m2)

Mobilized Unit Skin Friction

at Design Load (ton/m2)

S u ( U C ) t / m2

30

10

20

30

Bored Pile
Barrette
Calculated

30

40

40

50

50

60

60

Hard
Clay

-50
Second
Sand
-60

50

100

SPT-N (blows/30cm)

Figure 8 Comparison of calculated and mobilized unit skin friction of barrette and bored pile
6. 4 Effect of Construction Time on Shaft Capacity Reduction
Excessive construction time is one of the main parameters claimed to be responsible for shaft friction
capacity reduction of drilled shaft foundation constructed with wet process under bentonite slurry as
reported by various researchers. Total construction time of barrette and pile were 75 hours and 27
hours respectively. Though total construction time consumed for test barrette was almost 3 times more
than that of bored pile, there is no significant difference in shaft load transfer between them. The
measures adopted in construction of barrette to minimize the excessive filter cake formation along the
sand layers by proper desanding (continuous agitation), and retrenching with grab are considered the
main reasons contributed to this achievement. Comparison of developed unit skin friction values of
barrette and pile proves that difference in utilized bentonite viscosity as shown in Table 1 does not
have significant effect on the shaft capacity. These findings are in line with conclusions made by
Thasnanipan et al (1999) [4].
6.5 Effect of Shape on Shaft Load Transfer
Hosoi et al (1994) [5] concluded, from the results of numerical analysis that earth pressure acting on
the flat surface of diaphragm wall panel is larger than that of circular bored pile. Thasnanipan et al
(1999) [2] attempted to assess the effect of different aspect ratios (L/B) on the earth pressure
developed around the trenches by using finite element program and reported that no significant
difference in earth pressure was observed between the borehole (L/B=1) and the barrette (L/B=2).
Further research is necessary to evaluate this finding. No significant difference is found between the
maximum mobilized skin friction against displacement of rectangular-shape barrette and circularshape bored pile according to Figure 6a and 6b respectively. Displacements at maximum mobilized
skin frictions of barrette and bored pile fall within the range (0.5 % - 2 % of shaft diameter) reported by
Reese (1978) [6].

369

115

7. CONCLUSIONS
(1) Elastic deformation of both barrette and bored pile at relevant design load was found to be
negligible. Shaft resistance of barrette and bored pile was not fully mobilized at relevant design
load. These observations suggested that the selected size and length of barrette and bored pile
for specified design loads are sufficient for acting as friction piles.
(2) Mobilized shaft frictions of barrette and bored pile at maximum test load particularly in the first
sand layer are considerably higher than calculated values. Shaft friction capacity improvement
was considered to be contributed by base grouting which is in line with the findings reported by
various researchers.
(3) Despite the considerable difference in construction method applied and construction time
consumed (75 and 27 hours for barrette and bored pile respectively), similar characteristics of
shaft load transfer were observed between barrette and bored pile. This achievement indicated
that adopted measures (proper desanding and scraping of trench wall by grab prior to
reinforcement lowering) to alleviate the excessive growth of bentonite filter cake on the barrette
wall proved to be effective.
(4) Static load test up to 5290 ton conducted on barrette set the record as the highest load ever
tested for a single cast-in-situ deep foundation in Thailand.
ACKNOWLEDGEMENTS
The authors wish to express their appreciation to Dr. Wanchai Teparaksa and Mr. Kamol Singtogaw
for providing invaluable suggestions in the initial stage of the load test program. The authors also wish
to thank Mr. Chanchai Submaneewong for his assistance in preparing this paper.
REFERENCES
[1] Reese L.C., & ONeill M. W., Drilled Shafts : Construction Procedures and Design Methods,
ADSC: The International Association of Foundation Drilling, Dallas, Texas, USA, 1988
[2] Thasnanipan N., Anwar M. A., Maung A.W., Tanseng P., Performance Comparison of Bored and
Excavated Piles in the Layered Soils of Bangkok, Symposium on Innovative Solutions in
Structural and Geotechnical Engineering in Honor of Professor Seng-Lip Lee, Asian Institute of
Technology & National University of Singapore, 1999, p.p. 345-353.
[3] Teparaksa W., Thasananipan N., and Anwar M. A., Base Grouting of Wet Process Bored Piles in
Bangkok Subsoils, The Eleventh Asian Regional Conference on Soil Mechanics and
Geotechnical Engineering, Seoul, Korea, 1999, p.p. 269-272.
[4] Thasnanipan N., Anwar M. A. & Maung A. W., Review of the Shaft Capacity Degradation of
Bored Piles Constructed with Bentonite Slurry, Civil and Environmental Engineering Conference,
Asian Institute of Technology, Bangkok, Thailand, 1999, p.p. V-59 to V-68.
[5] Hosoi T., Yagi N., & Enoki M., Consideration to the Skin Friction of Diaphragm Wall Foundation,
rd
3 Intl. Conf. on Deep Foundation Practice incorporating PILETALK, Singapore, 1994
[6] Reese L. C., Design and Construction of Drilled Shafts, 12th Terzaghi Lecture, Proc. ASCE, Vol.
104, No. GT1, 1978, p.p. 95-116

116
370

GEOTECHNICAL SPECIAL PUBLICATION NO. 116

DEEP FOUNDATIONS 2002

An International Perspective on Theory,
Design, Construction and Performance
Volume One
PROCEEDINGS OF THE INTERNATIONAL
DEEP FOUNDATIONS CONGRESS 2002

SPONSORED BY
The Geo-Institute of the American Society of Civil Engineers
February 14-16, 2002
Orlando, Florida
EDITED BY
Michael W. ONeill
Frank C. Townsend

117

Performance of Wet-Process Bored Piles Constructed with

Polymer-Based Slurry in Bangkok Subsoil
Narong Thasnanipan*, Zaw Zaw Aye**, Chanchai Submaneewong*** and
Wanchai Teparaksa, Ph.D. ****
*Managing Director, **Project Manager, ***Geotechnical Engineer, SEAFCO Co.,
Ltd., 26/10 Rarm Intra 109 Rd., Bangchan, Klong Sam Wah, Bangkok 10510, Thailand,
Tel. (662) 9190090, Fax. (662) 9190098, [email protected], www.seafco.co.th
****Professor, Department of Civil Engineering, Chulalongkorn University, Bangkok,
Thailand, fcewtp@eng/chula.ac.th
Abstract
Conventionally, medium to large bored piles deep-seated in the multi-layered soils of
Bangkok were constructed by rotary drilling under wet process or slurry replacement
method, using betonite slurry as support fluid in boring process. As an alternative to
bentonite slurry, polymer-based slurry has been increasingly used in Bangkok recently.
Polymer-based slurry having environmentally friendly properties, is found to be a good
solution in minimizing the problems associated with the disposal of contaminated soils
as well as the site pollution. This paper presents the performance of medium to large
diameter bored piles constructed with polymer-based slurry in Bangkok. The static pile
load test results of polymer-based slurry bored piles are compared with the predicted
values using bentonite piles parameters. The measured frictional resistance of pile shaft
from instrumented pile load tests are also reported. Shaft resistance of bored piles
constructed with polymer-based slurry is found to be better than that of bored piles
constructed with bentonite slurry.
Introduction
Medium to large diameter bored piles with length ranging from 24 m to 60 m have been
used as foundations of the various structures in Thailand for over 3 decades. Rotary
drilling technique coupled with wet process method using bentonite slurry as support
fluid is commonly applied to construct these piles. As bentonite slurry had proven
acceptable to make qualified bored piles in the past, most of them were constructed only
by using this mineral based slurry. Recently, polymer-based slurry has been
increasingly used in Bangkok as an alternative supporting fluid in bored pile drilling. In
the initial stage, the primary reason for using this alternative slurry was to minimize the
problems associated with using classic bentonite such as site pollution and the disposal
of excavated soils which are usually contaminated. Construction records and static pile
load test conducted on bored piles constructed with polymer-based slurry proved that
there is no adverse effect on stability of borehole during drilling as well as capacity of
the pile from using this alternative supporting fluid in Bangkok subsoil. Dry anionic
partially hydrolyzed polyacrylamide (PHPA) polymer powder premixed with fresh
water and a small percentage of bentonite was used in the construction of bored piles
presented in this paper.
143
119

DEEP FOUNDATIONS 2002

144

Pore Pressure (kN/m 2)

Geotechnical
Parameters
0

Bangkok Soft
Clay

200

400

600

Weathered Crust

Su= 10-25 KPa

t = 14-16kN/m 3
-

10

20

Medium to
Stiff Clay

30

Medium
Dense Sand
(First Sand)

40

Hard Clay

50

Dense to Very
Dense Sand
(Second Sand)

Su= 50-14020- KPa

t = 17-21kN/m 3
-

30
SPT-N= 20-40
t = 20kN/m 3

Su> 200 KPa

t = 21KN/m 3

SPT-N= 50-100
t = 21KN/m 3

60

40

50

Line
tatic
ros
ine
Hyd
nL
dow
raw
cD
etri
zom
Pie
ual
Act

Depth (m)

10

60

Figure 1. Typical soil profile of Bangkok plain with piezometric drawdown conditions.
Subsoil Profile and Existing Pore Pressure Condition
Subsoil profile is relatively consistent at different localities in Bangkok. A typical
subsoil profile is characterized by the alternating layers of clay and sand deposits as
shown in Figure 1 and is briefly described below.
Weathered crust of 2 m thick is commonly found as the top layer. In urban areas
of Bangkok, this layer is covered by filled material. Soft to very soft, highly
compressible dark gray marine clay lies beneath weathered crust. Depending on the
location, this layer extends up to 12-18 m. About 2 m thick medium clay layer can be
observed between soft clay and underlying stiff clay. Generally stiff Clay layer occurs
directly underneath medium clay and its depth goes up to 22 m. Below stiff clay layer,
first sand layer 5-8 m in thickness can be found. This first sand layer, however, is
absent in some areas. Stiff to hard clay layer underlies first sand and it is found to be
about 5 m thick. Second sand layer generally occurs at depths between 45 to 65 m.
Existing pore water pressure conditions in upper part of Bangkok soft clay are
hydrostatic from nearly 1m below the ground level. Then the hydrostatic condition
changed to piezometric drawdown near bottom level of Bangkok soft clay as shown in
Figure 1. Excessive ground water abstraction from the deep aquifers in the past is
responsible for this significant drawdown of piezometric pressure (Nutalaya, 1981,
Thasnanipan et al., 1998).
Wet-Process Bored Pile Construction Method
In Bangkok, medium to large diameter (0.6 m to 1.8 m diameter) bored piles with depth
ranging from 24 m to over 60 m are constructed by wet process due to the prevailing
subsoil condition and limited application of dry process. Wet process method as its

120

DEEP FOUNDATIONS 2002

145

name implies makes the pile under wet condition by using drilling slurry. Temporary
casing of 14-15 m in length is also used as a support in soft clay layer. Drilling is
commonly commenced by dry process with auger applying rotary drilling action in the
soft clay layer and stiff clay layer. Before reaching to the first sand layer slurry is fed to
the borehole and drilling is continued with the bucket to the final depth. A cleaning
bucket is commonly used to clean the bottom of the borehole prior to reinforcement
cage installation. Tremie concreting is necessary for the piles installed under wet
process.
Polymer-Based Drilling Slurry
The primary function of drilling slurry is to prevent the borehole instability as deepseated bored piles are required to install through water bearing and caving soils.
Bentonite slurry was mainly used in the past as it adequately fulfilled the requirement.
Polymer-based slurry has become increasingly popular as an alternative supporting fluid
in recent years. The main reason for switching to use polymer-based slurry is to
minimize the problems associated with environmental issues caused by the use of
bentonite slurry.
In the initial stage of using polymer-based slurry, the major concerns were the
potential negative impact on stability of the borehole and the shaft friction capacity as
this alternative slurry was not yet accustomed to the bored piling industry in Bangkok.
Trial boreholes conducted in the initial stage of introducing polymer-based slurry
revealed that borehole is well stablized if the right dosage is used. Dry anionic partially
hydrolyzed polyacrylamide (PHPA) polymer powder premixed with fresh water and a
small percentage of bentonite is mainly used for long bored piles in Bangkok. First, a
small dosage of bentonite powder is premixed with fresh water prior to polymer powder
introduction. The predetermined dosage of dry polymer is then added and mixed with
the betonite slurry. Premixed polymer-based slurry is stored in the steel tanks before
supplying to the bored pile drilling. The reason for adding bentonite for long bored piles
is to minimize the fluid loss which was observed from the trial boreholes drilled only
with pure polymer slurry. Depending on the thickness of sand layers present at site and
the length of pile, addition of bentonite may not be necessary. Close monitoring was
made by sonic caliper measurement (KODEN drilling monitoring equipment) to check
the stability of the borehole. The results of the sonic caliper measurement proved that
borehole can be well stabilized with the use of polymer-based slurry at total
construction time prior to concreting even over 24 hours. According to the cross-hole
sonic logging or sonic coring tests conducted on bored piles, there is no evidence that
degraded concrete remains along the pile shaft due to the use of polymer-base slurry.
Load tests were carried out to confirm the capacity of bored piles constructed
with the alternative drilling slurry. As most of the test piles constructed with polymerbased slurry gave smaller settlement values, particularly at maximum test load than
those obtained from piles constructed with bentonite slurry, confidence in using
polymer-based slurry was raised among the designers and the contractors.
Review of Shaft Friction Capacities Affected by Drilling Slurry
Shaft capacity degradation of wet-process bored piles may be attributed to three main
factors : reduction of effective stress at borehole interface, properties of slurry and

121

DEEP FOUNDATIONS 2002

146

excessive filter cake formation. Reduction of effective stress at borehole interface

caused by installation process of wet-process bored pile is theoretically explicable and
understandable but to the authors knowledge, detailed study on the issue has not been
well documented.
Properties of bentonite influence on the shaft capacity degradation has been
studied by many researchers. Thasnanipan et al. (1998) concluded from the case study
conducted on 11 bored piles in Bangkok subsoil that viscosity of bentonite slurry not
more than 55 sec/qt (Marsh Funnel Viscosity) does not have significant effect on the
shaft load transfer.
A number of researchers reported the reduction of shaft capacity caused by
excessive formation of the filter cake on borehole wall of bored piles constructed with
bentonite slurry. It is mainly contributed by long construction time without slurry
agitation as cited by them.
Few researches are available for the effect of polymer-based slurry on shaft
friction capacity of bored pile. Ata and ONeill (1998) concluded that unit side shear
resistances in both clay and sand layers of polymer test piles are higher than those of
bentonite piles. Bustamante and Gianeselli (1998) carried out the comprehensive study
on the performance of polymer bored piles. The authors reported that no reduction of
shaft friction was observed from load test results of polymer bored piles in comparison
with bentonite bored piles.
Littlechild and Plumbridge (1998) concluded from test bored piles conducted in
Bangkok, that the shaft friction resistance of bored piles constructed under polymer
slurry shows no significance change with construction times up to 42 hours. Based on
the test results of bored piles they also reported a slight reduction in shaft resistance
with decreasing polymer viscosity.
Design Parameters for Bored Piles
Following empirical formula are normally adopted to estimate the ultimate skin friction
capacity of bored piles.
layers.
where,

Total stress approach is used to calculate the shaft friction capacity of clay
fs

Cu

= .Cu
= adhesion factor
= undrained shear strength

Undrained shear strength, Cu is usually obtained directly from laboratory test

results for soft clay layer and indirectly estimated from SPT N values for deeper stiff
to hard clay layers.
Effective stress approach is applied to calculate the shaft friction capacity of
sand layers.
fs
= .v
where,
v
= effective overburden pressure

= shaft friction factor = Ks. tan

Ks
= coefficient of horizontal earth pressure

= coefficient derived from internal friction angle of sand layer

122

DEEP FOUNDATIONS 2002

147

Effective overburden pressure, v at existing piezometric drawn down condition

is usually used to calculate shaft friction capacity of sand layers. The design lines of
and of the previous research data which were developed from the instrumented load
test results of bored piles constructed with bentonite slurry are commonly used in
estimation of pile capacity. Hence, it is important to justify whether these design lines
are applicable for bored piles constructed with polymer- based slurry. Assessment is
made in the later section of this paper to justify this issue.
Load Test on Polymer-Based Slurry Bored Piles
Static pile load test on eleven test piles constructed by wet process with
polymer-based slurry are presented in this paper. Out of these, three piles were
instrumented with vibrating wire strain gauges and mechanical extensometers at
interface of different soil layers. All test piles except TP-8 were tested to minimum 2.5
times of pre-defined design load, which is commonly estimated in the preliminary
design stage.
Slurry Properties and Construction Records of Test Piles. Table 1 shows the
properties of polymer-based slurry and construction records of test piles. It is to be
noted that all these test piles were constructed by the same contractor. Hence it is
assumed that influence of other variables such as equipment as well as construction
procedure used in drilling, slurry mixing and quality control measures applied for all
piles would be minimized in evaluation of the load test results.
Table 1. Technical data of polymer-based slurry used for test piles
Pile
ID

Mixed Ratio
per m3 of water

Properties of Slurry
Prior to Concreting
Total Time Prior
to Concreting
Dia. x Depth Polymer Bentonite
Sand
(Hrs : Mins)
Density Viscosity*
(m)
(kg)
(kg)
pH Content**
(g/ml)
(sec/qt)
(%)
Pile
Dimension

TP-1

0.80 x 49

0.50

12

17 : 20

1.06

40

TP-2

0.80 x 50

0.50

12

12 : 20

1.03

45

0.5

TP-3

1.20 x 51

0.50

10

14 : 00

1.01

44

TP-4

1.80 x 62

0.50

10

23 : 00

1.04

51

TP-5

1.50 x 54

0.50

12

15 : 30

1.04

44

0.5

TP-6

1.00 x 51

0.22

7 : 00

1.02

36

TP-7

1.00 x 30

0.50

3 : 45

1.03

45

TP-8

0.80 x 41

0.50

10

6 : 55

1.01

44

0.1

TP-9

0.80 x 43

0.50

12

8 : 00

1.03

45

0.5

TP-10

0.60 x 26

0.50

2 : 30

1.02

43

TP-11

0.60 x 24

0.70

0 : 45

1.02

42

Note : *Marsh Funnel Viscosity, **API RP 13-B

123

DEEP FOUNDATIONS 2002

148

Prediction vs Performance of Test Piles. As none of the test piles except TP-5 was
tested to failure load, no attempt is made to determine the capacity of the piles from the
failure load criteria for further assessment. Instead the performance of test piles is
evaluated from load-settlement characteristics as described below.
As large quantities of tested-to-failure instrumented load test results of bored
piles constructed with bentonite slurry in Bangkok are available, load-settlement
behavior of this pile type can be accurately predicted using back-analyzed data. Loadsettlement behavior of all test piles constructed with polymer-based slurry presented in
this paper were predicted by taking the parameters obtained from back-analyzed data of
bentonite bored piles.
Predicted load-settlement curves are compared with those of actual test results as
shown in Figure 2 to 12. It is to be noted that the term gross settlement indicated in
the graphs represents the total pile head movement at applied load. The objective of
these comparisons is to assess from the actual load test results, whether or not use of
polymer-based slurry in bored pile construction affects the capacity of bored piles.
The summary of predicted and measured load-settlement data is presented in
Table 2. At the design load, measured pile head movement of most of the piles are
within 5 mm as can be seen in the table. It is also evident that movement measured from
the field load test is smaller than that of prediction for most of the test piles, and is more
clear at the applied load ranging from 1.5 to 3 times of pre-defined design load. This
finding pointed out that better capacity of bored pile constructed in Bangkok can be
achieved with use of polymer-based slurry. The pattern of the load-settlement curves
obtained from the filed tests suggested that majority of the applied load is resisted by
the shaft friction. Therefore, it is reasonable to conclude that overall shaft frictional
resistance of bored pile is improved with the use of polymer-based slurry.
SPT-N (Blows/ft)
1

10

Soft to
Medium
Clay

10

20

Stiff to Very
Stiff Clay

20

30
40
50
60

Medium to
Very Dense
Sand
Hard clay
Medium to
Very Dense
Sand
Very Dense
Sand

50

100

30
40
50

Gross Settlement (mm)

Depth (m.)

2000

Applied Load (kN)

4000

6000

8000

10000 12000

0
-5
-10
-15
-20
-25

Measurement
Prediction

-30

60

Figure 2. Load-Settlement curves of TP-1 (Dia. 0.80 x 49 m) : Prediction vs. Measurement.

124

DEEP FOUNDATIONS 2002

149

SPT-N (Blows/ft)
1

0
Soft Clay

20

30

40

50

Applied Load (kN)

100

2000

4000

6000

8000

10000

12000

0
10

Loose Sand
Stiff to Very
Stiff Clay

20
30

Dense to Very
Dense Sand

40

Hard Clay
Dense to
Very
Dense Sand

-10

Gross Settlement (mm)

Depth (m.)

10

50

-20
-30
-40

Measurement
Prediction

-50

50

-60

60

60

Figure 3. Load-Settlement curves of TP-2 (Dia. 0.80 x 50 m) : Prediction vs. Measurement.

SPT -N (Blows /ft)

1
0

50

Applied Load (kN)

100

Soft Clay
10

Gross Settlement (mm)

10

Depth (m.)

Stiff Clay
20

30

Hard Clay
Dense Sand

20
30

Hard Clay
40

40

50

Dense to
Very Dense
Sand

4000

8000

12000

16000

20000

50

-10
-20
Measurement
Prediction

-30
-40

60

60

Figure 4. Load Settlement curves of TP-3 (Dia. 1.20 x 51 m) : Prediction vs. Measurement.
STP-N (Blows/ft)
0

50

Applied Load (kN)

100

Depth (m.)

20
30

Stiff Clay
Dense
Sand

40

Hard Clay

50

Very
Dense
Sand

60
70

Gross Settlement (mm)

10

0
Soft to
Med. Clay

5000

10000

15000

20000

25000

30000

35000

-5

-15

Measurement
Prediction
-25

-35

Hard Clay

Figure 5. Load-Settlement curves of TP-4 (Dia. 1.80x62 m) : Prediction vs. Measurement.

125

DEEP FOUNDATIONS 2002

150

STP-N (Blows/ft)
0
0

Depth(m.)

20

Very Soft to
Soft clay
Stiff to Very
Stiff Clay

30
40

50
60

Applied Load (kN)

100

10

20

30

V. Dense Sand
Hard Clay
Medium
Dense to
Very Dense
Sand
Hard clay

5000

10000 15000 20000 25000 30000 35000

Gross Settlement (mm)

10

50

40

-25
-50
-75

Measurement
Prediction

-100
-125
-150

50

60

Figure 6. Load-Settlement curves of TP-5 (Dia. 1.50x54 m) : Prediction vs. Measurement.

SPT-N (Blow/ft)
0

50

Applied Load (kN)

100

Soft Clay

30
40

Gross Settlement (mm)

Depth (m.)

20

Stiff Clay
Dense Sand
V. Stiff Clay
Dense Sand

50
60

1500 3000 4500 6000 7500 9000 10500 12000

10

-5
-10
-15
-20

Measurement
Prediction

-25
-30
-35

Very Dense
Sand

70

Figure 7. Load-Settlement curves of TP-6 (Dia. 1.00x51 m) : Prediction vs. Measurement.

SPT-N (Blows/ft)
0

50

100

Applied Load (kN)

Depth ( m.)

10
15
20
25
30

Soft Clay
Soft to
Med. Clay
Stiff Clay
Hard Clay
Very
Dense Sand

1000

2000

3000

4000

5000

6000

7000

Gross Settlement (mm)

-5
-10
-15

Measurement
Prediction

-20
-25
-30
-35

35

Figure 8. Load-Settlement curves of TP-7 (Dia. 1.00x30 m) : Prediction vs. Measurement.

126

151

DEEP FOUNDATIONS 2002

STP-N (Blows/ft)
0

50

Applied Load (kN)

100

Depth (m.)

20
30
40

2000

4000

6000

8000

10000

12000

Gross Settlement (mm)

10

Soft - Med.
Stiff Clay
Medium to
Dense Sand
Stiff Clay
Dense to
Very
Dense Sand

50

-20
-40
Measurement
Prediction

-60
-80

-100

60

Figure 9. Load-Settlement curves of TP-8 (Dia. 0.80x41 m) : Prediction vs. Measurement.

STP-N (Blows/ft)
0
0
Soft to
Medium Clay

20

Stiff to Very
Stiff Clay

30

Medium to
Dense Sand

40
50

Applied Load (kN)

100

Hard Clay
Dense to
V. Dense Sand

Gross Settlement (mm)

Depth (m.)

10

50

10
15
20
25
30
35
40
45

1000

2000

3000

4000

5000

6000

7000

0
-5
-10

Measurement
Prediction

-15
-20

50

Figure 10. Load-Settlement Curves of TP-9 (Dia. 0.80x43 m) : Prediction vs. Measurement
SPT-N (Blows/ft)
0

50

100

Applied Load (kN)

0
Very Soft
to Soft Clay

Depth (m.)

10
15

Stiff to Very
Stiff Clay

20
25

Dense to Very
Dense Sand

Gross Settlement (mm)

500

1000

1500

2000

2500

3000

0
-2
-4
-6

Measurement
Prediction

-8

-10
-12

30
35

Figure 11. Load-Settlement curves of TP-10 (Dia. 0.60x26 m) : Prediction vs. Measurement

127

DEEP FOUNDATIONS 2002

152

SPT-N (Blow/ft)
0

50

100

Depth (m.)

10
15
20

Soft Clay

Gross Settlement (mm)

Medium Clay
Stiff Clay

25
30
35

Dense Sand
Hard Clay

500

Applied Load (kN)

1000

1500

2000

2500

0
-5
-10

Measurement
Prediction

-15
-20

40

Figure 12. Load-Settlement curves of TP-11 (Dia. 0.60x24 m) : Prediction vs. Measurement

Table 2. Summary of predicted and measured gross settlement

Pile
Dimension
Dia. x Depth
(m)

Design
Load,
DL
(kN)

Maximum
Test Load
(kN)

TP-1

0.80 x 49

3300

TP-2

0.80 x 50

TP-3

Pile ID

Gross Settlement at Applied Load (mm)

Prediction

Measured

DL

2xDL

2.5xDL

DL

2xDL

2.5xDL

9900

5.50

12.70

17.50

3.90

9.80

14.95

3300

9900

6.50

16.52

24.20

5.20

14.74

21.05

1.20 x 51

7200

18000

6.62

17.50

29.80

4.61

14.19

19.76

TP-4

1.80 x 62

11000

33000

5.62

14.07

20.55

4.00

9.88

12.96

TP-5

1.50 x 54

12000

32000

9.31

26.29

127.10

8.60

12.79

23.03

TP-6

1.00 x 51

4500

11250

5.50

13.91

19.80

2.49

7.10

9.15

TP-7

1.00 x 30

2600

6500

3.50

10.65

20.80

1.67

3.96

5.38

TP-8

0.80 x 41

3300

9900

5.72

15.90

30.20

2.75

8.05

11.60

TP-9

0.80 x 43

3300

6600

6.20

15.10

4.52

11.20

TP-10

0.60 x 26

1000

2500

2.56

6.35

9.50

1.61

3.76

4.55

TP-11

0.60 x 24

800

2000

2.59

7.84

14.30

1.09

3.19

4.22

Instrumented Pile Load Test. Shaft frictional resistance of polymer-based bored piles
in sand layers were of major interest to the authors as it is considered that drilling slurry
has less influence on clay layers according to the previous studies on bentonite bored
piles.
Three instrumented pile load tests were available at the time of this paper
preparation. Vibrating wire strain gauges (VWSG) and mechanical extensometers (ME)
were installed at the major soil boundaries for all 3 piles. It is to be noted that the lowest
level of strain gauges and extensometer of test piles TP-1, TP-2 and TP-3 are located at
1.5, 1.0 and 1.0 m above the relevant pile tip level respectively.

128

153

DEEP FOUNDATIONS 2002

Load distribution curves along the test pile shaft at various applied load are
shown in Figure 13 to 15. The values of unit skin friction developed by the maximum
test load at the different soil layers along the shaft of test piles in comparison with those
of ultimate unit skin friction calculated using the formula shown in the earlier section
are also illustrated in these figures.
Load (kN)

SPT-N (Blows/ft)
1

50

100

4000

8000

Unit Skin Friction (kN/m2)

12000

0.0

10

10.0

10

20

20.0

20

Medium Dense
to V.Dense
Sand

30

30.0

30

Hard Clay

40

40.0

40

50

50.0

50

60

60.0

60

50

100

150

Soft to

10

Depth (m.)

20

30

40

Medium Clay
Stiff to
Very Stiff
Clay

Medium to
V.Dense Sand

50
Very Dense Sand
60

Mobilized at 9900 kN
Qult (calculated)

Figure 13. Load distribution curves and unit skin friction of TP-1 (Dia. 0.80x49 m).

SPT-N (Blows/ft)
1
0

Depth (m.)

20

30

40

50

100

4000

8000

Unit Skin Friction (kN/m2)

0

12000

Loose Sand

10

10

10

Stiff to Very
Stiff Clay

20

20

20

Dense to Very
Dense Sand

30

30

30

Hard Clay

40

40

40

50

50

50

60

60

Soft clay

10

Load (kN)

50

100

Dense to Very

50

60

Dense Sand

60

Mobilized at 9900 kN
Qult (calculated)

Figure 14. Load distribution curves and unit skin friction of TP-2 (Dia. 0.80x50 m).

129

150

DEEP FOUNDATIONS 2002

154

0
Soft Clay
10

50

100

Unit Skin Friction (kN/m2)

Load (kN)

SPT-N (Blows/ft)

6000

12000

18000

10

10

10

20

20

20

30

30

30

40

40

40

50

50

50

60

60

60

50

100

150

200

250

Depth (m.)

Stiff Clay
20

30

V. Stiff to
Hard clay
Dense Sand
Hard Clay

40

Dense to
Very Dense
Sand

50

60

Mobilized at 18000 kN
Qult (calculated)

Figure 15. Load distribution curves and unit skin friction of TP-3 (Dia. 1.20x51 m).
A summarized comparison between measured total shaft friction resistance at
maximum test load and that of theoretically calculated values of 3 test piles is presented
in Table 3.
Table 3. Calculated and measured shaft resistance
Pile
ID

Pile
Dimension
Dia. x Depth
(m)

Design
Load
(kN)

Maximum
Test Load
(kN)

TP-1

0.80 x 49

3300

TP-2

0.80 x 50

TP-3

1.20 x 51

Total Shaft Resistance of Pile (kN)

Measured /
Calculated

Calculated

Measured at
Maximum Test Load

9900

6900

9710

1.41

3300

9900

6200

9670

1.56

7200

18000

10820

17060

1.58

Displacements required to mobilize ultimate shaft friction of piles are 0.5 - 2 %

of the shaft diameter (Reese, 1978). As the total pile head displacements of TP-1, TP-2
and TP-3 are 2.8, 3.9 and 1.6 % of the shaft diameter respectively, it is likely that the
shaft friction of these piles were fully mobilized at relevant maximum test load.
However, the pile displacement value required to reach peak shaft resistance depends
upon many conditions of the pile-soil interface and the stress strain behavior of
surrounding soils : soil characteristics, stresses of the ground surrounding the pile shaft,
construction methods of bored piles, time from construction to a loading test, loading
procedure and speed, etc. (Hirayama, 1990). Moreover, shaft friction development
against pile displacement is also influenced by the pile length. According to
Thasnanipan et al. (1998), based on the various instrumented static pile load test
conducted on long bored piles in multi-layered soil of Bangkok, shaft frictions of deeper
soil layers were not fully mobilized though the piles were declared to have failed due to
the excessive pile head movement. This finding may also be valid for bored piles
constructed with polymer-based slurry. Figure 16 illustrates the shaft friction versus pile
head movement of different soil layers obtained from test pile TP-1. As can be seen in
the figure, at the maximum test load where pile head movement reached 2.8 % of the

130

DEEP FOUNDATIONS 2002

155

shaft diameter, the shaft frictions of upper soil layers (soft clay and stiff to very stiff
clay) were fully mobilized and it is likely that medium dense sand (first sand) layer
almost reached to ultimate shaft friction resistance. However, it is most likely that very
dense sand (second sand) was not fully developed to the ultimate stage. The same
behavior is observed for test piles TP-2 and TP-3.
Shaft friction factors of the first and second sand layer of TP-1, TP-2 and TP-3
are back calculated from the load transfer values of relevant layers obtained from the
VWSG data. However, the back-calculated factors (named as pseudo- for purpose of
convenient reference in this paper) particularly for the second sand layer are the values
at applied maximum test loads and not at ultimate test loads. In other words, it is likely
that the back-calculated factors for the second sand layer are the values at partially
mobilized shaft friction of this layer. The back-calculation of value is mainly aimed to
get an idea on the magnitude of factors for polymer-based bored piles from the
currently available data in comparison with the design values commonly used for
bentonite bored piles constructed in Bangkok.
Figure 17 shows the values of pseudo- of polymer-based slurry piles obtained
from 3 instrumented pile load tests plotted on the graph of values derived from the
full-scale static pile load tests of bentonite bored piles in Bangkok subsoil from the
previous research works (Submaneewong, 1999). Submaneewong (1999) recommended
the design line of factors for bentonite bored piles in the first and second sand as
indicated in Figure 17. It is evident that pseudo- for the first sand layer of all 3 test
piles constructed under polymer-based slurry are significantly higher than those of
design values. The pseudo- values of the second sand layer are also slightly higher
than those of design line. It is most likely that values of second sand layers would be
higher than pseudo- values indicated in Figure 17 if the piles were tested to higher
loads or to failure, since shaft friction at this layer might not have been fully mobilized
at applied maximum test load as described in above paragraphs. These findings
suggested that polymer-based slurry bored piles produced better shaft friction resistance
in sand layers than that of bentonite bored piles. Further research is needed to determine
the values of the first and second sand layers at fully mobilized shaft friction
resistance.

Unit Skin Friction (kN/m 2)

200
Soft clay
Stiff to very stiff clay
Medium dense sand
Very dense sand

Stiff to very stiff clay

150

Medium dense sand

(First sand)

100
Very dense sand
(Second sand)

50

Soft clay

0
0

12

16

20

24

Pile Head Movem ent (m m )

Figure 16. Unit skin friction versus pile head movement of different soil layers (TP-1).

131

DEEP FOUNDATIONS 2002

156

1 .0

(Ks tan )

Submaneewong (1999) : Bored Piles (Tip in 2nd Sand)

Submaneewong (1999) : Bored Piles (Tip in 2nd Hard Clay)

0 .8

Chiewcharnsilp (1988)

0 .6

Polymer Bored Pile TP-1 (2nd Sand)

Polymer Bored Pile TP-2 (2nd Sand)
Polymer Bored Pile TP-3 (2nd Sand)

Polymer Bored Pile TP-1 (1st Sand)

Polymer Bored Pile TP-2 (1st Sand)
Polymer Bored Pile TP-3 (1st Sand)

Submaneewong
( 1999 )

0 .4

0 .2

0 .0
28 .0

30 .0

32 .0

34 .0

36 .0

38 .0

Angle of Internal Friction, '

Figure 17. Back-calculated values (pseudo- factor) of polymer bored piles at
maximum test load plotted on design line of bentonite bored piles constructed in
Bangkok subsoil .
Conclusions
(1) Performance of bored piles constructed with polymer-based slurry in Bangkok
subsoil have been presented using the results of eleven static pile load tests.
(2) The measurement of borehole profile by sonic caliper and construction records
proved that there is no adverse effect of using polymer-based slurry in Bangkok
subsoil as far as the stability of borehole is concerned.
(3) Pile head settlements at applied loads obtained from the field load test results of all
test piles constructed with polymer-based slurry are smaller than those of prediction
using bentonite piles parameters.
(4) Overall shaft resistance of bored piles constructed with polymer-based slurry is
higher than that of theoretically calculated values in the order of 1.5.
(5) The back-calculated factors of current polymer-based slurry test bored piles are
higher than those of design values commonly used in Bangkok for bentonite bored
piles.
(6) Above findings suggested that higher shaft friction resistance can be obtained if pile
is constructed with polymer-based slurry.
(7) Further research is needed to establish the design parameters for bored piles
constructed with polymer-based slurry in Bangkok subsoil.
Acknowledgement
The authors wish to express their appreciation to the project directors and engineers of
SEAFCO Co., Ltd. for providing the bored pile construction records. Thanks are also

132

DEEP FOUNDATIONS 2002

157

extended to EDE Co., Ltd. for providing the load test data. The authors also wish to
thank Miss. Siriphen Chaiyapak (former office engineer, SEAFCO Co., Ltd.) and Mr.
Artorn Kantawateera (trainee student, Khonkaen University) for their assistance in
preparing this paper.
References
Ata A. A. and ONeill M. (1998). Side-wall stability and side-shear resistance in
bored piles constructed with high-molecular-weight polymer slurry, Third
International Geotechnical Seminar on Deep Foundation on Bored and Auger Piles,
Ghent, Belgium, 19-21 October 1998, Balkema, Rotterdam, 111-117.
Bustamante M. and Gianeselli L. (1998). Performance of polymer slurries in large
diameter bored pile, Third International Geotechnical Seminar on Deep Foundation on
Bored and Auger Piles, Ghent, Belgium, 19-21 October 1998, Balkema, Rotterdam
119-127.
Harayama H. (1990). Load settlement analysis for bored piles using hyperbolic transfer
functions, Soil and Foundations, Vol. 30, No. 1, Japanese Society of Soil Mechanics
and Foundation Engineering, 55-64.
Littlechild B. and Plumbridge G. (1998). Effects of construction technique on the
behavior of plain bored piles constructed under drilling slurry, Seventh International
Conference and Exhibition on Piling and Deep Foundations, Deep Foundation Institute,
Vienna, Austria, 15-17 June 1998, 1.6.1-1.6.8.
Nutalaya, P. (1981). Subsidence Studies of the Bangkok Plain, AIT Research Report,
Asian Institute of Technology, Bangkok, Thailand.
Reese L. C., (1978). Design and Construction of Drilled Shafts, Twelfth Terzaghi
Lecture, Proc. ASCE, Vol. 104, No. GT1, 1978, 95-116.
Submaneewong C. (1999). Behavior of instrumented barrette and bored piles in
Bangkok subsoils, (in Thai), M.Eng. Thesis, Department of Civil Engineering, Faculty
of Engineering Chulalongkorn University, Bangkok, Thailand.
Thasnanipan N., Ganeshan B. and Anwar M.A. (1998). Effect of construction time
and bentonite viscosity on shaft capacity of bored piles, Third International
Geotechnical Seminar on Deep Foundation on Bored and Auger Piles, Ghent, Belgium,
19-21 October 1998, Balkema, Rotterdam, 171-177.
Thasnanipan N., Tanseng P. and Anwar M. A. (1998). Large diameter bored piles in
multi-layered soils of Bangkok, Third International Geotechnical Seminar on Deep
Foundation on Bored and Auger Piles, Ghent, Belgium, 19-21 October 1998, Balkema,
Rotterdam, 511-518.

133

GEOTECHNICAL SPECIAL PUBLICATION NO. 116

DEEP FOUNDATIONS 2002

An International Perspective on Theory,
Design, Construction and Performance
Volume Two
PROCEEDINGS OF THE INTERNATIONAL
DEEP FOUNDATIONS CONGRESS 2002

SPONSORED BY
The Geo-Institute of the American Society of Civil Engineers
February 14-16, 2002
Orlando, Florida
EDITED BY
Michael W. ONeill
Frank C. Townsend

135

Barrette of Over 50,000 kN Ultimate Capacity

Constructed in the Multi-Layered Soil of Bangkok
Narong Thasnanipan*, Zaw Zaw Aye** and Wanchai Teparaksa, Ph. D.***
*Managing Director, **Project Manager, SEAFCO Co., Ltd., 26/10 Rarm Intra 109
Rd., Bangchan, Klong Sam Wah, Bangkok 10510, Thailand, Tel. (662) 9190090, Fax.
(662) 9190098, [email protected], www.seafco.co.th
***Professor, Department of Civil Engineering, Chulalongkorn University, Bangkok,
Thailand, fcewtp@eng/chula.ac.th
Abstract
A static load test carried out on barrette having cross sectional size of 1.5 m x 3.0 m
seated at 57 m below ground level is presented in this paper. Over 50,000 kN load was
applied to evaluate the performance of this fully instrumented barrette. The load applied
on barrette with a reaction frame tower capacity of 60,000 kN was claimed to be one
of the highest test loads in the region. The test results were compared with those from
the instrumented load test on bored pile of diameter 1.5 m with the same length, being
located 30 m away. Since bored pile and barrette were constructed by different method
and the parameters such as construction time, pile shape, slurry properties etc, were
considerably different, results from the load tests have provided a unique opportunity to
assess the extent of difference in behavior.
Introduction
In the area of foundation engineering, Bangkok, the capital of Thailand, is well known
for its weak soils. As such, most of the structures are supported by piled foundations.
Construction of heavy structures such as sky scrapers, elevated expressways, overpass
bridges, underground car parks and subway stations commonly require the high
capacity foundations. Bored cast-in-place piles are mainly used to fulfill this
requirement. However, in some projects bored piles are not feasible due to the site
constraints, applicable construction method and/or extensive bearing capacity
requirements. In such cases, the use of barrette foundations would make a suitable
alternative. Barrettes with dimension ranging from 0.80 m x 2.7 m to 1.5 m x 3.0 m for
safe working load capacity from 11,000 to 23,000 kN have been used in some major
projects. To assess the performance of large-capacity foundation elements in layered
subsoil of Bangkok, instrumented load testing is compulsory. This paper presents the
static load test results of an instrumented barrette tested up to 52,900 kN for the
foundation of a fifty-storey building in Bangkok. The test results, particularly load
transfer characteristics, and shaft friction capacity were compared with those of bored
pile diameter 1.50 m with the same length located 30 m away.
1073

137

1074

DEEP FOUNDATIONS 2002

Project Summary
The project is located in the prime area of Bangkok on Rama IX road. The design of
fifty-storey office complex called for the high-capacity piled foundation : 560 bored
piles of 1.2 m and 1.5 m diameter and 24 barrettes having cross section size of 1.5 m x
3.0 m. Bored piles and barrettes were seated at an approximate depth of 57 m in the
sand layer. Barrettes were designed to support the large load for towers elevator shafts
as bored piles were not feasible in this case. Base grouting was applied for barrettes and
bored piles at the locations of high column load mainly in the central tower area.
Instrumented static pile load tests were carried out for one barrette and one 1.5 m
diameter bored pile. The design safe working loads for the base grouted barrettes and
bored piles are 23,000 kN and 11,750 kN respectively. The barrette load test was
conducted on July 8, 1998 with maximum applied load of 52,900 kN.
Subsoil Profile and Ground Water Condition
Similar to other localities in Bangkok a typical subsoil profile at the site is characterized
by the alternating layers of clay and sand deposits as shown in Figure 1. Undrained
shear strength (Su) from unconfined compression tests and standard penetration tests
(SPT) are indicated with depth in Figure 1. It is to be noted that unconfined
compression tests were conducted only for soft clay layer whereas SPT tests were
carried out for all the layers below soft clay. Soil investigation from five boreholes at
different location reveals that subsoil layers along the site are relatively consistent. The
soil layers present at site are briefly described below.
S u ( U C ) kN/m2
DEPTH (m)
0

50

Pore Pressure (kN/m 2 )

100

0
Weathered
Crust

S oft Clay

1
0

-10

Stiff Clay

-30

Medium to
Dense Sand

400

600

Ac

-20

200

Hy
d ro
s ta
tic
L in

ra w
ic D

dow
in
nL

5
0

e tr

Dense Sand
H ard C lay

4
0

zom

-50

ie
lP

Hard
Clay

tu a

-40

Dens e to Very Dense Sand 6

-60
0

50

100

S P T - N ( b l o w s / 3 0 cm)

Figure 1. Soil profile at site and the typical ground water condition of Bangkok

138

1075

DEEP FOUNDATIONS 2002

Soft, highly compressible dark gray marine clay lies beneath a two meter thick
weathered crust and extends up to 13.5 m. A stiff clay layer occurs directly underneath
soft clay and its depth goes up to 26 m. Below the stiff clay layer, 10 m thick medium to
dense sand layer also known as the first sand layer can be found. Hard clay underlies
the first sand and it is found to be about 16 m thick. Dense to very dense sand (the
second sand layer) occurs at depths between 51 to 72 m. Within the second sand layer, a
thin layer of hard clay is found.
Figure 1 depicts the present typical piezometric profile of Bangkok. As can be
seen in the figure, existing pore water pressure conditions in upper part of Bangkok soft
clay are hydrostatic from nearly 1m below the ground level and then the hydrostatic
condition changed to piezometric drawdown below the bottom level of Bangkok soft
clay. Excessive ground water abstraction from the deep aquifers in the past is
responsible for this significant drawdown of piezometric pressure (Nutalaya, 1981,
Thasnanipan et al., 1998).
Installation Method and Process
Both the barrette and bored pile were constructed by wet process under bentonite slurry.
Bentonite slurry conformed to the locally accepted specification (AASHTO 1992) was
used. The actual properties of the bentonite slurry used are given in Table 1 along with
those specified by the project. Single stage base grouting was applied 24 hours after
concreting for both barrette and bored pile. The detailed procedures utilized in the
construction of two different foundation structures are outlined below.
Table 1. Comparison of bentonite slurry properties
Project
Properties Specified
Values
Density
(g/cc)
Marsh
Funnel
Viscosity
(sec/qt)
API Sand
Content
(%)
pH value

Measured Actual Properties

Barrette
Bored Pile
Before feeding After Recycling Before feeding After Recycling
to the borehole
& Before
to the borehole
& Before
concreting
Concreting
(near borehole
(near trench
base)
base)

1.03-1.20

1.10

28-45

36

< 4.0
8-12

1.17

1.08

1.10

49

33

36

1.0

1.1

0.1

0.8

A mechanical cable-suspended grab was used for barrette construction.

Excavation of the trench was carried out by the cyclic process of lifting and lowering of
the grab under gravity and tangential force of the clamshell operated by cables. A guide
wall with inside clear dimensions slightly larger than the nominal size of the barrette
was used to guide the grab during initial bites. Bentonite slurry was introduced to the
trench as soon as initial excavation commenced. The excavation was continued under

139

1076

DEEP FOUNDATIONS 2002

the bentonite slurry to the final depth. Since time consumed in the preparation of
instrumentation was relatively long, circulation of the bentonite slurry was continuously
done to keep the slurry agitated, which also helped to alleviate the growth of filter cake
by minimizing the actual exposure time. Prior to lowering the reinforcement cage the
trench was once again occupied by grab to scrape the trench walls to remove, if any,
filter cake formed on the walls. This attempt is in line with the recommendation made
by Reese and ONeill (1988). It is authors opinion that, if due to some unforeseen
reasons, reinforcement cage lowering have to be delayed for considerable period of
time, it is a good practice to use the grab again to scrape the trench walls. This measure
eliminates any foreseeable negative impacts caused by unexpected delays. After
lowering the rebar cage, tremie concreting was done.
Hydraulic rotary drilling was employed for bored pile excavation. Different
from barrette excavation, temporary casing of 15 m length was used as a support in Soft
Clay layer for bored pile drilling to assure the stability of the borehole. Firstly, an auger
was used to drill within the temporary casing and to the bottom of the stiff clay layer.
Bentonite slurry was then supplied to the borehole and drilling was continued with a
bucket down to final depth of excavation. Before lowering the reinforcement cage a
special cleaning bucket was used to scrape the borehole walls and to clean the base.
Reinforcement cages were then lowered inside the borehole and concreting was carried
out by tremie method. Time consumed in different construction activities of test
barrette and bored pile is shown in Table 2. Equipment used in the excavation of
barrette and bored pile are shown in Figure 3.
Table 2. Time consumed in different construction activities of test barrette and bored pile

Time Consumed (hours)

Activities

Barrette
19
36
10
2.5
7.5
75

Drilling
Recycling and Desanding
Reinforcement Cage Lowering
Delay
Concreting
Total Time Consumed

Bored Pile
7
4
7
4.5
4.5
27

Pile Toe Grouting Process

Pile toe grouting was conducted on both test piles. A special grouting pipe system (a
grouting circuit) in the form of U-loop consisting of 0.5 PE pipes connected with 0.5
steel pipe at the bottom was used. The steel pipe horizontally placed at the bottom of the
U-loop is perforated and wrapped by a rubber sleeve "Manchette". Three grouting
circuits (3 loops) were used for barrette whereas two grouting circuits (2 loops) were
employed for bored pile. The grouting circuits were attached to the reinforcement cage.
Twenty-four hours after concreting, the pile toe was cracked by pressurized
water to flush and open the grouting circuit so as to make a way for forth coming grout.
Cement grouting was commenced soon after completion of the water cracking. The
preset grouting criteria was to stop the grout injection when the grout volume reached to

140

1077

DEEP FOUNDATIONS 2002

the maximum target volume (1500 and 750 liter for barrette and bored pile respectively)
or the maximum target grouting pressure (40 kN/m2) was achieved, whichever came
first. The grouting for test barrette and bored pile was stopped by the volume criteria
since the maximum grouting pressure achieved for both piles was only 20 kN/m2.
BARRETTE

BORED PILE

3.0 m.

DIA. 1.50 m.

1.5 m.

Po re P re ss ure (kN/m 2 )
Depth (m)
0

Weathered Crust
GUIDE WALL

-10

S t iff Clay

Medium to
Dense Sand

dro
t ic
st a
Li n
e

dow
i ne
n L

-57.0 m.

-57.0 m.

r aw
ic D

Dense Sand
Hard Clay
Dense to Very
Dense Sand

5
0

e tr

Hard Clay

z om

4
0

Hy

ie
lP

-60

tu a

-50

600

Ac

-40

400

1
0

BENTONITE SLURRY
-30

200

TEMPORARY
STEEL CASING

Soft Clay

-20

6
0

Figure 2. General configuration of test barrette and bored pile

Mechanical cable-suspended
grab used for barrette excavation

Auger used for upper part of bored

pile drilling

Bucket used for bored pile

drilling under slurry

Figure 3. Excavation and drilling tools of barrette and bored pile

141

1078

DEEP FOUNDATIONS 2002

Major Differences of Two Test Piles

Basic installation parameters of the test barrette and bored pile are considerably
different as shown in Table 3. The general configuration of test barrette and bored pile
with soil profile at site is illustrated in Figure 2. The influence of the major differences
indicated in Table 3 on capacity of piles are discussed in later sections of this paper.
Table 3. Major differences in installation process of barrette and bored pile
Major Differences
Test Pile
Type

Equipment and
Drilling Process

Total
Construction
Method / Use of Construction
Time
Temporary
Casing

Shape

Properties
of
Bentonite

Barrette

Mechanical grab
with excavation
under gravity and
grab action

None

75 hours

Rectangular

Different
as
indicated
in Table 1

Bored
Pile

Hydraulic rotary
piling rig with
rotary drilling
action

15 m Steel
Casing

27 hours

Circular

Different
as
indicated
in Table 1

Load Test Program

Pile load tests were carried out in accordance with ASTM D 1143-81 Item 5.1
Standard Loading Procedure. The pile and barrette were tested under three and four
cycles of loading and unloading respectively.
Test Frame Configuration. Test barrette and bored pile layout is presented in Figure 4.
General view of the barrette and bored pile test frame set up is presented in Figure 5.
With overall height of 8 m, reaction frame utilized for static load testing on barrette in
this project was claimed to be one of the biggest of its kind in the region. Four barrettes
were used as anchoring system. Four numbers of built-up steel girders supported on
each side by two 1st level cross beams were used as main beams to achieve the
maximum capacity of 60,000 kN. First level beams were supported against the second
level cross beams. Second level cross beams were anchored against surrounding
barrettes using anchor blocks at the top. Specially fabricated transfer-girders were used
to distribute the tension force transferred from the tie-bars to dowel bars above the
anchor barrette heads. Sixteen hydraulic jacks each having 5,000 kN capacity were
placed between the test barrette cap and the main beams of the reaction frame.
Test pile layout of bored pile was similar to those of other static bored pile load
tests in Thailand as it was considered that preset maximum test load would not require
special test frame set up. Steel test frame anchored against four bored piles was used in
bored pile load test. Different from the test frame configuration of the barrette, only 1
layer of cross beams was required for that of bored pile.

142

DEEP FOUNDATIONS 2002

5.00

1079

5.00

2n d C ro ss
B eam
8.00
4.00

1st C ro ss
B eam

4.00

5.00

AN C H O R
B AR R E TTE

An ch o r P ile

P ile C ap

H yd rau lic
Jacks

8.00

TE S T
B AR R E TTE

5.00

M ain B eam

4.00

M ain B eam

Test P ile

H yd rau lic Jacks

4.00

C ro ss B eam

2n d C ro ss
B eam

1st C ro ss B eam
2n d C ro ss
B eam

Bored Pile Test Frame Layout

Barrette Test Frame Layout

Figure 4.

Barrette and bored pile test frame configuration

View of Barrette Test Frame

View of Bored Pile Test Frame

Figure 5. General view of the reaction frame for static load test on barrette and bored pile

Instrumentation and Monitoring System. Vibrating wire strain gauges (VWSGs) and
Mechanical Extensometers (ME) were fixed at five levels along the shafts at the known
interface boundaries of different soil layers. At each level, four sets of VWSGs and one
set of ME were installed for the pile whereas six sets of VWSGs and two sets of ME
were installed for the barrette.
Direct measurement from four dial gauges placed in diametrically opposite
positions having equidistance from the test pile axis was used as a main monitoring
system of test pile head movement. Precise leveling and piano wire were also utilized as
backup for pile head movement measurement. Additional two dial gauges were also

143

Conference Papers

NINTH INTERNATIONAL
CONFERENCE AND EXHIBITION
ON PILING AND DEEP
FOUNDATIONS
3 5 June 2002
NICE Acropolis
France

Performance of Toe-Grouted Large Diameter Bored Pile in Sand-Gravel

Formation of Chiang Mai City, Northern Thailand
Narong Thasnanipan, Zaw Z. Aye, Chanchai Submaneewong & Dr. Wanchai Teparaksa

Sponsored by
Deep Foundations
Institute

145

PERFORMANCE OF TOE-GROUTED LARGE BORED PILE IN SAND-GRAVEL FORMATION OF CHIANG

MAI CITY, NORTHERN THAILAND
Narong Thasnanipan, Zaw Z. Aye, Chanchai Submaneewong, Seafco Company Limited, Bangkok, Thailand
Dr. Wanchai Teparaksa, Department of Civil Engineering, Chulalongkorn University, Bangkok, Thailand

Performance of diameter 1.5 m toe-grouted bored pile founded at depth 44 m in sand-gravel

formation of Chiang Mai City is presented in this paper. Toe grouting was applied by using
tube-a-manchette method. Effectiveness of toe grouting was observed from the standard
penetration tests conducted immediately below pile toe through grouting pipes at two stages,
prior to grouting and three weeks after completion of grouting. Improvement of pile end
resistance by toe-grouting is demonstrated with a series of back analyses carried out by loadtransfer method employing hyperbolic transfer function. Construction experience of large
diameter deep-seated bored piles in multi-layered subsoil and practical aspects of toe-grouting
are also briefly discussed.
INTRODUCTION

of Precambrian age (BAUM et al. 1970). Rhodes et al.

(1997) named the fluvial deposits poorly exposed along
the western edge of Chiang Mai Basin as Mae Rim
Formation and stated that this formation is folded into
gentle domes and basins. Based on the discovery of a
distinct intraformational angular unconformity, the
authors confirmed that the Mae Rim Formation
accumulated during the opening of the Chiang Mai
Basin. According to the record of deep borehole drilled
up to 1184 m, the basin deposits comprise claystone of
Tertiary age overlain by thick layers of unconsolidated
sand, clay and gravel of Quaternary age.

Chiang Mai founded 800 years ago is now known to be

the tourist city of the northern Thailand. It is located
about 700 km north of Bangkok at latitude 18.75 N and
longitude 99 E. Bored piles have been used as
foundations of heavy structures in Chiang Mai. Most of
the bored piles constructed in Chiang Mai are of small to
medium diameter short bored piles : diameter ranging
from 0.60 to 0.80 m with toe depths of about 20 m.
Large diameter deep-seated bored piles were rarely
constructed in the city as high-capacity foundations were
not in great demand in the past. Hence, behavior and
performance of large diameter deep-seated bored piles
in this city have not yet been studied. In recent years,
elevated bridges and underpasses are required in the
urban area and the vicinity of Chiang Mai to alleviate the
increasingly congested traffic condition.
In some
projects, design of bridge pier called for a single pile per
column concept to facilitate the available space. Large
diameter deep-seated bored piles are necessary in such
cases. Construction experience of diameter 1.5 m toegrouted bored piles founded at depth 44 m in sand
gravel formation of Chiang Mai City is presented in this
paper. A total of 33 bored piles were constructed for
bridge foundation of elevated interchange.
The
instrumented static pile load test was carried out up to
25,000 KN on one of the working piles as a proofloading test, which was claimed to be the highest static
pile load test ever tested in the city of Chiang Mai.

Intrasuta (1983) presented the geologic setting of

Chiang Mai area based on the lithology and structural
features of each unit. The geological map of Chiang Mai
and the surrounding area modified from the work of
Intrasuta (1983) is shown in Figure 1.
SITE GEOLOGY AND SUBSOIL PROFILE
The project is located in the inner part of the lowland
floodplain area.
The site subsurface geology is
characterized by thick alluvial deposits composed of
alternating series of sand and clay layers overlaying
gravel beds. A relatively thin layer of gravel formation is
observed between sand and clay layer at depth about 14
-17 m.
Figure 2 illustrates the generalized soil profile along the
site derived from 5 boring logs with indication of SPT-N
values obtained from each borehole. As can be seen in
the figure a large variation in soil profile along the site is
observed.
Above 35 m depth, the thickness and
occurrence of strata radically varies from one borehole
to another. However, below this depth the top and
bottom boundary elevation of stratum 11 (sand and
gravel mixture formation) is relatively uniform across the
site. Pile toe was designed to be founded in this layer.
The simplified description of soil stratum along with SPTN values and undrained shear strength, Su (for clay
layers) are presented in Table 1. Ground water level
observed from the soil investigation boreholes was about
1 m below the existing ground.

REGIONAL GEOLOGY AND MORPHOLOGY

Geomorphologically, Chiang Mai is characterized by flat
floodplain occupying the most part of the city and the
eastern area bordered by the mountain ranges in the
west. The main river, Mae Ping flows through the city
center from north to south. Geologically, the lowland
floodplain of the city is situated in the Chiang Mai Basin
which was believed to be a horst-basin type structural
basin formed by tectonic activities during Tertiary period.
Western mountain ranges are formed by so-called
Chiang Mai-Tak gneiss belt composed of medium to
high grade gneiss, schist, calc silicate rocks and granite

147

Figure 1 Geological map of Chiang Mai and adjacent areas (modified from Intrasuta, 1983)
B H -2

BH-1
0

10

DEPTH (m.)

15

20

25

6
11
14
11
9
16
16

35

24
56
51
26
24
40
40
32

53
57
33
27

35
43
STR ATUM 10

4 9 C LAYEY S AN D
30
50
ST R AT U
62
M 10
C LA YE Y
33
SA N

40

45

STR ATUM 3
LO O SE SAN D

2 3 STR ATUM 5
1 8 M ED IU M TO
1 5 VER Y D EN SE SAN D

13
17
10
23
17
15
51

29
50
30

STR ATUM 2
M ED IU M TO
STIFF S ILTY C LAY

B H -3

15
10
18
13
11
12

56
40
87
90
50/5"
50/0.5"

37
33
33
41
47
37
69
31
50
93/11"
50/0"

9
9
4
3
5

STR ATUM 1
LOOSE CLAYEY
TO SILTY SAND

26
29
40
33
55
20
32
2 6 STR ATUM 10 C LAYEY SAN D
40
52
58
53
67
40
95/11"
STR ATUM 11
50
78
50/5"
50/0"

TE S T P ILE

SAND AND
G R AV EL M IXTUR E

STR ATUM 12 VER Y D EN SE G R AVE L

B H -5

4
8
5
16
10

19
19
17
15
EL
46
UM 6
G R AV
STR AT
DENSE
3 7 M E D IU M T O
24

UM 8
E
STR AT M TO D EN S S AN D
M E D IU Y T O S IL T Y
E
Y
A
CL
STR ATUM 9
VER Y S TIFF TO H AR D
SILTY C LAY

B H -4

6
3

STR ATUM 4
SO FT TO M ED IU M STIFF
D AR K G R EY C LAY

16
20
58
42
21
58
3 5 STR ATUM 7
VER Y S TIFF SILTY CLAY
68
51
46
47
42
44
22
53
39
44
55
55
31
54
50/5"

18
13
12
16
15
8
7
27
23
15
24
64
56
18
31
49
25
49
42
40
28
55
50
36
57
41
50/5"
50/6"
50/3"
50/1"

85
50/5"
50/0"

50

Figure 2 Generalized soil profile along the project (the numbers denoted behind the borehole are SPT-N values)

148

Table 1 Summarized description of soil stratum

Stratum
No.
1
2
3
4
5
6
7
8
9
10
11
12

Description

Clayey sand
(brownish grey)
Medium to very
stiff silty clay
Loose sand
Soft to medium
stiff silty clay
Medium to very
dense sand
Gravel
Very stiff clay
Medium to dense
clayey sand
Very stiff to hard
silty clay
Clayey sand
Very dense sandgravel formation
Very dense gravel

SPT
N

Unit
Weight
(KN/m3)
18.0

4 to 16

19.0

8 to 18

90-120

18.5
18.0

3 to 5
3 to 7

35-60

19.5

15 to 46

21.0
20.0
21.0

21 to 68
21 to 38
26 to 40

150-200
-

22.0

22 to 69

> 200

22.0
22.5

26 to 49
40 to 90

50 / 0

manchette were fixed in the reinforcement cage as all

bored piles were required to have sonic logging test.
The sonic logging tubes were also utilized as toegrouting tubes. The arrangement and configuration of
grouting circuits are presented in the later section. Once
the reinforcement cage installation was done, tremie
pipes complete with air lifting accessories were lowered
and air lifting operation commenced.
Air lift was
continuously perform until the arrival of the first concrete
truck to minimize the accumulation of sediments at the
pile base if not entirely eliminated. The depth of the pile
was re-measured to ensure that the base of the pile was
reasonably cleaned.
Hundred percent cleaning or
removal of undesired sediments was most unlikely
particularly for the long bored piles in the presence of
thick sand formation and layers of gravel. Concreting
was done by tremie method. All the contract piles
including the test and anchor piles were constructed by
the procedure described above. The properties of
bentonite slurry measured from the test pile at different
stages are shown in Table 2.

Su
2
(KN/m )

Y'

BORED PILE CONSTRUCTION METHOD

All the contract bored piles including the test pile were
constructed by rotary drilling under wet process or slurryreplacement method, using bentonite slurry as support
fluid in boring process. Depending on the presence of
loose material, temporary steel casings of length ranging
from 3 to 11 m were used to stabilize the borehole in top
soil layer. After installing the temporary casing by
hydraulic vibrator (Vibro-hammer) soil inside the casing
was drilled out by an auger. The borehole was filled with
bentonite slurry prior to advancing the drilling below the
casing. Drilling bucket was then used to excavate the
soil under the slurry. Upon reaching to 15-19 m depth,
verticality of the borehole was checked by Koden drilling
monitor (sonic caliper method). Koden test at this depth
also helped to assess the stability of borehole in gravel
layer (stratum 6). After the careful examination of the
borehole verticality monitoring results, drilling was
continued to the final depth. Additional Koden tests
were performed to ensure the verticality of the borehole
at regular interval before reaching to the final depth if
excessive inclination was found in the first test.

Depth (m) x

x'

x'
Y
Y

Top of
slurry

Y'

0.0

Clayey
sand
Stiff clay

5.0

Fine sand

10.0

Coarse sand
Gravel with
sand
Very stiff
clay with
trace of
sand

15.0
20.0
25.0
30.0
35.0

Dense sand
Sand-gravel
formation

Since the sand content in the slurry after drilling to the

final depth was higher than that of specified (API 4%),
recycling and desanding were carried out to screen out
the excessive sand in the bentonite and to maintain the
properties of slurry. Up to 1m3 of sand in the slurry was
removed by the recycling process. This represented
about 0.5 m height of sediment at the base of pile. After
completion of recycling, a final Koden test was
conducted. Figure 3 shows the borehole profile obtained
from the final Kodent test conducted on test pile. Pile
base was then cleaned by cleaning bucket to remove the
accumulated loose material deposited on the base of the
pile during drilling process. After the cleaning, the depth
of the pile was measured and recorded. Reinforcement
steel cages were then installed section by section in the
borehole. Four sonic logging access tubes (internal
diameter 54 mm black steel pipes) complete with tube-a-

40.0

Figure 3 Borehole profile of test pile from Koden Test

(soil profile shown in the figure is based on the record of
a visual inspection of soil from pile drilling)
Table 2 Properties of bentonite at different stages
Density
(g/ml)

Marsh Funnel
Viscosity
(sec/qt)

API Sand
Content
(%)

pH

Before drilling

1.05

31

After recycling

1.07

33

0.5

After cage
lowering

1.10

30

Stage

149

SONIC LOGGING TEST

were fixed to the reinforcement cage and installed prior

to concreting as described in the earlier section. In the
initial stage of piling, manchette opening and water
cracking was done 24 hours after casting of piles and
cement-grout injection was performed after 7 days or
more upon completion of sonic logging test. However,
with this process difficulty was found in cement grout
filling process as grout circuit was blocked by intrusion of
sand and very small gravels into the pipes through
perforations since manchettes were broken out during
earlier water cracking.

The construction specification called for the integrity

assessment of all the piles by sonic logging test. Sonic
logging test was normally carried out at a minimum of 7
days after completion of pile concreting. Figure 4 shows
the sonic logging result of the lower part of the test pile
tested 7 days after pile concreting.

75

1500

75

Pile W all
Main Reinforcem ent
Sonic Tube (ID =54 m m )
Horizontal Pipe (ID. =40 m m )
with Manchette

Rubber Sleeve
(Manchette)

Pile Base

S EC T IO N O F G R O U T C IR C U IT
Horizontal Pipe (ID.=40 m m )
Sonic Tube (ID.=54 m m )

Perim eter of
Bored Pile

Rubber Sleeve
(Manchette)
75

Figure 4 Sonic logging test profile of lower part of the

test pile (depth indicates in the figure are measured from
the top of sonic tubes)

1500

75

P LAN O F PILE B ASE

Figure 5 Grout circuit arrangement and layout (spiral

reinforcement is not shown in the figure for clarity)

PILE TOE-GROUTING
Original Specification

Modification of Grouting Process

The original project specification called for the use of drill

and-grout method for toe-grouting. Drill-and-grout
method involves coring through the pile base up to
20 cm into the soil beneath the pile base. Coring is to be
carried out by drilling tools using sonic tubes as access
to the base of the pile. After completion of drilling, sonic
tubes are to be cleaned and filled with grout followed by
grout injection from top of the tubes. This method was
considered inappropriate as it is likely that uplift pressure
at pile toe level blow out sand gravel mixture into the
sonic tubes which in turn would cause the undesired
consequences ; blockage of the sonic tubes and
progressive collapse of the soil below pile base.

Therefore it was decided to change the sequence by

performing water cracking immediately prior to the actual
cement grouting. To facilitate this change, the level of
manchette was modified by placing just above pile tip
level (in the initial stage the manchettes were placed
about 10 to 20 cm above the pile tip level). After
completion of the sonic logging test, manchette opening
and water cracking was carried out followed by the filling
of grout in the grout tubes. Each grout loop was filled
with cement grout from the top by using one pipe as inlet
and another as outlet. Cement grout was injected at
each loop with low pressure from inlet pipe until the
cement grout comes out from the outlet pipe. Once both
outlet pipes were observed with discharge of cement
grout, two outlet grout pipes were connected to the grout
pump and pressurized grout injection started. Toegrouting was performed through all 4 pipes
simultaneously. Grouting was stopped when the injection
pressure reached to 20 bar.
The target grouting
pressure was set at 20 bar (2,000 KN/m2) for two main
reasons ; (i) to avoid excessive hydrofracturing of
surrounding soil as pile base was embedded in highly
permeable sand- gravel formation, (ii) the main purpose
of the toe-grouting in this project was to improve the soft
toe condition of the pile. Compacting the remaining
loose materials at the base of the pile and compressing
the distressed soil immediately below the pile base with
the target pressure was considered sufficient to improve
the soft toe condition. Total volume of grout required per

Alternative Method (Tube-a-manchette Method)

The piling contractor therefore proposed tube-amanchette technique as an alternative toe-grouting
method. In this method a system of grouting consisted
of two U-shape loops formed by two pairs of inside
diameter 54 mm main vertical steel pipes (sonic tubes)
attached with horizontal steel pipe of inside diameter
40 mm at bottom of each pair. The horizontal steel
pipes were perforated and wrapped with rubber sleeves
(manchettes). The bottom of the sonic tubes were
securely closed with a steel cap.
The position
arrangement and layout of grouting circuit is shown in
Figure 5. The grouting circuits completed with manchette
fabricated together with the lowest section of sonic tubes

150

As grout injection was applied under relatively low

pressure, upward migration of grout along the pile shaft
was unlikely. Following are the reasons.
Toe-grouting was carried out minimum 7 days after pile
concreting. Time gap between concreting and grouting
would have allowed some positive improvement of
bentonite skin (if any) along the pile shaft.
Sediments of approximately 5 cm were recorded for
most of the piles prior to concreting though intensive
air lifting was carried out. Possibly loose gravels were
also present at pile base since removal of relatively
heavy material by air lift is unlikely. Hence, it is most
likely that the base of the pile was loose and weaker
than not only the surrounding soil but also pile / soil
interface along the lower portion of pile shaft.
At 20 bar, a relatively low pressure, the grout would
first compact, compress and permeate the weakest
zone around the applied area (pile base as described
above). Grout would also bounded and cemented
loose sand and gravel at pile base.

pile was generally 350-500 liter, but up to 750 liter was

consumed for a few piles. No pile uplift was observed
during pressure grouting.
ASSESSMENT OF TOE-GROUTING BY SPT
In order to evaluate the effect of toe grouting, standard
penetration tests (SPT) were conducted immediately
below pile toe through sonic tubes at two stages, prior to
grouting and after completion of grouting. The tests
were performed on one working pile adjacent to the test
pile. Toe-grouting was carried out from one loop so that
the other loop (two sonic tubes) can be used for the
access of SPT sampler. The first SPT was conducted 7
days after casting of the pile whereas the second test
was performed 3 weeks after toe-grouting. Prior to
conducting the SPT at both stages, a steel cap at the
bottom of the sonic tubes were drilled through by
diamond core bit coupled with core barrel. Figure 6
shows the SPT-N values obtained from two different
stages in comparison with that of original obtained from
nearest borehole (BH-4). As can be seen in the figure, it
is clear that the density of the soil below pile toe has
been increased appreciably from soft toe condition by
pile-toe grouting. After completion of second SPT, an
additional toe-grouting was carried out from two sonic
tubes which were used for the access of SPT sampler as
described earlier.

Cement grout mixed with sand and some gravels were

observed from the first 5 cm of the collected sample
followed by pure sand-gravel mixture during standard
penetration test at pile base, which favored the assumed
mechanism described above. More investigations are
required to establish the exact mechanism of grout at
pile toe. Simplified version of the most likely geometry of
the pile base at different construction stages is
demonstrated in Figure 7.

20

SPT

2nd SP T

1s t SPT

nd

st

40

60

SPT

SPT-N

80

Original Value from BH - 4

100

S c hem atic sk etch o f pile b as e

P lan of P ile B ase

Original SPT-N of BH- 4

(nearest soil investigation
Before toe grouting ( 1 st SPT)

Borehole W all
Main Reinforcement
Sonic Tube (ID =54 m m )

After toe grouting ( 2 nd SPT)

0

Figure 6 Comparison of SPT-N values

Horizontal Pipe with

Manchette
Borehole Base

Loose Material
(Sedim ents)

S TAG E 1 (After B orin g)

MECHANISM OF TOE-GROUTING
Teparaksa et al. (1999) investigated the penetration of
grout into the sand layer from the toe-grouting of bored
piles in Bangkok. Toe-grouting was performed 24 hours
after concreting with maximum pressure 40 bar for
grouting the base of large diameter deep-seated bored
piles in medium to fine sand layer. From the series of
SPT and soil sampling performed below the pile toe, the
authors concluded that grout did not permeate into the
sand layer at pile base. Upward flow of grout along the
pile / soil interface is the most likely phenomenon of toegrouting in Bangkok as stated by them.

Convex
Concrete
Surface
Loose Material
(Sedim ents)

Pile
Concrete

P ossib le m ech an ism

(i) Concrete casting pressure
pushed loose m aterial
toward pile wall and
form ed convex surface
(ii) Loose material trapped
between bottom soil and
concrete

S TAG E 2 (After C on cretin g)

Convex
Concrete
Surface

In this case study, likely mechanism of toe-grouting

under maximum pressure 20 bar in sand-gravel
formation are;
Compaction of loose material at the pile base
Bounding and cementing the loose gravel at pile base
Compressing of distressed soil at pile base

Pile
Concrete

P ossib le m ech an ism

(i) Pressurized cem ent grout
com pacted loose m aterial
(ii) Grout cemented with
loose material and
form ed stiffer layer

Cem ent Grout

(cem ented with loose m aterial)

S TAG E 3 (After G rou tin g )

Figure 7 Possible geometry of pile base at different

construction stage

151

STATIC PILE LOAD TEST

Load transferred to the pile toe at maximum test load

was about 5500 KN (accounted about 22 percent of the
total applied load) as can be seen in Figure 9. At
maximum applied load, displacement at pile tip
measured by extensometer was 8.29 mm (0.5 % of pile
diameter).

Static pile load test (Maintain Load) was commenced 36

and 29 days after concreting and toe-grouting
respectively.
Additional 4 bored piles constructed
without toe-grouting were used as anchor piles. Vibrating
wire strain gauges (VWSGs) and Mechanical
Extensometers (ME) were fixed at five levels along the
shafts at the known interface boundaries of different soil
layers. At each level, two sets of VWSGs and one set of
ME were installed. The lowest level of instrument was
set at 70 cm above the pile toe level. The pile was
tested to maximum 25,000 KN, which is 2 times of the
design load (12,500 KN).

LOAD-DISPLACEMENT ANALYSIS OF THE PILE

Load-displacement behavior of pile was analyzed using
load-transfer method employing hyperbolic transfer
function proposed by Hirayama (1990). A computer
program written based on this method was used to
analyze the load-displacement behavior of the pile.
Figure 10 illustrates the hyperbolic stress-displacement
curves of bored cast-in-place piles presented by
Hirayama (1990). It should be noted that end resistance
qult indicated in Figure 10 represents physical ultimate
end resistance which is related to pile tip displacement in
hyperbolic function. Conventional definitions of ultimate
end bearing capacity for bored piles therefore do not
correspond to the qult indicates in this paper.

Load vs. Pile Displacement

The total applied load vs. pile head displacement of the
test pile is shown in Figure 8. Pile shaft and end
resistance in relation to the pile head displacement are
also included in the figure.
Load (KN)
0

5000

10000

15000

20000

25000

S kin friction : f

Pile Head Displacement

(mm)

1
4

12

16

Total Load Resistance vs. Pile Head Displacement

Shaft Resistance vs. Pile Head Displacement
End Resistance vs. Pile Head Displacement

End resistance : q

Load distribution along pile shaft at various applied loads

derived from the strain gauge data is shown in Figure 9.

D ep th (m )

15000

Maxim um Mobilized
U nit Shaft Friction
(KPa)
30000 0

10

10

20

20

30

30

40

40

50

50

50

100

f=

Z
a f + bf Z

Zref, f

qu lt

Load Distribution Mechanism

fu lt
2

P ile s h a ft d is p la c e m e n t : Z

Figure 8 Load-displacement curves of test pile

Applied Load
(KN)

f u lt

20

1
bf

af

150

1
be

ae

1
q u lt
2

q =

Ze
ae + beZe

Z ref, e
0
P ile tip d is p la c e m e n t : Z

Figure 10 Hyperbolic stress-displacement curves (After

Hirayama 1990)
One-dimensional finite element model constructed by
elastic bar elements used in the method proposed by
Hirayama is shown in Figure 11.
The detailed
discussion of the background principle and procedure to
determine the required parameters in using this method
can be seen in the published paper of Hirayama (1990).
The major advantage of the method is that input
parameters required can be determined from the results
of routine in-situ tests such as SPT and CPT. The
predicted load-displacement curves agreed well with
those of the actual measurement according to the
authors experience in Bangkok subsoil.

Figure 9 Load distribution along pile shaft of test pile

152

zn

n+1

Elem ent No.

z2
3

k f1

Node No.

Increm ental nodal displacem ent

Table 4 Input shaft and end resistance values used in

analyses

Tangential spring constant

In crem en tal ap p lie d lo ad :

k f2

k fn

n+1

Depth
(m)
0-4
4-12
12-14
14-16
16-18
18-22
22-24
24-30
30-32
32-36
36-42
42-44

ke

Preliminary Analyses
Two preliminary analyses were carried out. The input
soil parameters required in the analyses were
established based on the recommendation of Hirayama
(1990) as shown in Table 3 with the following key
assumptions.
Same shaft resistance values (derived from SPT-N
values of original site investigation) for both cases
Preliminary Case 1 : No pile toe improvement so that
normal toe condition exists lower factor (100) is
used in ultimate end resistance derivation.
Preliminary Case 2 : Toe-grouting improves pile toe
so that stiffer toe condition than case 1 higher
factor (300) is used in ultimate end resistance
derivation.

Ultimate end
Resistance, qult (KPa)

gravel

Assigned pile toe condition

qult

fult

qult

fult

qult

35
70
90
100
150
150
200
150
150
150
150
200

8.5

35
70
90
100
150
150
200
150
150
150
150
200

25.5

30
30
60
60
60
135
135
135
135
135
135
110

40

0.2 % of pile 0.2 % of pile 0.2 % of pile

diameter
diameter
diameter

5 % of pile 5 % of pile 5 % of pile

diameter
diameter
diameter

The load-displacement curves of preliminary and back

analyses along with that of actual measurement from
static pile load test are depicted in Figure 12.
Load (KN)
0

10000

Pile Head Displacement (mm)

5N ( 200 KPa)

20000

30000
Preliminary Case 1
Preliminary Case 2

Back Analysis
Measured

8
12
16

Case 1 : turns to plunging

failure over 23 ,000 KN

20
24

10N (150 KPa)

100N

7
7
18
20
30
24
60
50
44
40
40
85

fult

(KPa) (MPa) (Kpa) (MPa) (KPa) (MPa)

Pile toe condition

Normal toe Stiff toe
Stiffer toe
Note : SPT-N indicate in the table are average values of
relevant soil layers

Table 3 Factors used in estimation of shaft and end

resistance from SPT-N values for preliminary analyses
Case
Parameter
Soil Type
Case 1
Case 2
Sand and
gravel
clay

SPT
N

Sand
Clay
Sand
Gravel
Sand
Clay
Sand
Clay
Sand
Clay
Clay
Sandgravel
Assumed
displacement
required to fully
mobilize shaft
resistance
Assumed
displacement
required to fully
mobilize end
resistance

Figure 11 One-dimensional non-linear finite element

model of a pile used in analyses (after Hirayama 1990)

Ultimate shaft
Resistance, fult (KPa)

Soil
Type

Input data in analyses

Preliminary analysis
Back
analysis
Case 1
Case 2

28

300N

Figure 12 Load-displacement curves of preliminary and

back analyses in comparison with measured data

Normal toe Stiff toe

Note : Normal toe condition in above table represents nongrouted or plain pile toe condition

Assessment of Load-displacement Behavior

Back Analysis Based on Pile Load Test Results

Reviewing the analyses results shown in Figure 12 in

conjunction
with
Table
4,
load-displacement
characteristics of the test pile may be evaluated as
follows. Same pattern and magnitude of pile head
displacement as the actual measurement can be
observed for preliminary Case 1 before reaching to
18,000 KN applied load.
However, displacement
increases abruptly over 18,000 KN and turns to the

Fixing the ultimate shaft resistance derived from strain

gauges data obtained from static pile load test, a series
of trail runs were made to back calculate the ultimate
end resistance of the test pile. Table 4 shows the input
data of shaft and end resistance used in the preliminary
and back analyses.

153

CONCLUSIONS

plunging pattern at 23,000 KN, which suggests that the

pile would experience plunging failure before reaching to
the maximum test load of 25,000 KN, if the normal toe
condition for non-grouted or plain bored pile assumed in
Case 1 existed. Load-displacement pattern of Case 2 is
found to be very
similar to that of the actual
measurement. The factor (300N) adopted in the
derivation of end resistance shown in Table 3 is
therefore considered reasonable.

The practical aspects of toe-grouting in sand-gravel

formation of Chiang Mai City has been presented. Direct
observation from SPT conducted at pile base before and
after toe-grouting clearly confirms the improvement of
pile base by tube-a-manchette toe-grouting technique
employed in the project. Target pressure, 20 bar proved
to be sufficient for toe-grouting in sand-gravel formation.

With the use of maximum shaft friction values obtained

from the load test results and end resistance value of
40 MPa as input parameters, the load-displacement
curve of back analysis best fits the actual measurement
which means that the selected value of end resistance
in back analysis fairly represents that of actual
magnitude. In other words, the measured loaddisplacement curve strongly suggests that end
resistance of the pile is markedly high; detrimental soft
toe condition does not exist at pile base.
Load (KN)
0

10000

A practical method for examining the load against pile

displacement behavior is illustrated by a series of loaddisplacement curves with various pile end resistance
values. Load-displacement analyses with the application
of hyperbolic load-transfer method provided an effective
tool in assessment and evaluation of pile toe condition.
In addition to SPT conducted at pile base, improvement
of pile end resistance by toe-grouting is demonstrated by
back analyses results. Without applying the toe-grouting,
test pile is likely to experience plunging failure before
reaching to the maximum test load. Toe-grouting in this
case could therefore be considered as extra safety
measure to ensure the sufficient capacity of the pile.

Measured displacement
at 25 ,000 KN

20000

30000

40000

The back analyzed load-displacement curves also

demonstrate that the pile might be able to resist the load
up to 37,000 KN without reaching to the plunging failure.
Therefore, the specified design load 12,500 KN might
have been underestimated for the toe-grouted bored pile
in sand-gravel formation. Instead of conducting a proofloading test on the working pile, static pile load test
should have been performed on the pilot or trial pile with
the higher load up to failure to determine the ultimate
capacity of the pile and to optimize the pile design which
might provide some cost-saving for the project.

Pile Head Displacement (mm)

20
40
60

Input End Resistance 10 MPa

80

Input End Resistance 15 MPa

100

Input End Resistance 30 MPa

120

Input End Resistance 40 MPa

140

Input End Resistance 60 Mpa

160

Measured

REFERENCES

180

Figure 13 Load-displacement curves of back analyses

with various input end resistance values

BAUM, F., BRAUN, E. V., HAHN L., HESS, H., KOCH,

K. E., KRUSE, G., QUARCH H. and SEIBENHUNER M.
1970. On the geology of northern Thailand. Geologishe
Jahrbuch, Hannover, Beiheft 102.

To examine the influence of end resistance on the loaddisplacement mechanism, a series of additional back
analyses were carried out using different values of end
resistance with the fixed shaft resistance values
(obtained from the load test results) as presented in
Figure 13. It is evident from Figure 13 that loaddisplacement behavior is not excessively sensitive to
variation for end resistance values in the range of 30 to
60 MPa.
For end resistance less than 15 MPa,
displacement
increases
significantly and
loaddisplacement curve turns to a plunging failure pattern,
which is more significant for the case of 10 MPa as can
be seen in the figure. In summary, it should be pointed
out that test pile would have experienced plunging type
of failure if it had seated on the normal pile toe condition
(non-grouted or plain pile). It is therefore considered
that pile end resistance was effectively improved by the
toe-grouting method employed.

HIRAYAMA, H. 1990. Load settlement analysis for bored

piles using hyperbolic transfer functions. Soil and
Foundations, Vol. 30, No. 1, Japanese Society of Soil
Mechanics and Foundation Engineering, pp. 55-64.
INTRASUTA, T., 1983. Ground water potential in
st
unconsolidated sediments in Chiang Mai Basin. 1 Syp.
on geomorphology and quaternary geology of Thailand,
Bangkok, pp. 179-195.
RHODES, B. P., BLUM, J., DEVINE, T. and
RUANGVATANASIRIKUL, K., 1997. Geology of the Doi
Suthep metamorphic complex and adjacent Chiang Mai
Basin. Proc. of the International Conference on
Stratigraphy and Tectonic Evolution of Southeast Asia
and the South Pacific, Bangkok, Thailand, pp. 305-313.
TEPARAKSA, W., THASNANIPAN N., and ANWAR M.
A., 1999. Base grouting of wet process bored piles in
th
Bangkok subsoils. The 11 Asian Regional Confe. on
Soil Mechanics and Geotechnical Engineering, Seoul,
Korea, p.p. 269-272.

From Figure 13, it is reasonable to conclude that the pile

end resistance is in the range of 30 to 40 MPa. The loaddisplacement curves of end resistance 30 and 40 MPa
suggest that the higher load up to 37,000 KN can be
applied on pile without reaching to the plunging failure.

154

Under the Royal Patronage of His Majesty the King

Organized by :
The Engineering Institute of Thailand
Department of Civil Engineering, Khon Kaen University
with support from Petroleum Authority of Thailand and
Department of Civil Engineering from Universities in Thailand
In Corporation with :
Department of Agricultural Engineering and Engineering Alumni Association, Khon Kaen University

Advanced Technologies for the

Development of Civil
Engineering, Volume 2
Proceedings of the Eighth National Convention on
Civil Engineering

APPLICATION OF POLYMER-BASED SLURRY FOR WET-PROCESS BORED

PILES CONSTRUCTION IN MULTI-LAYERED SOIL OF BANGKOK
Zaw Zaw Aye1, Kamol Singtogaw2, Chanchai Submaneewong3
Project Manager1, Project Director2, Geotechnical Engineer3
SEAFCO Co., Ltd., Thailand

155

APPLICATION OF POLYMER-BASED SLURRY FOR WET-PROCESS BORED PILES

CONSTRUCTION IN MULTI-LAYERED SOIL OF BANGKOK

Zaw Zaw Aye1, Kamol Singtogaw2, Chanchai Submaneewong3

Project Manager1, Project Director2, Geotechnical Engineer3

SEAFCO Co., Ltd., Thailand

ABSTRACT: As an alternative to the classic bentonite slurry, polymer-based slurry has been increasingly
used for wet-process bored piles in Bangkok. Polymer-based slurry having environmentally friendly
properties, is found to be a good solution in minimizing the problems associated with the disposal of
contaminated soils as well as the site pollution. This paper presents the practical application of polymer-based
slurry for construction of medium to large diameter deep-seated bored piles in Bangkok. Important factors in
using polymer-based slurry in Bangkok subsoil are also discussed.

KEYWORDS : Polymer-based slurry, bentonite slurry, filtercake, fluid loss, wet-process bored pile
1. Introduction

2. Wet-Process Bored Pile Construction Method

Drilling slurry or supporting fluid plays a major role

in construction of wet-processed medium to large
diameter deep-seated bored piles in multi-layered
soil of Bangkok. Rotary drilling technique coupled
with wet-process method using bentonite slurry as
support fluid was commonly applied to construct
these piles for over two decades. As bentonite slurry
had proven acceptable to make qualified bored piles
in the past, most of them were constructed only by
using this mineral based slurry. Recently, polymerbased slurry (PBS) has been increasingly popular in
Bangkok as an alternative supporting fluid in wetprocess bored pile drilling. The primary reason, in
the initial stage, for using this alternative slurry was
to minimize the problems associated with using
classic bentonite such as site pollution and the
disposal of excavated soils which are usually
contaminated. This paper presents the application of
polymer-based slurry for construction of wet-process
bored piles in multi-layered soil of Bangkok.

Wet process method as its name implies makes the

pile under wet condition by using drilling slurry as
supporting fluid for borehole stability. Depending on
the subsoil condition, temporary casing of 14-18m in
length is also used as a support in soft clay layer.
Drilling is commenced by dry process with auger
applying rotary drilling action in the soft clay layer
and stiff clay layer. Before reaching to the first sand
layer, slurry is fed to the borehole and drilling is
continued with the bucket to the final depth. A
cleaning bucket or airlift is used to clean the bottom
of the pile prior to reinforcement cage installation.
Concreting is carried out by tremie method under the
slurry. Slurry from borehole is normally pumped
back to the storage tank during concreting and
reused for next piles after treating to get specified
properties. Depending on the level of design pile
cut-off, bored piles are normally overcast about 1 to
3m above cut-off level to avoid presence of
contaminated concrete within the design pile length.

157

3. Basic Principles of Different Drilling Slurries

It is important that practicing engineers involved in
bored pile construction understand the basic
principles on reaction of soil layers with different
drilling slurries so that the correct procedures and
methods can be applied in operation to avoid the
detrimental effect of wet-process bored piles.
Bentonite and polymer slurry react differently with
the soil layers. Pure bentonite slurry stabilized the
borehole wall by its density and formation of a
membrane or filtercake particularly in permeable
layers. Bentonite slurry is capable of maintaining the
soil particles in suspension. Therefore, soil particles
in bentonite slurry column stay in suspension for
long hours and slowly settled toward the borehole
bottom. Formation of filtercake reduces the slurry
filtration into the permeable layers, thick first and
second sand layers typically encountered in
Bangkok.
Unlike pure bentonite slurry, polymer-based slurry
does not suspend sand especially coarse grain
particles. Hence, in polymer slurry, sand particles
(mainly from cuttings during drilling) tend to fall out
of suspension and settle on the base of borehole
significantly sooner than that of pure bentonite as
evident from field observation. In Bangkok,
depending on subsoil profile and composition of
polymer-based slurry, 0.1m to 1m of sediment is
commonly observed on base of borehole in 2 to 8
hours. The sedimentation of sand particles is found
to be stable with time.
Polymer molecules tend to wrap around clay and silt
particles and produce small agglomerated structure.
By its own weight, the resulting agglomerated
structures tend to settle out of suspension slowly and
accumulate mushy sediments on the firstly settled
sand sediments at the bottom of the borehole. Some
of the agglomerated particles also tend to float
temporarily on the surface of polymer slurry stay in
the suspension. The polymer molecules then proceed
to attach themselves to larger and appear as a bulky
material that some observers have termed oatmeal
(ONeill and Reese, 1999 [4]). In Bangkok as small
percentage of bentonite is added in polymer-based
slurry, similar to polymer-clay system (Majano and
ONeill, 1993 [3]), it can be assumed that a thin
polymer-bentonite layer is formed in the soil-slurry
interface of permeable layers. This thin membrane
of polymer-based slurry is capable of reducing the
high filtration in sand layers of Bangkok. However,
overall volume of fluid loss for polymer-based bored
piles is found to be higher than those of bentonite
bored piles.

158

4. Practical Application of Polymer-Based

Drilling Slurry In Bangkok
In Bangkok, the primary reason for switching from
classic pure bentonite slurry to polymer-based slurry
in most cases was to minimize the problems
associated with environmental issues caused by the
use of bentonite slurry. In the initial stage of using
polymer-based slurry, the major concerns were the
potential negative impact on stability of the borehole
and the shaft friction capacity as this alternative
slurry was not yet accustomed to the bored piling
industry in Bangkok. However, no adverse effects
from the use of polymer-based slurry have been
experienced to date and numbers of projects have
been successfully completed with this environmental
friendly drilling slurry.
4.1 Mixed-Ratio and Procedure of Mixing
Dry anionic partially hydrolyzed polyacrylamide
(PHPA) polymer powder premixed with fresh water
and a small percentage of bentonite is mainly used
for long bored piles in Bangkok. First, a small
dosage of bentonite powder is premixed with fresh
water (bentonite 5-15kg per cubic meter of water)
and leave in the storage tank for minimum 12 hours.
The predetermined dosage of dry polymer (typically
polymer powder 0.5kg per cubic meter of premixed
bentonite slurry) is then added and mixed to get so
called polymer-based slurry. ONeill and Reese
(1999 [4]) cited that vigorous mixing of polymers
supplied in the dry form should be avoided, since
polymer chains can be broken down and the polymer
slurry rendered ineffective. Polymer-based slurry is
stored in the steel tanks before supplying to the
bored pile drilling. The reason for adding bentonite
for long bored piles is to minimize the fluid loss
(high filtration of slurry into formation soils), which
was observed from the trial boreholes drilled only
with pure polymer slurry. Depending on the
thickness and permeability of sand layers present at
site and the embedded length of the pile in sand
layers, addition of bentonite may not be necessary.
No specific method or formula is available to
determine the required best composition of polymerbentonite ratios. Therefore, at each site, different
mixed-ratios should be made on trial in the initial
stage to determine the most appropriate ratio. Table
1a and 1b present the mixed composition of
polymer-based slurry of bored piles constructed at
10 different locations. According to the loadsettlement curves obtained from static pile load tests,
observed settlements of all piles were less than those
of predicted values using bentonite pile parameters.

Table 1a. Technical data of polymer-based slurry

bored piles at different locations in Bangkok
Description
Pile Diameter (m)
Pile Length (m)
Mixed
Polymer
Ratio per
3
Bentonite
1m (kg)

Properties of slurry
prior to concreting

Total standing time prior

to concreting
(Hr : Mins)
Density
(g/ml)
Viscosity
(sec/qt)
pH
Sand
content
(%)
Approx.
First sand
Pile
embedded
Second
in sand
sand
layers (m)

0.80
49

T2

0.80
50

Pile ID
T3
1.20
51

1.80
62

1.50
54

0.50

0.50

0.50

0.50

0.50

12

12

10

10

12

14:00

23:00

15:30

T1

17:20 12:20

T4

T5

1.06

1.03

1.01

1.04

1.04

40

45

44

51

44

0.5

0.5

12

20

13

18

13

based slurry is commonly introduced only after

drilling reaches to stiff clay layer. However, the top
level of slurry is required to maintain about 1m
below the existing ground at all times once borehole
excavation reaches to the first sand layer. The
results of the sonic caliper measurement conducted
on hundreds of piles proved that borehole can be
well stabilized with the use of polymer-based slurry
at total construction time prior to concreting even
over 72 hours. Figure 1 shows the borehole profile
test conducted after 24 hours leaving the hole filled
with polymer-based slurry without any agitation.
Y'
EXISTING
GROUND 0
SOFT
CLAY

Properties of slurry
prior to concreting

Pile Diameter (m)

Pile Length (m)
Mixed
Polymer
Ratio per
3
Bentonite
1m (kg)
Total standing time (after
drilling completion)
prior to concreting
(Hr : Mins)
Density
(g/ml)
Viscosity*
(sec/qt)
pH
Sand
content
(%)
Approx.
First sand
Pile
embedded
Second
in sand
sand
layers (m)

1.00
51

1.00
30

T7

Pile ID
T8
0.80
41

0.60
26

0.60
24

0.50

0.50

0.50

0.50

0.50

10

7:00

3:45

6:55

2:30

0:45

1.02

1.03

1.01

1.02

1.02

36

45

44

43

42

0.1

12

12

15

11

T6

T9

X'

X'
Y

Y'

TOP OF
CASING
TOP OF
SLURRY

10
MEDIUM
CLAY 15

Table 1b. Technical data of polymer-based slurry

bored piles at different locations in Bangkok
Description

STIFF 20
CLAY

T10

25
DENSE 30
SAND
35

Figure 1 Profile of polymer-based pile: test

conducted after 24 hours leaving the hole (filled
with PBS) without agitation
4.3 Properties of Polymer-Based Slurry

* Marsh Funnel Viscosity

4.2 Borehole Stability
Ata & ONeill (1998, [2]) reported that polymer
slurry when introduced with the right dosage and
prior to the borehole reaching the piezometric level,
maintained the stability of the boreholes for more
than 18 hours without significant change of the
borehole diameter at any depth. In Bangkok subsoil,
as temporary casing of about 15m length is installed
to protect the collapse of soft clay layer, polymer-

It is important that properties of slurry used are

within the applicable ranges to avoid the detrimental
effect on the performance of bored piles. ONeill &
Reese (1999, [5]) cited that there is no perfect set of
slurry specifications that can be used on every job,
but specifications should be tailored to fit the
requirements of a particular job at a particular
location, where possible. Today, though use of
polymer slurry is widely acceptable in bored piling
industry, information on the desirable properties is
still extremely limited. Table 2 shows the ranges of
properties of polymer slurries specified by the
International Association of Foundation Drilling
Standards and Specification Committee (ADSC,
1999 [1]), those successfully used on several
projects as reported by ONeill and Reese (1999,
[4]) from the work of Majano et al. (1994) and the
properties commonly applied in Bangkok subsoil.

159

Table 2 Properties of polymer slurries specified by

ADSC and Polymer-based slurry used in Bangkok
Majano et al., 1994
ADSC Emulsified
Dry
(at 20 C)
PHPA
PHPA
(Vinyl)
1.02
0.995 to 0.995 to
Maximum
1.01
1.01

Common
Practice in
Bangkok
(PBS**)
Density
1.05
(g/cc)
Maximum
Viscosity 40 90*
33 - 45
50 - 120
40 to 90
(sec/qt)
(+/- 5)
pH
7 - 12
8 11.7 8 11.7
8 11
API Sand
1.0
0 to 1
0 to 1
Less than
Content Maximum
1%
*or as recommended by manufacturer and approved by
geotechnical engineer, ** Polymer-based slurry
Property

4.4 Integrity of Polymer-Based Bored Piles

According to sonic integrity and sonic logging or
sonic coring tests conducted on large numbers of
polymer-based bored piles in Bangkok, there is no
evidence that integrity of the pile is affected by the
use of polymer-based slurry.
4.5 Effective Base Cleaning for Polymer-Based
Bored Piles in Bangkok
ONeill & Reese (1999 [4]) pointed out that
vigorous mechanical agitation such as pumping by
centrifugal pumps can shear the polymer chains
severely. As soil-polymer reaction system is largely
depending on the polymer chemical chain structure,
excessive breaking of it would affect the
performance of the slurry. Therefore, high turbulent
means of base cleaning methods such as recycling
by centrifugal pumps should not be performed for
the polymer-based bored piles. Field experience
shows that base cleaning by air lift method cannot
effectively clean thick sediments at the base of the
polymer-based bored pile. In Bangkok, the effective
base cleaning for long bored piles may be achieved
by the following procedure.
(1) Carry out drilling to 1 to 2m above the final
design depth under polymer-based slurry for the
first bored pile
(2) Leave the borehole for 2 to 8 hours (depending
on the rate and stabilization trend of
sedimentation observed from trail borehole at
each site)
(3) Move the drilling rig to commence the drilling
of second bored pile
(4) Drilling rig come back to the first bored pile and
carry out final drilling for (remaining 1 to 2m)
with the cleaning bucket
(5) Immediate installation of reinforcement cage
and concrete pouring

160

Table 3 Summarized comparison of bentonite and

polymer-based slurry in application
Description
Mixing
method

Bentonite Slurry
Add bentonite
powder in water
and mixed by
Electric mixer
(or)
Water jet mixer
(using
submersible
pumps in tank
and mix with
jetting action)

Main function
in borehole
stability

Unit weight
(density) and
filtercake effect

Sand
suspension in
slurry column

Ability to suspend
sand particles

Desanding
process

Use desanding
equipment

Pile base
cleaning

Cleaning bucket
or air lift or
recycling
Possible
degradation of
shaft friction
capacity if
continuous
agitation is not
applied
Particulate
nature
Potentially
hazardous to
aquatic life
Not be disposed
in drainage
system or
sewers

Effect on long
exposure time

Environmental
impact

Polymer-based slurry
First add bentonite
powder in water
and mixed similar
to normal bentonite
mixing
Adding polymer
powder in
premixed bentonite
slurry and mix with
electric mixer
Vigorous mixing of
polymer powder
must be avoided
Unit weight
(density) and
binding of soil
particle by polymer
strand
Possible forming of
very thin filter cake
due to addition of
bentonite
Sand particles tend to
full out of suspension
and settle on base of
borehole.
Desanding equipment
is not required. Use
sedimentation tank to
allow fine sand and
silt settle before
reuse.
Mainly cleaning
bucket. Occasionally
by airlift
Shaft friction
degradation has not
been encountered
(from available
literatures and
Bangkok experience)
Non-particulate
nature
Non-hazardous
Can be disposed on
site in drainage
system or sewers

5. Review of Shaft Friction Capacities Affected

by Drilling Slurry
Properties of drilling slurries influence on the shaft
capacity has been studied by many researchers. A
number of researchers reported the reduction of shaft
capacity caused by excessive formation of the filter
cake on borehole wall of bored piles constructed

with bentonite slurry. It is mainly contributed by

long exposure time without slurry agitation as cited
by them.

Load (kN)
1

4000

Unit Skin Friction (kN/m 2)

8000 12000

0 .0

1 0.0

10

2 0.0

20

3 0.0

30

4 0.0

40

5 0.0

50

50

100

150

So ft to

20

M edium Clay

Stiff to
Very Stiff
Clay

Depth (m.)

Few researches are available on the comparison of

shaft friction capacity between bored piles
constructed with bentonite and polymer. Ata and
ONeill (1998 [2]) reported that developed shaft
friction factors (for sand) and (for clay) of bored
piles constructed with polymer were higher than
those recommended by USA-FHWA.

10

30

40

M edium Dense
to V.Dense
Sand

Hard Clay
M edium to

50

Mobilized at 9 9 0 0 kN

6. Performance of Polymer-Based Slurry Bored

Piles in Bangkok

60

Qult (calculated)

60

6 0.0

Figure 2. Load distribution curves and average unit

shaft friction of T1 (Dia. 0.80x49 m)

Instrumented static pile load test results of three test

piles constructed by wet-process with polymer-based
slurry in Bangkok are presented in this paper.

Lo ad (k N)

Unit Skin Friction (kN/m 2)

1
0

6.1 Slurry Properties and Construction Records of

Test Piles

6.2 Instrumented Pile Load Test Results

Shaft frictional resistance of polymer-based bored
piles in sand layers were of major interest to the
authors as it is considered that drilling slurry has less
influence on clay layers according to the previous
studies on bentonite bored piles.
Load distribution curves along the test pile shafts at
various applied loads of 3 test piles are illustrated in
Figure 2, 3 and 4. The values of average unit skin
friction developed by the maximum test load at the
different soil layers along the shaft of test piles in
comparison with those of ultimate unit skin friction
calculated using the empirical formula are also
illustrated in these figures. As can be seen in the
figures, developed average unit shaft friction in sand
layers are significantly higher than those of
theoretically calculated values using normally used
formula and parameters applied for bentonite bored
piles in Bangkok. This finding suggested that
polymer-based slurry bored piles produced better
shaft friction resistance in sand layers than that of
bentonite bored piles.

8000 12000

Lo o se Sand

10

10

Stiff to Very
Stiff Clay

20

20

Dense to V ery
Dense Sand

30

30

40

40

50

50

50

60

60

60

50

100

150

So ft clay

Depth (m.)

10

20

30

40

H ard Clay
Dense to V ery

Mobilized at 9 9 0 0 kN
Qult (calculated)

Figure 3. Load distribution curves and average unit

shaft friction of T2 (Dia. 0.80x50 m)
Load (kN)

1
0
0

Unit Skin Friction (kN/m 2)

0

6000 12000 18000

10

10

20

20

30

30

40

40

50

50

50

60

60

60

50

100

150

200

250

Soft Clay
10

Stiff Clay

Depth (m.)

Table 1a shows the properties of polymer-based

slurry and construction records of test piles (T1, T2,
and T3). It is to be noted that all these test piles were
constructed by the same contractor. Hence it is
assumed that influence of other variables such as
equipment as well as construction procedure used in
drilling, slurry mixing and quality control measures
applied for all piles would be minimized in
evaluation of the load test results.

4000

20

V. Stiff to
Hard clay

30

De nse Sand
Hard Clay

40

De nse to
D

Mobilized at 1 8 0 0 0 kN
Qult (calculated)

Figure 4. Load distribution curves and average unit

shaft friction of T3 (Dia. 1.20x51 m)
A summarized comparison between measured total
shaft friction resistances at maximum test load and
those of theoretically calculated values using
bentonite bored piles parameters of 3 test piles is
presented in Table 4.

161

Table 4 Calculated and Measured shaft resistance of

three instrumented test piles
Description
Diameter (m)
Length (m)
Design Load (kN)
Max. Test Load (kN)
Total shaft
resistance of
pile (kN)

Calculated

Measured at
max. test load
Measured
Calculated

T1
0.80
49
3300
9900

Test Pile ID
T2
T3
0.80
1.20
50
51
3300
7200
9900
18000

6900

6200

10820

9710

9670

17060

1.41

1.56

1.58

higher shaft friction resistance could be obtained

if pile is constructed with polymer-based slurry.
(5) Design parameters currently used for bored
piles in Bangkok were developed from bentonite
bored piles. These parameters are found to be
conservative for polymer-based bored piles.
Further research is needed to confidently
establish the design parameters for bored piles
constructed with polymer-based slurry in
Bangkok subsoil.
Acknowledgement

Thasnanipan et. al. (2002, [5]) reported that loadsettlement behavior of eleven polymer-based bored
piles constructed in Bangkok subsoil are better than
that of prediction using bentonite bored piles
parameters.
7. Main Factors on Shaft Friction Capacity
Improvement of Polymer-based Bored Piles
Shaft friction capacity improvement of bored piles
constructed under polymer-based bored piles in
comparison with those of bentonite bored piles is
considered to be contributed by the following
factors.
No thick mudcake or filtercake is formed by
polymer-based slurry
Polymer strands tend to bind the soil particle in
the formation which generates the drag forces and
cohesion (Unlike plate-shaped bentonite, polymer
molecules are hair-shaped strands)
8. Conclusion
(1) The practical application of polymer-based
slurry in construction of wet-process bored piles
in Bangkok subsoil has been presented.
(2) The measurement of borehole profile by sonic
caliper and construction records proved that
there is no adverse effect of using polymerbased slurry in Bangkok subsoil as far as the
stability of borehole is concerned.
(3) Overall shaft resistances of bored piles
constructed with polymer-based slurry are
higher than that of theoretically calculated
values using bentonite bored pile parameters in
the order of 1.5.
(4) The developed average unit shaft frictions in
sand layers of 3 instrumented bored piles
constructed under polymer-based slurry are
found to be higher than those of ultimate values
calculated by empirical formula using bentonite
piles parameters. This finding suggested that

162

The authors wish to express their appreciation to Mr.

Narong Thasnanipan, Managing Director, SEAFCO
Co., Ltd. for his encouragement to write this paper.
For Further Details
Please contact Mr. Zaw Zaw Aye, SEAFCO Co.,
Ltd. at 26/10 Rarm Intra 109 Road, Bangkok 10510
Tel.02-9190090, E-mail : [email protected]
References
[1] ADSC Standards and Specifications for the
Foundation Drilling Industry, ADSC, Revised 1999,
Incorporating ACI 336.1R-98, 1999.
[2] Ata A. A. and ONeill M. Side-wall stability and
side-shear resistance in bored piles constructed with
high-molecular-weight polymer slurry, Third
International Geotechnical Seminar on Deep
Foundation on Bored and Auger Piles, Ghent,
Belgium, 19-21 October 1998, Balkema, Rotterdam,
111-117, 1998.
[3] Majano R. E. and ONeill M. W. Effect of
mineral and polymer slurries on perimeter load
transfer in drilled shafts, A Report to ADSC,
University of Houston, Dept. of Civil Engineering,
UHCE, 1993.
[4] ONeill M. W. and Reese L. C. Drilled Shafts :
Construction Procedures and Design Methods, U.S
Department of Transportation Federal Highway
Administration, Publication No. FHWA-IF-99-025,
ADSC Publication No. ADSC-TL 4, 1999.
[5] Thasnanipan N., Z. Z. Aye, Submaneewong C.
and Teparaksa W. Performance of Wet-Process
Bored Piles Constructed with Polymer-Based Slurry
in Bangkok Subsoil, Proc. of the International Deep
Foundations Congress 2002, Geotechnical Special
Publication No. 116, Volume One, ASCE, February
14-16 2002, Orlando, Florida, USA, 143-157, 2002.

Fourth Regional Symposium on Infrastructure

Development in Civil Engineering, RSID4
Bangkok, Thailand
April 3rd 5th 2003

Organized by :
Department of Civil Engineering, Kasetsart University
Tokyo Institute of Technology
University of Philippines - Diliman

Development in construction of bored pile and barrette for

infrastructure projects in Thailand, A Country Report
Narong Thasnanipan, Kamol Singtogaw, Zaw Zaw Aye and Sumate Pravesvararat

163

DEVELOPMENT IN CONSTRUCTION OF BORED PILE AND BARRETTE FOR

INFRASTRUCTURE PROJECTS IN THAILAND, A COUNTRY REPORT
Narong Thasnanipan, Kamol Singtogaw, Zaw Zaw Aye and Sumate Pravesvararat
SEAFCO Co., Ltd. , 26/10 Rarm Intra 109 Road, Bangchan, Khlong Sam Wah
Bangkok 10510, Tel. 0-2919 0090, Fax. 0-2919 0098
E-mail : [email protected], Homepage : www.seafco.co.th

ABSTRACT
This paper presents the authors experiences in development of bored piles and
barrettes for infrastructure projects in Thailand with particular focused on construction issues.
Some literatures related to wet-processed bored piles and barrettes in Thailand published
through out the past 30 years are also summarized together with recent research works.
KEYWORDS : bored pile, barrette, deep foundations, bentonite, polymer-based slurry
INTRODUCTION
Cast-in-place deep foundations, particularly wet-processed deep-seated large diameter
bored piles and barrettes have been used for foundations of various infrastructure projects in
Thailand for over two decades. The versatility of the construction method and the high-load
capacity which in turn offered the constructibility and cost-saving are the main factors
contributed to the increasing use of deep-seated large diameter bored piles and barrettes. In
the initial stage there were a number of questions on design and construction aspects of these
foundation systems particularly in Bangkok subsoil. With the passage of time, the
construction equipment, installation techniques and testing of deep foundation elements have
been developed. Large numbers of instrumented full-scale static pile load tests were
conducted through out 1990s which provided better understanding on behavior of these deepseated foundations. Researches focused on the design parameters and methods were produced
based on these test results. The design parameters were so well established that wet-process
cast-in-place foundations became regarded as reliable foundations for not only in the property
sectors but also for practicing engineers involved in the infrastructure projects. This paper
presents the development of wet-processed bored piles and barrettes in Thailand. The
information contained in the paper are mainly from the infrastructure developments built in
Bangkok and adjacent areas.
CONSTRUCTION METHOD
Large diameter deep-seated wet-processed bored piles can be constructed by 2
methods, reverse-circulation and rotary drilling method. As construction procedure of historic
reverse-circulation method has been well documented in many literatures and it is less
frequently utilized nowadays in Thailand, only a rotary-drilling method is briefly presented in
this article as follows.
In rotary drilling method, a temporary casing of appropriate length (12 to 18m in
Bangkok depending on the thickness of soft clay) with required diameter (internal diameter
not less than that of design bored pile diameter) is first installed to ensure the stability of the
borehole in the top soft or loose soil layers. In some projects where the vibration is strictly

165

limited, standard length of casing with oscillator or short temporary casing of 5 to 8 m is

pushed down in combination with pre-boring process. Drilling is commenced by auger to drill
out the soil within the temporary casing and up to the top of the first water-bearing sand layer
or bottom of the casing in case of using short casing method. Drilling slurry or supporting
fluid is then supplied to the borehole and drilling is continued with a bucket down to the
design final depth of the pile. Before lowering the reinforcement cage, a special cleaning
bucket is used to clean the base of borehole. If bentonite slurry is used, recycling method by
air-lift or pump is applied as base cleaning process. Reinforcement cages are then lowered
into the borehole and concreting is carried out by tremie method.
A mechanical or hydraulic cable-suspended grab is commonly used for barrette
construction. Excavation of the trench is carried out by the cyclic-process of lifting and
lowering of the grab under gravity and tangential force of the clamshell operated by cables.
Different from bored pile construction, a guide wall of depth 1 to 1.5 m with inside clear
dimensions slightly larger than the nominal size of the barrette is used to guide the grab
during initial bites. Bentonite slurry is introduced to the trench as soon as initial excavation
commenced. The excavation is continued under the bentonite slurry to the final depth. After
recycling of slurry and lowering the rebar cage, concreting is done by tremie method.
OVERVIEW OF APPLICATION
Bored pile
The first wet-process large diameter bored pile was constructed for Pinklao Bridge in
Bangkok 30 years ago. Using reverse circulation method, diameter 1.50m bored pile was
installed up to 45m in the second sand layer. Three major bridges were constructed by bored
pile of reverse circulation method in Bangkok from early 1970 to 1980. The summarized
information of these bridges are tabulated below.
Table 1 First 3 major bridges constructed by large diameter wet-processed
bored piles in Bangkok
Project Name
Pinklao Bridge
Sathorn Bridge
New Memorial
Bridge

Year of
Construction
1971
1979
1982

Construction Method
Reverse circulation
Reverse circulation
Reverse circulation

Diameter
(m)
1.50
1.50
1.50

Depth
(m)
45
46
49

The first wet-processed bored pile of rotary-drilling method down to first sand layer
was constructed in Bangkok in late 1970 for high-rise building project, Royal Orchid Hotel.
Since then bored piles constructed by rotary-drilling method have been extensively used for
foundations of various heavy structures such as high-rise buildings, elevated expressways,
overpass-bridges, underground car park buildings, waste-water treatment plants and most
recently underground train stations of Bangkok first subway project. By mid 1980, bored pile
became the foundation of choice for heavy structures particularly in urban area of Bangkok.
The versatility of the construction method and the high-load capacity which in turn offered
the constructibility and cost-saving are the main factors contributed to the increasing use of
deep-seated large diameter bored piles and barrettes. Most of the early wet-process bored
piles (1980s) in Bangkok were constructed up to 50 m depth. Diameter of bored piles

166

commonly constructed in early days were ranging from 0.6 to 1.5 m. Table 2 below
summarizes the information of early major high-rise building projects in Bangkok.
Table 2 First high-rise building projects constructed by large diameter wet-processed bored
piles in Thailand (from 1979 to 1983)
Project Name
Royal Orchid Hotel
(1979)
Taiping Tower
(1980)
River City Hotel
(1982)
Asoke Tower
(1983)
Time Square
Building (1983)

Construction Method
Rotary-drilling with
auger and bucket
Rotary-drilling with
auger and bucket
Rotary-drilling with
auger and bucket
Rotary-drilling with
auger and bucket
Rotary-drilling with
auger and bucket

Diameter (m)
0.80 - 1.00

Depth (m)
33.0

0.80 - 1.50

32.0

0.80 1.00

27.5

0.80 - 1.50

50.0

0.80 - 1.50

50.0

Barrette
In some projects where bored piles were not feasible due to the site constraints,
applicable construction method and/or extensive bearing capacity requirements, the use of
barrette foundations would make a suitable alternative. Barrettes with dimension ranging from
0.80 m x 2.7 m to 1.5 m x 3.0 m for safe working load capacity from 11,000 to 23,000 kN
have been used in some major projects.
The first barrette in Bangkok was believed to be constructed in late 1970 for the
foundation of Bangkok Bank Head Office Building at Silom road with size of 0.6-0.8 x 2.5m
and toe depth of 33m in late 1970. Barrettes can be constructed with flexible layout plan for
both vertical and lateral loads. The layout pattern of barrettes can be arranged in a continuous
row or column, radial, alternating long and short axis of barrette and a combination of two or
more of such patterns. One type of equipment can be used for constructing both barrettes and
diaphragm walls in particular projects thus reduces the mobilization cost for additional
equipment. In addition to the large bearing capacity requirements, on site difficulties such as
limited head room where piling rigs cannot be utilized, under such situations like presence of
overhead high voltage power cables, existing overpasses or structures for elevated
expressways and planned subway stations, also demands the barrettes. A summary of
observation on barrettes of completed 26 projects completed with respect to selection criteria
is presented in Table 3.
Table 3 Summary of Barrette Selection
Criterion
High Load Capacity
Minimize Construction
Equipment
Limited Head Room for
Excavation
Combination with
Diaphragm Wall
Foundation as well as
portion of column

No. of Project
4
6

Remarks
Incorporated with bored piles
Alternative for bored piles

3
10

Under existing structures such as bridges,

elevated expressway and power lines
As diaphragm wall legs

Provision for the future requirement

167

Static load test up to 5290 ton conducted on barrette set the record as the highest load
ever tested for a single cast-in-situ deep foundation in Thailand (Thasnanipan et. al. 2002a)
PROBLEMS OF WET-PROCESSED BORED PILES AND BARRETTES
IN EARLY DAYS
Basic but extensive problems were experienced in early stages of bored pile construction in
Thailand as summarized below.
Limited availability and capacity of equipment
Lack of skills in operation of equipment (particularly drilling rig)
Adverse effect due to a slow rate of drilling (e.g. excessive formation of filter-cake by
using bentonite slurry)
Limited knowledge and less advance technique in control of bentonite slurry
Limited experience in construction method and related negative impact
Quality of concrete for tremie concreting method
Lack of experienced engineers and foremen
Improper construction and quality control specification and guidelines for deep-seated
piles in local soil condition
Improper design for constructibility
DEVELOPMENT IN CONSTRUCTION
Over the past 3 decades, along with the development of wet-processed piling
technology world-wide, equipment, construction method, design as well as better
understanding and realization of construction impact on the performance have been
significantly improved in Thailand. Table 4 summarizes the areas of improvement in bored
pile and barrette construction and main factors contributed to these.
Research works presenting some of these developments were published both locally
and internationally. The design, construction and behavior of bored cast in-situ concrete piles
in Bangkok Subsoil was presented by Thasnanipan et. al., (1998a). The construction and
performance of barrettes in Bangkok Subsoil was also reported by Thasnanipan et. al.,
(1998b). Effect of construction time and bentonite viscosity on shaft friction of bored piles
was also pointed out by Thasnanipan et al., (1998c). Teparaksa et. al., (1999) published the
base grouting of wet process bored piles in Bangkok subsoil which demonstrated the method
of base grouting and reported the grout spreading mechanism for long bored piles and
performance of base grouted bored piles.

(a)

(b)

Fig. 1 (a) Guide wall for cruciform barrette (b) View of cruciform barrette after installation of
reinforcement

168

Construction of cruciform barrette (shown in Figure 1) for monopole type highvoltage power transmission line was described in the work of Thasnanipan et. al., 2000a.
Above-mentioned research works reflect the development history of bored pile and barrettes
in Thailand and offered useful information to the local construction industry and perhaps to
the international deep foundation engineering society.
Table 4 Summary of development in bored pile and barrette construction
Area of
development /
improvement
Speed of
construction

In the past
(Early 1980s)

At present

Minimum 3 days
required to complete
diameter 1.5m tip 50m
bored pile

Less than 1 day to

complete diameter
1.5m tip 50m bored
pile

Pile size and

depth

Bored pile
Maximum diameter 2.0m
and common depth 2550m for bored pile.
Barrette
Limited in size and depth

Base grouting

Not available

Bored Pile
Maximum Diameter
2.0m and common
depth 25-60m
Barrette
Various sizes &
depth over 60m
Available

Application of
polymer-based
slurry (for bored
pile only)
Construction
impact on quality
and performance
Quality control in
construction
process
Quality control
test method and
interpretation

Not available

Extensively use

Not well understood

Improving

Not well established and

systematic

Well established
and systematic

Less methods available

and limited knowledge in
interpretation

Better equipment
available and better
knowledge in
interpretation

Main factor
contributed to
development
Better equipment,
operating skills as
well as improvedknowledge in
construction method,
management and
local soil condition
Better equipment,
operating skills as
well as improvedknowledge in
construction method,
management and
local soil condition
Equipment
availability and
advance technology
Material availability
and research works
Experience from past
projects and research
works
Experience from past
projects and research
works
Advance equipment,
experience from past
projects and research
works

Experience from past projects and extensive research works provided better
understanding on construction impact on quality and performance of these wet-processed
deep foundations. Marked difference between outcome quality of bored piles constructed in
1980s and 1990s can be observed by significant less defects found in the latter. It can also be
observed from load test results that bored piles constructed in late 1990 have higher capacity
than those of 1980s as shown in Table 5.

169

Table 5 Load test results of bored piles constructed in 1970 to 1980 and 1990s
Year of
construction

Project Name

Pile Dimension
(Dia. & Depth)

1971
1995

Pinklao bridge
Central Plaza
Pinklao
Taiping tower
High-rise building
at Ekamai Rd.

1980
1990

1.5m x 45m
1.2m x 45m

Design
Load
(ton)
410
700

Test
Load
(ton)
820
1400

Total settlement
at max. test load
(mm)
5
27

1.0m x 32m
1.0m x 32m

300
390

1125
975

118
23

Fig. 2 Progress of piling work for the Rama VIII main bridge
Bored piles have been extensively used for foundations of the majority of the elevated
expressways since 1991. Thousands of base-grouted bored piles were constructed for these
infrastructure projects including Second Stage Expressway (1991), Don Muang Tollway
Extension (1997), Bangna-Bang Pli-Bangprakong Expressaway (1998), and Wat Nakorn-In
project (2001). Over 700 bored piles of diameter ranging from 0.8m to 1.5m with depth from
35m to 54m were constructed (1999-2000) by polymer-based slurry for the foundation of
Rama VIII Bridge, one of the initiatives of his majesty the King Bhumibol Adulyadej.
Progress of piling works at the bank of Chao Phraya river for the main bridge of Rama VIII
Bridge is depicted in Figure 2.
The availability of more reliable and powerful equipment and tools for drilling makes
possible to construct the rock-socket bored piles. Figure 3 shows the drilling rig equipped
with rock-auger used for construction of highway bridge across Mool River in 2001, in the
north-eastern part of Thailand, where drilling was carried out through weathered-sandstone by
powerful rotary drilling rig with core barrel and rock-auger.

Fig. 3 Rock-socket bored pile construction for Mool River Project - Ubon Ratchathani Province

170

DEVELOPMENT IN DESIGN CONCEPT AND PARAMETERS

In the initial stage of introducing bored piles in Thailand, the design concepts and
parameters were mainly based on the available literatures from the research works carried out
in other parts of the world such as Tomlinson (1957), Skempton (1959), Broms (1966),
Bowles (1968), Meyerhof (1976) etc. The work of Chiruppapa (1968) was believed to be the
first research data available for design parameters of bored pile in Bangkok soft clay. The
author conducted the research based on 6 dry-processed small diameter bored piles with load
cells installed at the pile toe in AIT campus in Phathumthani. Though the piles were simply
constructed by dry-processed with casing method, this early-stage research study provided
some important information such as load-settlement behavior, adhesion factor () and bearing
capacity factor (N) of bored pile in Bangkok soft clay.
With the passage of time, design method and parameters for local subsoil were
improved as a result of research works carried out in 1980s. Differences between the behavior
of driven piles and bored piles were well realized from these researches. (Ng, 1983) presented
the load distribution characteristics of wet-processed bored piles founded in both first and
second sand layers of Bangkok based on the instrumented (strain gauges) pile load test
results. (Chiewcharnsilp, 1988) reported the shaft friction factor values of sand layers in
Bangkok based on the instrumented load test results. In addition to the researches previously
available (Chiruppapa, 1968; Suwanakul, 1969; Promboon, 1981; Ng, 1983; Chiewcharnsilp,
1988) Suchada (1989) determined the shaft friction factor (adhesion factor, ) of Bangkok
clay layers based on 11 bored piles of diameter ranging from 0.50 to 0.80m with the
embedded length varying between 21.50m and 46m. It should be noted that the design
parameters obtained from the research works of 1980s were mainly based on the estimation of
shaft friction loads from plain static pile load test results with numbers of assumption since
instrumented pile load test results were limited.
With the peak of construction-boom, large numbers of instrumented full-scale static
load tests on bored piles were conducted throughout 1990s which provided better
understanding on behavior of these deep-seated foundations. Researches focused on the
design parameters and methods were produced based on these test results. The design
parameters were more accurate and well established that wet-process cast-in-place
foundations became regarded as reliable foundations in construction industry of Thailand.
With more confidence on soil parameters and behavior of these deep foundations, the
designers designed to carry significantly higher loads for bored piles and barrettes constructed
in late 1990 than those of 1980s. (Thasnanipan et. al., 1999) reported the failure mechanism
of long bored piles in layered soil of Bangkok. The authors cited that for the bored piles
embedded in the multi layered soils of Bangkok, estimation of ultimate shaft friction capacity
needs to consider the brittle type of failure mechanism of stiff to hard clay layers and the
values selected need to be adjusted accordingly. Peak and residual values mobilized in
the stiff clay layers analyzed by the authors were plotted as shown in Figure 4 along with the
suggested curves by different researchers. It can be seen from the figure that the residual
value of 2nd stiff clay layer (undrained shear strength values of 25 ton/m2) at the maximum
test load drops below the curve suggested by (Suchada, 1989). So the values proposed in
Figure 4 for the stiff to hard clay layers gives overestimate of the ultimate shaft friction under
theses conditions.
Introduction of polymer-based slurry for wet-process bored piles marked a major
breakthrough for both construction and design engineers. Thasnanipan et al.,(2002b), reported
that bored piles constructed with polymer-based slurry have higher capacity than those
constructed with bentonite slurry. Figure 5 shows the shaft friction factors of sand layers
for polymer-based bored piles in comparison with the design line of bentonite bored piles.

171

1.20
Bored Cast Insitu Piles
1.00

Adhesion Factor ,

Tomlinson,1957
0.80
Suchada,1989
for bored piles in Bangkok soils

0.60

0.40

Peak
Stas and Kulhawy, 1984

0.20

Residual

0.00
0.00

5.00

10.00

15.00

20.00

25.00

30.00

Undrained Shear Strength (ton/m )

Fig. 4 Comparison of adhesion factor , suggested by different researchers with the actual
mobilized in the stiff clay layers (after Thansnanipan et. al 1999)
1 .0
Submaneewong (1999) : Bored Piles (Tip in 2nd Sand)
Submaneewong (1999) : Bored Piles (Tip in 2nd Hard Clay)
0 .8

Chiewcharnsilp (1988)
Polymer Bored Pile TP-1 (1st Sand)

(Ks tan )

Polymer Bored Pile TP-2 (1st Sand)

Polymer Bored Pile TP-3 (1st Sand)
0 .6

Polymer Bored Pile TP-1 (2nd Sand)

Polymer Bored Pile TP-2 (2nd Sand)

Submaneewong
( 1999 )

Polymer Bored Pile TP-3 (2nd Sand)

0 .4

0 .2

0 .0
28.0

30.0

32.0

34.0

36.0

38.0

Angle of Internal Friction, '

Fig. 5 Back-calculated values of polymer bored piles at maximum test load plotted on
design line of bentonite bored piles constructed in Bangkok subsoil
(after Thansnanipan et. al 2002b)
.
DEVELOPMENT IN QUALITY CONTROL TESTING
Improvement of testing equipment and powerful computer facilities are the key factors
contributed to the development in quality control testing. Interpretation skills relevant to local
soil condition and construction method of these tests were significantly improved in local
industry. For instance, in early 1980, interpretation of the sonic integrity (seismic test) test
results were needed to send to the specialists abroad which made the testing cost more
expensive. Significant cost-saving were achieved in some major projects of late 1990, as more
practical and precise interpretation were made to verify and establish the acceptance criteria in
proving the quality of suspected piles with anomalies.

172

Koden drilling monitoring system

In Thailand, prior to the availability of Koden testing equipment, borehole verticality
was checked by mechanical type equipment such as plumb float and adjustable globe.
However, the reliability of these methods was doubtful. Koden drilling monitor system was
believed to be first used in Thailand in 1980. The verticality of slurry-filled borehole can be
monitored rapidly from the electronic plot of continuous profile by Koden equipment. This
system is very useful for the verticality-control of long barrettes - unlike bored piles, guiding
by temporary casing is not available for barrette drilling.
Sonic integrity test
Sonic integrity test, also known as seismic test is the most common method of
integrity testing for both driven and bored piles in Thailand. Sonic integrity test is usually
selected for both quality check (control test) and retrospective investigation. It is the cheapest
in terms of cost and the simplest in terms of testing process. The main advantage of this test
is that since no particular preparation is necessary during the pile construction phase it is more
flexible to select which pile is to be tested. However, interpretation of sonic integrity testing
needs considerable experience and knowledge in testing, subsoil condition and construction
method. In many projects it is a part of the contractual requirement to conduct sonic integrity
test. Minimum 10 % to maximum 100 % of production piles are commonly tested. It is also
a reasonably acceptable method for applying as a retrospective investigation in determining
integrity of the pile. The signal characteristics and their interpretations of sonic integrity test
on piles founded in Bangkok subsoil were reported in details by Thasnasipan et al. (1998d).
Cross-hole sonic logging test
Sonic logging test is relatively expensive. It is mainly employed as a pre-planed site
quality control testing. The major advantage of this method is that test can be carried out
shortly after the pile construction. Hence, rectification measures can be implemented if the
pile is detected with defect while the foundation contractor is on site. However, this method is
generally not applicable if pile integrity is in question due to post-construction activities as
access tubes are usually grouted after completion of the test. The first sonic logging test was
believed to be conducted in 1982 for the wet-processed bored piles of Memorial Bridge
Project where bad concrete zones were detected at depth about 20m and 1m (Ng, 1983). Use
of Sonic logging test for checking pile integrity has increased in Thailand in recent years.
The results from sonic logging test conducted on model piles in Bangkok helped to extend the
knowledge of the signal characteristics and interpretation (Thasnanipan et. al., 2000b).
High strain dynamic load test
High strain dynamic integrity test has become a well-accepted method especially for
evaluating the pile capacity in todays foundation industry of Thailand and it is applied for
both driven and bored piles.. A large number of related technical papers and case histories of
the test have been published and it is a part of standards and specifications such as ASTM
D4945-89 (Standard Method for High-strain Dynamic Testing of Piles). Thasnanipan et. al.
2000b, reported the application of dynamic load testing in Thailand.
CONCLUDING REMARKS
Development of bored pile and barrettes in Thailand has been presented. According to
the authors experience as a contractor, development in both construction and design aspects
of these deep foundations in past decades are significant. In authors opinion, however, there
are many works to be done to improve further with particular focus on constructibility issues,
concrete technology for wet-processed bored piles and barrette, reliable but cost-effective

173

quality control testing and value-engineering. Starting from planning stage, site investigation,
design, construction and inspection should be integrated such that designers, contractors and
construction inspectors participating as a team with common goal. Appropriate and practical
specification should be established jointly by these parties for local soil condition and
construction method. Practical acceptance criteria should be developed to verify bored piles or
barrettes with suspected anomalies. Continuing education should be promoted for both
designers, inspection engineers and contractors. The Geotechnical Chapter of the Engineering
Institute of Thailand under the royal patronage of his majesty the King Bhumibol Adulyadej,
has started to establish the standard code of practice and guidelines for wet-processed bored
piles which will serve as a yardstick for the deep foundation industry upon completion in near
future.
REFERENCES
Bowles, J. E. (1968) Foundation analysis and design, McGRaw-Hill, New York
Broms B. B. (1966) Methods of calculating the ultimate bearing capacity of piles a summary, Soils-Soils, No.
18-19, pp. 21-32
Chiewcharnsilp, S. (1988) Static load transfer and predictions of carrying capacity of bored piles (in Thai), M.
Eng. Thesis, Chulalongkorn University, Bangkok, Thailand
Chiruppapa, P. (1968) Cast in-situ bored piles in Bangkok Clay, M. Eng. Thesis No. 213, Asian Institute of
Technology, Bangkok, Thailand
Meyerhof, G. G. (1976) Bearing capacity and settlement of pile foundations, Journal of soil mechanic and
foundation division, Proceeding ASCE Vol. 102, GT3, pp. 197-223
Ng, K. C. (1983) The construction problem and performance of large bored piles in second sand layer, M. Eng.
Thesis No. GT 82-26, Asian Institute of Technology, Bangkok, Thailand
Promboon, S . and Brenner, R.P. (1980), Large diameter bored piles in Bangkok clay, Proceedings of 10th
conference on soil mechanic and foundation engineering, Vol. 2, pp. 815-818
Suwanakul, V. (1969) Cast in-situ piles in Bangkok clay, M. Eng. Thesis No. 260, Asian Institute of Technology,
Bangkok, Thailand
Teparaksa W., Thasnanipan N., and Anwar M. A. (1999), Base grouting of wet process bored piles in Bangkok
subsoils, 11th Asian Regional Confer. on soil mechanics and geotech. Engg., Seoul, Korea, pp. 269-272.
Thasnanipan N., Teparaksa W., Maung A.W., Ganeshan B., (1998a) Design, construction and behavior of bored
cast in-Situ concrete piles in Bangkok Subsoil, Proceedings: Fourth International Conference on Case
Histories in Geotechnical Engineering, St. Louise, Missouri, U.S.A., p.p.281-287
Thasnanipan N., Maung A.W., Tanseng P., (1998b) Barrettes founded in Bangkok subsoils, construction and
performance, Thirteenth Southeast Asian Geotechnical Conference, Taipei, Taiwan, ROC, p.p.573-578.
Thasnanipan N., Ganeshan B., Anwar M.A., (1998c), Effect of Construction Time and Bentonite Viscosity on
Shaft Capacity of Bored Piles, Third International Geotechnical Seminar on Deep Foundation on
Bored and Auger Piles, Ghent, Belgium. p.p. 171-177.
Thasnanipan N., Maung A.W., Ganeshan B., (1998d), Sonic Integrity Test on Piles Founded in Bangkok
Subsoil-Signal Characteristics and Their Interpretations, Proceedings: Fourth International Conference
on Case Histories in Geotechnical Engineering, St. Louis, Missouri, U.S.A., p.p.1086-1092
Thasananipan N., M. A. Anwar Maung A. W (1999) Failure Mechanism of Long Bored Piles in Layered Soils
of Bangkok, Civil and Environmental Engineering Conference, Bangkok, Thailand, p.p. V-69 to V-73.
Thasnanipan N., Tanseng P., Maung A.W. & Anwar M. A. (2000a), Barrettes : A Versatile Foundation for
Transmission Line Towers, An International Conference on Geotechnical & Geological Engineering,
Melbourne, Australia, 19-24 November 2000
Thasnanipan N., Maung A. W. & Navaneethan T.., Z. Z. Aye (2000b), Non-Destructive Integrity Testing on
Piles Founded in Bangkok Subsoil, 6th International Conference on Application of Stress-Wave Theory
to Piles, September 2000, Sao Paulo, Brazil
Thasnanipan N., Zaw Zaw Aye and Wanchai Teparaksa (2002a) Barrette of Over 50,000 kN Ultimate Capacity
Constructed in the Multi-Layered Soil of Bangkok, Proceedings of the International Deep Foundations
Congress 2002, Geotechnical Special Publication No. 116, Volume Two, ASCE, Febuary 14-16 2002,
Orlando, Florida, USA, p.p. 1073-1087
Thasnanipan, N., Zaw Zaw Aye, Chanchai Submaneewong and Wanchai Teparaksa (2002b) Performance of
Wet-Process Bored Piles Constructed with Polymer-Based Slurry in Bangkok Subsoil, Proceedings of
the International Deep Foundations Congress 2002, Geotechnical Special Publication No. 116, Volume
One, ASCE, February 14-16 2002, Orlando, Florida, USA, p.p. 143-157

174

Proceeding of the 4th International Geotechnical Seminar on Deep

Foundations on Bored and Auger Piles
2-4 June 2003, Ghent, Belgium

Deep Foundations on Bored

and Auger Piles
Edited by:
Van Impe, W. F.

Behaviour of Polymer-based Slurry for Deepseated Bored Piles in Multi-layered Soil of

Bangkok
Narong Thasnanipan,
Boonyarak

Zaw

Zaw

SEAFCO Co., Ltd. Bangkok, Thailand

175

Aye

and

Thayanan

Behaviour of Polymer-based Slurry for Deep-seated Bored Piles in

Multi-layered Soil of Bangkok
Narong Thasnanipan, Zaw Zaw Aye and Thayanan Boonyarak
SEAFCO Co., Ltd. Bangkok, Thailand

ABSTRACT: This paper focuses the behaviour of polymer-based slurry both in the practical construction of
deep-seated bored piles and in the laboratory. The behaviour of polymer slurry with different mixed-ratios is
studied in the laboratory by means of model tests. Test models were constructed with the sand samples obtained from 25 to 40m depth, First and Second Sand layers of Bangkok. The model tests showed that significant strength improvement can be achieved in slurry-sand interface if polymer slurry (polymer 0.05% by
weight) is mixed with a small percentage of bentonite (1% by weight). The model tests also proved that filtration of slurry in permeable layers can be reduced by adding a small percentage of bentonite in polymer slurry.

1 INTRODUCTION

Depth (m)

10

Medium
Dense Sand
(First Sand)

40

Hard Clay

50

Dense to Very
Dense Sand
(Second Sand)

30
SPT-N= 20-40
t = 20kN/m 3

Su> 200 KPa

= 21KN/m 3
t

SPT-N= 50-100
t = 21KN/m 3

40

50

e
Lin

30

Su= 50-14020- KPa

t = 17-21kN/m 3

Line

Medium to
Stiff Clay

tatic

20

60

600

Su= 10-25 KPa

t = 14-16kN/m 3
-

wn
wdo
Dra

177

Bangkok Soft
Clay

400

ros
Hyd

Subsoil profile is relatively consistent at different localities in Bangkok. A typical subsoil profile is
characterized by the alternating layers of clay and
sand deposits as shown in Figure 1.

10

200

Weathered Crust

etric
zom
Pie

2 GENERAL DESCRIPTION OF BANGKOK

SUBSOIL

ual
Act

In Bangkok, deep-seated large diameter bored piles

of depths ranging from 35 to 60m and diameter of
800mm to 1800mm are commonly used for foundation of heavy structures such as high rise buildings,
elevated expressways, flyovers or overpass bridges
and recently for underground subway stations. Conventionally, these deep-seated bored piles in the
multi-layered soil of Bangkok were constructed by
rotary drilling under wet process or slurry replacement method, using pure bentonite slurry as support
fluid in boring process. As an alternative to bentonite slurry, polymer-based slurry has been increasingly used in Bangkok recently. Polymer-based
slurry with environmentally friendly properties, is
found to be a good solution in minimizing the problems associated with the disposal of contaminated
soil as well as the site pollution. This paper focuses
the behaviour of polymer-based slurry both in the
practical construction of deep-seated bored piles and
in the laboratory.

Pore Pressure (kN/m 2)

Geotechnical
Parameters

60

Figure 1. Typical subsoil profile and piezometric profile of

Bangkok (Thasnanipan et. al. 2002)

3 WET-PROCESS BORED PILE

CONSTRUCTION METHOD
Wet process method as its name implies makes the
pile under wet condition by using drilling slurry as
supporting fluid for borehole stability. Depending on
the subsoil condition, temporary casing of 14-18m in
length is also used as a support in soft clay layer.
Drilling is commenced by dry process with auger
applying rotary drilling action in the soft clay layer
and stiff clay layer. Before reaching to the first sand

layer, slurry is fed to the borehole and drilling is

continued with the bucket to the final depth. A
cleaning bucket or airlift is used to clean the bottom
of the pile prior to reinforcement cage installation.
Concreting is carried out by tremie method under the
slurry. Slurry from borehole is normally pumped
back to the storage tank during concreting and reused for next piles after treating to get specified
standard properties. Depending on the level of design pile cut-off, bored piles are normally overcast
about 1 to 3m above cut-off level to avoid presence
of contaminated concrete within the design pile
length.

layer is formed in the soil-slurry interface of permeable layers. This thin membrane of polymer-based
slurry is capable of reducing the high filtration in
sand layers of Bangkok. However, overall volume of
fluid loss for polymer-based bored piles is found to
be higher than that of bentonite bored piles according to the field observation of deep-seated bored pile
construction in Bangkok. Figure 2 (a) and (b) demonstrates the generalized behaviour of bentonite and
polymer slurry in borehole respectively.
Thick filter
cake is
formed

Suspended
soil particle
in bentonite
slurry

4 BEHAVIOUR OF SLURRIES IN OVERVIEW

Bentonite and polymer slurry react differently with
the soil layers. Pure bentonite slurry stabilized the
borehole wall by its density and formation of a
membrane or filtercake particularly in permeable
layers. Bentonite slurry is capable of maintaining the
soil particles in suspension. Therefore, soil particles
in bentonite slurry column stay in suspension for
long hours and slowly settled toward the borehole
bottom. Formation of filtercake reduces the slurry
filtration into the permeable layers, thick First and
Second sand layers typically encountered in Bangkok as shown in figure 1.
Unlike pure bentonite slurry, polymer-based
slurry does not suspend sand especially coarse grain
particles. Hence, in polymer slurry, sand particles
(mainly from cuttings during drilling) tend to fall out
of suspension and settle on the base of borehole significantly sooner than that of pure bentonite as evident from field observation. In Bangkok, depending
on subsoil profile and composition of polymer-based
slurry, 0.1m to 1m of sediment is commonly observed on base of borehole in 2 to 8 hours. The sand
sedimentation process under polymer-based slurry is
found to be stable with time. Polymer molecules
tend to wrap around clay and silt particles and produce small agglomerated structure. By its own
weight, the resulting agglomerated structures tend to
settle out of suspension slowly and accumulate
mushy sediments on the firstly settled sand sediments at the bottom of the borehole. Some of the agglomerated particles also tend to float temporarily on
the surface of polymer slurry stay in the suspension.
The polymer molecules then proceed to attach themselves to larger and appear as a bulky material that
some observers have termed oatmeal (ONeill and
Reese, 1999). In Bangkok as small percentage of
bentonite is added in polymer-based slurry, similar
to polymer-clay system Majano and ONeill, (1993)
, it can be assumed that a thin polymer-bentonite

Bentonite
slurry

Soft sediment
at base of
bored hole

(a)

Soil particle
is settling

Thin filter
cake is
formed

(b)

Sand
Particle

Polymer slurry
mix with very
little amount of
Bentonite

Dense sediment
at base of
bored hole

Polymer Chain
Structure

(b)
Figure 2. Generalized behaviour of slurries in borehole (a) bentonite slurry (b) polymer-based slurry

5 BEHAVIOUR OF SLURRIES IN
LABORATORY
Laboratory study was conducted to investigate the
followings properties of Bangkok sand layers in different types of slurries (Boonyarak, 2002).
Filtration through the sand layers
Shear strength of slurry-sand interface (by pocket
penetrometer)
Friction test for slurry-sand interface
5.1 Sand specimen preparation
The procedure involved in preparation of sand
specimen is outlined below.

178

 Sand samples were collected from the actual

bored piles drilling excavated soil of Bangkok
First and Second sand layers from depth about
25m and 40m below existing ground level respectively.
Basic properties (minimum and maximum density, grain size distribution, hydraulic conductivity) of collected sand were tested and recorded.
Determined the natural density of sand.
Sand samples were then mixed with water to obtain water content of 10%.
Poured sand into the testing mold.
Compacted the sand in the mold to the desired
natural density determined above
5.2 Slurry preparation and physical Tests
Slurries were mixed by propeller-type mixer following the procedure outlined below.
Mixed bentonite powder with clean water until
blended. Leave the slurry for few hours and remixed.
Mixed polymer powder with clean water until all
powder dissolved.
Mixed premixed bentonite and polymer slurry together (except for pure bentonite and polymer
slurry)
Tested basic properties of the slurry
The results of the physical tests of different
slurry types from the works of Boonyarak (2002) are
presented in Table 1.

5.3 Filtration tests

In order to study the laboratory behaviour with regard to the filtration rate of different slurries in sand
formation, series of filtration tests were performed.
The procedure involved in the test is outlined in the
following section. The test apparatus and arrangement used were similar to that of standard API filter
press as shown in Figure 3. The applied pressure
was predefined for the laboratory filtration test in
simulation of actual in-situ bored pile drilling condition the differential fluid pressure between the
slurry and pore pressure of formation soil at specified level, taking piezometric draw-down condition
of Bangkok sand layers (as shown in Figure 1) into
consideration.
Poured the slurry into the test chamber on preprepared sand specimen as described above
Applied predefined pressure through the pressure
inlet.
The slurry was flowed through the sand medium
by applied pressure depositing solids in suspension at the slurry-sand interface (filter cake formation)
The filtrate was collected in a graduated cylinder
The cumulative filtrate volume was recorded at
regular interval until the target test time (60 minutes)
Released air pressure and pumped out the slurry
from the chamber (sample was then ready for
penetrometer test)
PRESSURE
INLET

Table 1. Mixed ratios of different slurry types and results of

physical tests (Boonyarak, 2002)

Slurry
Type

Mixed ratios of
slurry (percentage to
water by weight)
Polymer Bentonite
(%)
(%)

Physical test results

1
2
3
4
5
6
7
8
9
10

0.05
0.05
0.05
0.08
0.08
0.08
0.10
0.10
0.10
0.15

1
1.5
2
0
1
2
0
1
2
0

Marsh
Slurry
Density Funnel
(g/cm3) Viscosity (sec)
1.007
41.8
1.012
64.3
1.019
71.3
1.000
43.3
1.008
49.8
1.020
98.3
0.999
45.5
1.008
58.6
1.020
110
0.999
50.0

11
12
13
14
15
16

0.15
0.15
0.20
0.20
0.20
0

1
2
0
1
2
5

1.009
1.019
1.000
1.008
1.020
1.050

71.8
120.4
60.3
83.2
160.3
70

SLURRY

pH

FILTER CAKE
(SLURRY-SAND INTERFACE)
SAND

9.16
9.20
9.38
7.78
8.87
9.13
7.69
9.14
9.35
7.75
9.28
9.30
7.81
9.18
9.32
9.12

FILTRATE

Figure 3. Picture and schematic of filtration test

The filtration test results of Bangkok First Sand

layer is presented in Figure 4. An influence of filter
cake is evident in Figure 4 for pure bentonite slurry
(Bentonite 5%) - the filtrate volume was very low
and became constant within a relatively shorter period. For the pure polymer slurry, filtration volume
was high and no constant rate of filtration was observed until the target testing time. This is apparently due to the absence of filter cake formation. In
case of polymer-based slurry (polymer 0.05% with

179

bentonite 1%), filtration volume was significantly

reduced in comparison with pure polymer slurry as
can be observed in the Figure. This finding is consistent with the field application in actual bored pile
construction for deep-seated bored piles in Bangkok.
4000
Bentonite 5%

Filtrate Volume (cm )

3500

Polymer 0.05% Bentonite 1%

3000

Polymer 0.15%

2500
2000
1500
1000

The relationship between shear strength of interface and dosage of polymer in slurry is depicted in
Figure 6. Based on the regions in which the shear
strength values fall, the test results can be categorized in 3 groups interface of pure bentonite slurry,
pure polymer slurry and polymer-bentonite slurry.
The results indicate that interface of polymerbentonite slurry is significantly stronger than that of
pure bentonite and polymer. The pictures showing
the condition of interface produced by pure polymer
slurry and polymer-bentonite slurry can be seen in
Figure 7 (a) and (b) respectively. Stronger surface
condition of polymer-bentonite interface is evident
from these pictures.

500
0
0

10

20

30
Time (min)

40

50

60

Figure 4. Filtration of different slurry types through Bangkok

First Sand sample.

5.4 Shear strength of slurry-soil interface

Shear strength of slurry-sand interface formed by
different types of slurries was tested by means of
pocket penetrometer. These tests were conducted
immediately after completion of filtration test described above. Picture and schematic of pocket
penetrometer test are shown in Figure 5.

(a) Interface of slurry-sand (pure polymer)

Pocket
Penetrometer
slurry-sand
interface

Figure 7. Condition of slurry-soil interface after the filtration

test with (a) pure polymer slurry (b) polymer-bentonite slurry

Sand

Figure 5. Picture and schematic of pocket penetrometer test applied on slurry-soil interface

Interface Shear Strength (kN/m )

250

200
Polymer + Bentonite Slurry
Pure Bentonite
Slurry

150

Pure Polymer
Slurry

100

50

Bentonite 0%

Bentonite 1%

Bentonite 1.5%

Bentonite 2%

Bentonite 5%

0
0

0.05

0.1

0.15

0.2

(b) Interface of slurry-sand (polymer-bentonite)

0.25

Polymer Dosage (%)

Figure 6. Test results of penetrometer test applied on slurry-soil

interface

5.5 Friction tests

Though it is almost impossible to model the field
condition of shaft friction resistance influence by
slurry, attempt was made to study the frictional resistance of slurry-sand interface in the lab. A series
of friction tests were conducted using the apparatus
shown in Figure 8 following the procedure outlined
below.
After completion of pocket penetrometer test the
sample was cut in such a way that less disturbed
slurry-sand interface was obtained.
Cement-mortar plate was prepared few days prior
to the test
Set up the sample in such a way that slurry-sand
interface was placed (upside-down position) directly on the top of cement-mortar as depicted in
Figure 9 (a).

180

Applied a range of normal and shear loads as

shown in schematic of Figure 9.
Recorded loads and displacement at regular
interval until friction load reached maximum
value or displacement reached 25% strain
The test was also conducted on pure sand
sample directly placed on the cement-mortar
following above procedure to make a rough
comparison of frictional resistance between
slurry-sand interface and pure sand against
mortar plate.

Frictional resistance of slurry-sand interface of

polymer-bentonite (polymer 0.1% + bentonite
1.5%) is slightly higher than that of pure bentonite (5% bentonite).
The difference of frictional resistance of slurrysand interface between polymer-bentonite (polymer 0.1% + bentonite 1.5%) and pure bentonite
(5% bentonite) was far less than expected. Likely
factor contributed to this result is that the contact
time between the bentonite slurry and sand sample was not long enough to develop an excessively thick filter cake.

Max. Friction (kN/m2)

100

2nd sand-mortar
polymer 0.1%+bentonit 1.5%

80

bentonite 5%

60
40
20
0
0

50

100

150

200

Normal Stress (kN/m2)

Figure 8. Laboratory friction test apparatus

N o rm a l
L oad
s lu rr y -s a n d
in te r fa c e

Figure 10. Friction test results of 3 different types of samples

S and

6 FIELD APPLICATION AND BEHAVIOUR OF

POLYMER-BASED SLURRY IN BANGKOK

F ric tio n
L oad
C e m e n t-m o rta r

(a)

(b)
Figure 9. (a) Schematic demonstration of friction test (b) photograph of friction test

The relationship between normal stress and maximum friction stress of different samples are plotted
and presented in Figure 10. Following points can be
drawn from this Figure.
Frictional resistance of slurry-soil interface of
both pure bentonite (5% bentonite) and polymerbentonite (polymer 0.1% + bentonite 1.5%) were
less than that of sand-mortar suggested that frictional resistance was reduced by application of
slurry.

181

In Bangkok, the primary reason for switching from

classic pure bentonite slurry to polymer-based slurry
in most cases was to minimize the problems associated with environmental issues caused by the use of
bentonite slurry. In the initial stage of using polymer-based slurry, the major concerns were the potential negative impact on stability of the borehole
and the shaft friction capacity as this alternative
slurry was not yet accustomed to the bored piling industry in Bangkok. However, no adverse effects
from the use of polymer-based slurry have been experienced to date and numbers of projects have been
successfully completed with this environmental
friendly drilling slurry.
6.1 Properties of polymer-based slurry
It is important that properties of slurry used are
within the applicable ranges to avoid the detrimental
effect on the performance of bored piles. ONeill &
Reese (1999) cited that there is no perfect set of
slurry specifications that can be used on every job,
but specifications should be tailored to fit the re-

quirements of a particular job at a particular location, where possible. Today, though use of polymer
slurry is widely acceptable in bored piling industry,
information on the desirable properties is still extremely limited. Table 2 shows the ranges of properties of polymer slurries specified by the International
Association of Foundation Drilling Standards and
Specification Committee (ADSC, 1999), those successfully used on several projects as reported by
ONeill and Reese (1999) from the work of Majano
et al. (1994) and the properties commonly applied
for bored piles in Bangkok subsoil.
Table 2. Properties of polymer slurries specified by ADSC and
Polymer-based slurry used in Bangkok
Property

Majano et al., 1994

ADSC
Emulsified Dry
(at 20 C) PHPA
PHPA
(Vinyl)
1.02
0.995 to
0.995 to
Maximum 1.01
1.01
40 90*
33 - 45
50 - 120

Common
Practice in
Bangkok
(PBS**)
1.05
Maximum
40 to 90
(+/- 5)
8 11
Less than
1%

No specific method or formula is available to determine the required best composition of polymerbentonite ratios. Therefore, at each site, different
mixed-ratios should be made on trial in the initial
stage to determine the most appropriate ratio. Table
3 presents the mixed composition of polymer-based
slurry of bored piles constructed at 5 different locations in Bangkok. According to the load-settlement
curves obtained from static pile load tests, observed
settlements of all piles were less than those of predicted values using bentonite pile parameters Thasnanipan et al. (2002).
Table 3. Technical data of polymer-based slurry constructed at
different localities in Bangkok
Description

Density
(g/cc)
Viscosity
(sec/qt)
pH
7 - 12
8 11.7
8 11.7
API Sand 1.0
0 to 1
0 to 1
Content
Maximum
*or as recommended by manufacturer and approved by geotechnical engineer, ** Polymer-based slurry

6.2 Mixed-ratio and procedure of mixing

Dry anionic partially hydrolyzed polyacrylamide
(PHPA) polymer powder premixed with fresh water
and a small percentage of bentonite is mainly used
for long bored piles in Bangkok. First, a small dosage of bentonite powder is premixed with fresh water (bentonite 5-15kg per cubic meter of water) and
leave in the storage tank for minimum 12 hours. The
predetermined dosage of dry polymer (typically
polymer powder 0.5kg per cubic meter of premixed
bentonite slurry) is then added and mixed to get so
called polymer-based slurry.
ONeill and Reese (1999) cited that vigorous mixing of polymers supplied in the dry form should be
avoided, since polymer chains can be broken down
and the polymer slurry rendered ineffective. Polymer-based slurry is stored in the steel tanks before
supplying to the bored pile drilling. The reason for
adding bentonite for long bored piles is to minimize
the fluid loss (high filtration of slurry into formation
soils), which was observed from the trial boreholes
drilled only with pure polymer slurry. Depending on
the thickness and permeability of sand layers present
at site and the embedded length of the pile in sand
layers, addition of bentonite may not be necessary.

Pile Diameter (m)

Pile Length (m)
Mixed Ratio per P*
1m3 slurry (% in
B**
weight)
Total standing time
prior to concreting
(Hr : Mins)
Density
(g/ml)
ProperViscosity
ties of
(sec/qt)
slurry
prior to
pH
concretSand
ing
content
(%)
Approx.
First sand
Pile embedded
in sand
Second
layers
sand
(m)
* polymer, **Bentonite

Pile ID
T1
T2
T3
T4
T5
0.80 0.80 1.20 1.80 1.50
49
50
51
62
54
0.05
0.05 0.05 0.05 0.05
1.2

1.2

1.0

1.0

1.2

17:20

12:20

14:00 23:00

15:30

1.06

1.03

1.01

1.04

1.04

40

45

44

51

44

0.5

0.5

12

20

13

18

13

6.3 Effective base cleaning for polymer-based

bored piles in Bangkok
Since bentonite and polymer slurry react differently
with the soil layers as described in section 4 above,
base cleaning method for two different slurries
should be selected based on their behaviour. ONeill
& Reese (1999) pointed out that vigorous mechanical agitation such as pumping by centrifugal pumps
can shear the polymer chains severely. As soilpolymer reaction system is largely depending on the
polymer chemical chain structure, excessive breaking of it would affect the performance of the slurry.
Therefore, high turbulent means of base cleaning
methods such as recycling by centrifugal pumps
should not be performed for the polymer-based
bored piles. Field experience shows that base cleaning by air lift method cannot effectively clean thick

182

sediments at the base of the polymer-based bored

pile since the sediments deposited at the base of the
pile are comparatively dense for bored piles constructed with polymer-based slurry. According to the
field observation on polymer-based bored piles, rate
of sedimentation at pile base was generally found to
be stabilized few hours after stopping the drilling
operation. Figure 11 presents the thickness of the
sediments at base of polymer-based bored piles recorded at regular interval. The records shown in
Figure 11 were made at the same site in Bangkok for
diameter 1000mm bored piles of about 55m in
length. The soil profile at the site is similar to that of
Figure 1, a typical subsoil profile of Bangkok. This
sedimentation behaviour of polymer-based bored
pile should be taken account in base cleaning
method and process.
40
35
30
25
Pile No.1
Pile No.2
Pile No.3
Pile No.4
Pile No.5

15
10
5
0
0

50

100

150

Soft Clay
10

Unit Skin Friction (kN/m2)

Load (kN)

20

6000 12000 18000

10

10

20

20

30

30

40

40

50

50

60

60

50

100

150

200

250

Stiff Clay
Depth (m.)

Thickness of sediment (cm)

45

slurry in the USA are higher than those of bentonite

piles. Similar finding was reported by Thasnanipan
et. al. (2002) from the results of instrumented test
pile in Bangkok.
Load distribution curves along the pile shaft at
various applied loads of instrumented test pile (T3 of
Table 3) constructed by polymer-based slurry is illustrated in Figure 12. The values of average unit
skin friction developed by the maximum test load at
the different soil layers along the shaft of test piles
in comparison with those of ultimate unit skin friction calculated using the empirical formula are also
illustrated in these figures. As can be seen in the figure, developed average unit shaft friction in sand
layers are significantly higher than those of theoretically calculated values using normally used formula
and parameters applied for bentonite bored piles in
Bangkok. This finding suggested that polymer-based
slurry bored piles produced better shaft friction resistance in sand layers than that of bentonite bored
piles.

200

Time after stopping drilling operation (min)

20

30

V. Stiff to
Hard clay

Dense Sand
Hard Clay

Figure 11. Sediments thickness with time at pile base recorded

from polymer-based bored piles

40

50

The effective base cleaning for long bored piles in

Bangkok, may be achieved by the following procedure.
Carry out drilling to 1 to 2m above the final design depth under polymer-based slurry for the
first bored pile
Leave the borehole for 2 to 8 hours (depending on
the rate and stabilization trend of sedimentation
observed from trail borehole at each site)
Move the drilling rig to commence the drilling of
second bored pile
Drilling rig come back to the first bored pile and
carry out final drilling for (remaining 1 to 2m)
with the cleaning bucket
Immediate installation of reinforcement cage and
concrete pouring
6.4 Shaft friction resistance of polymer-based bored
piles
Ata and ONeill (1998) concluded that unit shaft resistances of test piles constructed with polymer

183

60

Dense to
Very Dense
Sand

Mobilized at 18000 kN
Qult (calculated)

Figure 12. Load distribution curves and unit skin friction of instrumented test bored pile diameter 1.20m, 51m deep constructed by polymer-based slurry in Bangkok

7 DISCUSSION ON LAB TEST RESULTS AND

FILED PERFORMANCE
Polymer-based slurry used in construction of bored
piles in Bangkok is commonly composed of 0.05%
and 1% (by weight per 1m3 slurry) of polymer and
bentonite respectively. The performance of bored
piles constructed with this composition was proved
to be significantly better than that of bored piles
constructed with pure bentonite slurry. Using pure
polymer in the field, construction of deep-seated
bored piles in multi-layered soil of Bangkok is found
to be impractical particularly at the site where thick
sand layer presence due to the excessive loss of
slurry into the formation. Table 4 presents the findings of laboratory tests in comparison with field observation. Though perfect comparison of laboratory
findings and field observation is not possible, Table

4 provides some useful summary on the behaviour

of slurry in the laboratory and of practical application in the field.

which is proved to be performing better than classic

pure bentonite slurry in both construction and loading capacity aspects.

Table 4. Summarized comparison of parameters between lab

findings and field observation

REFERENCES

Parameters Laboratory
test
Filtration
Polymer
(0.05%) +
bentonite
(1%) significantly
reduces filtrate volume than
pure polymer
Shear
Slurry prostrength of duced by
slurrypolymer
sand inter- mixed with
face
small percentage of
bentonite has
significant
higher
strength than
pure polymer
or bentonite

Field
observation
Polymer
(0.05%) +
bentonite
(1%) significantly reduces filtrate
volume
(slurry loss)

Discussion
notes
Laboratory
results agree
well with
field observation

Slurry produced by
polymer
mixed with
small percentage of
bentonite has
significant
higher shaft
friction capacity than
that of pure
bentonite
Frictional Slurry proSlurry proresistance duced by
duced by
of slurry- polymer
polymer
sand inter- mixed with
mixed with
face
small persmall percentage of
centage of
bentonite has bentonite has
higher
significant
strength
higher shaft
(slightly) than friction capure polymer pacity than
or bentonite
that of pure
bentonite

Indirect comparison of lab

shear strength
vs. actual
shaft friction
measured
from instrumented pile
load test
agrees well

ADSC Standards and Specifications for the Foundation Drilling Industry. 1999, ADSC, Revised 1999, Incorporating
ACI 336.1R-98
Ata A. A. and ONeill M. 1998. Side-wall stability and sideshear resistance in bored piles constructed with highmolecular-weight polymer slurry, Third International Geotechnical Seminar on Deep Foundation on Bored and Auger Piles, Ghent, Belgium, 19-21 October 1998 : 111-117.
Balkema : Rotterdam,.
Boonyarak T. , Behaviour of Polymer Slurry for Wet-process
bored Piles Construction in Bangkok Subsoil, Master Thesis (in Thai), Chulalongkorn University, 2002
Majano R. E. and ONeill M. W. 1993. Effect of mineral and
polymer slurries on perimeter load transfer in drilled
shafts, A Report to ADSC, University of Houston, Dept. of
Civil Engineering, UHCE.
ONeill M. W. and Reese L. C. 1999. Drilled Shafts : Construction Procedures and Design Methods, U.S Department
of Transportation Federal Highway Administration, Publication No. FHWA-IF-99-025, ADSC Publication No.
ADSC-TL 4.
Thasnanipan, N., Aye, Z. Z., Submaneewong, C. and Teparaksa, W. 2002. Performance of Wet-Process Bored Piles
Constructed with Polymer-Based Slurry in Bangkok Subsoil, Proc. of the International Deep Foundations Congress
2002, Geotechnical Special Publication No. 116, Volume
One, American Society of Civil Engineers, February 14-16
2002 : 143-157. Orlando, Florida, USA : ASCE

Reduction of
frictional resistance is not
significant in
laboratory. In
lab, contact
time between
bentonite
slurry and
sand sample
was not long
enough to develop an excessively
thick filter
cake

8 CONCLUSION
Behaviour of polymer-based slurries in both laboratory and field have been presented. The laboratory
findings of this research work provided backup and
useful information for the actual construction of
bored piles in the field. The laboratory model tests
reveal that significant strength improvement can be
achieved in slurry-sand interface if polymer slurry
(polymer 0.05% by weight) is mixed with a small
percentage of bentonite (1% by weight). This mix
composition significantly reduced the filtrate volume
in the laboratory test in comparison with that of pure
polymer. Similar dosage of bentonite is commonly
added in preparing polymer-based slurry for construction of deep-seated bored piles in Bangkok

184

G EO T ECHN ICA L SPEC IAL PUBL IC AT ION NO . 124

GEOSUPPORT 2004

DRILLED SHAFTS, MICROPILING, DEEP

MIXING, REMEDIAL METHODS, AND
SPECIALTY FOUNDATION SYSTEMS

PROCEEDINGS OF SESSIONS OF THE GEOSUPPORT CONFERENCE :

INNOVATION AND COOPERATION IN THE GEO-INDUSTRY

January 29-31, 2004

Orlando, Florida

SPONSORED BY
International Associations of Foundation Drilling (ADSC)
The Geo-Institute of the American Society of Civil Engineers

EDITED BY
John P. Turner
Paul W. Mayne

Effectiveness of Toe-Grouting for Deep-Seated Bored Piles in Bangkok Subsoil

Narong Thasnanipan, Zaw Zaw Aye, Chanchai Submaneewong

185

Effectiveness of Toe-Grouting for Deep-Seated Bored Piles in Bangkok Subsoil

Narong Thasnanipan1, Zaw Zaw Aye2, Chanchai Submaneewong3
ABSTRACT: A comprehensive study on the effectiveness of two different toegrouting methods, known as tube--manchette and drill-and-grout, commonly
applied for deep-seated bored piles in Bangkok is presented in this paper. Mobilized
unit end bearing of bored piles constructed by two toe-grouting methods are
compared. The problems encountered in practical application of different methods
for large diameter deep seated bored piles are highlighted. Despite considerable extra
cost, various construction problems encountered and risks involved in successful
execution of grouting, drill-and-grout method does not offer particular advantage
over tube--manchette method as far as pile capacity improvement is concerned.
INTRODUCTION
In Bangkok, due to the prevailing subsoil and groundwater conditions, all the
deep-seated bored piles (depth ranging from 25 to over 60 m) are constructed by wetprocessed or slurry displacement method. Although the wet-processed method is
most suitable and its application has been well established in Bangkok subsoil, the
method itself cannot avoid undesired effects especially loosening and upheaval of
soil at pile base as well as accumulated suspension of loose materials or sediments
from drilling operation. These effects naturally may create a loose or soft pile toe
causing negative impact on the pile capacity or load vs. pile head movement
behavior. One of the special measures to improve the soft pile toe is to apply the
post-grouting at pile toe. Two different methods, known as tube--manchette and
drill-and-grout are commonly applied in Bangkok. A comprehensive study on the
effectiveness of two toe-grouting techniques applied for deep-seated bored piles in
Bangkok is presented in this paper.

Managing Director, 2Project Manager, 3Geotechnical Engineer

SEAFCO Company Limited, Bangkok, Thailand, [email protected]

187

SUBSOIL PROFILE AND EXISTING PIEZOMETRIC PROFILE

Subsoil profile is relatively consistent at different localities in Bangkok. A
typical subsoil profile is characterized by the alternating layers of clay and sand
deposits as shown in Figure 1. Pore water pressure profile illustrating the current
piezometric drawdown condition of Bangkok is also shown in Figure 1.
Pore Pressure (kN/m 2)

Geotechnical
Parameters
0

10

400

600

Su= 10-25 KPa

t = 14-16kN/m 3
-

10

Medium to
Stiff Clay

30

Medium
Dense Sand
(First Sand)

40

Hard Clay

50

Dense to Very
Dense Sand
(Second Sand)

Su= 50-14020-KPa
t = 17-21kN/m 3
-

30
SPT-N= 20-40
t = 20kN/m 3

Su> 200 KPa

= 21KN/m 3
t

SPT-N= 50-100
t = 21KN/m 3

60

40

50

Line
tatic
ros
ine
Hyd
nL
dow
raw
cD
etri
zom
Pie
ual
Act

20

Depth (m)

200

Weathered Crust
Bangkok Soft
Clay

60

Figure 1. Typical soil profile of Bangkok with piezometric drawdown

condition (Thasnanipan, 2002)
WET-PROCESS BORED PILE CONSTRUCTION METHOD
In Bangkok, medium to large diameter (0.6 m to 1.8 m diameter) bored piles
with depth ranging from 24 m to over 60 m are constructed by wet process due to the
prevailing subsoil and ground water condition. Temporary casing of 14-18 m in
length is typically used as a support in soft clay layer. Drilling is commenced by dry
process with auger applying rotary drilling action in the soft clay layer and stiff clay
layer. Before reaching to the first sand layer, supporting fluid or slurry is fed to the
borehole and drilling is continued with the bucket to the final depth. Depending on
the type of slurry used, either a cleaning bucket or airlift method is commonly
applied d to clean the bottom of the borehole prior to reinforcement cage installation.
Concreting is done by tremie method for wet-processed bored piles.
TOE GROUTING METHODS IN BANGKOK
Two grouting methods known as tube--manchette and drill-and-grout
method used in some major projects where the authors were involved are presented
in this paper. In Bangkok, both toe-grouting methods are mainly applied for the piles
with the tip embedded in the sand layer though there were few cases where it was
applied in clay layer.

188

Tube--manchette Toe-grouting Method (TAM Method)

This grouting method has been widely used in other parts of the world
according to the available published papers. A review of past and present toegrouting methods used throughout the world has been reported by Mullins et. al.
(2000). In Bangkok, the first toe-grouting for large diameter bored piles diameter
2.0m seated at depth 35m in the first sand layer was carried out by tube-a-manchette
method in 1985 for the construction of Rama IX cable stayed bridge (Morrison,
1987). Since then, tube-a-manchette method has been commonly used for toegrouting of wet-processed bored piles in Bangkok. It is estimated that over 5,000
bored piles have been constructed using this toe-grouting method in Bangkok.
Grout pipes configuration for tube--manchette method. In normal practice of toegrouting by this method in Bangkok, a system of grouting circuit consists of two Ushape loops formed by two pairs of PE pipes with manchette placed at 5 to 10 cm
above pile tip. The manchette is formed by the perforated steel pipe wrapped with
rubber sleeves as shown in Figure 2. The manchette allows the grouting material to
flow out of the pipe from the perforations but prevents return of grouting material
when the grout injection is stopped. PE tubes are fixed along the rebar cages to the
top. In some projects where the specification required cross-hole sonic logging tests,
same access tubes for the sonic test apparatus are used in lieu of PE pipes.

Pile Head

Vertical Steel Tubes or

Flexible PE Pipes

2 No. Horizontal
Steel Tubes with
perforation

Bored Pile
Wall

Rubber Sleeve
(Manchette)

Reinforcement
Cage

Perimeter of Bored
Pile (Bored Pile
Wall)

Pile Base
Horizontal
Steel Tube

Vertical Steel Tubes or

Flexible PE Pipes

Rubber
Sleeve

Figure 2. Position and layout configuration of grouting circuit in

Tube--manchette method

189

Grout Mixture Composition and Mixing Process. The typical cement grout consists
of cement and water with the ratio of 0.5-0.6 by weight. Bentonite or other reagent
can also be added but trial mix should be made to find out the characteristics of the
grout. Cement and water are added at right proportion in a high turbulence mixer and
mixed about 2 to 4 minutes. Cement mixture is then conveyed to the agitator. The
grout must be agitated continuously or at regular interval during the period between
the time of grout mixing and the time it enters the pump. Usually, first grout injection
should be started within 10 to 15 minutes after mixing the first batch to minimize the
stiffening of the cement grout during grouting operation. It is essential that grouting
team understands the required properties such as viscosity, density, setting time of
grout are met to effectively grout the toe of the pile. Grouting operation should be
completed within the initial setting time of the grout. Mixing rate, storage capacity
and grouting rate should be well planned to complete the grouting work within the
time limit.
Grouting procedure. The process involved in grouting by tube-a-manchettes method
is illustrated in Figure 3. First, the grout pipes are flushed with water to ensure that
the grouting loops are clear from any blockage. The pile toe is then cracked with
relatively high pressure water flow, 12 to 24 hours after concreting, to open the
manchettes and to make way for forth coming grout. The commencement of cement
grouting depending on the preset pressure criteria. If specified maximum pressure is
low (less than 20 bar / 2000 kN/m2), grouting is commenced immediately after the
pile toe cracking. Grouting is normally commenced 7 to 10 days after concreting if
higher maximum control pressure is specified to avoid damage to pile by high
pressure grouting. Grouting is stopped when the specified volume or pressure is
achieved. In current practice, pressures ranging from 20 to 60 bar (2000 to 6000
kN/m2) and grout volumes 500 to 1000 liters are normally applied.
IN

OUT

(1)
Flushing grout
tubes with water

WATER
INJECTION

GROUT UNDER
PRESSURE

(2)
Cracking with
water

(3)
Pressurized
grouting

(4)
Grouted pile base

Figure 3. Grouting process in Tube--manchette method

190

Drill-and-grout Method (DAG Method)

Toe-grouting by this method is in fact not frequently used in Bangkok. It is to
the authors knowledge, not commonly used in the current practice of toe-grouting of
bored piles world wide. To date, less than 500 bored piles have been completed in
Bangkok by this method.
Grout pipes configuration for drill-and-grout method. Depending on the size of
bored piles, three to four black steel pipes used for sonic test access tubes are mostly
utilized as grouting pipes in this method. Steel pipes are equally spaced and fixed
around inside perimeter of reinforcement cage as shown in Figure 4. The bottom of
steel pipes are securely closed with steel cap to prevent intrusion of foreign material
during installation and concreting.
Pile Head

Vertical Steel
Tubes

4 No. Vertical Steel

Tubes closed at
bottom-end

Pile
Wall
Reinforcement
Cage

Pile Base

Perimeter of Bored Pile

(Bored Pile Wall)

Pile Toe Coring

(After Concreting
and Sonic Testing)

Figure 4. Position and layout configuration of grouting circuit in

Drill-and-grout method
Grout Mixture Composition and Mixing Process. Cement to water ratio of 0.6-0.7
by weight is normally used in drill-and-grout method. Cement and water are filled at
right proportion and mixed about 2 to 4 minutes in a high turbulence mixer.
Bentonite is then added at 0.5 % of cement in weight and mixed for another 2
minutes and the readily mixed grout is conveyed to the storage tank equipped with
agitator. The control factors such as viscosity, density and setting time are also
equally important as mentioned in the tube--manchette method.
Grouting procedure. The procedure for drill-and-grout method presented in this
paper is for diameter 1.50m bored pile with 4 sonic access tubes. The sonic logging
test is performed minimum 7 days after casting of the pile. Coring is then carried out
through the pile base to the depth 20 cm below the pile toe. First, drilling rod
attached with core-barrel is lowered into the water-filled sonic access tubes (steel
pipes) and drilled through the bottom steel cap and the concrete at pile base. Drill rod
is then withdrawn to change core-barrel with the soil sampler and lower again to drill
the soil beneath the pile base about 20 cm. It is important that sonic tube is
completely filled with water during the drilling process to avoid excessive blow-in of
soil into the tubes. After completion of toe coring and sampling, the grout pipes are

191

flushed in a sequence to ensure that flow path exists between the tubes. To do this,
after initial cleaning, two pipes (inlet pipes) are connected to the pressure pump and
low pressure up to 10 bar (10 kN/2) is applied using clean water. The other two pipes
(outlet pipes) are filled with water and are left open. As the pressure from the pump
is gradually increased, water will start to flow from one or more of the other two
open pipes. Once the flow is observed from the first outlet pipe, it is closed off and
the water flushing continued until water flow is observed from the second pipe. The
process may be repeated in like sequence from each of the grout pipes until
satisfactory flow is observed from each pipe. Once the flushing of the preinstalled
grout pipes are completed, water-supply hoses are disconnected from 2 inlet grout
pipes and connected with the grout-supply hoses. Two outlet pipes are then opened
and cement grout is filled from two inlet pipes with low pressure until the grout
discharge from other 2 outlet pipes.
IN

OUT

GROUT
FILLING

GROUT UNDER
PRESSURE

PILE
TOE CORING

(1)
Pile toe coring

(2)
Flushing grout
pipes with water

(3)
Filling cement
grout in pipes

(4)
Pressurized
grouting

Figure 5. Grouting process in Drill-and-grout method

Initial discharged cement grout is normally contaminated with the in-situ
sand so that it is required to check the purity of outcoming grout. Once clean cement
grout is flowed out, 2 inlet pipes are immediately connected with the grout-supply
hoses and pressurized grouting commences. Grouting is started with the grout
injection rate (IR) between 25 to 30 liter per minute. Grout pressure development is
closely monitored from the pressure gauges installed at the top of each pipe. If the
grouting pressure does not reach to 20 bar (2000 kN/m2) after injecting the 300 to
500 liter of grout, the injection rate is reduced to 12-15 liter per minute. Otherwise
grouting is continued until the control criteria is achieved. Figure 5 illustrates the
working procedure of pile toe grouting by drill-and-grout method used in Bangkok
subsoil for 3 major projects where the authors involved.
Grout Spreading Mechanism
Tube--manchette method. Teparaksa et al. (1999) investigated the penetration of
grout into the sand layer from the toe-grouting of bored piles in Bangkok by tube-manchette method. Toe-grouting was performed 24 hours after concreting with

192

maximum pressure 40 bar (4000 kN/m2) for grouting the base of large diameter
bored piles constructed by wet-processed method under bentonite slurry seated in
sand layer. From the series of SPT and soil sampling performed below the pile toe
and immediate vicinity of the shaft, the authors concluded that grout did not
permeate into the sand layer at pile base. Upward migration of grout along the pile
shaft is most likely in tube--manchette method, as grout injection was applied under
relatively high pressure within 24 hours after concreting of the piles, plane of
weakness was formed at pile-soil interface which is most likely to be contributed by
the following conditions.
Stress relaxation from the drilling process was not yet fully restored along the
shaft as time gap between pile drilling and grouting was relatively short
Presence of bentonite filter-cake along the shaft
Drill-and-grout method. Grout spreading mechanism of drill-and-grout method are
discussed in relation to the steps involved in grouting operation (illustrated in Figure
5) as summarized below.
Step 1 : drilling 4 boreholes of diameter 48 mm up to 20 cm below pile toe
(using water-filled sonic tubes as access pipes to the bottom of the pile)
Step 2 : water flushing the sonic tubes to remove the sand inside the tubes and to
form flow-paths between the tubes
Step 3 : filling the sonic tubes with the grout
Step 4 : pressurized grouting until pressure reaches 6,000 kN/m2 and maintain
for 5 minutes
Step 1 creates 4 boreholes in the soil underneath the pile base. These
boreholes become larger as pocket of sand between them is washed out during Step 2
and created a cavity or relatively large opening beneath the pile base prior to the
commencement of Step 3. This cavity is filled by grout in Step 3 which eventually
formed a mass grout bulb underneath the pile base as illustrated in Figure 5.
Pressurized grout injection in Step 4 thus compacted the grouted mass which in turn
densified the grout-soil interface. Basically grout did not enter soil pores but
remained in a compacted homogeneous mass. Upward migration of grout along the
pile shaft is considered unlikely since toe-grouting was carried out minimum 14 days
after pile concreting so that time gap between concreting and grouting would have
allowed some positive improvement of pile shaft condition (influence of stress
relaxation and bentonite filter cake).
PERFORMANCE OF TOE-GROUTED BORED PILES
Typical Load Transfer Mechanism of Toe-Grouted Bored Piles in Bangkok
Table 1 shows the measured mobilized shaft resistance of bored piles with
different toe grouting method at specific load in comparison with design working
load and calculated ultimate values using empirical formula. The details of the
empirical formula commonly used in Bangkok subsoil has been presented by

193

Thasnanipan et. al. (2002). As can be seen in the table under design working load,
shaft friction carried the large portion of the load (over 90% of applied load). Only
the small percentage of the total load was transferred to the pile tip. Therefore, even
with the toe-grouting application, the end bearing component of bored piles is still
under-utilized and thus remaining unmobilized end bearing capacity serves only as a
reserve that forms the additional factor of safety.
Table 1. Summary of mobilized shaft resistance at specific load in comparison with
design working load and calculated ultimate value
Test
Pile
No.
TT-1
TT-2
TT-3
TD-1
TD-2
TD-3

Total shaft resistance (KN)

Ratio of
Design
shaft
Grouting
Working
/
Q
fm
Pile Dia.
Calculated Measured Measured
resistance
Load,
DL
method
DL
at DL,
at 2 x DL,
xL
Ultimate
(kN)
(Qfm2 / Qfu)
(m)
Qfm
Qfm2
Value, Qfu
1.2 x 44.6
5000
6900
4480
8560
0.90
1.24
TAM
1.2 x 53.8
6500
9500
6120
10830
0.94
1.14
1.2 x 46.0
5000
7900
4730
9250
0.95
1.17
1.50 x 50.0
13000
14600
12200
22160
0.94
1.52
1.50 x 56.5 DAG
13000
16500
12590
24270
0.97
1.47
1.50 x 52.0
13000
16400
12570
24410
0.97
1.49

Mobilized Unit End Bearing

Figure 6 shows the mobilized unit end bearing vs. pile head movement of
TAM and DAG grouted bored piles. In general, as can be seen in the figure,
mobilized unit end bearing at specific pile head movement of piles constructed with
tube--manchette grouting method (TAM) is higher than those of piles constructed
with drill-and-grout method (DAG). It is also evident from Figure 6 that large pile
head movement is required to mobilize the end bearing despite toe-grouting was
applied for all piles. It should however to be noted that DAG grouted bored piles of
Figure 6 (TD1, TD2 and TD3) were not tested to failure.
Unit end bearing (kN/m2)

3000
2500
TT1 (TAM)

2000

TT2 (TAM)

1500

TT3 (TAM)

1000

TD1 (DAG)
TD2 (DAG)

500

TD3 (DAG)

0
0

10

15

20

25

Pile head movement (mm)

Figure 6. Mobilized unit end bearing of TAM and DAG grouted bored piles

194

Though it is not advisable to simply draw the conclusion on particular toegrouting method effects on performance of piles solely by mobilized unit end bearing
criteria from the available limited test results, the data provided in this research study
however could be taken as preliminary guidance to assess the effectiveness of the
different toe-grouting methods.
According to the recent full-scale instrumented static pile load tests carried
out in late 2002 and early 2003 on 13 drill-and-grout bored piles of diameter 1.50m
seated at depth over 50m (a complete set of data is not available while preparing this
paper), developed end bearing values of these piles were found to be scattered or
inconsistent despite the fact that grouting pressure of all piles were the same reached 60 bars (6,000 KN/m2). Therefore in authors opinion a careful justification
is to be made to design the deep-seated bored piles in Bangkok subsoil with high end
bearing values regardless of high pressure achievement in toe-grouting application.
Load vs. Pile Head Movement Comparison
Figure 7 shows the Load vs. Pile Head Movement of ungrouted, grouted
piles of tube--manchette and drill-and-grout methods. Table 2 summarized and
compared the load test results of the test piles of Figure 7.
Load (KN)
0

10000

20000

30000

40000

Pile Head Movement (mm)

0
10
20
30
40

A1 - Dia. 1.5x60.0m (TAM)

A2 - Dia. 1.5x60.0m (Ungrout)
B1 - Dia. 1.5x63.7m (TAM)
B2 - Dia. 1.5x63.7m (Ungrout)

50

C1 - Dia. 1.5x56.5m (DAG)

D1 - Dia. 1.5x52.0m (DAG)

60

Figure 7. Load vs. pile head movement of ungrouted and toe-grouted piles
Following conclusions can be drawn from Figure 7 and Table 2.
Limit loads at 20mm pile head movement are improved (14% and 21% for Project
A and B respectively) by tube-a-manchette toe grouting method (TAM) in
comparison with ungrouted piles.
No significant difference of load-settlement behavior is found for ungrouted piles
in comparison with TAM grouted and DAG grouted piles at design load and 2
times of design load.

195

Table 2. Summarized comparison of load test results

Diameter (m)
Length (m)
Qd (KN)
QL
(at 20mm total
settlement)
% increase in QL
Total pile
at Qd
head
1.5 Qd
movement 2 Qd
at

Project A
Project B
Project C
Pile No. A1 Pile No. A2 Pile No. B1 Pile No. B2 Pile No. C1
(TAM Grout) (Ungrout) (TAM Grout) (Ungrout) (DAG Grout)
1500
1500
1500
1500
1500
60.00
60.50
63.70
63.70
56.50
10,000
10,000
10,000
10,000
13,000
26,500
3.50
6.50
10.50

14.20

23,200

34,280

3.00
6.50
12.00

4.20
7.20
10.00

21.57

Project D
Pile No. D1
(DAG Grout)
1500
52.00
13,000

21,750

25,500

26,000*

3.75
6.00
9.70

N.A
6.10
11.90
21.50

N.A
5.80
11.00
17.50

Note : * At 17.50mm pile head movement

Therefore, despite the various construction problems encountered and risks

involved in successful execution of grouting, drill-and-grout method does not offer
particular advantages over tube--manchette method.
EFFECTIVENESS OF THE DIFFERENT METHOD IN PRACTICAL
APPLICATION
Problems Encountered in Practical Application of Different Method
Table 3 summarizes the most critical problems encountered in practical
application of different toe-grouting methods used in Bangkok. In summary, it is
obvious that higher risks are involved in the drill-and-grout method. The most critical
risk of this method for the long piles is to maintain a good verticality of the grout
access tubes.
Table 3. Summaries of the problems encountered in practical application of different
methods
Problems Encountered and Risks Involved to Complete
Toe-grouting
Toe Grouting
Unsuccessful or Verticality of sonic /
Unsuccessful
Method
ineffective
grout tube
coring of pile toe
cracking of
manchette
Tube-Possible but
Not required
Not required
manchette
rarely occur
Drill-and-grout Not required to
High verticality is
High risk. If coring
crack manchette extremely important. If
is unsuccessful, toe
access tube is inclined
grouting cant be
drilling rod can not be
performed.
lowered to the bottom of
the pipe for toe coring

196

Impact Caused by Different Toe-grouting Method on Bored Pile Construction

The procedure involved in preparation and execution of two methods of toegrouting poses different impact on bored pile construction. Since verticality is the
main governing criteria in the successful application of grouting by drill-and-grout
method as explained in above section the major impact caused in this method is due
to the need of significant extra time in the following activities to produce the good
verticality of the pile itself and the grout access tubes.
Drilling
Extensive borehole monitoring for verticality by sonic caliper method
Grout pipe installation
Due to the excessive time consumption in the above mentioned operations in
construction of the piles to further carry out the toe-grouting by drill-and-grout
method, to produce 1 bored pile of diameter 1.50 m seated over 50m below ground,
could take 2 to 3 days excluding time consumes for grouting work. For normal bored
pile of same size and length, it normally takes less than 24 hours to produce 1 pile.
Therefore considerable additional cost would incur for applying drill-and-grout
method due to the requirement of significant extra construction time.
Assessment on the Effective Toe-grouting Method in Practical Construction
Assessment on Constructibility. Constructibility is one of the most important issues
for most construction projects. If the construction method is impractical or if it is
involved various problems and high risk in actual execution, it should not be
employed in the first place. Careful justification on assessment of the constructibility
is essential in selection of the toe-grouting method and in establishing the technical
criteria. The party who is responsible for drawing the specification such as project
consultant or designer should take the following points into consideration.
The potential risks involve in the practical execution and consequential impact in
minimizing the risk imposed by the method itself (e.g. delay in construction, extra
cost)
Additional cost involved for toe-grouting against saving from higher bearing
capacity achivable by toe-grouting
Clear and practical acceptance and rejection criteria in case the problems are
encountered in actual execution
Assessment on the Performance of Pile. As presented in the earlier section, deepseated large diameter bored piles in Bangkok are mainly supported by shaft friction
and very small percentage of the load is transferred to pile toe under working load
and double of working load. Therefore, even with the toe-grouting application, the
end bearing component of bored piles is still under-utilized. Thus remaining
unmobilized end bearing capacity serves only as a reserve that forms the additional
factor of safety.

197

CONCLUSION
Based on the field application and measured data of instrumented pile load tests,
following conclusions can be drawn.
(1) Under design working load for large diameter deep-seated bored piles with toe
grouting in Bangkok subsoil, the large portion of load (over 90% of applied
load) is carried by the shaft friction. Only the small percentage of the total load
is transferred to the pile tip. No significant difference of load-settlement
behavior is found for non-grouted piles against TAM grouted and DAG grouted
piles at design load and 2 times of design load. Therefore, for deep-seated bored
piles in Bangkok subsoil, even with the toe grouting application, the end bearing
component of bored piles could be still under-utilized. Thus the remaining
unmobilized end bearing capacity serves only as a reserve which forms the
additional factor of safety.
(2) The drill-and-grout method has higher risks than the tube--manchette method in
successfully constructing the toe grouted bored pile. Considerable additional
cost would incur in applying drill-and-grout method due to the requirement of
significant extra construction time to complete a qualified pile.
(3) Despite the considerable extra cost, various construction problems encountered
and risks involved in successful execution of grouting, drill-and-grout method
does not offer particular advantage over tube--manchette method as far as pile
capacity improvement is concerned.
(4) The potential risks involve in the practical execution and consequential impact in
minimizing the risk imposed by the method itself should be taken into account in
selecting the effective toe-grouting method.
REFERENCES
Morrison, I.M. (1987) Bored piled foundation
for Chao Phya river crossing at Wat
Sai, Bangkok, Proceedings of the 9th South East Asian Geotechnical Society
Conference, Bangkok, pp. 6-207 to 6-218.
Mullins A. G., Dapp S. D. and Lai P. (2000) Pressure-Grouting drilled shaft tips in
sand New Technological and Design Developments in Deep Foundations,
Proceedings of the Sessions of Geo-Denver 2000, Geotechnical Special Pub. No.
100, ASCE, August 5-8, 2000, Denver, Colorado, USA, p.p. 1-17
Teparaksa W., Thasnanipan N., and Anwar M. A. (1999). Base grouting of wet
process bored piles in Bangkok subsoils 11th ARC. on soil mechanics and
geotechnical Engineering., Seoul, Korea, pp. 269-272.
Thasnanipan, N., Zaw Zaw Aye, Chanchai Submaneewong and Wanchai Teparaksa
(2002) Performance of wet-process bored piles constructed with polymer-based
slurry in Bangkok subsoil, Proceedings of the International Deep Foundations
Congress 2002, Geotechnical Special Publication No. 116, Volume One, ASCE,
February 14-16 2002, Orlando, Florida, USA, p.p. 143-157

198

ASCE Geo-Institute Geotechnical Special Publication in Honor of

Professor Gorge Gobble
27-31 July 2004, Los Angeles, California, USA

GeoTrans 2004

Current Practice and Future Trends of Deep

Foundations in Bangkok, Thailand
Narong Thasnanipan, Zaw Zaw Aye Kamol Singtogaw and
Sumate Pravesvararat
SEAFCO Co., Ltd. Bangkok, Thailand

199

CURRENT PRACTICE AND FUTURE TRENDS

OF DEEP FOUNDATIONS IN BANGKOK, THAILAND
Narong Thasnanipan, Zaw Zaw Aye, Kamol Singtogaw & Sumate Pravesvararat
SEAFCO Co., Ltd. , 26/10 Rarm Intra 109 Road, Bangchan, Khlong Sam Wah
Bangkok 10510, Tel. 02-919 0090, Fax. 02-9190098
E-mail : [email protected], Homepage : www.seafco.co.th
ABSTRACT
This paper presents the authors experiences in development of deep foundations,
bored piles and barrettes in particular in Thailand. Some literatures related to wet-processed
bored piles and barrettes in Thailand published through out the past 30 years are also
summarized together with recent research works.
KEYWORDS : bored pile, barrette, deep foundations, bentonite, polymer-based slurry
INTRODUCTION
Cast-in-place deep foundations, particularly wet-processed deep-seated large diameter
bored piles and barrettes have been used for foundations of various infrastructure projects in
Thailand for over two decades. The versatility of the construction method and the high-load
capacity which in turn offered the constructibility and cost-saving are the main factors
contributed to the increasing use of deep-seated large diameter bored piles and barrettes. In
the initial stage there were a number of questions on design and construction aspects of these
foundation systems particularly in Bangkok subsoil. With the passage of time, the
construction equipment, installation techniques and testing of deep foundation elements have
been developed. Large numbers of instrumented full-scale static pile load tests were
conducted through out 1990s which provided better understanding on behavior of these deepseated foundations. Researches focused on the design parameters and methods were produced
based on these test results. The design parameters were so well established that wet-process
cast-in-place foundations became regarded as reliable foundations for not only in the property
sectors but also for practicing engineers involved in the infrastructure projects. This paper
presents the development of wet-processed bored piles and barrettes in Thailand. The
information contained in the paper are mainly from the infrastructure developments built in
Bangkok and adjacent areas.
CONSTRUCTION METHOD
Large diameter deep-seated wet-processed bored piles can be constructed by 2
methods, reverse-circulation and rotary drilling method. As construction procedure of historic
reverse-circulation method has been well documented in many literatures and it is less
frequently utilized nowadays in Thailand, only a rotary-drilling method is briefly presented in
this article as follows.
In rotary drilling method, a temporary casing of appropriate length (12 to 18m in
Bangkok depending on the thickness of soft clay) with required diameter (internal diameter
not less than that of design bored pile diameter) is first installed to ensure the stability of the
borehole in the top soft or loose soil layers. In some projects where the vibration is strictly
limited, standard length of casing with oscillator or short temporary casing of 5 to 8 m is

201

pushed down in combination with pre-boring process. Drilling is commenced by auger to drill
out the soil within the temporary casing and up to the top of the first water-bearing sand layer
or bottom of the casing in case of using short casing method. Drilling slurry or supporting
fluid is then supplied to the borehole and drilling is continued with a bucket down to the
design final depth of the pile. Before lowering the reinforcement cage, a special cleaning
bucket is used to clean the base of borehole. If bentonite slurry is used, recycling method by
air-lift or pump is applied as base cleaning process. Reinforcement cages are then lowered
into the borehole and concreting is carried out by tremie method.
A mechanical or hydraulic cable-suspended grab is commonly used for barrette
construction. Excavation of the trench is carried out by the cyclic-process of lifting and
lowering of the grab under gravity and tangential force of the clamshell operated by cables.
Different from bored pile construction, a guide wall of depth 1 to 1.5 m with inside clear
dimensions slightly larger than the nominal size of the barrette is used to guide the grab
during initial bites. Bentonite slurry is introduced to the trench as soon as initial excavation
commenced. The excavation is continued under the bentonite slurry to the final depth. After
recycling of slurry and lowering the rebar cage, concreting is done by tremie method.
OVERVIEW OF APPLICATION
Bored pile
The first wet-process large diameter bored pile was constructed for Pinklao Bridge in
Bangkok 30 years ago. Using reverse circulation method, diameter 1.50m bored pile was
installed up to 45m in the second sand layer. Three major bridges were constructed by bored
pile of reverse circulation method in Bangkok from early 1970 to 1980. The summarized
information of these bridges are tabulated below.
Table 1 First 3 major bridges constructed by large diameter wet-processed
bored piles in Bangkok
Project Name
Pinklao Bridge
Sathorn Bridge
New Memorial
Bridge

Year of
Construction
1971
1979
1982

Construction Method
Reverse circulation
Reverse circulation
Reverse circulation

Diameter
(m)
1.50
1.50
1.50

Depth
(m)
45
46
49

The first wet-processed bored pile of rotary-drilling method down to first sand layer
was constructed in Bangkok in late 1970 for high-rise building project, Royal Orchid Hotel.
Since then bored piles constructed by rotary-drilling method have been extensively used for
foundations of various heavy structures such as high-rise buildings, elevated expressways,
overpass-bridges, underground car park buildings, waste-water treatment plants and most
recently underground train stations of Bangkok first subway project. By mid 1980, bored pile
became the foundation of choice for heavy structures particularly in urban area of Bangkok.
The versatility of the construction method and the high-load capacity which in turn offered
the constructibility and cost-saving are the main factors contributed to the increasing use of
deep-seated large diameter bored piles and barrettes. Most of the early wet-process bored
piles (1980s) in Bangkok were constructed up to 50 m depth. Diameter of bored piles
commonly constructed in early days were ranging from 0.6 to 1.5 m. Table 2 below
summarizes the information of early major high-rise building projects in Bangkok.

202

Table 2 First high-rise building projects constructed by large diameter wet-processed bored
piles in Thailand (from 1979 to 1983)
Project Name
Royal Orchid Hotel
(1979)
Taiping Tower
(1980)
River City Hotel
(1982)
Asoke Tower
(1983)
Time Square
Building (1983)

Construction Method
Rotary-drilling with
auger and bucket
Rotary-drilling with
auger and bucket
Rotary-drilling with
auger and bucket
Rotary-drilling with
auger and bucket
Rotary-drilling with
auger and bucket

Diameter (m)
0.80 - 1.00

Depth (m)
33.0

0.80 - 1.50

32.0

0.80 1.00

27.5

0.80 - 1.50

50.0

0.80 - 1.50

50.0

Barrette
In some projects where bored piles were not feasible due to the site constraints,
applicable construction method and/or extensive bearing capacity requirements, the use of
barrette foundations would make a suitable alternative. Barrettes with dimension ranging from
0.80 m x 2.7 m to 1.5 m x 3.0 m for safe working load capacity from 11,000 to 23,000 kN
have been used in some major projects.
The first barrette in Bangkok was believed to be constructed in late 1970 for the
foundation of Bangkok Bank Head Office Building at Silom road with size of 0.6-0.8 x 2.5m
and toe depth of 33m in late 1970. Barrettes can be constructed with flexible layout plan for
both vertical and lateral loads. The layout pattern of barrettes can be arranged in a continuous
row or column, radial, alternating long and short axis of barrette and a combination of two or
more of such patterns. One type of equipment can be used for constructing both barrettes and
diaphragm walls in particular projects thus reduces the mobilization cost for additional
equipment. In addition to the large bearing capacity requirements, on site difficulties such as
limited head room where piling rigs cannot be utilized, under such situations like presence of
overhead high voltage power cables, existing overpasses or structures for elevated
expressways and planned subway stations, also demands the barrettes. A summary of
observation on barrettes of completed 26 projects completed with respect to selection criteria
is presented in Table 3.
Table 3 Summary of Barrette Selection
Criterion
No. of Project
Remarks
High Load Capacity
4
Incorporated with bored piles
Minimize Construction
6
Alternative for bored piles
Equipment
Limited Head Room for
3
Under existing structures such as bridges,
Excavation
elevated expressway and power lines
Combination with
10
As diaphragm wall legs
Diaphragm Wall
Foundation as well as
3
Provision for the future requirement
portion of column
Static load test up to 5290 ton conducted on barrette set the record as the highest load
ever tested for a single cast-in-situ deep foundation in Thailand (Thasnanipan et. al. 2002a)

203

PROBLEMS OF WET-PROCESSED BORED PILES AND BARRETTES

IN EARLY DAYS
Basic but extensive problems were experienced in early stages of bored pile construction in
Thailand as summarized below.
Limited availability and capacity of equipment
Lack of skills in operation of equipment (particularly drilling rig)
Adverse effect due to a slow rate of drilling (e.g. excessive formation of filter-cake by
using bentonite slurry)
Limited knowledge and less advance technique in control of bentonite slurry
Limited experience in construction method and related negative impact
Quality of concrete for tremie concreting method
Lack of experienced engineers and foremen
Improper construction and quality control specification and guidelines for deep-seated
piles in local soil condition
Improper design for constructibility
DEVELOPMENT IN CONSTRUCTION
Over the past 3 decades, along with the development of wet-processed piling
technology world-wide, equipment, construction method, design as well as better
understanding and realization of construction impact on the performance have been
significantly improved in Thailand. Table 4 summarizes the areas of improvement in bored
pile and barrette construction and main factors contributed to these.
Research works presenting some of these developments were published both locally
and internationally. The design, construction and behavior of bored cast in-situ concrete piles
in Bangkok Subsoil was presented by Thasnanipan et. al., (1998a). The construction and
performance of barrettes in Bangkok Subsoil was also reported by Thasnanipan et. al.,
(1998b). Effect of construction time and bentonite viscosity on shaft friction of bored piles
was also pointed out by Thasnanipan et al., (1998c). Teparaksa et. al., (1999) published the
base grouting of wet process bored piles in Bangkok subsoil which demonstrated the method
of base grouting and reported the grout spreading mechanism for long bored piles and
performance of base grouted bored piles.

(a)

(b)

Fig. 1 (a) Guide wall for cruciform barrette (b) View of cruciform barrette after installation of
reinforcement
Construction of cruciform barrette (shown in Figure 1) for monopole type highvoltage power transmission line was described in the work of Thasnanipan et. al., 2000a.

204

Above-mentioned research works reflect the development history of bored pile and barrettes
in Thailand and offered useful information to the local construction industry and perhaps to
the international deep foundation engineering society.
Table 4 Summary of development in bored pile and barrette construction
Area of
development /
improvement
Speed of
construction

Pile size and

depth

Base grouting
Application of
polymer-based
slurry (for bored
pile only)
Construction
impact on quality
and performance
Quality control in
construction
process
Quality control
test method and
interpretation

In the past
(Early 1980s)

At present

Minimum 3 days
required to complete
diameter 1.5m tip 50m
bored pile

Less than 1 day to

complete diameter
1.5m tip 50m bored
pile

Bored pile
Bored Pile
Maximum diameter 2.0m Maximum Diameter
and common depth 252.0m and common
50m for bored pile.
depth 25-60m
Barrette
Barrette
Limited in size and depth
Various sizes &
depth over 60m
Not available
Available
Not available

Extensively use

Not well understood

Improving

Not well established and

systematic

Well established
and systematic

Less methods available

and limited knowledge in
interpretation

Better equipment
available and better
knowledge in
interpretation

Main factor
contributed to
development
Better equipment,
operating skills as
well as improvedknowledge in
construction method,
management and
local soil condition
Better equipment,
operating skills as
well as improvedknowledge in
construction method,
management and
local soil condition
Equipment
availability and
advance technology
Material availability
and research works
Experience from past
projects and research
works
Experience from past
projects and research
works
Advance equipment,
experience from past
projects and research
works

Experience from past projects and extensive research works provided better
understanding on construction impact on quality and performance of these wet-processed
deep foundations. Marked difference between outcome quality of bored piles constructed in
1980s and 1990s can be observed by significant less defects found in the latter. It can also be
observed from load test results that bored piles constructed in late 1990 have higher capacity
than those of 1980s as shown in Table 5.
Table 5 Load test results of bored piles constructed in 1970 to 1980 and 1990s

205

Year of
construction

Project Name

Pile Dimension
(Dia. & Depth)

1971
1995

Pinklao bridge
Central Plaza
Pinklao
Taiping tower
High-rise building
at Ekamai Rd.

1980
1990

1.5m x 45m
1.2m x 45m

Design
Load
(ton)
410
700

Test
Load
(ton)
820
1400

Total settlement
at max. test load
(mm)
5
27

1.0m x 32m
1.0m x 32m

300
390

1125
975

118
23

Fig. 2 Progress of piling work for the Rama VIII main bridge
Bored piles have been extensively used for foundations of the majority of the elevated
expressways since 1991. Thousands of base-grouted bored piles were constructed for these
infrastructure projects including Second Stage Expressway (1991), Don Muang Tollway
Extension (1997), Bangna-Bang Pli-Bangprakong Expressaway (1998), and Wat Nakorn-In
project (2001). Over 700 bored piles of diameter ranging from 0.8m to 1.5m with depth from
35m to 54m were constructed (1999-2000) by polymer-based slurry for the foundation of
Rama VIII Bridge, one of the initiatives of his majesty the King Bhumibol Adulyadej.
Progress of piling works at the bank of Chao Phraya river for the main bridge of Rama VIII
Bridge is depicted in Figure 2.
The availability of more reliable and powerful equipment and tools for drilling makes
possible to construct the rock-socket bored piles. Figure 3 shows the drilling rig equipped
with rock-auger used for construction of highway bridge across Mool River in 2001, in the
north-eastern part of Thailand, where drilling was carried out through weathered-sandstone by
powerful rotary drilling rig with core barrel and rock-auger.

Fig. 3 Rock-socket bored pile construction for Mool River Project - Ubon Ratchathani Province
DEVELOPMENT IN DESIGN CONCEPT AND PARAMETERS

206

In the initial stage of introducing bored piles in Thailand, the design concepts and
parameters were mainly based on the available literatures from the research works carried out
in other parts of the world such as Tomlinson (1957), Skempton (1959), Broms (1966),
Bowles (1968), Meyerhof (1976) etc. The work of Chiruppapa (1968) was believed to be the
first research data available for design parameters of bored pile in Bangkok soft clay. The
author conducted the research based on 6 dry-processed small diameter bored piles with load
cells installed at the pile toe in AIT campus in Phathumthani. Though the piles were simply
constructed by dry-processed with casing method, this early-stage research study provided
some important information such as load-settlement behavior, adhesion factor () and bearing
capacity factor (N) of bored pile in Bangkok soft clay.
With the passage of time, design method and parameters for local subsoil were
improved as a result of research works carried out in 1980s. Differences between the behavior
of driven piles and bored piles were well realized from these researches. (Ng, 1983) presented
the load distribution characteristics of wet-processed bored piles founded in both first and
second sand layers of Bangkok based on the instrumented (strain gauges) pile load test
results. (Chiewcharnsilp, 1988) reported the shaft friction factor values of sand layers in
Bangkok based on the instrumented load test results. In addition to the researches previously
available (Chiruppapa, 1968; Suwanakul, 1969; Promboon, 1981; Ng, 1983; Chiewcharnsilp,
1988) Suchada (1989) determined the shaft friction factor (adhesion factor, ) of Bangkok
clay layers based on 11 bored piles of diameter ranging from 0.50 to 0.80m with the
embedded length varying between 21.50m and 46m. It should be noted that the design
parameters obtained from the research works of 1980s were mainly based on the estimation of
shaft friction loads from plain static pile load test results with numbers of assumption since
instrumented pile load test results were limited.
With the peak of construction-boom, large numbers of instrumented full-scale static
load tests on bored piles were conducted throughout 1990s which provided better
understanding on behavior of these deep-seated foundations. Researches focused on the
design parameters and methods were produced based on these test results. The design
parameters were more accurate and well established that wet-process cast-in-place
foundations became regarded as reliable foundations in construction industry of Thailand.
With more confidence on soil parameters and behavior of these deep foundations, the
designers designed to carry significantly higher loads for bored piles and barrettes constructed
in late 1990 than those of 1980s. (Thasnanipan et. al., 1999) reported the failure mechanism
of long bored piles in layered soil of Bangkok. The authors cited that for the bored piles
embedded in the multi layered soils of Bangkok, estimation of ultimate shaft friction capacity
needs to consider the brittle type of failure mechanism of stiff to hard clay layers and the
values selected need to be adjusted accordingly. Peak and residual values mobilized in
the stiff clay layers analyzed by the authors were plotted as shown in Figure 4 along with the
suggested curves by different researchers. It can be seen from the figure that the residual
value of 2nd stiff clay layer (undrained shear strength values of 25 ton/m2) at the maximum
test load drops below the curve suggested by (Suchada, 1989). So the values proposed in
Figure 4 for the stiff to hard clay layers gives overestimate of the ultimate shaft friction under
theses conditions.
Introduction of polymer-based slurry for wet-process bored piles marked a major
breakthrough for both construction and design engineers. Thasnanipan et al.,(2002b), reported
that bored piles constructed with polymer-based slurry have higher capacity than those
constructed with bentonite slurry. Figure 5 shows the shaft friction factors of sand layers
for polymer-based bored piles in comparison with the design line of bentonite bored piles.

207

1.20
Bored Cast Insitu Piles
1.00

Adhesion Factor ,

Tomlinson,1957
0.80
Suchada,1989
for bored piles in Bangkok soils

0.60

0.40

Peak
Stas and Kulhawy, 1984

0.20

Residual

0.00
0.00

5.00

10.00

15.00

20.00

25.00

30.00

Undrained Shear Strength (ton/m )

Fig. 4 Comparison of adhesion factor , suggested by different researchers with the actual
mobilized in the stiff clay layers (after Thansnanipan et. al 1999)
1 .0
Submaneewong (1999) : Bored Piles (Tip in 2nd Sand)
Submaneewong (1999) : Bored Piles (Tip in 2nd Hard Clay)
0 .8

Chiewcharnsilp (1988)
Polymer Bored Pile TP-1 (1st Sand)

(Ks tan )

Polymer Bored Pile TP-2 (1st Sand)

Polymer Bored Pile TP-3 (1st Sand)
0 .6

Polymer Bored Pile TP-1 (2nd Sand)

Polymer Bored Pile TP-2 (2nd Sand)

Submaneewong
( 1999 )

Polymer Bored Pile TP-3 (2nd Sand)

0 .4

0 .2

0 .0
28.0

30.0

32 .0

34.0

36 .0

38.0

Angle of Internal Friction, '

Fig. 5 Back-calculated values of polymer bored piles at maximum test load plotted on
design line of bentonite bored piles constructed in Bangkok subsoil
(after Thansnanipan et. al 2002b)
.
DEVELOPMENT IN QUALITY CONTROL TESTING
Improvement of testing equipment and powerful computer facilities are the key factors
contributed to the development in quality control testing. Interpretation skills relevant to local
soil condition and construction method of these tests were significantly improved in local
industry. For instance, in early 1980, interpretation of the sonic integrity (seismic test) test
results were needed to send to the specialists abroad which made the testing cost more
expensive. Significant cost-saving were achieved in some major projects of late 1990, as more
practical and precise interpretation were made to verify and establish the acceptance criteria in
proving the quality of suspected piles with anomalies.

208

Koden drilling monitoring system

In Thailand, prior to the availability of Koden testing equipment, borehole verticality
was checked by mechanical type equipment such as plumb float and adjustable globe.
However, the reliability of these methods was doubtful. Koden drilling monitor system was
believed to be first used in Thailand in 1980. The verticality of slurry-filled borehole can be
monitored rapidly from the electronic plot of continuous profile by Koden equipment. This
system is very useful for the verticality-control of long barrettes - unlike bored piles, guiding
by temporary casing is not available for barrette drilling.
Sonic integrity test
Sonic integrity test, also known as seismic test is the most common method of
integrity testing for both driven and bored piles in Thailand. Sonic integrity test is usually
selected for both quality check (control test) and retrospective investigation. It is the cheapest
in terms of cost and the simplest in terms of testing process. The main advantage of this test
is that since no particular preparation is necessary during the pile construction phase it is more
flexible to select which pile is to be tested. However, interpretation of sonic integrity testing
needs considerable experience and knowledge in testing, subsoil condition and construction
method. In many projects it is a part of the contractual requirement to conduct sonic integrity
test. Minimum 10 % to maximum 100 % of production piles are commonly tested. It is also
a reasonably acceptable method for applying as a retrospective investigation in determining
integrity of the pile. The signal characteristics and their interpretations of sonic integrity test
on piles founded in Bangkok subsoil were reported in details by Thasnasipan et al. (1998d).
Cross-hole sonic logging test
Sonic logging test is relatively expensive. It is mainly employed as a pre-planed site
quality control testing. The major advantage of this method is that test can be carried out
shortly after the pile construction. Hence, rectification measures can be implemented if the
pile is detected with defect while the foundation contractor is on site. However, this method is
generally not applicable if pile integrity is in question due to post-construction activities as
access tubes are usually grouted after completion of the test. The first sonic logging test was
believed to be conducted in 1982 for the wet-processed bored piles of Memorial Bridge
Project where bad concrete zones were detected at depth about 20m and 1m (Ng, 1983). Use
of Sonic logging test for checking pile integrity has increased in Thailand in recent years.
The results from sonic logging test conducted on model piles in Bangkok helped to extend the
knowledge of the signal characteristics and interpretation (Thasnanipan et. al., 2000b).
High strain dynamic load test
High strain dynamic integrity test has become a well-accepted method especially for
evaluating the pile capacity in todays foundation industry of Thailand and it is applied for
both driven and bored piles.. A large number of related technical papers and case histories of
the test have been published and it is a part of standards and specifications such as ASTM
D4945-89 (Standard Method for High-strain Dynamic Testing of Piles). Thasnanipan et. al.
2000b, reported the application of dynamic load testing in Thailand.
CONCLUDING REMARKS
Development of bored pile and barrettes in Thailand has been presented. According to
the authors experience as a contractor, development in both construction and design aspects
of these deep foundations in past decades are significant. In authors opinion, however, there
are many works to be done to improve further with particular focus on constructibility issues,
concrete technology for wet-processed bored piles and barrette, reliable but cost-effective

209

quality control testing and value-engineering. Starting from planning stage, site investigation,
design, construction and inspection should be integrated such that designers, contractors and
construction inspectors participating as a team with common goal. Appropriate and practical
specification should be established jointly by these parties for local soil condition and
construction method. Practical acceptance criteria should be developed to verify bored piles or
barrettes with suspected anomalies. Continuing education should be promoted for both
designers, inspection engineers and contractors. The Geotechnical Chapter of the Engineering
Institute of Thailand under the royal patronage of his majesty the King Bhumibol Adulyadej,
has started to establish the standard code of practice and guidelines for wet-processed bored
piles which will serve as a yardstick for the deep foundation industry upon completion in near
future.
REFERENCES
Bowles, J. E. (1968) Foundation analysis and design, McGRaw-Hill, New York
Broms B. B. (1966) Methods of calculating the ultimate bearing capacity of piles a summary, Soils-Soils, No.
18-19, pp. 21-32
Chiewcharnsilp, S. (1988) Static load transfer and predictions of carrying capacity of bored piles (in Thai), M.
Eng. Thesis, Chulalongkorn University, Bangkok, Thailand
Chiruppapa, P. (1968) Cast in-situ bored piles in Bangkok Clay, M. Eng. Thesis No. 213, Asian Institute of
Technology, Bangkok, Thailand
Meyerhof, G. G. (1976) Bearing capacity and settlement of pile foundations, Journal of soil mechanic and
foundation division, Proceeding ASCE Vol. 102, GT3, pp. 197-223
Ng, K. C. (1983) The construction problem and performance of large bored piles in second sand layer, M. Eng.
Thesis No. GT 82-26, Asian Institute of Technology, Bangkok, Thailand
Promboon, S . and Brenner, R.P. (1980), Large diameter bored piles in Bangkok clay, Proceedings of 10th
conference on soil mechanic and foundation engineering, Vol. 2, pp. 815-818
Suwanakul, V. (1969) Cast in-situ piles in Bangkok clay, M. Eng. Thesis No. 260, Asian Institute of Technology,
Bangkok, Thailand
Teparaksa W., Thasnanipan N., and Anwar M. A. (1999), Base grouting of wet process bored piles in Bangkok
subsoils, 11th Asian Regional Confer. on soil mechanics and geotech. Engg., Seoul, Korea, pp. 269-272.
Thasnanipan N., Teparaksa W., Maung A.W., Ganeshan B., (1998a) Design, construction and behavior of bored
cast in-Situ concrete piles in Bangkok Subsoil, Proceedings: Fourth International Conference on Case
Histories in Geotechnical Engineering, St. Louise, Missouri, U.S.A., p.p.281-287
Thasnanipan N., Maung A.W., Tanseng P., (1998b) Barrettes founded in Bangkok subsoils, construction and
performance, Thirteenth Southeast Asian Geotechnical Conference, Taipei, Taiwan, ROC, p.p.573-578.
Thasnanipan N., Ganeshan B., Anwar M.A., (1998c), Effect of Construction Time and Bentonite Viscosity on
Shaft Capacity of Bored Piles, Third International Geotechnical Seminar on Deep Foundation on
Bored and Auger Piles, Ghent, Belgium. p.p. 171-177.
Thasnanipan N., Maung A.W., Ganeshan B., (1998d), Sonic Integrity Test on Piles Founded in Bangkok
Subsoil-Signal Characteristics and Their Interpretations, Proceedings: Fourth International Conference
on Case Histories in Geotechnical Engineering, St. Louis, Missouri, U.S.A., p.p.1086-1092
Thasananipan N., M. A. Anwar Maung A. W (1999) Failure Mechanism of Long Bored Piles in Layered Soils
of Bangkok, Civil and Environmental Engineering Conference, Bangkok, Thailand, p.p. V-69 to V-73.
Thasnanipan N., Tanseng P., Maung A.W. & Anwar M. A. (2000a), Barrettes : A Versatile Foundation for
Transmission Line Towers, An International Conference on Geotechnical & Geological Engineering,
Melbourne, Australia, 19-24 November 2000
Thasnanipan N., Maung A. W. & Navaneethan T.., Z. Z. Aye (2000b), Non-Destructive Integrity Testing on
Piles Founded in Bangkok Subsoil, 6th International Conference on Application of Stress-Wave Theory
to Piles, September 2000, Sao Paulo, Brazil
Thasnanipan N., Zaw Zaw Aye and Wanchai Teparaksa (2002a) Barrette of Over 50,000 kN Ultimate Capacity
Constructed in the Multi-Layered Soil of Bangkok, Proceedings of the International Deep Foundations
Congress 2002, Geotechnical Special Publication No. 116, Volume Two, ASCE, Febuary 14-16 2002,
Orlando, Florida, USA, p.p. 1073-1087
Thasnanipan, N., Zaw Zaw Aye, Chanchai Submaneewong and Wanchai Teparaksa (2002b) Performance of
Wet-Process Bored Piles Constructed with Polymer-Based Slurry in Bangkok Subsoil, Proceedings of
the International Deep Foundations Congress 2002, Geotechnical Special Publication No. 116, Volume
One, ASCE, February 14-16 2002, Orlando, Florida, USA, p.p. 143-157

210

Proceedings of the 16th Southeast Asian Geotechnical Conference

8-11 May 2007, Kuala Lumpur, Malaysia

16th Southeast Asian

Geotechnical Conference
Edited by:
Ooi Teik Aun, Yee Yee Weng and Ting Wen Hui

Development and Achievements of Deep-seated

Bored Piles and Barrettes Construction in
Thailand for the Past Forty Years, A Country
Report
Narong Thasnanipan, Zaw Zaw Aye,
Submaneewong and Thayanan Boonyarak
SEAFCO Public Co., Ltd. Bangkok, Thailand

211

Chanchai

Development and Achievements of Deep-seated Bored Piles and Barrettes

Construction in Thailand for the Past Forty Years, A Country Report
Narong Thasnanipan, Zaw Zaw Aye, Chanchai Submaneewong and Thayanan Boonyarak
Seafco Public Company Limited, Bangkok, Thailand
[email protected]

Abstract: This paper presents the development and achievements of deep-seated bored piles and barrettes construction in Thailand
based on the authors experiences, in the past four decades. Literature related to bored piles and barrettes in Thailand published
through out the past forty years is summarized together with recent research works. Problems and difficulties in construction of wetprocessed bored piles and barrettes in early days are reported. The areas of improvement, advancement and achievement in bored pile
and barrette construction and main factors contributed to these developments are also presented. An attempt is also made to briefly
discuss the future trend of bored pile and barrette foundations.

1 INTRODUCTION
Pore Pressure (kN/m 2)

Geotechnical
Parameters
0

Depth (m)

10

Bangkok Soft
Clay

200

400

600

Su= 10-25 KPa

t = 14-16kN/m 3
-

10

30

Medium
Dense Sand
(First Sand)

40

Hard Clay

50

Dense to Very
Dense Sand
(Second Sand)

Su= 50-14020- KPa

t = 17-21kN/m 3
-

30
SPT-N= 20-40
t = 20kN/m 3

Su> 200 KPa

= 21KN/m 3
t

SPT-N= 50-100
t = 21KN/m 3

40

50

60

e
Lin
tic
sta
e
dro
Lin
Hy
wn
do
raw
ic D
etr
om
iez

20

Medium to
Stiff Clay

60

0
0

Weathered Crust

lP
tua
Ac

During the past four decades, owing to the acceleration of

development in mega infrastructure and numerous tall
building projects, techniques and practice of cast-in-place
deep foundations, particularly wet-process deep-seated large
diameter bored piles and barrettes have experienced enormous
development and achievements in Thailand. The versatility of
the construction method and the high-load capacity which in
turn offered the constructibility and cost-saving are the main
factors contributed to the increasing use of those cast-in-place
deep foundations. In the initial stage, there were a number of
questions on design and construction aspects of these
foundation systems particularly in Bangkok subsoil. With the
passage of time, the construction equipment, installation
techniques and testing of deep foundation elements have been
developed. Large numbers of instrumented full-scale static
pile load tests were conducted through out the 1990s mainly in
mega projects, which provided better understanding on
behavior of these deep-seated foundations.
Research focused on the design parameters and methods
were produced based on these test results. The design
parameters became so well established that wet-process castin-place foundations became regarded as reliable foundations
for practitioners involved in construction industry in Thailand.
This paper presents the past and current practice of
wet-processed bored piles and barrettes in Thailand together
with a brief discussion on future trend of these cast-in-place
deep foundations. The information contained in the paper is
mainly from the mega projects constructed in Bangkok and
adjacent areas.

Fig. 1. Typical soil profile of Bangkok with piezometric

drawdown condition (Thasnanipan, et al., 2002)
3. CONSTRUCTION METHOD IN GENERAL
Due to the prevailing subsoil and groundwater conditions,
deep-seated bored piles of toe depth over 24m are constructed
by wet-process or slurry displacement method. Wet-process
bored piles can be constructed by 2 methods, reversecirculation and rotary drilling. As construction procedure of
historic reverse-circulation method has been well documented
in various studies and since it is less frequently utilized
nowadays in Thailand, only a rotary-drilling method is briefly
presented in this article as follows.
In the rotary drilling method, a temporary casing of
appropriate length (12 to 18m in Bangkok depending on the
thickness of soft clay) with required diameter (internal
diameter not less than that of design bored pile diameter) is
first installed to ensure the stability of the borehole in the top
soft or loose soil layers. In some projects where the vibration
is strictly restricted, a standard length of casing with oscillator
or short temporary casing of 5 to 8 m is pushed down in
combination with the pre-boring process. Drilling is
commenced by auger to drill out the soil inside the temporary
casing. Auger-drilling is commonly continued up to the top of

2. SUBSOIL AND EXISTING PIEZOMETRIC PROFILE

Subsoil profile and the present piezometric drawdown
condition of Bangkok are presented in Fig. 1 below. A typical
subsoil profile is relatively consistent in different localities in
Bangkok. It is characterized by alternating layers of clay and
sand deposits as shown in Fig. 1.

213

the first water-bearing sand layer or bottom of the casing when

using the short casing method. Drilling slurry or supporting
fluid is then supplied to the borehole and drilling is proceeded
with a bucket down to the design final depth of the pile.
Before lowering the reinforcement cage, a special cleaning
bucket is used to clean the base of borehole. If bentonite slurry
is used, recycling method by air-lift or pump is applied as the
base cleaning process. Reinforcement cages are then lowered
into the borehole and concreting is carried out by tremie
method.
A mechanical or hydraulic cable-suspended grab is
commonly used for barrette construction. Excavation of the
trench is carried out by the cyclic-process of lifting and
lowering of the grab under gravity and tangential force of the
clamshell operated by cables (mechanical grab) or hydraulic
action (hydraulic grab). Different from bored pile
construction, a guide wall of depth 1 to 1.5 m with inside clear
dimensions slightly larger than the nominal size of the barrette
is used to guide the grab during initial bites. Bentonite slurry
is introduced into the trench as soon as the initial excavation
commenced. The excavation is continued under the bentonite
slurry to the final depth. After recycling of slurry and lowering
the rebar cage, concreting is done by tremie method.

the construction method and the high-load capacity which in

turn offered the constructability and cost-saving, are the main
factors contributing to the increasing use of deep-seated large
diameter bored piles and barrettes. Most of the early wetprocess bored piles (1980s) in Bangkok were constructed up to
50 m depth. Sizes of bored piles constructed in early days
ranged from diameter 0.6 to 1.5 m. Summarized information
of early major high-rise building projects in Bangkok is
presented in Table 2.
4.2

In some projects where bored piles were not feasible due to

the site constraints, applicable construction methods, and/or
extensive bearing capacity requirements, the use of barrette
foundations would make a suitable alternative. Barrettes with
dimensions ranging from 0.80 m x 2.7 m to 1.5 m x 3.0 m for
safe working load capacity from 11 000 to 25 000 kN have
been used in some major projects.
The first barrette in Bangkok was believed to have been
constructed in the late 1970s for the foundation of the
Bangkok Bank Head Office Building at Silom road with the
size of 0.6-0.8 x 2.5m and the toe depth of 33m. Barrettes can
be constructed with flexible layout plans for both vertical and
lateral loads. The layout pattern of barrettes can be arranged
in a continuous row or column, radial, alternating long and
short axis of barrette and a combination of two or more of
such patterns. To minimize the need of additional equipment,
a mechanical grab mounted on crawler crane can be used for
constructing both barrettes (as piled-foundation) and
diaphragm walls (as earth retaining structure) in some suitable
projects without employing a bored piling rig, which could
save the extra mobilization cost. In addition to the large
bearing capacity requirements, on site difficulties such as
limited head room where piling rigs cannot be utilized, under
such situations like presence of overhead high voltage power
cables, existing overpasses or structures for elevated
expressways and planned subway stations, also demands the
use of barrettes. Static load test up to 52 900 kN conducted on
barrette set the record as the highest load ever tested for a
single barrette foundation in Thailand (Thasnanipan et. al.
2002a). A summary of completed 35 projects constructed with
barrettes with respect to selection criteria is presented in
Table 3.

4. OVERVIEW OF APPLICATION
4.1 Bored piles
Using tripod rigs, small diameter dry-processed bored pile
construction was started in early 1970. The first wet-process
large diameter bored pile was constructed for Pinklao Bridge
in Bangkok 36 years ago. Using reverse circulation method, a
1.50m diameter bored pile was installed up to 45m in the
second sand layer. Three major bridges were constructed by
bored pile using the reverse circulation method in Bangkok
from early 1970 to 1980. The summarized information of
these bridges is tabulated in Table 1 below.
Table 1. First 3 major bridges constructed by large diameter
wet-processed bored piles in Bangkok.
Project
Name
Pinklao
Bridge
Sathorn
Bridge
New
Memorial
Bridge

Year
of
Construction
1971
1979
1982

Construction
Method
Reverse
circulation
Reverse
circulation
Reverse
circulation

Diameter
(m)
1.50

Depth
(m)
45

1.50

46

1.50

49

Barrette piles

5. PROBLEMS AND DIFFICULTIES IN

CONSTRUCTION OF WET-PROCESS BORED PILES
AND BARRETTES IN EARLY DAYS
Basic but extensive problems were experienced in early stages
of bored pile construction in Thailand as summarized below.
Limited availability and capacity of equipment
Lack of skills in operation of equipment (particularly the
drilling rig)
Adverse effect due to a slow rate of drilling (e.g. excessive
formation of filter-cake by using bentonite slurry)
Limited knowledge and less advanced techniques in the
control of bentonite slurry
Limited experience in construction method and related
negative impact
Quality of concrete used for tremie concreting
Lack of experienced engineers and foremen

The first wet-processed bored pile utilizing the rotarydrilling method down to first sand layer was constructed in
Bangkok in the late 1970s for high-rise building project, the
Royal Orchid Hotel located at the bank of Chao Phraya river.
Since then bored piles constructed by rotary-drilling method
have been extensively used for foundations of various heavy
structures such as high-rise buildings, elevated expressways,
overpass-bridges, underground car park buildings, waste-water
treatment plants and most recently underground train stations
of Bangkoks first subway project. By mid 1980s, bored piles
became the foundation of choice for heavy structures
particularly in the urban area of Bangkok. The versatility of

214

Improper construction and quality control specifications

and guidelines for deep-seated piles in local soil
Improper design for constructability

of barrettes in Bangkok Subsoil was also reported by

Thasnanipan et al., (1998b). Effect of construction time and
bentonite viscosity on shaft friction of bored piles was also
pointed out by Thasnanipan et al., (1998c). Thasnanipan et
al., (2004a) also reported a comprehensive study on the
effectiveness of two different toe-grouting methods, known as
tube--manchette and drill-and-grout, applied in Bangkok.

Table 2. First high-rise building projects constructed with

large diameter wet-processed bored piles in Thailand (from
1979 to 1983).
Project
Name
Royal
Orchid
Hotel
(1979)
Taiping
Tower
(1980)
River City
Hotel
(1982)
Asoke
Tower
(1983)
Time Square
Building
(1983)

Construction
Method
Rotary-drilling
with auger and
bucket

Diameter (m)
0.80 - 1.00

Depth
(m)
33.0

Rotary-drilling
with auger and
bucket
Rotary-drilling
with auger and
bucket
Rotary-drilling
with auger and
bucket
Rotary-drilling
with auger and
bucket

0.80 - 1.50

32.0

0.80 1.00

27.5

0.80 - 1.50

50.0

0.80 - 1.50

50.0

Table 4. Summary of development in bored pile and barrette

construction.
Area of
development
/
improvement
Speed of
construction

In the past
(Early
1980s)

At present

Minimum 3
days
required to
complete
diameter
1.5m tip 50m
bored pile

Less than 1
day to
complete
diameter 2.0
m tip 60m
bored pile

Pile size and

depth

Bored pile
Maximum
diameter
1.5m and
common
depth 2550m for
bored pile.
Barrette
Limited in
size and
depth

Bored Pile
Maximum
Diameter
2.0m and
common
depth 2560m
Barrette
Various sizes
& depth over
60m

Base
grouting

Not available

Available

Application
of polymerbased slurry
(for bored
pile only)
Construction
impact on
quality and
performance
Quality
control in
construction
process
Quality
control test
method and
interpretation

Not available

Extensive
use

Not well
understood

Improving

Not well
established
and
systematic
Fewer
methods
available and
limited
knowledge
in
interpretation

Well
established
and
systematic
Better
equipment
available and
better
knowledge
in
interpretation

Table 3. Summary of Barrette Selection in major projects.

Selection Criterion

No. of
Project

Remarks

High Load Capacity

Incorporated with bored

piles

Minimize
Construction
Equipment
Limited Head Room
for Excavation

Alternative for bored

piles

Combination with
Diaphragm Wall

16

Under existing structures

such as bridges, elevated
expressway and power
lines
As diaphragm wall legs

Foundation as well
as portion of
column

Provision for for the

futures requirement

6. IMPROVEMENT IN CONSTRUCTION
Over the past three decades, along with the development of
wet-processed piling technology in other parts of the world,
equipment, construction technique and design methods as well
as better understanding of construction impact on the
performance of this type of foundation have significantly
improved in Thailand. Table 4 summarizes the areas of
improvement in bored pile and barrette construction and main
factors contributed to these developments. Some of these
improvements were published both locally and internationally.
The design, construction and behaviour of bored cast in-situ
concrete piles in Bangkok Subsoil were presented by
Thasnanipan et al., (1998a). The construction and performance

215

Main factor
contributed
to
development
Better
equipment,
operating
skills as well
as
improvedknowledge
in
construction
method,
management
and local
soil
condition
Better
equipment,
operating
skills as well
as
improvedknowledge
in
construction
method,
management
and local
soil
condition
Equipment
availability
and advance
technology
Material
availability
and research
Experience
from past
projects and
research
Experience
from past
projects and
research
Advance
equipment,
experience
from past
projects and
research

Experience from past projects and extensive research works

provided better understanding of construction impact of
quality and performance of these wet-processed deep
foundations. Marked difference between outcome quality of
bored piles constructed in 1980s and 1990s can be observed
by significant fewer defects found in the latter. It can also be
observed from load test results that bored piles constructed in
late 1990 have higher capacities than those of 1980s as shown
in Table 5.
Table 5. Load test results of bored piles constructed in 1970
to 1980 and 1990s.
Year
1971
1995
1980
1990

Project
Name
Pinklao
bridge
Central
Plaza
Pinklao
Taiping
Tower
High-rise
building at
Ekamai
Rd.

Pile size
(Dia. &
Depth)
1.5 m x
45 m
1.2 m x
45 m
1.0 m x
32 m
1.0 m x
32 m

Design
Load
(kN)
4100

Test
Load
(kN)
8200

max
(mm)

7000

14000

27

3000

11250

118

3900

9750

23

Fig. 2b. Rama VIII cable-stayed bridge and connecting

elevated flyovers.
The availability of more reliable and powerful equipment
and tools for drilling makes it possible to construct the rocksocket bored piles by rotary drilling method. Fig. 3(a) and 3(b)
show the drilling rig equipped with a rock auger used for
construction of highway bridge across Mool River in 2001, in
the north-eastern part of Thailand. Drilling was carried out
through weathered-sandstone by powerful rotary drilling rig
with core barrel and rock auger.

Bored piles have been extensively used for foundations of

the majority of the elevated expressways since 1991.
Thousands of base-grouted bored piles were constructed for
these infrastructure projects including Second Stage
Expressway, constructed in 1991; Don Muang Tollway
Extension,
constructed
in
1997;
Bangna-BangpliBangprakong Expressaway constructed in 1998; and
Wat Nakorn-In project, constructed in 2001. Over 700 bored
piles of diameter ranging from 0.8m to 1.5m with depth from
35m to 54m were constructed between 1999 and 2000 using
polymer-based slurry for the foundation of Rama VIII Bridge,
one of the initiatives of his majesty the King Bhumibol
Adulyadej. Progress of piling and superstructure works of the
Rama VIII Bridge at the bank of Chao Phraya river is depicted
in Fig. 2(a) and 2(b).

Fig. 3 a. Rock-socket bored pile construction for Mool River

Project Ubon Ratchathani Province

Fig. 2a. Rama VIII cable-stayed bridge construction more

than 700 deep-seated bored piles were constructed using
polymer-based slurry.

Fig. 3 b. Rock-auger for Mool River bored pile construction

216

Bored pile of diameter 2m founded at 62m depth for the

foundation of Rama 8 bridge, a cable-stayed bridge across
Chaophaya river constructed in early 2005 was the largest
deep-seated bored pile in Bangkok installed by rotary drilling
method at that time. In the same project, both conventional
static pile load tests and bi-directional load tests were carried
out on 2m diameter bored piles. Simplified version of load vs.
settlement of these pile tests are presented in Fig.4.
Apparently, from the test results, the test piles could carry a
significantly higher load than the original pre-defined safe
design load of 1600 ton. Thasnanipan et. al. (2006) reported
the experiences gained in practical construction of these
largest diameters bored piles at the river bank. Pictures of
some activities during construction of bored piles in the said
project are presented in Figs. 5(a), 5(b), 6(a), 6(b), 7(a) and
7(b).
Fig. 5b View of progress of bored pile diameter 2m drilling at
the bank of Chaophaya River

LOAD (kN)

10000

20000

30000

40000

50000

Settlement (mm)

5
10
15
20
25

Conventional Static Pile Load Test

Bi-directional Load Test

30

Fig.4. Load-settlement relationship of diameter 2.0m bored

piles (conventional static pile load test and bi-directional load
test)

Fig. 6a View of progress in rebar cage preparation for bidirectional load test

Fig. 5a View of progress in temporary casing installation for

bored pile diameter 2m at the bank of Chaophaya River

Fig. 6b View of progress in installation of rebar cage

217

Fig. 8b View of cruciform barrette after installation of

reinforcement
Fig. 7a View of conventional static pile load test

Due to the extensive bearing capacity requirements, Tshape barrettes (Thickness=1m, Width =3m) of 55m deep
were used in one of the recent major projects. View of the
guide wall and reinforcement installation of T-shape barrette
are shown in Fig. 10(a) and 10(b). Fig. 11 shows the T-shape
barrettes after exposing to the cut-off level.
The above-mentioned works reflect the development and
achievement history of bored piles and barrettes in Thailand
and offered useful information to the local construction
industry as well as to that of the international deep foundation
construction.

Fig. 7b View of bi-directional pile load test

Deep-seated T-shape and L-shape barrettes were
constructed as leg-piles for diaphragm wall in recently
completed projects in Bangkok. Construction of cruciform
barrette (shown in Fig. 8(a) and 8(b)) for monopole type highvoltage power transmission line was described by
Thasnanipan et al., (2000a).
The availability of more powerful and efficient hydraulic
grabs offers faster construction of deep-seated barrettes in
comparison with mechanical cable-hung grabs. Fig. 9 shows
the construction of barrettes using hydraulic grab.

Fig. 9 View of barrette construction by hydraulic grab

Fig. 10a View of guide-wall construction for T-shape barrette

Fig. 8a Guide wall for cruciform barrette

218

instrumented (strain gauges) pile load test results.

Chiewcharnsilp (1988) reported the shaft friction factor
values of sand layers in Bangkok based on the instrumented
load test results. In addition to the literature previously
available (Chiruppapa, 1968; Suwanakul, 1969; Promboon,
1981; Ng, 1983; Chiewcharnsilp, 1988), Pimpasugdi (1989)
determined the shaft friction factor (adhesion factor, ) of
Bangkok clay layers based on 11 bored piles of diameters
ranging from 0.50 to 0.80m with the embedded length varying
between 21.50m and 46m. It should be noted that the design
parameters obtained from the research works of 1980s were
mainly based on the estimation of shaft friction loads from
plain static pile load test results with numbers of assumptions
since instrumented pile load test results were limited.
With the peak of construction-boom in Thailand,
particularly in Bangkok, large numbers of instrumented fullscale static load tests on bored piles were conducted
throughout 1990s which provided better understanding on
behavior of these deep-seated foundations. Researches focused
on the design parameters and methods were published based
on these test data. With improved design methods the wetprocess cast-in-place foundations came to be regarded as
reliable foundations in the construction industry of Thailand.
With more confidence on soil parameters selection and better
understanding on behaviour of these deep foundations, the
designer designed higher load capacity bored piles and
barrettes in late 1990 than those in 1980s. Thasnanipan et. al.,
(1999) reported the failure mechanism of long bored piles in
layered soil of Bangkok. The authors cited that for the bored
piles embedded in the multi layered soils of Bangkok,
estimation of ultimate shaft friction capacity needs to consider
the brittle type of failure mechanism of stiff to hard clay layers
and the values selected need to be adjusted accordingly.
Peak and residual values mobilized in the stiff clay layers
analyzed by the authors were plotted as shown in Fig. 12
along with the suggested curves by different researchers. It
can be seen from the figure that the residual value of 2nd
stiff clay layer (undrained shear strength values of 25 ton/m2),
at the maximum test load, drops below the curve suggested by
Pimpasugdi (1989).So the values proposed by the
author in Fig.12 for the stiff to hard clay layers overestimate
the ultimate shaft friction under these conditions.
Introduction of polymer-based slurry for wet-process bored
piles marked a major breakthrough for both construction and
design engineers. Thasnanipan et al.,(2002b), reported that
bored piles constructed with polymer-based slurry have higher
capacity than those constructed with bentonite slurry. Fig. 13
shows the shaft friction factors of sand layers for polymerbased bored piles in comparison with the design line of
bentonite bored piles. The higher load capacity of bored piles
constructed with polymer-based slurry allows the use of a
single, large diameter deep-seated bored pile in place of a
group of smaller size shallow-seated bored piles or driven
piles. Fig. 14 depicted the static pile load test result of 1.80m
diameter bored pile of 60m depth constructed with polymerbased slurry in urban area of Bangkok. The test was conducted
in late 2003. The maximum applied load 48 000 kN in this test
is believed to be the highest static pile load test performed on
a single bored pile in Thailand. Apparently, from the test
result, the pile (diameter 1.80m x 60m deep) could carry a
significantly higher load than the original pre-defined safe
design load of 16 000 kN.

Fig. 10b View of reinforcement installation for T-shape

barrette

Fig. 11 View of T-shape barrette after excavation to the final

depth
7. IMPROVEMENT IN DESIGN CONCEPT AND
PARAMETERRS SELECTION
In the initial stage of introducing bored piles in Thailand, the
design concepts and parameters were mainly based on the
available literature from research carried out in other parts of
the world such as Tomlinson (1957), Skempton (1959), Broms
(1966), Bowles (1968), Meyerhof (1976) etc. The work of
Chiruppapa (1968) was believed to be the first research data
available for design parameters of bored piles in Bangkok soft
clay. The author conducted the study based on 6 dry-processed
small diameter bored piles with load cells installed at the pile
toes on AIT campus in Phathumthani. Though the piles were
simply constructed using the dry-process with casing method,
this early-stage study provided some important information
such as load-settlement behaviour, adhesion factor () and
bearing capacity factor (N) of bored piles in Bangkok soft
clay.
With the passage of time, design method and selection of
parameters for local subsoil were improved as a result of
research works carried out in 1980s. Differences between the
behaviour of driven piles and bored piles were well realized
from these studies. Ng (1983) presented the load distribution
characteristics of wet-processed bored piles founded in both
the first and second sand layers of Bangkok based on

219

1.20

performance monitoring and quality control testing.

Interpretation skills relevant to local soil conditions and
construction methods of these tests were significantly
improved in local industry. For instance, in early 1980, the
sonic integrity (seismic test) test results were needed to send
to the specialists abroad for interpretation which in turn made
the testing cost more expensive. Significant cost-saving were
achieved in some major projects of late 1990s, as more
practical and precise interpretation were locally made to verify
and establish acceptance criteria in proving the quality of
suspected piles with anomalies.

Bored Cast Insitu Piles

1.00

Adhesion Factor ,

Tomlinson,1957
0.80
Suchada,1989
for bored piles in Bangkok soils

0.60

0.40

Peak
Stas and Kulhawy, 1984

0.20

Residual

9. KODEN DRILLING MONITORING SYSTEM

0.00
0.00

5.00

10.00

15.00

20.00

25.00

30.00

In Thailand, prior to the availability of Koden testing

equipment, borehole verticality was checked by mechanical
type equipment such as plumb float and adjustable globe.
However, the reliability of these methods was doubtful. The
Koden drilling monitor system was believed to be first used in
Thailand in 1980. The verticality of slurry-filled borehole can
be monitored rapidly from the electronic plot of the
continuous profile by Koden equipment. This system is very
useful for the verticality-control of long barrettes - unlike
bored piles, guiding by temporary casing is not available for
barrette drilling.

Undrained Shear Strength (ton/m )

Fig. 12 Comparison of adhesion factor, suggested by

different researchers with the actual mobilized in the stiff clay
layers (after Thansnanipan et. al 1999)
1.0

Submaneewong (1999) : Bentonite bored Piles (2nd Sand)

Submaneewong (1999) : Bentonite Bored Piles (2 nd Hard Clay)
Chiewcharnsilp (1988)

0.8

Polymer Bored Pile TP-1 (1st Sand)

(Ks tan

Polymer Bored Pile TP-2 (1st Sand)

Polymer Bored Pile TP-3 (1st Sand)

0.6

Polymer Bored Pile TP-1 (2nd Sand)

Design Line for

bentonite bored piles
Submaneewong (1999)

Polymer Bored Pile TP-2 (2nd Sand)

Polymer Bored Pile TP-3 (2nd Sand)

0.4

10. SONIC INTEGRITY TEST

0.2
0.0
28

30

32

34

Angle of Internal Friction, '

36

38

Fig. 13 Back-calculated values of polymer bored piles at

maximum test load plotted on design line of bentonite bored
piles constructed in Bangkok subsoil
(after Thansnanipan et. al 2002b)
Applied Load (KN)

Pile head displacement (mm)

0
0
10

5000

10000 15000 20000 25000 30000 35000 40000 45000 50000

Cycle1
Cycle2
Cycle3

20

Sonic integrity test, also known as seismic test is the most

common method of integrity testing for both driven and bored
piles in Thailand. Sonic integrity test is usually selected for
both quality check (control test) and retrospective
investigation. It is the cheapest in terms of cost and the
simplest in terms of testing process. The main advantage of
this test is that since no particular preparation is necessary
during the pile construction phase, it is more flexible to select
which pile is to be tested. However, interpretation of sonic
integrity testing needs considerable experience and knowledge
in testing, subsoil condition and construction method. In many
projects it is a part of the contractual requirement to conduct
sonic integrity test. Minimum 10 % to maximum 100 % of
production piles are commonly tested. It is also a reasonably
acceptable method as a retrospective investigation in
determining integrity of the pile. The signal characteristics and
their interpretations of sonic integrity test on piles founded in
Bangkok subsoil were reported in details by Thasnasipan et al.
(1998d).

30

11. CROSS- HOLE SONIC LOGGING TEST

40

The first sonic logging test was believed to be conducted in

1982 for the wet-processed bored piles of the Memorial
Bridge Project where bad concrete zones were detected at
depth about 20m and 1m Ng, 1983. Use of the Sonic logging
test for checking pile integrity has increased in Thailand
particularly in mega projects. Cross-hole sonic logging test is
relatively expensive. It is mainly employed as a pre-planned
site quality control testing. The major advantage of this
method is that test can be carried out shortly after pile
construction. Hence, rectification measures can be
implemented if the pile is defective while the foundation
contractor is on site. However, this method is generally not

50

Fig. 14 Static pile load test result of 1.80m diameter bored

pile of depth 60m in Bangkok
8. DEVELOPMENT IN PERFORMANCE MOINTORING
AND QUALITY CONTROL TESTING
Improvement of testing equipment and powerful computer
facilities are key factors that contributed to the development in

220

applicable if pile integrity is in question due to postconstruction activities, as access tubes are usually grouted
after completion of the test. The results from sonic logging test
conducted on model piles in Bangkok helped to extend the
knowledge of the signal characteristics and interpretation
(Thasnanipan et. al., 2000b). Thasnanipan et al., (2004b) also
demonstrated the application of cross-hole sonic logging test
in identifying over-cast length of bored piles prior to exposing
the pile heads.

Thailand. However, in the authors opinion, there is much

work to be done with particular focus on constructability
issues, concrete technology for wet-processed bored piles and
barrette, reliable but cost-effective quality control testing and
value-engineering.
Starting from the planning stage, site investigation, design,
construction and inspection should be integrated so that
designers, contractors and construction inspectors can
participate as a team with a common goal. Appropriate and
practical specifications should be established jointly by these
parties for local soil conditions and construction methods.
Practical acceptance criteria should be developed to verify
bored piles and barrettes with suspected anomalies.
Continuing education should be promoted for designers,
inspection engineers, and contractors. The Geotechnical
Chapter of the Engineering Institute of Thailand under the
royal patronage of his majesty King Bhumibol Adulyadej, has
started to establish the standard code of practice and
guidelines for wet-processed bored piles which will serve as a
yardstick for the deep foundation industry upon its completion
in the near future.

12. HIGH STRAIN DYNAMIC LOAD TEST

High strain dynamic integrity test has become a well-accepted
method especially for evaluating the pile capacity in todays
foundation industry of Thailand and it is applied for both
driven and bored piles. A large number of related technical
papers and case histories of the test have been published and it
is a part of standards and specifications such as ASTM
D4945-89 (Standard Method for High-strain Dynamic Testing
of Piles). Thasnanipan et al. (2000b), reported the application
of dynamic load testing on piles in Thailand.

REFERENCES

13. FURURE TRENDS OF CAST-IN-PLACE DEEP

FOUNDATIONS IN THAILAND

Bowles, J. E. 1968. Foundation analysis and design,

McGRaw-Hill, New York
Broms B. B. 1966. Methods of calculating the ultimate
bearing capacity of piles a summary, Soils-Soils, No. 1819 : 21-32
Chiewcharnsilp, S. 1988. Static load transfer and predictions
of carrying capacity of bored piles (in Thai), M. Eng.
Thesis, Chulalongkorn University, Bangkok, Thailand
Chiruppapa, P. 1968. Cast in-situ bored piles in Bangkok
Clay, M. Eng. Thesis No. 213, Asian Institute of
Technology, Bangkok, Thailand
Meyerhof, G. G. 1976. Bearing capacity and settlement of pile
foundations, Journal of soil mechanic and foundation
division, Proceeding ASCE Vol. 102, GT3 : 197-223
Ng, K. C. 1983. The construction problem and performance of
large bored piles in second sand layer, M. Eng. Thesis No.
GT 82-26, Asian Institute of Technology, Bangkok,
Thailand
Pimpasugdi Suchada 1989. Performance of Bored, Driven and
Auger Press Piles in Bangkok Subsoils, M. Eng. Thesis
No. GT 88-12, Asian Institute of Technology, Bangkok,
Thailand
Promboon, S . and Brenner, R.P. 1980. Large diameter bored
piles in Bangkok clay, Proceedings of 10th conference on
soil mechanic and foundation engineering, Vol. 2 : 815818
Sakaret Naka 2004. Bidirectional pile test on bored piles in
Bangkok subsoils (in Thai), M. Eng. Thesis,
Chulalongkorn University, Bangkok, Thailand
Suwanakul, V. 1969. Cast in-situ piles in Bangkok clay, M.
Eng. Thesis No. 260, Asian Institute of Technology,
Bangkok, Thailand

As the practitioners in construction industry have gained more

experience and confidence in using higher capacity bored piles
and barrettes, these types of deep foundations are expected to
be of more in demand in the future. The higher load capacity
achievable by deep-seated large bored piles and barrettes will
allow the use of a single foundation element in place of a
group of smaller size shallow-seated bored piles and driven
piles. It is expected that improvement in installation
techniques will be of key advance in the future of deep
foundation industry in Thailand.
It is anticipated that barrettes will be more popular in the
future for the following reasons.
The availability of more powerful and efficient hydraulic
grabs which will offer faster construction of deep-seated
barrettes
Less noise which will offer main advantages in congested
and sensitive neighbourhood
Less equipment requirement that will offer advantageous
construction in limited space
More versatility (e.g. able to construct in the area with
limited headroom)
Better in safety aspects as less equipment are required
(especially for the construction along the public roads,
subway stations, elevated expressway etc.)
14. CONCLUDING REMARKS
In Thailand, according to the authors experience as a deepfoundation contractor, development and achievement in both
construction and design aspects of wet-process deep bored
piles and barrette foundations in the past four decades were
significant. With recognition of technical and economic
advantages of using these high capacity cast-in-place
foundations by local practitioners, it is expected that they will
be more popular in the future construction industry of

221

Thasnanipan Narong, Zaw Zaw Aye, Thayanan Boonyarak

and Natamon Kampananon 2004b. Quantifying Pile Head
Condition Before Basement Excavation by Cross-hole
Sonic Logging Tests, The Seventh International
Conference on Application of Stress-Wave Theory to
Piles, Kuala Lumpur, Malaysia, August 9 11, 2004 :
233-238.
Thasnanipan N., Manosuthisan C., Boonyarak T., Chanchad
C., 2006. Experience in Construction of Largest Diameter
Deep-seated Bored Piles at the Bank of Chaopraya River,
Proceedings of the Eleventh National Convention on Civil
Engineering, Phuket, April 20 22, 2006 (in Thai)

Thasnanipan N., Teparaksa W., Maung A.W., Ganeshan B.,

1998a. Design, construction and behavior of bored cast inSitu concrete piles in Bangkok Subsoil, Proceedings:
Fourth International Conference on Case Histories in
Geotechnical Engineering, St. Louise, Missouri, U.S.A. :
281-287.
Thasnanipan N., Maung A.W., Tanseng P. 1998b. Barrettes
founded in Bangkok subsoils, construction and
performance, Thirteenth Southeast Asian Geotechnical
Conference, Taipei, Taiwan, ROC :.573-578.
Thasnanipan N., Ganeshan B., Anwar M.A. 1998c. Effect of
Construction Time and Bentonite Viscosity on Shaft
Capacity of Bored Piles, Third International Geotechnical
Seminar on Deep Foundation on Bored and Auger Piles,
Ghent, Belgium : 171-177.
Thasnanipan N., Maung A.W., Ganeshan B. 1998d. Sonic
Integrity Test on Piles Founded in Bangkok SubsoilSignal Characteristics and Their Interpretations,
Proceedings: Fourth International Conference on Case
Histories in Geotechnical Engineering, St. Louis,
Missouri, U.S.A. : 1086-1092
Thasananipan N., M. A. Anwar Maung A. W 1999. Failure
Mechanism of Long Bored Piles in Layered Soils of
Bangkok, Civil and Environmental Engineering
Conference, Bangkok, Thailand : V-69 to V-73.
Thasnanipan N., Tanseng P., Maung A.W. & Anwar M. A.
2000a. Barrettes : A Versatile Foundation for
Transmission Line Towers, An International Conference
on Geotechnical & Geological Engineering, Melbourne,
Australia, 19-24 November 2000
Thasnanipan N., Maung A. W. & Navaneethan T.., Z. Z. Aye
2000b. Non-Destructive Integrity Testing on Piles
Founded in Bangkok Subsoil, 6th International Conference
on Application of Stress-Wave Theory to Piles, September
2000, Sao Paulo, Brazil
Thasnanipan N., Zaw Zaw Aye and Wanchai Teparaksa.
2002a. Barrette of Over 50,000 kN Ultimate Capacity
Constructed in the Multi-Layered Soil of Bangkok,
Proceedings of the International Deep Foundations
Congress 2002, Geotechnical Special Publication No.
116, Volume Two, ASCE, Febuary 14-16 2002, Orlando,
Florida, USA : 1073-1087.
Thasnanipan, N., Zaw Zaw Aye, Chanchai Submaneewong
and Wanchai Teparaksa 2002b. Performance of WetProcess Bored Piles Constructed with Polymer-Based
Slurry in Bangkok Subsoil,
Proceedings of the
International Deep Foundations Congress 2002,
Geotechnical Special Publication No. 116, Volume One,
ASCE, February 14-16 2002, Orlando, Florida, USA :143157.
Thasnanipan, N., Zaw Zaw Aye and Chanchai Submaneewong
2004a. Effectiveness of Toe-grouting for Deep-seated
Bored Piles in Bangkok Subsoil, Proceedings of sessions
of the Geosupport Conference: Innovation and
Cooperation in the Geo-Industry, Geotechnical Special
Publication No. 124, ASCE, January 29-31, 2004,
Orlando, Florida, USA : 561-572.

222

Proceedings of Geotechnical and Foundation Conference 2012

26-27 September 2012, Bangkok, Thailand

Geotechnical and Foundation

Conference 2012

Barrette for Highrise Buildings and Heavy

Structures in Bangkok Subsoil
Narong Thasnanipan, Zaw Zaw Aye, Phongpat Kitpayuck and
Nutthapon Thasnanipan
SEAFCO Public Co., Ltd. Bangkok, Thailand

Thayanan Boonyarak
Hong Kong University of Science and Technology

223

Barrette for Highrise Buildings and Heavy Structures

in Bangkok Subsoil
Narong Thasnanipan1, Zaw Zaw Aye1, Thayanan Boonyarak2, Phongpat Kitpayuck1
Nutthapon Thasnanipan1
1
Seafco Public Company Limited: [email protected]
2
Hong Kong University of Science and Technology: [email protected]
Based on the authors experiences, this paper is intended to summarize the development of
deep-seated barrettes construction in Thailand, in comparison with that of the past. Literature
related to wet-processed bored piles and barrettes in Thailand published through out the past
30 years is summarized together with recent research works. Results of the static load test
performed on barrettes in three projects are summarized. Increasingly use of barrettes for the
highrise building projects is described. Recent application of heavy structures barrettes for
elevated mass rail transport project, MRT Purple Line is also presented. An attempt is also
made to briefly discuss the future trend of deep foundations in Thailand.
INTRODUCTION
During the past three decades, owing to the acceleration of development in mega
infrastructure projects, techniques and practice of cast-in-place deep foundations, particularly
wet-processed deep-seated large diameter bored piles and barrettes have experienced
enormous progress in Thailand. The versatility of the construction method and the high-load
capacity which in turn offered the constructability and cost-saving are the main factors
contributed to the increasing use of those cast-in-place deep foundations. In the initial stage,
there were a number of questions on design and construction aspects of these foundation
systems particularly in Bangkok subsoil. With the passage of time, the construction
equipment, installation techniques and testing of deep foundation elements have been
developed. Large numbers of instrumented full-scale static pile load tests were conducted
through out the 1990s mainly in mega projects, which provided better understanding on
behavior of these deep-seated foundations.
Research focused on the design parameters and methods were produced based on
these test results. The design parameters became so well established that wet-process cast-inplace foundations became regarded as reliable foundations for practitioners involved in
construction industry in Thailand. This paper focus the past and current practice of barrettes
in Thailand together with a brief discussion on future trend deep foundations in Bangkok
subsoil. The information contained in the paper is mainly from the highrise buildings and
infrastructure projects constructed in Bangkok and adjacent areas.

225

SUBSOIL AND EXISTING PIEZOMETRIC PROFILE

Subsoil profile and the present piezometric drawdown condition of Bangkok are
presented in Figure 1 below. A typical subsoil profile is relatively consistent in different
localities in Bangkok. It is characterized by alternating layers of clay and sand deposits as
shown in Figure 1.
2
Pore Pressure (kN/m )

Geotechnical
Parameters

Bangkok Soft
Clay

400

600

Su= 10-25 KPa

t = 14-16kN/m 3
-

10

50

ine

SPT-N= 50-100
t = 21KN/m 3

e
Lin

Dense to Very
Dense Sand
(Second Sand)

40

tic L

50

Su> 200 KPa

= 21KN/m 3
t

wn
wdo

Hard Clay

ta
ros

40

30
SPT-N= 20-40
t = 20kN/m 3

D ra

Medium
Dense Sand
(First Sand)

H yd

30

Su= 50-14020- KPa

t = 17-21kN/m 3

c
etri

Medium to
Stiff Clay

m
iezo
al P

20

60

200

Weathered Crust

u
Act

Depth (m)

10

60

Figure 1. Typical soil profile of Bangkok with piezometric drawdown condition

(Thasnanipan, et. al., 2002)
BARRETTE CONSTRUCTION METHOD IN BANGKOK SUBSOIL
Due to the prevailing subsoil and groundwater conditions, barrettes are constructed by
wet-process or slurry displacement method.
Both mechanical and hydraulic cable-suspended grab are used for barrette
construction. In recent years, use hydraulic grabs are more common. Excavation of the trench
is carried out by the cyclic-process of lifting and lowering of the grab under gravity and
tangential force of the clamshell operated by cables (mechanical grab) or hydraulic action
(hydraulic grab). Different from bored pile construction, a guide wall of depth 1 to 1.5 m with
inside clear dimensions slightly larger than the nominal size of the barrette is used to guide
the grab during initial bites. Bentonite slurry is introduced to the trench as soon as the initial
excavation commenced. The excavation is continued under the bentonite slurry to the final
depth. After recycling of slurry and lowering the rebar cage, concreting is done by tremie
method.

226

GUIDE WALL
TREMIE PIPE

REINFORCEMENT
CAGE
BENTONITE
SLURRY

EXCAVATION

INSTALLATION OF
REINFORCEMENT

CONCRETING

Figure 2. Schematic of barrette construction sequences

OVERVIEW OF BARRETTE APPLICATION IN CONSTRUCTION PROJECTS
In some projects where bored piles were not feasible due to the site constraints,
applicable construction methods, and/or extensive bearing capacity requirements, the use of
barrette foundations would make a suitable alternative. Barrettes with dimensions ranging
from 0.80 m x 2.7 m to 1.5 m x 3.0 m for safe working load capacity from 11,000 to 23,000
kN have been used in some major projects.
The first barrette in Bangkok was believed to be constructed in the late 1970s for the
foundation of the Bangkok Bank Head Office Building at Silom road with size of 0.6-0.8 x
2.5m and toe depth of 33m. Barrettes can be constructed with flexible layout plans for both
vertical and lateral loads. The layout pattern of barrettes can be arranged in a continuous row
or column, radial, alternating long and short axis of barrette and a combination of two or more
of such patterns. To minimize the need of additional equipment, a mechanical or hydraulic
grab mounted on crawler crane can be used for constructing both barrettes (as piledfoundation) and diaphragm walls (as earth retaining structure) in some suitable projects
without employing a bored piling rig, which could save the extra mobilization cost. In
addition to the large bearing capacity requirements, on site difficulties such as limited head
room where piling rigs cannot be utilized, under such situations like presence of overhead
high voltage power cables, existing overpasses or structures for elevated expressways and
planned subway stations, also demands the barrettes. Static load test up to 52,900 KN
conducted on barrette set the record as the highest load ever tested for a single barrette
foundation in Thailand (Thasnanipan et. al. 2002a). A summary of completed projects
constructed with barrettes with respect to selection criteria in 2003 in comparison with 2012 is
presented in Table 1. As can be seen in the table, significant increase of barrette foundation
can be realized particularly in the project with high load capacity requirement. It is also
evident that barrette in combination with diaphragm wall is increasingly popular in the
construction industry of Thailand.

227

Table 1 Summary of Barrette Selection in major projects

Selection Criterion
Accumulated number of projects
as of 2003
as of 2012
August
High Load Capacity
4
21
Minimize Construction
Equipment
Limited Head Room for
Excavation

14

Combination with
Diaphragm Wall
Foundation as well as
portion of column

10

25

Remarks
Incorporated with bored
piles
Alternative for bored piles
Under existing structures
such as bridges, elevated
expressway and power lines
As diaphragm wall legs
Provision for the future
requirement

General configuration of barrettes as leg piles integrated with diaphragm wall and bored piles
of building with deep basement is shown in Figure 3.

Figure 3 Schematic of barrettes as leg piles integrated with diaphragm walls and bored piles

PROBLEMS AND DIFFICULTIES IN CONSTRUCTION OF BARRETTES IN

EARLY DAYS
Basic but extensive problems were experienced in early stages of barrette construction in
Thailand as summarized below.
Limited availability and capacity of equipment
Lack of skills in operation of equipment (particularly the drilling rig)

228

Adverse effect due to a slow rate of drilling (e.g. excessive formation of filter-cake by
using bentonite slurry)
Limited knowledge and less advanced techniques in control of bentonite slurry
Limited experience in construction method and related negative impact
Quality of concrete for tremie concreting method
Lack of experienced engineers and foremen
Improper construction and quality control specifications and guidelines for deepseated piles in local soil
Improper design for constructability

IMPROVEMENT IN CONSTRUCTION OF BARRETTE

Over the past three decades, along with the development of wet-processed piling
technology in other parts of the world, equipment, construction technique and design methods
as well as better understanding of construction impact on the performance of this type of
foundation have significantly improved in Thailand. Table 2 summarizes the areas of
improvement in bored pile and barrette construction and main factors contributed to these
developments. Some of these improvements were published both locally and internationally.
The design, construction and behavior of bored cast in-situ concrete piles in Bangkok Subsoil
was presented by Thasnanipan et. al., (1998a). The construction and performance of barrettes
in Bangkok Subsoil was also reported by Thasnanipan et. al., (1998b). Effect of construction
time and bentonite viscosity on shaft friction of bored piles was also pointed out by
Thasnanipan et al., (1998c). Thasnanipan et. al., (2004a) also reported a comprehensive study
on the effectiveness of two different toe-grouting methods, known as tube--manchette and
drill-and-grout, applied in Bangkok.

(a)

(b)

Figure 4. (a) Guide wall for cruciform barrette (b) View of cruciform barrette after
installation of reinforcement
Construction of cruciform barrette (shown in Figure 4) for monopole type highvoltage power transmission line was described in the work of Thasnanipan et. al., (2000a).
Construction of deep-seated T-shape and L-shape barrettes as leg-piles for diaphragm wall are
now more common in Bangkok.

229

Above-mentioned research works reflect the development history of bored pile and
barrettes in Thailand and offered useful information to the local construction industry and
perhaps to the international deep foundation engineering society.
Table 2 Summary of development in barrette construction
Area of
development /
improvement
Speed of
construction

In the past
(Early 1980s)

At present

Main factor
contributed to
development
Better equipment,
operating skills as
well as improvedknowledge in
construction method,
management and
local soil condition
Better equipment,
operating skills as
well as improvedknowledge in
construction method,
management and
local soil condition
Equipment
availability and
advance technology
Experience from past
projects and research

Minimum 3 days
required to complete
barrette 1.5mx3m of toe
depth 57m

Less than 2 day to

complete barrette
1.2mx3m toe depth
60m

Pile size and

depth

Barrette
Limited in size and depth

Barrette
Various sizes &
depth over 60m

Base grouting

Not available

Available

Construction
impact on quality
and performance
Quality control in
construction
process
Quality control
test method and
interpretation

Not well understood

Improving

Not well established and

systematic

Well established
and systematic

Experience from past

projects and research

Fewer methods available

and limited knowledge in
interpretation

Better equipment
available and better
knowledge in
interpretation

Advance equipment,
experience from past
projects and research

Experience from past projects and extensive research works provided better
understanding of construction impact of quality and performance of these wet-processed deep
foundations.
The availability of more powerful and efficient hydraulic grabs offers faster
construction of deep-seated barrettes in comparison with mechanical cable-hung grabs. Figure
5 shows the construction of barrettes using hydraulic grab.

230

Figure 5. View of barrette construction by hydraulic grab

IMPROVEMENT IN DESIGN CONCEPT AND PARAMETERS SELECTION

Usually the designers use the same design concept and parameters for barrettes and
bored piles constructed under bentonite slurry. Following paragraphs describe the design
practice of both elements commonly applied in the past and improvement with time.
In the initial stage of introducing bored piles in Thailand, the design concepts and
parameters were mainly based on the available literature from research carried out in other
parts of the world such as Tomlinson (1957), Skempton (1959), Broms (1966), Bowles
(1968), Meyerhof (1976) etc. The work of Chiruppapa (1968) was believed to be the first
research data available for design parameters of bored piles in Bangkok soft clay. The author
conducted the study based on 6 dry-processed small diameter bored piles with load cells
installed at the pile toes on AIT campus in Phathumthani. Though the piles were simply
constructed using the dry-process with casing method, this early-stage study provided some
important information such as load-settlement behavior, adhesion factor () and bearing
capacity factor (N) of bored piles in Bangkok soft clay.
With the passage of time, design method and selection of parameters for local subsoil
were improved as a result of research works carried out in 1980s. Differences between the
behavior of driven piles and bored piles were well realized from these studies. Ng (1983)
presented the load distribution characteristics of wet-processed bored piles founded in both
the first and second sand layers of Bangkok based on instrumented (strain gauges) pile load
test results. Chiewcharnsilp (1988) reported the shaft friction factor values of sand layers in
Bangkok based on the instrumented load test results. In addition to the literature previously
available (Chiruppapa, 1968; Suwanakul, 1969; Promboon, 1981; Ng, 1983; Chiewcharnsilp,
1988), Pimpasugdi (1989) determined the shaft friction factor (adhesion factor, ) of Bangkok
clay layers based on 11 bored piles of diameters ranging from 0.50 to 0.80m with the
embedded length varying between 21.50m and 46m. It should be noted that the design
parameters obtained from the research works of 1980s were mainly based on the estimation of
shaft friction loads from plain static pile load test results with numbers of assumptions since
instrumented pile load test results were limited.

231

With the peak of construction-boom, large numbers of instrumented full-scale static

load tests on bored piles and a few barrettes were conducted throughout 1990s which
provided better understanding on behavior of these deep-seated foundations. Researches
focused on the design parameters and methods were published based on these test data. With
improved design methods the wet-process cast-in-place foundations came to be regarded as
reliable foundations in the construction industry of Thailand. With more confidence on soil
parameters selection and better understanding on behavior of these deep foundations, the
designer designed higher load capacity bored piles and barrettes in late 1990 than those in
1980s. Thasnanipan et. al., (1999) reported the failure mechanism of long bored piles in
layered soil of Bangkok. The authors cited that for the bored piles embedded in the multi
layered soils of Bangkok, estimation of ultimate shaft friction capacity needs to consider the
brittle type of failure mechanism of stiff to hard clay layers and the values selected need to
be adjusted accordingly. Peak and residual values mobilized in the stiff clay layers analyzed
by the authors were plotted as shown in Figure 10 along with the suggested curves by
different researchers. It can be seen from the figure that the residual value of 2nd stiff clay
layer (undrained shear strength values of 25 ton/m2), at the maximum test load, drops below
the curve suggested by Pimpasugdi (1989).So the values proposed by the author in Figure
6 for the stiff to hard clay layers overestimate the ultimate shaft friction under these
conditions.

1.20
Bored Cast Insitu Piles
1.00

Adhesion Factor ,

Tomlinson,1957
0.80
Suchada,1989
for bored piles in Bangkok soils

0.60

0.40

Peak
Stas and Kulhawy, 1984

0.20

Residual

0.00
0.00

5.00

10.00

15.00

20.00

25.00

30.00

Undrained Shear Strength (ton/m )

Figure 6 Comparison of adhesion factor , suggested by different researchers with the actual
mobilized in the stiff clay layers (after Thansnanipan et. al 1999)

232

1.0

Submaneewong (1999) : Bentonite bored Piles (2nd Sand)

Submaneewong (1999) : Bentonite Bored Piles (2 nd Hard Clay)
Chiewcharnsilp (1988)

0.8

Polymer Bored Pile TP-1 (1st Sand)

(Ks tan

Polymer Bored Pile TP-2 (1st Sand)

Polymer Bored Pile TP-3 (1st Sand)

0.6

Polymer Bored Pile TP-1 (2nd Sand)

Design Line for

bentonite bored piles
Submaneewong (1999)

Polymer Bored Pile TP-2 (2nd Sand)

Polymer Bored Pile TP-3 (2nd Sand)

0.4
0.2
0.0
28

30

32

34

36

38

Angle of Internal Friction, '

Figure 7 Back-calculated values of polymer bored piles at maximum test load plotted on
design line ( values in sand layer) of bentonite bored piles based on back-calculated
(after Thansnanipan et. al 2002b)

DEVELOPMENT IN PERFORMANCE MOINTORING

AND QUALITY CONTROL TESTING OF BARRETTES
Improvement of testing equipment and powerful computer facilities are key factors
that contributed to the development in performance monitoring and quality control testing.
Interpretation skills relevant to local soil conditions and construction methods of these tests
were significantly improved in local industry. For instance, in early 1980, the sonic integrity
(seismic test) test results were needed to send to the specialists abroad for interpretation which
in turn made the testing cost more expensive. Significant cost-saving were achieved in some
major projects of late 1990, as more practical and precise interpretation were locally made to
verify and establish acceptance criteria in proving the quality of suspected piles or barrettes
with anomalies.
Koden drilling monitoring system
In Thailand, prior to the availability of Koden testing equipment, borehole verticality
was checked by mechanical type equipment such as plumb float and adjustable globe.
However, the reliability of these methods was doubtful. The Koden drilling monitor system
was believed to be first used in Thailand in 1980. The verticality of slurry-filled borehole can
be monitored rapidly from the electronic plot of the continuous profile by Koden equipment.
This system is very useful for the verticality-control of long barrettes - unlike bored piles,
guiding by temporary casing is not available for barrette drilling.

233

Sonic integrity test

Sonic integrity test, also known as seismic test is the most common method of
integrity testing for both driven and bored piles in Thailand. Sonic integrity test is usually
selected for both quality check (control test) and retrospective investigation. It is the cheapest
in terms of cost and the simplest in terms of testing process. The main advantage of this test
is that since no particular preparation is necessary during the pile construction, phase, it is
more flexible to select which pile is to be tested. However, interpretation of sonic integrity
testing needs considerable experience and knowledge in testing, subsoil condition and
construction method. In many projects it is a part of the contractual requirement to conduct
sonic integrity test. Minimum 10 % to maximum 100 % of production piles are commonly
tested. It is also a reasonably acceptable method as a retrospective investigation in
determining integrity of the pile. The signal characteristics and their interpretations of sonic
integrity test on piles founded in Bangkok subsoil were reported in details by Thasnasipan et.
al. (1998d).
Cross-hole sonic logging test
The first sonic logging test was believed to be conducted in 1982 for the wetprocessed bored piles of the Memorial Bridge Project where bad concrete zones were detected
at depth about 20m and 1m Ng, 1983. Use of the Sonic logging test for checking pile
integrity has increased in Thailand particularly in mega projects. Cross-hole sonic logging
test is relatively expensive. It is mainly employed as a pre-planned site quality control testing.
The major advantage of this method is that test can be carried out shortly after pile
construction. Hence, rectification measures can be implemented if the pile is defective while
the foundation contractor is on site. However, this method is generally not applicable if pile
integrity is in question due to post-construction activities, as access tubes are usually grouted
after completion of the test. The results from sonic logging test conducted on model piles in
Bangkok helped to extend the knowledge of the signal characteristics and interpretation
(Thasnanipan et. al., 2000b). Thasnanipan et. al., (2004b) also demonstrated the application of
cross-hole sonic logging test in identifying over-cast length of bored piles prior to exposing
the pile top.
High strain dynamic load test
High strain dynamic integrity test has become a well-accepted method especially for
evaluating the pile capacity in todays foundation industry of Thailand and it is applied for
both driven and bored piles. A large number of related technical papers and case histories of
the test have been published and it is a part of standards and specifications such as ASTM
D4945-89 (Standard Method for High-strain Dynamic Testing of Piles). Thasnanipan et. al.
(2000b), reported the application of dynamic load testing on piles and barrettes in Thailand.
STATIC PILE LOAD TESTS ON BARRETTES
Figure 8, 9 and 10 show the load vs. pile head movement and load-transfer mechanism
of two standard rectangular-shape barrettes and T-Shape barrette respectively.

234

Applied Load (Tons)

Pile Head Settlement (mm)

1,000

2,000

3,000

4,000

5,000

6,000

20

40

60
Barrette 1.5x3.0 m, depth 57.5 m
80

(a)

Load (Ton)

Su (UC) t/m2
5

10

1000

2000

3000

4000

5000

Weathered
Crust
Soft
Clay

10

20

First Stiff
Clay

30

First
Sand

40

Hard
Clay

50

3.00 m

Barrette

Dense Sand
Hard Clay
Second
Sand

60
0

50

Tip Level
-57.5 m

100

60

SPT-N (blows/30cm)

VWSG
Mechanical
Extensometer

(b)
Figure 8 (a) Load vs. pile head movement and (b) load-transfer mechanism of base-grouted
barrette 1.50x3.00m founded at 57.5 m in Bangkok

235

Applied Load (Tons)

Pile Head Settlement (mm)

1,000

2,000

3,000

4,000

5,000

6,000

7,000

10

20

30

Barrette 1.2x3.0 m, depth 65 m

40
Figure 9 Load vs. pile head movement of barrette 1.20x3.00m founded at 65 m in Bangkok
Applied Load (Tons)

Pile Head Settlement (mm)

1,000

2,000

3,000

4,000

5,000

6,000

10

20

30

Barrette 1.0x2.0x3.0 m, depth 55 m

40

Figure 10 Load vs. pile head movement of T-shape barrette (1.00x2.00x3.00m) founded at
55m in Bangkok (constructed in tested in 2008)

236

Table 3 Summary of load-settlement curve from three case studies

Test No. 1
Test No. 2
(1.5m x 3.00m)
Barrette Test 2
Description
Base-grouted
(1.2m x 3.0m)
Dimension

1.50m x 3.00m

1.2m x 3.0m

Test No. 3
T-shape barrette
(1.0m x 2.0m x
3.0m)
1.0m x 2.0m x 3.0
m (T)
-55.0
770
5.00
Bentonite
6

Tip Level (m)

-57.5
-65.0
Shaft friction area (m2)
518
546
End Bearing area (m2)
4.50
3.60
Slurry Used
Bentonite
Bentonite
Total sand layer thickness
18
33
along the shaft (m)
Design load (ton)
2,300
2,500
2,600
Maximum pile head movement
62
25
28
at maximum applied load (mm)
Load at 5mm pile head
2,515
2,470
2,600
movement (ton)
Load at 20mm pile head
5,290
6,100
5,270
movement (ton)
Load at fully mobilized shaft
5,060
6,470
4,600
friction (ton)a
Pile head movement at fully
20
24
17
mobilized shaft friction (mm)
Unit end bearing at maximum
65b
130b
70b
test load (ton/m2)
Note : a. Assumed shaft friction fully mobilized at plunging point of load-settlement curve
b. Unit end bearing not fully mobilized
Though it is not advisable to remark major conclusions from 3 barrette tests, following
summary can be drawn.
Barrette head displacement (settlement) at design load is less or equal to 5mm
Barrette head displacement (settlement) at assumed fully mobilized shaft friction is in
the order of 1.3 to 2% of shorter width of barrette (thickness)
End bearing does not contribute to the ultimate capacity of barrette
Thickness of sand layer along the pile shaft may have influence on the barrette shaft
friction capacity (as can be seen in Test No. 2 where sand layer thickness along the
shaft is significantly thicker than other 2 test barrettes)
BARRETTES FOR HIGH-RISE BUILDING PROJECTS
More use of barrettes for highrise buildings can be observed in recent years. Points
below are considered key factors contributed to increasing use of barrettes for highrise
buildings particularly in congestive urban area.
Buildings foot-prints are relatively small in comparison with the heights in majority of
the highrise buildings. Hence higher load capacity foundation is required within
available area
Equipment utilized for barrette construction are relatively quiet and less vibration so
that disturbance to the neighbouring area is less than that of piling equipment

237

Less equipment require for barrette construction one set of barrette excavation
equipment can be used for construction of barrette and diaphragm wall for the project
with limited space
Removal of excavated soil from the site can be easier as barrette grab can dispose
directly into the dump trucks without needing to store on site provided that dumping
trucks are able to move in and out from the site during excavation time (important
issue in most of the congestive area where traffic restriction is stringent)
Bored pile
T-shape barrette

Figure 11 Mat foundation of a high-rise building supported by barrette and bored pile

Figure 11 shows an example of mat foundation of a high-rise building supported by

combination of barrette and bored pile. As explained earlier, one of major benefits of using
barrette to support high-rise building is larger load bearing capacity of barrette compared with
that of bored pile. Barrette can be designed for the high load concentration area while bored
pile can be used in less concentrated load area as shown in the figure. In addition, barrette
provides a versatile option to be constructed in various shapes such as T-shape (Fig. 10), Xshape or cruciform (Fig. 4) and L-shape to meet requirement of load bearing and bending
moment capacity.
BARRETTES FOR HEAVY STRUCTURES - ELEVATED TRAIN PROJECT

Bored piles have been extensively used for foundations of the majority of the elevated
expressways since 1991. Thousands of base-grouted bored piles were constructed for these
infrastructure projects including Second Stage Expressway, constructed in 1991; Don Muang
Tollway Extension, constructed in 1997; Bangna-Bang Pli-Bangprakong Expressaway
constructed in 1998; and Wat Nakorn-In project, constructed in 2001. Over 700 bored piles of
diameter ranging from 0.8m to 1.5m with depth from 35m to 54m were constructed between
1999 and 2000 using polymer-based slurry for the foundation of Rama VIII Bridge, one of the
initiatives of his majesty the King Bhumibol Adulyadej.
However, in 2010, large numbers of barrettes were constructed for new MRT project
called Purple Line. Originally, the majority of foundations to support heavy elevated-

238

structures were designed for large bored piles of diameter 1m, 1.2m and 1.5m depending on
the loading requirement. After awarding the contracts to the main contractors, the value
engineering option using barrettes to replace bored piles proposed by the piling contractor
was accepted (as shown in Figure 12 and 13). Apart from the safety aspects, the key reasons
of changing from bored piles to barrettes are that (i) construction of barrettes requires less
equipment on the middle of the congestive road than bored piles (ii) pile cap (footing) for
barrette is much smaller than that of bored pile. Construction of over 1000 barrettes of
different sizes, 1x3m, 1.2x3m and 1.5x3m with toe depth ranging from 50m to 65m were
completed in the beginning of 2012.

Figure 12. Schematic of bored piles (original design) and proposed option of barrettes with
equivalent capacity size of footing for barrette (20m2 in plan area) is significantly less than
that of bored pile (60m2 in plan area)

Figure 13. Construction of barrette for Elevated Train Project in Bangkok

Comparison of equipment and construction aspects bored pile and barrette is shown in Table
4 below.

239

Table 4. Comparison of equipment and construction aspects of barrette and bored pile for
construction on the public roads

Item

Description

Barrette

Bored Pile

Total high-tech equipment requirement

High

Medium

Experience and skill of operation crew

High

Medium

Risk involved (safety aspect)

Medium

High

4
5

Cost of equipment
Area requirement on road

High
Less

Medium
More

Site and road cleanness

Excellent

Average

Traffic problem due to method of soil disposal

Less

More

LATER LOAD TEST ON T-SHAPE BARRETTE AND DIAMETER 2M BORED PILE

The comparison of field test results and 3D finite element analysis on full-scale
instrumented static lateral load tests on T-shape barrette (Figure 14) and bored piles
constructed in Bangkok were presented by Submaneewong C (2009). Figure 15 presents the
Load-movement graphs of barrettes and bored piles under lateral loading. The findings
reported by the authors provide good reference for future design of similar elements subjected
to lateral loading, particularly result of stiffness ratio vs shear strain level of barrette and
bored pile as can be seen in Figure 16.
Soil Description

1
0
5
10
15

Depth below Ground Surface (m)

20

(a) Section and plan configuration of

lateral load test on T-shape barrettes

Very Soft to
Soft Clay
Medium Stiff Clay
Medium Dense Sand

15
20
-24.00

30

-40.00

40

Hard Clay
Hard Clay

55
60

-12.00

35

40

50

-7.00

10

25

Very Stiff Clay

35

45

-1.50

Stiff to Very Stiff Clay

25
30

1
0

45
50
Pile Tip at
- 55.00m

55

Very Dense Sand

1.00m

2.00m

65
70
75

S-3

S-4

Hard Clay

S-5
1.00m

3.00m

80

S-1
S-2

S-6
Inclinometer Tube
Strain Guage

(b) Soil profile and instrumentation

details of tested T-shape barrettes

Figure 14. Lateral load test configuration for T-shape barrettes and instrumentation details

240


5000

Load (kN)

4000
3000
T-shape barrette (T4BP13)

2000

T-shape barrette (T4BP6)

Bored pile (T4P13)
Bored pile (T4P14)

1000
0
0

10

20

30

40

50

60

70

80

90

100

Movement at Pile Head (mm)

Figure. 15. Movement at pile head under lateral load of barrettes and bored piles

700
600
500

G/Su

400
300
200
100
0
0.01

0.1

Shear Strain (%)

Figure 16. Strain level of T-barrette and bored pile (reported by Submaneewong 2009)
plotted on stiffness ratio vs shear strain for Bangkok Soft Clay reported by Prinzl and Davies
(2006)

241

FURURE TRENDS OF CAST-IN-PLACE DEEP FOUNDATIONS IN THAILAND

As the practitioners in construction industry have gained more experience and
confidence in using higher capacity bored piles and barrettes, these types of deep foundations
are expected to be of more demanding in the future. The higher load capacity achievable by
deep-seated large bored piles and barrettes will allow the use of a single foundation element
in place of a group of smaller size shallow-seated bored piles and driven piles. It is expected
that improvement in installation techniques will be of key advance in the future of deep
foundation industry in Thailand.
It is anticipated that barrettes will be more popular in the future for the following
reasons.

The availability of more powerful and efficient hydraulic grabs which will offer faster
construction of deep-seated barrettes
Less noise which will offer main advantages in congestive and sensitive
neighbourhood
Less equipment requirement which will offer advantages construction in limited space
More versatility (e.g. able to construct in the area with limited headroom)
Better in safety aspects as less equipment are required (especially for the construction
along the public roads, subway stations, elevated expressway etc.)

CONCLUDING REMARKS
In Thailand, according to the authors experience as a deep-foundation contractor,
development in both construction and design aspects of wet-process barrette foundations in
past decades were significant. With recognition of technical and economic advantages of
using these high capacity cast-in-place foundations by local practitioners, it is expected that
they will be more popular in the future construction industry of Thailand. However, in the
authors opinion, there is much work to be done with particular focus on constructability
issues, concrete technology for wet-processed bored piles and barrette, reliable but costeffective quality control testing and value-engineering.
Starting from the planning stage, site investigation, design, construction and
inspection should be integrated so that designers, contractors and construction inspectors can
participate as a team with a common goal. Appropriate and practical specifications should be
established jointly by these parties for local soil conditions and construction methods.
Practical acceptance criteria should be developed to verify bored piles and barrettes with
suspected anomalies. Continuing education should be promoted for designers, inspection
engineers, and contractors.
REFERENCES
Bowles, J. E. (1968) Foundation analysis and design, McGRaw-Hill, New York
Broms B. B. (1966) Methods of calculating the ultimate bearing capacity of piles a
summary, Soils-Soils, No. 18-19, pp. 21-32

242

Chiewcharnsilp, S. (1988) Static load transfer and predictions of carrying capacity of bored
piles (in Thai), M. Eng. Thesis, Chulalongkorn University, Bangkok, Thailand
Chiruppapa, P. (1968) Cast in-situ bored piles in Bangkok Clay, M. Eng. Thesis No. 213,
Asian Institute of Technology, Bangkok, Thailand
Meyerhof, G. G. (1976) Bearing capacity and settlement of pile foundations, Journal of soil
mechanic and foundation division, Proceeding ASCE Vol. 102, GT3, pp. 197-223
Ng, K. C. (1983) The construction problem and performance of large bored piles in second
sand layer, M. Eng. Thesis No. GT 82-26, Asian Institute of Technology, Bangkok,
Thailand
Pimpasugdi Suchada (1989) Performance of Bored, Driven and Auger Press Piles in Bangkok
Subsoils, M. Eng. Thesis No. GT 88-12, Asian Institute of Technology, Bangkok,
Thailand
Promboon, S . and Brenner, R.P. (1980), Large diameter bored piles in Bangkok clay,
Proceedings of 10th conference on soil mechanic and foundation engineering, Vol. 2,
pp. 815-818
Sakaret Naka (2004) Bidirectional pile test on bored piles in Bangkok subsoils (in Thai), M.
Eng. Thesis, Chulalongkorn University, Bangkok, Thailand
Submaneewong C. (2009) Behavior of T-shape barrette and bored pile subject to Vertical and
Lateral Loading, Ph.D Dissertation, Chulalongkorn University, Bangkok, Thailand
Suwanakul, V. (1969) Cast in-situ piles in Bangkok clay, M. Eng. Thesis No. 260, Asian
Institute of Technology, Bangkok, Thailand
Thasnanipan N., Teparaksa W., Maung A.W., Ganeshan B., (1998a) Design, construction and
behavior of bored cast in-Situ concrete piles in Bangkok Subsoil, Proceedings: Fourth
International Conference on Case Histories in Geotechnical Engineering, St. Louise,
Missouri, U.S.A., p.p.281-287.
Thasnanipan N., Maung A.W., Tanseng P., (1998b) Barrettes founded in Bangkok subsoils,
construction and performance, Thirteenth Southeast Asian Geotechnical Conference,
Taipei, Taiwan, ROC, p.p.573-578.
Thasnanipan N., Ganeshan B., Anwar M.A., (1998c), Effect of Construction Time and
Bentonite Viscosity on Shaft Capacity of Bored Piles, Third International
Geotechnical Seminar on Deep Foundation on Bored and Auger Piles, Ghent,
Belgium. p.p. 171-177.
Thasnanipan N., Maung A.W., Ganeshan B., (1998d), Sonic Integrity Test on Piles Founded
in Bangkok Subsoil-Signal Characteristics and Their Interpretations, Proceedings:
Fourth International Conference on Case Histories in Geotechnical Engineering, St.
Louis, Missouri, U.S.A., p.p.1086-1092

243

Thasananipan N., M. A. Anwar Maung A. W (1999) Failure Mechanism of Long Bored Piles
in Layered Soils of Bangkok, Civil and Environmental Engineering Conference,
Bangkok, Thailand, p.p. V-69 to V-73.
Thasnanipan N., Tanseng P., Maung A.W. & Anwar M. A. (2000a), Barrettes : A Versatile
Foundation for Transmission Line Towers, An International Conference on
Geotechnical & Geological Engineering, Melbourne, Australia, 19-24 November
2000
Thasnanipan N., Maung A. W. & Navaneethan T.., Z. Z. Aye (2000b), Non-Destructive
Integrity Testing on Piles Founded in Bangkok Subsoil, 6th International Conference
on Application of Stress-Wave Theory to Piles, September 2000, Sao Paulo, Brazil
Thasnanipan N., Zaw Zaw Aye and Wanchai Teparaksa (2002a) Barrette of Over 50,000 kN
Ultimate Capacity Constructed in the Multi-Layered Soil of Bangkok, Proceedings of
the International Deep Foundations Congress 2002, Geotechnical Special Publication
No. 116, Volume Two, ASCE, Febuary 14-16 2002, Orlando, Florida, USA, p.p. 10731087.
Thasnanipan, N., Zaw Zaw Aye, Chanchai Submaneewong and Wanchai Teparaksa (2002b)
Performance of Wet-Process Bored Piles Constructed with Polymer-Based Slurry in
Bangkok Subsoil, Proceedings of the International Deep Foundations Congress
2002, Geotechnical Special Publication No. 116, Volume One, ASCE, February 14-16
2002, Orlando, Florida, USA, p.p. 143-157.
Thasnanipan, N., Zaw Zaw Aye and Chanchai Submaneewong (2004a) Effectiveness of Toegrouting for Deep-seated Bored Piles in Bangkok Subsoil, Proceedings of sessions of
the Geosupport Conference: Innovation and Cooperation in the Geo-Industry,
Geotechnical Special Publication No. 124, ASCE, January 29-31, 2004, Orlando,
Florida, USA, p.p. 561-572.
Thasnanipan Narong, Zaw Zaw Aye, Thayanan Boonyarak and Natamon Kampananon
(2004b), Quantifying Pile Head Condition Before Basement Excavation by Cross-hole
Sonic Logging Tests, The Seventh International Conference on Application of StressWave Theory to Piles, Kuala Lumpur, Malaysia, August 9 11, 2004. p.p. 233-238.
Thasnanipan N., Manosuthisan C., Boonyarak T., Chanchad C., (2006) Experience in
Construction of Largest Diameter Deep-seated Bored Piles at the Bank of Chaopraya
River, Proceedings of the Eleventh National Convention on Civil Engineering, Phuket,
April 20 22, 2006

244

PROCEEDINGS OF THE THIRTEENTH

SOUTHEAST ASIAN GEOTECHNICAL
CONFERENCE
November 16-20 1998, Taipei, ROC

Performance of a Braced Excavation in Bangkok Clay,

Diaphragm Wall Subject to Unbalanced Loading Conditions
N. Thasnanipan, A. W. Maung, P. Tanseng and S.H. Wei
Seafco Co., Ltd., Bangkok, Thailand

Sponsored by;
Chinese Institute of Civil and Hydraulic Engineering
Southeast Asian Geotechnical Society

245

Thirteenth Southeast Asian Geotechnical Conference, 16-20 November, 1998, Taipei, Taiwan, ROC

Performance of A Braced Excavation in Bangkok Clay,

Diaphragm Wall Subject to Unbalanced Loading Conditions
N. THASNANIPAN

SEAFCO Co., Ltd., Bangkok, Thailand

A. W. MAUNG

SEAFCO Co., Ltd., Bangkok, Thailand

P. TANSENG

SEAFCO Co., Ltd., Bangkok, Thailand

S.H. WEI

SEAFCO Co., Ltd., Bangkok, Thailand

SYNOPSIS
0.8m thick cast-in-situ diaphragm wall having toe depth of 28m with two level temporary bracing was used
to construct the basements of a structure which is located near the river and surrounded by buildings, including a historical one
in Bangkok, Thailand. Due to the site condition, unbalanced lateral loading on the wall was expected and an excavation down
to -12.7m below the existing ground level was carried out with instrumentation, consisting of (8) inclinometer tubes installed
in the wall panels, settlement plates around excavation zone and tiltmeters and beam sensors on the existing structures. This
paper presents computer model analysis and performance of the wall including results of instrumentation. Behaviour and
performance of the wall is compared with those of other projects in Bangkok area.
INTRODUCTION
In Bangkok, the growing land price and need for space has
necessitated deeper and larger basement excavations, even in
some unfavorable subsoil and site conditions and in limited
spaces. Subsoil conditions in Bangkok is generally a very
soft clay of 12m to 18m thick layer underlain by stiff to hard
clay and series of sand layers. Excavation in such soft soil
requires efficient retaining structures and cast-in-situ
diaphragm walls have therefore come in use frequently.
This paper presents performance of a bracing excavation
with diaphragm wall adjacent to the river and surrounding
structures, including a historical building. This historical
building, having archeological and cultural values, not only
limited the height of the building but also influenced the
construction time. Since the location of planned building is
in a current limited height zone in which up to 4 storey
height is permitted, 3 level basement was included to
increase the usable floor area.
SITE CONDITION AND SUBSOIL
The building site is located nearby the Chao Phaya River
and surrounded by a historical building and other existing
structures supported by pile foundation (Fig. 1). Accurate
information on foundation of these existing buildings was
not available. The diaphragm wall along the river (D1) was
constructed about only 4.0m away from the existing old
river wall. The river bed near the river wall is about 2.2m-

Fig. 1 Layout of project site and instrumentation

3.0m in depth, sloping towards the mid-stream to a depth of
about 10m to 12m. River water level is about 1.6m below
the ground level during dry season and sometimes in the
rainy season is above the ground level, causing inundation.
A primary site investigation was carried out by drilling two
boreholes and two field vane shear tests. Prior to design for
bracing and basement excavation, drilling of additional two
boreholes and two field vane shear tests were performed to
check variability of subsoil conditions. The subsoil
655

247

0m

6m

10m

20m

SPT-N Value (blows/ft)

0
0

40 60

80 100 0.0

Su (t/m2)
2.0

4.0

Weathered
Clay

RIVER
5

10

20

Soft to
Very Soft
Clay

RIVER WALL

Med. Clay

15

FOUNDATION
PILES

0.80 THK.
D-WALL

0.80 THK.
D-WALL

20

Stiff to
Very Stiff
Clay

25

Dense to
Very
Dense
Sand

30

Fig. 2 Basement section and soil profile

properties obtained from the boreholes and test data were
summarized in Table 1.

For the wall sections adjacent to floor slab openings,

especially in the water storage area, additional
reinforcements were provided for bending stress in
horizontal direction.

Table 1. Summary of subsoil properties

Soil Type

Soft Clay
Medium Clay
Stiff Clay
Dense Sand

Layer
Top in
Depth
m

0.-3.0
12.7
14
25

?s

kN/m

35-78
30
22-34
14-25

16-19
19
19-21
20-23

cu

SPT

kPa

30
71
43-300
-

PILE FOUNDATIONS

14-52
>39

A total of 165 bored piles consisting of 0.8m, 1.0m, 1.2m

and 1.5m in diameter being founded in depth of 48m was
constructed to support the structure. Out of 165 piles, 5
numbers of 0.8m diameter piles and 8 numbers of 1.0m
diameter piles were incorporated with diaphragm wall panels
as a leg to carry the load transferred through the wall.

DIAPHRAGM WALL
A 800mm thick diaphragm wall was designed for excavation
down to 12.7m below the ground with two levels of
temporary bracing. The diaphragm wall toe was embedded
down to 28m to achieve the overall stability of the
excavation which is right on the river bank (Fig. 2).
Originally excavation stages were modeled by using a onedimensional finite element computer program (Nonlinear
Beam Column Analysis). Soil elements were modeled as
spring and wall elements were modeled as beam. Four types
of wall were designed to suit the temporary construction
stages and permanent conditions. The walls were reinforced
to withstand bending stresses up to 1000 kN.m/m in vertical
direction. The maximum movement of the wall was
expected to be 24.2mm.

INSTRUMENTATION
Due to the locality and particular conditions of the site,
more instrumentation than those used by other projects in
Bangkok were installed and monitored. Three types of
instrumentation - Eight inclinometer tubes were installed in
diaphragm wall panels to monitor displacement of wall, ten
settlement plates of 1.0m and 5.0m in depth around
excavation zone, and ten tiltmeters and four vertical beam
sensors on surrounding buildings were installed to monitor
tilting. Layout of instrumentation is presented in Fig. 1 .

656

248

BASEMENT EXCAVATION
The principal steps of basement excavation and construction
sequence are described below;
1. Construct capping beam and excavation to 2.5m below
ground level.
2. Install first level bracing at -1.5m and pre-load the struts
3. Excavate down to -7.0m
4. Install second level bracing at -6.5m and pre-load the
struts
5. Continue excavation down to the final depth at -12.7m
6. Construct mat foundation (Basement 3), Basement 2 and
remove second level bracing
7. Construct Basement 1 and remove the first level bracing
Since the excavation work is located adjacent to the river
and surrounded by old buildings, the diaphragm walls are
subject to three different lateral load conditions resulting
from (1) full depth of the earth, (2) steep downward slope of
riverbed, and (3) full depth of earth with possible surcharges
from the adjacent buildings. In particular, the walls
alongside the river (D1) and the opposite walls (D2), were
expected to undergo an unbalanced loading condition.
During temporary bracing design, two dimensional
computer modeling were carried out to study the effects of
unbalanced lateral loading on the wall and bracing. An
additional two dimensional model analysis was carried out
prior to designing a temporary bracing. The model analysis
indicated relatively less movement of Wall D1 towards
excavation (Fig. 3). However, the following measures were
adopted in excavation work to prevent potentially adverse
behavior of wall D1 and to keep the lateral wall movements
within tolerable limits;
1. Using a simple, but efficient temporary bracing system
2. Pre-loading of the struts (200kN/m and 400kN/m for first
and second levels respectively) on the one end of the
strut on Wall D2 only (Fig. 1).
3. Excavating the soil in front of Wall D1 side first at any
excavation stage
4. Frequent monitoring of wall movements
5. Minimizing construction time
During the initial excavation to 2.5m for installation of first
level bracing, a historic foundation was unexpectively
discovered. Excavation was suspended for about 3 months
and resumed after further excavation was permitted by the
archaeological department.
Excavation for final depth was made in the rainy season
and a berm made of sand bags for flood protection was
constructed around the perimeter of the wall. After lean
mixed blinding concrete had been cast at the final

Fig. 3 Deformed mesh of two-dimensional model

excavation level, the river water level rose to a maximum
level about 0.5m above the ground level.
TEMPORARY BRACING SYSTEM
A simple cross-lot bracing system with continuous wale
beams was used to support the wall during basement
construction stage (Fig. 1). 20m long steel king posts of
H300x300 sections were driven into stiff clay to support the
bracing system and temporary deck for excavation and
construction equipment. A summary of steel sections used
in bracing system is presented in Table 2.
Table 2. Summary of temporary bracing system
Bracing
Level
First
Second

Wale Beam

Strut

1xWF400x400 WF350x350
2xWf400x400 WF400x400

Expected
Force
(kN/m)
484.0
789.0

Between wale beams and wall, lean mixed concrete were

poured into the gaps between wale beams and wall to
achieve a good load transfer. Pre-loading was carried out
using hydraulic jacks. During pre-loading, the movements
of the strut were measured. Horizontal movements of
0.82mm to 9.06mm and 0.15mm to 29.83mm in the
direction of jacking for first level struts and second level
struts respectively were recorded. For second level struts,
the vertical movement of up to 1.68mm was measured at the
jacking position.
INSTRUMENTATION RESULTS
Inclinometer Monitoring - Wall movements measured from
inclinometers and predicted displacements by onedimensional finite element analysis are shown in Fig. 4.
Movements of Wall D1 and a section of Wall D3 close to
the river, near the garden are found to be cantilever shape
indicated by I-6, I-7 and I-8 respectively while movements
657

249

Displacement (mm)
0

20

Displacement (mm)
40

0.00

RIVER

within the range (about 0.1% to 1.2% of excavation depth)

of projects completed in Bangkok area.
(Max. Wall Movement.)/(Excavation Depth), %

of Walls D2 and D3 have inward bulging shape indicated by

I-1 to I-5.
40

20

-1.50

5.00

-6.50

10.00

Depth (m)

-12.70
15.00

PREDICTED
OBSERVATION

20.00

25.00

Inclinometer
No. I-7

30.00

3.0
Excavate to 7.0m

2.5

Excavate to 12.7m

2.0
0.9

1.5

1.0
1.1

1.0

1.4
0.64

2.0

0.5

0.42

3.0

0.24
0.14

Inclinometer
No. I-2

0.0
10

Displacement (mm)
0

10

20

0.00

30

40

30

20

10

Depth (m)

10000

Settlement Monitoring - Ground settlement of up to 16mm

was observed after completion of excavation works. The
maximum settlement was found to be near the location of
the inclinometer I-6 which indicates the largest lateral
movement at the top of the wall.

-6.50

10.00

-12.70
15.00

PREDICTED
OBSERVATION

20.00

25.00

30.00

1000

Fig. 5 Wall movement and excavation depth ratio

-1.50

5.00

100

System Stiffness, (EI)/(w/h4)

Displacement (mm)

40

Inclinometer
No. I-3

Inclinometer
No. I-5

Fig. 4 Predicted and measured lateral wall movements

Generally the predicted and measured lateral wall
movements are in a good agreement, except for the top
portion of wall in which the measured wall movements
exceed the predicted movements. The differences in wall
movement were found to be caused by the following;

Tiltmeter and Beam Sensor Monitoring - Readings of

tiltmeter and beam sensors installed in the surrounding
structures are shown with the progress of excavation in Fig.
6. A comparison between results from tiltmeter and beam
sensor and typical values for maximum building slope or
settlement for damage risk assessment is presented in Table
3.
Table 3.
Comparison between observed values and
maximum damage risk assessment (Lake et. al.))
Max. Slope
of
Building

1. Advancement of excavation in front of Wall D1 further

than other walls, allowing Wall D1 to stand longer prior
to strut installation than the others
2. Wall D1 was in cantilever condition for about 3 months
after initial excavation for first level bracing.
3. No direct nor immediate pre-loading on Wall D1.
4. Slight over excavation to install first level strut for all
Walls.
The walls are in fix end conditions as fixity of walls
indicated by all inclinometer reading are found to be at depth
of about 20.0m which is 8m above the walls toe depth.
Maximum lateral wall movements after installation of
struts were in the range of 0.14% to 0.64% of the
corresponding excavation depths (Fig. 5) and were also

Risk
Category 1

<1/500

Max.
Settlement
of building
(mm)
<10

Tiltmeter

1/2082*

Vertical
Beam Sensor
Risk
Category 2

1/3165*

1/500

10 - 50

Description
Negligible : superficial
damage unlikely

Slight: possible
superficial damage
which is unlikely to
have structural
significant

* monitoring results for the project presented in this paper

658

250

No building settlements nor cracks on the buildings were

observed till completion of the basement construction. The
buildings were evaluated to be in risk category 1. The risk
assessment and monitoring results confirm a good
performance of the braced excavation.
Time (days)

Depth (m)

50

100

150

200

250

Delay in strut installation and construction time affect the

wall movements, especially at initial excavation stage before
installation of first level bracing.

Install 2nd strut

Excav. near the river

Excav. away from the river

15

T (mm)

0.10
T

0.05

-0.05

50

100

150

0.50

200

Wall D1 which is alongside the river was found to be

minimally affected by the unbalanced load conditions due to
deep embedment (about 8m below the fixity) of the wall and
presence of the old river wall. This was also shown by the
two dimensional analysis (Fig. 3).

TILTMETER

250

EL-1 (5th Floor of B3)

EL-2 (5th Floor of B3)
EL-3 (7th Floor of B2)
EL-4 (2nd Floor of B1)
EL-5 (2nd Floor of B1)

EL

EL (mm)

T-1 (B3)
T-2 (B3)
T-3 (B3)
T-4 (B3)
T-5 (B3)
T-6 (B2)
T-7 (B1)
T-8 (B1)
T-9 (B1)
T-10 (B1)

0.00

0.25

The predicted movements of the wall in cantilever condition

at initial excavation stage in soft clay layer is considerably
small due to the high modulus of soft clay adopted. In this
case, construction time and sequence of excavation needed
to be strictly controlled to keep the wall within the
movement limit predicted by the one dimensional analysis.

VERTICAL
BEAM
SENSOR

0.00
-0.25

Max. Horiz. Movement (mm)

-0.50

50

100

150

200

70

250
I-1
I-2
I-3
I-4
I-5
I-6
I-7
I-8

60
50
40
30

Two dimensional analysis with facilities to model the

geometry of site is required for the diaphragm wall subject
to unbalanced lateral loading conditions.

INCLINOMETER

20
10
0

Pre-loading is an effective way to reduce further wall

movement after strut installation by providing a good intact
between wall and supporting system. This is clearly seen in
measured wall movements compared between I-7 and I-2,
and I-3 and I-5 which are located in the walls generally
facing each other (Fig. 4 and 1)

Install 1st strut

10

-0.10

DISCUSSION

50

100

150

200

250

Fig. 6
Tilting of adjacent buildings and horizontal
movement of diaphragm wall with construction time
PLANE STRAIN BACK ANALYSIS
A two-dimensional analysis was carried out by using finite
element program named PLAXIS (Veermeer and
Brinkgreve, 1995). A trial and error procedure in which the
values of undrained Youngs modulus (Eu) were adjusted
iteratively was adopted to obtain the best fit between
calculated and measured lateral wall movements. The
excavation stages were modeled by using plane strain
analysis. Mohr-Coulomb constitutive model was used for
clay and sand. Walls and strut were modeled as plate
element and elastic-perfectly plastic spring element
respectively. Pre-loading in struts are also included in the
analysis.
Results of analysis show that modulus value of soft clay is
about 500Su and for stiff clay modulus value is about
2,000Su. For one-dimensional analysis, the spring stiffness
used is equivalent to Eu=800Su for both soft clay and stiff
clay.

251

CONCLUSION
Braced excavation using diaphragm wall subject to
unbalanced loading due to adjacent river and surrounding
buildings, was successfully achieved with proper
instrumentation and monitoring.
Performance of the wall based on the instrumentation results
are presented and discussed.
Back analysis was carried out with two dimensional
modeling to determine the soil modulus. The soil modulus
obtained from back analysis and the modulus adopted in
wall designed were compared.
ACKNOWLEDGEMENT
The authors express their appreciation to the colleagues,
especially to Mr. Ganeshan Baskaran and Mr. Muhammad
Ashfaq Anwar for their invaluable suggestion and assistance
in the preparation of this paper. Initial analysis works
carried out by Dr. Vichai Vitayasupakorn and Mr.Young
Zou are acknowledged.
659

REFERENCES
Lake L. M., Rankin W. J. and Hawley J. (1996) Prediction
and Effects of Ground Movements Caused by Tunnelling
in Soft Ground beneath Urban Area CIRIA Report No.
30, Construction Industry Research and Information
Association, UK.
Vermeer, P. A. and Brinkgreve, R. B. .J. (1995). Finite
Element Code for Soil and Rock Analyses, Balkema,
Rotterdam.
Mana I. A. and Clough G. W. (1981) Prediction of
Movements for Braced Cuts in Clay, Journal of
Geotechnical Engineering Division, ASCE, Vol. 107,
No. GT6, pp. 36-42.
Gan C. H. (1997). Review and Analysis of Ground
Movements of Braced Excavation in Bangkok Subsoil
using Diaphragm Walls, M. Eng Thesis, Asian Institute
of Technology , Bangkok, Thailand.
Tanseng, P. (1997). Instrumented Deep Excavations in
Bangkok Subsoil , M. Eng Thesis, Asian Institute of
Technology, Bangkok, Thailand.

660

252

Fourth International Conference on

Case Histories in Geotechnical

Engineering
March 8 12, 1998, St. Louis, Missouri, USA

Editor : Shamsher Prakash

Prediction and performances of short embedded cast

in-situ diaphragm wall for deep excavation in Bangkok
subsoil
Thasnanipan N. , Maung A. W. and Wei S. H.
SEAFCO Co., Ltd., Bangkok, Thailand

Teparaksa W.

Chulalongkhon University, Bangkok, Thailand

University of Missouri-Rolla
Rolla, Missouri

253

686

Proceedings: Fourth International Conference on Case Histories in Geotechnical Engineering, St. Louis, Missouri,
March 9-12, 1998.

PREDICTION AND PERFORMANCES OF SHORT EMBEDDED CAST IN-SITU

DIAPHRAGM WALL FOR
DEEP EXCAVATION IN BANGKOK SUBSOIL
Dr. Wanchai Teparaksa
Civil Engineering Dept.
Chulalongkorn University
Bangkok Thailand

Narong Thasnanipan
SEAFCO CO. Ltd.
Bangkok 10320
Thailand

Aung Win Maung

SEAFCO Co., Ltd.
Bangkok 10320
Thailand

Wei Shixin
SEAFCO Co., Ltd.
Bangkok 10320
Thailand

Paper No. 5.04

ABSTRACT
This paper presents a case history of performances and predictions of a diaphragm wall which is subject to serve differently from its
initial purpose for which the wall was constructed. The updated requirements associated with the additional excavation created a
situation whereby the wall would become short embedded and behave like free-end supported wall. As the wall had been already
constructed during the time of modification it was only possible to modify the excavation method and bracing system. Accordingly,
the wall stability and possible toe movement were analyzed, to accommodate the updated requirements, by using finite element
methods. Continuous inclinometer monitoring has been carried out during the construction and results are being compared and
analyzed for predicted values. Performance of the wall based on comparison between the inclinometer monitoring results during
different excavation stages and predicted results, are discussed. It has been found that excavation depth for first bracing layer and
construction period are very important for diaphragm wall performance. Large initial movements in the wall strongly influence the
wall movements in the successive excavation works. Construction practice plays a major role in deep excavation work.
KEY WORDS
Diaphragm wall, Lateral movements, Inclinometer, Settlements, Free-end, Bracing, Excavation, Critical height, Cantilever

255

687

INTRODUCTION
Economic growth in Thailand during the past decade has been
forcing the construction industry to look for all feasible
methods to enhance the infrastructure of the country
especially for the Bangkok city. Huge number of high-rise
buildings with multilevel basements are being erected within a
short time possible to accommodate the booming economy.
Value of urban land is increasing in many folds beyond the
limits. Unlike in developed countries, the building rules and
regulations of this city are only in documented form and
hardly established all in practice in a short time. As a result,
developers and owners enjoy relatively more freedom to
modify the structural layout, even during and after
construction stages, to facilitate their growing demands which
are being conceptualized within a short time duration in the
fast developing city, aiming to maximize the area utilization.
The project reported in this paper is an example for such case.

Fig. 1 Soil Profile at the Site

The soil properties used for Diaphragm wall design are
summarised in the table below

The project site is located in the vicinity of central business

area of Bangkok. The structure is a multi-storied building for
a business complex. The load of the structure would be
carried by bored cast-in-situ bored piles and barrette piles with
toe depth of 60m below the ground level, embedded into the
second sand layer of the Bangkok subsoil.

Table 1 Unit weight and Undrained shear strength of soil

layers.
Unit Weight

(t/m3)
1.6

Undrained Shear
Strength
Su (t/m2)
1-2.9

Medium Clay CH

1.93

3.5

Stiff Clay CH

2.0

13.7-17.95

Soil Type

The constructed cast in-situ diaphragm wall is 0.8m thick and

the toe is embedded into stiff clay at a depth 18m below the
ground level. Initially the wall was designed for construction
of four basements including mat foundation of 3.5m thick.
The maximum excavation depth was set to be 14m and a three
level of temporary bracing system was considered to construct
the basements with conventional bottom up construction
method.

Soft Clay CH

SITE CONDITION

Embedded depth below the final excavation for the mat

foundation was designed to be 4m. After construction of the
wall it was decided to increase the number of basements to be
five for part of the building area (Tower area) and excavation
depth needed to be increased by 1.8m from the originally
designed depth, leaving the wall embedment depth of 2.2m
only.

The project site is located in the corner of a main road and a

street of the central business district in Bangkok. The planned
building is facing both the main road and street. The building
plan was set out in accordance with building code having a
setback of 6m from an adjacent property boundaries. Behind
it are two story buildings located more than 10m away. Deep
excavation is ideal in such site condition. Figure 2 shows the
layout plan of the building.

SUBSOIL CONDITION
DIAPHRAGM WALL DESIGN

The sub soil conditions at the project site is not uncommon to

usual conditions observed in Bangkok comprising a 15m thick
soft to medium clay layer on the top underlain by stiff clay to
a depth of about 22m. Below the stiff clay is a series of
alternating layers of hard silty clay, sandy clay and dense
sand.

Originally diaphragm wall was designed with a 3 level

temporary bracing to allow a maximum excavation depth of
14m below the existing ground for construction of mat
foundation (see Fig. 3). Computer programs of finite element
methods (WALLAP and CRISP) and finite difference method
(FLAC) were used to simulate the staged excavation and
predict the wall behaviour and performance. In the computer
modeling, undrained strength of soil with total stress
parameters derived from laboratory and field test results were

256

688

used. 0.8m thick Diaphragm wall with toe depth of 18m was
found to be adequate enough for the planned basement
construction in such soil condition.

Fig. 3 The Original Construction Sequence with three

temporary bracing levels
MODIFIED CONSTRUCTION SEQUENCE
Fig. 2 Layout Plan of Diaphragm Wall & Piles

Finite element computer modeling was again carried out for

the desired excavation depth. The model analysis showed that
4 level temporary bracing was necessary with strut preloading. It also indicated the diaphragm wall in free-end
condition with some degrees of toe movement in a range of
23mm. The maximum predicted wall movement at the top was
77mm.

Estimated maximum bending moment of 90 t-m/m on the

excavation side and 56 t-m/m on soil side of the wall. The
maximum deflection of the wall predicted was about 3537mm with some degrees of toe movement up to 11mm.
Bored piles of 1.20m in diameter were in-cooperated within
and below the toe of the wall to support the column load and
dead load of the wall. Foundation piles (dia. 1.20m),
including bored piles and barrette piles (2.70mx0.80m) were
founded at the depth of 60m in dense sand layer. Four
inclinometer tubes I-1 to I-4 (20m to 25m in length) were
installed in the wall panels with bored pile legs to observe the
behaviour and performance of the wall.

To prevent progressive toe movement with time, followings

were recommended for excavation works.
(i) frequent monitoring of diaphragm wall movements when
excavation close to the final excavation depth.
(ii) casting a reinforced lean concrete slab with a thickness of
20cm and 6m wide in front of the wall just after reaching the
excavation depth of -13.8m.

Diaphragm wall and foundation piles were constructed in

parallel. After completion of diaphragm wall construction,
but construction of bored piles were still in progress, overall
design of the building was revised to suit the requirements of
the property market. Number of basement floors was
increased from 4 to 5 in the main tower area. As a result, the
diaphragm wall constructed was deemed to support for the
maximum excavation depth of 15.8m.

(iii) excavation to -15.8m must be carried out leaving a 12m

wide soil berm with lean mixed concrete slab on top of the
berm all around the wall.
Apart from four inclinometer tubes installed in the wall, a
total of 16 sets of settlement observation stations were
installed around the wall perimeter. Each station comprised
two to three settlement points located at 3m intervals and 3m
away from the wall.

Firstly diaphragm wall design and construction stages were

then reviewed.
Since the wall has been constructed,
excavation sequence and bracing system can only be modified
so that the modified deeper excavation can be done without
impairing the structure and stability of the wall.

257

689

in the soil of active side about 2-3m away from the wall.
After installation of struts and pre-loading them, the lateral
movements of the wall ceased off. An increment of 2mm4mm was recorded with further excavation stages (See Fig. 5).
Struts from first to fourth levels were pre-loaded 40 t/m, 75
t/m, 75 t/m and 40 t/m respectively, 20%-25% of design load
of bracing system, prior to further excavation. The preloading of the strut was very effective to reduce lateral
movement of the wall and ensuring good intact between wall
and bracing system.

Fig. 4 The Modified Construction Sequence

The risk of over excavation was carefully assessed not only
for diaphragm wall but also for the existing buildings near by
the site. Review of the site condition indicated that the
existing buildings surrounding the project site were more than
10m away from the excavation zone and they were found
subject to no risk.
Fig. 5 Maximum lateral movements at the top of wall vs.
Construction Time.

BRACING SYSTEM
The bracing system included continuous waling beams along
the diaphragm wall and longitudinal and transverse struts.
The spacing of strut was generally 6.4m to 6.8m. Table no. 2
shows the type of wide flange structural steels used for the
bracing.
Table 2 Structural steel sections used for temporary bracing
Bracing
Layer
1st

2W300x300x94kg/m

2-W300x300x94kg/m

2nd

2W350x350x137kg/m

2W350x350x137kg/m

rd

2W350x350x137kg/m

2W350x350x137kg/m

W350x350x137kg/m

2W350x350x137kg/m

4th

Strut

Wale

LATERAL MOVEMENTS OF THE WALL

Monitoring of wall movements was carried out by reading the
inclinometer tubes profiles installed in the wall at every major
construction stages. The wall constructed was designed for
initial excavation of -2.5m to install the first bracing layer.
However the excavation has been done to about -3.0m to 3.2m at once and a movement of 40mm was observed. Time
delay during construction of capping beam and over
excavation deeper than the critical height (Hcr = 2C/) was
unsupported around three months resulted an increase in wall
movements to 55mm -70mm. Tension cracks were developed

Fig. 6A.
Lateral Movements of the wall shown by
Inclinometer Tubes I-1 and I-2.
The pattern of movements of the wall shown by the
inclinometer readings is similar to that of cantilever condition.
Because of large movements were induced Initially, braced
mode was not developed. After reaching the final excavation
depth the observed movements of the wall at the toe level

258

690

were in a range of 1.2mm to 16mm while the predicted toe

movement was about 23mm. Predicted and observed wall
movements is tabulated as below;
Table 3 Predicted and observed wall movements
Location
on the
Wall

Predicted
for 15.8m
Excavation
(mm)
77

Observed
on Site
(mm)

Top

Predicted
for 14.0m
Excavation
(mm)
35-37

Toe

11

23

1.2-16

55-98

The maximum lateral movements were found in I-4. At the

final excavation stage, movements of the wall increased
considerably below the first bracing level down to toe of the
wall.
It was noted that in I-4 area the longitudinal and
transverse struts of the bracing system become diagonal to the
wall sections due to the geometry of wall layout. The stiffness
of the bracing in this particular zone is considered to be much
less than that of others.

Fig. 7 Relationship between maximum ground settlements

and maximum lateral wall movements

The walls fixity was found to be at depth 20m-22.5m from

inclinometer readings when the final excavation reached.

Based on the deformation of the wall measured from the

inclinometer readings, the bending moments were computed
from the well known equation to check structural performance
of the wall

ANALYSIS ON PERFORMANCE OF THE WALL

M=Ec.Ief.[dS/dy]
where

Ec
Ief
S
y

(1)

= Modulus of Elasticity of concrete

= Effective moment of inertia of the wall section
= slope given by inclinometer reading
= elevation of the slope given

Generally the bending moments computed were within the

design moments and the performance of the wall is found to
be satisfactory. However a direct comparison between
bending moment envelopes from computer simulation and
bending moments derived from inclinometer readings shows
some differences, especially at the strut levels and below the
depth of 12m. These differences can be due to the following;
1. Stiffness of medium to stiff clay layer used based on the
field and laboratory tests for simulation is much lower than
the actual in-situ soil stiffness. In simulation for the wall
design, influence on soil stiffness below the final excavation
level by bored piles was not considered.
2. Efficiency and stiffness of bracing system, including preloading forces. In the original design calculation, much lower
pre-loading forces were used.

Fig. 6B Lateral movements of the wall shown by Inclinometer

Tubes I-3 and I-4.
From the settlement observation data, it was found that the
maximum ground settlements behind the wall were
approximately 34% to 50% of the lateral wall movements,
except for the area near the I-4 where the ground settlements
equal the lateral wall movements.

3. large initial movements during the wall in cantilever

condition.
4. Inclinometer tubes (20-25m in length) were installed in
diaphragm wall panels with bored pile legs. The bored pile

259

691

The relationship between wall movement and system stiffness

(Fig. 9) shows that because of large movements during the
wall is in cantilever condition, the excavation to 5.5m with
one bracing deviates from the normal trend of braced
excavation.

legs provide the fixity of the wall and increase the bending
moments at a location near the bottom of the wall.
5. Moment redistribution in the wall is considered not fully
activated.

Fig. 10 Relationship between Normalized Maximum lateral

movement versus factor of safety against basal heave.
The normalized lateral movements of the wall shown in Fig.
10 is within the range of 0.5-1.5% which is slightly higher
than those of similar projects in Bangkok subsoil. The higher
movement ratio occurred during excavation in the soft clay
layer with slight over excavation and delay in first strut
installation. It is observed that the first excavation depth and
duration in bracing layer installation are very critical for the
performance of the wall at further excavation stage.

Fig 8 Design envelope of bending moments from computer

modeling and bending moments computed from inclinometer
readings at final excavation

CONCLUSION
1. Movements in first excavation stage is governing the
performance of Diaphragm wall for further excavation.
2. First excavation deeper than the critical height of the soil
needs to be supported immediately.
3.
Apart from support system and
soil properties,
construction practice influenced the movement of the wall in
this particular project.
4. A successful deep excavation with short embedded
diaphragm wall has been achieved using a combination of
prediction by computer simulations and instrumented field
observations.

Fig. 9
Relationship amongst system stiffness, nondimensionalized wall movement and factor of safety against
basal heave .

260

ACKNOWLEDGEMENT

692

The authors express their appreciation to the colleagues,

especially to Mr. Ganeshan Baskaran, Mr. Young Zou, Mr.
Muhammad Ashfaq Anwar, Mr. Pornpot Tanseng and Mr.
Chetsadar Plongkrathok for their assistance in the preparation
of this paper.
REFERENCES
Balasubramaniam A.S., Bergado D.T., Chai J.C. and Sutabutr
T. [1994]. Deformation Analysis of Deep Excavations in
Bangkok Subsoils XIII Internation Conference on Soil
Mechanic and Foundation Engineering , New Delhi, India. pp.
909-914.
Day P. W. [1994]. Factors Influencing the Movement of
Retaining Structures XIII Internation Conference on Soil
Mechanic and Foundation Engineering , New Delhi, India. pp.
109-114.
Gan C. H. [1997]. Review and Analysis of Ground
Movements of Braced Excavation in Bangkok Subsoil using
Diaphragm Walls. M. Eng Thesis, Asian Institute of
Technology. Bangkok, Thailand.
Mana I. A. and Clough G. W. [1981] Prediction of
Movements for Braced Cuts in Clay, Journal of Geotechnical
Engineering Division, ASCE, Vol. 107, No. GT6, pp. 36-42.
Peck R. B. [1969] Deep Excavation and Tunneling in Soft
Ground. Proceeding of 7th International Conference on Soil
Mechanics and Foundation Engineering, Mexico. State of Art
Volume, pp. 225-290.
Saxena S. K. [1975]. Measured performance of a rigid
concrete wall at the World Trade Centre. Diaphragm Walls
and Anchorages. Institution of Civil Engineers, London, UK.
pp. 107-112.

261

PROCEEDINGS OF THE 5TH INTERNATIONAL SYMPOSIUM ON FILED MEASUREMENTS

IN GEOMECHANICS FMGM99/SINGAPORE/ 1 3 DECEMBER/ 1999

Filed Measurements
In Geomechanics
Edited by

C.F.Leung, S.A.Tan & K.K.Phoon

Department of Civil Engineering, National University of Singapore

Lessons from the collapse during construction of an inlet

pumping station : Geotechnical instrumentation aspects
Teparaksa W.,

Chulalongkhon University, Bangkok, Thailand

Thasnanipan N., Maung A.W. and Tangseng P.

Seafco Company Limited Bangkok, Thailand

A.A. BALKEMA / ROTTERDAM / BROOKFIELD / 1999

263

Field Measurements in Geomechanics, Leung, Tan & Phoon (eds) 1999 Balkema, Rotterdam, ISBN 90 5809 066 3

Lessons from the collapse during construction of an inlet pumping station :

Geotechnical instrumentation aspect
W. Teparaksa
Department of Civil engineering, Chulalongkorn University,Bangkok, Thailand

N. Thasnanipan, A. W. Maung and P. Tanseng

Seafco Co., Ltd., Bangkok, Thailand

ABSTRACT: In the wastewater treatment plant, an Inlet Pumping Station (IPS) is one of the most important
structures to collect and pump all wastewater to the treatment plant. The IPS presented in this paper consists
of 20m deep underground chambers. A reinforced concrete diaphragm wall system 1.0m thick and 25m. deep
was used in construction of the IPS. Four Inclinometer tubes were installed in the diaphragm wall to monitor
the lateral wall movements during excavation work for basement construction. When excavation approached
the final depth, excessive lateral wall movements were observed, but they were not taken seriously. Later,
some cracks on the capping beam and noise from the bracing system were noticed. Then the IPS collapsed
and was buried under sliding soil mass. At the time IPS collapsed five main bracing layers and only two of
the five intermediate bracing layers had been installed. This paper emphasizes the importance of instrumentation. Additionally it presents the results of investigation into the causes of the collapse, as well as the reconstruction work.
1. INTRODUCTION
nal
Ca
ng
isti
Ex

Reconstructed
Canal

Inlet Pumping
Station (IPS)
Storm Pumping
Station (SPS)

Chao Phraya River

Bangkok metropolis is sectioned into four parts for

the implimentation of four mega projects, or
"phases" that comprise a wastewater treatment system to serve the rapidly growing industrial and domestic needs. Three of four phases are under construction. All projects are a turnkey system where
the main contractors are responsible for both design
and construction. The engineers will be the consultants as representatives of the owner to review the
conceptual and detailed design and to supervise the
construction. The aim of this paper is to highlight
the importance of instrumentation in deep underground construction, as proved by the collapse of
the Inlet Pumping Station (IPS) at the second phase.
This paper also briefly describes its reconstruction.

Canal Wall
Pump Sta
tion

Treatment Plant
(STP)

Figure 1. Layout of Inlet Pumping Station

The IPS has two distinct chambers, a pumping

chamber and an inlet/bypass chamber, all built underground. The pumping chamber is a circular shaft
while the inlet/bypass chamber forms a box on the
north side of the pumping chamber. The pumping
chamber is 20.3m in diameter and 20.2m deep. It
consists of a central shaft 8.75m. in diameter and an
annulus 4.875m. wide. The inlet chamber is essentially a large box receiving the inlet sewers and also
housing screens for removal of debris. The inlet
chamber is 11.5m x 9.0m approximately in plan.
The bypass chamber is adjacent to an inlet chamber
and divided by a concrete wall from base level at

2. PROJECT DESCRIPTIONS
The second phase Bangkok wastewater project presented herein was started in 1995 to serve the central area of Bangkok City. The project was planned
to be completed by 1999 (ie. within 4 years). In the
sewage treatment plant (STP), the IPS is one of the
most important structures to collect and pump all
wastewater to the sewage treatment plant. Figure 1
presents the location of IPS on the sewage treatment
plant (STP) site.
247

265

19.5m below ground to roof level. The inlet/bypass

chamber has three levels. The base level is at
19.50m below the ground. Levels 1 and 2 are at
10.5m and 5.5m depth respectively. The roof slab is
at 1.0m depth.
The IPS was located at the northwest corner of
the STP site, approximately four to ten meters from
a canal. The canal is protected by concrete walls
supported by battered concrete piles. Private residences consisting of wooden house were located
close to the property line in the west.
Space limitation necessitated reinforced concrete
diaphragm walls 1.0m thick and about 25.0m deep
as permanent structures of the IPS. The diaphragm
walls were also used as temporary structures during
excavations up to 22.0m deep. The one-meter-thick
wall was cast in panels with lengths of around 2.5 3.0m using the bentonite slurry trench construction
method. The connection between adjoining panels
was constructed by using U-shaped stop-ends with
water bars.
For excavation work, five levels of temporary
steel bracing were designed to support the straight
wall portion and an additional five intermediate
bracing layers was planned to frame the edges
where the straight wall meets the circular enclosure
(Fig. 2). Four inclinometer tubes were installed at
critical points of the diaphragm wall to observe lateral movements during excavation for underground
construction.
The bracing system was classified as temporary
and thus no submission was made for review. The
final temporary bracing arrangement (Fig. 3) was
changed from the initial design in Figure 2. In the
initial design, two lacing beams shown as a dashe
line in Figure 2 were planned to strengthen the strutting system. However these were cancelled in the
final temporary bracing system (Fig. 3).

Legend

-2.20

Main strut +Diagonal Strut

-3.95

Intermediate Strut
-5.70

I-2

-7.95

-10.20
-11.95

I-1
Diagonal
Struts
Main Strut +
Intermediate Strut

I-4

I-3

-13.70

-15.45
-17.20
-19.10
-21.0

B+
A+

I-3

SECTION A-A

-21.0

SECTION B-B

Figure 2. Temporary bracing system (Initial design)

Figure 3. Final Temporary bracing system (after KanokNukulchai et al., 1998).

Table 1. Summary of soil properties
Depth
from
0.0
3.0
16.0
20.0
>30

3. SOIL CONDITIONS
At the start of the project, the site consisted of very
soft clay (mud) about 2 - 3m deep. The soft clay
apparently extended westward, into the residential
area of the wooden houses. When work on the project commenced, mud was removed from the site up
to the property boundaries. Approximately a 3.0m
thick of sand backfill was placed on the site. Then
the soil conditions consisted of 3.0m thick sand
backfill, followed by soft to medium dark grey clay
approximately 16.0m deep. Then stiff to very stiff
silty clay was encountered at about 30.0m depth,
overlaying the first dense silty sand layer. The sensitivity of the soft to medium clay was in the order
of 2.5 to 6.0. A piezometric level was found at
about 21.0m below ground level, as a result of deep
well pumping. A summary of general soil properties and soil stiffness is presented in Table 1.

m
to
3.0
16.0
20.0
30.0

Soil
Description
Fill
Soft Clay
Stiff Silty Clay
Very Stiff Silty Clay
Dense Silty Sand

t
kN/m3
18.0
16.0
19.0
20.0
20.0

Su
kN/m2
12-30
100-150
175-250
-

4. COLLAPSE OF INLET PUMPING STATION

The reinforced concrete diaphragm wall and installation of four inclinometers were completed on 15th
January 1997 by the diaphragm wall sub-contractor.
After completion of the diaphragm wall, the same
sub-contractor pointed out to the main contractor the
importance of an adequate bracing system and
staged construction sequences. The excavation and
strutting was carried out by another subcontractor,
while the other permanent structures such as capping ring beam were constructed by the main contractor. The capping ring beam had been casted to
two thirds of its height by the main contractor on 8
June 1997. Two inclinometer tubes (I-1 and I-2)
were found damaged and this was reported. But the
248

266

damaged tubes were left unrepaired and could not

be used for monitoring.
The excavation began in mid-June, 1997. Backhoes were used for excavating inside the shaft and
all materials were collected with a bucket. The
bucket was hung from a service crane located behind the diaphragm wall perimeter near I-1 location
(collapsed zone) on the ground. The excavation and
installation of struts was carried out step by step.
However, only 5 diagonal bracing layers, 5 main
struts and two out of five intermediate struts were
installed without instrumentation.
Lateral wall movements were recorded by I-3,
located at one edge of circular enclosure and I-4, located in the intact zone (Fig.2). Wall movements in
A-axis direction of the inclinometer access tube I-3
were indicative of the wall in cantilever/unsupported
condition while those in B-axis direction were indicative of the wall being supported (Fig. 4a).
Displacement (mm)
0

50

Displacement (mm)
-100
0.00

100

5.00

5.00

10.00

10.00
Depth (m)

Depth (m)

-50
0.00

15.00

-50

50

15.00

20.00

20.00
20-Jun-97
4-Jul-97
12-Jul-97
18-Jul-97
25-Jul-97
5-Aug-97
13-Aug-97

25.00

30.00

20-Jun-97
4-Jul-97
12-Jul-97
18-Jul-97
25-Jul-97
5-Aug-97
13-Aug-97

25.00

30.00

A-Axis

B-Axis

ments indicated that the edges of the wall between

the straight wall and the circular enclosure had been
under torsional stress. It was reported that tension
cracks on the capping beam were observed in late
July 1997. I-4 recorded wall movements with a
maximum of 13mm (Fig. 4b). Movements of the
wall toe could not be measured due to lack of the
inclinometer tube penetration into the soil below the
wall toe level.
After observation of tension cracks on the capping beam, immediate installation of the remaining
intermediate struts to support the edges of the circular enclosure and straight wall area were recommended. Occurence of more cracks on the capping
ring beam and noise from the bracing system was
noticed. At this stage excavation depth inside the
circular enclosure had been reached the final depth.
However, only two of the five intermediate bracing
layers had been installed at the time the IPS collapsed. Figure. 5 shows the photo of the bracing
system taken one day before collapse. The IPS collapsed in the morning on 17 August 1997.
After the collapse, soil mass caved in. Several adjacent wooden houses were also damaged and slipped
down (Fig. 6a). The soil mass buried the collapsed
portions of the wall, the steel bracing, and the excavation equipment (Fig. 6b). Some sections of the
wall on the western side collapsed while the remaining wall sections stood in place. The collapse also
sheared off the capping beam at the collapse location, causing severe cracking in the capping beam
section, and opening panel joints on remaining sections of the wall. No human casualties were reported in the incident as the loud noise prior to
the

Figure 4a. Inclinometer reading of I-3

Displacement (mm)
-10

Displacement (mm)
10

-20

0.00

0.00

5.00

5.00

10.00

10.00
Depth (m)

Depth (m)

-20

15.00

20.00

10

20.00

4-Jul-97
12-Jul-97

25.00

5-Aug-97

27-Jun-97
4-Jul-97
12-Jul-97
25-Jul-97

25-Jul-97

30.00

15.00

27-Jun-97

25.00

-10

30.00

A-Axis

5-Aug-97

B-Axis

Figure 4b. Inclinometer reading of I-4

However, the wall movements in A-axis direction were considerably large and thus the main
beams supporting the edges of the circular enclosure
seemed not effective. The two differential move249

267

Figure 5. Bracing system seen on 16, Aug. 1997 (after Thasnanipan, 1997)

Figure 6a. After collapse soil mass moved in (after Thasnanipan, 1997)

Figure 7. Underground construction of similar shape to IPS

(after Katzenbach et al. 1998)

posed walls. Phien-Wej & Sriruanthong (1999) reported that at the final excavation depth in IPS reconstruction, the toe of the diaphragm wall in the
collapsed zone was found in original position and
there was no evidence of basal heave.
The importance of temporary bracing has been
emphasized in various design guides, manuals and
reports by Padfied & Mair (1984), William & Waite
(1993) and Feld & Carper (1996). In the case of
improper temporary bracing, this issue was identified as the main cause of collapse. Results of the
analysis by means of FEM (Kanok-Nukulchai, et al.
1998) concluded that the main strut supporting the
edges of the circular enclosure was grossly overstressed. The report also stated that these struts
would have progressively reached their capacity at
the initiation of the collapse and may have become
totally ineffective. It was clearly indicated by the
photographs taken just one day before the collapse
of the IPS (Fig. 8).
Figure 8 shows separations of the built-up walling beams and struts, which would reduce the capacity of the monolithic section by 30%. The three
lower intermediate struts required at the edges of the
circular enclosure were not present eventhough the
excavation works reached the final depth (Fig. 5).
This improper bracing also contributed to the cracks
observed on the capping beam. Additionally, lateral
movements of the walls, shown by inclinometer

Soft Clay

Stiff Clay

Figure 6b. Soil mass buried everything inside the pit. (after
Phien-Wej et al. 1999)

collapse provided warning to the people working inside.

5. CAUSE OF COLLAPSE
The cause of collapse has been investigated by various parties involved. The preliminary causes of the
collapse were considered to be:
Improper shape of IPS permanent structure
Failure of diaphragm wall
Failure of internal temporary bracing during
excavation
Failure due to improper sequence of excavation
Failure due to poor workmanship in temporary
works
Three main causes of collapse are identified; the
improper shape of the IPS, failure of the diaphragm
wall, and improper temporary bracing.
In the case of improper shape of the IPS, similar
underground structure had been completed in Frankfurt, Germany (Katzenbach et al. 1998) as shown in
Figure 7. Pit excavation was about 20.0m deep and
a similar temporary strutting system was also used.
Evidence also indicated that collapse of the IPS
was not caused by the diaphragm wall. It was found
that the results of analysis of the wall designed by
both the designer and diaphragm wall sub-contractor
were in conformity with completed projects in
Bangkok. The completed projects have been reported by Teparaksa et al. (1998) and Thasnanipan
(1998). No defect or leakage was found on the ex-

Figure 8. Separating of the built-up walling beams and struts

in layer 3 (after Thasnanipan, 1997)
250

268

251
Displacement (mm)
-50

50

-100

100

0.00

0.00

5.00

5.00

10.00

10.00

Displacement (mm)
-50

50

Legend
Grouting
Zone

I-3

Depth (m)

Depth (m)

Inclinometer
Tube

15.00

I-6
I-2

15.00

Existing Diaphragm Wall

1.00m thk.

I-4

20.00

20.00

Outer Diaphragm Wall

1.50m thk.
25.00

I-1

25.00
FEM Analysis

Excavate 16.5 m

Excavate 16.5 m

Excavate 19.0 m

Excavate 19.0 m
30.00

30.00

A-Axis

I-5

B-Axis

Figure 9. Prediction vs. actual wall behavior (I-3)

monitoring from I-3 indicated that the walls were in

cantilever mode. The wall behavior did not reflect
predicted behavior indicating inefficiency of the
bracing system (Fig. 9). The cracks occurred in the
capping beams confirmed cantilever wall movements.
6. REMEDIAL ACTION / RECONSTRUCTION

Figure 10. Location of inclinometers and treated area by Jet

grouting (after EIT 1999)
Legend
Inclinometer Tube
VW Strain Gauge

I-3

I-6

Reconstruction of the IPS was started at the end of

1997. The process involved construction of an outer
diaphragm wall perimeter to enclose the collapsed
IPS location. The IPS could not be relocated due to
presence of affiliated structures. Before starting the
outer loop diaphragm wall, Jet Grouting techniques
were used to stabilize the soil in the surrounding
area for purposes of diaphragm wall construction as
well as for excavation works (Fig. 10). Jet grouting
and diaphragm wall construction was undertaken by
another subcontractor. Jet grouting stabilized only
soft clay up to about 16.0m deep. The outer loop
diaphragm wall which was designed as a temporary
structure was 1.5m thick and about 32.0 m. deep.
During reconstruction, all parties realized the importance of instrumentation. Six inclinometers were
installed, five numbers in the wall panels, and one in
the jet grouted area in the collapsed zone (Fig. 10).
During excavation, two bracing layers with heavily built-up steel sections were used as diagonal
struts at depths of 2.0m and 11.5m (Fig. 11). In
both layers, strain gauges were also installed on two
main struts and readings were recorded each time a
depth of every 2.0m. had been excavated or every 2
days.
Installation and monitoring of (VWSG) on struts
provided information on their performance during
construction so that necessary action against instability could be taken.

269

Strain Gauge
Location

I-2
Strain Gauge
Location

I-4

I-1

I-5

Figure 11. Arrangement of strutting system with stain gauges

(after EIT 1999)

Figure 12. Excavation in progress for IPS reconstruction

Displacement (mm)

-50

-25

25

50

Feld, J. and Carper K.L (1996) Construction failure (2nd edition) John Wiley & Sons. Inc.
Kanok-Nukulchai, W. and Phien-Wej, N. (1998). Investigation
on Collapse of an Inlet Pumping Station- Looking into
modeling, Design and Construction Considerations, ACECOMS News and Views, July - September, AIT, Thailand
Katzenbach, R., Moormann, C., and Quick H (1998), A new
concept for the excavation of deep building pits in inner
urban areas combining top/down method and piled-raft
foundation, 7th Int. Conf. and Exhibition on Pile and Deep
Foundation, Vienna, Austria.
Padfied, C. J. and Mair, R. J. (1984) Design of Retaining walls
Embedded in stiff clay CIRIA Report 104.
Phien-Wej, N. (1991) Types of Design and limitations of various retaining cofferdam system, Technical seminar in
Foundation and underground construction works, organized by EIT, Feb. (in Thai) Bangkok, Thailand
Phien-Wej, N. and Sriruanthong, M. (1999) Remedial Construction of Inlet Pump Station, wastewater Project, Yannawa, Civil Engineering Magazine, EIT. Jan- March. (in
Thai.)
STS (1995) Preliminary factual report of Geotechnical Investigation for Bangkok Wastewater Treatment on Sewage
Treatment Plant, Bangkok, Thailand
Teparaksa, W., Thasnanipan, N., Muang, A. W. and Wei, S.
H. (1998) Prediction and Performances of Short embedded
cast in-situ diaphragm wall for deep excavation in Bangkok subsoil. 4th Intl. Conf. On Case Histories in Geotechnical Engineering St. Louis, Missouri, USA.
Thasnanipan, N. (1997) The Cause of failure of the cofferdam
under soil excavation and bracing installation stage at an
Inlet Pumping Station, Sewage treatment plant, SEAFCO,
Internal Report (in Thai).
Thasnanipan, N., Muang, A.W, Tanseng, P. and Wei, S.H.
(1998) Performance of A Brace Excavation in Bangkok
clay, Diaphragm wall subject to unbalanced loading
conditions. 13th Southeast Asian Geotechnical Conference.
Taipei, Taiwan, ROC.
William, B. P. and Waite, D. (1993) The Design and
Construction of Sheet Piled Cofferdams CIRIA special
publication 95, Thomas Telford.

Displacement (mm)

75 0

25

50

75

100

0.00

-2.00

5.00

Depth (m)

10.00

-11.50

15.00

20.00

-21.20

25.00

30.00

35.00

29-Mar-98
13-Apr-98

13-Apr-98

17-May-98

17-May-98

30-May-98

30-May-98

21-Jun-98

21-Jun-98

2-Aug-98

2-Aug-98

13-Sep-98

13-Sep-98

Inclinometer No. I-6

-32.00

Inclinometer No. I-2

Figure 13. Deflections recorded by I-2 & I-6

Figure 12 shows the status of excavation and strutting before casting the new IPS structures inside the
temporary diaphragm wall. The lateral wall movements were recorded by I-2 and I-6 with maximum
wall deflections of about 50.0mm 70.0 mm respectively (Fig. 13). The reconstruction work was thus
successfully completed with proper instrumentation.
7. CONCLUSIONS & RECOMMENDATIONS
It was learnt that ignorance of instrumentation for
deep underground construction can cause a serious
disaster.
Adequate instrumentation should be provided for
deep excavation to avoid remedial work.
Monitoring and maintenance of instrumentation
should be carried out throughout the work.
Parties involved in deep underground construction should have enough knowledge in geotechical
works and instrumentation.
Proper instrumentation enabled the successful
reconstruction of IPS.
8. ACKNOWLEDGEMENT
The authors wish to express their appreciation to the
colleagues, especially to Mr. Thiruchelvam
Navaneethan for his invaluable advice and assistance in preparation of this paper.
9. REFERENCES
EIT (1999), Case study of problems related to settlement and
failures Geotechnical Engineering in the Past 2 years,
Subcommittee of Geotechnical Engineering, Engineering
Institute of Thailand (EIT)

252

270

PROCEEDINGS OF THE 5TH INTERNATIONAL SYMPOSIUM ON FILED MEASUREMENTS

IN GEOMECHANICS FMGM99/SINGAPORE/ 1 3 DECEMBER/ 1999

Filed Measurements
In Geomechanics
Edited by

C.F.Leung, S.A.Tan & K.K.Phoon

Department of Civil Engineering, National University of Singapore

Behavior and performance of diaphragm walls under

unbalanced lateral loading along the Chao Phraya River
Thasnanipan N., Maung A.W. and Tangseng P.
Seafco Company Limited Bangkok, Thailand

Teparaksa W.,

Chulalongkhon University, Bangkok, Thailand

A.A. BALKEMA / ROTTERDAM / BROOKFIELD / 1999

271

Field Measurements in Geomechanics, Leung, Tan & Phoon (eds) 1999 Balkema, Rotterdam, ISBN 90 5809 066 3

Behavior and Performance of Diaphragm Walls under Unbalanced

Lateral Loading along the Chao Phraya River
N. Thasnanipan, A. W. Maung and P. Tanseng
Seafco Co., Ltd., Bangkok, Thailand

W. Teparaksa

Department of Civil Engineering, Chulalongkorn University, Bangkok, Thailand

ABSTRACT: This paper presents the behavior and performance of diaphragm walls constructed for basements of two buildings along the bank of the Chao Phraya River bank, Bangkok, Thailand. These buildings
were constructed at different periods under one contract. A case history on basement excavation of the first
building has been reported previously by Thasnanipan et al. (1998). For the second building, more numbers
of instrumentation including VWSGs in a diaphragm wall panel and pressure gauges in struts to monitor
stress in the wall and the performance of bracing system were used respectively. The construction sequence
adopted for excavation, performance of the walls and bracing based on instrumentation results are discussed.
1 INTRODUCTION

2 SITE AND SUBSOIL CONDITIONS

The building sites are located 40.0m apart, and separated by an existing 5 storey building between them.
They are located along the river and surrounded by
existing structures including a historical building
(Fig. 1). The existing buildings were supported by
pile foundation. The distance between diaphragm
walls along the river and the existing old river wall
267

273

Chao Phraya River

First Building

Excav. Depth = 12.70m

g
uildin
rey B
5 Sto

Retaining structures and supported systems for deep

excavation in urban areas, particularly in the vicinity
of a river, are often subjected to unbalanced lateral
load due to asymmetric loading and surcharges. The
various publications and design guides by GCO no.
1/90 (1996), Williams & Waite (1993) and Padfield
& Mair (1991) give special guidance on overall stability of retaining structures and strutting systems in
such conditions. Case histories on braced excavations under unbalanced/non-symmetrical lateral
loading also have been reported by Thasnanipan et
al. (1998), De Rezende Lopes (1985) and Kotoda et
al. (1990).
This paper presents two basement excavation
works up to 12.7m below the ground level for two
separate buildings using braced diaphragm walls
along the bank of the Chao Phraya River, Bangkok.
The first one had previously been reported by Thasnanipan et al. (1998). Construction of the second
building commenced about one year after completion of the first one.

buildings respectively. They are also situated in the

heart of an old established culturally-significant
zone. Space and building height limitations necessitate multi-level basements to increase usable floor
areas although the planned buildings are right on the
river bank.

Second Building

5 Storey
Building

Excav. Depth = 9.70m

5 Storey
Building

5 Storey
Building

Figure 1. Layout of building sites

At the building sites, the river is about 205m wide

and 10-12m deep mid-stream. The riverbed near the
river wall is about 2.2-3.0m in depth with a gentle
slope. The river water level in dry season is about
1.6m below ground level and in the rainy season
sometimes rises above ground level, causing a flood.
Primary site investigations included one borehole
60m deep and one field vane shear test for each site.
For the first building, prior to designing temporary
bracing and basement excavation work, drilling of
additional three boreholes 20m deep and two field
vane shear tests were carried out to check subsoil

variations. For the second building, two additional

60.0m deep boreholes were dug. The shear strength
of soft clay in the planned second building site is
higher than the soft clay in the first site. However,
there is no significant variation in subsoil conditions
within each site, particularly along the depth of
planned diaphragm walls. Subsequently the profile
of the sites and surcharges from surrounding buildings are of major concern for unbalanced lateral
loading conditions in excavation works. Table 1
shows a summary of subsoil properties obtained
from the boreholes and test data.
Table 1. Summary of soil properties
Layer
W
s
Cu
SPT
Soil
Top in
3
Type
Depth m %
kN/m kPa
N
Soft Clay
0-3.0
35-78 16-19 30
Med. Clay
12.7
30
19
71
Stiff Clay
14
22-34 19-21 43-300 14-52
Dense Sand
25
14-25 20-23 35-50
Silty Clay
36.5
17-21 20-23 175-240 30-45
Dense Sand*
42-45
20-26 21
>58
* Some 4-5m thick hard clay seams present at depth 48-52m

5 INSTRUMENTATION

+14.80

3 5

10

+11.30
+7.80
Vane Shear Strength (kPa)

+4.30

-3.20
-5.80
-8.40

-28.00 (D-Wall tip)

-48.00 (Pile Tip)

+0.70

Bored
Pile

Foundation piles having pile toe at 48m depth were

constructed using bored piling under bentonite
slurry. Quantities of piles are tabulated in Table 2.

Pile Dia.
Building
800mm
1000mm 1200mm 1500mm
First
32*
18*
31
6
Second
17*
33
5
23
* Some piles were incorporated in diaphragm wall panels.

For both buildings, 800mm thick cast in-situ concrete diaphragm walls of 28m toe depth with two
level temporary bracings were designed for basement excavation. The maximum excavation depths
for the first and second buildings were 12.7m and
9.7m respectively. 28m deep walls were necessary
for overall stability of excavation as they were located on the riverbank.

0.80m Thick
D-Wall

4 PILE FOUNDATIONS

Table 2. Quantity of foundation piles

3 DIAPHRAGM WALLS

+0.80

nite element program predicted maximum wall

movements up to 24.2mm for the first building.
Based on the performance of walls for the first
building, a maximum wall movement of 51.0mm
was allowed during excavation for the second building. The diaphragm walls for its basements were
analyzed with elasto-plastic soil model using
WALLAP program.

Weathered Clay

Soft to
Medium
CLAY

20

40

60

80

SPT-N (Blows/ft)
0

50

100

0.00
5.00

As construction sites were located in a very sensitive

urban area and subjected to unbalanced lateral loading conditions, various types of instrumentation
were installed and systematically monitored. Layouts of instrumentation for the first and second
buildings are shown in Figures 3a and 3b respectively.
For the second building, 5 levels of vibrating wire
strain gauges (VWSG) in pairs were installed in one
diaphragm wall panel and 2 sets of earth pressure
gauges were installed in struts to observe stress in
the diaphragm wall and strut forces respectively.
Types of instrumentation are presented in Table 3.

10.00

Table 3. Quantity of Instrumentation

15.00

Stiff to Very
Stiff CLAY 20.00
25.00

Medium to
Very Dense 30.00
SAND

First Bldg.
Second Bldg.

35.00

Figure 2. Basement section and soil profile (Second Building)

Different wall types were classified by design according to load conditions, slab supporting locations
and geometry of ground profiles. The walls were
designed generally as vertical beams to withstand
bending moments up to 1100 kN-m/m. In the slab
opening areas, additional horizontal bars were provided for secondary bending moments. A non-linear
Beam-Column Analysis using a one-dimensional fi-

Type of
Instrumentation
Inclinometer

First
Building
8

Second
Building
6

Measurements
made
Wall deflections

Settlement Plate

10

20

Tiltmeter

10

10

Vertical Beam
Sensor

Ground settlements
Tilting of Buildings
Tilting of buildings

VWSG

1 location*

Strain in rebars

Earth Pressure
2x2
Sturt forces
Gauge
* A pair of VWSG in 5 layers along the depth of wall panel

268
274

Tiltmeter

I-7

Vertical Beam Sensor

Surface & Deep Settlement Plates

1. Using a simple and efficient temporary bracing

system
2. Pre-loading on one end of struts on the opposite
walls
3. Excavating first soil in front of the wall closest to
the river at any excavation stage
4. Frequently monitoring of wall movements
5. Minimizing construction time

Chao Phray
a River

Inclinometer

D1

I-6

Bored Piles
Preload Position

I-8

Historical
Building
(2 Storey)

I-5

I-2

D2

B1

I-3

D3

I-1

D2
0

10

5 Storey Building

D3

D3

B3

I-4

B2

15

7 Storey Building

Scale

Figure 3a. Layout of temporary bracing and instrumentation

(First Building)
Chao Phraya River
D4

S4

I-1

D10

S10

D9

S9

S3
D3

I-2
T1

EL5

lding
ey Bui
5 Stor

S1

I-3

I-6

T9

S8
D8

EL1
T2

Figure 4. Behavior of walls shown by defromed mesh of twodimensional model

T10

D1

S2
D2

5 Storey
Building

T8
EP-1

D7

EL4

T3
S5
T4

I-5

S6

5 Storey Building

D5

D6

I-4
T5

S7

VWSG-1
EL2

T6

Legend
I = Inclinometer in D-Wall
T = Tilt plate

D = Deep settlement plate (5.0m)

EP = Earth pressure gauge

EL = Vertical beam sensor

VWSG = Vibrating wire strain gauge

S = Surface settlement plate (1.0m)

Pre-load position

T7

EL3

5 Storey Building

10

20

Figure 3b. Layout of temporary bracing and instrumentation

(Second Building)

6 EXCAVATION WORK
Conventional bottom-up method with two levels of
temporary bracing was adopted for excavation to
construct foundation and basement floors of both
buildings. Prior to excavation work for the first
building, a two-dimensional computer analysis was
carried out using PLAXIS finite element computer
program to study possible behavior of the wall system under unbalanced lateral loading conditions.
Such conditions were considered to be resulting
from (1) full depth of the earth, (2) a step sloping
river bed and (3) full depth of earth with possible
surcharges from the adjacent buildings.
A major concern was that the diaphragm walls
alongside the river would be thrust from the opposite
walls. These bore a higher lateral load through axial
force of struts as excavation progressed in stages.
This wall behavior was indicated by computer modeling (Fig. 4), and previous reports by De Rezende
Lopes (1985) and Kotoda et al. (1990). To prevent
any adverse wall behavior alongside the river, the
following measures were taken;

275

269

Moreover, an observational construction approach

(Ikuta et al. 1994), using the "most probable" conditions and parameters in the design with a contingency plan for "most unfavorable" conditions was
employed. Firstly a monitoring plan was established
before excavation. Secondly the instrumentation
monitoring data were used to trigger the contigency
plan. Generally predicted wall movements were set
as primary trigger values for the contingency plan.
After successful completion of the first building, the
same approach was also adopted in excavation work
for the second building. For the first level bracing
of the second building, single strut arrangement (less
rigid but economical compared to that of the first
building) was adopted with provision of strut-force
monitoring.
A simple cross-lot bracing system with continuous
wale beams was used for both buildings. 20m and
19m-long, steel king posts of H300 x 300 in section
were used to support the working platform and bracing system for the first and second buildings respectively.
In both cases, the initial excavation to 2.5m for installing the first level bracing revealed historic foundations. Temporary bracing and excavation works
Table 4. Summary of steel sections for the temporary bracing
Building Bracing Bracing
Strut
Design Strut
Level Elevation Sections
Force kN/m
First
I
-2.0
2 x WF350 x 350
484.0
II
-6.5
2 x WF400 x 400
789.0
Second
I
-2.0
1 x WF400 x 400
279.4
II
-7.0
2 x WF350 x 350
332.1

were delayed for about 3 and 2 months for the first

and second buildings pending permission by the archaeological department for further excavation.

Time (Days)
0.0

100

200

Depth (m)

2.0

300
B1

Excavation Depth

4.0

B2

6.0
B3

8.0

7 INSTRUMENTATION RESULTS

400

10.0
Max. horiz. movement
(mm)

I-4

60
I-2

40
20
0
3000

Design Line

2500
Strut Load (kN)

Construction activities and the corresponding results

of instrumentation monitoring with construction
time are presented in Figures 5a and 5b. These figures, particularly for the second building indicate a
good relationship between construction activities on
one hand and responses of the walls, existing buildings and ground on the other hand. It can be observed that significant changes in these responses
generally occured during the initial excavation stage
in which walls were unsupported.

80

2000

First Strut

1500
1000
500

Second Strut

0
1200

50

100

200

250

Excav. near the river

Excav. aw ay f rom the river

10

T-1 (B3)
T-2 (B3)
T-3 (B3)
T-4 (B3)
T-5 (B3)
T-6 (B2)
T-7 (B1)
T-8 (B1)
T-9 (B1)
T-10 (B1)

Tension = positive

Vertical Beam Sensor

EL-1 (5th Floor of B3)
EL-2 (5th Floor of B3)
EL-3 (7th Floor of B2)
EL-4 (2nd Floor of B1)
EL-5 (2nd Floor of B1)

0.00

-0.25

-0.50
100
80
60
40

800
400

EF (4.5m)
EF (7.0m)
EF (9.5m)
EF (13.0m)
EF (17.0m)

0
-400

Settlement (mm)

Displacement
(mm)

0.25

Excavation Face

1200

-300
0.50

50% of ultimate tensile strain

SF (4.5m)
SF (7.0m)
SF (9.5m)
SF (13.0m)
SF (17.0m)

Strain (micro strain)

-100

Soil Face

-400

Inclinometer
I-1
I-2
I-3
I-4
I-5
I-6
I-7
I-8

20

Surface settlement plates

S6 (Land side)

-20

S9 (River side)

S7 (Land side)
S10 (River side)

-40
-60
20

20
0

Figure 5a. Instrumentation results with construction time (First

Building)

During the period of delay, monitoring of

instrumentation was frequently carried out to
compare the monitoring data and the trigger values
for planning a contingency plan against possible
risks to adjacent structures from unsupported
excavations to 2.5m depth.
7.1 Inclinometer monitoring

Settlement (mm)

50% of ultimate tensile strain

400

Tilt Meter
100

-200

Max. Horiz
. Movement
(mm)

800

Differential in Tilting
(Second)

Differential in Tilting
(second
)

15
200

Deep settlement plates

D5 (Land side)
D6 (Land side)
D9 (River side)

-20

D10 (River side)

-40
-60
200
100

Tilt plates
T5
T6

0
-100
-200
10

Displacement (mm)

Depth(m)

150

Tension = positive

0
0

Strain (micro strain)

Time (days)

Vertical beam sensor

EL2
EL3

0
-5
-10

For the first building, generally the predicted and

measured lateral wall movements were in good
agreement, except for those at the top portion of
walls (Fig. 6a). The wall-top movements exceeded
predicted movements in the case of both buildings
due to delay in installing the first level bracing, leaving the walls in cantilever condition for about 2
months or more.

276

270

Figure 5b. Instrumentation results with construction time

(Second Building)

For the second building, wall movements were

monitored weekly, sometimes every 2-3 days when
necessary. At I-6 location, wall movements were
found to reach the trigger/predicted value, especially

at the top of wall when further excavation reached

7.5m depth after installation of the first level bracing.

Displacement (mm)
0.00

RIVER

20

Displacement (mm)
40

40

20

-1.50

5.00

-6.50

10.00

Depth (m)

-12.70
15.00

PREDICTED
OBSERVATION

20.00

7.2 Pressure gauges

Readings from the pressure gauges installed on
struts of the second building were recorded daily.
After removal of second level bracing upon completion of base slab and during construction of basement 2 floor, measured strut force of the first level
bracing reached the trigger value. However, in general the readings indicated that the bracing system
used was adequate.

25.00

Inclinometer
No. I-7

30.00

Inclinometer
No. I-2

Figure 6a. Lateral wall movements (First Building)

Displacement (mm)
0.00

20

40

60

Displacement (mm)

80

80

60

40

20

-2.00
River
5.00

-7.00

Day 133
Day 155

Day 67
Day 133

-9.70

10.00

Depth (m)

Table 5. Predicted, allowed and measured strut forces (Second

Building).
Item
First level
Second level
kN/m
kN/m
Predicted by WALLAP 354.5
307.3
Allowed
279.4
332.1
Measured
331.3
206.2

Day 67

Day 221

15.00

Day 155
Day 221

Prediction

7.3 Vibrating Wire Strain Gauges

20.00

Day

25.00

30.00

I-4 suggested that the wall alongside the river had

been pushed against the retaining soil by the opposite wall (Fig. 6b).
All inclinometer readings for both buildings indicated that the walls were in fixed-end condition with
fixity at depths of about 20.0m. Maximum wall
movements after installation of bracing ranged from
0.14% to 0.64% of the corresponding excavation
depth for the first building, and 0.27% to 0.83% for
the second building. These were also within the
range (about 0.1% to 1.2% of excavation depth) of
wall movements in other projects completed in
Bangkok area.

67
133
155
221

Inclinometer

No. I-2

Strains developed in the wall of the second building

were measured by using VWSGs attached to main
reinforcement bars of the wall panel where I-4 was
also located. A comparison between the measured
strains and the allowed strains of the wall was made
as shown in Figure 5. The comparison indicated that
strains induced in the wall were well below the allowed values.

Description
Excavate to -2.5m
Excavate to -7.5m
Excavate to -9.70m
Remove strut at -7.00m

Inclinometer

No. I-4

Figure 6b. Lateral wall movements (Second Building)

Lateral wall movement reached a maximum rate of

6.5mm/day. Then a tension crack on the ground surface appeared about 8.0m away from the wall. A
close examination of the bracing system indicated
that one strut had swayed slightly about 130mm at a
distance of 15.0m from the wall. Moreover, the sections of the strut had been connected at an angle. An
additional strut was then immediately stacked on the
defective one and then wall movements were found
to cease. Installation of the second level bracing
was immediately carried out for this area.
Regarding unbalanced lateral loading condition,
the wall movements of the first building were found
unaffected by this condition (Fig. 6a). However, for
the second building, monitoring results from I-2 and

277

271

7.4 Settlement Plates

For the first building, a maximum settlement of
16mm was recorded. For the second building, in
most cases, ground settlements (surface and deep
settlements at 1.0m and 5.0m depth) had been found
to stop after installation of the first bracing. In some
cases, the surface settlements stopped only after installation of the second level bracing, resulting in
maximum settlements up to 30-50mm. The large
settlements were found to be associated with surcharges from stockpiles of construction materials
and in one case, from unsupported concrete steps,
under demolition, which had become detached from
the adjacent buildings.

9 CONCLUSION

7.5 Tiltmeter and Beam sensor

Readings are also shown in Figure 5. Results from
monitoring and typical values for maximum building
slope or settlement for damage risk assessment are
compared in Table 6.

Table 6. Comparison between observed values and maximum

damage risk assessment (Lake et. al.)
Max. Slope
of
Building
Risk
Category 1
Tiltmeter
Vertical
Beam Sensor
Risk
Category 2

<1/500
1/2082*
1/2008**
1/3165*
1/985**
1/500

Max.
Settlement
of building
mm
<10
10 - 50

Description
Negligible: superficial damage unlikely

Slight: possible superficial

damage which is unlikely to
have structural significant

* monitoring results for the first building,

** for the second building presented in this paper

With proper instrumentation, two basement excavations using braced diaphragm walls subjected to unbalanced lateral loading from the river and adjacent
buildings were completed.
Performance of the walls based on the instrumentation results are presented and discussed.
For underground construction work, monitoring
with instrumentation plays a major role, especially
when delay occurs during excavation. Moreover,
overall construcion cost, time and risk can be minimised with the use of instrumentation.
10 ACKNOWLEDGEMENT
The authors express their appreciation to Mr. Thiruchelvam Navaneethan for his assistance in the
preparation of this paper. We would also like to acknowledge initial analysis work carried out by Dr.
Vichai Vitayasupakorn and Mr. Young Zou.
REFERENCES

8 DISCUSSION
The planned excavation sequence and deep embedment (about 18m and 8m below the excavation and
the fixity respectively) of the walls are considered to
have contributed to minimizing the effect of unbalanced load conditions, particularly for the walls
alongside the river.
Instrumentation results, particularly from the
monitoring of wall movements have given effective
warning against inproper installation of bracing system and excavation sequences.
This allowed
improvement or modification of the bracing system
where necessary.
No damage was found in adjacent buildings, as
confirmed by instrumentation results, particularly
those from vertical beam sensors and tiltmeters.
In comparing bracing systems for the first and second building works, stiffness of the first level bracing used for the second building is estimated to be
only 50% of that for the first building while expected strut force is 25% less. Moreover single strut
arrangement in the first level bracing for the second
building is less rigid compared to dual strut arrangement which can act as a composite beam for
the first building. Instrumentation on struts confirmed that bracing system used was adequate.
In both cases, early installation of bracing, particularly for the first level is very important for
minimizing wall top movements, subsequently reducing accumulated wall movements for further excavation.

278

272

De Rezende Lopes, F. (1985). The problem of asymmetricsurcharges in strutted excavations. Ground Engineering, vol.
18, no.2, pp 31-35. GCO Publication no.1/90 (1996). Review of design methods for excavations Geotechnical Engineering Office, Civil Engineering Department, Hong
Kong.
GCO Publication no.1/90 (1996). Review of design methods
for excavations. Geotechnical Engineering Office, Civil
Engineering Department, Hong Kong.
Ikuta, Y., Maruoka M., Aoki M. and Sato E. (1994). Application of the observational method to a deep basement excavated using the top-down method. Geotechnique 44, No. 4,
pp 655-664.
Kotoda, K., Kazama, S., Ogawa, H. and Kato, K. (1990). Behavior of earth retaining structure for excavation subjected
to non-symmetrical lateral loading. Tenth Southeast Asian
Geotechnical Conference, 16-20 April 1990, Taipei, Taiwan. pp. 539-544.
Lake, L. M., Rankin, W. J. and Hawley, J. (1996). Prediction
and effects of ground movements caused by tunnelling in
soft ground beneath urban area. CIRIA Report No. 30,
Construction Industry Research and Information Association, UK.
Padfield, C. J. and Mair, R. J. (1991). Design of retaining
walls embedded in stiff clay. CIRIA Report 106. Construction Industry research and Information Association. London, UK.
Peck, R. B (1969). Advantages and limitations of the observational method in applied soil mechanics. Geotechnique 19,
No. 2, pp. 171-187.
Thasnanipan, N., Maung A. W., Tanseng P. and Wei S. H.
(1998). Performance of a braced excavation in Bangkok
clay, diaphragm wall subject to unbalanced loading conditions. Thirteenth Southeast Asian Geotechnical Conference, 16-20 November, 1998, Taipei, Taiwan. Pp. 655-660.
Williams, B. P and Waite D. (1993). The design and construction of sheet-piled cofferdams. CIRIA Special Publication
98, Thomas Telford Publications.

Under the Royal Patronage of His Majesty King Bhumibol Adulyadej

Sixth Cycle Celebration of

His Majesty the King of Thailand and
th
40 Anniversary of the Asian Institute of Technology
Civil and Environmental Engineering Conference
New Frontiers and Challenges
8-12 November 1999
Bangkok, Thailand

Conference Proceedings
Volume 2 (Part I)

Editors:
D. G. Lin
D. T. Bergado
N. Phien-wej
P. Nutalaya
A. S. Balasubramaniam

Analysis of lateral wall movement for deep excavation in

Bangkok subsoils
Teparaksa W.
Chulalongkhon University, Bangkok, Thailand
N. Thasnanipan & Pornpot Tanseng
Seafco Company Limited Bangkok, Thailand

279

Civil and Environmental Engineering Conference

New Frontiers and Challenges, 8-12 November 1999, Bangkok, Thailand

ANALYSIS OF LATERAL WALL MOVEMENT FOR DEEP BRACED

EXCAVATION IN BANGKOK SUBSOILS
Dr. Wanchai Teparaksa
Department of Civil Engineering, Chulalongkorn University
Bangkok 10330, Thailand
Narong Thasnanipan & Pornpot Tanseng
SEAFCO Co.Ltd., Bangkok, Thailand
ABSTRACT
This paper presents the case histories and back analysis of diaphragm walls for deepbased excavations in Bangkok subsoils. Three case studies with different boundary conditions
are detailed as under unbalanced loading, underpass project under very narrow clearance under
the first stage expressway, and short embedded wall for modified deeper excavation. The
back-figured values for the soil stiffness in terms of Youngs modulus are calculated as Eu/Su
= 500 and 2000 for the soft Bangkok clay and stiff clay, respectively. The results have later on
been confirmed with the first self-boring pressuremeter tests in Bangkok subsoils.
INTRODUCTION
Deep excavations in Bangkok City are still necessary even the country has faced sever
economic crisis. There are still required the necessary civil works to serve the demands and
also to solve the environmental problems such as the first underground subway project, and
the wastewater treatment schemes. Subsoils condition in Bangkok is generally consists of a
very soft dark gray of about 12-14 m thick layer underlain by stiff and very stiff clay to about
23-25 m depth then followed by the first dense silty sand layer. Excavation in such soft dark
gray clay requires efficient retaining structures such as temporary sheet pile for shallow
excavation, and cast in-situ reinforced concrete (RC) diaphragm wall as permanent structures
for deep excavation. There are a lot of diaphragm wall activities under construction and under
excavation stages for the first blue line subway project, Metropolitan Rapid Transit (MRT).
This subway project consists of 18 underground stations, about 20 km long twin tunnels
running from North of Bangkok at Bangsue station and turn around the business area, bus
terminal and ended at the main state railway station namely Hua Lumpong station (Teparaksa,
1999). Excavation depth for the underground stations is between 22-33 m.
This paper aims to evaluate the soil stiffness parameters of the soft and stiff Bangkok
clay for RC diaphragm wall based on back analysis of three case studies with different
construction boundaries.
SOIL PARAMETERS FOR DIAPHRAGM WALL DESIGN
The RC diaphragm wall can be designed by various methods such as Finite Difference
Method, Beam on Elastic Foundation Method, and also the Finite Element Method (FEM).
The most well known and commonly used method is the two dimensional (2D) FEM method
or plain strain method. The 3D FEM analysis is rarely used. The FEM analytical method can
simulate the soil-structure interaction, and can predict the lateral wall movement as well as the

II ~ 67

281

ground surface settlement. The appropriate soil modelings as well as the input soil parameters
are the key issues to simulate the system. The most important soil parameter is the stiffness of
each soil layer, which is normally specified in terms of Youngs modulus (E). The E value is
very important especially for the soft clay layer that lead to induce the lateral wall movement
and ground settlement during excavation works. Normally, E is assumed to be linear and
determined from the conventional triaxial compression test and is very low as compared to the
values obtained from the empirical correlation. This is because of the high strain levels at
which the E is measured in the conventional triaxial test. The E value is recently known to be
dependent on the order of shear strain level. Mair (1993) compared the soil stiffness at various
shear strain levels for different structural systems as well as various kinds of laboratory tests as
shown in Figure 1. In case of concrete diaphragm walls, the shear strain is of the order of
0.01% to 0.10% only.
The research works on the soil behaviors at the low strain level in soft Bangkok clay
was presented by Shibuya et al (1997) based on laboratory tests as well the in-situ field vane
shear tests. Authors presented the soil stiffness in terms of Maximum Shear Modulus (Gmax).
Gmax is about 300*Su (300 times of Undrained Shear Strength) to 500*Su and if converted to
the Undrained Youngs modulus (Eu) will be equal to 900*Su to 1500*Su for Poisson ratio ()
of 0.5. Teramast (1998) also presented the soil stiffness at low strain levels based on the
Bender Element Test as Gmax = 400*Su to 570*Su or Eu = 1200*Su to 1710*Su. As shown
in Figure 1, Mair (1993) mentioned that the Bender Element Test is suitable for very low
strain levels such as applied to the dynamics or earthquake problems. Most practical value of
Eu for Bangkok subsoils is based on the empirical correlation proposed by Duncan &
Buchigani (1976) as shown in Figure 2. The Eu values were presented relative to the
undrained shear strength, which is dependent on the order of Over Consolidation Ratio (OCR).
For Bangkok soft clay where OCR is about 1.6, Eu/Su = 300-600.

Typical strain Ranges:

Stiffness G

Retaing Walls
Foundations
Tunnels

0.0001
0.001
Bender

0.01

0.1

10

Shear Strain s: %

Resonant Coulomn
Local

Special Triaxial
Conventional

Figure 1 : Typical Strain Range (Mair, 1993)

Figure 2: Empirical Youngs modulus

Values, Duncan & Buchigani, 1976.

BACK ANALYSIS CASE STUDIES

As mentioned above that the soil stiffness varies and depends on the shear modulus.
However, the values from laboratory tests give a high variation with the empirical correlation.
Therefore, the back analysis from case histories of diaphragm wall construction is really

282

II ~ 68

necessary to verify the appropriate soil stiffness. The back analysis was carried out for three
case studies at the Thammasat University project, Dindang underpass project and Sathorn
Complex project. Using the 2D-FEM program namely PLAXIS (Brinkgreve & Brand,
1996) back analysis for the said projects have been carried out. The analytical plain strain
concept and the Mohr-Coulomb soil model or bi-linear modeling was used.
Thamasart University Project
The building site is located nearby the Chao Phaya River and surrounded by a
historical building seated on shallow raft foundation as shown in Figure 3. The diaphragm
wall along the river was constructed about 3-4 m away from the existing old and shallow river
wall as shown in the building section in Figure 4. The river bed near the river wall is about 3
m in depth and sloping towards the mid-stream to a depth of about 10-12 m (Thasananipan et
al, 1998). This riverbank lead to induce the unbalanced lateral earth pressure on the
diaphragm wall system between land side and riverside. A 800 mm thick diaphragm wall was
designed for excavation down to 12.7 m below the ground with two levels of temporary
bracing at 1.5 m and 6.5 m depth. The diaphragm wall toe was embedded down to 28 m to
achieve the overall stability of the excavation which is right on the river bank (Fig. 4). A site
investigation was carried out by two boreholes and four field vane shear tests, the subsoil
properties are summarized in Table 1.
Table 1. Summary of subsoil properties
Soil Type
Crust
Soft Clay
Medium Clay
Siff Clay
Dense Sand

Depth
(m)
0.0-2.5
2.5-12.7
12.7-14.0
14.0-25.0
>25.0

Wn
(%)

50
70-85
50-55
20-25
-

Unit Weight, s
(kN/m3)
19
26
17
20
20

Cu
(kPa)

30
15-30
30-45
130-250
-

SPT-N
(Blows/ft)

14-40
>39

Instrumentation
Due to the locality and particular conditions of the site, more instrumentation than those used
by other projects in Bangkok were installed and monitored. Three types of instrumentation Eight inclinometer tubes are installed in diaphragm wall panels to monitor displacement of
wall, ten settlement plates of 1.0m and 5.0m in depth around excavation zone, and ten
tiltmeters and four vertical beam sensors on surrounding buildings were installed to monitor
tilting. Layout of instrumentation is presented in Fig. 3. Since the excavation work is located
adjacent to the river and surrounded by old buildings, the diaphragm walls are subject to the
unbalanced lateral load The model analysis indicated relatively less movement of Wall D1
towards excavation in the land side and move less in the river side as well as induce forward
movement of river wall to the river (Fig. 5). This analysis led the engineer to solve the
damages of the river wall due to unbalance loading by using the following construction
methods.

283

II ~ 69

Chao Praya Riv

er

I-7

D1

I-6

I-8

Tiltmeter

B3

Inclinometer
Surface & Deep Settlement Plates

D3

Vertical Beam Sensor

Preload Position

I-5

I-2

D2

B1

Historical Building
(2 Storey)

I-3

10

15

Scale

B2

D3

I-1

5 Storey Building

D3

Bored Piles

D2

I-4

7 Storey Building

Figure 3: Project Layout

0m

6m

10m

20m

0
RIVER

5
10

RIVER
WALL

15
20

0.80 THK.
D-WALL

FOUNDATION
PILES

0.80 THK.
D-WALL

25
30
35

Figure 4: Section perpendicular to the river

1. Arrange the strut as a simple straight grid line
2. Pre-loading of the struts (200kN/m and 400kN/m for first and second levels
respectively) on the one end of the strut on the land side wall.
3. Excavating the soil in front of Wall river side first at any excavation stage

Figure 5 : Deformed Mesh of Two-Dimensional model

284

II ~ 70

Analytical Results
The analysis was trial by changing the soil stiffness and compares the results with the
field performance. Figure 6 shows the results of FEM analysis compared with inclinometer
measurements based on soil stiffness of Eu/Su = 500 and 2000 for soft Bangkok clay and stiff
clay, respectively. . Movements of Wall D1 close to the river are found to be cantilever shape
indicated by I-7 while movements of Walls D2 and D3 have inward bulging shape indicated
by I-1 to I-5. Generally the predicted and measured lateral wall movements are in a good
agreement, except for the top portion of wall in which the measured wall movements exceed
the predicted movements. The differences in wall movement were found to be caused by the
following;

Advancement of excavation in front of Wall D1 further than other walls, allowing Wall D1
to stand longer prior to strut installation than the others.
Wall D1 was in cantilever condition for about 3 months after initial excavation for first
level bracing.
0

SPT (blow/ft)
20 40
60

Su FV (kN/m 2)

80 0

10

20

30

Excavate
7.00 m

Excavate
2.00 m

40

0.00

0.00

5.00

5.00

5.00

10.00

10.00

10.00

15.00

15.00

15.00

20.00

20.00

20.00

25.00

25.00

0.00

Excavate
12.70 m

Crust

Med.
Clay

Stiff to
Very
Stiff Clay

Depth (m)

Soft to
Very
Soft Clay

Field

25.00

PLAXIS

Dense to
Very Dense
Sand
30.00

10

20

30

Displacement (mm)

40

30.00

10

20

30

30.00
40

Displacement (mm)

10

20

30

40

Displacement (mm)

Figure 6 : Comparison between calculations and observation (Thammasat University Project)

Dindang Underpass Project

The Dindang underpass was the first Bangkok underpass project at Dindang intersection.
Underpass the main road beneath the main foundations of the expressway. This project is also
very interesting, specially designed equipment for D-wall construction was employed under
very narrow clearance under the first stage expressway (Thasananipan et al. 1996). There was
one temporary bracing layer at 2 m. depth and excavation to 4.9 m. depth. Figure 7 shows the
results of FEM analysis compared with inclinometer measurements based on soil stiffness of
Eu/Su = 500 and 2000 for soft Bangkok clay and stiff clay, respectively. The predicted and
measured lateral wall movements are in a good agreement.

285

II ~ 71

SPT (blow/ft)
0

20

40

60

20

40

60

Excavate
4.90 m

Excavate
2.7 m

Su FV (kPa)
80 0

80

0.00

0.00

5.00

5.00

10

10.00

10.00

15

15

15.00

15.00

Stiff
to Very Stiff
20
Clay

20

20.00

Crust

Soft to Very Soft

Clay
10

Medium Stiff Clay

20.00
Field
PLAXIS

25

25

25.00

25.00
-10

10

Displacement (mm)

20

-10

10

20

Displacement (mm)

Figure 7 : Comparison between prediction and observation (Dindang Underpass Project)

Sathorn Complex project

The project site is located in the vicinity of central business area of Bangkok. The
structure is a multi-storied building for a business complex. The load of the structure would
be carried by bored cast-in-situ bored piles and barrette piles with toe depth of 60m below the
ground level, embedded into the second sand layer of the Bangkok subsoil. The constructed
cast in-situ diaphragm wall is 0.8m thick and the toe is embedded into stiff clay at a depth 18m
below the ground level. Initially the wall was designed for construction of four basements
including mat foundation of 3.5m thick. The maximum excavation depth was set to be 14m
and a three level of temporary bracing system was considered to construct the basements with
conventional bottom up construction method. Embedded depth below the final excavation for
the mat foundation was designed to be 4m. Figure 8 shows the layout plan of the building
After construction of the wall it was decided to increase the number of basements to be
five for part of the building area (tower area) and excavation depth needed to be increased by
1.8m from the originally designed depth, leaving the wall embedment depth of 2.2m only.
The first author was invited to solve such a short embedded diaphragm wall (Teparaksa et al,
1998). The temporary bracing was changed from three layers to be four layers with using the
soil berm of 12 m. wide to prevent the excessive wall movement as shown in Figure 9.

286

II ~ 72

I-3
Excavation -15.5m
Bored Pile
Barrette
Inclinometer Tube

I-4
I-2
89.35

Excavation -14.0m

I-1

120.25

Figure 8: Layout of Project

+1.70

16

-1.50

15

-2.40

14

-4.20

13

-5.10

-6.90

11
-9.00

-11.80

-12.30

10

-9.60

9
8

-14.00

-13.30

-18.00

12

-15.50

SCHEMATIC SECTION OF BASEMENT CONSTRUCTION

-18.00

LEGEND
12

CONSTRUCTION STAGE
TEMPORARY BRACING

Figure 9: Schematic Section

Subsoil Conditions
The sub soil conditions at the project site is not uncommon to usual conditions observed in
Bangkok comprising a 15m thick soft to medium clay layer on the top underlain by stiff clay
to a depth of about 22m. Below the stiff clay is a series of alternating layers of hard silty clay,
sandy clay and dense sand. The soil properties used for Diaphragm wall design are
summarized in the table 2 belows:
Table 2 Soil Properties used for D-Wall Design
Soil Type

Unit Weight
s (kN/m3)

Soft Clay, CH
Medium Clay, CH
Stiff Clay, CH

16
18
20

287

II ~ 73

Undrained Shear
Strength, Su
(kPa)
10-29
35
137-180

Analytical Results
.
The wall constructed was designed for initial excavation of -2.5m to install the first
bracing layer. However the excavation has been done to about -3.0m to -3.2m at once and a
movement of 40mm was observed. Time delay during construction of capping beam and over
excavation deeper than the critical height (Hcr = 2C/) was unsupported around three months
resulted an increase in wall movements to 55mm -70mm. Tension cracks were developed in
the soil of active side about 2-3m away from the wall. After installation of struts and preloading them, the lateral movements of the wall ceased off. Struts from first to fourth levels
were pre-loaded 40 t/m, 75 t/m, 75 t/m and 40 t/m respectively, 20%-25% of design load of
bracing system, prior to further excavation. The pre-loading of the strut was very effective to
reduce lateral movement of the wall and ensuring good intact between wall and bracing
system. Figure 10 shows the comparison of the results of FEM analysis compared with
inclinometer measurements based on soil stiffness of Eu/Su = 500 and 2000 for soft Bangkok
clay and stiff clay, respectively. The measuring data was neglect the first movement due to the
delay of the first cantilever diaphragm wall around three months. The predicted and measured
lateral wall movements are in a good agreement.
Horiz. Displacement, mm
-10

10

20

30

40

0.00

-2.40
-5.10

5.00

Depth, m

-9.00
10.00

-11.80
-14.00
-15.50

15.00

-18.00
20.00

25.00

Inclinometer Tube
Penetrate into Pile Leg

PLAXIS
I-1
I-2

Figure 10 : Comparison of wall movement from FEM analysis and inclinometer results
COMPARISONS OF ANALYSIS WITH PRESSUREMETER TEST
The trial soil stiffness based on the back analysis shows that the Young modulus of soft
Bangkok clay and stiff clay are Eu/Su = 500 and 2000, respectively. Recently, six self-boring
pressuremeter tests were performed along the route of Metropolitan Rapid Transit (MRT)
northern contract by the Cambridge In-situ of Little Eversden (1997). The self-boring
pressuremeter tests were carried out by reload/loading cycles in order to determine the soil
stiffness at various shear strain levels. Figure 11 shows the results of pressuremeter test and

288

II ~ 74

1500

1 5 00

1250

1 2 50

1000

1 0 00

G /Su

G /Su

shows the variation of shear modulus with shear strain, Figure 11.1 is for soft clay while
Figure 11.2 is for first stiff clay. For normal practice of diaphragm wall construction for deep
basements in Bangkok soft clay, the shear strain is in the order of 0.1-0.2%, where the G/Su is
about 160 or Eu/Su =480. In case of the first stiff silty clay layer where shear strain is in the
order of 0.05-0.1%, the G/Su is about 340 or Eu/Su = 1020. The soil stiffness of the soft clay
from pressuremeter test results is very close to the results of back analysis, however, the
values of stiff clay from pressuremeter test results is lower than the results of back analysis,
this is might due to the effect of bored pile stiffness inside of the excavation area. The
correlation from the self-boring pressuremeter was the non-linear soil model that will be used
to carry out the future back analysis.

750

500

500

250

250

0
0 .0 1

0 .1 0

0 .01

1 .0 0

0 .10

1 .00

Sh ear Strain (%)

Sh ear S train (%)

750

Figure 11.1: Self-boring Pressuremeter

Test Result of Bangkok Soft Clay

Figure 11.2: Self-boring Pressuremeter

Test Result of Bangkok Stiff Clay

CONCLUSIONS
The analysis and the design of the diaphragm wall are normally based on the
assumption of soil stiffness. Back analysis to verified the soil stiffness based on three
different cases as under unbalanced loading, underpass project under very close to the
foundations of the first stage expressway, and short embedded wall. The results shows that the
soil stiffness in terms of Young modulus is Eu/Su = 500 and 2000 for soft Bangkok clay and
stiff clay, respectively. These values were also corresponding to the results of self-boring
pressuremeter tests of the first blue-line subway project in Bangkok.
REFERENCES
Brinkgreve, R. and Brand, P.A.(1996): Application of PLAXIS for soil and rock plasticity,
Short Course on Numerical Analysis in Geotechnical Engineering, AIT.
Cambridge In-situ (1997): Self-boring pressuremeter MRTA initial system project, North
Contract.
Duncan, J.M. and Buchigani, A.L.(1976): An engineering manual for settlement studies,
Geotechnical Report of Civil engineering Department, university of California at
Berkeley.

289
II ~ 75

Mair,R.J.(1993): Development in geotechnical engineering research, application to tunnels and

deep excavations, Proceedings of the Institution of Civil Engineers and
Civil
engineering.
Shibuya, S., Hanh, L.T., Wilailak K., Lohani T.N., and Tanaka H. (1997): Characterizing
stiffness and strength of soft Bangkok clay from in-situ and laboratory tests, First Int.
Conf. on Site Characteristics.
Thasananipan N., Muang A. Singtokaew K. (1996): Experiences in construction of diaphragm
wall for underpass project, Seminar on Underground Construction96, Engineering
Institute of Thailand.
Thasananipan N., Teparaksa W, Muang A. and Baskaran G., (1998): Design, construction and
behavior of bored cast in-situ concrete pile in Bangkok subsoils, Proc. 4th Int. Conf. on
Case Histories in Geotechnical Engineering, St. Louis, Missouri, USA.
Thasananipan N., Muang A., Tanseng P. and Wei S. H., (1998): Performance of a braced
excavation in Bangkok Clay, diaphragm wall subject to unbalanced loading conditions,
13th Southeast Asian Geotechnical Conference, Taipei, Taiwan ROC.
Teparaksa W., Thasananipan N., Muang A., and Wei S. (1998): Prediction and performances
of short embeded cast in-situ diaphragm wall for deep excavation in Bangkok subsoils,
Proc. 4th Int. Conf. on Case Histories in Geotechnical Engineering, St. Louis,
Missouri, USA.
Teparaksa W. (1999): Geotechnical aspects of the Bangkok subway design and construction,
guest speaker for the Society of Professional Engineers Thailand (S.P.E.T.) member
meeting at British Club, 18 February. .
Theramast N. (1998): Characterization of pseudo-elastic shear modulus and shear strength of
Bangkok clay, M. Eng. Thesis, AIT.

290

II ~ 76

GeoEng2000

An International Conference on
Geotechnical & Geological Engineering
19-24 November 2000
Melbourne Exhibition and Convention Centre
Melbourne, Australia

Diaphragm wall and barrette construction for Thiam Ruam Mit

Station Box, MRT Chaloem Ratchamongkhon Line, Bangkok
Thasnanipan N., Maung A.W. and Ganeshan B.
SEAFCO Co., Ltd., Bangkok, Thailand

ISBN No. 1-58716-068-4

291

GeoEng2000 An International Conference on Geotechnical and Geological Engineering 19-24 November 2000, Melbourne, Australia

DIAPHRAGM WALL AND BARRETTE CONSTRUCTION FOR

THIAM RUAM MIT STATION BOX, MRT CHALOEM
RATCHAMONGKHON LINE, BANGKOK
Narong Thasnanipan1, Aung W. Maung2 and Ganeshan Baskaran3
ABSTRACT
Construction experience of cast-in-situ diaphragm walls and foundation barrettes for the Thiam Ruam Mit
Station, Chaloem Ratchamongkhon line of Metropolitan Rapid Transit Authority (MRTA) is briefly reported
in this paper. The construction activities in the midst of heavily congested traffic and in the vicinity of underground utilities to meet the specified quality and safety controls are also presented.
INTRODUCTION
The Chaloem Ratchamongkhon line is the first underground mass rapid transit project in Bangkok, Thailand being constructed under the supervision of Metropolitan Rapid Transit Authority (MRTA). The project
involves the construction of about 22km-long, twin bored, single track tunnels, 18 station boxes and cutand-cover tunnels. The project was divided into two separate contracts, South and North. The south contract
was awarded to BCKT Joint Venture consisting of Bilfinger + Berger Bauaktiengesellschaft, Ch. Karnchang
Public Co., Ltd., Kumagi Gumi Co., Ltd. and Tokyu Construction Co., Ltd. The north contract was awarded
to ION Joint Venture, comprising Italian Development Public Co., Ltd., Obayashi Corporation and Nishimatsu Construction Co., Ltd. This paper presents experience on construction of diaphragm walls and foundation barrettes for Thiam Ruam Mit Station (station no. 12) of the North contract where heavy traffic conditions prevail.
MRT (Underground Structures-North)
MRT (Underground Structures-South)
MRT Interchange Station
MRT Station
SRT System (Hopewell)
BMA System (BTS)

BANG SU STATION

CH
AO
PH
RA
YA

RIV
ER

LEGEND :

THIAM RUAM MIT STATION

DEPOT

HUA LAMPHONG STATION

The station is one of the biggest stations

and located on the Ratchadaphisek Road. The
station box is enclosed by principally 1.0m
thick side walls and 1.20m thick end walls
32.0-42.0m deep. A Tunnel Boring Machine
(TBM) launching shaft was included in the
north side of the station. To facilitate topdown construction, 63 Barrettes (1.2mx3.0m)
embedded 44.5-55.0m deep in conjunction
with pre-placed stanchions at the top were
used.

Figure 1. Layout of MRT Chaloem Ratchamongkhon line

Three basement levels were planned for the station. The excavation depth inside the station box was up
to 24.5m for tunneling and basement slab construction. As deep excavation required stringent tolerances for
the diaphragm walls and barrettes, a high level of quality control was necessary.
1

Managing Director, Seafco Co., Ltd. Bangkok, Thailand

Project Manager, Seafco Co., Ltd. Bangkok, Thailand
3
Project Engineer, AVOPILING Pty., Ltd. NSW 2087, Australia
2

293

Construction of diaphragm walls and barrettes of the Thiam Ruam Mit station was carried out by an experienced local foundation contractor in which the authors worked while international contractors undertook
similar work for the remaining 17 stations.
SUBSOIL CONDITIONS
0m

10m

SOUTH

The subsoil profile, as shown in Fig. 2 consists of

made-ground up to 2.5m thick for pavement underlain
by a series of clay and sand layers. Below the madeground is 12.0m to 14.8m-thick Bangkok Soft Clay.
The first stiff clay layer occurs below 14.0m depth and
extends to 21.0m, overlying the first sand layer. The
sand layer is about 20.0m thick at this location and has
two thin (about 2-4m thick) very stiff clay layers at the
top. Below the sand layer is stiff to hard clay layer
which extends beyond 60.0m in depth. Within the clay
layer, from 48.0m depth, a 14m thick sand layer occurs
in the north and lenses out to the south. The properties
of soil layers are summarized in Table 1.

NORTH

STATION BOX

SOFT CLAY

TUNNEL

20m
TUNNEL
30m
SILTY SAND
40m

50m

STIFF CLAY
SILTY SAND

60m

Figure 2. Profile of soil layers and station box.

Table 1. Summary of soil properties
Soil Type
Depth to Top
of Layer (m)
Made-Ground
0.0
Bangkok Soft Clay
0.2-2.6
Stiff Clay
14.2-17.3
Sand
24-28.6
Stiff to Hard Clay
40.8

w
%
40
64-100
16-51
16-20

Cu
kPa
30-45
15-31
107-250
61-410

s
kN/m3
17.0-19.0
14.6-16.8
18.3-21.8
20.8-21.1
21.0-21.1

SPT- N
Blow/ft
5-9
13-17
27-59
11-61

The depth of ground water table at the site generally varies from 0.9m to 2.2m. Piezometers installed in
the project area indicate the pore pressure is hydrostatic below 23m depth due to pumping of ground water.
SITE CONDITION AND PREPARATION

2-SC

R.S.TOWER

2-SC

2-SC

2-SC

OPEN AREA

Mr. P FRESH
MILK

THAI
INSURANCE

2-SC

4-CF

ENTRANCE 12.1

4-CF
ENTRANCE 12.2

C
L

C
L
C
L
C
L

VENT.

ENTRANCE
12.4

SHAFT 1

ORAN
GE LIN
E

TO RAMA IX ROAD

TBM
DRIVE
SHAFT

RATCHADAPHISEK
RD.

C
L

27.000

TO LATPHRAO ROAD

C
L

ENTRANCE 12.3
VENT.
SHAFT 2

359.000

Figure 3. Layout of station box

Traffic Management
The construction site was on a 35.0m-wide main road with 4 inbound and 4 outbound traffic lanes. During the construction period a maximum of only two traffic lanes could be occupied at a time, allowing 3 or 4
lanes for each traffic direction. The width of the working area available was about 11.0m. Sometimes, the

294

width of the work area needed to be widened locally to position the grab crane for diaphragm wall panel excavation.
Utility Diversion
The site is bounded by some underground utilities, including a 1.0m diameter water pipe and telecommunication cables in the close vicinity of the work area, particularly on the east side. Above ground, 69kV electricity lines, telephone cables, drainage and sewage pipes (1.5m dia.) lay on both sides. The water pipe ran
along the planned diaphragm wall at about 3.0m depth and thus it was diverted by the other subcontractor to
the west, outside the station box prior to construction. The face-to-face clearance distance between the diverted water pipe and the outer face of the station wall was only 35cm due to space constraints. A reinforced
concrete wall to protect the pipe was constructed and it formed the bottom part of the guide wall for diaphragm wall panel excavation.
Two duct banks of the telecommunication cables were located 5-10cm away from the wall face at a depth
of 3.04m. 15mm thick steel plates were used in these locations as both guide walls and protection systems
for the duct banks.
CONSTRUCTION SPECIFICATION
Generally construction specifications were in line with the commonly practised specification for foundation works. Regarding construction tolerances, verticality of the walls and barrettes is specified within 1:200
while for pre-founded steel stanchions, the verticality allowed is within 1:400.
Table 2. Test and compliance values for bentonite
slurry
Property
Density
Viscosity
Fluid loss
Sand content
pH

Test
Specification
API RP13B1 Section
3
4
5
7
9

Freshly
Mixed
< 1.05g/ml
29-50
< 30ml
< 1%
8-12

Prior to
Placing
Rebar &
Concrete
< 1.10g/ml
29-50
< 30ml
< 2%
8-12

Properties of supporting fluid for wall and

barrette excavation were maintained within the
ranges as shown in Table 2. In addition, design
assumptions required that total construction
time below the soft clay layer was limited to
minimize the effects of bentonite on shaft friction and in order to achieve optimum shaft friction capacity of foundation elements.

CONSTRUCTION PHASES
Throughout the project, the main construction schedule for the station comprised several interfacing and
parallel work activities such as for utilities and traffic diversions, temporary decking, substructure work for
station box and the TBM driving shaft which other subcontractors undertook. This meant that guide wall
construction in particular needed to follow utility diversion as the same area was to be excavated for relocation of the utilities, removal of abandoned water pipes and construction of guide walls. The temporary decking and excavation for the TBM launching shaft followed the diaphragm wall construction.
DIAPHRAGM WALL AND BARRETTE CONSTRUCTION
Equipment and Plant
Mechanical rope-suspended grabs with crawler cranes (80tons) and service cranes (50-80tons) were used
as main construction equipment. Grabs, 1.0mx2.0-3.0m and 1.2mx3.0m were used to cut the required dimensions of panel excavations. Bentonite slurry was used as supporting fluid for panel and barrette excavation. Silos having a total of 300 cu.m storage capacity and de-sanding and de-silting units (80cu.m/hr capacity) were used to supply the slurry required for up to two-panel excavation at a time.

295

29.200

15.000

OF TNL.

TBM
LAUNCHING
SHAFT
4.500

Figure 4. General view of the site

6.300

C
L

C
L

OF TNL.

3.000

4.000

6.300

4.850

Figure 5. Panel layout of TBM launching shaft

Guide Walls
Generally 1.5m deep guide walls were used for panel excavation. In the location of underground water
pipes and duct banks, up to 3.0m-deep guide walls were used. Individual panels were sized with the following considerations: (1) type of reinforcement, (2) size of grabs used, (3) thickness and depth of panels, (4)
location of slab and wall openings for entrances and TBM, (5) stability of trench in connection with location
of underground utility and (6) location of public accesses and streets.
Panel Layout
The walls were divided into 184 panels with horizontal lengths varying from 2.5m to 4.85m. L-shaped
and T-shaped panels were used to form box structures of the station and TBM launching shaft (see Fig. 5).
Sequence of Construction
Construction activities took place day and night. To maintain the traffic flow, comply with local authority regulations, construction sequence and activities were planned to allow concrete pouring and spoil removal to be done before and after rush hours, in most cases at night. Generally panel excavation was carried
out at least 6.0m away from the recently cast panel or 16hrs after casting of the adjacent panels to avoid any
damage induced by excavation.
Supporting Fluid
Locally available bentonite powder was used for preparation of slurry. This product had been used for
more than a dozen completed diaphragm wall projects,. Approximately 496 tons of bentonite powder was
consumed for excavation work. The last portion of slurry displaced by concrete or contaminated with cement was usually discarded. Soil volumes of about 26,274cu.m and 10,740cu.m were excavated for diaphragm walls and barrettes respectively.
Reinforcement
Steel bars of SD50 were used for the main reinforcement bars. The reinforcement cages were 30.0m to
42.2m long and up to 4.5m wide, with 4 levels of box-out for slab connections for diaphragm wall panels.
For barrettes, cages were 48.1 to 54.6m long. 4 to 5 cage sections with lengths of 6-12m were joined to form
one continuous cage for individual panels and barrettes. The cage sections were fabricated off site and no
more than two were joined for transporting to the site. Cage sections were connected prior to and during
lowering into the trench. In order to place the box-out in position and to achieve an exact match of the cage
sections, markings were made on the main bars of each cage. U-shaped bolts were used for cage connections. Couplers were generally used for slab connection and both couplers and bent-out bars were used for
beam connections at the wall openings for entrances and slab openings. The large panel cages weighed up to
34.8 tons.
For panels with glass fibre reinforcement polymer (GFRP) for openings of TBM break-through, a temporary steel frame was necessary to hold the cage section due to different stiffness between GFRP and steel.
The frame was then cut away section by section while lowering the cage. Diaphragm wall panels were rein-

296

forced with 136-218 kg/m2 of steel bars while barrettes were reinforced with about 50kg/m3 of steel bars.
The wastage of steel was about 7% in which most quantity was consumed for cage hanging bars and frames
for soft eye openings. A total of 4,756.15 tons of reinforcement bars was used for diaphragm wall and barrettes. Steel pipes 150mm in diameter were installed inside the diaphragm wall panels at 5 locations as a
void former for inclinometer access tubes. 28 sets of vibrating wire strain gauges and total jack-out pressure
cells were also installed in the wall panels of TBM launching shaft.
Concrete Casting
Ready mixed concrete grades 40 (cube strength 40MPa at 28 days) and 35 (35MPa at 28days) to BS5328
were used for diaphragm walls and barrettes respectively. Two sets of tremie pipes were used in pouring
concrete for both diaphragm walls and barrettes. The largest single concrete pour was about 232cu.m for Lshaped panels. Total concrete volumes of 27,480cu.m and 6,611cu.m were poured for diaphragm walls and
barrettes respectively. Concrete over-consumption in diaphragm wall caused by old, sand back-fill in the
area of abandoned water pipes was estimated to be an average of 7.8% with a concrete wastage of up to
38.0% occurring in one panel. However, the average concrete wastage was 5.0%. For barrettes, the average
concrete wastage was up to 14%. The high wastage compared to building and elevated highway projects (up
to 7%) was caused by overcasting of concrete well above the cutoff level of 22m depth to ensure that sound
concrete reached above the cutoff. Aggregates were used to backfill the open trench of barrettes above the
concrete after casting.
Installation of Stanchion in Barrettes
Conventional plunging method was not adopted in view of the site conditions, low cutoff level, the stanchion design and horizontal bar arrangements in the reinforcement cage. Two specially designed installation
frames with adjustable screws were used for installation of stanchions to ensure high accuracy for verticality
and plan position. A removable 10m-long, steel column was fixed on top of the stanchion to check the verticality and alignment.
PRODUCTION RATES
The construction time required for a typical panel and barrette is detailed
in Table 3.
Table 3. Average construction time for typical diaphragm wall panels
(1.0x4.5x35.0m, cutoff 0.0m) and barrettes (1.2x3.0x44.5m, cutoff 22.5m).
Activities
Excavation
Checking verticality with Koden
Desanding
Cage installation
Stanchion installation
Tremie preparation
Concrete pouring

Diaphragm wall
Hours
18
1
4
2.5
2.5
4

Barrette
Hours
16.5
2
5
4.5
3
3.5
3.5

Figure 6. Guide frame for steel

stanchion installation in barrettes
Throughout the construction period, a total of about 867hrs, approximately 15% of construction time (excluding mobilization and site preparation period) was consumed by equipment maintenance and repair. Grab
teeth were broken and changed often during chipping off the concrete overbreak behind the stop-end plate
caused by loose material below the guide walls. Additionally, steel wire ropes for the grab also needed to be
changed frequently as they were worn out by both the chiseling action and broken concrete pieces. 17% of
construction period was rainy, but this usually had no significant effect on the production rate. A flash flood

297

occurred at one time by a heavy downpour. Standby time (14% of construction period) took place and was
contributed by (1) obstructions by overbreak concrete during panel excavation and stop-end plate removal
and (2) standing time after panel excavation reached below the soft clay to allow subsequent construction activities such as spoil removal and concrete delivery taking place outside rush hours. Traffic diversions were
carried out in multi-phases and took at least 5% of construction period.
QUALITY, SAFETY AND ENVIROMENTAL CONTROLS
As the alignment of diaphragm wall and barrettes is critical, every excavated panel and trench was
checked with Koden monitoring equipment. Monitoring was carried out after excavation reached about
22.0m (ie. at the levels of base slab and barrette cutoff) and toe levels. Necessary corrections to trench verticality were made during excavation if required. Panel ends, especially for those panels with soft-eyes for
TBM break-through were also checked to achieve high levels of accuracy for positioning. During excavation, reference posts were used for checking the grab position and observing rope position relative to the
trench sides at all times. Prior to lowering the reinforcement cage into the trench, any mudcake built-up on
the trench faces and panel joints was scraped off by a grab to which brushes were attached. Finally, any
sediment or loose materials deposited at the bottom of the trench were also removed by using the grab.
Reference bars attached to the top of reinforcement cage were extended above ground level to check the
position of the cage with survey equipment during installation of the cage into the trench.
Delivered concrete was checked for slump and cohesiveness prior to casting. As per specifications, adequate embedding length of tremie pipes in the concrete and slurry properties were maintained to achieve
good-quality concrete casting. Plugging materials and shutters were introduced in the tremie pipes to separate first concrete pour and slurry in the trench. Samples of reinforcement bars and concrete were taken as
specified for testing in the lab. 37 barrettes were provided with 6 steel tubes each for sonic logging to test
the barrette concrete quality. No significant anomalies were detected by sonic logging test, and integrity and
quality of barrette were found adequate.
As the site was in a busy public area, site and public accesses were kept clean all the time. Flagmen were
also provided for traffic control at each construction zone. All workers were inducted in safety procedure
prior to assuming duty. Construction activities were also supervised by a full-time safety officer. The safety
officer ensured that all labor used mandatory personal protection equipment and observed the safety regulations. All heavy equipment and cranes were checked for safety and certified for operation prior to use and
for regular maintenance. Additionally, regular safety patrols and inspection were also conducted by the main
contractor and supervising engineers.
CONCLUSION
Construction work on the public road, involving various activities parallel with different
teams/subcontractors demanded comprehensive construction sequences and planning to meet the targeted
milestones. Proper coordination, planning, supervision and strict safety and environmental control by the
parties involved thus brought about the successful completion of diaphragm wall and barrette construction on
one of the busiest roads in Bangkok without any serious accident. When the paper was finalized, excavation
inside the station box had been completed and it revealed good quality of workmanship without requiring
any significant remedial work.
ACKNOWLEDGEMENT
The authors wish to express their appreciation to the MRTA and Nishimatsu Construction Co. Ltd. for
their permission to publish this paper.
REFERENCES
Thasnanipan, N. Maung, A. W. and Tanseng, P. (1998), "Barrettes Founded in Bangkok Subsoils, Construction and Performance", Thirteenth Southeast Asian Geotechnical Conference, Taipei, Taiwan, ROC, November 16-20, 1998. pp. 573-578.

298

Under the Royal Patronage of His Majesty the King

Organized by :
The Engineering Institute of Thailand
Chulalongkhon University
Department of Civil Engineering
Department of Environmental Engineering
Department of Survey Engineering
Department of Water Resources Engineering
In Corporation with :
Department of Civil Engineering from Universities in Thailand

Engineering for Sustainable

Development
Proceedings of the Seventh National Convention on
Civil Engineering

Monitoring of diaphragm wall displacement and associated ground

movement, brace excavation adjacent to historical building at the bank of
Chao Phraya River
Pornpot Tanseng
Suranaree University of Technology, Nakornratchasima, Thailand
Zaw Zaw Aye and Chanchai Submaneewong
SEAFCO Co., Ltd., Bangkok, Thailand

ISBN 974-13-0107-3

299

MONITORING OF DIAPHRAGM WALL DISPLACEMENT AND ASSOCIATED GROUND

MOVEMENT, BRACED EXCAVATION ADJACENT TO HISTORICAL BUILDING
AT THE BANK OF CHAO PHRAYA RIVER

Pornpot Tanseng1, Zaw Zaw Aye2 and Chanchai Submaneewong3

1

Lecturer
Suranaree University of Technology, Nakornratchasima, Thailand
2
Project Manager, 3Geotechnical Engineer
SEAFCO Company Limited, Bangkok, Thailand

(Diaphragm wall)

(Unbalanced loading condition)

ABSTRACT: Modern

buildings in Bangkok frequently require the excavation works for the basement facility
adjacent to the existing structures. Diaphragm walls have been commonly used as permanent retaining walls for
excavation works in various projects. Comprehensive monitoring system is of essential for any excavation works
particularly for those adjacent to the sensitive buildings. This paper presents the monitoring of diaphragm wall
displacement and associated movement of the ground and adjacent structures including historical building using
a number of instruments for basement construction works located at the bank of Chao Phraya River. The bracing
system and construction sequence adopted for excavation with the consideration of the presence of an
unbalanced loading condition are also briefly discussed. Building damage risk assessment using a simple
approach was carried out based on the instrumentation results and compared with the damage criteria established
by published research work. A course of action undertook in response to the observed critical movement is also
presented.

KEYWORDS: Diaphragm wall, Instrumentation, Unbalanced Loading

For further details, contact Mr. Zaw Zaw Aye, SEAFCO Co., Ltd. 26/10 Prayasuren Road, Bangchan,
Khlong Sam Wah, Bangkok 10510, Thailand

301

1.

INTRODUCTION
Excavation works in soft ground inevitably induced the ground movements which in turn may
damage the adjacent existing structures. Diaphragm walls are commonly used as permanent
retaining walls for basement excavation of many projects in Bangkok. It is necessary to use this
type of rigid retaining wall in excavation work to minimize the ground movement particularly
where the work is to be carried out in close proximity to the existing sensitive buildings. This
paper presents the monitoring of diaphragm walls displacement and associated movement of the
ground and the adjacent structures including historical building using a number of instruments
for completed construction works located at the bank of Chao Phraya River, Bangkok. Design
and construction aspects of the project have been reported by Thasnanipan et al. (1999).

2.

PROJECT LOCATION AND CONSTRAINTS IMPOSED ON SITE

The project is situated in the heart of an old established culturally-significant area. The project
consisted of basement excavation works for two separate buildings with the support of braced
diaphragm walls.
Chao Phraya River
I1-7

I1-6

D1-1

Chao Phraya River

D2-4 S2-4

D1-6
T1-1
E1-2
T2-1

B3

T1-2

I1-8
D1-3

S2-3
D2-3

S2-1
I2-6
E2-1 S2-2
T2-2
D2-2

T2-10
E2-5
I2-3
T2-9
B6
S2-8
D2-8

Second Building

Building

T1-3
I1-5

S2-10
S2-9

D2-1

5 Storey

ing
First Build

I2-1

D2-10
D2-9

T1-4

I1-2
D1-2 T1-8
E1-4
E1-5

Historical
Building
(2 Storey)

T1-10

T1-9

E1-1

S2-5

I2-5

S2-6 D2-5 I2-4

T1-5
T2-4

I1-3

B1

T2-8
D2-7 E2-4

T2-3

B4

T2-5 E2-2

T2-6

T2-7 E2-3
0

D1-8
D1-2
B2

D2-6 S2-7

6 10

B5

20

I1-4
Legend

T1-6
E1-3
Building

Inclinometer in D-Wall
Tilt plate
Vertical beam sensor
Surface settlement plate (1.0m)

7 Storey

Deep settlement plate (5.0m)

VW Strain gauge
Earth pressure gauge
Pre-load location

Figure 1 Layout of excavation sites and instrumentation

Two basements are located 40m apart, separated by an existing 5 storey building. As shown in
Figure 1, construction site is located along the bank of Chao Phraya River and surrounded by
existing buildings including a historical building. Along the river, the distance between
diaphragm walls and existing old river wall is about 4m and 8m for the first and second buildings
respectively. Locations of the excavation sites themselves posed some constraints, which called
for the need of careful consideration in establishing the design principles and sequence of
construction. The major constraints imposed on sites are outlined below.
2.1 Loading condition
Thasnanipan et al. (1999) reported the presence of unbalanced loading condition. As the
excavation areas are located immediately next to the river and surrounded by existing buildings,
the walls alongside the river and the opposite walls were particularly expected to undergo an
unbalanced lateral loading condition. No case history data of similar condition in the area were
available at the time of design and construction of the basements.
2.2 Risk of damage to the existing buildings
The existing buildings located in the proximity of the excavation areas are of old structures.
According to the available information, the buildings are supported by the piled foundation.
However, no detailed information such as type, size and length of the piles was available. The
most critical structure is the historical building which is a predominantly masonry and only 5m
away from the outer face of the diaphragm wall of the first excavation site. As the building is
important for its historical heritage, a damage category worse than negligible (as defined by
Rankin 1988) would not be acceptable.

302
GTE - 54

3.

SUBSOIL INVESTIGATION
Two stages of site investigation were made. Preliminary stage of site investigation was
performed during the tendering stage. Second stage of soil investigation was conducted prior to
designing temporary bracing and basement excavation work. As subsoil properties obtained from
the boreholes and test data were well reported by Thasnanipan et al. (1999), they are not
presented again in this paper.

4.

DIAPHRAGM WALL AND BORED PILES

For both buildings, 800mm thick cast in-situ concrete diaphragm walls of 28m toe depth with
two level temporary bracings were designed for basement excavation. The maximum excavation
depths for the first and second buildings were 12.7m and 9.7m respectively. Deep diaphragm
walls of toe depth 28m were necessary to ensure the overall stability considering the location of
the basement excavation at the bank of the river. Bored piles of diameter ranging from 800mm to
1500mm founded at 48m depth were constructed by wet process using bentonite slurry.

5.

INSTRUMENTATION AND MONITORING PROGRAM

Instrumentation played major role in this project. The primary objective of the instrumentation
program was to monitor the performance of the excavation to ensure that diaphragm wall as well
as the entire excavation system were stable and that adjacent structures were not adversely
affected. Furthermore, the instrumentation program was established to provide the main feedback
in application of observational method. Types of instrumentation used in the project are
presented in Table 1. Layouts of instrumentation for the first and the second buildings are shown
in Figure 1.
Table 1 Quantity of Instrumentation and parameters measured
Type of Instrumentation
First Building
Second Building
Inclinometer
8
6
Settlement Plate
10
20
Tiltmeter
10
10
Vertical Beam Sensor
5
5
VWSG
1 location*
Earth Pressure Gauge
2x2
* A pair of VWSG in 5 layers along the depth of wall panel

6.

Monitored Parameters
Wall deflections
Ground settlements
Tilting of Buildings
Tilting of buildings
Strain in rebars
Strut forces

EXCAVATION METHOD AND BRACING SYSTEM

Conventional bottomup method with two levels of temporary bracing was applied for both
buildings excavation. As reported by Thasnanipan et al. (1999), to prevent adverse wall behavior
alongside the river due to the presence of unbalanced lateral loading condition, pre-loading was
applied at one end of struts on the land side wall with simple and efficient temporary bracing
system. As the excavation work of the first building was completed 1 year before the
commencement of the second building, monitoring results of the first excavation work provided
an ample opportunity to review the design assumption and fine tune the parameters used in the
analysis of the diaphragm wall for the second building. The major modification was to use single
strut for the first level of the second building (less rigid but economical compared to that of the
first building) with the provision of strut-force monitoring, according to Thasnanipan et al.
(1999). Summarized information of excavation and bracing systems of two buildings is tabulated
below.
Table 2 Summary of excavation and bracing systems
Building
First

Excavatio
n Depth
-12.79m

Second

-9.70m

Bracing
Level
I
II
I
II

Bracing
Elevation
-2.0
-6.5
-2.0
-7.0

Strut
Sections
2 x WF350 x 350
2 x WF400 x 400
1 x WF400 x 400
2 x WF350 x 350

303
GTE - 55

Design Strut Force

kN/m
484.0
789.0
279.4
332.1

7.

SETTLEMENT PREDICTION
Due to the presence of unbalanced loading condition, non-symmetric pattern of settlement
distribution around the perimeter of excavation could be expected. However, considering the fact
that pre-loading was to be used for the land side diaphragm wall but not for the riverside wall,
settlement distribution was assumed to be symmetrical. Available methods were reviewed for
predicting the ground movement with the consideration of two main factors such as, (1) simple
and practical in application (2) enable to correlate with the predicted and measured diaphragm
wall deflection.

7.1 Prediction of Surface Settlement

The method proposed by J. E. Bowels (1990) was selected as it meets the required criteria
mentioned above. Bowels suggested the ground settlement induced by excavation as a function
of ground loss due to the deflection of the retaining wall. Bowel demonstrated the calculation of
settlements at specified distance by assuming parabolic variations of settlement within the
influence distance. Using predicted diaphragm wall deflection, surface settlement behind the wall
was computed by empirical formulas proposed by Bowels.
7.2 Prediction of Sub-surface Settlement
A simplified prediction of sub-surface settlement was made based on the calculated surface
settlement described above. First, subsurface settlement influence line was constructed. As
shown in Figure 2, settlement influence zone is assumed to decrease with depth from Do at the
surface and zero at the wall toe. With the assumption of linear relationship between the volume
of deflected wall shape and the volume of settlement trough at any depth within settlement
influence zone, subsurface settlement at different depths were calculated.

Sio

D O = HW+ W
Surface settlement trough
volume, V To
S WO =

HE

Dy
Siy

Sub-surface settlement
trough volume, VTy
Settlement
influence zone
deflected volumn V y at depth (H w- Y)

Swy

HW

D-wall

SWO

W
Total deflected
shape volume, V O

4V O

DO

x
S io = SWO *

DO
D
DY = Y * O
HW
V
VTY = VY * TO
VO

Figure 2 Demonstration of settlement prediction from diaphragm wall deflection values

8. INSTRUMENTATION RESULTS
8.1 Inclinometer Monitoring
The predicted and observed diaphragm wall deflections of the first building and second buildings
are presented in Figure 3a and 3b respectively. As can be seen in the figures, observed wall
movements exceeded the predicted values at the top portion of the walls for both buildings. The
delay in installing the first level bracings, leaving the walls in cantilever conditions for the long
period was the main reason of the large movement induced by the excavation at that stage.

304

GTE - 56

RIVER

20

Displacement (mm)

Displacement (mm)

Displacement (mm)
0.00

40

40

20

0.00

-1.50

20

40

60

Displacement (mm)

80

80

60

5.00

-6.50

-7.00
-9.70

10.00

-12.70
15.00

Depth (m)

Depth (m)

PREDICTED
OBSERVATION

Day 67
Day 133

Day 221

10.00

15.00

Day 155
Day 221

Prediction

20.00

Day

25.00

67
133
155
221

25.00

30.00

20

Day 67
Day 133
Day 155

20.00

40

-2.00
River

5.00

Inclinometer
No. I-7

Inclinometer
No. I-2
30.00

Inclinometer

No. I-2

st

Figure 3a. Lateral wall movements (1 Building)

Description
Excavate to -2.5m
Excavate to -7.5m
Excavate to -9.70m
Remove strut at -7.00m

Inclinometer

No. I-4

Figure 3b. Lateral wall movements (2nd Building)

At Inclinometer No.I-6 of the second building, wall movement was found to reach the trigger
value at top portion when excavation reached 7.5m depth after the first level bracing installation.
Lateral wall movement reached a maximum rate of 6.5mm/day and a tension crack on the ground
was found about 8m away from the wall. A close examination of the bracing system indicated
that one strut had swayed slightly about 130mm. Immediate actions were taken by stacking
additional strut on the defective one and installing of the second level bracing carried out at that
area. The wall movements were found to cease after these actions.
The wall deflections of the first building excavation found unaffected by the unbalanced lateral
loading condition. However, for second building, monitoring results from inclinometer I-2 and I4 suggested that the wall alongside the river was pushed against the retaining soil by the opposite
wall.
8.2 Settlement and Building Movement
Figures 4a, 4b and 4c present the construction activities with time and corresponding monitoring
data. As can be seen in the figures, monitoring results reflected the construction activities in
general. It was also proved that performance and response of the diaphragm wall were largely
influenced by the construction activities. It can be observed that significant changes in the
responses were mainly occurred during the initial excavation stages in which walls were
unsupported for long period. Monitoring data particularly that of settlement and building tilt
measurement provided ample opportunity to assess the risk of the existing buildings damage. As
the measured data were far less than the critical values, the excavation works were proceeded
with confidence.
8.3 Strut Force Measurement
Daily monitoring was carried out for pressure gauges installed on struts. The comparison of
predicted and measure strut forces are presented in Table 3. In general, the measured values
indicated that the bracing system used was adequate.
Table 3 Comparisons of Predicted and Measured Strut Force

Bracing Level
First Level
Second Level

Predicted Force
(KN/m)
354.5
307.3

Measured Force (KN/m)

331.3
206.3

305

GTE - 57

Excav atio n R e cord

0

Excavation Record

Time , D ays

50

100

150

200

250

50

Time, Days
100

150

200

250

300

350

Exc. De pth, m

2
4

Exc. Depth, m

6
8
10

Excav. near the river

12

Excava. aw ay from the river

14

Tilt M e te r D ata
200

Strut Load Data

300

Strut Load (KN)

Tilt, Se cond

100

T 1 - 1 (B 3 )
T 1 - 6 (B 2 )
T 1 - 8 (B 1 )

-1 0 0

-2 0 0

a) Tilt meter results with construction time

(1st Building)
Excavation Record
50

150

200

250

300

100

Second Strut

50

Surface Settlement Data

350

0.0
Exc. Depth, m

150

10

Time, Days
100

First Strut

200

2.0
4.0
6.0

Settlement, mm

250

8.0

0
-10

S6
S7
S9
S10

-20
-30
-40

10.0

Tilt Meter and EL Sensor Data

Tilt, Second

100
0
-100

E2-2 (B4)
T2-4 (B4)
T2-8 (B5)
T2-10 (B6)

-200

b) Tilt meter results with construction time

(2nd Building)

Max. Hor. Movement (mm)

Inclinometer Data

200

80

I-4

60
40

I-2

20
0
0

50

100

150

200

250

300

350

(c) Strut load, settlement and max. wall movement

with construction time (2nd Building)

Figure 4 Monitoring results of instruments with construction time for First and Second Buildings

9.

BUILDING DAMAGE RISK ASSESSMENT

A simple approach was adopted for building damage assessment. Settlement and slope of the
buildings due to the excavation works were predicted assuming the green filed condition.
Neglecting restraints from foundation and structure, it was assumed that buildings follow the
ground settlement trough at foundation level (estimated pile tip level). Using predicted and
measured deflection values of diaphragm walls, surface and subsurface settlement profiles were
prepared as shown in Figures 5 and 6. Building slopes likely to be induced by differential
settlement were then predicted as presented in Table 4.

306

GTE - 58

Distance from Wall (m)

Distance from Wall (m)

0

10

15

20

25

30

35

40

0
-5

-5

-10

-10

Settlement (mm)

Settlement (mm)

10

15

20

25

30

35

40

-15
-20

-15
-20

surface settlement

-25

settlement at depth 7m
settlement at depth 14m

settlement at depth 14m

-30

surface settlement

-25

settlement at depth 7m

-30

b) From measured deflection values

a) From predicted deflection values

Figure 5 Prediction of surface and subsurface settlement induced by 1st building excavation
Distance from Wall (m)

Distance from Wall (m)

5

10

15

20

25

30

35

40

-10

-10

-20

-20

Settlement (mm)

Settlement (mm)

-30
-40

10

15

20

30

35

40

-30
-40
surface settlement

surface settlement
-50

settlement at depth 7m

-50

25

settlement at depth 14m

settlement at depth 7m
settlement at depth 14m

-60

-60

a) From predicted deflection values

b) From measured deflection values

Figure 6 Prediction of surface and subsurface settlement induced by 2nd building excavation

Table 4 Predicted Building Slope Calculated from D-Wall Deflection

Building
No.

Depth of
Settlement
Considered*

Predicted Building Slope Calculated

from D-Wall Deflection
From
From Inclinometer
Prediction
Measurement
1/1778
1/2000
1/1778
1/2133
1/1583
1/1827
1/1583
1/2879
1/1563
1/1667
1/1667
1/2778
1/880
1/880
1/978
1/1257
1/944
1/1700

Maximum Tilt
Measured by Tilt
Meter / EL Beam

Building Slope
Criteria for
Negligible Damage**

At surface
1/2082
At depth 5m
At surface
B2
1/4125
At depth 7m
At surface
B3
1/2578
At depth 7m
At surface
B4
1/1412
At depth 7m
At surface
B5
>1/5000
At depth
1/1771
1/4250
14m
At surface
1/926
1/926
B6
1/3437
At depth
1/2256
1/4400
14m
Note : (1) * assumed foundation level (2) ** unlikely to have superficial damage (Rankin 1988)
B1

< 1/500
< 1/500
< 1/500
< 1/500
< 1/500
< 1/500

As can be seen in Table 4, the measured maximum tilt of the buildings reasonably agree with
predicted building slopes derived from inclinometer measurement, particularly if the buildings
were assumed to be settled by subsurface settlement at assumed pile tip levels. All buildings

307

GTE - 59

were predicted to be in negligible risk category as classified by W.J. Rankin (1988). Field
inspection confirmed that no damage was caused in all buildings from the minor movements
induced by excavation works.
10. CONCLUSIONS
Instrumentation played a major role in execution and successful completion of excavation works
adjacent to the existing sensitive buildings under unbalanced loading condition. Effective bracing
system, excavation sequence and adequate embedment of the retaining walls were the main
factors contributed in minimizing negative impact of the unbalanced loading condition.
A simple approach in prediction of ground settlement and assessment of the risk of building
damage has been presented. The measured maximum tilt of the buildings reasonably agreed with
the predicted values. It is necessary to use the rigid diaphragm wall in excavation work to
minimize the ground movement particularly where the work is to be carried out in close proximity
to the existing sensitive buildings.
REFERENCES
1. Bowels J.E., Foundation Analysis and Design, 4th Edition, McGraw-Hill, Singapore, 1990.
2. Rankin W.J., Ground Movement Resulting from Urban Tunnelling : Prediction and Effects,
Engineering Geology Special Publication No.5, p. 79-92, 1988.
3. Thasnanipan, N., Maung A. W., Tanseng P. & Teparaksa W., Behavior and Performance of
Diaphragm Walls under Unbalanced Lateral Loading along the Chao Phraya River , Field
Measurements in Geomechanics, Leung, Tan & Phoon (eds), p. 267-272, Balkema, 1999.

308

GTE - 60

Proceedings of the 12th Asian Regional Conference on Soil Mechanics

and Geotechnical Engineering
4-8 August 2003, Singapore

th

12 Asian Regional
Conference on Soil Mechanics
and Geotechnical Engineering

Performance
of
Buttress-Support
Thin
Diaphragm Wall for Underground Car Park in
Bangkok
Narong Thasnanipan,
Submaneewong

Zaw

SEAFCO Co., Ltd. Bangkok, Thailand

309

Zaw

Aye

and

Chanchai

Performance of Buttress-Support Thin Diaphragm Wall for Underground

Car Park in Bangkok
Narong Thasnanipan, Zaw Zaw Aye, Chanchai Submaneewong
SEAFCO Company Limited, Bangkok, Thailand, [email protected]

Abstract : This paper presents the geotechnical aspect of the construction of two-level underground car park building located in the culturally and historically significant area of Bangkok. Performance of buttressed-support diaphragm wall of 0.60m width is reported
based on the inclinometer monitoring results. Intensive modification of construction sequence in actual work execution with value
engineering options different from tender stage design is demonstrated along with application of observational method.

1 INTRODUCTION

3 MAJOR CONSTRAINTS AND DESIGN PRINCIPLE

Diaphragm walls have been utilized for underground permanent

retaining structures in Bangkok for many years. The common
width of the diaphragm wall in Bangkok ranges from 0.80m to
1.20m. Thin diaphragm wall of width less than 0.80m is rarely
used in this growing-metropolis. Performance of buttress-support
diaphragm wall of 0.60m width for two level underground car
park building construction is discussed in this paper.

Location of the project site itself in the vicinity of sensitive

buildings posed some constraints, which called for the need of
careful consideration in establishing the design principles and sequence of construction. The architectural and utility aspects of
the project called for the design of the basement with a number
of openings from the ground surface to the final basement slab
level to facilitate the ventilation system as shown in figure 1.
Hence the roof slab can not be physically utilized as bracing in
most of the area where the diaphragm wall is to be acting as a
cantilever retaining wall in the permanent stage. It was analyzed
in the preliminary analyses that the deflection of the diaphragm
wall of 0.60 m width would be large if it was to be fully cantilevered. As the project is located in a sensitive area, ground
movement induced by large deflection was unfavorable. It was
therefore decided to use buttress to minimize the diaphragm wall
deflection as shown in figure 2.

CITY HALL
CITY
HALL

OPENING

OPENING

SIRIPONG ROAD

OPENING

OPENING

OPENING

SHOP-HOUSES

OPENING

MAHUNNOP ROAD

SHOP-HOUSES
DINGSOR ROAD

The project is a two-level underground car park located in the

center of Rattanakosin Island, the heart of an old established, historically and culturally significant area of Bangkok. As shown in
Figure 1, construction site is surrounded by numbers of sensitive
structures - in the south, Wat Suthat, one of Thailands most important temples and the Historical Giant-swing, in the north, the
City Hall and at South-east corner, the Historical Brahmin Temple. Rows of old shop-house buildings are closely located in the
east and west boundaries of the project.
The project owner, Bangkok Metropolitan Authority (BMA)
awarded the semi-turnkey basis construction contract to
SEAFCO Co., Ltd. as a contractor. The contract consists of 3 major scope of works : (1) Construction of diaphragm wall, barrette
and bored piles (2) Excavation works including temporary bracing design and installation (3) Construction of the entire basement structure having car park area of 18,552 m2 and roof-level
park of 10,936m2 plus cut-and-cover tunnel, underpass access to
the City Hall. Since the project area was being used as a gradelevel car parking space prior to the award of the contract, the key
requirement was to construct the underground car park in two
phases to construct Phase 1 while leaving space for car parking
in Phase 2, and to utilize semi-finished underground car park of
Phase 1 during construction of Phase 2.
Diaphragm wall having 600mm width founded at 16m below
ground level (B.G.L) was constructed simultaneously with dryprocessed bored piles of diameter 600mm with toe depth 20m
B.G.L. Barrettes having same toe depth as bored piles were installed at 8m spacing along with diaphragm wall panels. Sheet
pile wall (14m deep) was used as a temporary retaining wall at
the boundary of Phase 1 and 2 as shown in figure 2.

BRAHMIN TEMPLE

2 PROJECT OVERVIEW

OPENING

OPENING

BAMRUNG MUANG ROAD

GIANT-SWING

Fig. 1. Layout plan of the project showing adjacent buildings

311

I-5

I-4

first strut and final excavation level for 600mm diaphragm wall)
due to large volume of excavation work involved.

NORTH D-WALL

EAST D-WALL

WEST D-WALL

6 POST TENDER DESIGN REVIEW AND MODIFICATION

FOR PHASE 1

I-6

PHASE 2

Prior to the commencement of excavation works, value engineering review of the temporary works was undertaken by the contractors new team of design engineers together with the construction team. The main objectives were to minimize the
material and construction sequence involved in temporary works
so as to accelerate the excavation time thereby saving overall
costs without compromising the safety aspect. After conducting a
series of re-analyses with different conditions major modifications were made : (1) To change first strut level to 1.8m from
the original tender stage design level 1.0m (2) To use only 1
temporary strut, omitting second level raking strut with the provision of sloping soil berm against diaphragm walls. Soil berm
was to remove after completion of base slab construction in the
majority of area minimizing the elapsed time of partially unstrutted diaphragm wall between temporary strut and the final
excavation level.

124 m

SHEET PILE

See details in (b)

I-3

I-1

PHASE 1

I-2

SOUTH D-WALL

78 m

(a)
D IA P H R A G M W A L L

RC BUTTRESS

BARRETTE

W ALE BEAM

0 .3 0

T IE B E A M

7 APPLICATION OF OBSERVATIONAL METHOD

P IL E C A P

SLAB

(b)
+0.20 ROOF SLAB

0.00
CAPPING
BEAM
RC BUTTRESS

G.W.L

TIE
BEAM

-9m

-12m

-16m

-2.70 SLAB (B1)

PILE CAP
-5.60 (B2)
SOFT CLAY
Su = 18 kN/m2
Eu = 7000 kN/m2, = 16.5 kN/m3

WALE
BEAM

DIAPHRAGM
WALL

MEDIUM CLAY
Su = 30 kN/m2
Eu = 19250 kN/m2, = 17.5 kN/m3
STIFF CLAY
Su = 60 kN/m2
Eu = 45000 kN/m2, = 19 kN/m3

FOUNDATION
PILES
DIA.600MM.

BARRETTE
-20m.

(c)
Fig. 2 (a) Layout of diaphragm wall and piles (b) plan of diaphragm wall and buttress (c) sectional view of the structures
4 SUBSOIL CONDITION
Similar to other localities in Bangkok a typical subsoil profile at
the site is characterized by thick Bangkok soft clay layer at the
top followed by thin layer of medium clay, and stiff clay layers
as depicted in figure 2 (c). The first sand layer is found below
26m from the existing ground level.

In order to assess the most probable condition assumed in value

engineering design and to take necessary actions if monitoring
results reveal most unfavorable conditions (i.e. actual deflection
of diaphragm wall reaches maximum acceptable limit), observational method was implemented on the followings basis.
Reviewed the design parameters together with critical conditions posed on sites as well as most critical stage of excavation
Designed and predicted the performance of diaphragm wall using the most probable conditions and parameters.
Established the trigger criteria based on predicted diaphragm
wall deflection
Predefined the practical contingency plan for most unfavorable conditions where wall deflection reaches trigger levels
Set out the instrumentation program with the consideration of
above factors.
Monitored the performance of diaphragm wall. Compared the
monitoring results with the predicted and trigger values and reassessed
Implemented the contingency measures if monitoring result
reaches action level of trigger values
Comprehensive and robust monitoring program was set up as
a key element in application of the observational method and to
ensure that modified construction sequence would not have adverse effect in temporary stage and on permanent design. A total
of 6 inclinometers (3 in each phase) were installed in diaphragm
wall together with some survey points.
HORIZONTAL STRUT

5 TENDER STAGE DESIGN

RAKING STRUT

D-WALL
-16m
-20m

BERM
BARRETTE

The designers involved in the tender stage design made a fairly

conservative design with two levels temporary bracing as shown
in figure 3. Uncertainty of the performance of a thin diaphragm
wall in soft clay layer was the likely reason to adopt the conservative design in the tender stage. It is not unreasonable to adopt
the conservative design considering rheological property of soft
marine clay and likely long elapsed time of un-strutted diaphragm wall (relatively long un-strutted span of 5.6m between

SLAB

D-WALL

BORED PILE

Fig. 3 Tender stage bracing system diaphragm wall was designed to be supported by 2 struts (horizontal bracing and raker)

312

SLAB

HORIZONTAL STRUT

HORIZONTAL STRUT

SLAB

8 PREDICTION AND PERFORMANCE OF PHASE 1 DIAPHRAGM WALL

BERM
D-WALL

RAKING STRUT
BARRETTE

SHEET PILE

-16m

-14m

-16m
-20m

-20m

BORED PILE

SOUTH D-WALL vs SHEET PILE

EAST & WEST D-WALL

Fig. 4 Bracing system for Phase 1diaphragm wall

Cum. Displacement , mm.

Cum. Displacement , mm.

20

40

60

80

100

8
10

Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
with buttress
without buttress
Tender Stage

12
14
16
18
20

Depth, m.

Depth, m.

(a) East diaphragm wall

20

40

60

80

100

8
10
12
14
16
18

9 MODIFICATION OF PHASE 2 BRACING SYSTEM

Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
with buttress
without buttress
Tender Stage

20

(b) South diaphragm wall

Fig. 5 Phase 1 - Monitored diaphragm wall deflection at different stages with predicted maximum deflection profiles of different condition (with buttress, without buttress and tender stage)

Stage

East Diaphragm wall

Excavate to 2.2m and
installed temporary strut

2
3
4
5

South Diaphragm wall

Excavate to 2.2m and
installed temporary strut

Excavate to 4m

Excavate to 4m

Excavate to 6.6m with

berm

Excavate to 6.6m with

berm

Removal of berm

Removal of berm

Removal of temporary
strut after completion of
buttress

Removal of temporary
strut after completion of
buttress

Fig. 6 Phase 1 excavation work in progress with Wat Suthat in

background (Phase 2 area was being used as car parking space)

Monitoring results of the Phase 1 excavation work provided an

ample opportunity to review the design assumption, fine tune the
parameters used in the analysis of the diaphragm wall for the
Phase 2 and made modification of construction sequence. The
major modifications are : (1) Removal of soil-berm at East and
West diaphragm wall in shorter duration than that of Phase 1, and
(2) Using raking struts instead of horizontal strut for North diaphragm wall a shown in figure 7.
10 PERFORMANCE OF PHASE 2 DIAPHRAGM WALL
Figure 8 (a) depicts the measured deflection of east diaphragm
wall. As can be observed in figure 8 (a) in comparison with Figure 5 (a), the maximum deflection of east diaphragm wall in
Phase 2 is larger than that of Phase 1. The likely reasons of this
observation are;
Un-strutted elapsed time for first temporary bracing in Phase
2 was longer than that of Phase 1.
In Phase 1, horizontal struts were installed in north-south direction which temporary kingpost columns and strut were integrated in crisscross pattern with east-west struts - providing
complete-support more rigid bracing system. Whereas in
Phase 2, horizontal struts were installed only in east-west direction without having crisscross pattern with north wall
having less rigid bracing system than that of Phase 1.

Table 1. Summarized description of stages shown in figure 5

1

Figure 5 (a) shows the maximum predicted diaphragm wall

deflection at different stages of excavation monitored by inclinometer No.1 (I-1 at East wall of Phase 1) together with predicted maximum deflection profile of 3 different conditions tender stage design (2 temporary struts), modified design with
buttress and modified design without buttress. It can be observed
from figure that measured lateral movement pattern of diaphragm
wall agreed well with that of prediction for modified design with
buttress - meaning buttress-support has some influence on wall
deflection. Deflection profile of South diaphragm wall which
braced against temporary sheet pile wall at Phase 1 and 2
boundaries is presented in figure 5 (b). As can be seen in figure 5
(b), South diaphragm wall deflection is significantly higher than
that of east wall, which is likely to be caused by the fact that
South diaphragm wall is braced with more flexible sheet pile
wall. Description of stages shown in the legend of figure 5 is
summarized in table 1.

With assurance of diaphragm wall performance from monitoring results of Phase 1, original plan of using horizontal struts
for North diaphragm wall was modified by using raking struts instead of horizontal struts. As can be seen in figure 8 (b), deflection of North diaphragm wall (with raking strut support) is significantly higher than that of East diaphragm wall (with
horizontal strut support). The main reason of larger movement of
north diaphragm wall is due to the fact that it was supported only
by the berm for the long period (about 52 days) before completion of raking struts so that soil-rheology effect took place during
the long elapsed un-strutted period. Time-dependent deflection
pattern due to soil-rheology effect can be observed in North diaphragm wall as illustrated in figure 9. North diaphragm wall
moved progressively toward excavation before completion of
raking struts (at 52 days) as can be seen in figure.

313

SLAB

HORIZONTAL STRUT

COMPLETED PHASE 1
STRUCTURE

TIE BEAM CONSTRUCTED IN BAYS

RAKING STRUT
SLAB

TEMPORARY KING-POST

CAPPING BEAM

BERM

RAKER

D-WALL

BARRETTE

-16m

DIAGONAL STRUT

DIAPHRAGM WALL

D-WALL

BERM

TEMPORARY WALER

BUTTRESS

BERM

TEMPORARY
STEEL BEAM
PILE
CAP

-16m

-20m

-20m
EAST & WEST D-WALL

COM

S
TED
PLE

LAB

NORTH D-WALL with raker support

Fig. 7 Bracing system of Phase 2 diaphragm walls

Cum. Displacement , mm.
20

40

60

80

Cum. Displacement , mm.

0

100

10

Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
with buttress
without buttress

12
14
16
18

Depth, m.

Depth, m.

Fig. 10 Perspective of North diaphragm wall constructed in bays

20 40 60 80 100 120 140

11 IMPLEMENTATION OF CONTINGENCY PLAN

10

Stage 1
Stage 2
Stage 3
Stage 4
Prediction

12
14
16
18
20

20

(a) East diaphragm wall

(b) North diaphragm wall

Fig. 8 Phase 2 - Monitored diaphragm wall deflection at different stages with predicted maximum deflection profiles of different condition
Cum. Displacement , mm.

Since deflection of North diaphragm wall (Phase 2) approached

trigger levels, monitoring frequency was increased and the following contingency measures were implemented on site.
Poured 15cm thick 1m wide lean concrete on the top of the
berm along North diaphragm wall to provide bearing-effect
Installed additional king-post and diagonal struts attached to
the raking struts to provide more rigid support against diaphragm wall
Soil-berm was removed locally in bays followed by construction of wale beam, tie beam and buttress as shown in figure 10
Movement of diaphragm wall movement was observed to be
decreased by the above actions. No significant ground settlement
was observed in the vicinity of the North diaphragm wall.

20 40 60 80 100 120 140

0
2
4

12 COST AND TIME SAVING FROM VALUE ENGINEERING OPTIONS

Depth, m.

6
8
10
12
14
16
18
20

0 days
30 days
52 day (raker installed)
56 days
74 days
Prediction

Fig. 9 Phase 2 Time dependent wall deflection of North diaphragm wall diaphragm wall was supported only by soil-berm
for 52 days before completion of raker installation.
Table 2 Summarized description of stages shown in figure 8
Stage

East Diaphragm wall

North Diaphragm wall

Excavate to 2.2m and

installed temporary strut

Excavate to 4m

Excavate to 2.2m at dwall, and to 6.6m with

sloping berm
Installed raking strut

Excavate to 6.6m with

berm
Removal of berm

4
5

Removal of temporary
strut after completion of
buttress

Removal of berm

Significant cost and time saving were achieved from the value
engineering options coupled with observational method implemented for both Phase 1 and 2. The major savings were achieved
by less operation and material utilized in the following elements
of temporary works.
Cancellation of 2nd level raking struts against diaphragm wall
for both phases
Modification of bracing system using raking struts instead of
horizontal struts for North diaphragm wall in Phase 2
13 CONCLUSION
This research study reveals that a permanent diaphragm wall
coupled with effective design and construction method supplemented by observational method could offer a logistically and financially attractive solution in construction of underground car
park without disturbing the environment in the prominent historical area of Bangkok.

Removal of raker after

completion of buttress

314

Proceedings of the 15th Southeast Asian Geotechnical Conference

22-26 November 2004, Bangkok, Thailand

15th Southeast Asian

Geotechnical Conference
(15 SEAG)
Edited by:
Sambhandharaksa, S.
Bergado, D. T.
Boonyatee, T.

Stability of a Trial Trench Excavated under

Polymer Slurry in Bangkok Soft Clay
Narong Thasnanipan, Aung
Boonyarak and Zaw Zaw Aye

Win

SEAFCO Public Co., Ltd. Bangkok, Thailand

315

Maung,

Thayanan

Stability of a Trial Trench Excavated under Polymer Slurry in Bangkok Soft

Clay
Thasnanipan, N., Maung, A. W., Boonyarak, T., and Aye, Z. Z.

Seafco Public Company Limited, 26/10 Rarm Intra 109 Road, Bangchan, Khlong Sam Wah, Bangkok 10510,
Thailand.
[email protected]

Abstract: This paper presents findings on trench stability resulting from the use of polymer slurry. A trial trench 0.60x4.50x18.00m
was excavated for diaphragm wall construction in Bangkok soft clay, using polymer as an alternative to bentonite slurry. Under
consideration was its superior environmental advantages for soft clay soils typical of the region, notwithstanding a limited capacity to
suspend soil particles. The trench was excavated with a cable suspended mechanical grab, filled with polymer slurry and left open for
4 days. Field measurement included recording of slurry properties, settlement surveys, inclinometer monitoring, periodical checking of
the trench profile with Koden drilling monitoring equipment and trench depth sounding. Polymer slurry properties in the trench were
found to change sharply within 12 hours, with density, pH and viscosity gradually decreasing after 12 hours. Sedimentation occurred
within a few hours after completion of trench excavation and a 7.00m thick deposit was indicated at the bottom of trench after 10.5hrs.
The deposit was cleared with the excavation grab but sedimentation recurred. The trench profiles plotted by Koden drilling monitoring
equipment showed trench surface failure starting at depths between 7.00m and 7.50m on one of the long sides of the trench face 18
hours after trench excavation, with inclinometer readings indicating maximum soil yielding of about 6mm after 10 hours. A major
collapse of the trench surface was observed approximately 36 hours after trench completion, with ground settlement of up to 9mm.
Based on field measurement results and characteristics of the polymer used, the suitability of using polymer slurry is discussed.

1. INTRODUCTION

2. SITE CONDITIONS

Polymeric fluids have been in use as support slurries to provide

stability during excavation for bored piles and diaphragm walls.
The merits of such polymeric fluids have been evaluated and
reported by Beresford. J.J and et al in 1989. Polymeric slurries
are environmentally compactable and acceptable. The used
polymer fluid can be disposed directly into the municipal sewer
without harmful consequences to water courses. Excavated
materials can also be safely dumped in unregulated landfill sites.
As the polymer slurry offers considerable advantages in terms of
extending operation capabilities, logistics, environmental
compatibility and overall cost-saving, as well as enhancement to
frictional capacity of bored piles by providing intimate contact
between pile shaft and surrounding soil, particularly in
cohesionless soil without forming a filter cake, it has been used
in bored piling work in Bangkok and elsewhere in Thailand
(Ref). For diaphragm wall work, bentonite slurry is still used as
the main supporting slurry in Thailand. Polymeric slurry usually
purges itself of soil particles and particle settling takes place
whilst in the trench. Thus a special process to remove the
particles deposited at the bottom of trench is mandatory as a
conventional mechanical grab will not effectively remove the
sediments in such cases. For some basement excavation projects
in Bangkok subsoils with a maximum excavation depth of up to
10.0m, diaphragm wall systems constructed predominantly in
clay layers are required. Hence the use of polymer was
considered a practical option for diaphragm wall excavation in
clay. A trial trench (0.6x4.5x18m) was then excavated under
polymer slurry in soft clay at Bangchan, north-eastern Bangkok,
to evaluate the performance of the slurry.
Performance of trial trenches under bentonite slurry in soft
clay have been reported previously by Dibiagio and Myrvoll
(1972) and Thasnanipan et al (1998).

A 25m deep borehole was drilled to investigate subsoil

conditions at the trial trench location prior to trenching. The
investigation indicated that a soft clay layer extended up to
13.5m below the existing ground level, overlaying a 3.0m thick
stiff clay layer with SPT values ranging from 10 to 20.
Sensitivity of soft clay ranged from 2.3 to 3.5. Below the stiff
clay layer was very stiff clay 4.5m thick with SPT values ranging
from 17 to 27, underlain by a fine to medium sand layer at a
depth of 22.0m. SPT values obtained in the sand layer ranged
from 27 to 54.

317

Undrained Shear Strength (kN/m2)

Depth (m) Soil Description
0

20

10

40

60

80

100

120

20

30

40

50

60

Atterberg Limit
0

20

40

%
60

80

100

120

Crust

5
Soft Clay
10
M. Stiff Clay
15

20

Stiff Clay

Very Stiff Clay

M. Dense Sand
25

UCS
SPT

SPT N (Blow/ft)

Natural water content

Liquid Limit
Plastic Limit

Fig. 1. Subsoil conditions at the trial trench site (Bangchan).

Natural water content in the soft clay generally ranges from 6080%. However, at depths of 5m and 8m the natural water
content is slightly above or about 100% respectively.
Groundwater level observed in the borehole was at about 2.5m
below the ground.

3. TRIAL TRENCH

the equipment was in a range of +/- 0.2% (in this case

the accuracy was within 1-2mm).
Slurry testing this was carried out on the viscosity,
density and pH of the slurry sample taken from the
trench at a depth of 9.0m.
Ground settlement monitoring at 6 settlement points Monitoring was carried out approximately every 2 hrs
during excavation and every 6hrs after excavation.
Checking trench depth Depth sounding at 3
locations, the center and near both ends of the trench
was carried out every 30minutes after excavation.
Monitoring lateral soil movement with inclinometer
An inclinometer access tube was installed in the soil
investigation borehole to a depth of 24.0m.
Monitoring was carried out every about 2 hrs during
excavation and at 2-6hrs of elapsed time after
excavation.

The trial trench was located about 1.0m away from the
borehole. A concrete guide wall of 1.5m deep was constructed
with continuous reinforcement and cast against firm ground.
The trial trench was 0.6m wide, 4.5m long and 18.0m deep. For
field monitoring, all monitoring points, comprising 6 settlement
survey points using one inclinometer tube were located on the
opposite site of the excavation equipment to avoid disruption
caused by the excavation activities during trenching operation.
Locations of settlement survey points and inclinometer tube are
shown in Figure 2.

Trial trench 0.60 x 4.50 m

I1

P2

P3

P4

P5

P6

5. TESTING AND MONITORING RESULTS

1m
2m

2m

2m

2m

Legend
Settlement Point
P5
I1

Inclinometer

Fig. 2. Layout of field monitoring locations.

Trenching was carried out on 16th June 1999 at about 14:00hrs
under commercially available polymer slurry and completed at
about 23:30hrs on the same day. A cable suspended mechanical
grab (0.6mx2.0m) was used for trench excavation. 3 bites
excavation process was used to achieve the required trench
dimensions.
The polymer type used for trial trenching was a water-soluble
synthetic polymer supplied as a concentrated liquid. As per by
supplier recommendations, a mix ratio of 0.5kg of polymer per 1
cu.m. of slurry was used for slurry preparation. The properties
of slurry were tested for compliance values specified for
excavation prior to introducing it to the trenching process.
The slurry level in the trench was maintained up to ground
level at all times. The rate of raising the grab during excavation
was controlled to minimize the suction action which would
impact on the trench stability. The trench profile was checked
with Koden drilling monitoring equipment when excavation
reached a depth of 9.0m. The polymer slurry was also tested for
its quality. Trenching was then continued to the final depth of
18.0m. The verticality of the trench indicated by the drilling
monitoring equipment was within 1:180.
4. FIELD TESTING PROGRAM
Field monitoring and slurry testing were generally carried out
when trenching reached a depth of 9.0m and the final depth of
18.0m, and then every 6hrs after trenching was completed. The
following field testing was performed.

Checking trench profile with Koden drilling

monitoring equipment the equipment was stationed
in a fixed position on the guide wall to monitor the
changes in trench profiles. The recording accuracy of

The trial trench excavation was carried out on 16th of June 1999
at 14:00hrs and completed at 23:30hrs. Sedimentation occurred
within a few hours after completion of the trench and sediment
built up to 11.0m in depth (i.e. deposit thickness was 7.0m) after
10.5hrs. The estimated rates of settlement of soil particles
ranged from 1.0m to 1.4m per hour in thickness for the first 4hrs
after trenching and then gradually decreased. Cleaning of the
trench base was carried out 14hrs after trench completion.
However, sedimentation recurred and built up to a depth of
13.0m after 12hrs.
24hrs after base cleaning (36hrs after
excavation), sedimentation built up to a depth of 6.9m due to
trench face failure.

Time after excavation (hrs)

0

12

24

36

48

60

72

84

96

0.00

Trench Depth below ground level (m)

P1

5.00

10.00

after trench face collapsed

15.00

after trench base cleaning

20.00

Fig. 3. Plots of sedimentation with. time.

The density of polymer slurry in the trench was found to
change sharply within 12 hrs from 1.30g/ml to 1.08g/ml and then
gradually decrease. 36hrs after trench completion, this reached
1.0g/ml. Similarly, viscosity of the slurry changed from 43sec to
32sec within 12 hrs and decreased to 29sec after 36hrs. A Marsh
Funnel viscometer was used for indicative measurement of
relative changes in viscosity only, rather than to evaluate the
slurry performance. The pH value also changed from 8 to 7 after
12 hrs. After 36 hrs, the properties of the slurry remained
constant till the trench was backfilled. It is considered that after
36hrs the properties of slurry indicated that the slurry in the

318

trench turned to water, purging all polymer constituents with soil

particles.

P1

P2

P3

P5

P4

P6

Distance from trench side (m)

50

0
1

40

Settlement (mm)

30

20

Density (kN/m )

10

pH
0
0

12

24

36

48

60

72

84

96

Fig. 4. Plots of polymer slurry property changes with time.

2.00

3.00

4.00

5.00

12 Hrs
24 Hrs

36 Hrs

48 Hrs

60 Hrs

72 Hrs

96 Hrs

84 Hrs

6.00

Excavation width 0.60m

Depth below ground level (m)

Soil Displacement (mm)

2 Hrs

Fig. 6. Plots of ground settlement readings.

Plots from the inclinometer readings indicated lateral

movements of soil up to 5.8mm at a depth of 3.5m. However,
actual soil movements at the face of the trench may have been
greater than those measured by the inclinometer located 1.0m
away. A maximum ground settlement of 9.0mm was observed
2.0m away from the trench after major failure of the trench face
opposite site settlement points. No ground settlement was
observed beyond the distance of 8.0m.
1.00

10

Elapse time (hrs)

0.00

Time after
trench
excavation

Viscosity (sec)

Excavation width 0.60m

Excavation width 0.60m

Trench base -6.8 m

Trench base -11.9m

0.0

Depth below Ground Level (m)

Trench base -18.0m

5.0

0 Hr after excavation

36.0 Hrs after excavation

96.5 Hrs after excavation (end of test)

Fig. 7. Profiles of trial trench plotted by Koden drilling

monitoring equipment at different times.

10.0

Excavated to -9.0m
15.0

Excavated to -12.0m

6. DISCUSSION

Excavated to -15.0m
0 Hr after trenching
4 Hrs

20.0

6 Hrs
12 Hrs
79 Hrs

25.0

Fig. 5. Lateral soil displacements plotted from inclinometer

readings.
The trench profile plotted by the monitoring equipment
shows that on the opposite side of instrumented monitoring
points (inclinometer and settlement survey points), trench
surface failure started at depths of 7.0-7.5m about 18hrs after
trench completion. The surface failure progressively increased
and a major collapse was observed at depths between 2.0m and
8.0m after 38hrs. The trench surface failure was found to be a
wedge failure, extending more than 1.0m into the soil face.
Likewise, after 42hrs had elapsed, a localized collapse was also
observed on the instrumented monitoring side of the trench at
6.0m below ground level, increasing progressively to a depth of
about 5.0m below ground level till the monitoring was
terminated at 96hrs after excavation.

319

Excavation of the trial trench was mainly in clay layer, which is

non-permeable. Moreover, clay normally maintains a relatively
stable ground condition apart from elastic deformation due to
reduction of confining pressure.
The movement can be
controlled by hydrostatic pressure conceptually in diaphragm
wall excavation. Compared to bentonite, polymer contributes
relatively little additional density to the water in which it is
prepared, and is just as effective in holding back the plastic
deformation of the ground.
Empirical calculations were carried out for trench stability
analysis based on the available soil test data with planned trench
dimensions prior to excavation. The analysis indicated that
polymer fluid pressure with a density of 1.0 t/m3 (reduced to
density of water) was still adequate to maintain the stability of
the trench. However, high natural water content approximating
the liquid limit or exceeding moisture content of 100% in soft
clay at depths between 3.5m and 9.0m was a major concern for
trench stability. In the past, occurrence of localized failure or
trench collapse in such soil conditions under bentonite slurry was
experienced when pore pressure and/or density of slurry
significantly changed after heavy rain.

Instability of trench was not encountered in the trial trenches

which were excavated in soft clay and under-consolidated
marine deposits under bentonite slurry as reported by Dibiagio &
Myrvoll (1972) and Cowen et al (2001) respectively. The soil
properties and movements of the trial trenches, including the
trial trench reported herein are compared where possible and
presented in Tables 1 and 2 respectively.
Table 1. Comparison of Soil Properties at Trial Trench Sites
Test Location
Depth
Soil
Studenterlunden
Khlong
m
Properties
Norway
Prapa,
North
Bangkok
1.60-1.86
1.60-1.80
0-10
t/m3
4.0-8.0
1.3-2.5
Su t/m3
WC %
40-45
40-70
LL %
45-50
60-70
3
1.82-1.91
Below
10-20 t/m
2.0-5.0
14.0m is
Su t/m3
WC %
40-42
medium
LL %
40-50
dense sand
1.81-1.96
20-30 t/m3
4.0-6.0
Su t/m3
WC %
40-50
LL %
40-55

Bangchan,
North-East
Bangkok
1.37-1.80
1.6-2.0
40-105
50-85
1.6-2.0
2.0-18.7
21-71
43-84
Below
22.0m is
medium
dense sand.

Table 2. Comparison of Ground movements

Item

Studenterlunden Khlong Prapa, Bangchan,

Norway
North
North-East
Bangkok
Bangkok
1.0x5.0x28.0
0.8x2.7x14.0 0.6x4.5x18.0

Trench
Dimension in
m
Max. Ground
8mm
Settlement
Max. Soil
55
20-30
yield in mm
(after 31 days)
(after 24 hrs)
Supporting
Bentonite,
Bentonite
Fluid
Water after 32days

It is considered that high natural water content of the soil,

which approximated the liquid limit or exceeded 100% posed a
risk to trench stability and contributed to the failure of the trial
trench.
An efficient cleaning mechanism is considered necessary to
remove the soil particles settling at the bottom of the trench
under polymer slurry.
ACKNOWLEDGEMENT
The authors wish to acknowledge the contribution of EDE
Company limited for field monitoring and assistance in trial
trenching at Khlong Prapa and Bangchan sites.
RERERENCES
Beresford, J. J., Cashman, P. M., and Hollanby, R. G. 1989. The
Merits of Polymeric Fluids as Support Slurries.
Proceedings of the International Conference on Piling and
Deep Foundations, London, 15-18 May 1989. pp 3-10.
Cowan, S., Littlechild, B. and Green, E. 2001. Trial Diaphragm
Wall Panels in Under-Consolidated Marine Deposits. HKIE
Annual Seminar Geotechnical Deformations and
Movements 2001. pp. 179-192.
Dibiagio, E. and Myrvoll, F. 1972. Full Scale Field Test of a
Slurry Trench Excavation in Soft Clay. Proceedings of 5th
ECSMFE, vol.1, pp. 461-471. Madrid.
Thasnanipan, N., Maung, A. W. and Tanseng, P. 1998. Barrettes
Founded in Bangkok Subsoils, Construction and
Performance. Thirteenth Southeast Asian Geotechnical
Conference, 16-20 November 1998, Taipei, Taiwan: pp
573-578.

9mm
6
(after 10hrs)
Polymer

Most of the monitoring points for Bangchan trial trench were

located on one side of the long trial trench face. The trench
profile as shown by drilling monitoring equipment indicated that
the collapse of the trench face occurred initially at the side
opposite the monitoring points. It is considered that the recorded
maximum values of results may not have represented the actual
maximum values of soil movements occurring around the trench,
particularly on the side of the trench without monitoring
instrumentation due to lack of monitoring data.
7. CONCLUSIONS
Based on the field monitoring results obtained from the trial
trench excavated under polymer in Bangkok soft clay, the
following conclusions were made.
The trial trench was found stable up to 18hrs after completion
of excavation under polymer fluid at this site and after that the
stability of trench deteriorated with time.

320

Proceedings of International Symposium on Underground Excavation

and Tunneling
2-4 February 2006, Bangkok, Thailand

Urban Tunnel Construction

for Protection of
Environment
Edited by:
Phienwej, N.

Application of Top-down Construction Method

for Deep Excavations in Bangkok
Thasnanipan N., Aye Z. Z., Boonyarak T. and Submaneewong
C.
SEAFCO Public Co., Ltd. Bangkok, Thailand

321

APPLICATION OF TOP-DOWN CONSTRUCTION METHOD FOR DEEP

EXCAVATIONS IN BANGKOK
Narong Thasnanipan, Zaw Zaw Aye, Chanchai Submaneewong, Thayanan Boonyarak
Seafco Public Company Limited, Bangkok, Thailand
ABSTRACT
Top-down construction method was used for construction of deep underground
structures of major projects in Bangkok, including 18 subway stations of the first Bangkok
MRTA. This paper presents authors experiences in the practical installation of the different
stanchion types in large diameter deep-seated bored piles and barrettes in Bangkok. The
problems encountered during the actual installation of the stanchions and solutions to them
are also discussed. Application of top-down method in recently completed deep basement
structure is also reported highlighting the benefits offered by the method in comparison with
conventional bottom-up method. This paper demonstrates that the top-down method is the
most appropriate solution in construction of the deep underground structures for protecting
the modern urban environment.
1. INTRODUCTION
Top-down construction method as the name implies, is a construction method, which
builds the permanent structure members of the basement along with the excavation from the
top to the bottom. Top-down method is mainly used for two types of urban structures,
underground structures such as car parks, underpasses and subway stations and tall buildings
with deep basements. Top-down method is the most appropriate solution in minimizing the
duration and disruption to surface traffic and other urban activities, for the construction of the
subway stations, underground car parks and underpasses where the structure is located
directly underneath the existing roads. For tall buildings with deep basements, a full
application of top-down method enables superstructure to build concurrently with excavation
and construction of the basement giving a significant advantage in reducing the overall
construction time.
2. TYPICAL SUBSOIL PROFILE OF BANGKOK
Bangkok subsoil is featured by alternating clay and sand layers of thick Quaternary
deposits. As characteristics and properties of Bangkok subsoil layers have been well
documented in many literatures, only a general description of Bangkok subsoil is presented in
this article as follows.
Top layer consists predominantly of Fill-Materials, Clayey Sand or Silty Clay with
some cement block rubble and rock fragments, is commonly found up to 2 to 4m depth. Soft
to very soft, highly compressible dark gray marine clay lies beneath top layer and in some
areas it lies under weathered crust layers of about 2m thick. Depending on the location, Soft
Clay layer is extended up to 12-18m. About 2m thick Medium Clay layer can be observed
between Soft Clay and underlying Stiff Clay. Generally Stiff Clay layer occurs directly
underneath Medium Clay and its depth goes up to 22m. Below Stiff Clay layer, First Sand
layer 5-8m in thickness can be found. This First Sand layer, however, is absent in some areas.
Stiff to Hard Clay layer underlies First Sand and it is found to be about 5m thick. Second

323

Sand layer generally occurs at depths between 45 to 65m. Typical subsoil profile of Bangkok
area is shown in Figure 1.

Pore Pressure (kN/m 2)

Geotechnical
Parameters
0

200

400

600

Su= 10-25 KPa

t = 14-16kN/m 3
-

10

40

Hard Clay

50

Dense to Very
Dense Sand
(Second Sand)

30
SPT-N= 20-40
t = 20kN/m 3

60

50

Line
tatic
e
Lin
wn
wdo
Dra

SPT-N= 50-100
t = 21KN/m 3

40

tric

Su> 200 KPa

= 21KN/m 3
t

ros
H yd

Medium
Dense Sand
(First Sand)

me

30

Su= 50-14020- KPa

t = 17-21kN/m 3

zo
Pie

Medium to
Stiff Clay

ual

20

Act

Depth (m)

10

Bangkok Soft
Clay

0
0

Weathered Crust

60

Figure 1. Typical subsoil profile of Bangkok piezometric drawdown condition.

3. BORED PILE AND BARRETTES CONSTRUCTION METHOD
Wet process or slurry displacement method is always used for construction of deep-seated
large bored piles and barrettes in Bangkok. For bored piles, temporary casing of 14-15m in
length is used as a support in soft clay layer. Soil inside the casing is normally excavated by
auger applying rotary drilling method and drilling is continued with the bucket under
polymer-based slurry from the top of sand layer to the final depth. A mechanical cable-hunggrab or hydraulic grab mounted on crawler crane is used for excavation of trench with slurry
support in barrette construction. Guide walls having a minimum depth of 1.2m are commonly
used for initial guiding of grab excavation at the top of the trench. Different from bored pile,
bentonite slurry is commonly used for barrette excavation in Bangkok. Tremie concreting is
necessary for casting both bored piles and barrettes.
Stanchions are mainly installed in the large diameter deep-seated bored piles and barrettes
founded in first or second sand layer. Bored piles of diameter 1.50m to 1.8m and rectangularshape barrettes of 0.80x2.8m to 1.5x3.0m (thickness x width) founding at depth between 50m
and 60m are commonly used to accommodate the stanchions.

324

4. ADVANTAGES OF TOP-DOWN CONSTRUCTION METHOD

Application of top-down method offers three main advantages as outlined below.

It allows early commencement of super-structure construction without the need to

wait for excavation reaching the bottom level. Overall construction duration can be
significantly less than that of conventional method thereby saving the cost.

It does not require temporary bracing system so that time requirement and cost for
temporary works (both material and labor costs) are eliminated, in turn significantly
saving cost.

Minimize soil movement induced by excavation works which is an important factor

for some sensitive locations.
5. STRUCTURAL MEMBERS REQUIRED FOR TOP-DOWN CONSTRUCTION
Design and construction principles for top-down method primarily call for two major
structural elements.

Columns with sufficient capacity must be pre-founded in bored piles or barrettes to

sustain the construction load and to utilize as part of bracing system.

Excavation for basement must be carried out with the support of permanent
retaining wall so that basement floor slabs can be utilized as lateral bracing.
Diaphragm wall of 0.8m to 1.2m in thickness with sufficient embedment in firm soil
layers is commonly used as a retaining wall whereas prefabricated steel columns known as
stanchions embedded in either large diameter deep-seated bored piles or barrettes are utilized
as structural columns. Figure 2 illustrates the top-down construction method with utilization
of stanchions and diaphragm wall for underground subway station.

Figure 2. Top-down construction with stanchion and diaphragm wall

325

6. TYPES OF STANCHION AND THEIR APPLICATION IN BANGKOK

Pre-fabricated H section steel columns and steel built-up-section columns are mainly
used as pre-founded structural columns or stanchions. Pre-cast reinforce concrete columns are
very seldom used (Manoharn S. & Aye Z. Z. , 1994). Commonly used stanchion types and
their application in Thailand can be viewed in published paper of Thasnanipan et. al (2000).
7. ALLOWABLE TOLERANCE FOR POSITION OF STANCHIONS
Allowable tolerance for stanchions is usually called by the designer and it is mainly
governed by the structural tolerance of the steel as well as required position of the finished
columns. In most projects, allowable vertical and horizontal deviation of stanchion are
specified as 25 to 50mm whereas verticality is required between 1:200 and 1:400. There are a
number of constraints to get highly accurate position of heavy and lengthy stanchions which
have to be installed in deep-seated foundation piles constructed by wet process in Bangkok
subsoil as described in the following sections.
8. STANCHION INSTALLATION METHODS
Stanchion installation method is usually selected by the piling contractor who take into
consideration three main factors such as installation depth, size of stanchion and size of bored
or barrette piles. Though installation details may be different from one contractor to another,
stanchion installation can be categorized under two main methods, pre-concreting installation
or placing stanchion prior to concreting or post-concreting or plunging installation.
8.1 Pre-concreting or pre-placing installation method
This method was applied in Thailand Cultural Center Station of the M.R.T Chaloem
Ratchamongkhon Line. Stanchion was installed prior to tremie concreting, immediately after
completion of barrette excavation and reinforcement lowering. The main reason for selection
of this method in the mentioned project was due to the relatively large size of the barrettes. In
some projects stanchion is attached to the last section of reinforcement and installed together
with the reinforcement. General construction steps involved in this method are demonstrated
in Figure 3.
Concrete

Guide frame

Stanchion

Stanchion

Stanchion

Slurry

Slurry

Tremie pipe

Temporary
casing

Temporary
casing

Temporary
casing

Pile cut-off
level

Pile cut-off
level

Pile cut-off
level

Pile cut-off
level

concrete

concrete

Complete drilling and

reinforcement installation

Stanchion installation

Concreting
(1 tremie for bored pile and
2 tremie for barrette)

Backfill

Complete casing extraction

and backfilling

Figure 3. General construction sequence of pre-concreting installation method

326

Specially designed guide frame used in installation of H-section steel column of

419x407x390kg/m, 24.7m long, inserted into barrettes applying pre-concreting method in
Bangkok MRT project is shown in Figure 4. Effectiveness of guide frame plays one of the
important roles in achieving positional accuracy of stanchion.

Figure 4 Guide frame used for pre-concreting installation in Thailand Cultural Center Station
of the M.R.T Chaloem Ratchamongkhon Line
Figure 5 shows the installation of a 24m long built-up steel stanchion into the pile
borehole prior to concreting.

Figure 5 Stanchion length 24m attached to the top portion of reinforcement cage installed into
borehole prior to concreting

327

7. 2 Post-concreting installation or plunging method

In this method, stanchion is installed immediately after completion of bored pile
concreting process. General construction sequence involved in this method is demonstrated in
Figure 6. Guide frame is used to install the stanchion at the correct position.
Concrete
Guide frame

Tremie pipe

Stanchion

Stanchion

Slurry

Slurry

Slurry

Backfill

Temporary
casing

Temporary
casing

Temporary
casing

Pile cut-off
level

Pile cut-off
level

Pile cut-off
level

Concrete

Complete drilling and

reinforcement installation

Concreting by tremie
method

Installing stanchion by
plunging into concreted pile

Concrete

Complete casing extraction

and backfilling

Figure 6. General construction sequence of pre-concreting installation method

REVIEW OF THE STANCHION INSTALLATION METHODS USED IN OTHER PARTS
OF THE WORLD
The stanchion installation methods used in other parts of the world have been presented
in some published papers. Among these, Findlay (1989) reviewed a number of stanchion
installation methods used in construction of large diameter bored piles for top-down
construction particularly in UK. The author reported that steel columns can be placed with
better accuracy by dry process bored piling method than wet process method (with support
fluid).
Table 1. Summary of stanchion installation method from published literatures
Reported by
Location
Stanchion installation method
Hollingsworth (1991)
London, UK
Post-concreting
Ressi di Cervia and Tamaro Boston, USA
Pre-concreting
(1991)
Crawley and Stones (1996)
London, UK
Pre-concreting
Arz (1989)
Germany
Pre-concreting
From the available literatures it is noted that pre-concreting method is more common in
use than post-concreting method.
PROBLEMS ENCOUNTERED IN POST-CONCRETING INSTALLATION METHOD
The major problem commonly encountered in this installation method is inability to
install the stanchion at the design position due to one or combination of the following causes.

Inclination of the borehole / trench

328

Stanchion can not be inserted up to the required depth as concrete becomes hard due
to the premature setting

During lowering, stanchion is stuck by reinforcement cage and due to the hardening
of the concrete, extraction of stanchion becomes impossible for reinstallation
As the installation of the stanchion is often associated with many unforeseen problems,
it is likely in many cases that concrete becomes stiff or prematurely set during the installation.
In Bangkok, premature setting of concrete is usually found to be attributed by;

long delivery time due to traffic conjunction

inappropriate mix

severe weather

disruption in concreting / equipment breakdown

PROBLEMS ENCOUNTERED IN PRE-CONCRETING INSTALLATION METHOD

Using this method can minimize some problems encountered during installation
provided that adequate clearance is available between pile reinforcement and the stanchion for
tremie pipes lowering. Inclination of borehole is also important for this installation method.
Proper planning in concrete ordering is required to avoid unnecessary waiting of concrete
trucks for stanchion placing which may affect the total batching time limitation. As stanchion
has to be hung in the position until completion of concrete placement, appropriate and
sufficient capacity of the hanging and positioning device should be provided to avoid falling
and deviation of stanchion in the borehole. Temporary casing should not be used as a hanging
system without additional support for the large stanchion.
FACTORS EFFECTING THE POSITIONAL ACCURACY OF THE STANCHION
It is hardly possible to install a stanchion so that its position and verticality is always as
designed. As-built position of stanchion is found to be generally influenced by the installation
method. Skill and experience of the contractor plays a major role in achieving positional
accuracy of the stanchion provided that the installation is practical with the design elements
such as size of foundation piles in relation to size of stanchion. The factors affecting the
positional accuracy for two different methods were reported in details by Thasnanipan et. al
(2000).

Figure 7. View of stanchions embedded in bored piles at final excavation level

329

STANCHION POSITION EFFECTED BY EXCAVATION

In Bangkok, post-installation movement (horizontal movement) of the stanchion is
frequently encountered and it is mainly caused by excavation induced soil displacement
particularly at initial stage of excavation where diaphragm wall deflection is characterized by
large cantilever rotation within Soft Clay layer. In some projects, deviated stanchions had to
be pushed or jacked back to the original position and restrained by means of temporary
support until permanent slab was cast.
APPLICATION OF TOP-DOWN METHOD IN CONGESTIVE URBAN AREA
Top-down method was successfully applied in Bangkok for a fifteen-storey commercial
building with five-level underground car park located in an old established commercial area
of Bangkok. Construction site is bounded by rows of old shop-house buildings estimated to
be supported by short piles. Hence, ground movement induced by large deflection of retaining
wall was unfavorable. Limited space availability on site and congestive neighborhood also
posed another major constraint on construction logistic. By the local traffic regulation,
concrete trucks and other heavy vehicles entering to the site were restricted during rush-hours
of both morning and afternoon. These factors called for the need of careful consideration in
establishing the design principles and sequence of construction. A numbers of construction
sequences were considered. After having detailed review and conducting a series of analyses
with different conditions, decision was made to apply the top-down construction method.
The main objectives were ;
to minimize the material and construction sequence involved in temporary works so as
to accelerate the construction time
to minimize the logistic problem imposed on site due to limited space in congestive
neighborhood
to minimize soil movement induced by excavation works which is important factor for
the project adjacent to the old buildings of weak foundation
DIAPHRAGM WALL, BARRETTES AND BORED PILE CONSTRUCTION
Diaphragm wall having 800mm width founded at 24m below ground level (B.G.L) was
constructed simultaneously with wet-processed bored piles of diameter ranges from 800mm
to 1800mm with toe depth 60m B.G.L. Barrettes having same toe depth as bored piles were
constructed along with diaphragm wall panels. Depending on the loading conditions, different
sizes of permanent steel stanchions were installed by post-concreting method. Table 2 shows
the detailed of foundation elements constructed in the project. Figure 8 shows layout of
diaphragm wall and bored pile with stanchion. Figure 9 shows the aerial view of excavation
works.

330

Table 2 Detail of foundation elements in each project

Type and Size of
Allowable Pile Permanent Steel Stanchion Size
Foundation
Capacity (ton)
2 No. H-350x350mm
Barrette 0.80 x 3.00m
1750
H-400x400mm + 2 No. H-200x200mm
Bored pile dia. 1.80m
1600
Bored pile dia. 1.80m
Bored pile dia. 1.80m
Bored pile dia. 1.65m
Bored pile dia. 1.50m
Bored pile dia. 1.20m
Bored pile dia. 1.80m
Bored pile dia. 1.50m
Bored pile dia. 0.80m

1600
1600
1450
1250
950
1600
1450
500

+ Steel Plate 80mm

H-400x400mm + 2 No. H-200x200mm
H-400x400mm
H-400x400mm
H-400x400mm
H-400x400mm

No stanchion
No stanchion
No stanchion

Total
Number
6
3
15
17
2
17
8
3
9
16

Boundary Line

Bored Pile with Stanchion

D-Wall

Figure 8. Layout of bored pile with stanchion and diaphragm wall

Figure 9. Arial view of the top-down construction in progress concrete batching plant was
set up on site to minimize the concrete supply interruption in congestive area with severe

331

COMPARISON BETWEEN TOP-DOWN AND BOTTOM-UP METHOD

Master Construction Schedule of the major activities of top-down and bottom-up method is
illustrated in a simple bar-chart form as shown in the figure below. Apparently, top-down
construction method offered significant time saving as can be seen in the figure. This
schedule comparison is based on the construction sequences by top-down and bottom-up
method as demonstrated in Figure 10. Figure 11 shows construction progress after excavation
reached to final level.
Method

Major Activity

Top-down

D-wall and piling

Sub-structure
Super-structure

Bottom-up D-wall and piling

Sub-structure
Super-structure

Duration
(month)
4.5
10
12

Construction Period (Month)

12
16
20
24

32

36

4.5
16
13

Figure 10 Master construction schedule of major construction activities of top-down and

bottom-up method
Apart from significant time saving, the cost associated with temporary works was
significantly minimized by using top-down method. If bottom-up method was utilized, it
would have required, 3 layers temporary bracing as shown in Figure 12. A total weight of
steel H-beam type temporary bracings required for 3 layers was estimated to be 1,070 ton.
Significant amount of scaffolding usage was also minimized by applying top-down method.

Figure 11 View of construction progress after excavation reached to final level

332

Floor level

Floor level
1st Bracing

Soft Clay

B1

-3.92

Soft Clay

B1

-3.92

2nd Bracing

B2

-8.20

B3

-11.29

B4

-14.30

B5

-17.39

Stiff Clay

Hard Clay

-19.10

-24.00

Soft Clay 2

Medium Stiff Clay

B2

-8.20

B3

-11.29

rd

3 Bracing

B4

-14.30

B5

-17.39

Stiff Clay

Hard Clay

Top-down construction

D-wall 0.8 m. thk.

Medium Stiff Clay

D-wall 0.8 m. thk.

Soft Clay 2

-19.10

-24.00

Bottom-up construction

Top-Down Method
Stage
Construction Sequence
Stage
1
1
Excavate to -1.7m and install
temporary bracing at -1.5m
2
2
Excavate to -4.5m and construct B1
Slab
3
3
Remove 1st bracing at -1.5m
Excavate to -11.8m and construct B3
4
Slab
4
Excavate to -19.1m and construct
5
Mat Foundation
5
Construct B4, B2 and Ground Floor
6
6
Slab
7
Construct above floors
8
7
9
10
11
12
13

Bottom-up Method
Construction Sequence
Excavate to -1.7m and install first
level temporary bracing at -1.5m
Excavate to -7.2 m and install 2nd
level temporary bracing at -7.0m
Excavate to -13.0m and install 3rd
level temporary bracing at -12.8.0m
Excavate to -19.1 m and construct
mat foundation
Construct B4 Slab and columns
Remove 3rd level bracing
Construct B3 Slab and columns
Construct B2 Slab and columns
Remove 2nd level bracing
Construct B1 Slab and columns
Remove 1st level bracing
Construct Ground Floor Slab
Construct above floors

Figure 12 Summarized construction sequence of top-down and bottom-up method

333

COMPARISON ON DIAPHRGM WALL BEHAVIOR

Behavior of diaphragm wall with two different construction methods was studied. Figure 13
depicts the results of bending moment, shear force and displacement of diaphragm wall
compared for top-down and bottom-up construction methods. Followings were concluded
from the comparison.
Reinforcement required for diaphragm wall with top-down method would be less than
that of bottom-up method which in turn offered cost saving on material and labour
Diaphragm wall deflection for top-down method was significantly less than that of
bottom-up method which would minimize risk of damage to adjacent existing
buildings

Predicted and measured diaphragm wall deflection is shown in Figure 14 As can be seen in
the figure, actual measured deflection values are less than those of prediction. The shape of
the deflection curves of prediction and actual measurement at each construction stage is fairly
identical which suggests that design parameters and assumption used in analysis are of
applicable.
Shear force (t/m)

Bending moment (t-m / m)

0

-200 100
0

-5

-5

-5

-10

-10

-10

-100

-20
-25
-30

-50

-100

-15
-20

Top-down
Bottom-up

20

40

60

Depth (m)

-15

100

Depth (m)

Depth (m)

200

50

Cumulative Displacement (mm)

-15
-20

-25
Top-down
Bottom-up

-30

-25
-30

Figure 13 Bending moment, shear force and displacement of diaphragm wall

comparison for top-down and bottom-up construction methods

334

Top-down
Bottom-up

80

Cumulative Displacement , mm.

0

10

20

30

40

Construction Stage
1. Excavate to -1.70 m
2. Install 1st bracing and Exc. to
-4.30 m
3. Construct B1 and remove
1st bracing
4. Excavate to -11.80 m and
construct B3
5. Excavate to -19.10 m

1-P

5
3-A

1-A

3-P

2-A

2-P

10
Depth, m.

4-A

4-P

5-A

15

20

25

30

A-Actual Movement
P-Predicted movement

Figure 14 Comparison between predicted and actual movement of diaphragm wall in

each stage of construction
CONCLUSION
Application of top-down construction method in Bangkok subsoil has been presented
with particular emphasis on the two different stanchion installation methods. According to the
authors experience, pre-concreting installation method provided fewer problems in practical
installation and achieved better positional accuracy. The designers such as structural
engineers and architects should be aware of the accuracy achievable by practical installation
method and take it into consideration at the design stage. Stanchion installation contractor
should select the appropriate method, equipment as well as experienced personnel plus a wellformed plan with the consideration of all potential problems to achieve the successful
construction of foundation structure, which is of primary importance for top-down technique.
Practical application of top-down method in recently completed project is presented
highlighting the key advantages over bottom-up method. Significant cost saving from shorter
construction period and minimum usage of material for temporary works offered by top-down
construction method was remarkable achievement. In short, top-down method is the most
appropriate solution in construction of the deep underground structures for protecting the
modern urban environment.

335

REFERENCES
Arz P. 1989. Foundations of the Ko-Galerie in Dusseldorf Using Top-down Building
Technologies, Proceedings of the International Conference on Piling and Deep
Foundations, London, 15-18 May 1989 : 185-192. Rotterdam: Balkema.
Crawley J.D. & Stones C. S. 1996. Westminister Station Deep Foundation and Top-Down
Construction in Central London, Geotechnical Aspects of Underground Construction in
Soft Ground, London : 93-97. Rotterdam : Balkema.
Findlay J. D., Wren G.E. 1989. Review of the Methods Used to Construct Large Diameter
Bored Piles for Top Down Construction. Proceedings of the International Conference
on Piling and Deep Foundations, London, 15-18 May 1989 : 199-210. Rotterdam:
Balkema.
Finsted J. A. 1991. Royal Christiana Hotel : Basement with Permanent Sheet Pile Wall, Upand-down Method, 4th International DFI Conference : 387-392. Rotterdam: Balkema.
Hollingsworth J.R. 1991.
Diaphragm Walls, Load Bearing Piles and Piled Soil
Reinforcement for a Deep Top-down Basement Construction, 4th International DFI
Conference : 79-84. Rotterdam : Balkema.
Manoharn S. & Aye Z. Z. (1994) , Bored Pile with King-post and Secant Pile Wall
Construction for Top-Down Method Central Sukhothai Hotel Haddyai, Latest
Technology in Construction Technology Conference, Bangkok, Thailand
Thasnanipan N., Maung A. W. & Z. Z. Aye, (2000) Practical Installation of Stanchions for
Top-Down Construction in Bangkok Subsoil, Year 2000 Geotechnics, Geotechnical
Engineering Conference, The Imperial Queens Park Hotel, Bangkok, Thailand, 27-30
November 2000
Yong K.S., Tsen C.N. & Tang S.K. 1998. Performance of a Deep Excavation with Top-down
Construction, 13th Southeast Asian Geotechnical Conference, 16-20 November 1998.
Taipei, Taiwan ROC : 587-593

336

Proceedings of International Symposium on Underground Excavation

and Tunneling
2-4 February 2006, Bangkok, Thailand

Urban Tunnel Construction

for Protection of
Environment
Edited by:
Noppadol Phienwej

Construction of Diaphragm Wall for Basement

Excavation Adjacent to Tunnels in Bangkok
Subsoil
Narong Thasnanipan, Aung Win Maung, Zaw Zaw Aye,
Chanchai Submaneewong and Thayanan Boonyarak
SEAFCO Public Co., Ltd. Bangkok, Thailand

337

International Symposium on Underground Excavation and Tunnelling

2-4 February 2006, Bangkok, Thailand

Construction of Diaphragm Wall for Basement Excavation

Adjacent to Tunnels in Bangkok Subsoil
Narong Thansnanipan1, Aung Win Maung1, Zaw Zaw Aye1, Chanchai Submaneewong1
Thayanan Boonyarak1
1

Seafco Public Company Limited, Thailand

ABSTRACT
In July 2005, an earth retaining structure with diaphragm wall was completed in the vicinity of
Bangkoks first subway system, the MRTA Chaloem Ratchamonkhon Line. Diaphragm wall design
analysis and ground movements associated with diaphragm wall construction were reviewed. The
review was carried out in accordance with technical requirements for engineering works within the
specified protection zone of MRTA structures. Preliminary diaphragm wall analyses indicated that
lateral movements of the wall should not exceed 26.54mm. A 6.0mm maximum movement allowance
is specified for MRTA structures. The construction work was undertaken in compliance with
technical requirements for engineering works within the specified protection zone of MRTA
structures. Two inclinometer tubes with 20.0m in length were installed in the ground between the
planned diaphragm wall and the south-bound tunnel. Surface settlements markers were also
established in line with the inclinometer tube location on the pavement. Ground movements were
monitored during diaphragm wall panel construction. Based on the monitoring results, the estimated
maximum ground movement at the tunnel location due to wall panel construction was less than
0.5mm, indicating no impact on the tunnels by diaphragm wall construction.
1. INTRODUCTION
Deep excavation work adjacent to existing structures have become a common construction activity in
most major cities as utilization of underground spaces, including underground car parks, basements,
mass transit subway stations and tunnels increase. With mass transit systems in place, properties
located along the commuter lines are considered the most favorable areas to be developed with
maximum utilization of areas available. In turn, deep basement construction adjacent to the existing
subway tunnels is also often necessary.
At present in Bangkok, such deep excavation work in the vicinity of the subway tunnels is
considered relatively uncommon compared to major cities in developed countries. It is anticipated that
this phenomenon will change in the near future. This paper presents an unprecedented case in
Thailand of deep excavation adjacent to a subway tunnel. The proposed hotel tower reported herein
was planned to be constructed with 2 basements using cast in-situ diaphragm wall located 7.8m to
9.6m away from the south-bound tunnel of the Mass Rapid Transit Authority, MRTA Chaloem
Ratchamonkhon Line. The main foundation for the superstructure consists of 66 bored piles (1.5m
diameter) and 7 barrettes (1.0x3.0m) embedded to depths of about 57.0m in the second sand layer of
Bangkok Suboil. For construction of basements and foundation mat of 3.5m in thickness, diaphragm
walls (0.8mx20m deep) with 2 levels of temporary bracing are planned as a retaining system for a
general excavation to -10.50m. The perimeter diaphragm wall is supported by 24 bored pile of 0.8m
in diameter. Due to the presence of the MRTA tunnels in the close vicinity of the project, a review

339

371

was carried out on the expected performance of the retaining system, with regards to technical
requirements for the engineering works within the protection zone of the MRTA.
In this paper, predicted diaphragm wall deflections and associated expected tunnel movements
are presented. Pre-established geotechnical instrumentation program and recommended contingency
measures are also described. Monitoring results of ground movement during different stages of
diaphragm wall panel construction is also reported.
2. SITE CONDITIONS
The project site is located on the eastern side of Asoke Road, Sukhumvit, Bangkok, Thailand. On the
eastern side are two one storey houses, 3 storey office building and open space car parking. The site is
flanked by a 12 storey office building on the northern boundary and a 5 storey residential building on
the southern boundary.
The MRTA twin tunnels with an outside diameter of 6.30m are located in front of the site about
15.0m below Asoke Road. Sound bound tunnel is adjacent to the site being 9.623m to 7.815m away
from the planned diaphragm wall. The layout plan of the site is presented in Figure 1.

Figure 1. Layout plan of the project

3. SUBSOIL CONDITIONS
A site investigation was carried out with 3 boreholes drilled at the project site. The borehole
information indicated that a soft clay layer about 10.0m in thickness occurred below 2.0m thick fill. A
medium stiff clay layer with Su 2.9-3.9t/m2 was encountered at a depth of 12.0m, overlaying a stiff
clay layer at a depth of 15.0m. A 2.0-2.5m thick sand layer was intercepted at a depth of 19.5m
overlying very stiff clay layers. Two dense sand layers with thickness of about 7.5m and 12.0m

340
372

occurred at depths of about 42.5m and 54.0m respectively in hard silty clay layers. A summary of
subsoil conditions obtained from borehole data is presented in Table 1.1.
Table 1. Summary of Subsoil Conditions
Depth below
Description of Material
Ground (m)
0-2.0
Pavement/top soil
2.0-3.0
Soft Clay 1
3.0-12.0
Soft Clay 2
12.0-15.0
Medium Stiff Clay
15.0-17.0
Stiff Clay
17.0-19.5
Stiff to Very Stiff Clay
19.5-22.0
Medium Dense Sand
22.0-35.0
Very Stiff Clay
35.0-42.5
Stiff Clay
42.5-50.0
Dense Sand
50.0-54.0
Hard Clay/ Silty Clay
54.0-66.0
Dense to Very Dense Sand
66.0-80.0*
Hard Silty Clay
Note:* at the end of borehole

Thickness
(m)
2.0
1.0
9.0
3.0
2.0
5.0
2.5
13.0
7.5
7.5
4.0
12.0

Su (UC)
(t/m2)
2.0
1.2
3.5
5.0
12.0
20.0
12.0
29.9
-

SPT N
(blow/ft)
-

18
20
30
18
>52
45
>60
>45

4. TECHNICAL REQUIREMENTS FOR ENGINEERING WORKS WITHIN PROTECTION

ZONE OF MRTA
Over 6 to 10 years, deep excavations with braced diaphragm walls have been carried out in Soft
Bangkok Clay for multi-level basements, underground car parks, subway stations and deep shafts for
ventilation and launching TBMs. After completion of the subway system, protection zones for the
MRTA subway structures were established and technical requirements for future works planned within
the protection zone were also issued. There are two protection zones, named A and B. Protection
zone A is a zone located immediately above the MRTA structure, plus 3.0m clearance distance
measured laterally from the outer faces of MRTA structure. Zone B is located outside the 3.0m
clearance distance from the MRTA structure, but within the influence zone based on a line drawn from
a point at the center level of MRTA tunnel on the 3m clearance boundary line at 1:1 slope to the
ground for MRTA structures with base level no deeper than 20.0m depth. With the base level of
MRTA tunnel located deeper than 21.0m below the ground, the influence zone below 21.0m is
determined using a line with a slope 2:1 vertical to horizontal ratio from the clearance boundary line at
the center level of the structure. Figure 2 illustrates the MRTA protection zones and Figure 3 shows a
typical section of the project reported herein together with the protection zone.
The technical requirements generally cover construction method and technique allowed and
approval, surcharge load limits, clearance distance to the structure and movement limits induced by the
future works.
The most critical requirement for working adjacent to the MRTA structure is the differential
movement resulting from the work shall not produce final distortion in the track or its plinth in excess
of 3mm in 6.0m (1:2000), or a total movement in the MRTA structure or track not exceeding 6mm in
any plan.
Due to the above mentioned requirements, the diaphragm wall design analysis was reviewed with
finite element computer modeling to predict the movement of the tunnels caused by basement
excavation work. Subsequently, impact assessment was carried out and necessary precautions and
safety measures were outlined.
An assessment of works associated with diaphragm wall and bored piling for the project to meet
the technical requirements for engineering works within the protection zone is summarized in Table 6.

341
373

Figure 2. Configuration of MRTA Protection Zone

Figure 3. A Typical Section of Project

5. DIAPHRAGM WALL ANALYSIS AND DESIGN
Diaphragm wall analysis and design was primarily carried out using the WALLAP computer program.
Stability of wall, bending stresses and wall deflections were computed in relation to planned basement
excavation sequence. Due to the presence of the MRTA subway tunnels about 7.8m away from the
planned excavation, 2 dimensional analyses using the PLAXIS program was carried out to compute
the response/displacement of the tunnels and ground settlements caused by the excavation works.

342
374

Based on the primary analysis with WALLAP, 0.8m thick diaphragm wall embedded to a depth of
20.0m with two levels of temporary bracing at -1.50m and 6.20m was required.
WALLAP has been used by the authors for analysis of braced excavations with diaphragm walls
and sheet piled walls in Bangkok subsoils and elsewhere. More than 12 projects have been completed
by the authors. Additionally 9 underground stations and a number of ventilation shafts of MRTA
southern part has been completed with diaphragm walls designed by using the program (Schulz C.
2001).
Prior to determination of tunnel movements using PLAXIS, the same basic model (without
foundation piles and tunnel) was used in analysis using both WALLAP and PLAXIS programs for a
comparison between analysis results to establish a correlation for further analysis using PLAXIS
program. The comparison of analyses indicated that the predicted maximum wall deflection computed
by PLAXIS is about 39.0mm which is close to the 43.0mm computed by WALLAP. Figure 4 shows
results of analysis from both WALLAP and PLAXIS programs.

Shear force (t/m)

Bending moment (t-m / m)

50

-50

-100

-150

60

40

20

-20

0.00

-40

-5

-5

-5

-10

-10

-10

-15

Depth (m)

Depth (m)

Depth (m)

100

Displacement (m)

-15

-20

-20

Wallap

Plaxis

0.04

0.06

0.08

0.10

-15

-20

-25

-25

0.02

Wallap

Plaxis

-25

Wallap

Plaxis

Figure 4. Comparison of Diaphragm Wall Analysis Results from WALLAP and PLAXIS
5.1 Wall Deflection and Ground Movements
The characteristics of diaphragm wall movements in the completed projects in Bangkok have been
reviewed by Phienwej et al (1998). It was reported that in terms of the ratio of hmax /H, the typical
range of the maximum wall movement hmax fell within 0.3-0.8% of excavation depth H.
For the planned project reported herein, primarily predicted maximum wall movement of
34.00mm, about 0.31% of the excavation depth (11.0m) by WALLAP was considered in the
reasonable range and thus control of the wall deflection within the predicted range could be practically
achievable, with adequate lateral supports and proper construction sequence and time controls if there
is no existing subway tunnel nearby.
However, for calculation of lateral and vertical displacements of the ground, including those of
the existing MRTA tunnels, a two dimensional simulation is necessary and thus PLAXIS program was
selected. A number of computer simulations were carried out using the PLAXIS program to meet the
technical requirements, particularly tunnel displacement criteria.

343
375

1
B 1 11
1 st Temp. Bracing

B2

Note:

Figure 5. Computer modeling with PLAXIS

Excavate to -6.5m

Mat/Base Slab

10

2nd Temp. Bracing

Excavate to -1.8m

Excavate to -11.0m

10

Remove Temp. Bracing

Figure 6. Basement construction stages

Initially computer modeling, using PLAXIS to estimate the displacement of the tunnels without soil
improvement to the soft clay layer inside the excavation, indicated that tunnel displacement exceeded
the limit of 6.0mm specified for the MRTA structure or tracks.
A series of computer models with different wall configurations such as increased wall thickness
and embedment were carried out. It was found that the wall system did not significantly control the
displacement of the tunnels within the specified limit, but the expected tunnel movement is generally
governed by the wall deflection. Moreover, the computer modeling indicated that excavation for
installation of the 2nd bracing layer and subsequent removal of the bracing after completion of base
slab influenced the overall ground displacement and wall deflection, due to the presence of a thick soft
clay layer between at depths of 3.0m to 12.0m. To effectively reduce wall deflection and ground
displacement, improvement of the soft clay layer was considered necessary. We considered also that
the site conditions allowed the soft clay to be improved to the strength of stiff clay with deep cement
mixing inside the excavation zone close to the south-bound tunnel. The improve soil block could act
as a counterfort to the wall during installation of the 2nd bracing layer. Additionally the improved soil
block could also be used as a berm or contingency support for the wall during final excavation depth.
Predicted lateral and vertical displacements of the ground and tunnel movements are summarized
and presented in Table 2.
Table 2. Summary of Maximum Movements Computed by PLAXIS Program
Maximum Movement of
Maximum Movement of
ground (mm)
Tunnel (mm)
Model Type
Horizontal
Vertical @
Specified
Predicted
@ D-wall
ground level
Limit
Simple model without soil
improvement
Improved Soft Clay layer @3.013.5m, 5.0m wide inside
excavation.

40.57

38.22

6.88

<6.00

26.54

25.79

5.97

<6.00

5.2 Comparision with Similar Projects elsewhere

Publication of other cases resembling that of the planned project (Case A) in Bangkok has not been
found and thus published papers on the similar/relevant projects elsewhere are reviewed and an
assessment of movements in the MRTA structure or tracks induced by the planned excavation is
carried out and summarized as below.

344
376

Table 3. Summary of Subsoil Conditions at Project Sites

Case
Project Name
Location
No.
A

Bangkok

Ref.
In this paper

New Dhoby Ghaut

Station

Singapore

Cheung Sha Wan Gov.

Office

Hong Kong

People Park Station

Shanghai

New World Masion

Shanghai

12.0m thick soft clay overlying stiff

clay.
25.0m thick very soft marine clay
overlying dense to very dense
decomposed granite
2-3m thick fill, 3m thick organic clay
over 2m thick fluvial and 2m thick
residual soil overlying sandstone.
11.0m thick fill over 11.0m thick
marine deposit, overlying alluvium and
decomposed granite
30.0-40.0m thick soft clay and silty
clay interbeds
Silty clay and clay layers

Hong Kong Square

Shanghai

Silty clay and clay layers

The George Green

Cut&Cover

London

4-4.5m thick made ground over clay

Planned Project on
Asoke Road
Inland Revenue
Headquarter

Subsoil

Singapore

Yang et al
(1998)
Niu et al
(2003)
Chu et al
(2001)
Jianhang
(1998)

Higgins et al
(2001)

Although, general thickness of clay deposits vary with the projects, the nature of the work is
similar, it is noteworthy that deep excavation work was carried out in soft soils adjacent to existing
subway tunnels.
Table 4. Summary of Excavation Configurations
Max.
Case
Project Name
Excavation
No.
Depth (m)
A
B
C
D

Planned Project on Asoke

Road
Inland Revenue
Headquarter
New Dhoby Ghaut Station

Distance to
Outer Face of
Tunnel (m)

Depth to
Tunnel
Crown (m)

-18.0

16.0

11.0

-30.0

4.5

11.5

-18.5

<5.0

19.0-22.0

-23.067

12.9

6.959

-13.0

3.0

8.0

-11.0

7.8

15.0

Cheung Sha Wan Gov.

Office
People Park Station

New World Mansion

Hong Kong Square

-14

7.0

The George Green

Cut&Cover

-9.0

1.0m to below
excavation

13.7

345
377

Wall Type
0.8mx20m
D-Wall
1.0mx24.5m
D-Wall
0.8+1.0mx35.0m
Secant pile wall
D-Wall
D-Wall
26m D-Wall
0.8x24.0m
D-Wall
1.0mx19.0m
Secant pile wall

Table 5. Summary of Performance of Walls

Case
No.

Project Name

Planned Project
on Asoke Road

Inland Revenue
Headquarter
New Dhoby
Ghaut Station
Cheung Sha Wan
Gov. Office
People Park
Station
New World
Mansion
Hong Kong
Square

C
D
E
F
G
H

The George
Green
Cut&Cover

Excavation
Method
Bottom Up
2 level Strut +
improved soil berm
Top-down

Max. Wall
Deflection (mm)

Max. Tunnel Displacement

(mm)

Predicted
26.54

Actual
-

Predicted
5.97

Actual
-

Specified
6.0

30

25

11

15

17

15
<7@ track
18

20

Bottom Up
7 level Strut
Top-down

24
<25

<25

Bottom Up
5 level Strut
Bottom Up

35

38

<12

Bottom Up
& improved soil
berm + 3 level strut
Top-down

<40

<40

<15
-

11

<10

<10

<10

<10

12

Review of the above completed projects, with the exception of the planned project (Case A),
indicates that the specified limit of tunnel displacement ranges from 10mm to 20mm. Details of tunnel
lining and specified limits of the railway track movement of these projects were not specifically
reported. Only in Case No. C (Ref. Niu et al 2003), maximum movements of 7mm and 15mm were
recorded for both the railway track and tunnel respectively.
Based on the projects mentioned above, controlling the maximum movements of the MRTA
tunnel and railway track induced by the planned excavation for the project, within the specified limits
is possible. However the excavation work must be carried out by competent contractors with
contingency plans such as that outlined previously, with the geotechnical instrumentation and
monitoring mentioned in the following section as a minimum requirement.
5.3 Geotechnical Instrumentation and Monitoring
The following instrumentation is considered essential and monitoring must be carried out throughout
the basement construction period.
Inclinometer - 2 inclinometers in the ground between the tunnel and planned diaphragm wall and
(for diaphragm wall construction) 4 to 5 inclinometers in the diaphragm wall around the excavation.
Convergent Bolts & Tape Extensometer If necessary, adequate quantities of convergent bolts
may be installed on the tunnel segments to monitor the convergent movement with tape extensometer
at each construction stage throughout the basement excavation period.
Settlement Plates/markers A minimum of 4 settlement points on the ground across the tunnels
needs to be installed in line with each inclinometer location and surveyed at each construction stage
throughout the construction period.
Survey Movement of tracks must be surveyed at each construction stage throughout the
basement excavation period.
Contingency Plan and Actions - Since the predicted displacement of the tunnels is very close to
the specified limits, monitoring of adjacent structures and ground movements during the staged

346

378

excavation is necessary, with checking against the predicted movements at every stage of construction.
A contingency plan with trigger values of displacements being based on the predicted values must be
prepared and necessary action must be taken immediately if the measured movements of diaphragm
wall and tunnels exceed the trigger values set. It was suggested that a contingency plan including, but
not limited to, the following be prepared and carried out accordingly when required.
Preloading of struts as necessary
Localized excavation and monitoring
Provision of an additional bracing layer
Speeding up the construction.
6. SUMMARY OF ASSESSMENT
An assessment of works associated with bored piling and diaphragm wall work to meet Technical
Requirements for Engineering Works within the Protection Zone of the MRTA Chaloem
Ratchamonkhon Line is summarized in the table below.
Table 6. Summary of Assessment of Basement Excavation with Braced Diaphragm Wall
Specified
Item
Description
Provision
Assessment/Remarks
Requirement
1
2
2.1

Diaphragm Wall
Bored Piling
Bracing and Excavation
Calculations

2.1.1

Engineering Work in
Zone B
Surcharge at ground level

2.1.2

Surcharge on tunnel

2.1.3

Differential Movement

2.1.4

Clear distance between

pile/d-wall and MRTA
structure
Additional surcharge load
by piles

2.1.5

2.1.6

Piling

2.1.7
2.1.8

Use of vibratory method

Future
additional
surcharge in Zone B
Approval for excavation
adjacent
to
MRTA
structures

2.1.9

2.1.10
2.1.11

Site investigation
boreholes for location of
MRTA structure
Soil Improvement

Method Statements

Signed with
Chartered Engineer

Not > 50kPa (5t/m2)

Not
>
25KPa
(2.5t/m2)
< 1:2000
< 6mm in total
> 3m or (5) pile
diameter
Not > 50kPa
designed GL

Bored piling with

auger under slurry
Not within 10m
Use 1:1 slope of load
distribution
>3.5m near tunnels
or station
>1.5m near others
Suitable
instrumentation
Fully grouted
approved by the
authority
Approval submission

347
379

Yes
Yes
outlined
Yes

Conformed

15kPa
(1.5t/m2)
15kPa
(1.5t/m2)
< 6mm
> 6m or (7) pile
diameter

Used in D-Wall analysis and

Design
Used in D-Wall analysis and
Design
Use observational approach
and instrumentation program
layout plan and drawings
conform.

Yes

Pile cutoff is below 11.0m

and generally outside Zone B.
D-Wall is supported by pile
legs to 57-60.0m.
Method statement conforms.

> 10m
Yes

Method statement conforms.

Conformed

~ 11.0m with
braced D-Wall.
Instrumentation
program
outlined
-

Approval is obtained.

Yes (by Main

contractor)

Documents
prepared
submitted for approval

&

Documents
prepared
&
submitted for approval
The site is about 8m away
from the MRTA structure.
It will be inside the perimeter
D-wall as analysed.

Moreover we also concluded that the structural design of the diaphragm wall as analyzed and designed
with the WALLAP program was adequate for the planned excavation work. However the maximum
deflection of the wall adjacent to the existing tunnels had to be reduced to 20mm, to limit the
displacement of the tunnel within the specified displacement as indicated by computer modeling with
PLAXIS program. Additionally, soft clay adjacent to the south-bound tunnel inside the excavation at
depths between 3.0m and 13.5m must be treated to improve the shear strength of the soil to that of stiff
clay layer (about 10t/m2). We intended the soil improvement to minimize the wall movement and acts
as a contingency measure in excavation. The observational approach is strongly recommended for a
deep excavation in the proximity of exiting tunnels.
7. DIAPHRAGM WALL CONSTRUCTION
A review assessment of ground movements induced by diaphragm walling was also carried out during
diaphragm wall design and analysis. Ground movements associated with diaphragm wall trenching in
Bangkok clay has been studied and reported (Thasnanipan et al, 2004). It was reported that an
inclinometer (24m long) was installed 1.0m away from the trial trench face for monitoring of soil
movement during and after trenching. Additionally 6 surface settlement points with 2.0m spacing
were also installed and monitored. The trial trench 0.6x4.5x18.0m was excavated under polymer
slurry. The monitoring results indicated that vertical ground settlement did not extend beyond 8.0m
from the trench face. The maximum vertical settlement of 9mm was found to occur at about 2.0m
away from the trench face. The inclinometer readings which were recorded 1.0m away from the
trench face indicated that the maximum lateral soil movement of 5.8mm occurred at the depth of about
3.5m below the ground. At 15.0m depth, the lateral movement of soil 1.0m away from the trench was
about 1.0mm after 12 hours of trench excavation. Based on the above information, for the diaphragm
wall construction adjacent to the existing MRTA tunnels located about 7.8m away from the outer face
of planned diaphragm wall of the project site and at depths of about 15.0m below ground level, the
following statements were made;
1. Vertical displacement of ground due to diaphragm wall trenching was expected to be within
8.0m from the trench face and thus not affect the tunnels which were located beyond 7.8m
from the trenching.
2. Lateral soil displacement within 1.0m from the trench face at a depth of 15.0m depth was
about 1.0mm. It was expected that the displacement would decreasing distance from the
trench face. At the tunnel location, no lateral soil displacement was expected as the southbound tunnel was about 8.0m from the trench face.
3. For trench stability, bentonite slurry was used instead of polymer slurry.
Diaphragm wall construction was carried out with a conventional rope-suspended mechanical
grab under bentonite slurry in July 2005. Two inclinometer tubes were installed between the southbound tunnel and the planned diaphragm wall, about 4.0m from the outer face. Since the subway
tunnels were beneath Asoke Road, settlement marks were established on the pavement in line with the
inclinometer location and crossing the tunnel alignment. Monitoring of ground movements was
carried out during construction of diaphragm wall panels nearest to the inclinometer tubes. Figure 8
shows the layout of inclinometer and settlement marks. For construction of bored piles, 15m long
temporary casing was used. The piles located within protection zone B were constructed after
completion of adjacent diaphragm wall panels to protect from the ground vibration induced by
temporary casing installation and extraction.
7.1 Soil Movement during Diaphragm Wall Construction
Soil movement during diaphragm wall panel construction were monitored (Table 7). Results of
settlement survey at the settlement marks indicated no vertical movements of the ground.

348
380

Table 7. Summary of Monitoring Results (I-6 & I-7): Soil Movements at Inclinometer Locations
Movement @ 15.0m depth
Construction Activities
(mm)
Remarks
I-6
I-7
Top 2.0m in depth was
D-Wall panel excavation to -10.0m
-0.87
-0.40
protected by concrete
D-Wall panel excavation to -20.0m
1.04
-0.81
guide walls.
After de-sanding
0.12
-0.38
After concrete pouring
-1.91
-2.18
24hrs after concrete pouring
-2.34
-1.62
36/48hrs after concrete pouring
-1.13
-2.55
Note: - movement towards tunnel (opposite to direction of excavation)
INCLINOMETER I-7 (A AXIS)

INCLINOMETER I-6 (A AXIS)

Cum. Displacement , mm.
-20

-15

-10

-5

10

Cum. Displacement , mm.

15

-20

20

10

10

10

15

20

MRTA Northbound Tunnel

Asoke Road

Depth, m.

20

20

D-wall excavate to -10 m

D-wall excavate to -20 m
After de-sanding
After concreting
24 hrs After concreting
48 hrs After concreting"

25

nel

I-6

I-7

9.625m

15

Tun
MRTA Southbound

7.815m

15

25

-5

Depth, m.

-10

-15

I-4

D-wall excavate to -10 m

D-wall excavate to -20 m
After de-sanding
After concreting
24 hrs After concreting
36 hrs After concreting"

ilding

Figure 7. Inclinometer Monitoring Results of I-6 and I-7

LEGEND
Inclinometer
Settlement Point
D-Wall with Pile Leg
D-Wall with Barrette Leg
Barrette

12 Storey O

I-3

Figure 8. Layout of Instrumentation

for Diaphragm Panel Construction

The following conclusions were made based on the above monitoring results.
1. No vertical displacement was observed at the surface settlement points; vertical displacement
of ground due to diaphragm wall trenching seemed to be within 2.0m from the trench face and
would not affect the tunnels which were located 7.8m beyond the trenching.
2. Maximum lateral soil displacement at 4.0m from the trench face at 15.0m depth was about
1.0mm towards excavation. It was expected that the displacement would decrease with
distance increasing from the trench face. At the tunnel location, lateral soil displacement was
estimated to be less than 0.5mm based on the extrapolation of displacement data.
3. After concrete casting of the panels, lateral soil movements were found to be towards the
tunnels, possibly from pressure of wet concrete. Maximum movement at depth 15.0m at the
inclinometer location was about 2.18mm after concrete pouring. After 36hrs of concrete
casting, the maximum movement observed was about 2.55mm. At the tunnel location, lateral
soil displacement was estimated to be less than 1.6mm based on the extrapolation of
displacement data.
4. After the diaphragm wall panel was completed, residual movement of the soil towards the
tunnels was observed. Due to the direction of this movement, it could be excluded from
tunnel movement measurements induced by diaphragm wall deflection during basement
excavation work.

349

381

8. CONCLUSION
At the time of drafting this paper, soil improvement with soil-cement columns was completed and
excavation work has commenced with specified instrumentation and monitoring programs. An
independent third party consultant has been appointed by the main contractor to review, inspect and
supervise the excavation work.
Diaphragm wall and foundation bored pile construction was completed, in compliance with
technical requirements for engineering work within the MRTA protection zone. Review and
assessment of the deep excavation with diaphragm wall adjacent to the existing tunnels indicated that
the project is feasible with adequate monitoring and contingency plans as outlined.
It is also concluded that diaphragm wall construction in close vicinity of existing subway tunnels
was successfully completed with adequate instrumentation and monitoring.
Further research is recommended on the actual performance of the diaphragm wall when
basement construction data and complete monitoring results are available.
REFERENCES
CIRIA Report 185 (1999), The observational method in ground engineering: Principles and
applications, CIRIA, 6 Storeys Gate, Westminster, London, UK.
Chapman T. J. P., Nicholson D. P and Luby D. (2001), Use of the observational method for the
construction of piles next to tunnels, Proceedings of the International conference on Response
of Buildings to Excavation Induced Ground Movements held at Imperial College, London, UK.,
July 17-18, 2001, pp 213-223.
Chu R. P. K., Yau P. K. F., Leung D. H. K., and Mok K. H. (2001), Integrate approach for deep
excavation along soft ground (MRTC) tunnels, Proceedings of the Third International
Conference on Soft Soil Engineering, Hong Kong, December 6-8, 2001, pp 391-396.
Higgins K G, Paterson J., Moriaty J., Potts D. M., and Jardine R. J. (2001), The effect of an
excavation in a stiff fissured clay on existing tunnels in close proximity, Proceedings of the
International Conference on Response of Buildings to Excavation Induced Ground Movements
held at Imperial College, London, UK., July 17-18, 2001, pp 313-324.
Jianhang L. (1998), Design and construction of excavation for Shanghai Metro, Big Digs Around
the World, Proceedings of Sessions of Geo-Congress 98, Geotechnical Special Publication No.
86, Boston, Massachusetts, USA, October 18-21, 1998, pp 103-123.
Mair R. J., Taylor R. N., and Bracegirdle A., (1993), Subsurface settlement profiles above tunnels in
clays, Geotechnique 43, No. 2, pp 315-320.
Niu J. X.,Wong I. H., and Makino M., (2003), Effects of deep excavation on adjacent transit tunnels
and station, Proceedings of the Twelfth Asian Regional Conference on Soil Mechanics and
Geotechnical Engineering, Singapore, August 4-8, 2003, pp 883-886.
Phienwej N., Hock G. C., and Balasubramnaiam A. B., (1998), Ground movements in deep
excavations with concrete diaphragm walls in Bangkok soils, Proceedings of the Thirteenth
Southeast Asian Geotechnical Conference, Taipei, Taiwan ROC, November 16-20, 1998, pp
637-642.
Schulz C., (2001), Geotechnical challenges in the construction of Thailands first subway,
Proceedings of the Fourteenth Southeast Asian Geotechnical Conference, Hong Kong,
December10-14, 2001, pp 413-418.
Thasnanipan, N., Maung, A. W., Boonyarak, T., and Aye, Z. Z. (2004), Stability of a trial trench
excavated under polymer slurry in Bangkok soft clay, 15th Southeast Asian Geotechnical
Conference, Bangkok, Thailand, November 22-26, 2004.
Yang K. S., Tsen C. N., and Tang S. K., (1998), Performance of a deep excavation with top-down
construction, Proceedings of the Thirteenth Southeast Asian Geotechnical Conference, Taipei,
Taiwan ROC, November 16-20, 1998, pp 587-593.

350
382

Proceedings of International Symposium on Tunnelling for Urban

Development
14-15 December 2007, Pattaya, Thailand

International Symposium on
Tunnelling for Urban
Development
(IS-Pattaya 2007)

Application of Observational Method

Diaphragm Wall Support Excavation
Zaw Zaw Aye, Chanchai Submaneewong,
Boonyarak and Chakkrit Chanchad
SEAFCO Public Co., Ltd. Bangkok, Thailand

351

in

Thayanan

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand

Application of Observational Method in Diaphragm Wall Support

Excavation
Zaw Zaw Aye1, Chanchai Submaneewong2, Thayanan Boonyarak3 and Chakkrit Chanchad4
1,2,3,4

Seafco Public Co., Ltd., Bangkok, Thailand

Abstract: Application of observational method in managing geotechnical risk in diaphragm wall

support excavation in Bangkok is presented in this paper. Intensive modification of construction
sequence in actual work execution with value engineering options different from tender stage design
along with application of observational method in Diaphragm-wall-support two level underground car
park building located in the historically significant area of Bangkok is firstly demonstrated. Simple
approach of observational method in risk assessment of buildings and structures backup by extensive
instrumentation in tunneling and deep excavation works of Contract No. 1 of the M.R.T Chaloem
Ratchamongkhon Line, the first underground mass transit system project of Bangkok (MRTA) is also
reported. Key Factors contributed in successful application of observational method are discussed.
1. INTRODUCTION
Observational Method (OM) is not a very popular term among geotechnical engineers in Thailand,
though the method has been applied in some major projects. The main reason of unpopularity of the
method in Thailand may be due to the fact that traditionally there is less option opened for design
modification as the project progresses. Geotechnical engineers may be more concerned on misused
and limitation of the method and do not prefer to take risk. It is important to discuss among
geotechnical engineers with regard to overall philosophy, key requirement, limitation and practical
approach in application of observational method.
Tunneling works and deep excavation works have a high degree of geotechnical risk that needs
to be identified, assessed, minimized and managed. Deep excavation and tunneling works generally
interact with natural ground or geo-material and incorporating their characteristics as structural
elements of their own stability. This interaction is more significant in underground excavation and
tunneling projects than in other civil engineering projects. In this paper, application of observation
method in managing geotechnical risk in two different projects is presented.
2. OBSERVATIONAL METHOD AND GEOTECHNICAL RISK MANAGEMENT
Historically, engineers used observational method to deal with uncertainties in the ground or
geotechnical risks in early age of civil engineering. Peck (1967) pointed out the advantages of
observational method by giving an example on a dam across the Svir River in Russia, from a
documented record of the work of Graftio (1936).
Professor R. B. Peck set out procedures for the observational method (OM) as applied in soil
mechanics in the Ninth Rankine Lecture (Peck, 1969). Peck described the limitation and drawbacks of
observational method. Powderham (1996) reviewed the main features of the observational method of
Peck and summarized the key requirements as follows;
(1) It must be possible to alter the design during construction
(2) The contractual condition must be compatible and allow design to be directly related to actual
construction method
(3) An acceptable level of risk must be identified and controlled. In particular this requires a
planned course of action for every foreseeable eventuality

353

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
(4) Critical observation must be identified and obtained
3. OBSERVATIONAL METHOD, RISK MANAGEMENT AND VALUE ENGINEERING
Figure 1 shows a generalized link between the elements of observational method and risk management
with integration of value engineering.

Geotechnical
Risk Management

Observational Method

Design & Plan

Identify Risk
Value Engineering

Analysed Degree
of Risk

Minimize
& Control

Construction Control
Monitoring & Review
Trigger Criteria
Exceeds

Yes

No
Implement
Planned Action

Figure 1. The elements of observational method and risk management with integration of value
engineering.
4. VALUE ENGINEERING AND APPLICATION OF OM IN BASEMENT EXCAVATION
WITH BUTTRESS-SUPPORT THIN DIAPHRAGM WALL
This project demonstrated the benefits of value engineering option powered by effective application of
observational method in historically significant area of Bangkok. The project is a two-level
underground car park located in the center of Rattanakosin Island, the heart of an old established,
historically and culturally significant area of Bangkok. The project owner, Bangkok Metropolitan
Authority (BMA) awarded the semi-turnkey basis construction contract Lam Kon Muang
Underground Car Park to SEAFCO Co., Ltd. as a contractor. The contract consists of 3 major scope
of works: (1) Construction of building foundation and retaining structure - diaphragm wall, barrette
and bored piles (2) Excavation works including temporary bracing design and installation (3)
Construction of the entire two-level underground car park building having car park area of 18,552 m2
and roof-level park of 10,936m2 plus cut-and-cover tunnel, underpass access to the City Hall.
Geotechnical aspects highlighting the application of observational method and performance of
buttress-support diaphragm wall of 0.60 m width for two-level underground car park building is
discussed in this paper.
5. PROJECT REQUIREMENT AND MAJOR CONSTRAINTS
Since there is a limited availability of car parking space in the surrounding congested neighborhood
and the project site was being used as a grade-level car parking space prior to the award of the
contract, the key requirement was to construct the underground car park in two phases to construct
Phase 1 while leaving space for car parking in Phase 2, and to utilize semi-finished underground car
park of Phase 1 during construction of Phase 2. This requirement posed the need of temporary
retaining wall between Phase 1 and 2.

354

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
Construction site is surrounded by numbers of sensitive structures as shown in Figure 2, - in the
south, Wat Suthat, one of Thailands most important temples and the Historical Giant-swing, in the
north, the City Hall and at South-east corner, the Historical Brahmin Temple. Rows of old shop-house
buildings are closely located in the east and west boundaries of the project. Location of the project site
itself in the vicinity of sensitive structures and buildings therefore posed some constraints, which
called for the need of careful consideration in establishing the design principles and sequence of
construction.
CITY HALL

OPENING

SIRIPONG ROAD

OPENING

OPENING
OPENING

SHOP-HOUSES

SHOP-HOUSES
DINGSOR ROAD

OPENING

SHOP-HOUSES

OPENING

MAHUNNOP ROAD

OPENING

OPENING

BAMRUNG MUANG ROAD

GIANT-SWING

WAT SUTHAT (TEMPLE)

Figure 2. Layout plan of the project showing adjacent buildings

Location of the project site itself in the vicinity of sensitive structures and buildings therefore
posed some constraints, which called for the need of careful consideration in establishing the design
principles and sequence of construction. The architectural and utility aspects of the project called for
the design of the basement with a number of openings from the ground surface to the final basement
slab level to facilitate the ventilation system as shown in figure 3.
Hence the roof slab can not be physically utilized as bracing in most of the area where the
diaphragm wall is to be acting as a cantilever retaining wall in the permanent stage. It was analyzed in
the preliminary analyses that the deflection of the diaphragm wall of 0.60 m width would be large if it
was to be fully cantilevered. As the project is located in a sensitive area, ground movement induced by
large deflection was unfavorable. It was therefore decided to use buttress to minimize the diaphragm
wall deflection as shown in figure 3.

355

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand

D IA P H R A G M W A L L

RC BUTTRESS

BARRETTE

W ALE BEAM

0 .3 0

T IE B E A M

P IL E C A P

SLA B

(a)

I-5

N O R T H D -W A L L

+0.20 ROOF SLAB

0.00
CAPPING
BEAM
RC BUTTRESS

G.W.L

S H E E T PIL E

EAST D-WALL

WEST D-WALL

PH A S E 2

I-6

2
SOFT CLAY

-9m

124 m

I-4

S ee d etails in (b )

I-3

PH A S E 1

S O U T H D -W A L L

-12m
I-1

PILE CAP
-5.60 (B2)

TIE
BEAM

2
WALE
BEAM

SOFT CLAY

DIAPHRAGM
WALL (tip=-16m)

MEDIUM CLAY

STIFF CLAY

-15m
-16m

-20m.

I-2

78 m

-2.70 SLAB (B1)

VERY STIFF CLAY

FOUNDATION PILES
DIA. 600mm (tip=-20m)

BARRETTE
(tip=-20m)

-20m.

(c)

(b)

Figure 3. (a) Plan of diaphragm wall and buttress (b) layout of diaphragm wall and piles (c) sectional
view of the structures

6. DIAPHRAGM WALLS BARRETTES AND BORED PILES CONSTRUCTION

Diaphragm wall having 600mm width founded at 16m below ground level (B.G.L) was constructed
simultaneously with dry-processed bored piles of diameter 600 mm with toe depth 20 m below ground
level. Barrettes having same toe depth as bored piles were installed at 8m spacing along with
diaphragm wall panels. Sheet pile wall (14m deep) was used as a temporary retaining wall at the
boundary of Phase 1 and 2 as shown in Figure 4.
7. SUBSOIL CONDITION AND DESIGN PARAMETERS
Typical subsoil profile at the site is characterized by thick Bangkok soft clay layer at the top followed
by thin layer of medium clay, and stiff clay layers.

356

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand

Undrained Shear Strength (kN/m )

Depth from ground surface (m)

10
15
20

Soft clay

35

Stiff clay
Very Stiff clay

150

20
40
60
SPT N-Value (blows / ft)

200

5.00

15.00
20.00
25.00

Medium Dense
Sand

30.00
35.00
40.00

40
45

100

Medium stiff clay 10.00

25
30

50

0.00

0
5

Dense sand

45.00
50.00

50

Su BH-1
Su BH-3
SPT BH-1
SPT BH-3

80

Su BH-2
Design (Su)
SPT BH-2
Design (SPT)

Figure 4. Plotted Su and SPT N-value with design lines

8. ORIGINAL TENDER STAGE DESIGN
The designers involved in the tender stage design made a fairly conservative design with two levels
temporary bracing as shown in Figure 5. Uncertainty of the performance of a thin diaphragm wall in
soft clay layer was the likely reason to adopt the conservative design in the tender stage. It is not
unreasonable to adopt the conservative design considering time-dependent consolidation property of
soft marine clay and likely long elapsed time of un-strutted diaphragm wall (relatively long un-strutted
span of about 6.0 m between first strut and final excavation level for 600 mm diaphragm wall) due to
large volume of excavation work involved.

HORIZONTAL STRUT

RAKING STRUT

D-WALL

BERM
D-WALL

BARRETTE

-16m

SLAB

BORED PILE

-20m

Figure 5. Tender stage bracing system diaphragm wall was designed with soil-berm and 2 struts
support (horizontal bracing and raker)

357

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
9. VALUE ENGINEERING OPTION FOR PHASE 1
Value engineering review of the temporary works was undertaken by the contractors new in-house
design engineering team prior to the commencement of Phase 1 excavation works. Rigorous attention
to detail of the design concept and constructability was made in the pre-construction discussions
between design engineers and construction team.
The main objectives for value engineering options were to minimize the material and
construction sequence involved in temporary works so as to accelerate the excavation time thereby
saving overall costs without compromising the safety aspect. After conducting a series of re-analyses
with different conditions major modifications were made: (1) To lower the first strut level to 1.8 m
from the original tender stage design level 1.0 m (2) To use only 1 temporary strut, omitting second
level raking strut with the provision of sloping soil berm against diaphragm walls. Soil berm was to
remove after completion of base slab construction in the majority of area minimizing the elapsed
time of partially un-strutted diaphragm wall between temporary strut and the final excavation level.
Modified bracing system of Phase 1 as an outcome of value engineering review is illustrated in Figure
6.

HORIZONTAL STRUT

SLAB

HORIZONTAL STRUT

BERM

RAKING STRUT

D-WALL

SHEET PILE

BARRETTE

-16m

SLAB

-16m
-20m

-20m

-14m
BORED PILE

SOUTH D-WALL vs SHEET PILE

EAST & WEST D-WALL

Figure 6. Value engineering option - Modified Bracing system for Phase 1

10. IMPLEMENTATION OF THE OBSERVATIONAL METHOD
During the review of tender stage design for Phase 1, it was recognized that two-phase excavation
works in this project was ideally suitable for application of the observation method. The flexibility of
the contractual requirement in temporary design which allowed the contractor to modify the design
and construction method also provided the favor for the observational method.
In order to assess the most probable condition assumed in value engineering design and to
take necessary actions if monitoring results reveal most unfavorable conditions (i.e. actual deflection
of diaphragm wall reaches maximum acceptable limit), the observational method was implemented on
the followings basis.
Reviewed the design parameters together with critical conditions posed on sites as well as most
critical stage in excavation and basement construction work
Predicted the performance of diaphragm wall with most probable as well as most unfavorable
conditions and parameters.
Established the trigger criteria based on predicted diaphragm wall deflection
Predefined the practical contingency plan for most unfavorable conditions where wall deflection
reaches trigger levels
Set out the instrumentation program with the consideration of above factors

358

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
Monitored the performance of diaphragm wall. Compared the monitoring results with the predicted
and trigger values and reassessed
Implemented the contingency measures if monitoring result reaches action level of trigger values
Process of OM in Phase 1 is illustrated in a flowchart in figure 7.
OM in Phase 1
Value
Engineering
Option

Design & Plan

Review &Modification of Tender Stage Design

Construction Control

Reassessment
Process

Monitoring & Review

Trigger Criteria
Exceeds

No
Implement
Planned Action

Yes

Figure 7. Flowchart showing the OM process involved in Phase 1

Figure 8 shows the predicted diaphragm wall lateral displacement or deflection of Phase 1 (east,
west and south diaphragm wall) at two conditions together with trigger levels and tender stage
prediction. It should be noted that the diaphragm wall deflection was predicted to be maximum or
most critical after removing the horizontal temporary strut.
Most probable condition was established for the predicted deflection of diaphragm wall with
full influence of buttress support assuming buttress effectively supports as permanent strut in
diaphragm wall analysis model. Most unfavorable condition was set out for the predicted deflection of
diaphragm wall without considering influence of buttress buttress was excluded in the model.
Diaphragm wall reinforcement was designed based on the most unfavorable condition.
Cum. Displacement , mm.
0
0
2

20
NORMAL

40

60
ALARM

80

100

ACTION

MOST
PROBABLE
PREDICTION

Depth, m.

6
MOST
UNFAVORABLE
PREDICTION

8
10
12
14
16
18
20

Tender Stage
Prediction with buttress :
Alarm Trigger Level
Prediction without buttress
Action Trigger Level

Figure 8. Prediction of diaphragm wall performance in pre-construction stage with trigger criteria in
comparison with tender stage prediction

359

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
In establishing the criteria for trigger values, it was necessary to consider the broad context in
which diaphragm wall exists, design assumption and concept, likely behaviour of the wall itself or its
predicted performance and effectiveness of selected temporary bracing system. Trigger levels were
established to provide the design team and construction team, an opportunity for early review and
resetting of the monitoring frequency as well as for implementing the contingency measure as
necessary.
In general, exceeding alert trigger levels must initiate a review of design data, construction
progress and monitoring frequencies with the consideration of possible measures to limit further
deflection. Exceeding action trigger levels must initiate further review of above mentioned points and
if necessary to initiate a planned course of action or contingency measures.
Effective and good communications between the design team and construction crew were made
all along the excavation stages with clear responsibilities in construction control. The contingency
measures included immediate backing filling of excavated soil and installing of temporary struts in
the critical area.

11. MONITORING PROGRAM

In planning the monitoring system and program, it is important to consider the parameters to be
measured which reflect the actual performance of the diaphragm wall support excavation. It is also
necessary to take account the practical measurement applicable for established trigger criteria, number
and frequency of measurement required to carry out a meaningful interpretation of wall behaviour
which would be integrated in the implementation of observational method.
Comprehensive and robust monitoring program was set up as a key element in application of the
observational method and to ensure that modified construction sequence would not have adverse effect
in temporary stage and on permanent design. A total of 6 inclinometers (3 in each phase) were
installed in diaphragm wall together with some survey points. In order to make effective use of the
established trigger levels, an adequate number of measurement were carried out at appropriate
frequency. Typical monitoring frequencies for inclinometer set up as guideline for the project is
outlined below.

Measured immediately before commencing excavation in the vicinity of instrument

Minimum readings of 2 times a week while excavation in progress

Minimum readings of 1 time a week when no excavation the vicinity of instrument

Minimum readings of 3 times a week when measured deflection values exceeded alert trigger
levels

Minimum reading 1 time a day when measured deflection values reached action trigger levels
As the most critical stage was predicted at the time horizontal temporary bracings were removed,
a full attention was paid to the inclinometer monitoring with the following special criteria and
frequencies.

First temporary strut removal was to carry out at the diaphragm wall panel where inclinometer
was located

Measured immediately before removal of temporary strut at the closest distance to the instrument.

Measured every 6-8 hours immediately after removing the first strut

Second strut to be removed must be the one located immediately adjacent to first strut which had
removed

Not to remove the second strut until inclinometer measured deflection values had stabilized

Not to remove more struts unless measured deflection values were stabilized and within alarm
trigger levels
In addition to inclinometer measurement, diaphragm wall movement was also monitored by the
survey points strategically marked on the wall panels. Ground settlement and surface cracks behind
the diaphragm wall were also visually checked by the construction team as daily basic.

360

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
12. PERFORMANCE OF PHASE 1 DIAPHRAGM WALL
After carrying out the comprehensive desk studies and establishment of systematic monitoring
program presented above, Phase 1 excavation work was carefully commenced. Figure 9 shows the
maximum accumulated diaphragm wall deflection at different stages of excavation monitored by
inclinometer No.1 (I-1 at East wall of Phase 1) together with trigger levels and predicted maximum
deflection profile of 3 different conditions - tender stage design (2 temporary struts), modified design
with buttress and modified design without buttress. It can be observed from figure that measured
lateral movement pattern of diaphragm wall agreed well with that of prediction for modified design
with buttress - meaning buttress-support has significant influence on wall deflection.
Cum. Displacement , mm.
0

20

40

60

80

ALARM

NORMAL

100

ACTION

MOST PROBABLE PREDICTION

2
4

MOST
UNFAVORABLE
PREDICTION

Depth, m.

6
8

Stage 1
Stage 2

10

Stage 3

12

Stage 4
Stage 5

14

Tender Stage
16

Alarm Trigger Level

Action Trigger Level

18

Prediction without buttress

20

Figure 9.
levels

Phase 1 East Wall-Monitored diaphragm wall deflection at different stages with trigger
Cum. Displacement , mm.
0

20

40

60

80

100

0
NORMAL

ACTION

ALARM

MOST PROBABLE PREDICTION

4
6
MOST UNFAVORABLE
PREDICTION

Depth, m.

8
10
12

Stage 1
Stage 2
Stage 3
Stage 4

14
16
18
20

Stage 5
Tender Stage
Alarm Trigger Level
Action Trigger Level
Prediction without buttress

Figure 10. Phase 1 South Wall - Monitored diaphragm wall deflection at different stages with trigger
levels

361

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
Deflection profile of South diaphragm wall which braced against temporary sheet pile wall at
Phase 1 and 2 boundaries is presented in Figure 10. As can be seen in Figure 10, South diaphragm
wall deflection is significantly higher than that of east wall, which is likely to be caused by the fact
that South diaphragm wall is braced with more flexible sheet pile wall. description of stages shown in
the legend of Figure 9 and 10 is summarized in Table 3.
Table 3. Description of stages shown in Figure 9 & 10
Stage
Description
1

Excavate to 2.2m and installed temporary strut

Excavate to 4m

Excavate to 6.6m with berm

Removal of berm

Removal of temporary strut after completion of buttress

Figure 11. Phase 1 excavation work in progress with historical Buddhist temple Wat Suthat in
background (Phase 2 area was being used as car parking space)
13. OM IN PHASE 2 - MODIFICATION OF BRACING SYSTEM
Monitoring results of the Phase 1 excavation work provided an ample opportunity to review the
design assumption, fine tune the parameters used in the analysis of the diaphragm wall for the Phase 2
and made modification of construction sequence. The major modifications are: (1) Removal of
soil-berm at East and West diaphragm wall in shorter duration than that of Phase 1, and (2) Using
raking struts instead of horizontal strut for North diaphragm wall a shown in Figure 12.

362

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand

HORIZONTAL STRUT

SLAB

COMPLETED PHASE 1
STRUCTURE

RAKING STRUT
SLAB

D-WALL

BERM

D-WALL

BARRETTE

-16m

-16m

-20m

-20m
EAST & WEST D-WALL

Figure 12.

BERM

NORTH D-WALL with raker support

Bracing system of Phase 2 diaphragm walls

Process involved in application of OM in Phase 2 is shown in figure 13.

Value
Engineering
Option

Used data &

knowledge
gained from
Phase 1
Calibrated
design parameters

OM in Phase 2
Design & Plan
Modification of Design
Based on Actual Performance of Phase 1

Reassessment
Process

Construction Control
Monitoring & Review
Trigger Criteria
Exceeds

Yes

No
Implement
Planned Action

Figure 13. Flowchart showing the OM process involved in Phase 2

14. PERFORMANCE OF PHASE 2 DIAPHRAGM WALL
Figure 14 depicts the measured deflection of east diaphragm wall of Phase 2. As can be observed in
Figure 14 in comparison with Figure 9, the maximum deflection of east diaphragm wall in Phase 2 is
larger than that of Phase 1.

363

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
The likely reasons of this observation are;
Un-strutted elapsed time for first temporary bracing in Phase 2 was longer than that of Phase 1.
In Phase 1, horizontal struts were installed in north-south direction which temporary kingpost
columns and strut were integrated in crisscross pattern with east-west struts - providing
complete-support more rigid bracing system. Whereas in Phase 2, horizontal struts were installed
only in east-west direction without having crisscross pattern with north wall having less rigid
bracing system than that of Phase 1.
Cum. Displacement , mm.
0
-2

20
NORMAL

40

60
ALARM

80

100

ACTION
MOST PROBABLE PREDICTION

2
4

Depth, m.

6
MOST UNFAVORABLE
PREDICTION

8
Stage 1

10

Stage 2

12

Stage 3
Stage 4

14

Stage 5

16

Alarm Trigger Level

Action Trigger Level

18

Prediction without buttress

20

Figure 14. Phase 2 East Wall - Monitored diaphragm wall deflection at different stages with trigger
levels
With assurance of diaphragm wall performance from monitoring results of Phase 1, original
plan of using horizontal struts for North diaphragm wall was modified by using raking struts instead.
As can be seen in Figure 15, deflection of North diaphragm wall (with raking strut support) is
significantly higher than that of East diaphragm wall (with horizontal strut support). The main reason
of larger movement of north diaphragm wall is due to the fact that it was supported only by the berm
for the long period (about 52 days) before completion of raking struts so that soil-berm became soften
during the long elapsed un-strutted period.
Table 4. Description of stages shown in Figure 14 & 15
Stage East Diaphragm wall
North Diaphragm wall
1
Excavate to 2.2m and installed
Excavate to 2.2m at d-wall, and to 6.6m with
temporary strut
sloping berm
2
Excavate to 4m
Installed raking strut
3
Excavate to 6.6m with berm
Removal of berm
4
Removal of berm
Removal of raker after completion of buttress
5
Removal of temporary strut after
completion of buttress

364

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
Cum. Displacement , mm.
0
-2

20

40

60

80

NORMAL

100
ALARM

120

140

ACTION

MOST PROBABLE PREDICTION

2
4

Depth, m.

6
MOST UNFAVORABLE
PREDICTION

8
10
12
14
16
18

Stage 1
Stage 2
Stage 3
Stage 4
Alarm Trigger Level
Action Trigger Level
Prediction without buttress

20

Figure 15. Phase 2 North Wall - Monitored diaphragm wall deflection at different stages with trigger
levels
Time-dependent deflection pattern due to softening and deformation of soft clay can be
observed in North diaphragm wall as illustrated in Figure 17. North diaphragm wall moved
progressively toward excavation before completion of raking struts (at 52 days) as can be seen in
Figure 17.

Figure 16. View of buttress-support diaphragm wall prior to temporary bracing removal

365

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
Cum. Displacement , mm.
0

20

40

60

80

100 120 140

0
2
4

Depth, m.

6
8
10
12
14
16
18

0 days
30 days
52 day
56 days
74 days
Prediction

20

Figure 17. Phase 2 Time dependent wall deflection of North diaphragm wall diaphragm wall was
supported only by soil-berm for 52 days before completion of raker installation.

15. IMPLEMENTATION OF CONTINGENCY PLAN

Since deflection of North diaphragm wall (Phase 2) approached action trigger levels, monitoring
frequency was increased and the following contingency measures were implemented on site.
Poured 15cm thick 1m wide lean concrete on the top of the berm along North diaphragm wall to
provide bearing-effect
Installed additional king-post and diagonal struts attached to the raking struts to provide more rigid
support against diaphragm wall
Soil-berm was removed locally in bays followed by construction of wale beam, tie beam and
buttress as shown in Figure 18.
TIE BEAM CONSTRUCTED IN BAYS

TEMPORARY WALER

BUTTRESS

DIAGONAL STRUT

DIAPHRAGM WALL

TEMPORARY KING-POST

CAPPING BEAM

RAKER

BERM

TEMPORARY
STEEL BEAM
PILE
CAP

TED
PLE
COM

B
SLA

Figure 18. Perspective of North diaphragm wall constructed in bay

366

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
Movement of diaphragm wall was observed to be decreased and eventually stabilized by the
above actions. No significant ground settlement was observed in the vicinity of the North diaphragm
wall.

Figure 19. View of raker and soil-berm support soil-berm was removed locally bay by bay

Figure 20. View of buttress-support diaphragm wall after removal of temporary bracing

367

International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)

Pattaya City, Thailand
16. TIME AND COST SAVING FROM VALUE ENGINEERING OPTIONS AND THE
APPLICATION OF OBSERVATIONAL METHOD
Significant cost and time saving were achieved from the value engineering option coupled with
observational method implemented for both Phase 1 and 2. The major savings were achieved by less
operation and material utilized in the following elements of temporary works.
Cancellation of 2nd level raking struts against diaphragm wall for both phases
Modification of bracing system using raking struts with soil-berm support instead of horizontal
struts for North diaphragm wall in Phase 2
17. KEY FACTORS CONTRIBUTED
OBSERVATIONAL METHOD

IN

SUCCESSFUL

APPLICATION

OF

Outcome of a through desk study at post-tender stage provided an effective value engineering
option which offered significant cost and time saving for overall construction program. Effective and
good communications between the design team and construction crew played a key role in successful
application of observational method to completion of the project. Systematic monitoring program with
clear defined trigger criteria was also the important element in implementing the observational method.
This case study reveals that a thin permanent diaphragm wall coupled with effective design and
construction method supplemented by the observational method and robust monitoring program could
offer a logistically and financially attractive solution in construction of underground car park without
disturbing the environment in the prominent historical area of Bangkok.
18. REFERENCES
Clayton C. R. I., 2001. Managing Geotechnical Risk. Improving Productivity in UK Building and
Construction, Institute of Civil Engineers and Thosmas Telford Ltd. UK
Clough, G.W., and ORourke, T.D., 1990. Construction Induced Movements of In-situ Walls. ASCE
Geotechnical Special Publication No. 25, pp. 439-470
Clough C.W. and T.D. ORourke, 1990. Deep Excavation and Tunneling. 7th ICSME, Mexico City
Graftio H., 1936. Some Features in Connection with the Foundation of Svir 3 Hydro-electric Power
Development, Proceedings of the first international conference in Soil Mechanics, Cambridge,
Mass., 1, pp. 284-290
OReilly M.P and New B.M., 1982. Settlements above tunnels in United Kingdom their magnitude
and prediction. Tunneling 82, pp. 173-181, London, IMM
Peck R.B., 1969. Deep Excavation and Tunneling in Soft Ground. 7th ICSME, Mexico City, State of
the Art 3, 225-290
Peck, R. B., 1969. Advantages and limitations of the observational method in applied soil mechanics,
Geotechniques 19, No.2, pp. 171-187
Powderham, A. J., 1996. An overview of the observational method : development in cut and cover
and bored tunneling projects, The Observational Method in Geotechnical Engineering, The
institute of civil engineers, Thomas Telford Publishing, London, pp. 7-24
Zaw Zaw Aye, Dhiraj Karki and Christian Schulz, 2006. Ground Movement Prediction and Building
Damage Risk-Assessment for the Deep Excavations and Tunneling Works in Bangkok Subsoil,
Proceedings of the International Symposium on Underground Excavation and Tunneling, Urban
Tunnel Construction and Protection of Environment, EIT, Bangkok, Thailand, pp. 281-297

368

Proceedings of Geotechnical and Foundation Conference 2012

26-27 September 2012, Bangkok, Thailand

Geotechnical and Foundation

Conference 2012

Diaphragm Wall Support Deep-Excavations in

Bangkok
Narong Thasnanipan, Zaw Zaw Aye, Aung Win Maung,
Chanchai Submaneewong
SEAFCO Public Co., Ltd. Bangkok, Thailand

Thayanan Boonyarak
Hong Kong University of Science and Technology

369

Diaphragm Wall Support Deep-Excavations

In Bangkok Subsoil
Narong Thasnanipan1, Zaw Zaw Aye1, Aung Win Maung1, Thayanan Boonyarak2
Phongpat Kitpayuck1, Nutthapon Thasnanipan1
1
Seafco Public Company Limited: [email protected]
2
Hong Kong University of Science and Technology: [email protected]

ABSTRACT
Diaphragm wall support deep excavation works of some major projects including Bangkok first subway
project are presented in this paper. Available literatures related to diaphragm wall and deep excavation works
in Bangkok are summarized together with recent research works. Problems and difficulties in construction of
diaphragm walls and deep excavation works in early days are reported. The areas of development in
diaphragm wall construction and main factors contributed to these are also presented. An attempt is also made
to briefly discuss the future trend of diaphragm wall construction and deep excavation works in Bangkok. This
paper is intended to serve as a source of reference for the practitioners in the construction industry with regard
to the application of diaphragm walls in deep excavation works in urban area of Bangkok.
INTRODUCTION
As in other major cities, growing land price and need of underground space for commercial and infrastructure
developments necessitated deeper underground excavation works in Bangkok. Diaphragm wall as embedded
earth-retaining structure is one of the most suitable solutions to facilitate the deep excavation works in urban
area of the growing metropolis. The common thickness of diaphragm wall in Bangkok ranges from 0.80m to
1.20m. Though not very common, in some cases, thick diaphragm wall of 1.50m and thin diaphragm wall of
0.50 to 0.60m were also constructed. Toe depths of diaphragm walls are in the range of 16m to over 30m
depths depending on the final elevation of the excavation. This paper aims to provide the summaries of the
technical data on diaphragm wall construction and deep excavations in Bangkok with particular focus on the
progressive development in past decades. Literatures published throughout past 20 years are summarized
along with the recent research works. The information and data presented in this paper are based on the
authors personal experience, observation from other completed projects and from published literatures.
SUBSOIL AND PIEZOMETRIC PROFILES IN BANGKOK
Subsoil profile and the present piezometric drawdown condition of Bangkok are presented in Fig. 1 below. A
typical subsoil profile is relatively consistent in different localities in Bangkok. It is characterized by
alternating layers of clay and sand deposits as shown in the Fig. 1.

371

Pore Pressure (kN/m 2)

Geotechnical
Parameters
0

Bangkok Soft
Clay

400

600

Su= 10-25 KPa

t = 14-16kN/m 3
-

10

Dense to Very
Dense Sand
(Second Sand)

SPT-N= 50-100
t = 21KN/m 3

40

50

e
Lin

50

Su> 200 KPa

= 21KN/m 3
t

ine

Hard Clay

tic L

40

30
SPT-N= 20-40
t = 20kN/m 3

ta
ros

Medium
Dense Sand
(First Sand)

wn
wdo
Dra

30

Su= 50-14020- KPa

t = 17-21kN/m 3

Hyd

Medium to
Stiff Clay

tric
me
iezo

20

60

200

Weathered Crust

P
ual
Act

Depth (m)

10

60

Fig. 1, Subsoil and piezometric profile of Bangkok

THE NEEDS OF DIAPHRAGM WALL FOR DEEP EXCAVATIONS IN BANGKOK
Generally, requirement and selection of basement retaining walls are based on the following factors.

Depth of final excavation of basement shallow or deep excavation

Subsoil condition
Construction sequence and permanent condition of basement structure
Environmental conditions and constraints effect of retaining wall installation and basement
excavation to neighbouring properties and structures
Constructability availability of space and physical condition of existing site

(Peck et. al. 1977) defined that deep excavations are those whose depths are deeper than 6m. Sheet
piles, contiguous-pile walls, secant pile walls and sinking caisson are alternative retaining structures available
in Bangkok for deep excavations. Prevailing subsoil (existence of very soft to soft clay) and ground water
condition are the main factors for using diaphragm walls for deep basement excavation in Bangkok. For most
of the deep excavation works in urban environment, advantages offered by diaphragm walls weigh more
favourably for both technical and economical reasons while other methods have distinct limitations. Major
advantages of diaphragm wall are listed below.

Can be used as permanent structural wall

Water retainable

Can be installed to deeper depths and for load bearing element

Less temporary propping needed

Can be applied for top-down construction method

Rigid structure so that ground movement induced by basement excavation is less than other flexible
retaining wall

Vibration and noise generated from installation of diaphragm wall is less than other methods

Key limitations of diaphragm wall can be summarized as;

Diaphragm wall itself is water retainable structure but not 100% water-proof. Additional measures are
required if basement required high-grade water-proof wall
Availability of limited numbers of specialist diaphragm wall contractors
Finished wall surface is influenced by the subsoil
2

372

DIAPHRAGM WALL CONSTRUCTION METHOD

The diaphragm wall construction method allows for the formation a reinforced concrete wall beneath the
ground surface. The construction of the diaphragm wall is carried out from ground level, panel by panel
forming series of cast-in-place underground RC walls. RC Guide walls of 1m to 1.5m deep are mainly used to
assist the trenching operation. Each panel is excavated by using cable operated mechanical clamshells
suspended from crawler cranes or by more modernized hydraulic grabs. Throughout the excavation Bentonite
slurry is maintained at the top of the excavated trench for wall stability. On completion of excavation, the
Bentonite slurry which may have become contaminated with soil, is cleaned by recycling through de-sanding
equipment. The reinforcement cages are then lowered into the slurry filled trench, with each unit spliced to
the other, to form a continuous cage to the required depth. Tremie pipes are then installed to the base of the
panel and concrete is cast from the panel toe up to the required cut-off. During casting the displaced Bentonite
slurry is drawn off and stored for reuse. Adjacent panels are then excavated using the same procedure. Stopends are used to provide the formation shear-key and continuity between adjoining panels. Fig. 2 is a
summarized demonstration of process involved in diaphragm wall construction. The selection of panel width
is mainly depending upon the available grab size, ground condition and geometric constraints of the basement
footprint. Panel width of 3m to 6m are commonly use in Bangkok. Where practical, it is advisable to use same
panel width for the ease of preparation and handling of the reinforcement cage.

CONCRETE
PULL OUT

GL
GUIDE WALL
WL

STOP-END TUBE

BENTONITE

REINFORCEMENT

FINISHED PANEL

PANEL EXCAVATION

BENTONITE

REINFORCEMENT

EXCAVATED PANEL

FINISHED PANEL

GUIDE WALL

EXCAVATED PANEL

GUIDE WALL

STOP-END TUBE

STAGE 1
EXCAVATION

STAGE 2
REINFORCEMENT

STAGE 3
CONCRETING

STAGE 4
WITHDRAW STOP-END TUBES

Fig. 2 Diaphragm wall construction process

OVERVIEW OF DIAPHRAGM WALL PROJECTS IN BANGKOK
The first diaphragm wall in Bangkok was believed to have been constructed in the late 1970s for the basement
retaining walls of the Bangkok Bank Head Office Tower on Silom Road. Table 1 summarized the diaphragm
wall projects constructed in the early days.

373

Table 1, List of the early diaphragm wall projects in Bangkok

Diaphragm Wall
Dimension (m)
Project Name
Thickness
Depth
Bangkok Bank HQ
0.50
14.00
International Trade Center
0.82
26.00
Srivara Hightech Tower
1.02
26.00
Silom Precious Tower
1.02
28.00
Jewelry Trade Center
0.82
28.00
Oriental Tower
0.82
28.00
Sukhumvit 33 Tower
0.82
26.00
Tharmasat University Library Building
0.80
28.00

Excavation
Depth (m)

Year of
construction

7.00
18.50
17.20
20.00
15.60
15.50
14.00
12.70

1977
1985
1991
1992
1992
1992
1992
1994

DEPTH BELOW EXISTING GROUND (m)

Fig. 3 illustrates the embedded length of diaphragm wall in relation to the excavation depth of the
projects listed in Table 1. It is to be noted that except shallow excavation required for Bangkok Bank Head
Office, diaphragm wall embedded deeper into hard clay layer in all projects constructed in the early days. Fig.
4 shows the embedded length of diaphragm wall in relation to the excavation depth of the projects listed in
Table 2.
From 1991, with the booming construction industry, usage of diaphragm walls for deep basements
of the high-rise buildings in Bangkok has been significantly increased. In year 1997, due to the economic
crisis, with significant decline of property sector, diaphragm wall construction was concentrated only for
infrastructure projects.

Fig. 3, Diagram showing the excavation depths in relation to embedded length of diaphragm walls in early
days (1977 to 1994)

374

DEPTH BELOW EXISTING GROUND (m)

Fig. 4, Diagram showing the excavation depths in relation to embedded length of diaphragm walls from 1997
to 2008
Table 2, List of some diaphragm wall projects constructed from 1997 to 2008 in Bangkok (Basement
Excavations Deeper than 10m)
Diaphragm Wall Dimension (m)
Excavation
Year of
Project Name
Depth (m)
Construction
Thickness
Depth
Charoen Pump Station
Sathorn Pump Station
MWA GMC-7A
Bank of Thailand
Hampton Condominium
Central World Plaza
Boe Bae Bumrung Meung Plaza
Esplanade
Por Teck Tung Office
Noble Remix
Life@Sathorn
Royal Maneeya
Sathorn Square
Renaissance
The Sukhothai Residences
Bangsue Wastewater Treatment
Grand Rama 9 Square
MahaNakhon Hill & Tower
MahaNakhon Retail Cube
UBC III & EM2

0.80
0.80
0.80
0.80
0.80
1.00
0.80
0.80
0.80
0.80
0.80
1.00
0.80

20.00
20.00
21.00
20.00
20.00
18.00
24.00
18.00
16.00
20.00
18.50
30.00
21.00

16.50
16.20
16.00
15.20
12.20
9.00
19.00
10.00
8.00
13.00
10.50

1.00-1.20
1.20-1.50
1.00-1.50
0.6-0.80
0.80
0.80-1.00

20.00-22.0
24.70
22-25.00
16.00-18.0
22.0
18.0-22.0

17.80

15.50

20.15
15.40
18.10
8.00-18.00

1997
1997
2002
2003
2003
2004
2004
2005
2006
2007
2008
2005
2007
2008
2008
2009
2010
2011
2011
2011

375

m
Sil
o

Lu
mp
ini

Bo
nk
ai

Pet
cha
bu
ri
Su
kh
um
vit
Sir
iki
t

Sa
my
an
Hu
aL
am
po
ng

it

/m

IX
Ra
ma

am
ph
en

Ru
am

Th
iam

Pra
cha
rat
B

Su
thi
san

Ra

tch
ada

n
La
dP
rao

Ph
aho
lyo
thi

Su
e
Ba
ng

DEPTH BELOW EXISTING GROUND (m)

Ka
mp
hea
ng
Ph
et
Mo
Ch
it

Extensive application of diaphragm wall for deep excavations can be observed between 1997 and
2001 for the first Bangkok subway project where 18 deep stations and associated structure plus some sections
of cut-and-cover tunnels were constructed by diaphragm wall. (Phienwej et. al., 2006) discussed the difference
between the length of diaphragm wall to depth of excavation between two contracts of the first Bangkok
subway, Chaloemratchamongkhon Line. The authors cited that it is partly due to the difference in design
criteria adopted. Fig. 5 illustrates the embedded length of diaphragm wall in relation to the excavation depth of
MRT Stations. Excavation depth over 30m carried out in Silom Station set the deepest excavation ever done in
Bangkok subsoil.

Fig. 5- Diagram showing the excavation depths in relation to embedded length of diaphragm walls of 18
subway stations of the first Bangkok Mass Rapid Transit Project
DIAPHRAGM WALL CONSTRUCITON PROBLEMS IN THE EARLY DAYS
Basic but extensive problems were experienced in early stages of diaphragm wall construction as summarized
below.

Lack of experienced engineers and foremen

Limited availability and capacity of equipment

6

376

Lack of skills in operation of equipment

Limited knowledge and less advanced techniques in the control of bentonite slurry
Limited knowledge and experience in construction method and related negative impact
Quality of concrete used for tremie concreting
Improper construction and quality control specifications and guidelines for local soil
Improper design for constructability
Limited availability of performance monitoring equipment (e.g. efficient and reliable inclinometer and
other instruments)

DEVELOPMENT IN DIAPHRAGM WALL CONSTRUCTION

Over the past three decades, along with the development of diaphragm wall construction technology in other
parts of the world, equipment, construction technique and design methods as well as better understanding of
construction impact on the performance of embedded retaining wall have significantly improved in Thailand.
Table 3 summarizes the areas of improvement in diaphragm wall construction and main factors contributed to
these developments.
Table 3, Summary of development in diaphragm wall construction in Bangkok
Area of development/
Main factor contributed to development
improvement
Speed of construction

Better equipment, operating skills as well as improved-knowledge in

construction method, management and local soil condition

Size and depth of

diaphragm wall

Better equipment, operating skills as well as improved-knowledge in

construction method, management and local soil condition

Slurry management

Experience from past projects and research studies

Construction impact on
quality and performance

Experience from past projects and research studies

Quality control in
construction process

Experience from past projects and research studies

Quality control test method

and interpretation

Advance equipment, experience from past projects and research studies

In some projects where headroom is limited, special diaphragm wall grabs and equipment were
employed. Fig. 6 shows the specially-designed diaphragm wall equipment used in the project with limited low
headroom.

Fig. 6, Specially designed baby-grab with short-boom for limited low-headroom area
7

377

Diaphragm wall is an ideal solution for deep circular shafts. Though alternative method such as
sinking caisson may offer cheaper direct cost, significant time saving and maximum safety can be achieved by
using diaphragm wall in construction of deep shafts. View of diaphragm wall support deep shaft is presented
in Fig. 7.

Fig.7. Deep circular shaft constructed by diaphragm wall

Diaphragm walls were only solution for deep excavations required for subway stations of the first
Bangkok MRT. (Thasnanipan et. al., 2000) presented the construction of diaphragm wall for one of the largest
subways stations of the MRT. Construction parameters and technical details of the diaphragm wall were
reported by the authors.
DEVELOPMENT IN DEEP EXCAVATION SUPPORTED BY DIAPHRAGM WALLS
In Bangkok, most of the diaphragm wall support deep excavations are carried out by multi-propping or of
braced-excavation. Depending on the excavation depths and sequence, bracing span-lengths commonly ranges
from 4m to 8m. Improvements in diaphragm wall support deep excavation works are obvious and significant
in past decade. Developments in the following areas are notable.

Application of top-down construction method in diaphragm wall support deep excavation works

More complex excavation works to deeper depths

More understanding on excavation induced ground movement and risk of damages to adjacent
structures

Application of ground improvement method to minimize the ground movement

Application of observational method integrated with value engineering option and comprehensive
instrumentation program

More understanding on water-retainable capacity of diaphragm wall

Application of top-down construction method in urban area of Bangkok was presented by
(Thasnanipan et. al., 2006). Along with technical aspects of the top-down application for 19.10m deep
excavation works, the authors highlighted the cost saving gain from shorter construction period and minimum
usage of material for temporary bracing works. View of excavation process with permanent slabs having
braced the excavation can be seen in Fig. 8.

378

Fig.8. Application of top-down method for deep excavation work in Bangkok (Thasnanipan et. al., 2006)
All 18 deep underground stations of the first Bangkok MRT Subway were constructed between 1997
and 2000 by top-down method using diaphragm walls as permanent structural support. Fig. 9 shows the view
of subway station construction with application of top-down method. Skeleton slabs provided large openings
for which offered logistically significant advantage in deep excavation works.

Fig.9. View of subway station construction with application of top-down method

Complex and challenging deep excavation works at Silom MRT Station was reported by (Hall et. al.
2001). The authors demonstrated the complicated construction sequence involved in the deepest excavation
works in Bangkok subsoil supported by NS-Box Diaphragm Wall. The process involved in underpinning of
existing flyover was also described by the authors.
Ground movement prediction and building risk damage assessment for the deep excavation
works can be found in the works of (Aye et. al., 2006). With demonstration of the procedure used in staged
assessment on risk of damage by excavation induced ground movement, the authors proposed a simple method
9

379

to predict both vertical and horizontal surface and subsurface ground movement based on the deflection
profile of diaphragm wall. Fig. 10 shows the proposed method for prediction of subsurface ground movement.

Sio

D-wall

DDoO=2.5H
= HW+ W E

Surface settlement trough

volume, V To
S WO

W
Total deflected
shape volume, V O

HW

HE

Dy
Sub-surface settlement
trough volume, VTy
Settlement
influence zone
deflected volumn V y at depth (H w- Y)

Swy
Siy

SWO

4VO
DO

x
Sio SWO *
DO
D
DY Y * O
HW
VTY VY *

VTO
VO

Fig.10. Demonstration of subsurface settlement prediction from diaphragm wall deflection values proposed by
(Zaw Zaw Aye et. al., 2006)

Fig.11. Typical section of diaphragm wall support deep excavation adjacent to MRT tunnel showing
protection zone, (Thasnanipan et. al., 2006)
(Thasnanipan et. al. 2006) reported an unprecedented case on use of soil-cement columns as ground
improvement to minimize ground movement induced by diaphragm wall support deep excavation adjacent to
existing Bangkok MRT tunnels. The authors presented the technical requirement of the deep excavation work
in the protection zone of MRT tunnels along with actual monitoring results of similar projects in other parts of
the world. This integration of ground improvement application to diaphragm wall deep excavation works is
expected to be more popular in the future as need of basement facilities are increasing along the existing MRT
tunnels. The performance of diaphragm wall and impact on tunnels meeting all technical criteria requireed
imposed by the Metropolitan Rapid Transit Authority of Thailand (MRTA) was subsequently reported by
(Teparaksa et. al., 2006)

10

380

Geotechnical
Risk Management

Observational Method

Design & Plan

Identify Risk
Value Engineering

Analysed Degree
of Risk

Minimize
& Control

Construction Control
Monitoring & Review
Trigger Criteria
Exceeds

Yes

No
Implement
Planned Action

Fig.12. The elements of observational method and risk management

with integration of value engineering in deep excavations,( Aye et. al., 2006)
Application of observational method integrated with value engineering and risk management for
deep excavation works was demonstrated by (Aye et. al., 2006). The authors concluded that well organized
plan with systematic approach of observational method backup by extensive instrumentation and close
cooperation among the involved parties were the key factors contributed to successful completion of the
reported projects.
(Thasnanipan et. al., 2008) presented the comprehensive study of water-resisting deep basement in
Bangkok soil. The authors highlighted the preventive measures for water leakage commonly occurred in
basement with reference to BS8102. The authors also recommended the designers to consider additional
measures to prevent water leakage and dampness for diaphragm wall support excavation as diaphragm wall
itself is not fully water-proof.
(Pongrujikorn, 2006) reported the development of underpass construction in urban area of Thailand
with emphasis on the projects in Bangkok. The author stated that diaphragm wall has been proved to be a
reliable option for most cases in underpass construction.
IMPROVEMENT IN DESIGN PARAMETERS
In the initial stage of introducing diaphragm wall in Thailand, the design concepts and parameters were mainly
based on the available literature from research carried out in other parts of the world. With the passage of time,
design method and selection of parameters for local subsoil were improved as a result of research works
carried out in 1990s. Differences between the behaviour of flexible pile walls and diaphragm walls were well
realized from these studies. Development in powerful computer facilities and commercially available FEM
programs are key factors contributed to the advancement in analysis of diaphragm wall support deep
excavations.
Detailed advance FEM analyses of diaphragm wall support deep excavations in Bangkok subsoil can
be found in the work of (Balasubramaniam et. al., 1992). The authors demonstrated the FEM simulation for
both top-down and conventional braced excavation with temporary support. The authors concluded that
behavior of supported excavation in soft clay is mainly controlled by the cantilever mode of deformations and
it is 80% of the maximum deflection of diaphragm wall.
11

381

With the peak of construction-boom in Thailand, particularly in Bangkok, large numbers of

diaphragm wall support deep excavations were systematically monitored by geotechnical instrumentation
throughout 1990s which provided better understanding on behavior of this type of embedded retaining wall.
Researches focused on the design parameters and methods were published based on these data. With improved
construction and design methods, diaphragm wall came to be regarded as reliable excavation support retaining
structure in the construction industry of Thailand. With more confidence on soil parameters selection and
better understanding on behaviour of diaphragm wall, significant cost savings were achieved by more efficient
design. This can be seen by shorter embedded depths of diaphragm walls in recent projects than those in early
days as shown in Fig. 4 in comparison with Fig. 3.
(Teparaksa et. al., 1999), back-analysed the soil stiffness parameters of the soft and stiff Bangkok
clay for diaphragm wall support excavation based on three cases studies. The authors proposed the soil
stiffness values for soft and stiff clay in terms of Young modulus in relation to undrained shear strength which
provided useful reference for the design engineers involved in diaphragm wall and deep excavation works.
IMPROVEMENT IN APPLICATION OF INSTRUMENTATION
In early days, use of instrumentation in diaphragm wall support excavations was limited. Increasing needs of
deep excavation works adjacent to sensitive buildings and structures called for more comprehensive
instrumentation and monitoring program.
(Thasnanipan et. al., 1998) reported the use of inclinometer, tiltmeter, vertical beam sensor, vibrating
wire strain gauge, settlement plate and earth pressure gauge in monitoring of diaphragm wall support deep
excavations and adjacent sensitive buildings at the bank of Chao Phraya River. The important role of
instrumentation for deep excavation in presence of unbalance loading condition was emphasized by the
authors.
Ignorance of instrumentation data resulted in catastrophe collapse of deep excavation for Inlet
Pumping Station was discussed by (Teparaksa et. al., 1999). The authors presented the findings of
investigation into the causes of failure showing excessive movement recorded by inclinometer prior to the
collapse.
Extensive application of geotechnical instrumentation was more evident in construction of 18 deep
subway stations for the first Bangkok MRT between 1997 and 2001. (Aye et. al., 2006) summarized the
instrumentation program used in monitoring of ground movement induced by deep station excavations. The
authors discussed the systematic approach of damage risk assessment of adjacent existing buildings and
structures with application of extensive instrumentation.
(Thasnanipan et. al., 2004) reported the effective application of inclinometer monitoring data in
construction of underground car park in the historical area of Bangkok. The authors described that thin
diaphragm wall with buttress support (Fig. 13) applied in this project was the first of its kind in Bangkok as
described by the authors.

Fig.13. View of thin diaphragm wall supported by buttress (Thasnanipan et. al., 2004)
12

382

The authors presented a good data set of diaphragm wall movement profile showing time-dependent
progressive deflection pattern due to softening and deformation of soft clay caused by delay in installation of
raker support as shown in Fig. 14. The authors demonstrated by comparing actual monitoring results with preestablish trigger levels contributed to the understanding of the condition of the excavation as work progressed.
This approach served as the basis to judge whether contingency actions are to be implemented. To ensure the
safety of the excavation, the contingency measures exercised upon diaphragm wall deflection reaching preset
trigger levels were reported by the authors.
The actual monitoring data of different type of instrumentation presented in above literatures served
as valuable information and technical references for all parties involved in diaphragm wall and deep
excavations industry.

Cum. Displacement , mm.

0

20

40

60

80

100 120 140

0
2
4

Depth, m.

6
8
10
12
14
16
18

0 days
30 days
52 day
56 days
74 days
Prediction

20

Fig.14. Time dependent wall deflection of diaphragm wall diaphragm wall was supported only by soil-berm
for 52 days before completion of raker installation (Thasnanipan et. al., 2004)
OBSERVATION OF GROUNDWATER CONDITION IN RECENT PROJECTS
Recently, deep excavations deeper than 18.0m in depth at two projects for basement constructions being
5.25km apart in Bangkok subsoil encountered presence of groundwater after reaching about 17.5m. Subsoil
information obtained by soil boring in these projects indicated a series of stiff silty and sandy clay between
15m and 24m, overlying a dense sand layer. Construction progress was significantly affected by presence of
groundwater in silty clay. Fully saturated silty clay turned liquid when disturbed by excavation activities,
causing difficulties in casting a concrete blinding layer for foundations construction. Dewatering and ground
improvement to the subgrade layer below foundation mat and footings needed to be carried out. Bearing
capacity of disturbed silty clay with high water content is very low so that cobble size concrete blocks and
crushed rocks to be compacted in to the soil for improving the subgrade. Then a lean concrete blinding layer
was cast on the subgrade part by part while dewatering.
A general piezometric profile in Bangkok shown in Figure 1 indicates a drawdown at -20m. Additional
peizometric readings from the piezometers installed at depths in the range of 28.0m to 33.0m and at 40m from
other sites in Bangkok indicate that piezometrice levels are around -15.0m and -23.0m respectively.
Piezometers installed at depths 28.0m to 33.0m shows piezometric levels increase to -12.0m within six
months. However, the groundwater encountered in the above mentioned sites was considered to be perched
water in semi permeable soils layer being vertically cut off by diaphragm walls, with an assumption that the
groundwater level was reduced gradually by pumping out.
13

383

Fig.15. Groundwater encountered in basement excavations deeper than -17.50m.

CONCLUSIONS
According to the authors experience as deep-foundation contractors and observation as researchers,
development in both construction and design aspects of diaphragm wall and deep excavations in Bangkok in
past decades were significant. With recognition of technical and economical advantages of using this type of
embedded retaining walls by local practitioners, it is expected that they will be more popular in the future
construction industry of Thailand. However, in the authors opinion, there is much work to be done with
particular focus on constructability issues, concrete technology for diaphragm walls, reliable but cost-effective
quality control testing and value-engineering. Starting from the planning stage, site investigation, design,
construction and inspection should be integrated so that designers, contractors and construction inspectors can
participate as a team with a common goal. Appropriate and practical specifications should be established
jointly by these parties for local soil conditions and construction methods. As a commercial and political
center of the country with a rapidly growing population, Bangkok will continue to witness more underground
construction for its infrastructure development in which diaphragm wall support excavation will clearly play a
great role.
REFERENCES
Balasubramaniam, A.S, Longanathan, N., Fernando, G.S., Indraranta, B., Phien-wej, N., Bergado, D.T.,
Hongjo, Y. (1992) Advance Geotechnical Analysis, Finite Element Analysis Coupled with Critical State
Theories Using CRISP Computer Program, Asian Institute of Technology, Thailand
Hall, J.T. , Munson, S.R., Sakai, K. (2006) Silom Underground Station-Bangkok Design for Unusual Planning
Challenges, Proceedings of the Symposium
on Underground Construction of Bangkok MRT
Chaloemratchamongkhon Line, EIT, MRTA, Bangkok, Thailand, pp. 58-68
Peck, R.B., Hanson, W. E. and Thornburn, T.H. (1977). Foundation Engineering, John Wiley & Sons, New
York
Phienwej, N., Photayanuvat, C., Suwansawat, S. and Mcconochie, D., (2006) Geotechnical Aspects in Design
and Construction of Chaloemratchamongkhon Underground MRT, Proceedings of the International
Symposium on Underground Excavation and Tunneling, Urban Tunnel Construction for Protection of
Environment, 2-4 February 2006, Bangkok, Thailand, pp. 171-180

14

384

Pongrujikorn, T., (2006) Development and Use of Underground Strucutre as Underpass in Urban Area in
Thailand, Proceedings of the International Symposium on Underground Excavation and Tunneling, Urban
Tunnel Construction for Protection of Environment, 2-4 February 2006, Bangkok, Thailand, pp. 441-448
Teparaksa, W., Thasnanipan, N., Maung, A. W. and Tanseng, P. (1999), Lessons from the collapse during
construction of an inlet pumping station: Geotechnical instrumentation aspect, 5th International Symposium
on Field Measurements in Geomechanics, Singapore, p.p. 247-252.
Teparaksa, W., Thasananipan, N., and Tanseng, P. (1999), Analysis of Lateral Wall Movement for Deep
Braced Excavation of Bangkok Subsoils, Civil and Environmental Engineering Conference, Bangkok,
Thailand, p.p. II-67 to II-76.
Teparaksa W., Boonsong C., Submaneewong C., Wisuttipat N., (2006) MRTs Tunnel Deformation due to
Deep Basement Excavation in MRTs Protection Zone, Proceedings of the 11th National Convention on Civil
Engineering, 20-22 April 2006, Phuket, Thailand (in Thai), pp. GTE-095
Thasnanipan, N., Maung, A.W., Tanseng, P., Wei, S.H., (1998), Performance of a Braced Excavation in
Bangkok Clay, Diaphragm Wall Subject to Unbalanced Loading Conditions, Thirteenth Southeast Asian
Geotechnical Conference, Taipei, Taiwan, ROC, p.p.655-660.
Thasnanipan, N., Maung, A.W. and Ganeshan, B., (2000), Diaphragm Wall and Barrette Construction for
Thiam Ruam Mit Station Box, MRT Chaloem Ratchamongkhon Line, Bangkok, An International Conference
on Geotechnical & Geological Engineering, Melbourne, Australia, 19-24 November 2000, pp. 157
Tanseng P., Aye, Z. Z. and Submaneewong, C. (2001), Monitoring of Diaphragm Wall Displacement and
Associated Ground Movement, Braced Excavation Adjacent to Historical Building at the Bank of Chao
Phraya River, Proceedings of the Seventh National Convention on Civil Engineering, Bangkok, Thailand, pp.
GTE-53-60
Thasnanipan, N. , Aye, Z. Z. and Submaneewong, C. (2004), Construction of Diaphragm Wall Support
Underground Car Park in Historical Area of Bangkok Proceedings of Fifth International Conference on Case
Histories in Geotechnical Engineering, April 13-17 2004, New York, U.S.A. pp. 5.66
Thasnanipan, N. , Aye, Z. Z., Submaneewong, C. and Boonyarak, T. (2006) Application of Top-down
Construction Method for Deep Excavations in Urban Area of Bangkok, Proceedings of the International
Symposium on Underground Excavation and Tunneling, Urban Tunnel Construction for Protection of
Environment, 2-4 February 2006, Bangkok, Thailand, pp. 383-396
Thasnanipan, N., Maung, A.W., Aye, Z. Z., Submaneewong, C. and Boonyarak, T. (2006) Construction of
diaphragm wall for basement excavation adjacent to tunnels in Bangkok subsoil, Proceedings of the
International Symposium on Underground Excavation and Tunneling, Urban Tunnel Construction for
Protection of Environment, 2-4 February 2006, Bangkok, Thailand, pp. 371-382
Zaw Zaw Aye , Dhiraj Karki and Christian Schulz [2006]. Ground Movement Prediction and Building
Damage Risk-Assessment for the Deep Excavations and Tunneling Works in Bangkok Subsoil, Proceedings of
the International Symposium on Underground Excavation and Tunneling, Urban Tunnel Construction and
Protection of Environment, EIT, Bangkok, Thailand, , pp. 281-297

15

385

Proceedings of Geotechnical and Foundation Conference 2012

26-27 September 2012, Bangkok, Thailand

Geotechnical and Foundation

Conference 2012

Underpass Construction in Congestive Area of

Bangkok
Zaw Zaw Aye, Phongpat Kitpayuck
Sedtha Mahasantipiya, Nutthapon Thasnanipan
SEAFCO Public Co., Ltd. Bangkok, Thailand

Thayanan Boonyarak
Hong Kong University of Science and Technology

387

Underpass Construction in Congestive Area of Bangkok

Zaw Zaw Aye1, Thayanan Boonyarak2, Phongpat Kitpayuck1
Sedtha Mahasantipiya1, Nutthapon Thasnanipan1
1

Seafco Public Company Limited: [email protected]

Hong Kong University of Science and Technology: [email protected]

Diaphragm wall has been used as permanent retaining wall for a vehicle underpass at major
intersection in urban area of Bangkok besides deep basement structures of high-rise buildings,
subway stations, deep shafts and other forms of underground structures,. This paper presents the
key technical features of recently completed vehicle underpass along one of the busiest roads in
Bangkok. The design aspects of earth-retaining structure integrated with top-down construction
method are reported. Predicted and measured deflections of diaphragm wall at different stages of
construction are presented. Practical construction method adopted under the existing flyover is
highlighted. Difficulties encountered during the construction are briefly described. This project
marked the first application of post-tension roof slabs along with top-down construction
technique for vehicle underpass construction by cut-and-cover method in Thailand.
1.

Introduction

To combat appalling traffic congestion, Bangkok Metropolitan Authority has been implementing
various measures in both urban and sub-urban areas. Flyovers are frequently used to improve the
traffic flow along the road at many major intersections. Though underpass is not commonly
used, it is the only solution in some critical traffic jam areas. The first underpass in the form of
depressed-road for rail-road junction was believed to be constructed about 35 years ago at
Bangsue, northern outskirt of Bangkok. Several years later, vehicle underpass was constructed
using cast-in-place diaphragm wall by Bangkok Metropolitan Authority (BMA) at Din Daeng
intersection in 1994. Since then more than 20 underpasses have been constructed by BMA and
other agencies in Bangkok and other major cities in the country.
This paper presents the design and construction of recently completed vehicle underpass along
one of the busiest roads in Bangkok. The presenting project in this paper demonstrated the
effective use of underground space in congestive urban areas of Bangkok and exemplified good
solutions for environmental preservation.
2.

Project Requirement and Major Constraints

The project works involved construction of a vehicle-underpass and associated utility diversion
and road expansion at one of the major intersections in the eastern part of Bangkok. Figure 1
shows the alignment and location of the underpass overlaid on aerial photo of the site (modified
from GoogleEarth). Four-lane underpass was to construct in the form of 758 m long cut-andcover tunnel along six-lanes existing Srinakarin Road. As can be seen in the figure, at the
intersection, tunnel section of underpass was to construct under existing flyover of Udomsuk
Road.

389

As the main six-lanes Srinakarin Road is one of the major roads linking eastern parts and inner
Bangkok, traffic density along this road is high. Presence of two major shopping malls in the
close proximity to the intersection is also another factor contributed to congestive traffic
condition in the area.
Location of the project in the heavy traffic area and need of construction under the existing
flyover with limited headroom therefore posed major constraints, which called for the need of
careful consideration in establishing the design principles and sequence of construction. As the
construction is to be carried out along the major public road, ground movement induced by cutand-cover tunnel excavation must be minimized.

Srinakarin Rd

Udomsuk Rd

Flyover

4-Lanes Vehicle Underpass

Total length of underpass 758 m

Figure 1 Alignment and location of vehicle underpass (aerial photograph modified from
GoogleEarth)
3.

Sub-soil Condition

Similar to other localities in Bangkok a typical subsoil profile at the site is characterized by the
alternating layers of clay and sand deposits as shown in Figure2 . Undrained shear strength (Su)
and SPT N-value obtained from 5 SI boreholes were plotted and design line was derived as
depicted in Figure 2. Typical subsoil profile at the site is characterized by thick Bangkok soft
clay layer at the top followed by thin layer of medium clay underlying by alternative of stiff clay
and dense sand layers. Undrained shear strength of soft clay is obtained by using unconfined
compressive strength test from Shelby tube sample. In stiff to hard clay layer, correlation of
undrained shear strength with SPT N-value is adopted where undrained shear strength is equals
to 6.85N (kPa) for high plasticity clay (Sambhandaraksa & Pitupakorn. 1985). Angle of internal
friction () of dense sand is estimated to be 32 by using SPT correlation (Peck et al., 1974).The
design soil parameters are tabulated in Table 1.

390

1
0

0.00

Fill

5.00

Depth from ground surface (m)

10
15
20
25
30
35
40
45
50

Undrained Shear Strength, Su (kN/m2)

50
100
150
200

Soft clay

10.00

Medium stiff clay

15.00
20.00

Very stiff
clay

25.00
30.00

Dense sand

35.00

Very stiff to
hardclay

40.00

Very dense sand

45.00
50.00

20

40

60

80

SPT N-Value (blows / ft)

Su BH-1
Su BH-3
Su BH-5
SPT BH-1
SPT BH-3
SPT BH-5

Figure 2 Soil profile, undrained shear strength and SPT N-value

Su BH-2
Su BH-4
Design Su
SPT BH-2
SPT BH-4
Design (SPT)

Table 1 Design soil parameters

Depth
Soil type
(m)
Fill
0-2
Soft clay 1
2-10
Soft clay 2
10-14
Medium stiff clay
14-18
Stiff clay
18-20

t
(kN/m3)
18.5
16.5
17.5
19.0
19.0

Su
(kPa)
25
17
22
40
80

(degree)
0
0
0
0
0

Very stiff clay

20-25

19.5

120

Hard clay

25-27

20.0

180

Dense sand

27-32

20.0

32

Hard clay

32-40

20.0

180

Very dense sand

40-45

20.0

32

391

4.

Review of Tender Stage Design and Modification

Review of the practical construction method was undertaken by the contractors in-house design
team. Rigorous attention to detail of the design concept and constructability was made in the preconstruction discussions among the client, consultants, design engineers and construction team.
Original tender called for the conventional construction method to complete the excavation to
final excavation depth and construct the roof-slab in the tunnel section of the underpass by using
precast pre-stressed concrete panels. Installation of 2.5m x 18.8m heavy precast panels
underneath the existing flyover was considered difficult and high risk involved. After reviewing
different options with consideration of the constraints imposed on site and constructability, topdown construction with post-tension slabs as roof-slab of the tunnel section was selected.
4.1 Numerical analysis for diaphragm wall design
Figure 3 illustrates an example of finite element mesh of deep approach with sump pit. In other
zone, basic geometry of the mesh follows that of excavation depth, wall depth and construction
sequence of each zone. Numerical analyses are carried out by using PLAXIS V7.2 (Brinkgreve
and Vermeer, 1998) finite element program. Plane strain total stress analyses using elasto-plastic
with Mohr-Coulomb failure criterion constitutive model are adopted. Undrained shear strength
of clay (Su) and internal friction angle of sand () are selected according to subsoil properties as
previously discussed. Youngs modulus ratio (Eu/Su) of each clay ranges from 250 to 750 which
is estimated based on expected shear strain of diaphragm wall. Boundary condition consists of
roller support applied at both vertical sides and pin support applied at the bottom of the mesh.
Numerical procedure follows the construction sequence in the field conditions. The primary
objective of numerical analysis is to verify structural capacity and overall stability during
construction. It should be noted that time dependent behavior of clay is not simulated.

Depth (m)

10

25

45
Figure 3 Finite element mesh of deep approach with sump pit

392

5.

Construction of Diaphragm Wall, Barrettes and Bored Piles

Figure 4 shows the diaphragm walls, bored piles and barrettes layout plan. Diaphragm wall was
designed as earth-retaining wall for the entire section of the underpass. Diaphragm wall having
800mm thick founded at 25m below ground level (B.G.L) was constructed in three phases. Both
Phase 1 (north section) and Phase 2 (south section) were constructed in two stages (different
working stages for east and west diaphragm walls) to meet the traffic diversion requirement.
North section of the underpass diaphragm wall was constructed in Phase 1 in the initial stage of
the project. In Phase 2, south section of diaphragm wall was constructed prior to construction of
cross-wall for a deep sump-pit - diaphragm wall with 6m deep cut-off-level perpendicular to
longitudinal alignment of the main underpass diaphragm wall.
Diaphragm wall
0.8m x 25m
Centerline of
Udomsuk Rd Flyover

Barrette 0.8 x 2.7 m x 29 m deep

constructed underneath flyover
Bored pile 1.0 m x 29 m deep

Phase 1

Phase 2
Phase 3
(under flyover)
Note: only diaphragm wall in tunnel zone is shown for clarity

Figure 4 Diaphragm wall, bored pile and barrette layout plan of vehicle underpass
As the tunnel section of the underpass at the intersection was to be constructed under limited
headroom due to the presence of existing flyover, careful planning was needed for Phase 3
diaphragm wall construction. Main factors were the limited time available for traffic diversion at
the intersection and the need of special equipment to practically construct the diaphragm wall
under the flyover where headroom was only 6m. Specially-designed mechanical grab was used
to construct the diaphragm wall in this low-headroom section as shown in Figure 5. Another
major challenge involved in construction of diaphragm wall under the existing flyover was the
obstruction of existing water supply pipe at 8m below ground level along the diaphragm wall
alignment. Removing the pipe buried 8m below ground level by open-cut or sheet-pile supported
excavation option was considered time consuming, unpractical as well as risky, taking the
consequences involved into account (unpractical and risky to have 8m open-cut excavation in
soft clay, unpractical to install 16m deep sheet pile low headroom. After reviewing different
options to remove the obstructed water supply pipe, decision was made to remove the pipe by
sending a man inside the pipe - to cut the obstructed part in smaller pieces and close the pipe to
stop the flow of bentonite into the pipe during diaphragm wall construction. Risk assessment and
safety measures with contingency plan were made before actual pipe removal was done. The
obstructed section of the buried pipe was successfully removed.

393

Figure 5 Special-design diaphragm wall grab working under existing flyover

Figure 6 demonstrates the 3D perspective of stanchions embedded in bored pile and barrettes.
Total 43 bored piles of 1000mm diameter founded at depth 29m below ground level were
designed to support roof-slab of tunnel section. Bored piles were constructed by wet-processes
method using polymer-based slurry. Steel stanchions were installed in bored piles to facilitate
top-down construction. As it was not practical to utilize bored pile drilling machine underneath
the flyover, within the low-headroom section of the underpass, three barrettes were installed to
replace bored piles.

Figure 6 Three-Dimensional perspective of stanchions embedded in bored pile and barrettes

394

6.

Application of top-down method integrated with post-tension slab

Figure 7 shows the plan of cut-and-cover tunnel constructed by using top-down method
integrated with post-tension slab. Total length of tunnel zone is approximately 180 m along the
longitudinal profile of underpass or about 25% of overall underpass length.
As discussed in the previous section, top-down construction requires modifications from the
original design. Instead of using temporary prop to support the diaphragm wall during
excavation of tunnel section, roof-slab was used as temporary support. Four openings were
provided at an evenly spacing of about 40 m for access of soil excavation underneath the roofslab.
Post-tension
roof slab
y40 m

y40 m

Slab opening
during construction

Figure 7 Plan view of tunnel zone showing post-tension roof slab constructed prior to
excavation and openings for excavation works
Construction sequences of tunnel zone are shown in Figure 8. Major construction activities in
each stage are given as follows.
Stage 1: Construct diaphragm wall and bored pile with embedded temporary stanchion
Stage 2: Construct upper part of median wall (spine-beam) on the top of stanchion
- Construct roof slab and apply post-tension force
- Excavate to formation level
Stage 3: Construct base slab
- Construct median wall between temporary stanchions
- Remove temporary stanchion and construct the remaining median wall

395

Stage 1

Stage 2

Stage 3

Figure 8 Construction sequences in tunnel zone

6.1 Comparison of roof slab between original tender design and alternative design
Comparison between tender design and alternative design of roof slab are shown in Figure 9.
Figure 9a shows precast prestress slab in the tender design. The width of each precast slab is 2.5
m. A gap of 0.2 m between each precast slab is filled with concrete after placing of precast slab.
Figure 9b shows post-tension slab used as alternative in this project. Each construction joint is
located at a distance of about 12.5 m. By using alternative design, numbers of joints were
significantly reduced. A major benefit of reduction of number of joint is possibility of water
leakage from roof slab through joint is dramatically reduced. In addition, using the roof slab as
permanent prop provide significant larger axial stiffness compared with when temporary steel
prop is used. The evidence of increasing axial stiffness on performance of diaphragm wall is
discussed later in the following section.

Tendon
CONSTRUCTION JOINT

CONSTRUCTION JOINT

(a)

(b)

CONSTRUCTION JOINT

Grouting hose

(c)

Typical tendon

(d)
Figure 9 (a) Precast prestress slab; original design (b) post-tension slab; actual application (c)
typical tendon profile (d) tendon and grouting hose

396

Before starting excavation underneath the roof slab, tension is applied to tendon embedded
inside the roof slab (Fig. 9c). Subsequently, grout is injected via grouting hose. Figure 9d shows
a typical tendon profile which is similar in both original and alternative design. A profile of
tendon was designed to resist loading condition of a beam with end and middle supports. In
order to apply roof slab as a permanent prop for excavation, rigorous structural capacity
checking for combined bending moment and axial force in the slab is required. Figure 10a and
10b shows a typical construction of post-tension roof slab.

(b)

(a)

Figure 10 (a) Tendon installation for roof-slab (b) grouting after post-tensioning the slab
6.2 Excavation underneath the roof slab
Figure 11a and 11b show slab opening and temporary prop in tunnel zone. Spacing between each
opening was determined by optimizing efficiency of excavation and providing space to facilitate
the construction. Excavation was carried by using a small excavator inside the tunnel (Fig. 11a)
or using a long reach boom excavator on the roof slab (Fig. 11b). Presence of roof slab provides
additional space and increase efficiency of excavation works in comparison with the original
design where temporary working platform is to be used.

(b)
(a)
Figure 11 Slab opening with temporary support (a) using a small excavator inside the tunnel; (b)
using a long reach boom excavator on the roof slab

397

6.3 Temporary support of the roof slab

Figure 12a shows temporary support of roof slab. Load from the roof slab is transferred to the
upper part of the median wall which works as a spine-beam. Additional reinforcement in the
upper part of median wall was required during construction stage. The upper part of the wall was
supported by stanchions made of steel H 350x350-137 kg/m embedded in 1.0 m diameter bored
pile. The remaining part of median wall was constructed after completion of base slab (Fig. 12b).
After completion of median wall, temporary stanchions were removed since vertical load from
roof slab was designed to be transferred through the median wall.

Stanchion

Upper part of
median wall
Stanchion
Median wall

(b)

(a)

Figure 12 (a) Temporary stanchions supporting post-tension roof slab (b) median wall and
stanchion
7. Performance of diaphragm wall
Figure 13 shows location of instrumentation in this project. Instrumentation consists of
inclinometer, surface settlement point, strain gauge and piezometer. Instrumentation in
diaphragm is installed in three zones which are tunnel, deep approach and deep approach with
sump pit. Each zone of monitoring is selected based on its construction sequences. Table 2
summarizes construction sequence in each zone according to the location of instrumentation
which is T1, T2, D1, D2 and DS1. In this paper, only some of measured results are reported for
conciseness of paper.
D1

T1
NO-43

NO-44

NO-45

NO-46

NO-47

NO-48

NO-49

NO-50

NO-51

NO-52

NO-53

NO-54

NO-55

NO-56

NO-57

NO-58

NO-59

NO-60

NO-61

NO-63

NO-62

NO-64

G
3.000

3.000

CWN-2

4.500

4.500

CWN-3

CLE-69

SO-70

SO-71

SO-72

SO-73

SO-74

SO-75

SO-76

SO-77

SO-78

SO-79

SO-80

SO-81

SO-82

SO-83

SO-84

SO-85

SO-86

SO-87

SO-88

SO-89

SO-90

SO-91

SO-92

SO-93

SO-94

SO-95

Sonic Logging Test

SO-96

SO-97

Sonic Logging Test

SO-98

3.350

3.350

CWS-7

CWS-8

3.350

CWS-9

3.000

3.000

Sonic Logging Test

CWS-10

G
NI-44

S2

Sonic Logging Test

S1
NI-43

CLE-68

CWS-6

CWN-4

CWN-5

NO-67

NO-66

NO-65

Sonic Logging Test

CWN-1

NI-45

G
S1

NI-46

NI-47

NI-48

NI-49

NI-50

NI-51

NI-52

NI-53

NI-54

NI-55

NI-56

NI-57

NI-58

NI-59

S3

NI-60

NI-61

S2 NI-62

S1

P
NI-63

NI-64

NI-65

NI-66

NI-67

CLW-68

CLW-69

SI-70

SI-71

SI-72

SI-73

SI-74

SI-75

SI-76

SI-77

SI-78

SI-79

SI-80

SI-81

SI-82

SI-83

SI-84

SI-85

SI-86

SI-87

SI-88

SI-89

SI-90

SI-91

S4

S5

S5

S6

S6

DS1

SI-92

SI-93

S2 SI-94
S3

S3

S4

S4
S5
S6

D2

T2

Figure 13 location of instrumentation

398

Sonic Logging Test

I
P

SI-95

SI-96

SI-97

SI-98

Table 2 Summary of construction sequence in each zone

Tunnel (T1 & T2)
Deep approach (D1 & D2)

Deep approach with sump pit

(DS1)
Construct roof slab
Excavate to -1.5 m
Excavate to -1.5 m
Excavate to final depth Install temporary bracing at Install temporary bracing at
(-6 m)
-1.0 m
-1.0 m
Construct base slab
Excavate to final depth (-6 m)
Excavate to -5.5 m
Construct base slab
Install temporary bracing at -5 m
Remove temporary bracing
Excavate to final depth (-9.5 m)
Construct sump pit slab
Construct base slab
Remove temporary bracing
7.2 Lateral movement of diaphragm wall
Lateral movement of diaphragm wall in three zones are shown in Figure 14. Measured results
from three major construction stages which are excavation to final depth, base slab construction
and four months after completion of base slab are shown. Computed results at the final stage of
construction in each zone are compared with the measured results. Maximum lateral
displacement of diaphragm wall is 24, 41 and 30 mm in tunnel, deep approach and deep
approach with sump pit zone, respectively.
By comparing diaphragm wall performance in two different construction sequences with the
same diaphragm wall geometry and excavation depth, it is apparent that lateral movement in
bottom-up construction (Fig. 14b) is about 2 times larger than that in top-down construction
(Fig. 14a). Not only magnitude of wall movement is larger in bottom-up excavation compared
with that of top-down, but also the rate of movement. Rate of wall movement after final
excavation depth reached in deep approach is about one order of magnitude larger than that in
tunnel. The major reason of this observation is considered to be contributed by higher stiffness
of prop at the top of diaphragm wall. Prop stiffness in tunnel section which is the post-tension
roof slab is about 16 times higher than that of temporary steel prop used in deep approach
section. A major difference in wall deflection was observed at the top of the wall. In tunnel
section, only a small deflection was observed at the top of the diaphragm wall and maximum
wall deflection was observed at 9 m below ground surface. In deep approach section, movement
at the top of the wall was about 24 mm due to shortening of the bracing. Subsequently, earth
pressure is transferred to the formation level (6 m below ground surface) causing maximum
movement at depth 7 m.
In deep approach with sump pit (Fig. 14c), although it was constructed by using bottom-up
method and excavation depth (9.5 m) is deeper than that of deep approach - maximum wall
movement is about 73% of that in deep approach. Two major reasons are identified to explain
this observation. First, there are two levels of temporary steel prop in deep approach with sump
pit compared resulting in 25% smaller average prop spacing (H/number of prop) compared with
that in deep approach. Second, excavation length of sump pit is only 18 m in longitudinal
direction of the wall which is almost the same as excavation width (17 m). Thus geometry

399

effects may contribute to reducing of lateral wall movement compared with plane strain
condition in deep approach.
Computed result overestimates wall movement in tunnel zone while it underestimates that in
deep approach zone. The differences are expected to be due to two major causes. First, constant
soil stiffness is adopted. The reduction of soil stiffness with increasing shear strain is not
simulated. Second, time dependent soil properties are not modeled. It is apparent that time
dependent behaviour in deep approach is the major component of wall movement. Nevertheless,
computed results reasonably agree with measured result in terms of overall trend.

10

20

30

40

50

10

20

Depth (m)

10

Displacement (mm)

Displacement (mm)

Displacement (mm)

30

40

50

10

20

30

40

50

0.34
mm/day
0.14
mm/day

0.03
mm/day

15

20

25

Measured [Before
base slab]
Measured [Base slab
construction]
Measured [4 months
after base slab]
Computed [Final
stage]

Measured [Before
base slab]

Measured [Before
base slab]

Measured [Base slab

construction]
Measured [4 months
after base slab]

Measured [Base slab

construction]
Measured [4 months
after base slab]

Computed [Final
stage]

Computed [Final
stage]

(b)

(a)

(c)

Figure 14 Lateral wall movement in (a) tunnel zone ; top-down method (T1); (b) deep approach
bottom-up method with 1-layer bracing (D2); (c) deep approach with sump pit bottom-up
method with 2-layers bracing (DS1)
7.3 Comparison of measured wall movement with other case histories
Figure 15 shows maximum measured wall movement by using two different construction
methods and sequences reported previously in other case histories (Phienwej et al., 1995).
Previous study conducted by Phienwej et al., 1995 suggests that maximum lateral wall
movement using bottom-up construction can be up to 0.4% of excavation depth (H). On the
other hand, wall movement in top-down construction is estimated to be only 0.2%H.
In this study, movement in bottom-up construction in deep approach and top-down construction
in tunnel exceeds 0.4%H and 0.2%H line, respectively. Wall movement using top-down
construction is closer to 0.27%H line proposed by Wang et al., 2010 based on measured field
results in Shanghai soft clay. Formation level of excavation in this study is in soft clay and only
one prop is provided. Thus, passive resistance is mainly from soft clay causing larger lateral
movement compared with other case histories. Another possible reason of large lateral wall

400

movement is mainly due to the construction time or time-dependent movement as construction

of underpass naturally takes more time due to traffic diversion and other factors.

Max. lateral movement, max (mm)

60
50
40
30
20
Bottom-up [Phienwej et al., 1995]
Top-down [Phienwej et al., 1995]

10
0

Soft clay
depth

10

Bottom-up [This study]

Top-down [This study]

15

Max. excavation depth, H (m)

20

25

Figure 15 Measured lateral wall movement compared with other case studies in Bangkok
[Modified from Phienwej et al., 1995]
7.4 Bending moment of diaphragm wall
Figure 16 shows comparison of measured and computed bending moment of diaphragm wall.
Measured bending moment is from vibrating wire strain gauges (VWSG) installed at
reinforcement of diaphragm wall at depth of 6 and 17.5 m where bending moment is expected to
be large. Measured strain is larger than 150 thus wall flexural stiffness reduction factor of
80% is adopted. Deduced bending moment is from two times differentiation of measured lateral
wall movement (Fig. 16) using the same stiffness reduction factor as measured from strain
gauge.
In tunnel zone (Fig. 16a), measured bending moment at depth 6 m is slightly smaller than
deduced bending moment suggesting that deduced bending moment is reasonable. It should be
noticed that deduced bending moment exceeds computed bending moment from 8 to 13 m below
ground surface due to smaller stiffness in soft to medium clay in numerical analysis.
In deep approach zone (Fig. 16b), measured bending moment at 6 m from strain gauge is about
30% larger than that in tunnel zone. Deduced bending moment in deep approach also confirm
the measured result. Bending moment from depth 3 to 7 m significantly exceeds the computed
bending moment due to lacking of time dependent properties of clay in numerical analysis.

401

At depth 17.5 m, measured, deduced and computed bending appears to be in a good agreement.
In addition, deduced bending moments in both tunnel and deep approach zones are still within
allowable bending moment capacity of diaphragm wall.

Bending moment (kN.m/m)

Bending moment (kN.m/m)

-500
0

500

1000

-500

500

1000

Depth (m)

10

15

20

Measured [VWSG]

Measured [VWSG]

Deduced [INC]

Deduced [INC]

Computed [FEM]

Computed [FEM]

Allowable

Allowable

25

(b)
(a)
Figure 16 Bending moment of diaphragm wall in (a) tunnel zone (T1); (b) deep approach zone
(D2)
7.5 Excess pore water pressure
Measured excess pore water pressure normalized with initial pore water pressure is shown in
Figure 17. Pore water pressure is measured from standpipe piezometer installed at a distance of 1
m away from diaphragm wall at sump pit with tip of piezometer at 10 m. Initial pore water
pressure is about 70 kPa which is considered to be reasonable if hydrostatic pore water
distribution is assumed given ground water table is at 3 m below ground surface (Teparaksa,
1999). Wall displacement at depth of 10 m is shown to compare with response of piezometer.
During the first 120 day after initial reading, although there is constant increasing of wall
movement, there is no significant change of pore water pressure. This observation may be due to
unloading of horizontal stress causes negative excess pore water pressure and shearing in soft
clay causes positive excess pore water pressure. Consequently, two components counteract each
other causing only a slight change of excess pore water pressure. From day 120 to 150, lateral

402

wall displacement increase significantly when excavation reaches to final depth of 10 m.

Subsequently, there is a sharp increment of excess pore water pressure from day 160 to 200. This
is likely to be caused by shearing in compression mode dominates effect of horizontal stress
unloading. After base slab is constructed (at day 210), wall movement and excess pore water
pressure appear to be stable.

Wall movement

0.5

25
20

0.3

15

0.2

10

0.1
0.0

Base slab
construction

0.4

Excavate
to 10 m

u/u0

30

Excess pore water pressure

50

100

150

200

250

Lateral displacement (mm)

0.6

Construction time (day)

Figure 17 Normalized excess pore water pressure at 10 m deep outside of diaphragm wall at
sump pit (DS1)
8. Conclusion
Design and construction of vehicle underpass in the congestive area of Bangkok is presented.
Inclinometer measurement confirmed that top-down construction with permanent slab (stiffer
bracing) induced less deflection than that temporary steel bracing with lower stiffness. This
project reveals that a permanent diaphragm wall coupled with effective design modification of
associated structure members and construction method could offer a logistically and technically
attractive solution in construction of vehicle underpass in the congestive traffic area of Bangkok.
Careful consideration in post-tender review of technical options and selecting the most practical
construction method backup by sound design input and good performance of construction team
are important factors contributed to successful completion of the project.

403

References
Brinkgreve, R.B.J., and Vermeer P.A. 1998. PLAXIS v7.2 Manual, A.A. Balkema, Rotterdam,
Netherlands
Peck, R.B., Hanson, W.E., and Thornburn, T.H. 1974. Foundation Engineering. New York: John
Wiley&Sons.
Phienwej, N., Akawanlop, K and Balasubramaniam, A.S. 1995. Comparative evaluation of
ground movements associated with braced - excavation in Bangkok soft clay, 10th Asian
Regional Conference on Soil Mechanics and Foundation Engineering, Beijing, 341-344
Sambhandaraksa, S., and Pitupakorn, W. 1985. Prediction of prestressed concrete pile capacity
in stiff clay and clayey sand. Proc. of 8th Southeast Asian Geotechnical Conference,
Kuala Lumpur, Malaysia, (3), 58-63
Teparaksa, W. 1999. Principle and Application of Instrumentation for the First MRTA Subway
Project in Bangkok. 5th International Symposium on Field Measurements in
Geomechanics, Singapore, 411-416
Wang, J. H., Xu, Z. H. and Wang, W. D. 2010. Wall and ground movements due to deep
excavations in Shanghai soft soils. Journal of Geotechnical and Geoenvironmental
Engineering, 136 (7), 985-994

404

Fourth International Conference on

Case Histories in Geotechnical

Engineering
March 8 12, 1998, St. Louis, Missouri, USA

Editor : Shamsher Prakash

Sonic integrity test on piles founded in Bangkok

subsoil Signal characteristics and their
interpretations
Thasnanipan N. , Maung A. W. and Ganeshan B.
SEAFCO Co., Ltd.

University of Missouri-Rolla
Rolla, Missouri

405

1086
Proceedings: Fourth International Conference on Case Histories in Geotechnical Engineering, St. Louis, Missouri,
March 9-12, 1988.

SONIC INTEGRITY TEST ON PILES FOUNDED IN BANGKOK SUBSOIL

SIGNAL CHARACTERISTICS AND THEIR INTERPRETATIONS
Narong Thasnanipan
SEAFCO Co., Ltd.
Bangkok 10320
Thailand

Aung Win Maung

SEAFCO Co., Ltd.
Bangkok 10320
Thailand

Ganeshan Baskaran
SEAFCO Co., Ltd.
Bangkok 10320
Thailand

Paper No. 11.02

ABSTRACT
Sonic integrity testing is a common method used to test the integrity of piles in Bangkok. An acoustic technique equipped with
computer software, developed by TNO, PDI, IFCO, etc., is commonly used to test the piles. This paper presents the signal
characteristics and their interpretations from sonic integrity test performed on mainly bored piles varying in length, size, construction
method and founded soil strata. Toe reflection of small bored piles (diameter of 0.35m to 0.60m) constructed in Bangkok subsoil can
be observed for pile length of up to 20m to 24m . Beyond 35m below the ground , visibility of toe reflection is uncommon for large
bored piles (diameter of 0.80m and larger). Commonly predicted defects and frequency of occurrences are tabulated. Data sets have
been selected from the Data Bank of a foundations specialists organization to which the authors are associated with. Recently
completed and current projects in Bangkok have been selected so that the data set could be well represented by over 8000 piles. 285
piles (3.3%) out of 8,689 piles have been interpreted for poor concrete or inclusions, cracks and size reduction. Among these, cracks
contribute the major portion (2.2% or 194 piles). Cracks are mainly induced by either improper excavation adjacent to the piles or
trimming the pile head by ill fated ways. Small diameter piles are the most suffered by cracks. Prominent sectional variations have
been indicated by the signals at depth where temporary casing ends. Intermittent variations also are common at the interface of soft
clay and medium to stiff clay layers.
KEYWORDS
Integrity, Sonic signal, Toe reflection, Pile impedance, Discontinuity

407

1087
INTRODUCTION

For large bored piles of 0.80m to 1.50m in diameter, piles are

constructed to depths of 30m to over 60m. Wet process is
commonly used and piles are founded in sand layers.

Recent economic growth is being embraced by high-rise

buildings in Bangkok, the city of Thailand, supplemented by
other fast developing structures such as elevated highways and
railways, bridges etc. As the Bangkok subsoil suffers a lot of
constraints especially by the presence of thick soft clay on the
top, almost every structure relies on pile foundation. Tens of
thousands of piles are being constructed and tested annually in
this city. For old structures, short friction piles (6-8m long)
floating in the soft clay layer were commonly used. In modern
practice, driven concrete piles with tips extending to the first
sand layer or to stiff clay layer at a depth about 20-30m are
normally employed for light to medium sized structures. For
heavy structures or high-rise buildings deep large diameter
(0.80m-1.5m) bored piles with tip extending to more than 45m
are commonly used. Safety and overall stability of any finished
structure highly depend on the foundation and assessing the
quality of foundation is an inevitable aspect at the early stages
of the structure. An acoustic technique equipped with computer
softwares, developed by TNO, PDI, IFCO, etc., is commonly
used to test the pile integrity.

Compressive cylindrical strength of concrete used in bored

piles in Bangkok soil is in a range of 240-280 ksc (24-28 Mpa)
PRINCIPLES OF SONIC INTEGRITY TESTING
The test method is based on the time-domain, one dimensional
wave theory. The compressive wave is propagated axially by
the blow of hand-held impact device on to the pile head. The
wave travels down along the pile shaft and reflects upward at
the pile toe. In case where pile impedance variations occur due
to changes in pile cross sectional area or properties of pile
materials, or presence of discontinuities, part or whole of down
ward traveling wave reflects at the impedance variations and
returns to the pile top before the first reflection from the pile
toe. The reflected wave is detected by an accelerometer
mounted on the pile top. Signal traces of time-velocity are
usually recorded. As piles are installed in the ground the
reflected signal is usually influenced by shaft friction
contributed by the surrounding soil layers.

TYPICAL SUBSOIL CONDITION IN BANGKOK

The Bangkok subsoil generally consists of 12-18m soft marine
clay deposit underlain by medium to stiff clay. The medium
dense to dense first silty fine sand is found below the stiff clay
layer at depths 20m to 35m. The dense second sand layer
occurs at about 50m depth. Between these sand layers, stiff to
hard clay layers are present.

The formulae commonly used in analysis of integrity are;

CONSTRUCTION METHOD

Impedance of pile z (pile resistance to change in stress wave

velocity) is given by

L= c. T/2

(1)

Where L is pile length, c is velocity of stress wave and T is

travel time of the wave from pile top to toe and reflect back to
the top.

Two types of construction method, dry and wet process are

commonly practiced in bored piling in Bangkok Soil.

z = E.A/c

(2)

where A is cross-sectional area of piles and E is Youngs

Modulus of the pile material

Dry process uses steel temporary casing (about 15m in length)

to protect the soft clay from caving in. The bore is drilled
either auger or using bucket. For small piles, shells with drop
weight is used to make the bore. Concrete is poured into
borehole which is in dry condition after placing the
reinforcement cage.

E = .c2

(3)

from Equation (2) and (3)

z = c.A

(4)

where is density of pile.

Wet process too normally uses steel temporary casing of 1415m in length. A borehole is made by rotary method using
auger or auger bucket under bentonite slurry. Concrete is
poured with tremie method.

The length of pile is determined from the travel time of the

reflected wave from the pile toe and the wave velocity in
concrete. The velocity wave can be estimated from the
concrete strength of piles or from the reference pile with
known length. In the same way the location of discontinuities
and irregularities in piles can be also determined. Equation (4)
shows that the change in pile impedance could be due to
variation in pile cross-sectional area or in concrete quality.

Dry process is employed mainly for small piles of 0.35m to

0.60m in diameter. Length of the piles are up to 25m and
extends further if necessary, depending on subsoil condition,
absence of sand layer with groundwater, etc.

408

1088
CHARACTERISTICS OF SONIC TEST SIGNAL IN
BANGKOK SUBSOIL

In some cases, even though pile toe is extended beyond 35m in

depth, the toe reflection can be observed in the test signals of
toe grouted piles having L/D within 40.

Soil influence

Figure 3 is an example for such case where the bored piles

were base grouted and of 40m in length (toe at 55m, trim level
at -15m). However, it should be noted that these reflections are
prominent only for few piles. Influence on toe reflection by
grouted toe also depends on stiffness of the soil, underneath
the toe, which has been grouted.

Influence by different soil strata on signals have been observed

in many cases. Stiff clay layer at depth 15-25m (undrained
shear strength of 10-22 t/m2.) influences prominently and
higher degree of signal reflections have been observed in small
bored piles than that of large bored piles. Figures 1 and 2
explain these conclusions.

Soft Clay Very Stiff Clay

Soft Clay

Hard Clay Dense Sand

Med. Very Stiff Dense

Clay Clay
Sand
5.3cm/s
0

12

Pile 49
19 Sep 95

15
18
4000 m/s
exp: 20

21

24

27
f:8

30

10

15

20

Pile G30
27 Jan 97

10.6cm/s
V2-88c
Sr

25
30
4000 m/s
exp: 20

35

40

45
f:8

50

V2-88c
Sr

Fig. 3 Toe reflection observed from a toe grouted pile .(pile

1.00m x 40.0m , cutoff 15.0m, toe depth 55.0m)

Fig. 1. Influence of soil friction on sonic signals (a pre-cast

driven pile 0.40x0.40x25.0m in clay layers)

Size variations
Toe reflections

Size variations interpreted by signals are often consistent with

casting records and drilling monitoring results available.

Toe reflections are clear when the toe depth is 20-22m for
piles of 0.35m-0.50m in diameter and up to 24m for 0.60m
piles. In general, when pile length over diameter ratio (L/D) is
within 40, toe reflections are clear and hence the pile length is
predicted. However, for piles founded in depth deeper than 3035m the toe reflection is uncommon as damping of signals is
caused by deeper soil layers. Further, magnitude of toe
reflection is prominent for larger piles compared to small piles
with same length and constructed at same site. Figure 2 shows
the signal characteristics of different size piles constructed just
a few meters apart with same length.

Figure 4A and 4B show the comparison of borehole profiles

and sonic test records.

5.4cm/s
0

Pile 322
10 Jan 95

10

15

20

25

30

35

4000 m/s
exp: 20

40

45
f:8

50

V2-88c
Sr

23.2m
11.0cm/s
0

12

15

Pile 6

4000m/s

12 Sep 95

exp: 20

18

21

24

27

30

f:8

V2-88c
Sr

22.0m

7.1cm/s
0

12

15

Pile 24

4000m/s

12 Sep 95

exp: 20

18

21

24

27
f:8

30

Soft Clay

V2-88c
Sr

Med.
Stiff to
Dense Stiff
Clay Hard Clay Sand Clay

Dense Sand

Fig 4A Size increase in pile shaft at 13m is shown by both

sonic integrity test and drilling monitoring records. (Pile
1.0m, cutoff 5.00m, toe depth 45m.)

Fig. 2 Toe reflections from different pile sizes, pile no. 6 (

0.50m x 22.0m) and pile no. 24 ( 0.35m x 22.0m)

409

1089
Figure 6 shows bored pile sections formed at the lower end of
temporary casing. These bored piles were cast together with a
steel column for construction of deep basements with
diaphragm walls using top-down method. These piles were
exposed when excavation reached to 15m below the ground
level.

4.3cm/s
0

10

15

20

25

Pile 99

4000 m/s

25 Nov 94

exp: 20

Soft Clay

30

35

40

45

50

V2-88c

f:8

Med. Stiff to Hard Dense Stiff

Clay
Clay
Sand Clay

Sr

Dense Sand

Fig 4B Uniform pile size along the pile shaft is shown by both
drilling moitoring record and sonic integrity test record. (Pile
0.80m, cutoff 3.0m, toe depth -45m.)
Common causes for size variations in bored piles are;

Fig. 6 The photograph illustrates the pile size changes caused

by difference in diameter of temporary casing and auger used
in pile construction.

1. Effects from temporary casing

- slight dimension changes of casing and bucket or auger
- inadequate length to protect the soft clay
- surface intact change at transition at the casing end
2. Delayed feeding of slurry and improper slurry level
maintenance for wet process
3. Long time duration of maintaining open bore
4. Non continuous concrete pouring
5. Soil conditions such as inclusion of sand lenses in clay
layers.

The test signal in Fig. 7 shows size decrease caused by

inadequate casing length used in soft clay layer. Such defect is
found often in piles constructed by dry process using casings
of inadequate length to protect the entire soft clay layer of
varying thickness.

Med.
Clay

Soft Clay

Stiff Clay Dense

Sand

11.0cm/s
0

12

Pile 66
25 Jun 97

15
4000 m/s
exp: 20

18

21

24

27
f:8

30

V2-88c
Sr

Fig. 7 The signal shows decrease in pile size at 10.6m (pile

0.35m x 26.0m)

Med. Stiff Clay Dense Sand

Clay

Very Soft Clay

Very Stiff Clay

10.0 cm/s

10.5cm/s
0
Pile 30
2 Jul 97

Med.
Clay

Very Stiff Clay

Figure 5 shows reduction of pile size at about 14m. A

completed bored pile would have the size of outer casing at the
top portion while the lower portion would be of bucket/auger
size which is slightly less than the casing size. This size
variation has been indicated by the signals at the transition
zone. However dimensional variations by improper concrete
slumping upon extracting the temporary casing and gradual
dimensional changes at the transition zone are hardly indicated
by the signal.

12

15

4000 m/s
exp: 20

18

21

24

27
f:8

30

V2-88c

Pile 168
11 Nov 95

Sr

12

16

20
4000 m/s
exp: 20

24

28

32

36
f:8

40

V2-88c
Sr

Fig. 8 Pile with localize size increase at13.5m indicated by the

sonic signal, (pile 0.80m, cutoff 1.4m, toe depth 35.85m))

Fig. 5 Pile size decrease or change at 14m shown by the

signal (Pile 0.80m x 21.0m)

410

1090
In contrast to the size reduction, caused by smaller
auger/bucket size than the casing size, relative size increase
also have been observed. Casting records for piles indicated in
Fig. 8 noted about 30% over break of concrete volume.

Med. Stiff Clay Dense

Clay
Sand

Very Stiff Clay

Discontinuities/cracks

8.9 cm/s
0

Common causes attributed for these are;

1. disruption in concrete pouring
2. contamination of concrete with bentonite slurry or soil
3. lifting up of insufficiently workable concrete or hardened
concrete upon extracting the temporary casing

12

15

Pile 50

4000 m/s

4 Apr 97

exp: 20

18

21

24

27

30

f:8

V2-88c
Sr

Fig. 10A Sonic test signal showing size decrease and or a

discontinuity at 7.5-9.0m and poor quality pile section near
the to (Pile 0.35m x 20.0m)

Figure 9A shows a discontinuity in pile shaft caused by

disruption during concrete pouring. Concrete supply was
interrupted by poor coordination between the piling contractor
and concrete suppliers. By the time to resume the concreting
the lower section has been hardened. The discontinuity found
was later closed by coring and cement grouting. Figures 9A
and 9B show the signal interpretation before and after closing
the discontinuity.

Med. Stiff Clay Dense

Clay
Sand

Very Stiff Clay

3.5 cm/s
0

12

15

Pile TEST

4000 m/s

25 Jun 97

exp: 20

18

21

24

27

30

f:8

V2-88c
Sr

Fig. 10B Sonic signal on the failed test pile showing a

construction joint between pile cap and pile top at 1.40m. Size
decrease or a discontinuity is also indicated.
14.2 cm/s
Pile 1

12

16

24 Apr 97

20
24
4000 m/s
exp: 20

28

32

36
40
f:8

V2-88c
Sr

Figure 11A shows presence of poor quality concrete at the pile

top due to contamination of concrete with bentonite slurry.
Such defect is often found in piles with shallow cutoff level
and hence adequate over cast length to have a sound concrete
at cut off level, sometimes, cannot be made. Normally 1.0m to
2.0m over-casting of concrete is practiced.
After trimming back the pile top, sonic test was performed and
the test signal indicates the sound integrity of the same (see
Fig. 11B)

Fig 9A Sonic signal showing a discontinuity at 11.0m caused

by interruption in concrete pouring(pile 0.80mx38.8m) and
multiple reflections from the discontinuity.

7.6 cm/s
0

12

16

20

Pile 1 #

4000 m/s

2 Jun 97

exp: 20

24

28

32

36
f:8

40

V2-88c
Sr
22.8 cm/s

Fig. 9B Sonic integrity test shows the discontinuity at 11.0m

has been closed by grouting.

12

Pile 28
15 Jan 97

15

18

21

24

4000 m/s
exp: 20

27

30

V2-88c
Sr

f:8

Fig. 11A The signal showing poor quality concrete at pile top
section (Pile 0.60m x 18.0m)

Figures 10A and 10B show the characteristics of signals

observed at a site where dry process was adopted. Many piles
at this particular site were interpreted for poor integrity
especially at the bottom portion. From the construction records
it was noted that boring was done in water bearing sand layer
without casing or bentonite slurry causing concrete
contamination with water and/or segregation.

8.8 cm/s
0
Pile 28
18 Jan 97

Further, static pile load test results indicate the test pile (Fig.
10B) failed well before designed maximum test load. Figure
10B also shows the joint between pile cap and pile.

12

15

18

4000 m/s
exp: 20

21

24

27
f:8

30

V2-88c
Sr

Fig. 11B The signal acquired after trimming top portion. The
reflections from the lower part of the pile and toe become
visible.

411

1091
Physical pile damage due to excavation or improper chipping
to trim level
Most discontinuities or cracks in bored piles after installation
of piles can be caused by construction activities associated
with basement excavation adjacent to them and improper
trimming to design cut-off level.

12

18

24

Pile 681
8 Apr 96

30

36

Dense Sand

42

48

54

60

Sr

Fig. 12 Sonic signal showing a discontinuity (crack) at about

7.8m, caused by lateral soil movements.

10

20

30
1

40
2

50
3

ms
L/c

-750

30
Ton / m

0.1
2.2
1.0

Sonic integrity test is an effective method to assess the pile

integrity. Computerized acoustic technique developed by
TNO, PDI and IFCO is commonly used in Bangkok for both
driven piles and bored piles. About 3.3% of 8689 piles have
been interpreted for either size reduction or poor quality of
concreting or cracks/discontinuities.

For Msd
Vel Msd

750

cutoff level near the ground

or not enough overcast
Excavation works for pile
trimming and Construction
activities
Inadequate casing length or
soft clay layer variation in
thickness

CONCLUSION

29-Apr-96
CAPWAP (R) Version 1 994 -1

1500
Ton

Poor concrete at
pile top (0-3m)
Cracks/
discontinuities

A total of 1491 piles were found with localize increments in

pile size. All of size variations including size reductions are
found to be at the transition interface of temporary casing and
soil layer, where soil strata changes also occur.

V2-88c

f:8

4000 m/s
exp: 20

From the above assessment, most of the defects are found to be

in small piles with a diameter of 0.60m or smaller. Study of
piling and excavation plans show that piles with a crack
indicated by the sonic integrity were located in the vicinity of
excavation boundaries.

6.2 cm/s
0

Causes

Size reduction

The test signal in Fig. 12 shows a crack or discontinuity in the

pile caused by soil movements occurred by adjacent
excavation work.
Soft Clay Med. Very Stiff Dense Sand
Clay Clay

Type of Defect

Pile
Shaft Resistance
Distribution
(Toe 178.8 Ton)
Pile Forces at Rut

1000
Ton

Toe reflections are generally clear for small piles with

slenderness ratio of 40. But only in few cases, especially toe
grouted, toe reflection has been observed for longer piles with
toe embedded below 30-35m.

Fig. 13 High strain dynamic load test signal showing a crack

in the pile at about 8.0m.
An inadequate retaining system for excavation work has
caused the lateral movement to the pile top about 80cm and a
crack has developed in this pile. The crack was also detected
by the high strain dynamic load test performed to determine
the pile capacity of the cracked pile (see Fig. 13).

Signal interpretations have been compared with soil

investigation results and construction records. Causes for
defects could be identified and conformed to casting records.
Minor defects, such as small cracks can be predicted in many
instances and nature of them need to be verified by further
visual inspection.

From Fig. 12 and Fig. 13 the duration of pulse in the sonic

integrity test signal is about half of that in dynamic test signal.
The short duration (high frequency) pulses can resolve
reflections from narrow impedance changes. In such case
presence of a crack in the pile is more clearly indicated by
sonic integrity test signal (low strain) than the dynamic test
signal (high strain).

The predicted defects and possible causes verified from the

construction records would help the piling and excavation
contractors to be aware of the causes of defects associated with
construction practice.

An assessment of pile defects based on the sonic integrity test

records on 8,689 bored piles in Bangkok subsoil and their
construction records including soil conditions was carried out.
The assessment indicates that integrity of 285 piles (3.3%) out
of 8689 piles has been suspected. They are summarized in the
table below;

ACKNOWLEDGEMENT
The authors express their appreciation to E.D.E. Co. Ltd. in
supplying test records and to Mr. Sujit Onsri and Mrs. Piyamitr
Sura-auggool for their assistance in the preparation of this
paper.

412

1092
REFERENCES
CIRIA PG72 [1977] Review of problems associated with the
construction of cast-in-place concrete piles Construction
Industry Research & Information Association.
CIRIA PG4 [1977] Integrity Testing of Piles: A Review
Construction Industry Research & Information Association.
Fleming, W. G. K., Reiding, F., and Middendorp, P. [1985]
Faults in Cast in place Piles and Their Detection,
Proceedings of the Second International Conference on
Structural Faults and Repair, The Institution of Civil
Engineers, Westminster, London, pp. 301-310.
Rausche F, Likins E. G and Hussein M. [1988] Pile Integrity
by Low and High Strain Impacts, Application of Stress-Wave
Theory to Piles, Third International Conference. Ottawa,
Canada, pp. 44-55.
Seitz, J.M. [1984] Low Strain Integrity Testing of Bored
Piles, Application of Stress-Wave Theory to Piles, Second
International Conference, Stockholm, Swedin, pp. 94-102.
Sliwinski, Z. J. and Fleming, W.G.K. [1983] The Integrity and
Performance of Bored Piles, Advances in Piling and Ground
Treatment for Foundations, The Institution of Civil Engineers,
London, UK., pp. 153-165.

413

PROCEEDINGS OF THE SIXTH INTERNATIONAL CONFERENCE ON THE APPLICATION

OF STRESS-WAVE THEORY TO PILES / SAO PAULO / BRAZIL/ 11 13 SEPTEMBER 2000

Application of Stress-Wave
Theory to Piles
Quality Assurance on Land and Offshore Piling
Edited by

Sussumu Niyama

Institute for Technological Research IPT, Sao Paulo, Brazil

Jorge Beim

PDI Engenharia, Rio de Janeiro, Brazil

Sonic Integrity test of piles-integrity effected by basement excavation

in Bangkok soft clay
N. Thasnanipan, A.W. Maung, P. Tangseng & Z. Z. Aye
Seafco Company Limited Bangkok, Thailand

A.A. BALKEMA / ROTTERDAM / BROOKFIELD / 2000

415

Sonic integrity test of piles-integrity effected by basement excavation in

Bangkok Soft Clay
N. Thasnanipan, A. W. Maung, P. Tanseng & Z. Z. Aye
Seafco Company Limited, Bangkok, Thailand

ABSTRACT: Modern buildings in Bangkok nowadays are designed with basement facility and different elevation of pile foundation which require excavation works in the immediate vicinity of the constructed bored
piles. Adverse effects on piles from adjacent excavation works in terms of excessive movement of piles and
pile damages have been observed in some projects. Assessment of bored pile damages using sonic integrity
test is focused in this paper with demonstration of test results from three case histories. Method of remedial
works for damaged piles and additional testing for pile capacity justification where available are also briefly
discussed. Stress induced in piles due to the soil movement caused by excavation works is also examined
with the aid of computerized finite element program. Location of cracks detected by sonic integrity tests is
correlated with the analyzed bending moment in pile, which exceeds the cracking moment capacity of pile.

1 INTRODUCTION

2 GENERAL DESCRIPTION OF BANGKOK

SUBSOIL

Due to the prevailing condition of subsoil (existence

of considerably weak stratum at top part) piles are
commonly used for foundation of buildings and
structures in Bangkok. Deep-seated large diameter
bored piles of over 35 m depth with 0.8 1.8m in diameter have been mainly used for heavy structures
such as high rise buildings, elevated expressways,
flyovers or overpass bridges and more recently for
underground subway stations of the first Bangkok
MRTA system. As a part of pre-planned quality assurance regime and retrospective investigation, integrity of the piles are tested to obtain the information with regard to the potential deficiencies of the
constructed piles which may have formed during actual pile construction process or may have been attributed by other activities after construction of the
piles. Designs of the modern buildings in Bangkok
frequently call for the basement facility and for the
different elevation of pile foundation, which require
excavation work mainly in top soft clay layer. In
some projects, adverse effect from excavation works
adjacent to constructed piles has been observed.
This paper presents the assessment of bored pile
damages which were induced by adjacent excavation
using sonic integrity testing method.

Bangkok subsoil consisted of alternating clay and

sand layers of Quaternary deposits extending down
to about 550m depth where bedrock generally exists
(Balasubramanium, 1991). Made-Ground consists
predominantly of Fill-Materials, Clayey Sand or
Silty Clay with some cement block rubble and rock
fragments, is commonly found up to 4m depth. Soft
to very soft, highly compressible dark gray marine
clay lies beneath Made-Ground and in some areas it
lies under weathered crust layers of 2m thick. Depending on the location, this layer is extended up to
12-18m. About 2m thick Medium Clay layer can be
observed between Soft Clay and underlying Stiff
Clay. Generally Stiff Clay layer occurs directly underneath Medium Clay and its depth goes up to 22m.
Below Stiff Clay layer, First Sand layer 5-8m in
thickness can be found. This First Sand layer, however, is absent in some areas. Stiff to Hard Clay
layer underlies First Sand and it is found to be about
5m thick. Second Sand layer generally occurs at
depths between 45 to 65m. The Generalized Fence
Diagram of Subsoil covering the central area of
Bangkok is illustrated in Figure 1.
163

417

the technical requirement of the particular projects

ranging from minimum 10% to maximum 100% of
total constructed bored piles. A sonic integrity tester
with built-in computer having high quality signal
acquisition system and computer programs developed by TNO, PDI, IFCO etc. are commonly used
for pile integrity testing.
5 OVERVIEW OF EXCAVATION INDUCED
PILE DAMAGE IN BANGKOK
In general, pile defects are caused at two stages, during the pile construction process and after construction of the piles (post-pile-construction).
Thasnanipan et al. (1998b) reported that integrity
of 285 bored piles (3.3 % of 8,689) were found to be
of doubtful quality according to an assessment on
the results of sonic integrity test with the additional
information obtained from the pile construction records of bored piles in Bangkok subsoil. The results
of the findings are summarized below.

Figure 1. Generalized fence diagram of Bangkok subsoil

3 BORED PILE CONSTRUCTION IN

BANGKOK
In Bangkok, bored piles are constructed mainly by
two methods, dry and wet process. Dry process is
applied commonly for small piles of diameter 0.35
to 0.60m with relatively shorter lengths. In dry
process, a pile is bored to the required depth within
impermeable layers (usually up to 25m) by either rotary or percussion method whilst temporary casing
(about 15m in length) is installed in soft clay layer
for protection of soil caving. After installing reinforcement cage, concrete is poured into the borehole
under dry condition. Tripod method, shell with drop
weight is also commonly used in dry process.
Wet process method as its name implies makes
the pile under wet condition by using drilling slurry
(bentonite or polymer). Temporary casing of 1415m in length is also used as a support in soft clay
layer and soil inside the casing is normally excavated by auger applying rotary drilling method.
Drilling is continued with the bucket under the
slurry from the top of sand layer to the final depth.
Tremie concreting is necessary for pile installed under wet process. Bored piles constructed by wet
process are generally of large diameter (0.8m to
1.8m diameter) deep-seated (30m to over 60m) and
are normally founded in either first or second sand
layer. Bored piles having compressive cylindrical
strength of concrete in a range of 240-280 ksc (2428 MPa) with reinforcement ranging 0.5% to1.2% of
pile sectional area, are commonly used in Bangkok.

Table 1. Summary of pile assessment on 8,689 bored piles

(Thasnanipan et al. 1998b)
Type of Defect

Causes

Poor concrete at
pile top (0-3m)

Cutoff level near the ground

or inadequate overcast

0.1

Size reduction

Insufficient casing length or

soft clay layer variation in
thickness

1.0

Cracks/discontinuit
ies

Excavation works for pile

trimming and construction
activities

2.2

As can be seen in the above table, higher percentage

of defects is caused by post-pile-construction activities and more commonly by movement of surrounding soils induced by excavation works in the vicinity
of the piles.
Modern buildings in Bangkok frequently require
the excavation works for basement facility and for
the use of different elevations of pile foundation.
Excavation works are mainly carried out by using
cut slope, sheet pile wall, diaphragm wall and secant
pile wall. Since most of the bored piles are designed
mainly to carry an axial load, lateral and tensile
force which may impose on pile during basement
excavation are sometimes not considered or overlooked. The lateral displacement of ground caused
by excavation naturally induced an additional lateral
force and bending moment in the pile. Once the induced bending moment exceeds the cracking moment capacity, piles are subject to be cracked. Unexpected tensile force acting upon the pile due to an
excessive heave or uplift force induced by excavation can also cause cracking of pile. In Bangkok,
pile cracks caused by external forces are normally
found to be at the level where Soft and Medium to

4 SONIC INTEGRITY TESTING IN BANGKOK

Sonic integrity test is widely used for integrity testing of both driven and bored piles in Bangkok. It is
employed as a part of quality control and or as a retrospective investigation when some problem becomes apparent. For a quality control regime, the
number required for integrity testing depends upon
164

418

Stiff Clay boundary is present. Figure 2 illustrates

the typical features of external forces, which cause a
crack in the piles. Pile damage associated with sheet
pile wall excavation is presented in this paper.
TYPICAL SOIL PROFILE

SOFT CLAY

MEDIUM CLAY
STIFF CLAY

Basically, following formulae are used in analysis of

sonic integrity test.
L = c T/2

(1)

where, L is a length from pile head to reflected surface, c is a velocity of stress wave and T is a travel
time of the wave for length L.
From this formula, length L, can be computed
from the travel time of reflected wave and the wave
velocity of concrete which may be estimated from
concrete strength of piles or from the reference pile
of known length.

SAND

crack induced
by large lateral
movement

crack induced
by sliding
failure

z = E A/c
E = c2

crack induced
by uplift
force

Figure 2. Typical features of external forces causing pile

cracks in Bangkok subsoil

(2)
(3)

where, z is an impedance of pile, E is Young

Modulus of pile material, A is cross-sectional area of
pile, c is the propagation velocity of the stress wave
and is density of pile.
From Equations (2) and (3)

6 APPLICATION OF SONIC INTEGRITY TEST

IN PILE-INTEGRITY CHECKING

z = cA

Sonic integrity test is simple, rapid, efficient and

cost effective in examination of the pile integrity. It
is reasonably reliable for the integrity testing of the
pile especially for the upper part where potential of
damage is higher as far as defect caused by the external effects (lateral displacement of soft clay due
to adjacent excavation works, unexpected force from
improper pile head trimming) is concerned.
6.1 Basic principles of sonic integrity test
Sonic integrity testing also called as low-strain integrity testing examines the response of a pile to a
light external impulse force. Either time-domain or
frequency-domain can be used in analysis of measured data by sonic integrity test. In Bangkok, timedomain method is mainly applied in interpretation of
test data.
Sonic integrity test is undertaken by striking the
head of the pile with a light hand-held hammer and
recording the response of the pile to this impulse
blow by means of a sensor or accelerometer placed
in good contact with the pile head. Stress wave theory is a basic principle of this low-strain integrity
testing. The hammer blow induces a compressive
stress-wave into the pile, which propagates axially
along the pile shaft and reflects back toward the pile
head at a change of impedance within the pile. In
the case where pile impedance variation occurs due
to changes in pile cross sectional area or properties
of pile materials or presence of discontinuities, part
or whole of down ward wave reflects at the impedance variations and returns to the pile head before
the first reflection from the pile toe.
165
419

(4)

As can be seen from Equation (4), the change in pile

impedance could be due to variation in pile crosssectional area or in concrete quality.
6.2 Interpretation of sonic integrity test
In general the typical pile features detectable by
sonic integrity test includes;
Pile size variation
Pile toe reflection (for pile length verification)
Pile material variation
Soil influence
Discontinuity and/or cracks
The typical sonic integrity signals of above features have been reviewed and presented by Thasnanipan et al. (1998a). As some features have a
similar pattern to each other, it is always important
to check and compare the test data of pile in question with those of other piles within the same job site
to establish site signature. In some cases correlations should be made with similar projects where
damaged piles were investigated with firmed evidence (coring or excavation to detect damage level).
Interpretation should also be made with sound
knowledge of pile construction technique, subsoil
condition and other factors (problem encountered
during and post pile construction) which may influence the sonic test signals. Pile construction records
are also useful in interpretation of piles with detected anomaly. Conclusion should be made after
careful and thorough review of test results in conjunction with other available information. Further

investigation (if applicable excavate down to level of

detected crack for visual inspection) may require before final conclusion is made.
As a main focus of this paper, the typical features
associated with the pile damage (discontinuity/
cracks) detected by sonic integrity test is discussed
in detail with the data obtained from three cases as
presented in the following sections.

movement as well as settlement of ground including

tension cracks of more than 300mm in width were
observed at some locations. Minor damages were
also investigated at the adjacent properties. Temporary raking struts were immediately installed in response to these excessive ground movements caused
by sheet pile failure. A typical scheme showing excavation works of Case I is presented in Figure 3.

6.3 Pile damage assessment

From the results of sonic integrity test, degree of
damage or level of crack may be assessed. The assessment of crack generally involved setting of
amplification in the same range for all piles and
comparison of input impulse, amplitude of anomaly
and intensity of reflection from the anomaly (i.e.
multiple reflection). Additional information obtained
from further investigation as described in above section is also used in pile damage assessment.
7 CASE STUDIES
Figure 3. A picture showing slope excavation with sheet pile
support Case I

7.1 Case I Sheet pile excavation supported by soil

berm

7.1.4 Sonic integrity test and pile damage assessment

More than 50% of piles were tested with sonic integrity test. Test results were thoroughly reviewed and
site signature, most common features of the
majority of piles, is established. It is to be noted that
test signals from piles of less potential for damage
(least effect from excavation) are selected in establishment of site signature or reference signal of
good piles. Based on the site signature of good
piles, defect piles were identified and assessed. Table 2 shows the summary of pile damage assessment
by sonic integrity test.

7.1.1 Subsoil condition

A 12m thick soft clay layer occurs beneath fill material and weathered crust of 1.0 to 2.5m thick. Unconfined shear strength (Su) of soft clay layer ranges
1t/m2 at top and 2.5t/m2 at bottom whereas the natural water content is about 60-80%. Medium clay
layer of 1 to 2m in thickness, having shear strength
of 2.5t/m2 to 3.5t/m2 underlies soft clay layer. Below medium clay, at the depth between 18 to 27m,
stiff clay with traces of fine sand having SPT N
value of 15 to 37 is observed. Thick layer of fine
sand interbedded with clayey sand occurs underneath stiff clay layer.

Table 2. Summary of pile damage assessment by sonic integrity test Case I

7.1.2 Pile foundation

Pile foundation of eight high rise buildings in this
project comprises 904 cast-in-place bored piles of
diameter 1.0m and 1.5m, founded at 59m below
ground level. Pile tips are embedded in dense sand
layer and the entire length of piles are reinforced
with steel reinforcement of 0.72% of pile sectional
area for the top part and reduced gradually to 0.23%
for the bottom section. To support the underground
water tanks, 88 pre-cast concrete piles (I-30cm) having 21m length were also driven down to stiff clay
layer.

Pile
Dia.
(m)

No. of piles with defect

Defect depth
below ground
level (m)

Least
prominent
crack

Less
prominent
crack

Prominent
crack

1.0

6.0 13.8

1.2

10

6.1 7.6

1.5

9.9

The typical sonic integrity test results of undamaged

pile and damaged pile classified as least prominent
crack, less prominent crack and prominent crack are
shown in Figure 4. Majority of piles with crack detected were located in the periphery of the excavation zone. Piles, particularly with prominent cracks,
were observed to be deviated from their original location up to 600mm towards excavation.

7.1.3 Excavation method and observed failure

Excavation was initially carried out by using sheet
pile (Type FSP III) with soil berm support or slope.
Before reaching to 8.1m maximum depth of excavation, sheet pile failure and associate severe lateral
166

420

7.2 Case II Pile damage caused by sheet pile

excavation with one level temporary support
7.00cm/s

12

18

24 30

Pile 890
10 Apr 96

5.8cm/s

12

18

24 30

42

48

54

36

12

18

24 30

Pile 894
10 Apr 96

36

42

48

8.8cm/s

12

18

24 30

36

54

42

48

54

60

42

48

54

v2-88c
sr

60

7.2.2 Pile foundation

402 bored piles of diameter 1m and 1.5m, founded at
55m below ground, being embedded in dense sand
layer were used as foundation in this project. Reinforcement steel bars of 0.5% pile cross sectional area
were provided for the entire length of bored piles.

v2-88c
sr

f : 12

4000m/s
exp: 20

7.2.1 Subsoil condition

Soft Clay layer extends to 15m below ground surface with undrained shear strength increases from
1t/m2 at the top and 2t/m2 at the bottom. Medium
Clay layer of 2 to 3m in thickness is found between
Soft Clay and the underlying Stiff Clay. Stiff Clay
layer occurs up to depth 42m with the SPT N values ranges between 14 and 40.

v2-88c
sr

f : 12

4000m/s
exp: 20

Pile 681
9 Apr 96

60

f : 12

4000m/s
exp: 20

Pile 680
8 Apr 96

5.2cm/s

36

4000m/s
exp: 20

60

f : 12

v2-88c
sr

7.2.3 Excavation method and observed failure

Sheet pile wall of 14m in length with 1 level bracing
was utilized for maximum 8m excavation. Though
the design required for installation of struts at 1m
below ground level, the actual excavation was carried out up to 3m depth without installing the temporary struts causing sheet piles deflected in large
magnitude.

Figure 4. Seismic test results of good pile, pile with least

prominent, less prominent and prominent crack (top to bottom
in order) Case I

7.1.5 Remedial measures and pile capacity justification

Bored piles with cracks were cored to the depth below the crack and grouted with non-shrink cement.
Sonic integrity test was conducted few days after
rectifying the pile. Sonic integrity signals of the defected pile tested before and after the remedial work
is presented in Figure 5. After the rectification
works, high strain dynamic load test was performed
on two piles with large horizontal deviation and suspected prominent cracks. The dynamic test results
indicate that tested piles could be capable of carrying the design load with factor of safety higher than
1.5.

7.2.4 Sonic integrity tests and pile damage assessment

Sonic integrity test was performed for all piles after
trimming to the designed cutoff level. It is evident
that piles with crack detected by sonic integrity test
were mostly located in the vicinity of the excavation
boundary.

3.5cm/s
Soft Clay

6.2cm/s

Med. Very Stiff

Clay Clay

Dense Sand

Dense Sand

2.2cm/s
0

12 18 24 30 36 42 48 54 60
4000m/s
f:8
exp : 20

Pile 681
8 Apr 96

4.1cm/s
Pile 681A
29 May 96

12 18 24 30 36 42 48 54 60
4000m/s
exp : 20

f:8

12 18 24 30 36 42 48 54 60
4000m/s
exp. 20

Pile 165
13 May 97

v2-88c

Pile 193
13 May 97

sr

f: 12

12 18 24 30 36 42 48 54 60
4000m/s
exp. 20

f: 12

v2-88c
sr

v2-88c
sr

Figure 6. Sonic integrity test results of good pile and pile with
prominent crack Case II

v2-88c
sr

A summary of the number of damaged piles and

their degree of damage assessed are presented in Table 3.

Figure 5. Sonic integrity test results on defected pile before

and after remedial work Case I
167

421

Table 3. Summary of pile damage assessment by sonic integrity test Case II

Pile
Dia.
(m)

No. of piles with defect

Least
prominent
crack

Less
prominent
crack

Prominent
crack

Defect depth
below ground
level (m)

1.0

7.5 - 9.4

1.5

14

8.4 - 21.0

7.2.5 Remedial measures

Damaged piles were cored to the level below the defected crack and grouting was performed.

Figure 7. Layout plan showing part of bored piles, sheet pile

wall and bracing system Case III

7.3.4 Sonic integrity tests and pile damage assessment

Though sonic integrity test was employed as a part
of quality control regime in the original plan, it was
eventually applied for a retrospective investigation
after encountering many piles with defect. Hence all
piles were undergone sonic integrity test. According
to the test results, 84% of piles located within the
excavation zone were identified to be with crack.
Sonic integrity test results of good pile, pile with
less prominent crack, and pile with prominent crack
are illustrated in Figure 8. As can be seen in Figure
8, for pile with prominent crack, a distinct and sharp
reflection is evident at depth about 5.5m below pile
top. Repetition of reflection at multiples of this
depth can also be observed for this pile. These features are comparatively not distinct for the pile with
less prominent crack. An excessive basal heave, insufficient reinforcement to resist the tensile stress
and relatively small pile size are considered to be the
main reasons causing the cracks in this case. Table
4 shows a summary of pile damage assessment by
sonic integrity test.

7.3 Case III Pile damage caused by sheet pile

excavation with two level temporary support
7.3.1 Subsoil condition
Underneath the top 2m of weathered crust, Soft Clay
layer can be observed up to 15m below the ground
surface. Medium Clay layer underlies Soft Clay
with the thickness of 2m. Undrained shear strength
of Soft Clay ranges between 1t/m2 and 3t/m2
whereas it is about 50t/m2 in Medium Clay. Beneath
Medium Clay, Stiff Clay layer is found up to 25m
where the boundary of Hard Clay is investigated.
Within Stiff Clay stratum softer zone of Medium
Clay is identified at the depth between 19 and 21m
characterized by reduction in SPT values which appear consistent with increases in moisture content.
7.3.2 Pile foundation
A total of 143 bored piles of 0.6m diameter were installed with pile tip at about 26m below the ground
to support a residential apartment. Dry process was
employed to construct the piles. Cut-off level of
bored piles ranges from 0.85m to 5.95m depth. Entire lengths of all piles are fully reinforced with steel
bars grade SD40 of 0.35% of pile sectional area.

20.2cm/s

7.3.3 Excavation method and observed failure

Temporary retaining wall using FSP III sheet pile
type of 14m in length with two level bracing placed
at 1m and 3.5m below ground level was employed
for 6m depth excavation. A part of pile layout
within excavation area showing sheet pile wall and
bracing system is illustrated in Figure 7. No major
failure in terms of large tension cracks and ground
settlement was observed at the surface adjacent to
the excavation though some excessive sheet pile deflection was visually inspected at few locations.
Hence, it was not expected by the visual inspection
at site that excavation work has effected the integrity
of piles.

Pile 103
17 Feb 93

13.3cm/s

12 15 18 21 24 27 30

4000m/s
exp : 20

Pile 48
17 Feb 93

f:8

12 15 18 21 24 27 30

4000m/s
exp : 20

f:8

18.8cm/s
0

Pile 50
17 Feb 93

12 15 18 21 24 27 30

4000m/s
exp : 20

f:8

V2-88c
sr

V2-88c
sr

V2-88c
sr

Figure 8. Case III - Seismic test results of good pile, pile with
less prominent and prominent crack (top to bottom in order)
168

422

Table 4. Summary of pile damage assessment by sonic integrity test Case III
Pile
Dia.
(m)

Prominent
crack

Defect depth
below ground
level (m)

33

5.5 13.8

No. of piles with defect

Least
prominent
crack

0.60

Less
prominent
crack

38

Stiffness of foundation piles per linear meter (E.I

per pile spacing in row) was used as a parameter for
plate elements, where E and I is Young Modulus and
moment of inertia of pile respectively. The soil between piles in a row is ignored. Mohr-Coulomb constitutive model is adopted using undrained soil parameters derived from soil investigation at site and
associated lab test results.

7.3.5 Remedial measures and pile capacity justification

Coring was carried out up to the depth below the detected crack to verify the pile damage and then it
was grouted with non-shrink cement. To justify the
capacity of damage piles, high strain dynamic load
test was carried out on 25 piles in which 24 piles
with detected cracks and 1 pile without crack detection. Summary of dynamic load test is presented in
Table 5.

9 CORRELATION BETWEEN RESULTS FROM

FINITE ELEMENT AND SONIC INTEGRITY
TEST

Table 5. Summary of dynamic load test results of damaged

piles Case III

Table 6. Moment capacity of piles and induced moment in

piles due to excavation simulated by finite element method

Mobilized capacity / Design ultimate

capacity of 240 ton

The analysis results indicate that bending moment in

piles induced by excessive ground movement due to
excavation works are higher than that of cracking
capacity of pile in all cases. Table 6 presents the
summary of the analyses.

No. of
pile

Less than 50%

Less than 10%

More than 100%

2
22

Case

Pile
size
(m)

Rebar
(%)

Cracking
moment
(t.m)

I
II

1.0
1.5

0.75
0.50

30
102

III

0.60

0.35

6.5

Ultimate
moment
(Whitney)
(t.m)
79
188
9.2

Moment by
FEM model
(t.m)
149.7
117.2
255.7*
13.3

8 FINITE ELEMENT ANALYSIS WITH

COMPUTER MODEL SIMULATION

Note: (*) moment in pile located outside of sheet pile wall

within the active side of soil mass.

Finite element analysis was carried out using twodimensional modeling. With the aid of PLAXIS
computer program staged excavation was simulated
and pile/soil movements as well as bending stress
induced in the piles were examined for above presented projects. Model profile of Case III used in
FEM analyses is presented in Figure 9.

As can be seen in Table 7, the crack location in pile

indicated by sonic integrity test is generally consistent with the location of computed bending moment,
which exceeds the cracking moment capacity of
piles for all three cases.

1.52.5

17.00

0.0 1.0 2.0 3.0 4.0 0

0.0

5.0

Su (t/m2)

SPT (blows/ft)
20

40

Table 7. Locations of crack and computed bending moments

in piles in the vicinity of excavation boundary
Integrity test
depth of crack in
piles near excavation boundary (m)

12.7 & 13.7

12.7 &13.7
11.5 & 12.0

60

Crust

Depth of maximum bending

moment by
FEM (m)
18.0

Depth of bending
moment exceeding piles cracking
moment from
FEM (m)
11.0 & 13.0

15.0
15.5

12.8 & 14.5

13.0 & 14.0

Sheet pile
Soft CLAY

10.0

15.0

Case

II
III

Medium
CLAY

Very Stiff
CLAY
20.0

Medium
CLAY

25.0

Hard
CLAY

It is evident that detected cracks are located near or

at the depths where boundary of Soft and Medium or
Stiff Clay is present.
In case III, FEM analysis results suggest that
piles are subject to some degrees of basal heave as
illustrated in Figure 10. This is found to be caused
by inadequate embedded length of sheet pile. As can
be seen in Figure 11, the results also show that

30.0

Figure 9. Finite element model profile of Case III

169

423

bending moment induced by excavation exceeds the

ultimate moment capacity of pile . Sonic integrity
test results of some piles indicate the presence of
more than one crack at some depths, suggesting piles
would have experienced both bending and tensile
stress due to the excavation.

Observation of actual construction works on sites

suggests that over excavation prior to installation of
support and inadequate retaining system are the
main causes leading to excessive ground movement,
a primary reason of pile damage. Excavation contractors should be well aware of the consequences of
excessive soil movement induced by improper construction practice.
The designer of the foundation should also be
aware of the potential problem on site and should
take the actual construction practice into consideration in the design. For instance, if piles are expected
to experience the excessive tensile or bending stress
from the most practical excavation method, sufficient reinforcement should be provided in the design
at the first place. The foundation designer should
closely be involved in the design of the retaining
system and construction control on site during the
excavation period so that any possible negative effect can be minimized if not entirely avoided.

Figure 10. Finite element analysis results showing large degree

of basal heave caused by excavation with inadequate sheet pile
embedded length - Case III

REFERENCES

Bending Moment
(T-m)
-10

10

20

Displacement
(mm)

20

Balasubramanium, A.S. 1991, Inaugural Lecture on Contributions in Geotechnical Engineering, Soil Mechanics and
Foundation Engineering, AIT, Bangkok, March 1991
Turner, M. J. 1997. Integrity Testing in Piling Practice, CIRIA
Report No. 144, Construction Industry Research and Association, London.
Thasnanipan, T., Maung, A. W. and Tanseng, P. 1998a. Damages to Piles associated with Excavation Works in Bangkok
Soft Clay, 6th International Conference on Problems of Pile
Foundations Building, Ufa, Russia. September 14-18, 1998.
pp. 91-98
Thasnanipan, T., Maung, A. W. and Baskaran, G. 1998b.
Sonic Integrity Test on Piles Founded in Bangkok Subsoil
Signal Characteristics and Their Interpretations, 4th International Conference on Case Histories in Geotechnical Engineering, St. Louis, Missouri. March 9-12, 1998. pp. 10861092.

EXCAVATION AND BRACING SYSTEM

40

0.00
1ST STRUT
2ND STRUT
5.00

SHEET PILE

BORED PILE
10.00

Depth
(m)
15.00

Cracking
Moment of
Pile
Interface
between Soft
and Stiff
Clay

20.00

SOFT TO MEDIUM
CLAY

STIFF CLAY

MEDIUM CLAY

STIFF CLAY
25.00
HARD CLAY
30.00

Figure 11. Computed bending moment and deflection of pile

located closest to sheet pile - Case III

10 CONCLUSIONS
With the demonstration of three case studies, effectiveness of sonic integrity test in pile damage assessment in Bangkok is presented in this paper. This
research study also proves that external forces due to
both lateral and uplift movement of the ground induced by adjacent excavation can cause considerable
damage to constructed bored piles.
170

424

PROCEEDINGS OF THE SIXTH INTERNATIONAL CONFERENCE ON THE APPLICATION

OF STRESS-WAVE THEORY TO PILES / SAO PAULO / BRAZIL/ 11 13 SEPTEMBER 2000

Application of Stress-Wave
Theory to Piles
Quality Assurance on Land and Offshore Piling
Edited by

Sussumu Niyama

Institute for Technological Research IPT, Sao Paulo, Brazil

Jorge Beim

PDI Engenharia, Rio de Janeiro, Brazil

Non-destructive integrity testing on piles founded in Bangkok subsoil

N. Thasnanipan, A.W. Maung, T. Navaneethan & Z. Z. Aye
Seafco Company Limited Bangkok, Thailand

A.A. BALKEMA / ROTTERDAM / BROOKFIELD / 2000

425

Application of Stress-Wave Theory to Piles, Niyama & Beim (eds) 2000 Balkema, Rotterdam, ISBN 90 5809 150 3

Non-Destructive integrity testing on piles founded in Bangkok subsoil

N. Thasnanipan, A. W. Maung, T. Navaneethan & Z. Z. Aye
Seafco Co., Ltd., Bangkok, Thailand

ABSTRACT: Pile foundation is compulsory in Bangkok, Thailand, to support almost all types of structure
due to prevailing subsoil condition. Large bored piles of 600mm to 1800mm in diameter and barrettes
(0.8mx2.7m-1.5mx3.0m) with toe level up to 60m depth are commonly used to support heavy structures covering high-rise buildings, subway stations, elevated highways, etc. In order to test or check the quality of
such high load carrying foundation piles constructed in multi-layered soil conditions, three main types of nondestructive integrity tests such as sonic integrity, sonic logging and high strain dynamic load tests are usually
employed. Sonic integrity test is the most common method in integrity testing on both driven and bored piles.
Sonic logging test is occasionally employed as a part of pre-planned site quality control system. High strain
dynamic load test is commonly used for pile capacity evaluation. This paper presents an overview of these
three commonly used non-destructive tests on piles in Bangkok subsoil. Some findings from the authors involvement in research works are also presented.

10

Su= 40-140KPa
-20
t = 17-21kN/m3
-30
SPT-N= 20-50
t = 20kN/m3

tati
ine

cL

-40

raw

Su= 250KPa
= 21KN/m3
t

cD

Hard Clay
(CL)

-10

etri

40

Su= 10-20KPa
t = 14-16kN/m3

ros

Dense to Very
Dense Sand
(SM)

60

Hyd

30

40

om

Medium to
Stiff Clay
(CH)

20

iez

20

0
0

Weathered Crust
Bangkok Soft
Clay (BSC)
(CH)

Pore Pressure (t/m2)

al P

2 TYPICAL SUBSOIL CONDITION IN

BANGKOK

Engineering
Parameters

u
Act

Three techniques namely sonic integrity, sonic logging and high strain dynamic load tests are the
commonly used non-destructive integrity tests applied in Bangkok, Thailand. The effectiveness and
suitability in terms of both cost and method of application itself are the primary reasons of the greatest
growth in use of these tests. An overview of sonic
integrity, sonic logging and high strain dynamic load
tests conducted on piles particularly bored piles
founded in Bangkok subsoil is presented in this paper.

38m. Second Dense to Very Dense Sand layer occurs at about 46m depth. Between these sand layers,
Hard Clay layer is present.

Depth (m)

1 INTRODUCTION

wn

-50

Do

Subsoil profile is relatively consistent at different

Dense to Very
50
localities in Bangkok. A typical subsoil profile is
Dense Sand
SPT-N= 50-100
characterized by the alternating layers of clay and
(SM)
t = 20KN/m3
sand deposits as soil succession shown in Figure 1.
60
Weathered Crust of 1m to 3m thick occurs as upper
most layer underlain by Very Soft to Soft Clay
Figure 1. Typical soil profile of Bangkok plain with piecommonly known as Bangkok Soft Clay (BSC).
zometric draw down conditions (after Thasnanipan et al.,
Thickness of Soft Clay layer varies from 12m to
1998a).
18m depending on the location. Underneath Soft
Clay, lies Medium to Stiff Clay. Dense to Very
The depth of bedrock is not well determined in
Dense Sand layer also known as First Sand layer is
the Bangkok area but has been reported to be at least
found below Stiff Clay layer at depths about 25m to
171
e

Lin

-60

427

550m deep (Balasubramaniam, 1991). Existing

pore water pressure conditions in upper part of BSC
are hydrostatic from nearly 1m below the ground
level. Then the hydrostatic conditions change to piezometric draw down near bottom level of BSC.
3 COMMON CONSTRUCTION METHOD
The common practice of bored pile construction
method in Bangkok can be divided into wet and dry
processes.
For the wet process, bored piles are constructed
using bentonite or polymer slurry as drilling fluid
over the full length of the piles, to ensure the stability of the excavation at all stages. Top 15m-thick
soft clay was temporarily cased. Drilling for the
piles was done with a conventional drilling auger or
auger bucket. To clean the sediments at the borehole base the airlift technique or cleaning bucket is
applied. After lowering the rebar cage, concrete is
poured with tremie method.
Generally the dry process is employed for small
piles (0.35m to 0.6m in diameter and 25m in length).
This method uses a steel temporary casing of about
15m in length to protect the soft clay from caving in.
Boring is carried out by rotary drilling or percussion
tripod rig.
Bored piles are commonly constructed with reinforcement of 0.5-1.2% of sectional area of pile.
4 COMMON DEFECTS CAUSED IN BORED
PILES
Common pile defects such as size reduction/necking,
discontinuity, soil/slurry inclusions, etc. may be
caused during pile construction. Most of the cracks
in bored piles caused at post construction stage are
usually induced by construction activities associated
with basement excavation and improper trimming of
pile head to design cut-off level. Causes of defects
in bored piles during construction are;

Inadequate length to protect the soft clay layer

Delayed feeding of slurry and improper slurry
level maintenance for wet process
Long time duration of maintaining open borehole
Non continuous concrete pouring or disruption
in concreting
Lifting up of insufficient workable concrete or
hardened concrete upon extracting the temporary
casing
Contamination of concrete with drilling slurry or
soil
Soft pile toe due to improper base cleaning

428

172

5 THE SELECTION OF SUITABLE METHOD

In general, non-destructive tests are conducted to
verify the quality of the pile. Turner (1997) highlighted the factors in selection of pile testing in
CIRIA Report 144 as follows;
The perceived nature of possible features or defects within the pile
The ability of the test method to detect the feature or defect under investigation
The cost of testing and examination
The ease of use and interpretation
Like other indirect measurement of pile integrity,
sonic integrity and sonic logging tests are only capable of identifying structurally significant features.
Rejection or acceptance of individual piles should
not rely only on the results of these tests. Further
investigation, engineering evaluation and judgement
are highly recommended to confirm the defect detected by these indirect techniques. Both methods
however are very useful for cost-effective screening
test to identify piles with potential defect.
Sonic integrity test is usually selected for both
quality check (control test) and retrospective investigation. It is the cheapest in terms of cost and the
simplest in terms of testing process. The main advantage of this test is that since no particular measures/preparation is necessary during the pile construction phase it is more flexible to select which
pile is to be tested. However, interpretation of sonic
integrity testing needs considerable experience and
knowledge in testing, subsoil condition and construction method.
Sonic logging test is relatively expensive. It is
mainly employed as a pre-planed site quality control
testing. The major advantage of this method is that
test can be carried out shortly after the pile construction. Hence, rectification measures can be implemented while the foundation contractor is on site.
However, this method is generally not applicable if
pile integrity is in question due to post construction
activities as access tubes are usually grouted after
completion of the test.
High strain dynamic load test is usually selected
to verify the load carrying capacity of piles. Pile integrity can also be determined by high strain dynamic load test. The major advantages of this test in
comparison with static pile load test are those of
cost, time and space requirement. In some projects,
dynamic load test was applied to justify the capacity
of pile after rectifying its defect detected by other integrity tests. Table 1 shows the suitability of test
method for different type of common defects.

Table 1. The suitability of test method for different type of defects.

Test method and suitability
Type of defects
for defect detection
Sonic
Sonic
High strain
integrity
logging
dynamic
load test
Crack
Suitable
Possible
Suitable
Pile size
Suitable
Possible
Suitable (if
change
severe defect)
Possible
Suitable
Suitable (if
Poor quality consevere decrete or Contamifect)
nated concrete
Soft pile toe
Not usually Suitable

6 SONIC INTEGRITY TEST

6.1 Overview of the test in Bangkok
Sonic integrity test, also known seismic test is the
most common method of integrity testing for both
driven and bored piles in Bangkok. In many projects
it is a part of the contractual requirement to conduct
sonic integrity test. Minimum 10 % to maximum
100 % of production piles are commonly tested. It is
also a reasonably acceptable method for applying as
a retrospective investigation in determining integrity
of the pile.
6.2 Interpretation of the test results
As has been reported in various literatures, it is to be
emphasized that sonic integrity test is highly dependent on the experience in both field testing and
interpretation. The signal characteristics and their
interpretations of sonic integrity test on piles
founded in Bangkok subsoil were reported by Thasnasipan et al. (1998b). The authors presented the interpretation of various features of sonic integrity test
signals in comparison with soil investigation data
and construction records.
6.3 Some findings from sonic integrity test in
Bangkok

fully developed so that damping effect or attenuation

of the sonic signal by shaft friction is relatively low.
52.0m

16.0cm/
s
0

173
429

18

24

30

36

42

48

54

60

f:8

4000m/s
exp: 20

v2-88c
Sr

Figure 2. Sonic signal showing a toe reflection of bored pile at

52.0m tested few days after concreting.

Use of temporary casing usually causes larger size at

the top portion (from pile top to casing bottom) for
completed piles in Bangkok subsoil. This is caused
by the size difference between casing and the drilling tool (bucket/auger) as the top section of pile
formed by the outside diameter of the casing is usually 5-10cm larger than the lower section of pile
with nominal diameter formed in the borehole made
by drilling tool (Figure 3). Test signal shown in
Figure 3 indicates a clear reflection at approximate
11m from pile top (about 15m below ground level)
with a repeat or multiple reflection at about 2 times
of that distance (approximate 22m from pile top). As
the polarity of the reflection is negative, the first interpretation is that this pile has decrease in cross section or crack at 15m below ground level. However,
with the consideration of pile construction method, it
is concluded that negative reflection is caused by
variation in pile cross section at the bottom end of
15m long temporary casing. The sonic signals with
anomaly acquired on those piles are often misinterpreted as indicative of defect piles, leading to an argument.
Med.
clay Stiff clay Dense sand

Very soft clay

Temporary casing
bottom level

Pile Cut-off level

Outer diameter of
casing collar

11.0 cm/s

Toe reflections are generally clear for the piles with

length to diameter ratio (L/D) within 40, but toe reflection was hardly detected for piles founded at
depth deeper than 30-35m as reported by Thasnanipan (1998b). However, it was found that toe reflection is detectable for some deep-seated bored piles
of about 50m in length if pile was tested at early age.
Figure 2 shows the significant toe reflection of
52.0m-long pile tested few days after concreting.
This is considered mainly due to the fact that at early
stage the bonding between soil and pile has not yet

12

pile 30
14 Jan 91

Pile 66
25 Jun 97

12

15 18
4000 m/s
exp : 20

Nominal dia./dia. of
drilling tool

21

24

27
f: 8

30 v2-88c
sr

Figure 3. A typical pile shape constructed in Bangkok soil and

resulted sonic integrity test signal profile.

For the analysis of sonic integrity test, velocity of

sonic wave in pile material (concrete) is typically assumed 4000m/s in most cases. Basically, lower
strength of concrete gives lower wave propagation
velocity. The trend of sonic wave velocity increment with pile age derived from a number of 600mm
diameter bored piles with known lengths is presented in Figure 4.

Velocity (m /s)

4300

overcome by drilling through the obstruction in the

worst case (but not possible if tube is bent), it should
be eliminated in the first place by careful fabrication, installation and protection of the tubes.

4100
3900
3700

7.2 Interpretation of test results

3500
3300
0

18

27

Days

36

45

54

63

Figure 4. A curve of sonic wave velocity vs. pile age in Bangkok sub soil (fc 35 MPa).

7 SONIC LOGGING TEST

7.1 Overview of the test in Bangkok
Use of Sonic logging test for checking pile integrity
has increased in Thailand in recent years. It is used
for cast-in-place foundation elements such as bored
piles, barrettes, diaphragm walls and caissons. Both
PVC and steel tubes can be used as access tubes.
Though cost of steel tube is higher than that of PVC,
steel tubes are considered more suitable than PVC
tubes for two reasons; (1) less potential for bending
and damage especially for deep-seated bored piles
and barrettes with heavy steel reinforcement and (2)
better bonding with concrete. These access tubes are
sometimes also used for pile base grouting. The access tubes are attached to the reinforcement cage,
which is installed prior to concreting as a normal
practice for cast-in-place piling.
Two access tubes are required as a minimum for
sonic logging test. For a good coverage of the test
pile, recommended number of tubes for different
sizes of bored piles is shown in Table 2.

The results of a series of laboratory tests to determine the effects of various defects or inhomogeneities within a concrete section were reported by Stain
and Williams (1991). The test results were carried
out on small panels constructed to model various
pile construction defects and anomalies. The results
from these panels were related to tests on control
panels formed from homogeneous concrete.
Faiella and Superbo (1998) reported the analysis
of the sonic logging test results of over 6800 piles
collected from 37 sites in Italy. Based on the results
of analyses, the authors presented the defect classification criteria for piles monitored by different number of tubes as illustrated in Tables 3 (a) and 3 (b).
Table 3 (a). Defect classification criteria monitored by means
of 2 tubes (Faiella and Superbo, 1998)
Ta/Ts
<1.15
1.15-1.45
1.45-2.0
>2.0

Ta/Ts
<1.15

1.45-2.0

Pa/Pt
Pa<Pt
Pa=Pt
Pa<Pt
Pa=Pt
Pa<Pt
Pa=Pt

A typical problem encountered in sonic logging test

is the access tube blockage which often comes to
know only at the time sonic testing is conducted one
week or more after pile construction. The blockage
of access tube is mainly caused by intrusion of concrete from leakage at either bottom cap or tube connection and intrusion of soil or other material from
the top cap of the tube. Sonic test instruments
(transducer and receiver) sometimes may not be able
to insert down to the bottom of the access tubes due
to the bend of the tube itself. Though the problem
associated with the access tube blockage can be
174
430

Type of defect
Non-homogeneous concrete
Light
Probably serious
Light
Probably serious
Probably serious
Light

Table 3 (b). Defect classification criteria monitored by means

of 3 and 4 tubes (Faiella and Superbo, 1998).

1.15-1.45

Table 2. Recommended number of tubes for different pile size

Pile Diameter
Minimum No. of
Tube spacing
(mm)
tubes
(degree)
D 750
2
180
750<D 1000
3
120
1000<D 1500
4
90
1500< D 2500
6
60
2500 < D
8
45

Tk/D
<3
>3
<1
>1
<0.5
>0.5

>2.0

Pa<Pt
Pa=Pt

Tk/D

Type of defect

Non-homogeneous concrete

<3
>3
<3
>3
<1
1-3
>3
<1
>1
<0.5
0.5-3
>3
<0.5
>0.5

Non-homogeneous concrete
Light
Light
Serious
Non-homogeneous concrete
Light
Serious
Light
Serious
Light
Serious
Serious
Light
Serious

where:
Ta = Travel time of sonic waves in the anomalous zone
Ts = Travel time of sonic waves in the sound concrete
Tk = Thickness of the anomalous zone
Pa = Number of measurement paths affected by sonic anomalies
Pt = Total number of travel paths for each pile
D =Pile diameter

Srivanavit et al. (1999) reported the interpretation of

test signals in comparison with the actual integrity of
9 model tests conducted in Bangkok. Model piles
are of 800mm diameter with 1m length. The summary of the model test results including the description of model piles as well as discussions by authors
are illustrated in Table 4.

Case
6

Table 4. The summary of sonic logging model test results

Case
Concrete & ma- Profile
Discussion of
terial properties
signal
1
Standard good
Sonic logging
concrete pile
profile shows
with clean tubes
continuos sigand clean water.
nal, no anomaly.

Good concrete
pile with dirty
tubes and clean
water.

Sonic logging
profile shows
defect at pile
head due to
bleeding of concrete.

Good concrete
pile with dirty
tubes and dirty
water (contaminated with bentonite slurry)
filled.

Sonic logging
profile shows
defect at pile
head due to
bleeding of concrete, similar to
case 2.

Good concrete
with Layer of
bentonite
slurry, sand,
gravel fill
(30cm.Thickness
whole Section) at
middle of the
pile.

Sonic logging
profile shows
signal loss from
0.4m to 0.7m
due to delay of
sonic wave.

Good concrete
with Clay on
surface of one
tube (50 cm.
from top of concrete).

Concrete & material properties

Concrete mixed
with bentonite
slurry.

Profile

Discussion of
signal
Sonic logging
profile shows
non-uniform
signal, not clear
signal due to
poor concrete
and inhomogeneous concrete.

Thin smear of
2mm. bentonite
slurry on tubes
(50 cm. from
top of concrete).

Sonic logging
profile shows
signal loss from
depth
0-0.5m (more
travel time).

Thick smear of
5mm. bentonite
slurry on tubes
(50 cm. from
top of concrete).

Sonic logging
profile shows
continuous signal loss and signal delay at
depth 0-0.5m,
similar to case
7.

Weak concrete
(high w/c ratio).

Though weak
(low strength)
concrete is used,
sonic logging
profile shows
continuous signal, because of
homogeneity of
the tested media.

7.3 Factors to be considered in interpretation of

sonic logging test

Sonic logging
profile shows
signal delay at
0.5m depth.

Turner (1997) pointed out the need of both theoretical knowledge and practical experience in interpretation of the test results in CIRIA Report 144. Suggestion is also made in this report that the anomaly
shown in sonic test signals can be caused not only
by changing in physical properties of the materials,
but also by factors within the measuring system itself. These factors are;

175
431

Free movement of the probes within access tubes

Mismatched probe positions especially at pile toe
Measurement resolution
Incorrect position of access tube
Air gaps or different material around access
tubes
Aggregate variation (in the case of base grouting)

It is to be noted from the authors suggestion based

on the above factors that only variations of transmit
time more than 15 to 20% of the norm should be regarded as warranting further investigation.

defect by other integrity tests. Figure 5 shows the

photo of dynamic load test carried out on barrette of
size (1.0x2.7m) with toe depth 48.94m located at the
bank of water supply canal. A 20 ton hammer was
used to activate the 2050 ton test load in this project.

8 HIGH STRAIN DYNAMIC LOAD TEST

8.1 Overview of the test in Bangkok
High strain dynamic integrity test has become a
well-accepted method especially for evaluating the
pile capacity in the foundation industry today. A
large number of related technical papers and case
histories of the test have been published and it is a
part of standards and specifications such as ASTM
D4945-89 (Standard Method for High-strain Dynamic Testing of Piles). In Bangkok, dynamic load
test is applied for both driven and bored piles. A
summary of the number of dynamic load test carried
out during 1991 to 1997 is shown in Table 5. It is to
be noted that quantities shown are collected from the
available sources and the actual tested numbers are
likely to be slightly more than those indicated in the
table. As can be seen in the table, application of dynamic load test increases year by year in Thailand.
The pile driving analyzer (PDA) with computer
software by Pile Dynamic INC, USA is mainly used
in Bangkok.

Figure 5. Dynamic load test carried out on a barrette

(1.0m x 2.7m x 48.94m).

8.2 Comparison and correlation of dynamic and

static pile load test
A comparison between dynamic and static load test
results has been reported by various researchers.
Seidel and Rausche (1984) reported the results of
dynamic and static load tests performed on drilled
shafts of the West Gate Freeway in Melbourne, Australia. A 20 ton hammer with drop heights between
1.6 and 2.5m was used for the dynamic tests of 12
shafts ranging from 1100mm and 1500mm in diameter and 35m to 64m in length. The authors reported
that dynamic activation of static pile resistance
forces exceeded 3000 ton for some 1500mm diameter shafts. Skin friction predictions from dynamic
load tests and values obtained from instrumented
shafts under static load tests were remarkably similar and pile head load-movement relationships obtained from both test methods were comparable as
reported by the authors.
Prebaharan et al. (1990) reported the results of
dynamic and static tests conducted on bored piles at
the Marina Bay Station of the Singapore Rapid
Transit System. According to the report, dynamic
load tests were carried out for bored piles of
1000mm diameter and 25m to 50m length founded
in Old Alluvium of Singapore Island. Agreement
between dynamic and static load tests results permitted replacement of 44 out of 51 tests with dynamic
load test from originally planned static load tests as
stated by the authors.
Vasinvarthana and Kampananon (1997) presented
the efficiency and reliability of dynamic load test
carried out in Bangkok. Figures 6 (a) and 6 (b) show
the load settlement curves of dynamic and static load
tests conducted in Bangkok on large and small di-

Table 5. Yearly minimum quantity of dynamic load test conducted in Thailand during 1991 to 1997
Year
No. of dynamic load test done
1991
84
1992
134
1993
242
1994
369
1995
464
1996
473
1997
695

For driven pile it is usually performed at two stages,

during driving (initial driving monitoring) and some
period after installation of pile (restrike test). For
bored piles, a specially designed pile cap is normally
required as an integral part of the pile head to avoid
pile head damage. Ram weights of 4 ton to 20 ton
are commonly used in Bangkok. The common ram
weights used for different pile sizes are shown in
Table 6.
Table 6. Common ram weights used for different size of piles
Pile size (m)
Ram weight (ton)
0.50
4.00
0.50 0.80
8.00
1.00 1.50
20.00

The dynamic load test is also occasionally used as

an additional investigation when pile is found with

176

432

ameter piles respectively. As can be seen in the figures, load-settlement characteristics obtained from
dynamic load test agree well with those of static load
test for both large and small diameter bored piles.

ACKNOWLEDGEMENT
The authors wish to express their appreciation to
EDE Co., Ltd. for providing the test data. Thanks
are also given to Mr. Veera Vasinvarthana and Mr.
Natomon Kampananon STS Engineering Consultants Co., Ltd. for giving permission for the use of
their previous works in this paper.

Settlem ent (m m )

-5

REFERENCES

-10

American Society for Testing and Materials 1989. ASTM

D4945-89 Standard method for high-strain dynamic testing
of piles, Philadelphia: American Society for Testing and
Materials.
Balasubramaniam, A. S. 1991. Inaugural lecture on contribution in geotechnical engineering soil mechanics and foundation engineering, AIT, Bangkok, 14 March 1991.
Faiella, D. & Superbo, S. 1998. Integrity non destructive tests
of deep foundations by means of sonic methods - Analysis
of the results collected on 37 sites in Italy, Van Impe &
Haegeman (eds), Proc. 3rd Int. Geotechnical Seminar on
Deep Foundations on Bored and Auger Piles, BAP III,
Ghent, Belgium, 19-21 October 1998: 209-213. Rotterdam:
Balkema.
Prebaharan, N., Broms, B., Yu, R. & Li, S. 1990. Dynamic
testing of bored piles, Proc. of the Tenth Southeast Asian
Geotechnical Conference, Volume 1, Taipei, ROC, 16-20
April 1990: 373-378.
Seidel, J. & Rausche, F. 1984. Correlation of static and dynamic pile tests on large diameter drilled shafts, Proc. 2nd
Int. Conf. on the Application of Stress Wave Theory on
Piles, Stockholm, Sweden, 27-30 May 1984: 313-318.
Stockholm: Swedish Pile Commission.
Srivanavit P., Hiranoon J. & Pimsarn S., 1999. Integrity tests
on deep bored piles constructed in Bangkok subsoil by
sonic logging method & interpretation, The engineering
technology exhibition and symposium, Organized by Engineering Institute of Thailand, Bangkok, 1-2 November
1999: 1-8. (in Thai).
Stain R. T. & Williams, H. T. 1991. Interpretation of sonic
coring results: A research Project, Proc. 4th Int. DFI Conf.
on Piling and Deep Foundations, Stresa, Italy, 7-12 April
1991: 633-640. Rotterdam: Balkema.
Thasnanipan, N., Baskaran, G., & Anwar, M.A. 1998a. Effect
of construction time and bentonite viscosity on shaft capacity of bored piles, Deep Foundations on Bored and Auger
Piles, Van Impe & Haegeman (eds), Proc. 3rd Int. Geotechnical Seminar on Deep Foundations on Bored and Auger
Piles, BAP III, Ghent, Belgium, 19-21 October 1998: 171177. Rotterdam : Balkema.
Thasnanipan, N., Maung, A. W. & Baskaran, G. 1998b. Sonic
integrity test on pile founded in Bangkok subsoil signal
characteristics and their interpretations, Proc. 4th Int. Conf.
on Case Histories in Geotechnical Engineering, St. Louis,
Missouri, 9-12 March 1998: 1086-1092.
Turner, M. J. 1997. Integrity testing in piling practice, CIRIA
Report 144, London: Construction Industry Research and
Information Association.
Vasinvarthana, V. & Kampananon, N. 1997. Efficiency and
reliability of dynamic load test, Seminar on Foundation
97, Organized by Engineering Institute of Thailand, Bangkok: 79-94. (in Thai).

-15
Dynamic load test
Static load test
-20
0

200

400
Load (ton)

600

800

Figure 6 (a). Load-settlement curves of dynamic and static load

tests on large diameter piles ( 0.80m x 50.0m).

Settlem ent (m m )

-5

-10

-15
Dynamic load test
Static load test
-20
0

20

40
60
Load (ton)

80

100

Figure 6 (b). Load-settlement curves of dynamic and static

load tests on small diameter piles ( 0.35m x 18.91m).

9 CONCLUSIONS
Use of sonic integrity, sonic logging and high strain
dynamic tests in Bangkok has been presented.
Though sonic integrity test is a simple and cost
effective method, the reliability of the test results is
highly dependent on the experience of the person in
both field-testing and interpretation.
The results from sonic logging test conducted on
model piles in Bangkok helps to extend the knowledge of the signal characteristics and interpretation.
High strain dynamic test has become a wellaccepted method especially for load testing of piles.
Agreement between static and dynamic load test results enhances the confidence in using the dynamic
load test as shown by rapid increment in number of
tests conducted in seven years in Thailand, mainly in
Bangkok.

177

433

THE SIXTH INTERNATIONAL CONFERENCE ON PROBLEMS OF PILE FOUNDATIONS

BUILDING, RUSSIA, SEPTEMBER 14-18, 1998

Problems of Pile
Foundations Building
Damages to piles associated with excavation works
in Bangkok soft clay
N. Thasnanipan, Maung A. W. and Tanseng P.
Seafco Company Limited Bangkok, Thailand

ISBN 5-88151-161-1

435

The 6th International Conference on Problems of Pile Foundations Building, Russia, September 14-18, 1998

DAMAGES TO PILES ASSOCIATED WITH EXCAVATION WORKS IN BANGKOK SOFT CLAY

Narong Thasnanipan
SEAFCO Co., Ltd.
Bangkok 10320
Thailand

Aung Win Maung

SEAFCO Co., Ltd.
Bangkok 10320
Thailand

Pornpot Tanseng
SEAFCO Co., Ltd.
Bangkok 10320
Thailand

ABSTRACT
In Bangkok, Thailand, pile foundations are commonly used to support all type of structures. The excavation for construction of
pile caps and mat foundations are carried out mostly in the soft clay layer. Sheet pile walls with temporary bracing systems are
commonly used for excavation works (depth ranging form 4.0-11.3m) in the built-up areas of central and outskirts of Bangkok
metropolis. Sometimes open cut excavation is adopted for shallow excavation (2.0-6.0m) in the areas where no buildings exist in
the close vicinity of the site. Lateral movements of soils take place in such excavations causing pile deviation and cracks to the
piles, especially located in the vicinity of boundary of excavation. Such defects have been found more frequently in smaller
diameter piles than the larger diameter piles. These tension cracks in piles are commonly observed above the level of interface of
soft and stiff clay layers where excavation induced bending moment in pile exceeds the piles cracking moment capacity. Sonic
integrity testing is usually employed to test the piles integrity. High strain dynamic load test is sometimes employed to check the
performance of piles with such defects. This paper presents four cases in which damages caused to piles by soil movement
associated with excavation works. Pile damages, especially tension cracks indicated by integrity testing, analysis on stresses in
pile due to the soil movements and behavior of retaining systems are also examined. Correlation between location of crack and
bending moment in pile which exceeds the pile cracking moment capacity also made.
KEYWORDS
Pile deviation, Lateral movement, Sonic integrity test, Dynamic load test, Pile damage, Bending moment, Cracking moment

91

437

INTRODUCTION

TYPES OF EXCAVATIONS

Almost every structure in Bangkok sits on pile foundations.

Excavation works for preparing the pile cap, footing and/or
mat are usually within the soft clay commonly found in
Bangkok. However for structures requiring many basement
levels and underground facilities, the excavation works are
undertaken to below the soft clay more than 20m from the
ground. Various soil retaining systems are adopted in such
excavation works to suit the site and soil conditions and
depth of excavation. Analysis and design for the retaining
systems sometimes are overlooked by the practicing
engineers for the possible movements, both vertical and
lateral, of soil adjacent to the excavation boundaries.
Moreover monitoring of soil movements, with exception of
existing structure near by, is often ignored. Such ignorance
leads to damage the piles already constructed or installed in
soft soil stratum. As a result secondary defects found in the
piles after completion of piling work, including pile top
deviations far from allowable tolerance become an argument
between the piling contractor, excavation subcontractor and
the employer or supervising engineer.

Common type of excavation found in Bangkok subsoil are

tabulated as below,
Table 1. Common Types of Excavations in Bangkok Subsoil
Type of Excavation Excavation
Site
below G.L.
Conditions
(m)
Cut Slope
3.0-6.0
open area, no
adjacent buildings
Sheet Piles Wall
4.0 to 11.3
adequate site
clearance
Jet Grouted Wall
5.0 to 10.5
adequate site
clearance
Secant or
4.0 to 20.5m
in built-up area
Contiguous Pile
Wall
Diaphragm Wall
7.0 to 23.0m
in built-up area
First two types of excavations are found to induce higher
lateral and vertical soil movements compared to the
remaining types (Ref. 2) and sometimes cause a localize
failure of the soil mass retained due to improper
construction practice. In this paper pile damage associated
with the first two types in four cases are studied and
presented.

PILE FOUNDATIONS
Both bored and driven piles are commonly used for
foundations of structures in Bangkok subsoil. The subsoil
generally consist of a 12-18m thick very soft clay layer
overlying medium to stiff clay layer. A sand layer occurs at
depth of 20m to 35m below the stiff clay layer. A second
layer of dense sand is generally ound below 50m where the
toe of deep bored piles are founded for high rise or heavy
structures. Stiff to hard clay layers occurs intermittently
between the sand layers. Large diameter bored piles having
diameter of 800mm to 1500mm are extended down to 35m
to 60m respectively being embedded in the sand layers. For
low rise buildings driven or small diameter bored pile of
35mm to 600mm are commonly used with footing for pile
groups or for individual piles. The length of pile is
generally to 20 to 28m embedded in stiff clay or first sand
layer.

INTEGRITY OF PILES
Sonic integrity test is usually employed to test the integrity
of piles after piles have been trimmed to the design cutoff
level. Generally a minimum of 10% to a maximum of 100%
of piles are tested. A sonic integrity tester with built-in
computer having high quality signal acquisition system has
been commonly used for pile integrity testing. However,
severity of the defect/crack or any irregularity in piles is
usually determined comparatively based on magnitude of
reflection amplitude of individual piles from such a
defect/crack or irregularity, with reference to the input
pulse. Figure shows the sonic test records of a good pile
and piles with crack in different severity from Cases IV.

The working load capacity of bored piles is usually limited

by the allowable stress of the concrete. Cutoff level of such
piles generally varies from about 2.0m to 6.0m below
ground surface. For the structures with deep basements,
bored piles are trimmed down to 20.0m.

CASE I - CUT SLOPE (Bangna-Trad Highway KM. 26)

Subsoil Condition - A 17m thick soft clay layer occurs with
unconfined average shear strength (Su), ranging from 0.69
t/m2 at the top to 2.20 t/m2 at the bottom of the layer. The
field vane shear test indicates in-situ shear strength of that
layer is 1.40t/m2 to 3.80t/m2 with sensitivity of 2.7 to 3.7.
The natural water content of soft clay layer exceeds 100%.
Below the soft clay layer is medium clay layer with Su value
of 4.0 t/m2 to 6.0t/m2. Then a stiff clay layer having Su
value of 13.0t/m2 is found at the depth of 21m-22m
overlying the first sand layer which extends to the depth of
about 34m to 37m. The second sand layer occurs at the

Generally bored piles are constructed with reinforcement of

0.5-1.2% of sectional area of pile and concrete having cube
strength of 30-32 MPa. Wet process, using bentonite slurry
is normally employed for drilling in water-bearing sand
layers and concrete pouring is done under the slurry. Dry
process normally used for construction of small diameter
bored piles with depth not greater than 26m.

92

438

Most piles with large crack detected were located in the

outer row of foundation piles ( Fig. 1) and located under the
traffic ramps of temporary working platform. It was found
that the crack in these piles were caused by localized soil
sliding which was occurred by the heavy traffic load. Pile
deviations in order of 50cm to 100cm towards excavation
were observed.

depths of about 52.0m - 61.0m. Between sand layers a

thick hard clay layer is present, but sporadically.
Pile foundation - A total of 428 concrete bored piles of
diameters 1.0m, 1.2m and 1.5m were installed with pile tip
at 53m to 65m below the ground, being embedded in second
dense sand layer to support a high rise structure. Cutoff
level of piles ranges from 4.6m to 5.9m in depth. Entire
length of piles were fully reinforced with steel
reinforcement of 1% and 0.5% of pile sectional area in top
33m and lower section of pile shaft respectively.

An inspection after the soil sliding revealed that king posts

were not installed to support parts of the ramps behind the
slope on the ground. Apart from the traffic ramp area, a
number of localized soil slidings up to 3m to 4m occurred in
some areas. They are found to be caused by loose materials
filled in bored hole above pile casting level during casting.
Less Prominent crack detected in some piles, particularly
within excavation zones are considered to be caused by
improper pile trimming method. A backhoe was used to
break off the over cast section by hitting the pile head after
chipping of the concrete cover at the cutoff level around the
pile shaft and separating the dowel bars form the overcast
section.

Soil Retaining System - Cut slope excavation with two

levels of 1:3 slope and a berm was adopted to achieve the
maximum excavation to -5.9m below the existing ground
level as the site is a large open space being located outside
built-up area. Slope stability analysis using effective stress
soil parameters indicated that a safety factor of 1.2 for
temporary condition can be achieved by driving 6m long I12 RC soldier piles at a spacing of 1.5m at the toe of the
bottom slope. Moreover, the ground level around the
perimeter of excavation was reduced down to 0.5m.
Excavation was carried out on temporary working platforms
supported by 24m long steel king posts (H300x300) driven
into the ground.

Soil around the suspected pile with prominent crack were

excavated down to detected crack level. These piles were
trimmed down to below the crack and re-constructed to the
cut off level. Tie beams were provided for the piles with
crack and horizontal deviation more than allowable
tolerance.

Pile Damages - All piles were tested with sonic integrity

test. Test results are summarized in the Table below;
Table 2. Summary of Pile Damage Detected by Sonic
Integrity Test - Case I
Pile
No. of Piles with
Defect
Dia.
Least
Less
Prominent
Depth
Prominent Prominent
Crack
below
(m)
crack
Crack
GL(m)
1.0
4
8
7
6.0-13.8
1.2
10
2
6.1-7.6
1.5
1
9.9

CASE II - SHEETPILE EXCAVATION SUPPORTED

BY SOIL BERM (Huay Kwang Area)
Subsoil Condition - A 12m thick soft clay layer occurs with
unconfined average shear strength (Su), ranging from 1.0
t/m2 at the top to 2.50 t/m2 at the bottom of the layer. The
natural water content of soft clay layer is about 60-80%.
Below the soft clay layer is medium clay layer with Su value
of 2.5 t/m2 to 3.5t/m2. Then a stiff clay layer with traces of
fine sand having SPT N value of 15 to 37 is found at the
depth of about 18m-overlying sand layers which occur
below 27m in depth.
Pile foundation - A total of 904 concrete bored piles of
diameters 1.0m and 1.5m were installed with pile tip at
about 59m below the ground, being embedded in dense sand
layer to support a total of eight high rise buildings located at
the particular size. Cutoff level of bored piles ranges from
5.1m to 8.1m in depth. Entire length of piles are fully
reinforced with steel reinforcement of 0.72% of pile
sectional area for the top and reduced gradually to 0.23% for
bottom section of pile shaft. A total of 88 pre-cast concrete
(I-30) piles with 21m in length was also driven to support
the underground water tanks. The cutoff level of the precast piles was 3.1m below the ground level.
Soil Retaining System - The foundation structure is about
10m to 20m away from the adjacent property boundaries.
Initially excavation with cantilever sheet pile wall (FSP III

Figure 1. Excavation Plan and Location of Damaged Piles

93

439

with 14m in depth) supported by a soil berm or slope was

adopted to achieve the maximum excavation to -8.10m
below the existing ground level. However after excavation
to some depths, due to sheet pile failure at some locations
and excessive vertical and lateral soil movements,
temporary raking struts were then immediately installed.
Tension cracks wider than 300mm were observed on the
ground about 4.0m away from the excavation zone. Minor
damages also occurred to the adjacent properties due to soil
movement.
Pile Damages - More than 50 % of piles were tested with
sonic integrity test.
Table 3 Summary of Pile damage Detected
Integrity Test - Case II.
Pile
No. of Piles with
Dia.
Least
Less
Prominent
Prominent Prominent
Crack
(m)
crack
Crack
I-30
49
14
1.0
1
23
4
1.5
1
1

Figure 2.1. A Typical Scene showing excavation works

Case II

by Sonic
Defect
Depth
below
GL (m)
6.2-8.0
13-16.6
15.6

CASE III SHEETPILE EXCAVATION WITH ONE

LEVEL TEMPORARY SUPPORT (Yannawa)
Subsoil Condition - In this project site, a 14-15m thick soft
clay layer occurs with unconfined shear strength (Su)
ranging from 1.0 t/m2 to 2.00 t/m2 at the top and bottom of
the layer respectively. Below is stiff to very stiff clay layers
extend down to about 42m in depth. The subsoil condition
below 36m from the ground level is variable, having a 13m
thick sand layer sporadically. A stiff to hard dark grey clay
is found at the depth between 40m and 49m. Generally a
thick dense sand layer occurs below 50m in depth.

Large numbers of piles with crack detected were located in

the periphery of the excavation zone (within the soil berm) .

Pile foundation - A total of 402 concrete bored piles of

diameters 1.0m and 1.5m were installed with pile tip at
about 55m below the ground, being embedded in dense sand
layer to support a building structure. Cutoff level of bored
piles ranges from 0.75m to 7.80m in depth. Entire length of
all piles are fully reinforced with steel reinforcement of
0.5% of pile sectional area.
Soil Retaining System - Due to different cutoff levels of
piles, sheet pile (14m) with one level of temporary bracing
at about 1m below the ground level was used for each
excavation zone. Figure 3 shows a portion of pile layout
which includes an excavation zone with braced sheet pile
wall. Prior to installation of temporary struts, excavation
was carried out to about 2.5 to 3m in depth. The over
excavation without support caused excessive movements of
soil and sheet pile wall along grid line C10 (Fig.3).
Deflected sheet piles were re-aligned and struts were
installed immediately with pre-loading.

Figure 2. Location of Damaged Piles (only a quarter of

Project Site shown here) Case II
Figure 2 shows a portion of the piling plan affected by soil
movements. Pile deviations of up to 600mm, particularly in
pile with prominent crack towards excavation were
observed.
Bored piles with crack were cored to the depth below the
crack and grouted. High strain dynamic load test was
carried out on two piles suspected with crack and large
horizontal deviation after remedial work. The dynamic test
results indicate that these piles could be capable of carrying
the design load with factor of safety higher than 1.5.

Pile Damages - All piles were tested with sonic integrity

test after pile were trimmed to the designed cutoff level.
Piles with crack detected were mostly located in the vicinity
of the excavation boundary (Fig. 3 and Ref. 6) . Moreover,
lateral pile deviation in a range of 100-840mm towards
excavation were also observed in those piles located in the
vicinity of the sheet pile wall ( Fig. 3).

94

440

Table 4. Summary of Pile damage Detected

Integrity Test - Case III.
Pile
No. of Piles with
Dia.
Least
Less
Prominent Prominent Prominent
(m)
crack
Crack
Crack
1.0
5
1.5
14

by Sonic
Defect
Depth
below
GL (m)
7.5-9.4
8.4-21.0

Damaged piles were cored to the depth below the crack and
grouted. Modification of reinforcement for the mat was
done for defect piles with large horizontal deviation.

Figure 4. Plan of Excavation and Location of Damaged

Piles (a Portion of Project Area) Case IV
Pile Damages - All piles were tested with sonic integrity
test. In this particular project, 84% of piles within the
excavation zone were found to be cracked (Fig. 4 ).
Table 5. Summary of Pile damage Detected by Sonic
Integrity Test - Case IV.
Pile
No. of Piles with
Defect
Dia.
Least
Less
Prominent
Depth
Prominent Prominent
Crack
below
(m)
Crack
Crack
GL (m)
0.6
38
33
6.5-13.8

Figure 3. Plan of Excavation and Location of Damaged

Piles and Pile Movements (a Portion of Project Area) Case
III
CASE IV - SHEETPILE EXCAVATION WITH TWO
LEVEL TEMPORARY SUPPORT (Sukhumvit Area)
Subsoil Condition - A 14-15m thick soft clay layer occurs
with unconfined shear strength (Su), ranging from 1.0 t/m2
at the top to 3.00 t/m2 at the bottom of the layer. The natural
water content of soft clay layer is generally about 37 - 87%.
Below is very thick stiff to very hard clay layers interbedding with a 3m thick medium clay layer at the depth of
19m.

Figure 5. Sonic Integrity

Test Records showing
Good pile, Pile with less
Prominent
Crack
and
Prominent
Crack
respectively - Case IV.

Pile foundation - A total of 143 concrete bored piles of

0.6m diameter were installed with pile tip at about 26m
below the ground to support a residential apartment. Dry
process was employed to construct the piles. Cutoff level of
bored piles ranges from 0.85m to 5.95m in depth. Entire
length of all piles are fully reinforced with steel
reinforcement of 0.35% of pile sectional area.

Five bored piles with crack suspected were cored to verify

the crack and grouted later. High strain dynamic load test
was also carried on a total of 25 piles in which 24 piles with
crack and 1 pile without crack. Summary of dynamic load
test results is presented in the Table 5.
Table 6. Dynamic Load Test Results of Damaged Piles Case IV
Mobilized Capacity/Design Ultimate
No. of
Capacity of 240 ton
Pile
less than 50%
1
less than 10%
2
More than 100%
22

Soil Retaining System - Type FSP III sheet pile (14m) with
two level of temporary bracing at about 1.0m and 3.5m
below the ground level was used. Figure 4 shows a portion
of pile layout which includes an excavation zone with
braced sheet pile wall.

95

441

From the results, only a few piles lead to low safety factor to
carry the designed load and also connected by mat.
Modification of reinforcement for the mat was done
according to redistributed load.

17.00
0

1.5 2.5

Su (t/m 2)

0.0

0.0

1.0

2.0

SPT (blows/ft)
3.0

4.0 0

20

40

60

Crust

Sheet pile

5.0

Soft CLAY

10.0

FINITE ELEMENT ANALYSIS WITH COMPUTER

MODEL SIMULATION

Very Stiff
CLAY

Two dimensional finite element computer modeling was

carried out using PLAXIS computer program (ref. 7) for
simulation of staged excavation process, pile/soil
movements and to analyze the associated bending stress in
the piles of the above projects. Stiffness of foundation piles
per linear meter (E.I/pile spacing in row) was used as a
parameter for plate elements, where E and I is Young
Modulus and moment inertia of pile respectively. The soil
between piles in a row is ignored.
Mohr-Coulomb
constitutive model is adopted with soil parameters derived
from the soil borehole data at the site. A plane-strain
analysis approach was adopted with undrained soil
condition.

20.0

10

6.00

32.00

9.00

18.00

Center
Line

Su (t/m2)
0

Slope
1:3

Slope
1:3

5.0

Mat Zone

4.0

8.0

50

30.0

20.
0

Sheet pile

14.0

20.0

4.
0

6.
0

SPT (Blows/ft)
0

2
0

4
0

Sheet pile
Soft to Vsoft
CLAY

Medium CLAY

Stiff to Vstiff
CLAY

Table 6. Moment Capacity of Piles and Excavation Induced

Moments in the Piles Closer to Excavation Boundary.
Ultimate
Moment
Case Pile Rebar Cracking
Moment
by FEM
Size
Moment (Whitney)
Model
(m)
(%)
(t.m)
(t.m)
(t.m)
I
1.0
1.0
30
100
86.4
109.5*
II
1.0
0.75
30
79
149.7
III
1.5
0.5
102
188
117.2
255.7*
IV
0.6
0.35
6.5
9.2
13.3
Note: 1. (*) moment in pile located in active side of soil
mass either outside of sheet pile wall or cut slope.

Su (t/m2)
SPT (Blows/ft)
2.5
5.0 0 20
40
60

Weathered
CLAY

Soft to Very
Soft CLAY

10.0

15.0

2.
0

DISCUSSION

100

MAT FOOTING

5.0

Su (t/m 2)

The modeling results indicate that bending moments in piles

induced by excavation are higher than the cracking moment
capacity of the piles in all cases. Summary of analysis is
presented in the Table below;

0.0
10

0.
0

Crust

Crust

Figure 6. Model Profile of Case I

CL

25.
0

Dense to
Very Dense
SAND

9.50

Figure 9. Model Profile of Case IV

35.0

40.50

15.
0

Stiff to Very
Stiff CLAY

0.0

1
0

10.
0

25.0
30.0

5.
0

Medium
CLAY

20.0

0.
0

Soft CLAY

10.0
15.0

Hard CLAY

Figure 8. Model Profile of Case III

SPT (Blows/ft)
0

Medium CLAY

25.0

Model profiles are presented in Figures 6 to 9. Figure no.

10 shows computed bending moments in pile and lateral
deflection of pile in Case IV.

0.0

Medium CLAY

15.0

Medium
CLAY

Stiff to Hard
CLAY

25.0

Figure 7. Model Profile of Case II

In Cases I and III, piles located in the active side of soil

mass are found to be under the bending stress about 10%
and 36% higher than the ultimate bending capacity of piles
respectively.
In Case II, the computed bending moments in the pile
closest to the sheet pile wall and piles in second row is 89%
and 27% respectively higher than the ultimate bending

96

442

capacity of piles. The sonic integrity test also detected

presence of crack in 35% of the piles which are located in
first two outer rows (see Fig. 2).

Bending Moment (T-m)

-20

-10

10

Displacement (mm)
20

40

20

0.00

In Case IV, the modeling results show that excavation

induced bending moments in pile exceed the ultimate
moment capacity of piles. This is confirmed by the
observed pile damages (see Fig. 5). Moreover the model
results suggest that these piles are subject to some degrees
of basal heave due to flexibility of and soil condition at
embedment of sheet pile wall. Sonic integrity test results
shows presence of more than one crack in some piles at
depths, suggesting piles would have experienced combined
stresses. Comparison between Case III and Case IV
illustrates that cracks are found more frequently in small
piles than large piles in a case of soil movements due to
excavation.

5.00

Depth (m)

10.00

15.00

Cracking
Moment of Pile

Interface
between Soft
and Stiff Clay

20.00

25.00

Figure 10. Computed

Bending Moment in
Pile
and
Piles
Deflection Closest to
Sheet Pile Wall - Case
IV

The crack location in piles indicated by sonic integrity test

generally is found to be consistent with the location of
computed bending moment which exceeds the cracking
moment capacity of piles for all cases.
Table 7. Locations of Crack and Computed Bending
Moments in Piles in the Vicinity of Excavation Boundary
Integrity Test
Depth of B.M
Depth of Crack
Depth of
by FEM
Case
in Piles
Max. B.M
exceeding
near Excavation
by FEM
Piles Cracking
Boundary
Moment
(m)
(m)
(m)
I
6.0 & 13.8
19.5
13.5
II
12.7 & 13.7
18.0
11.0 & 13.0
III
12.4 &19.0
15.0
12.8 & 14.5
IV
11.5 & 12.0
15.5
13.0 &14.0

Figure 11. Section Showing Displacements of Soil - Case

IV
Instrumentation such as installation of inclinometer and
strain gauges in piles and active side of soil mass is
suggested for the future projects to analyze behaviour and
performance of piles subject to lateral loads induced by
excavation.

FEM analysis indicates the maximum bending moment in

piles occurs at the level close to the interface between soft
clay and stiff clay layers. For lateral movements of piles
and soil, a direct comparison can not be made between
model and observed data as the accurate field monitoring
data are not available. Performance of soil retaining
systems by model analysis are summarized in the Table
below;
Table 8.
System
Case
I
II
III
IV

Summary of

Lateral Movements of Retaining

Type of
Retaining
(m)
Cut Slope
Sheet Pile
Sheet Pile
Sheet Pile

Level
of
Strut
1
1
2

Excav.
Depth
(m)
5.4
5.7
8.0
6.0

CONCLUSION
This paper presents four case studies on bored piles
damaged by excavation induced lateral soil movements in
Bangkok soft clay layer. Performance and response of pile
are studied by FEM modeling and confirmed with pile
integrity test results. Causes of pile damages found are also
discussed.

Max. Lateral
Movement
(mm)
117
175
56
52

In addition to FEM analysis, the followings were found

from the actual site construction activities,
1. Over excavation prior to installation of support
2. Inadequate retaining system which can effectively
control soil movement.
3. Inadequate support for construction traffic load (Ref. 4)
4. Improper pile trimming method (Ref. 4)
5. Inadequate pile design for lateral load during
construction
6. Inadequate reinforcement in piles to resist soil heave

For Case III, model analysis also confirmed that without

temporary bracing, sheet pile would fail when soil berm is
removed for trimming pile adjacent to the wall.

97

443

In designing of bored piles, the above mentioned causes,

including deviation of piles due to excavation induced soil
movement are to be taken into account.

to New Delhi 1936-1994), Bangkok, Thailand, pp 5.31 5.42

4. Sliwinski Z. J. and Fleming G.K. [1983]. The Integrity
and Performance of Bored Piles Advances in Piling and
Ground Treatment for Foundations, London, UK, pp 153165.

ACKNOWLEDGEMENT
The authors express their appreciation to the colleagues,
especially to Mr. Ganeshan Baskaran, Mr. Muhammad
Ashfaq Anwar, Mr. Tossaporn Sree-iam and Mr. Chetsadar
Plongkrathok for their assistance in the preparation of this
paper.

5. Phienwej N and Kanok-nukulchai W. [1989].

Excavation Induced Pile Movement and Damage - A Case
History in Bangkok, Piletalk - International 1989, Kuala
Lumpur, Malaysia. pp 234-244.
6. Chiruppapa P. [1995]. A Case Study of Pile Failure
Created by Excessive Lateral Deflection of Braced Sheet
Pile Wall During Basement Excavation. Annual Technical
Conference 1995, Engineering Institute of Thailand. pp CE81 - CE94.

REFERENCES
1. Phienwej N., Balasubramaniam A.S., and Sutabutr T.
[1996]. Performance of Retaining Structures for Deep
Excavation in Bangkok Soft Clay Short Course on
Deformation of Soils & Rocks and Displacement of
Structures, Ground Movment in Deep Excavation - Case
Histories in Bangkok Subsoil, Bangkok, Thailand. pp.14-20.

7. PLAXIS Manual [1995]. PLAXIS - Finite Element

Code for Soil and Roack Analyses Version 6, Edited by
Vermeer P.A. and Brinkgreve R.B.J.
8. Poulos H.G. and Chen L.T. [1997]. Pile Response due
to Excavation - Induced Lateral Soil Movement, Deep
Foundations, Excavations, Ground Improvements &
Tunneling, 30th Year Anniversary Symposium of The
Southeast Asian Geotechnical Society, Bangkok, Thailand.
pp 5.24-5.47.

2. Phienwej N., Akewantop, K. and Balasubramaniam

A.S., [1996].
Comparative Evaluation of Ground
Movements Associated with Braced Excavation in
Bangkok Soft Soil Short Course on Deformation of Soils &
Rocks and Displacement of Structures, Ground Movement
in Deep Excavation - Case Histories in Bangkok Subsoil,
Bangkok, Thailand. pp.21-35.

9. Nilson, A.H. and Winter G. [1991]. Design of Concrete

Structures, eleventh edition, McGraw -Hill, Inc.

3. Teparaksa W. [1994]. Prediction of Movement and

Behavior of Sheet Pile Wall for Deep Braced Cut
Excavation in Soft Bangkok Clay.
Symposium on
Developments in Geotechnical Engineering (From Havard

98

444

Proceedings of the 7th International Conference on Application of

Stress-Wave Theory to Piles
9-11 August 2004, Kuala Lumpur, Malaysia

7th International Conference

on Application of StressWave Theory to Piles

Quantifying Pile Head Condition before

Basement Excavation by Cross-Hole Sonic
Logging Tests
Narong Thasnanipan, Zaw Zaw Aye, Thayanan Boonyarak
SEAFCO Public Co., Ltd. Bangkok, Thailand

Natamon Kampananon
STS Instruments Co., Ltd.

445

QUANTIFYING PILE HEAD CONDITION BEFORE BASEMENT

EXCAVATION BY CROSS-HOLE SONIC LOGGING TESTS
Narong Thasnanipan1, Zaw Zaw Aye2, Thayanan Boonyarak3
and Natamon Kampananon4
ABSTRACT
In wet process bored piles construction, it is necessary that concrete is continued to be placed until good
quality of concrete is formed above the design pile cut-off level to avoid contamination of slurry within
the design pile length. Controlling the excessive overcast length to maintain the appropriate pile top level
is one of the major problems particularly for the piles with low cut-off level. It is also impossible to check
the quality of concrete at the top part of the piles prior to exposing the pile head for pile cap construction
or basement excavation. If the pile head condition of recently cast bored piles could be justified in the
initial stage of the project, it would be better and easier for the piling engineers to efficiently decide the
optimum overcast length of the pile which in turn would provide effective saving of unnecessary concrete
consumption. Application of sonic logging tests in justifying overcast pile length is demonstrated in this
paper.
Keywords: cross-hole sonic logging test, wet-processed bored piles, cut-off level, contaminated
concrete, pile integrity

1. INTRODUCTION
The first sonic logging test in Thailand was believed to be conducted in 1982 for the wet-processed bored
piles of the Memorial Bridge Project where bad concrete zones were detected at two levels, depth about
20m and 1m from the pile top as reported by Ng (1983). In the past, sonic logging test was only employed
as a pre-planed site quality control testing for large infrastructure projects in Bangkok. The primary reason
of using sonic logging test in only large scale projects was due to its relatively expensive cost since cost of
testing the entire length of deep-seated bored piles are high higher costs involved in installation of
lengthy access tubes and the test itself. With the passage of time, the application of this quality control
test is being extended. As interpretation skills relevant to local soil condition and construction method of
this test have been significantly improved in local industry, the piling engineers begins to use sonic
logging tests for further assessment of constructed bored piles.
2. WET-PROCESSED BORED PILE CONSTRUCTION METHOD
Due to the prevailing subsoil and groundwater conditions, deep-seated bored piles of toe depth over 24m
are constructed by wet-process or slurry displacement method. In the rotary drilling method, a temporary
casing of appropriate length (12 to 18m is commonly used in Bangkok depending on the thickness of soft
clay) with required diameter (internal diameter not less than that of design bored pile diameter) is first
installed to ensure the stability of the borehole in the top soft or loose soil layers.
1

President, 2Senior Vice President, 3Geotechnical Engineer SEAFCO Public Co., Ltd. ,
Bangkok, Thailand, [email protected]
4
Senior Geotechnical Engineer, STS Instruments Co., Ltd., Bangkok, Thailand

447

Drilling is commenced by auger to drill out the soil inside the temporary casing. Auger-drilling is
commonly continued up to the top of the first water-bearing sand layer or bottom of the casing. Drilling
slurry or supporting fluid is then supplied to the borehole and drilling is proceeded with a bucket down to
the design final depth of the pile. Before lowering the reinforcement cage, a special cleaning bucket is
used to clean the base of borehole. If bentonite slurry is used, recycling method by air-lift or pump is
applied as the base cleaning process. Reinforcement cages are then lowered into the borehole and
concreting is carried out by tremie method.
3. CLASSIFICATION OF THE SONIC LOGGING TEST ANOMALY
Based on the experience of testing specialist on over 3000 deep-seated bored piles in past 15 years in
Thailand coupled with the comparison between sonic logging test results and direct examination from
cored samples of zones detected with anomaly, the category and criteria for interpretation of cross-hole
sonic logging test results have been established as shown in Table 1. Figure 1 depicted the sonic signal
profile of 4 categories classified in Table 1.
Table 1. Category of sonic logging signals and criteria
Category Criteria
A
No or slight increase in transmission time not more than 10%
B
Minor increase in transmission time >10% but <25%
C
Transmission time more than Category B but transmission
time can still be defined
D
Transmission time can not be defined

Integrity
Homogeneous concrete
Minor anomaly
Major anomaly
Defect

Figure 1. Sonic profile showing different category (A, B, C and D) as classified in Table 1
4. PILE HEAD CASTING LEVEL OF WET-PROCESSED BORED PILES IN BANGKOK
In order to ensure the sound concrete at the specified pile cut-off level, wet-process bored piles are
commonly cast well above the cut-off level. Though most of the piling engineers wish to minimize the
overcast length, their primary concern is a formation of contaminated concrete at the design pile cut-off
level which would need the piling contractor to come back to the site to carry out the remedial work while
the superstructure contractor has commenced his work. There could be a claim from the superstructure

448

contractor for having delay to rectify the imperfect pile heads. Hence, most of the bored piling engineers
prefer to cast the pile significantly higher than specified pile top level as shown in Figure 2.

Figure 2. View of exposed diameter 1.20m bored piles for mat foundation - Most of the piles were
constructed with excessive overcast length (over 4m above design pile cut-off level)
The need of using the long temporary casing in Bangkok subsoil also makes it difficulty to maintain the
top of overcast at the desired level. Table 2 shows the casting tolerance above cut-off levels according to
The Institution of Civil Engineers (1996) United Kingdom.
Table 2. Casting tolerance above cut-off levels for wet-processed bored piles (ICE, UK, 1996)
Cut-off level below
Casting tolerance above cut-off level, m Condition
commencing surface, H, m
0.15 10.00
1.0 + H/12 + C/8
Piles cast under water or
Where C = length of temporary casing
support fluid
below the commencing surface
5. COMMON DEPTH OF ANOMALY FORMED BY CONTAMINATED CONCRETE IN
BORED PILES
As construction method and quality control process for wet process bored piles are being well established
in local construction industry of Thailand, defect (discontinuity or contaminated concrete) in lower part of
pile shaft rarely occurs in these days. Hence investigating the top part of pile is more important than
testing the entire length of the deep-seated piles in Bangkok subsoil. Moreover, if sonic access tubes are
installed only in the top part of bored piles, the costs associated with both the material and test process
itself would be significantly reduced and integrity testing by sonic logging method would be more
feasible. According to the authors experience from over 5,000 exposed bored piles in Bangkok subsoil, to
achieve good pile head condition within design cut-off level, bored piles constructed under polymer-based
slurry require shorter overcast lengths than those constructed under pure bentonite slurry. In authors
opinion, the tolerable overcast lengths of bored piles constructed under polymer-based slurry should be
less than those specified in Table 2. Two cases were studied to determine the common depth of significant
anomaly (Category C or D) of bored piles in Bangkok subsoil.
5.1 Case 1
According to the sonic logging test results obtained from 400 piles of diameter 1.50m with depth over
50m constructed under polymer-based slurry in Bangkok, only 2 piles (0.5% of total contract piles) were
detected with C Category in the design pile length. The level of the detected C category for these 2 piles

449

was found to be within 2m below cut-off level of the pile. The overcast length above cut-off level of 95%
of piles were within 2.5m.
5.2 Case 2
The sonic logging test conducted on 174 bored piles of diameter 1.0m with depth over 50m constructed
under polymer-based slurry reveals that no pile was detected with C category below the cut-off level. The
overcast length of 76% of piles were within 2.2m above design cut-off level.
Table 3. Summary of sonic test results from 2 cases bored piles constructed under polymer-based slurry
in Bangkok
Case
Total no. Pile diameter Number of piles detected with C
Maximum overcast
category below design cut-off level length of majority of pile
of piles
/ depth
by sonic logging test
from sonic logging test
result
(pile)
(%)
Case 1
400
1.50m / 52m
2
0.5
2.5m (95%)
Case 2
174
1.0 m / 50m
0
0
2.2m (76%)
This finding suggests that contaminated concrete rarely occur below the cut-off level if overcast length of
the pile is about 2.5m for polymer-based bored piles. Quality control of the construction process such as
slurry mixing process and composition, properties of slurry, base cleaning, maintaining adequate tremie
length, concrete quality and concreting procedure however must be according to the standard practice to
achieve this result. Practical application of polymer-based slurry for deep-seated wet-processed bored
piles in multi-layered soil of Bangkok has been reported by Thasnanipan et. al. (2003).

Contaminated
concrete only
found at top 1m

Figure 3. View of polymer-based bored pile diameter 1.8m with overcast length about 2m contaminated
concrete was found only at the pile top
6. QUANTIFYING THE PILE HEAD CONDITION BY SONIC LOGGING TEST PRIOR TO
BASEMENT EXCAVATION
Designs of the modern buildings in Bangkok frequently call for the basement facility which require design
cut-off level of the piles well below the existing ground level. In wet process bored piles construction, it is
necessary that concrete is continued to be placed until good quality of concrete is formed above the design
pile cut-off level to avoid contamination of slurry within the design pile length. Controlling the excessive
overcast length to maintain the appropriate pile top level is one of the most difficult task particularly for
the piles with low cut-off level. Checking the condition of concrete at the top part of the piles prior to
exposing the pile head for pile cap construction or basement excavation by direct inspection is impossible.

450

If the pile head condition of recently cast bored piles could be justified in the initial stage of the project, it
would be better and easier for the piling engineers to efficiently decide the optimum overcast length of the
pile which in turn would provide effective saving of unnecessary concrete consumption. Cross-hole sonic
logging test as an early indicator may be the ideal method to check the optimum overcast length of
recently constructed bored piles in controlling the excessive overcast length.
Sonic logging test could also be used as a performance-indicator for the piling contractor. In one major
project, during the excavation works for 8 to 10m deep basement construction, the main contractor
claimed that over 50% of exposed bored piles from total 174 contract piles were found to be with
excessive overcast length (over 4 to 6m from design pile cut-off level). All bored piles in this project were
constructed under polymer-based slurry and sonic logging test was conducted on every pile. The piling
contractor carefully reviewed all the sonic logging test results and found the followings;
76% of piles were within 2.2m overcast length
18% of pile were with overcast length between 2.2m and 3m
Only 10 piles (5%) exceeded overcast length of 3m.
Over 4m to 6m overcast length claimed by the main contractor was caused by the solid-like mixture of
backfill material mixed with highly contaminated concrete at the top of the piles as boreholes were
backfilled with crushed-rocks. Figure 4 shows the typical sonic logging test results of one of the piles
claimed by the main contractor as excessively overcast piles. As can be seen in the figure, homogenous
concrete was detected only up to 2m above design cut-off level whereas non-homogeneous (Category C or
D) was observed above 2m. Field observation also confirmed that majority of piles were with highly
contaminated at top 2m above design cut-off level. The main contractor finally accepted the proof of sonic
logging test result to justify the overcast length.

Level of overcast pile

top claimed by main
contractor - 4.5m
Top level of homogeneous
concrete -6.8m
Design cut-off
level -8.9m
Figure 4. Typical sonic logging test profile of bored pile claimed as excessively overcast (overcast length
4.4m above cut-off level) by the main contractor
7. RECOMMENDED PROCEDURE OF QUANTIFYING THE PILE HEAD CONDITION BY
SONIC LOGGING TEST
As presented in earlier section, if the construction method strictly follows the quality assurance and
control procedure (QA & QC), contaminated concrete is rarely formed within the design bored pile length.
It should be sufficient to test the pile integrity of the top 10 to 15m by sonic logging test so that cost
associated with both access tubes and the test itself can be significantly reduced. It is also not necessary to
cast the concrete with excessive overcast length as most of the modern bored piles are constructed under
polymer-based slurry. Followings are the recommended process in minimizing the formation of
contaminated concrete within the design pile length for wet-processed bored piles in Bangkok subsoil.

451

Carefully check and use the required temporary casing length considering the thickness of soft
clay layer (common thickness of Bangkok soft clay layer is 15 to 18m)
Pre-install the sonic logging access tubes covering only the top 15m for majority of the piles.
Small percentage of piles may have full length of sonic access tubes to the pile tip to randomly
check the integrity of the entire pile shaft
Check the borehole profile of each pile by drilling monitor (koden equipment) to roughly assess
the required concrete volume (to check the size and shape of borehole by drilling monitor to
estimate required concrete volume)
Strictly follow and control the slurry mixing process, slurry properties and quality check test
Use appropriate concrete mix with good quality and strictly follow and control the tremie
concreting process
Carefully withdraw the temporary casing
Construct the pile of higher design cut-off level first (if possible) and perform the sonic logging
test 10 days after concreting
If possible (for high cut-off level piles or cut-off level within 3m below existing ground)
excavate with sufficient support to visually inspect the pile head and compare with sonic logging
test. This process may not be necessary though it will be very useful to assess the sonic logging
test result to correlate with actual condition of pile top at early stage of pile construction
The piling engineer must carefully review the borehole profile, actually poured concrete volume
and sonic logging test of each completed pile to assess and justify the construction method, site
characteristic and quality assurance procedure for further improvement

8. CONCLUSION
Cross-hole sonic logging test is a useful quality control test to examine the integrity of the wet-process
bored piles in early stage of the project. Discontinuity in the form of contaminated concrete in the lower
part of pile shaft rarely occurs in these days as construction method and quality control process for wet
process bored piles have been well established in local construction industry in Thailand,. Therefore, it is
recommended that sonic logging test should be mainly used to investigate the integrity of the pile at the
top portion and a random check is performed only for few piles to examine the integrity of entire pile
shaft. An extended application of cross-hole sonic logging test in quantifying pile head condition prior to
exposing the pile has been presented. The piling engineers should pay more attention to control the
excessive overcast length to minimize the cost associated with cutting extra length of pile head and
disposal of the waste. The piling engineers should be trained to effectively apply the sonic logging method
not only as integrity-proof test but also as a tool to quantify the pile head condition at the early stage of the
piling work.
REFERENCES
Institution of Civil Engineers, 1996. Specifications for piling and embedded retaining walls, The
Institution of Civil Engineers and Crown, Thomas Telford Services Ltd. London, United Kingdom
Ng, K. C. 1983. The construction problem and performance of large bored piles in second sand layer, M.
Eng. Thesis No. GT 82-26, Asian Institute of Technology, Bangkok, Thailand
STS Instruments Company Limited, 2004. Proposal for sonic logging test (unpublished), Method
Statement for Performing and Interpretation of cross-hole sonic logging test submitted to SEAFCO Public
Company Limited, Bangkok, Thailand
Narong Thasnanipan, Zaw Zaw Aye, Thayanan Boonyarak, 2003. Behaviour of Polymer-based Slurry for
Deep-seated Bored Piles in Multi-layered Soil of Bangkok, Proceedings of the Fourth International
Geotechnical Seminar on Deep Foundation on Bored and Auger Piles, Ghent, Belgium.

452

Proceedings of the 29th Annual Conference on Deep Foundations

29 September 1 October 2004, Vancouver, British Columbia,
Canada

Emerging Technologies

Advantages and Limitation of Integrity Tests

for Large Diameter Bored Piles in Bangkok
Subsoil
Narong Thasnanipan, Zaw Zaw Aye, Thayanan Boonyarak
SEAFCO Public Co., Ltd. Bangkok, Thailand

Natamon Kampananon
STS Instrument Co., Ltd.

453

ADVANTAGES AND LIMITATION OF INTEGRITY TESTS FOR LARGE DIAMETER

BORED PILES IN BANGKOK SUBSOIL
Narong Thasnanipan, Zaw Zaw Aye, Thayanan Boonyarak, Seafco Public Co., Ltd., Bangkok, Thailand
Natamon Kampananon, STS Instruments Co., Ltd., Bangkok, Thailand

In Bangkok due to the prevailing condition of subsoil (existence of considerably weak

stratum at top part) deep-seated large diameter bored piles of over 50 m depth with 0.8 1.8m in diameter have been mainly used for heavy structures such as high rise buildings,
elevated expressways, flyovers and more recently for underground subway stations of the
first Bangkok MRTA system. As a part of pre-planned quality assurance regime and
retrospective investigation, integrity of the piles are tested to obtain the information with
regard to the potential deficiencies of the constructed piles which may have formed during
pile construction process or may have been attributed by other activities after construction
of the piles. An overview of three commonly used non-destructive tests such as sonic
integrity, sonic logging and high strain dynamic load tests on deep-seated piles in Bangkok
subsoil is presented in this paper. Advantages and limitations of each integrity test are
discussed.
INTRODUCTION

EXISTING

10

Bangkok Soft
Clay

Medium to
Stiff Clay

30

Medium
Dense Sand
(First Sand)

40

Hard Clay

50

Dense to Very
Dense Sand
(Second Sand)

Depth (m)

200

400

600

Su= 10-25 KPa

t = 14-16kN/m 3
-

10

20

60

0
0

Weathered Crust

Su= 50-14020- KPa

t = 17-21kN/m 3
-

30
SPT-N= 20-40
t = 20kN/m 3

Su> 200 KPa

= 21KN/m 3
t

SPT-N= 50-100
t = 21KN/m 3

40

50

Line
atic
rost
Line
Hyd
wn
wdo
Dra
etric
zom

AND

Pie
ual
Act

Due to the prevailing condition of subsoil deepseated large diameter bored piles of over 50 m
depth with 0.8 - 1.8m in diameter are mainly
used for heavy structures such as high rise
buildings, elevated expressways, flyovers or
overpass bridges and more recently for
underground subway stations of the first
Bangkok MRTA system. As a part of preplanned quality assurance regime and
retrospective investigation, integrity of the piles
are tested to obtain the information with regard
to the potential deficiencies of the constructed
piles which may have formed during actual pile
construction process or may have been
attributed by other activities after construction of
the piles. Three techniques namely sonic
integrity, sonic logging and high strain dynamic
load tests are the commonly used nondestructive integrity tests applied in Bangkok,
Thailand. The effectiveness and suitability in
terms of both cost and method of application
itself are the primary reasons of the greatest
growth in use of these tests.
SUBSOIL
PROFILE

Pore Pressure (kN/m 2)

Geotechnical
Parameters

60

Figure 1. Typical soil profile of Bangkok with

piezometric drawdown condition (Thasnanipan,
et. al., 2002)

PIEZOMETRIC

Existence of soft to very soft highly compressible

dark gray marine clay makes difficulties in foundation construction works. This notorious soil
layer lies beneath made-ground in most urban
area and in some areas it lies under weathered
crust layers of 2m thick. Depending on the location, this layer is extended up to 12-18m. About
2m thick Medium Clay layer can be observed

A typical subsoil profile is relatively consistent in

different localities in Bangkok. It is characterized
by alternating layers of clay and sand deposits of
Quaternary age extending down to about 550m
depth where bedrock generally exists.

455

between Soft Clay and underlying Stiff Clay.

Generally Stiff Clay layer occurs directly underneath Medium Clay and its depth goes up to
22m. Below Stiff Clay layer, First Sand layer 58m in thickness can be found. This First Sand
layer, however, is absent in some areas. Stiff to
Hard Clay layer underlies First Sand and it is
found to be about 5m thick. Second Sand layer
generally occurs at depths between 45 to 65m.
Typical subsoil profile and the present piezometric drawdown condition of Bangkok up to 60m
depth are presented in Figure 1.

soil/slurry inclusions. The major causes of these

defects are;
Inadequate length to protect the soft clay
layer (length of temporary casing)
Delayed feeding of slurry and improper slurry
level maintenance for wet process
Long time duration of maintaining open
borehole
Non continuous concrete pouring or
disruption in concreting
Lifting up of insufficient workable concrete or
hardened concrete upon extracting the
temporary casing
Contamination of concrete with drilling slurry
or soil
Soft pile toe due to improper base cleaning

CONSTRUCTION METHOD
Due to the prevailing subsoil and groundwater
conditions, deep-seated bored piles of toe depth
over 24m are constructed by wet-process or
slurry displacement method using rotary drilling
machine. A temporary casing of appropriate
length (12 to 18m in Bangkok depending on the
thickness of soft clay) with required diameter (internal diameter not less than that of design
bored pile diameter) is first installed to ensure
the stability of the borehole in the top soft or
loose soil layers. In some projects where the vibration is strictly limited, a standard length of
casing with oscillator or short temporary casing
of 5 to 8 m is pushed down in combination with
the pre-boring process. Drilling is commenced by
auger to drill out the soil inside the temporary
casing. Auger-drilling is commonly continued up
to the top of the first water-bearing sand layer or
bottom of the casing when using the short casing
method. Drilling slurry or supporting fluid is then
supplied to the borehole and drilling is proceeded with a bucket down to the design final
depth of the pile. Before lowering the reinforcement cage, a special cleaning bucket is used to
clean the base of borehole. If bentonite slurry is
used, recycling method by air-lift or pump is applied as the base cleaning process. Reinforcement cages are then lowered into the borehole
and concreting is carried out by tremie method.

Most of the defects or cracks in bored piles

caused at post construction stage are usually
induced by construction activities associated
with adjacent excavation work with inadequate
earth retaining system and improper trimming of
pile head to design cut-off level.
Thasnanipan et al. (1998b) reported that integrity
of 285 bored piles (3.3 % of 8,689) were found to
be of doubtful quality according to an
assessment on the results of sonic integrity test
with the additional information obtained from the
construction records of bored piles in Bangkok
subsoil.
The results of the findings are
summarized in Table 1. As can be seen in the
table, higher percentage of defects is caused by
post-pile-construction activities and more
commonly by movement of surrounding soils
induced by excavation works in the vicinity of the
piles.
Table 1. Summary of pile assessment on 8,689
bored piles (Thasnanipan et al. 1998b)

OCCURENCE OF POTENTIAL DEFECTS IN

DEEP-SEATED WET-PROCESSED BORED
PILE

Type of Defect

Causes

Poor concrete
at pile top (03m)

Cutoff level near the

ground or inadequate
overcast length
Insufficient casing length
or soft clay layer variation in thickness

0.1

Size reduction

In general, pile defects can be caused at two

stages, during the pile construction process and
after construction of the piles (post-pileconstruction). Pile defects that may arise during
the construction of wet-processed bored pile are
size reduction/necking, discontinuity, and

Cracks/discont
inuities

456

Excavation works for

pile trimming and construction activities

1.0

2.2

SELECTION OF SUITABLE TEST METHOD

defect detected by other integrity tests. Table 2

shows the suitability of test method for different
type of common defects.

Turner (1997) highlighted the factors in selection

of pile testing in CIRIA Report 144 as follows;
The perceived nature of possible features or
defects within the pile
The ability of the test method to detect the
feature or defect under investigation
The cost of testing and examination
The ease of use and interpretation

SONIC INTEGRITY TEST

Sonic integrity test, also known seismic test is
the most common method of integrity testing for
both driven and bored piles in Bangkok. In many
projects it is a part of the contractual requirement
to conduct sonic integrity test. Minimum 10 % to
maximum 100 % of production piles are
commonly tested.
It is also a reasonably
acceptable method for applying as a
retrospective investigation in determining
integrity of the pile.

Like other indirect measurement of pile integrity

testing methods, sonic integrity and sonic
logging tests are only capable of identifying
structurally significant features. Rejection or
acceptance of individual piles should not rely
only on the results of these tests. Further
investigation, engineering evaluation and
judgment are highly recommended to confirm
the defect detected by these indirect techniques.
Both methods however are very useful for costeffective screening test to identify piles with
potential defect.

Features Detectable by Sonic Integrity Test

for Deep-seated Bored Piles
Following negative features of deep-seated
bored piles can be commonly detected by sonic
integrity tests.
Discontinuity / crack
Size reduction or necking
Discontinuity / Contaminated concrete
The characteristic of sonic integrity signals of
above features have been thoroughly reviewed
and presented by Thasnanipan et al. (1998a).
As some features have a similar pattern to each
other, it is always important to check and
compare the test data of pile in question with
those of other piles within the same job site to
establish site signature.
In some cases
correlations should be made with similar projects
where damaged piles were investigated with
firmed evidence (coring or excavation to detect
damage level). Interpretation should also be
made with sound knowledge of pile construction
technique, subsoil condition and other factors
(problem encountered during and post pile
construction) which may influence the sonic test
signals. Pile construction records are also useful
in interpretation of piles with detected anomaly.
Conclusion should be made after careful and
thorough review of test results in conjunction
with other available information.
Further
investigation (if applicable excavate down to
level of detected anomaly for visual inspection)
may require before final conclusion is made.

High strain dynamic load test is usually selected

to verify the load carrying capacity of piles. Pile
integrity can also be determined by high strain
dynamic load test. The major advantages of this
test in comparison with static pile load test are
those of cost, time and space requirement.
Table 2. The suitability of test method for
different type of defects.
Type of
Test method and suitability
defects
for defect detection
Sonic
Sonic
High strain
Integrity
logging
dynamic
load test
Crack
Suitable
Possible Suitable
Pile size
Suitable
Possible Suitable (if
change
severe defect)
Possible
Suitable Suitable (if
Poor
severe dequality
fect)
concrete
or
Contaminated
concrete
Soft pile Not
Suitable
toe
usually

Discontinuity / Crack Detectable by Sonic

Integrity Test

In some projects, dynamic load test was applied

to justify the capacity of pile after rectifying its

Modern buildings in Bangkok commonly require

the excavation works for basement facility and

457

for the use of different elevations of pile

foundation. Excavation works are mainly carried
out by using cut slope, sheet pile wall,
diaphragm wall and secant pile wall. Since most
of the bored piles are designed mainly to carry
an axial load, lateral and tensile force which may
impose on pile during basement excavation are
sometimes not considered or overlooked.

Med.
Clay

Soft Clay

Very Stiff Clay

10.5cm/s
0

Pile 30
2 Jul 97

12

15

18

21

24

27

4000 m/s
exp: 20

30

V2-88c
Sr

f:8

TYPICAL SOIL PROFILE

Figure 3 Pile size reduction or necking detected

by sonic signal

SOFT CLAY

MEDIUM CLAY
STIFF CLAY

SAND

crack induced
by large lateral
movement

crack induced
by sliding
failure

crack induced
by uplift
force

Figure 3 shows reduction of pile size at about

14m. According to the construction record the
cause of the necking was believed to be caused
by use of inadequate 12m long temporary casing
at the particular subsoil condition where the soft
to medium clay layer extends up to 16m.
Contamination of concrete or Discontinuity
of Pile Material
A discontinuity in pile shaft caused by disruption
during concrete pouring can be seen in Figure 4.
Concrete supply was interrupted by poor
coordination between the piling contractor and
concrete suppliers.
As previously poured
concrete in the lower pile section had been
hardened when concreting resumed, the tremie
pipe was also blocked. Blockage was cleared
after completely pulling out the entire tremie from
borehole. Reinstallation of tremie pipe in the
borehole filled with bentonite slurry eventually
caused the formation of contaminated layer
between old and new concrete.

Figure 2. Typical features of external forces

causing pile cracks in Bangkok subsoil
Excessive lateral displacement of ground caused
by improper excavation naturally induced an
additional lateral force and bending moment in
the pile. Once the induced bending moment
exceeds the cracking moment capacity, piles are
subject to be cracked. Unexpected tensile force
acting upon the pile due to an excessive heave
or uplift force induced by excavation can also
cause cracking of pile. In Bangkok, pile cracks
caused by external forces are normally found at
the level where Soft and Medium to Stiff Clay
boundary is present (typically at 12 to 16m below
the existing ground level) as illustrated in
Figure 2.

14.2 cm/s

Size Reduction or Necking

Pile 1
24 Apr 97

Common causes for size reduction in bored piles

are;

12

16

20
24
4000 m/s
exp: 20

28

32

36
40
f:8

V2-88c
Sr

Figure 4. Sonic signal showing a discontinuity at

11.0m
caused by interruption in concrete
pouring - multiple reflections of signal can be
observed from the discontinuity

Insufficient length of temporary casing to

protect the soft clay
Delayed feeding of slurry and improper slurry
level maintenance for wet process
Long time duration of maintaining open bore
Non continuous concrete pouring
Soil conditions such as inclusion of sand
lenses in clay

The detected discontinuity was later rectified by

coring and cement grouting. Figures 5 shows
the sonic test signal of the same pile after
closing the discontinuity.

458

7.6 cm/s
0

12

16

20

Pile 1 #

4000 m/s

2 Jun 97

exp: 20

24

28

32

36

40

V2-88c

f:8

Sr

Based on the experience of testing specialist on

over 3000 deep-seated bored piles in past 15
years in Thailand along with the comparison
between sonic logging test results and direct
examination from cored samples of zones
detected with anomaly, the category and criteria
for interpretation of cross-hole sonic logging test
results have been established as shown in the
table below.
Table 5 Category of sonic logging
criteria
Category Criteria
A
No or slight increase in
transmission time not
more than 10%

Figure 5.
Sonic integrity test shows the
discontinuity at 11.0m has been closed by
grouting
CROSS-HOLE SONIC LOGGING TEST
Cross-hole sonic logging test can mainly be applied as a pre-planed site quality control testing.
Use of Sonic logging test for checking pile integrity has increased in Thailand in recent years. It
is used for cast-in-place foundation elements
such as bored piles, barrettes, diaphragm walls
and caissons. Both PVC and steel tubes can be
used as access tubes. Though cost of steel tube
is higher than that of PVC, steel tubes are considered more suitable than PVC tubes for two
reasons; (1) less potential for bending and damage especially for deep-seated bored piles and
barrettes with heavy steel reinforcement and (2)
better bonding with concrete. These access
tubes are sometimes also used for pile base
grouting. The access tubes are attached to the
reinforcement cage, which is installed prior to
concreting as a normal practice for cast-in-place
piling. Two access tubes are required as a minimum for cross-hole sonic logging test. For a
good coverage of the test pile, recommended
number of tubes for different sizes of bored piles
is shown in Table 4.

results and
Integrity
Homogeneous
concrete

Minor
increase
in
transmission
time
>10% but <25%

Minor anomaly

Transmission
time
more than Category B
but transmission time
can still be defined

Major anomaly

Transmission time can

not be defined

Defect

Table 4. Recommended number of tubes for

different pile size
Tube spacing
Pile Diameter
Minimum No.
(mm)
of tubes
(degree)
D 750

180

750<D 1000

120

1000<D 1500

90

1500< D
2500

60

2500 < D

45

Figure 6. Sonic profile showing A, B, C and D

categories

459

Turner (1997) pointed out the need of both

theoretical knowledge and practical experience
in interpretation of the test results in CIRIA
Report 144. Suggestion is also made in this
report that the anomaly shown in sonic test
signals can be caused not only by changing in
physical properties of the materials, but also by
factors within the measuring system itself.
These factors are;

Settlem ent (m m )

-5

-10

-15

Free movement of the probes within access

tubes
Mismatched probe positions especially at pile
toe
Measurement resolution
Incorrect position of access tube
Air gaps or different material around access
tubes
Aggregate variation (in the case of base
grouting)

Dynamic load test

Static load test
-20
0

20

40
60
Load (ton)

80

100

Figure 7. Load-settlement curves of dynamic and

static load tests on large diameter piles ( 0.80m
x 50.0m).
0

Settlem ent (m m )

It is to be noted from the authors suggestion

based on the above factors that only variations
of transmit time more than 15 to 20% of the
norm should be regarded as warranting further
investigation.

-5

-10

HIGH STRAIN DYNAMIC LOAD TEST

-15
Dynamic load test
Static load test

High strain dynamic load test can be employed

as a part of pre-planned quality assurance
regime and retrospective investigation. High
strain dynamic integrity test has become a wellaccepted method especially for evaluating the
pile capacity in the foundation industry today. A
large number of related technical papers and
case histories of the test have been published
and it is a part of standards and specifications
such as ASTM D4945-89 (Standard Method for
High-strain Dynamic Testing of Piles).

-20
0

200

400
Load (ton)

600

800

Figure 8. Load-settlement curves of dynamic and

static load tests on small diameter bored piles (
0.35m x 18.9m).
Pile integrity can also be determined by high
strain dynamic load test. Figure 9 and 10
illustrate the discontinuity (crack) detected at the
same pile by sonic integrity and high strain
dynamic load tests respectively.

Vasinvarthana
and
Kampananon
(1997)
presented the efficiency and reliability of
dynamic load test carried out in Bangkok.
Figures 7 and 8 show the load settlement curves
of dynamic and static load tests conducted in
Bangkok on large and small diameter piles
respectively. As can be seen in the figures,
load-settlement characteristics obtained from
dynamic load test agree well with those of static
load test for both large and small diameter bored
piles.

It is to be noted from Figure 9 and 10 that the

duration of pulse in the sonic integrity test signal
is about half of that in dynamic test signal. The
short duration (high frequency) pulses can
resolve reflections from narrow impedance
changes. In such case presence of a crack in
the pile is more clearly indicated by low strain
with sonic integrity test signal than the high
strain by dynamic test signal.

460

Soft Clay Med. Very Stiff Dense Sand

Clay Clay

Some defect features can be of similar signal

pattern and it can not detect the axial extent
of anomaly. Therefore crack may be
misinterpreted as contaminated concrete or
void which can lead to an argument between
piling contractor and excavation contractor

Dense Sand

6.2 cm/s
0

12

18

24

Pile 681
8 Apr 96

30

36

42

4000 m/s
exp: 20

48

54

60

V2-88c

f:8

Sr

Figure 9. Sonic signal showing a discontinuity

(crack) at about 7.8m, caused by excessive
lateral soil movement induced by adjacent
excavation work
29-Apr-96
CAPWAP (R) Version 1 994 -1
For Msd
Vel Msd

1500
Ton
750
10

20

30
1

40
2

50
3

ms
L/c

-750

30
Ton / m

Sonic logging tests are much more expensive

than sonic integrity tests as the cost of access
tubes and test itself are comparatively high for
deep-seated bored pile. It is mainly applicable as
a pre-planed site quality control testing. The
major advantage of this method is that test can
be carried out shortly after the pile construction.
Hence,
rectification
measures
can
be
implemented while the foundation contractor is
on site. However, this method is not applicable
for retrospective investigation. It is less useful if
pile integrity is in question due to post
construction activities as small horizontal cracks
are hardly detectable by this test method.
The major advantage of high strain dynamic load
test is its capability to verify both the load
carrying capacity and integrity of the pile. The
main advantages of this test in comparison with
static pile load test are effectiveness in terms of
cost, time and space requirement. It can be also
applied to justify the capacity of pile after
rectifying its defect detected by other integrity
tests. However, testing and interpretation of the
results must be made by qualified testing
specialists who have adequate knowledge in
local soil conditions and construction process of
wet-processed bored piles.

Pile
Shaft Resistance
Distribution
(Toe 178.8 Ton)
Pile Forces at Rut

1000
Ton

Figure 10. High strain dynamic load test signal

showing a crack in the pile at about 8.0m
DISCUSSION
Sonic integrity test can be applied for both
quality check (control test) and retrospective
investigation. It is the cheapest in terms of cost
and the simplest in terms of testing process. In
Bangkok a single testing crew can carried out
the test over 70 piles within a day at one site if
all piles tested are easily accessible and readily
prepared. The main advantage of this test is that
since no particular measures or preparation is
necessary during the pile construction phase it is
more flexible to select which pile is to be tested.
However, interpretation of sonic integrity testing
needs considerable experience and knowledge
in testing, subsoil condition and construction
method. The major limitations of sonic integrity
test are;
Signal characteristics is influenced by
damping effect due to the stiffer soil layers at
deeper depth. Hence toe reflection is hardly
detectable for piles founded at depth deeper
than 35m. Similarly signal reflections from
anomalies are generally clear for the piles
with length to diameter ratio (L/D) within 40

CONCLUSIONS
Integrity tests commonly applied for large
diameter bored piles in Bangkok are presented
highlighting the advantages and limitations of
each method. It is to be emphasized that
integrity test are highly dependent on the
experience of the testing firm in both field testing
and interpretation.
Interpretation should be
made with sound knowledge of pile construction
technique, subsoil condition and other factors
such as problem encountered during and post
pile construction which may influence the test
results and interpretation.
Pile construction
records are also useful in interpretation of the
piles with detected anomaly. Conclusion should
be made after careful and thorough review of
test results in conjunction with other available

461

information. Further investigation may require

before final conclusion is made.
REFERENCES
American Society for Testing and Materials
1989. ASTM D4945-89 Standard method for
high-strain dynamic testing of piles, Philadelphia:
American Society for Testing and Materials.
Thasnanipan, T., Maung, A. W. and Tanseng, P.
1998a. Dam-ages to Piles associated with
Excavation Works in Bangkok Soft Clay, 6th
International Conference on Problems of Pile
Foundations Building, Ufa, Russia. September
14-18, 1998. pp. 91-98
Thasnanipan, N., Maung, A. W. & Baskaran, G.
1998b. Sonic integrity test on pile founded in
Bangkok subsoil signal characteristics and
their interpretations, Proc. 4th Int. Conf. on Case
Histories in Geotechnical Engineering, St. Louis,
Missouri, 9-12 March 1998: 1086-1092.
Turner, M. J. 1997. Integrity testing in piling
practice,
CIRIA
Report
144,
London:
Construction Industry Research and Information
Association.
Vasinvarthana, V. & Kampananon, N. 1997.
Efficiency and re-liability of dynamic load test),
Seminar on Foundation 97, Organized by
Engineering Institute of Thailand, Bangkok: 7994. (in Thai).

462

Proceeding of the 8th International Conference on the Application of

Stress Wave Theory to Piles
8-10 September 2008, Lisbon, Portugal

8th International Conference

on the Application of Stress
Wave Theory to Piles
Edited by:
Jaime Alberto Dos Santos

Application of Low-strain Integrity Tests for

Quality Assurance of Deep Foundations in
Thailand
Aung Win Maung, Zaw Zaw Aye and Thayanan Boonyarak
SEAFCO Public Co., Ltd. Bangkok, Thailand

463

Application of Low-strain Integrity Tests for Quality Assurance of Deep

Foundations in Thailand
Maung A. Win, Aye Z. Z. and Thayanan Boonyarak
SEAFCO Public Company Limited, Bangkok, Thailand.

Keywords: Sonic integrity test, Sonic logging, Parallel seismic test

ABSTRACT: Low-strain integrity tests have been widely used in Thailand since the 1980s, when largediameter bored piles were introduced for the earliest high rise buildings in Bangkok. The most commonly
used integrity test methods on deep foundations in Thailand are sonic integrity tests (locally known as seismic
tests) and sonic logging tests. The sonic integrity test is specified as a compulsory quality assurance test for
bored piles in most projects. Additionally it is also carried out on some reference piles soon after piling work
is completed, particularly in projects which require deep excavation for further construction stages by other
than the piling contractor. This practice is adopted to prevent contractual dispute between the piling
contractor and the other construction contractor regarding pile damage contributed by soil movement
associated with deep excavation work. Sonic logging tests are commonly specified for all large bored pile
with diameters up to 2.00m and barrette piles for bridges and elevated highways and viaduct foundations. In
addition parallel seismic tests are now often utilized to determine pile length of old bridges and flyover
foundations to check alignment clearance for tunnels designed along the existing roads and canals. This
paper presents a brief history of low-strain integrity tests used in Thailand and discusses interpretation of test
signals in conjunction with pile construction method and subsoil conditions.
conditions with relatively shallow bedrock is
present. In order to assess the quality of foundation
piles, a number of low-strain integrity test methods
1 INTRODUCTION
have been applied according to pile type. A
monograph of the low-strain integrity testing
Thailand has been one of the economically fast
described in this paper has been published by Turner
growing nations in South East Asia since the 1980s,
M. J. (1997).
with the economic growth fuelling a boom in civil
construction. However, the majority of large scale
construction work has been mainly in Bangkok, the
2 FOUNDATION PILES
capital of Thailand, where over 10% of the total
population resides.
Bangkok is located in a
In Thailand, two major types of deep foundation
sediment basin comprising thick layers of soft clays
piling are pre-cast driven piles and bored piles. Preand a series of inter-bedding sand and clay layers in
cast piles are generally used outside urban or less
hundreds of meters of deep. The depth of bedrock is
built-up areas where there are no noise or vibration
not well determined in the Bangkok area but has
constraints for pile driving. In Bangkok, bored piles
been reported to be about 550m deep
(Balasubramaniam, 1991).
Deep foundations
with tips extending to the first sand layer or to stiff
consisting of large diameter bored piles and barrettes
clay layer at a depth of about 20 -30 m are normally
reaching a depth of 70m in such subsoil support
employed for light to medium-sized structures
heavy structures are therefore concentrated in
located in vibration-restricted areas. For heavy
Bangkok.
structures or high-rise buildings deep large diameter
With improvements in the network of highways
(0.80m-2.00m) bored piles with tips extending to
connecting all provinces over the last 20 years, civil
more than 45m are commonly used. Two types of
construction is now active in all major provincial
construction method, dry and wet process are
cities and tourist destinations. In these areas,
typically applied in bored piling.
foundation constructions are generally much
The dry process uses steel temporary casing
shallower than those used in Bangkok as better soil
(about 15m in length, especially in Bangkok) to

465

protect the soft soil from caving in. The bore is

drilled either using an auger or bucket. For small
piles, shells with drop-weight are used to make the
bore. Concrete is poured into a borehole which is in
dry condition after placing the reinforcement cage.
The wet process also uses steel temporary casing.
A borehole is made with rotary drilling using auger
or an auger bucket under bentonite or polymer
slurry. Concrete is poured with tremie method under
drilling slurry.
The dry process is employed mainly for small
piles of 0.35m to 0.60m in diameter. The length of
the piles are up to 25m and extends further if
necessary, depending on subsoil conditions, absence
of sand layer with groundwater, etc.
For large bored piles of 0.80m to 2.00m in
diameter and barrettes (up to cross sections of
1.5x3.0m), piles are constructed to depths of 30m to
over 60m. The wet process is commonly used and
piles are founded in sand layers. Compressive
cylindrical strengths of concrete used in bored piles
is in a range of 240-280 ksc (24-28 Mpa.).
3 SONIC INTEGRITY TESTS
3.1 Overview of sonic integrity testing in Thailand
The sonic integrity/echo test (locally known as
seismic test), performed with an accelerometer and
hammer, has been used in Thailand since the 1980s
when large-diameter bored piles were introduced for
the earliest high-rise buildings in Bangkok. In
addition to construction records and specification for
the bored pile construction for quality assurance, a
sonic integrity test was also employed. In the early
days of high-rise construction, the sonic integrity
test equipment showed a test record on the display
unit only. The record was photographed by a
Polaroid camera for later interpretation. The quality
of test records was visually poor and sometimes
comprehensive conclusions on test results were
difficult to make. A few years later, advancements
in both computer hardware and software enabled the
user to obtain high resolution data with signal
enhancement
and
digital
processing
for
interpretation of test data. Test results can thus be
accurately interpreted, particularly for types of
defect or anomalies and their position, and pile
length.
Other derivatives of low-strain integrity tests, such
as frequency response tests were introduced by a
few testing companies. However, the frequencybased testing was not as popular as the time-based
testing due to unfamiliarity with both analysis and
interpretation of the test results.

Figure 1. A photographic record of sonic integrity test result

taken by Polaroid camera in 1986.

In the period 1992-1997, with economic growth in

Asia, massive development in construction of multistorey buildings, elevated highways, bridges, etc.
took place in Bangkok. Large quantities of bored
piles and driven piles were installed annually. The
sonic integrity test was then widely used for quality
assurance of these piles.
After 1997, the construction industry in the
private sector was significantly affected by an
economic crisis. However infrastructure work such
as mass transit system and airport construction
projects was still active.
The sonic integrity test is used mainly to detect
defects, changes in the physical properties and
length of the bored piles. For driven piles, it is used
for detecting pile defects such as cracks and joint
splitting induced by over driving and/or pile
inclination.
3.2 Signal characteristics and interpretation
The sonic integrity test indicates toe reflections of
pile clearly for small piles with slenderness ratio of
40 . In few cases, especially toe grouted piles toe reflection has been observed for longer piles with toe
embedded below 30-35m.
In some cases, even though pile toe is extended
beyond 35m in depth, the toe reflection can be observed in the test signals of toe grouted piles having
L/D within 40 .
Figure 2 is an example of such a case where the
bored piles were base grouted and of 40m in length
(toe at 55m, trim level at -15m). However, it should
be noted that these reflections are prominent only for
a few piles. Influence on toe reflection by grouted
toe also depends on stiffness of the soil underneath
the toe, which has been grouted.

466

23.2m

Soft Clay Very Stiff Clay Hard Clay Dense Sand

11.0cm/s
0

5.3cm/s
0

10

15

20

Pile G30
27 Jan 97

25
30
4000 m/s
exp: 20

35

40

45
f:8

50

12

15

18

Pile 6

4000m/s

12 Sep 95

exp: 20

21

24

27

V2-88c

30

f:8

Sr

V2-88c
Sr

22.0m

Figure 2. Toe reflection observed from a toe grouted pile. (pile

1.0x40.0m, cutoff -15.0m, toe depth 55.0m)

7.1cm/s
0

In a few cases, it was found that toe reflection is

detectable for some deep-seated bored piles of about
50m in length if the pile was tested at an early age.
Figure 3 shows the significant toe reflection of
52.0m-long pile tested a few days after concreting.
Note young pile age, the bonding between soil and
pile has not yet fully developed so that damping
effect or attenuation of the sonic signal by shaft
friction is relatively low.

12

15

18

Pile 24

4000m/s

12 Sep 95

exp: 20

21

24

27

V2-88c

30

f:8

Sr

Figure 5. Toe reflections from different pile sizes, pile no. 6

(0.50mx22.0m) and pile no. 24 (0.35mx22.0m)

Typically, sonic integrity tests show size variation

in bored piles constructed in Bangkok. All size
variations including size reductions are found to be
at the transition interface of temporary casing and
soil layer, where soil strata changes also occur.

52.0m

16.0cm/
s
0

12

18

24

pile 30
14 Jan 91

30

36

42

48

54
f:8

4000m/s
exp: 20

60

Influence by different soil strata on signals has

been observed in many cases (Thasnanipan et al
1998). Stiff clay layer at depths of 15m-25m
(undrained shear strength of 10-22 t/m2)
significantly influences signal reflections. This was
in small rather than large bored piles. Figures 4 and
5 explain these conclusions.

Med. Very Stiff Dense

Clay Clay Sand

10.6cm/s
0
Pile 49
19 Sep 95

12

15
18
4000 m/s
exp: 20

21

24

27
f:8

30

Very Stiff Clay

Sr

Figure 3. Sonic signal showing a toe reflection of bored pile at

52.0m tested a few days after concreting.

Soft Clay

Med.
Clay

Soft Clay

v2-88c

V2-88c
Sr

Figure 4. Influence of soil friction on sonic signals (a pre-cast

driven pile 0.40x0.40x25.0m in clay layers)

467

10.5cm/s
0
Pile 30
2 Jul 97

12

15

4000 m/s
exp: 20

18

21

24

27
f:8

30

V2-88c
Sr

Figure 6. Pile size decrease or change at 14m as shown by the

signal (Pile 0.8mx21m)

Test signals shown in Figure 7 indicate a clear

reflection at approximately 11m from pile top (about
15m below ground level) with a repeat or multiple
reflections at about 2 times that distance
(approximately 22m from pile top). As the polarity
of the reflection is negative, the first interpretation is
that this pile has decreased in cross section or crack
at 15m below ground level. However, based on the
pile construction method, it is concluded that
negative reflection is caused by variation in pile
cross section at the bottom end of the 15m-long
temporary casing. The sonic signals with anomaly
acquired on those piles are often misinterpreted as
indicative of defective piles, leading to dispute.

Med.
clay Stiff clay

Very soft clay

Dense sand

Temporary casing
bottom level

Pile Cut-off level

4.3cm/s
0

Pile 66
25 Jun 97

12

10

15

20

25

4000 m/s

25 Nov 94

exp: 20

30

35

40

45

50

V2-88c

f:8

Sr

Nominal dia./dia. of
drilling tool

Outer diameter of
casing collar

11.0 cm/s

Pile 99

15
18
4000 m/s
exp : 20

21

24

27
f: 8

30 v2-88c
sr

Figure 7. A typical pile shape constructed in Bangkok soil

with resulting sonic integrity test signal profile.

Soft Clay

Figure 8 shows bored pile sections formed at the

lower end of temporary casing, confirming size
changes occur below the casing. These bored piles
were cast together with a steel column for
construction of deep basements with diaphragm
walls using top-down method. These piles were
exposed when excavation reached to 15m below
ground level.

Figure 8. The photograph illustrates pile size changes caused

by difference in diameter of temporary casing and auger used
in pile construction.

Size variations shown by signals are often

consistent with casting records and available drilling
monitoring results. Figures 9A and 9B show the
comparison of borehole profiles and sonic test
records.

Med. Stiff to Hard Dense Stiff

Clay
Clay
Sand Clay

Dense Sand

Figure 9B. Uniform pile size along the pile shaft is shown by
both drilling monitoring record and sonic integrity test record.
(Pile 0.80mx42.0m, cutoff -3.0m, toe depth 45.0m.)

Sonic integrity test can clearly reveal a crack or

discontinuity in piles. Most cracks or discontinuities
in bored piles after installation can be caused by
construction activities associated with basement
excavation adjacent to them and improper trimming
to design cut-off level. In the past, from the
assessment of pile defects, most of the defects are
found in small piles with a diameter of 0.60m or
less. Study of piling and excavation plans show that
piles with a crack indicated by the sonic integrity
were located in the vicinity of excavation boundaries
(Thasnanipan et al, 2000a). These piles were
cracked by soil movement due to both reinforcement
inadequate for resisting bending stresses and
improper soil excavation practice.
Since
geotechnical experience has been gained from
several construction projects with deep basements,
pile design has been improved over the years and
pile defects are significantly reduced.
The test signal in Figure 10 shows a crack or
discontinuity in the pile caused by soil movements
due to adjacent excavation work.

Soft Clay Med. Very Stiff Dense Sand

Clay Clay

Dense Sand

5.4cm/s
0

10

Pile 322
10 Jan 95

15

20

25

30

35

4000 m/s
exp: 20

40

45
f : 8

50

V2-88c
Sr

6.2 cm/s
0
Pile 681
8 Apr 96

12

18

24

30
4000 m/s
exp: 20

36

42

48

54
f:8

60

V2-88c
Sr

Figure 10. Sonic signal showing a discontinuity (crack) at

about 7.8m, caused by lateral soil movements.

Soft Clay

Med.
Clay

Stiff to
Hard Clay

Dense Stiff
Sand Clay

Dense Sand

Figure 9A. Size increase in pile shaft at 13m is shown by both

sonic integrity test and drilling monitoring records. (Pile
1.0mx40m, cutoff -5.0m, toe depth 45.0m.)

Figure 11A shows a discontinuity in the pile shaft

caused by disruption during concrete pouring.
Concrete supply was interrupted by poor
coordination between the piling contractor and

468

concrete suppliers. By the time the concreting was

resumed the lower section had hardened. The
discontinuity was later closed by core-drilling and
cement grouting. Figures 11A and 11B show the
signal interpretation before and after closing the
discontinuity.

14.2 cm/s
Pile 1

12

16

24 Apr 97

20
24
4000 m/s
exp: 20

28

32

36
40
f:8

V2-88c
Sr

Figure 11A. Sonic signal showing a discontinuity at 11.0m

caused by interruption in concrete pouring
(pile
0.80mx38.80m) and multiple reflections from the
discontinuity.

7.6 cm/s
0

12

16

20

Pile 1 #

4000 m/s

2 Jun 97

exp: 20

24

28

32

36

40

f:8

V2-88c
Sr

Figure 11B. Sonic integrity test shows the discontinuity at

11.0m has been closed by grouting.

Figure 12 shows the characteristics of signals

observed at a site where dry process was adopted.
Sonic integrity test signals were considered
indicative of poor pile integrity at the site.
Construction records revealed that boring was done
in water bearing sand layer without casing or
bentonite slurry causing concrete contamination with
water and/or segregation.

Med. Stiff Clay Dense

Sand
Clay

Very Stiff Clay

8.9 cm/s
0

12

15

Pile 50

4000 m/s

4 Apr 97

exp: 20

18

21

24

27
f:8

30

V2-88c
Sr

Figure 12. Sonic test signal showing size decrease and or a

discontinuity at 7.5-9.0m and poor quality pile section near the
toe (Pile 0.35x26.0m).

3.3 Current applications

The sonic integrity test has been accepted by
engineers and has been widely used in the deep
foundation works in Thailand for over 25 years. The
test is also applied to testing of barrette piles in
conjunction with sonic logging. For barrettes,
multiple testing points with signal stacking are used
for obtaining the test data.

469

Due to cost effectiveness, simplicity of the

method and ease of performing the test, the sonic
integrity test is now specified for testing 100% of
piles in most of the projects for quality assurance.
Additionally, a sonic integrity testing is also
carried out on some piles located in the boundary of
deep excavation zones soon after piling work is
completed. The test records use a reference pile to
document pile integrity prior to deep excavation
work to prevent dispute between piling contractor
and excavation contractor for any pile damage
caused by improper construction practice.
4 CROSSHOLE SONIC LOGGING
4.1 Overview of Crosshole sonic logging test in
Thailand.
The first crosshole sonic logging test is believed to
have been conducted in 1982 for quality insurance
of bored piles based on the test results for the
Memorial Bridge Project in Bangkok as reported by
Ng (1983). Usually the test is employed as quality
control testing for foundation piles and diaphragm
walls of infrastructure projects such as subways,
elevated highways and bridges. Due to the relatively
high costs involved in installation of a number of
steel pipes along the entire length of piles as sonic
emitter and receiver access tubes, it is employed for
large-scale projects. To be cost effective, use of the
access tubes for sonic logging tests is sometimes
extended for pile base grouting and for checking soil
conditions below the pile base.
During the 1990s, sonic logging tests were
extensively used for foundation piles of elevated
highways and bridges in Thailand. After 2000,
large bored piles up to 2.0m in diameter and
barrettes are now used for foundations of buildings,
viaducts and cable-stayed bridges.
Whereas
previously, bored piles of 1.5m in diameter were
routinely used for construction of large buildings.
Corsshole sonic logging tests have become
compulsory for quality insurance for these large
piles.
Two access tubes are required as a minimum for
the sonic logging test. A recommended number of
tubes for good coverage of different pile sizes is
shown in Table 1.
Table 1. Recommended number of tubes for different pile size
Pile Diameter
Minimum No. of
Tube spacing
(mm)
tubes
(degree)
D 750
2
180
750D 1000
3
120
4
90
1000D 1500
6
60
1500 D 2500
8
45
2500 D

4.2 Defect and Anomaly Classifications

Faiella and Superbo (1998) published defect
classification criteria for piles monitored by different
numbers of access tubes based on interpretation
from tests on over 6,000 piles from 37 sites in Italy.
In Thailand, it was based on the experience of
testing specialists who performed sonic logging tests
on over 3,000 piles in the past 15 years, coupled
with the analysis on both sonic logging test results
and core samples taken from the zones detected with
anomaly. The classification of anomalies was
established with signal transmission time criteria.
Figure 13 shows the sonic signal profiles of 4
categories classified in Table 2.
Table 2. Category of sonic logging signals and anomaly
classification, Thasnanipan et. al (2004).
Category Criteria
Integrity
Classification
A
No or slight increase in transmisHomogeneous
sion time not more than 10%
concrete
B
Minor increase in transmission
Minor
time >10% but <25%
anomaly
C
Transmission time more than Cat- Major
egory B but transmission time can anomaly
stilled be defined
D
Transmission time cannot be deDefect
fined

A review of sonic logging test results and pile

construction records of 148 large diameter bored
piles (dia. 2.0m, length 64.0m) was carried out to
study the detected anomalies and its causes.
Anomalies were detected in 14% of the piles, with
equal percentage of minor and major anomalies and
defects as classified in Table 2. Construction
records indicated these anomalies were associated
with relatively high sand content in drilling slurry
(the sand content was still under the limit of
specification). Actually the high sand content was
contributed by soil disturbance caused by removal of
underground obstruction the existing old pile at the
project pile location. Another major cause of
anomaly in pile was found to be due to congested
reinforcement arrangement. As the reinforcement
cage was considerably large, horizontal stiffener
rings were included for lifting at levels of pile cutoff
and every cage section lapping. Such reinforcement
with spiral bars of the pile reduced the spacing of
horizontal bars down to 32mm around the rebar cage
at the above mentioned levels for some piles.
Contaminated concrete and sediment were
considered to be trapped at these congested areas
during concrete pouring. This was later confirmed
when piles were trimmed to cut-off level and
trapping of slurry contaminated concrete was found
below the horizontal bars.
Based on the review of both records of pile
construction and sonic logging tests, a summary of
causes of anomaly is presented in Table 3.
Table 3. Summary of causes of anomaly in concrete of large
diameter bored pile found in a typical project.
Causes
Probability
High
Medium
Low
Reinforcement bar congestion
X
Relatively high sand content in
X
slurry
X
Concrete pouring time longer
than usual
Interruption and concrete pouring
X
X
Localized collapse of borehole
surface

Figure 13. Sonic profiles showings different category (A, B, C

and D) as classified in Table 2, (after Thasnanipan et al 2004).

4.3 Additional information from sonic logging test

results
As additional information, sonic logging test records
can also be used for quantifying overcast length and
volume of hard concrete above pile cut-off. In
project planning/bidding of prospective main
contractors this information gave the accurate
quantity of concrete to be trimmed (Thasnanipan et.
al 2004).

470

5 PARALLEL SEISMIC TEST

Parallel seismic test is used in Thailand for checking
the length and integrity of old foundation piles
where both construction records and pile head were
not accessible for investigation. Available records
indicate that the earliest parallel seismic test in
Thailand was performed in 2004.
The test is adopted for quality assurance of tunnel
design alignment since the tunnels are usually
planned along roads and canals. Bangkok has many
canals with bridges crossing over and flyovers at
road intersections. Frequency of parallel seismic
tests is increasing, as tunnels for subway, water
supply, storm-water drainage system and power

transmission lines are being designed and

constructed to pass adjacent to the foundation piles
within the available corridor and due to subsoil
conditions. To date parallel seismic tests were
performed on a total of 190 piles for 50 projects.

for investigation of the subsoil condition in the close

vicinity of the pile to be tested.

Table 4. List of projects using parallel seismic test for pile

integrity and length determination
Number of
Number of Piles
Description of Site
Project
Tested

Use of low-strained integrity tests in Thailand for

quality assurance of deep foundation piles has been
presented. Low-strain integrity tests are effective
methods to assess the pile integrity.
Signal
interpretations need to be compared with soil
investigation results and pile construction records.

Bridge
Old building
Building near Pile
Driving

2
47
1

15
145
30

6 CONCLUSIONS

ACKNOWLEDGEMENT
The authors wish to express their appreciation to
EDE Co., Ltd. and STS Engineering Consultants
Co., Ltd. for providing the test data and project
information.
REFERENCES

Figure 14.
1987)

Parallel seismic test (after Williams and Stain,

Figure 15. A typical test record of parallel seismic test (after

Williams and Stain, 1987)

Information on old pile length and integrity is

also used in carrying out geotechnical analysis for
studying the impact of tunneling, deep excavation
and pile driving works on structure supported by old
piles. To be cost effective, the parallel seismic test is
carried out using the same borehole which is drilled

471

Thasnanipan, N., Aye Z.Z., and Thayanan B. and Kampananon

N., (2004). Quantifying Pile Head Condition before
Basement Excavation by Cross-Hole Sonic Logging Tests,
Proceedings of 7th International Conference on Application
of Stress-Wave Theory to Piles, Kuala Lumpur, Malaysia.
9-11 August 2004. Pp. 233-238.
Thasnanipan, N., Maung A. W., Tanseng P., and Aye Z. Z.
(2000a). Sonic Integrity Test of Piles Integrity Effected
by Basement Excavation, Proceedings of 6th International
Conference on Application of Stress-Wave Theory to Piles,
Balkema, Rotterdam. Pp. 163-170.
Thasnanipan, N., Maung A. W., Navaneethan T., and Aye Z. Z.
(2000b). Non-destructive Integrity Testing on Piles
Founded in Bangkok Subsoil,
Proceedings of 6th
International Conference on Application of Stress-Wave
Theory to Piles, Balkema, Rotterdam. Pp. 171-177.
Thasnanipan, N., Maung A. W., and Baskaran B. (1998).
Sonic Integrity Test on Piles Founded in Bangkok Subsoil
Signal Characteristics and Their Interpretations,
Proceedings of 4th International Conference on Case
Histories in Geotechnical Engineering, St. Louis, Missouri.
9-12 March 1998. Pp. 1086-1092.
Faiella, D. & Superbo, S. (1998). Integrity Non-destructive
Tests of Deep Foundations by Means of Sonic Methods Analysis of the Results Collected on 37 Sites in Italy, Van
Impe & Haegeman (eds), Proc. 3rd Int. Geotechnical
Seminar on Deep Foundations on Bored and Auger Piles,
BAP III, Ghent, Belgium, 19-21 October 1998: 209-213.
Rotterdam: Balkema.
Turner, M. J. (1997). Integrity Testing in Piling Practice,
Construction Industry Reasearh and Information
Association, CIRIA Report 144.
6 Storeys Gate,
Westminster London.
Balasubramanium, A. S. (1991). Inaugural Lecture on
Contributions in Geotechnical Engineering, Soil Mechanics
and Foundation Engineering, Asian Institute of
Technology, Bangkok, March 1991.
Ng, K. C. (1983). The Construction Problem and Performance
of Large Bored Piles in Second Sand Layer, M. Eng.
Thesis No. GT 82-26, Asian Institute of Technology,
Bangkok, Thailand.

Proceedings of International Symposium on Underground Excavation

and Tunneling
2-4 February 2006, Bangkok, Thailand

Urban Tunnel Construction

for Protection of
Environment
Edited by:
Phienwej, N.

Ground Movement Prediction and Building

Damage Risk-Assessment for the Deep
Excavations and Tunneling Works in Bangkok
Subsoil
Zaw Zaw Aye

SEAFCO Public Co., Ltd. Bangkok, Thailand

Dhiraj Karki

CPG Consultants Pte. Ltd., Singapore

Christian Schulz

Bilfinger Berger AG , Bangkok Thailand

473

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

Ground Movement Prediction and Building Damage RiskAssessment for the Deep Excavations and Tunneling Works in
Bangkok Subsoil
Zaw Zaw Aye1 , Dhiraj Karki2 and Christian Schulz3
1

SEAFCO Public Company, Limited, Bangkok, Thailand, (former employee of JV BCKT)

CPG Consultants Pte. Ltd., Singapore, (former employee of JV BCKT)
3
Bilfinger Berger AG , Bangkok Thailand, (former employee of JV BCKT)
2

ABSTRACT
This paper presents the methods adopted in prediction of excavation and tunneling induced
ground movement and building damage risk-assessment carried out for the Contract No. 1 of
the M.R.T Chaloem Ratchamongkhon Line, the first underground mass transit system project
of Bangkok. The measured ground and associated movement of the buildings and structures
were within predicted values. All deep excavation works and tunneling were successfully
completed without any significant damages to the adjacent buildings and structures.
1. PROJECT OVERVIEW
The M.R.T Chaloem Ratchamongkhon Line is the first underground mass transit system
project in Bangkok developed by the Mass Rapid Transit Authority (MRTA) of Thailand.
Total length of the underground structure is 21.5km, comprises of 16 km of twin single-track
bored-tunnel, a total of 4 km long18 cut- and-cover stations, a twin 1.5km long cut-and-cover
approach tunnel to Depot and other associated structures. The project was commenced in late
1996 and opened for the public in 2004. The Contract No. 1, Underground Structure South southern portion of this initial system comprising 9 km of twin 6m outside diameter bored
tunnels, nine underground stations and cut-and-cover depot approach tunnel, was awarded to
the Joint Venture BCKT consisting of Bilfinger + Berger Bauaktiengesellschaft, Ch
Karnchang Public Company Limited, Kumagai Gumi Company Limited and Tokyu
Construction Company Limited. Figure 1 shows the full length of the Contract No. 1 in
simple schematic layout.
One of the major challenges of the Contract No. 1was the need of stacked-alignment
for a major portion of twin bored-tunnel underneath Rama IV road, one of the busiest roads of
Bangkok, which called for a stacked configuration at 3 stations, Lumpini, Si Lom and Sam
Yan. This requirement led the contractor to design and construct the deepest underground
structures of Bangkok at Si Lom Station - four-floor stacked-station of 32m deep right under
the existing flyover and adjacent to a number of sensitive structures. Deep excavation and
bored tunneling works in congested urban environment of Metropolitan Bangkok requires a
systematic process of ground movement prediction, building risk damage assessment and
protection. This paper presents the methods adopted in prediction of excavation and tunneling
induced ground movement and building damage risk-assessment carried out for the Contract
No. 1.

475

2. SUBSOIL CONDITIONS

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

Similar to other localities in Bangkok, a typical subsoil profile along the tunnel alignment is
characterized by the alternating layers of clay and sand deposits as shown in Figure 1.
Weathered crust of 2 m thick is commonly found as the top layer. In urban areas of Bangkok,
this layer is covered by fill material. Soft to very soft, highly compressible dark gray marine
clay lies beneath weathered crust or fill. Depending on the location, this layer extends up to
12-16 m. About 2 m thick medium clay layer can be observed between soft clay and
underlying stiff clay. Generally stiff Clay layer occurs directly underneath medium clay and
its depth goes up to 22 m. Below stiff clay layer, first sand layer of 5-8 m thickness can be
found. This first sand layer, however, is absent in some areas. Stiff to hard clay layer
underlies first sand and it is found to be about 5 m thick. Second sand layer generally occurs
at depths between 45 to 65 m.

Figure 1. Generalized soil profile along the tunnel alignment and stations
3. OVERVIEW OF GROUND MOVEMENT PREDICTION AND BUILDING

DAMAGE ASSESSMENT PROCESS APPLIED IN THE PROJECT

Staged Assessment for each station and tunnel sections was carried out according to the
requirement stipulated in the Outline Contract Design Specification (OCDS). Specified
references were used as basic literatures in ground movement prediction and building damage
assessment. For cut-and-cover excavation zone, the work of Peck (1969) and Clough and
ORourke (1990) were used whereas published papers of Burland et. al. (1977) and Boscardin
& Cording (1989) were applied for bored tunnels. Building damage assessment was carried
out in 3 stages; Stage 1 Assessment, Stage 2 Assessment, and Stage 3 Assessment.

476

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

Greenfield Site assumption was adopted in Stage 1 and 2 Assessment ignoring any
influence of the stiffness of the structure and type of supporting foundation. Stage 2
Assessment was further extended by Additional Stage 2 Assessment. In Additional Stage 2
assessment, the variation of settlement trough shape and magnitude with depth were
considered in identifying the risk of building damage. In Stage 3 assessment, a detailed
qualitative assessment of structure and vulnerability of each building or structure to damage
were taken into account, coupled with detailed structural analysis in assessing the buildings
falling in risk category 3 classified by Mair et. al. (1996).
Table 1: Building Damage Classification (after Burland et. al., 1977 and Boscarding and
Cording, 1989)
Risk
Category

Degree of
Damage

Description of Typical Damage

Negligible

Hairline cracks

Very Slight

Slight

Moderate

Fine cracks easily treated during

normal decoration
Cracks easily filled. Several slight
fractures inside building. Exterior cracks
visible.
Cracks may require cutting out and
patching. Door and windows sticking.

Severe

Very
Severe

Extensive repair involving removal and

replacement of walls, especially over
doors and windows. Windows and door
frames distorted. Floor slopes noticeably.
Major repair required involving partial or
complete reconstruction. Danger of
instability.

Approximate
Crack Width
(mm)

0.1 to 1
1 to 5
5 to 15 or a number
of cracks greater
than 3
15 to 25 but also
depends on number
of cracks

Max.
Tensile
Strain %
Less than
0.05
0.05 to
0.075
0.075 to
0.15
0.15 to
0.3
Greater
than
0.3

Greater than 25 but

depends on number
of cracks

4. SUMMARY OF GEOTECHNICAL INSTRUMENTATIONS

Instrumentation played a major role in the construction of this first underground mass transit
system project in congestive urban area with difficult soil condition. The primary objective of
the instrumentation program was to monitor the performance of the deep excavation and
tunneling operation to ensure safe execution of the construction works and that adjacent
structures were not adversely affected. Furthermore, the instrumentation program was
established to provide feedback in application of the observational method. Extensive sets of
instrumentation were installed around the predefined influence zones of station box
excavation, tunneling and associated building, structures and utilities. In addition to other
geotechnical instruments, inclinometers, extensometers, various types of settlement points,
tile-meter, crack meter and crack gauge were installed to monitor the ground and associated
building and structures movement.
5. BUILDING CONDITION SURVEY

477

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

Prior to commencement of the major works, all the existing buildings and structures within
the predefined influence zone were systematically surveyed. The primary objectives of the
surveys were;
To record pre-construction status of all buildings and structures within the influence
zone. The records included were age of building, building type, storey height,
expected foundation types, visible defects supported by sketches and photographs as
appropriate
To utilize the information collected from the building / structure condition survey in
the assessment of the buildings and evaluate sensitive structures with respect to
construction aspects and monitoring
To establish a benchmark to monitor the possible effect of construction.
Expected foundation types of different buildings were classified based on the height,
number of floors and age of the buildings. Building survey information was also used to
identify the types of instrumentation required.
6. PREDICTION OF GROUND MOVEMENT INDUCED BY STATION EXCAVATION
WORKS AND BUILDING DAMAGE ASSESSMENT

6.1 Prediction of ground movement induced by station excavation

Surface settlement induced by station excavation was predicted based on the dimensionless
design settlement profile behind the earth-retaining structures. The dimensionless design
settlement profile was developed from the results of various prediction methods such as
numerical analysis carried out by CRISP program, method proposed by Bowels (1990) and
Clough & ORourke (1990).
6.2 Stage 1Assessment
Based on the calculated surface settlement described above, contours of settlement around the
station and associated underground structures were plotted in the topographic plan with
buildings and other structures. Figure 2 depicts the surface settlement contours prepared for
one of the stations.

ENTRANCE B

ENTRANCE A

STATION

ENTRANCE C

40
30
20
10

478

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

Figure 2. Settlement contour in station excavation zone prepared for stage 1 assessment
Buildings falling outside 10mm settlement contour were eliminated from further
assessment. Remaining buildings with settlement greater than 10mm were short-listed for
Stage 2 Assessment.
6.3 Stage 2 Assessment
The dimensionless settlement profile for the station box construction were fitted using a
polynomial curve to identify the point of inflexion and divide the settlement profile into
sagging and hogging zone and calculated the tensile strain induced in the buildings from
excavation works using the method proposed by Mair et. al. (1996). The results were then
used to identify the category of risk to damage according to the classification of Building
Damage Classification presented in Table 1. Buildings classified within Slight risk category
of damage were eliminated from further assessment. Buildings falling into Moderate or
higher risk category of damage were further evaluated.
6.4 Additional Stage 2 Assessment
The settlement estimated in Second Stage Assessment presented above was based on surface
movement. For the buildings located in close proximity to the excavation founded on piles, it
was considered that subsurface settlement may be more critical. Stage 2 Assessment was then
further extended by Additional Stage 2 Assessment in which the variation of settlement
trough shape and magnitude with depth were considered in identifying the risk of building
damage. Settlement of the buildings due to the excavation works were predicted assuming the
green filed condition. Neglecting restraints from foundation and structure, it was assumed
that buildings follow the ground settlement trough at foundation level (estimated pile tip
level).
Available methods were further reviewed for predicting the surface settlement with
the consideration of two main factors; (1) simple and practical in application (2) enable to
correlate with the predicted and measured diaphragm wall deflection. The method proposed
by Bowels (1988) was selected as it meets the required criteria mentioned above. Bowels
suggested the ground settlement induced by excavation as a function of ground loss due to the
deflection of the retaining wall. Bowel demonstrated the calculation of settlements at
specified distance by assuming parabolic variations of settlement within the influence
distance. Using predicted diaphragm wall deflection, surface settlement behind the
wall was computed by empirical formulas proposed by Bowels.
6.4.1 Prediction of vertical sub-surface settlement
A simplified prediction of subsurface settlement was carried out based on the calculated
surface settlement described above. First, subsurface settlement influence line was
constructed. As shown in Figure 3, settlement influence zone is assumed to decrease with
depth from Do at the surface and zero at the wall toe. With the assumption of linear
relationship between the volume of deflected wall shape and the volume of settlement trough
at any depth within settlement influence zone, subsurface settlement at different depths were
calculated.

479

Urban Tunnel Construction for Protection of Environment

Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

W
Total deflected
shape volume, V O

S WO

Sio

Surface settlement trough

volume, V To
Dy

HW

HE
Y

Swy

Sub-surface settlement
trough volume, VTy
Settlement
influence zone
deflected volumn V y at depth (H w- Y)
Siy

D-wall

DDoO== H2.5
W+ WHE

S WO =

4V O
DO

S io = S WO *
DO
D
DY = Y * O
HW
VTY = VY *

VTO
VO

Figure 3. Demonstration of subsurface settlement prediction from diaphragm wall deflection

values
Based on the polynomial curve developed in Stage 2 Assessment, subsurface
settlement can be predicted by the similar approach described above.
6.4.2 Prediction of subsurface horizontal ground movement
Assuming a simple linear distribution of horizontal ground movement with distance,
prediction method for sub-surface horizontal ground movement was developed. The influence
distance of horizontal displacement behind the diaphragm wall was assumed to be 2.5 Hg at
ground level as shown in Figure 4. The influence depth of horizontal displacement is a depth
at zero diaphragm wall deflection, Hw, which can be obtained by extrapolating the diaphragm
wall deflection profile. Horizontal ground movement influence zone was constructed by
connecting the end points of Do and Hw as shown in the figure. Horizontal displacement is
assumed to decrease linearly with distance, maximum at wall face and zero at end of
influence zone.
Do = 2.5 Hg
Doi

Xi

Zi

D Oi =
Shi

S hi = S hwi *
Limit of
influence zone

D-wall
y

DO
* H Wi
HW

=
H Wi
HW Zi

Doi

Hwi

Hw

Shi
Shwi

Hg

Shwi

D Oi X i
D Oi

Deflection curve
extrapolated to zero

Figure 4 Demonstration of subsurface horizontal movement prediction from diaphragm wall

deflection values

480

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

6.4.2.1 Risk of damage based on maximum tensile strain

Based on the polynomial curve developed in Stage 2 Assessment and predicted subsurface
ground movement using method described in section 6.3.1, maximum tensile strain of the
building was predicted applying the method proposed by Mair et. al. (1996). Risk category of
to the building was then classified according to Table 1.
6.4.2.2 Risk of damage specified by angular distortion and horizontal strain
The chart proposed by Boscarding and Cording (1989), was used to determine the potential
damage of the building. The chart is a simple tow-dimensional plot of horizontal strains
versus angular distortion in which zones of potential damage to buildings are mapped.
Angular distortion () was calculated based on the difference between predicted settlement of
building one edge to another or two adjacent foundation supports. Subsurface settlement at
pile toe level was used in determining angular distortion.
= (SV1 SV2) / L
= vertical displacement at one end of the building
= vertical displacement at other end of the building
= building length

SV1
SV2
L

Horizontal strain induced by excavation was predicted assuming horizontal ground

movements at surface level decrease linearly with distance, maximum at excavation wall and
zero at end of influence zone. Average horizontal strain of the building can be computed by ;
where,

h = (Sh1 Sh2) / L
h
= average horizontal strain of building
Sh1
= horizontal displacement at surface at one end of the building
= horizontal displacement at surface at other end of the building
Sh2
L
= building length

Table 2 shows the summarized results of risk category obtained from two different methods
described above and assumed foundation types (without and with pile)
Table 2 Summary of risk category obtained from two different methods and foundation type
Building No. of Building
Block
stories Type

37

Masonry
Concrete
Masonry
Concrete
Masonry
Concrete
Masonry
Concrete

RC

Estimated
Pile Length

Distance
to edge
of wall

7-14m

18m

7-14m

17.5m

7-14m

32m

< 6m

2m

50m

8m

481

Risk category by
maximum tensile
strain
Without
With
pile
pile
0
0

Risk category by
chart of Boscarding
and Cording (1989)
Without
With
pile
pile
1
2

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

The predicted degree of damage of the buildings using simplified chart is more critical
than that of maximum tensile strain method as can be seen in above table. According to the
results obtained from of simplified chart, predicted degree of damage was higher if buildings
were assumed to be with piled foundation rather than without piled foundation.
6.5 Third Stage Assessment
Examples of the buildings and structures that required the stage 3 assessment as a result of
Stage 2 assessment are listed below.
Table 3 The buildings required Stage 3 Assessment in station excavation zones

No. of Storey /
Preliminary
Stage 2 Calculated Total
Risk Category /
Building Type
Estimated Pile Length
Tensile Strain (%)
Degree of Damage
4 / RC*
8m
0.121 %
2 / Slight
22 / RC
>22m
0.159 %
3 / Moderate
4 / RC
7-14m
0.197 %
3 / Moderate
2 / Steel Frame
15-22m
0.288 %
3 / Moderate
5 / RC, Masonry
15-22m
0.166 %
3 / Moderate
2 / RC, Masonry
<6m
0.29 %
3 / Moderate
3 / RC, Masonry
7-14m
0.242 %
3 / Moderate
1 / RC, Masonry
0.191 %
3 / Moderate
6m
* Stage 3 Assessment of this building was required due to the pre-existing damage of the building.

A process of Stage 3 Damage Assessment involved;

Collection of the record of pre-construction defects/damages of the building from
Building Condition Survey Reports
Conducting detailed structural Survey and assessing the current state of the building
Condcuting damage assessment

6.5.1 Collection of building condition survey

The building condition survey records were collected and primary review of the state of the
building prior to commencement of construction works.
6.5.2 Detailed Structure Survey
In order to understand the building response to the ground movement and to carry out the
qualitative damage assessment of the building, details of the structure should be known in
advance. A structural survey was carried out by experienced structural engineer to obtain the
detailed information of structural members.
6.5.3 Damage Assessment
Structural survey information was the key factor in the selection of methodology for the Stage
3 Assessment. Based on the pre-construction state building condition survey and detailed
structural survery, the method for damage assessment was selected. Two main approaches
were applied.

482

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

Conservative approach, mainly applied in additional stage 2 assessment with some

refinements
Frame analysis using FEM structural software

6.5.3.1 Stage 3 Assessment by conservative approach

Conservative approach applied in Stage 3 Assessment involved consideration of two separate
criteria, namely ;
The risk of damage to the building specified by angular distortion and horizontal strain.
The risk of damage to the building in terms of magnitude of building slope with reference
to the limiting values specified by various researchers.
The following assumptions were made in the conservative approach of Stage 3
assessment.
Buidling movement is influence by excavation induced settlement of both main station
box and associated underground structures (e.g. Entrance Building)
Building settlement was casued by subsurface settlement at foundation level as buildings
are supported by piled foundation.
No allowance was made for structural stiffness of the building.
No allowance was made for possible effects of soil-structure interaction.
A summary of risk category determination in Stage 3 Assessment based on the predicted
and monitored settlement of the building adjacent to station excavation work is shown in the
Table 4 below.
Table 4 Summary of risk category determination based on Boscarding and Cording (1989)
Data Taken
From

Horizontal Strain

Sh1
Sh2
(mm) (mm)

Settlement
31
Prediction
Actual
31*
Monitoring Data

L
(m)

Angular Distortion

Sv1 Sv2
S
(mm) (mm) (mm)

L
(m)

Risk
Category

Degree of
Damage

24

18.6 0.0004

48

23

25

18.60

0.0013

slight

24*

18.6

0.0004
*

53

14

39

18.60

0.0020

slight

*Predicted horizontal movement was used in computing horizontal strain for actual
miniotring data.
From the above table, following conclusions can be made.
Predicted angular distortion broadly agrees with that of actual monitoring data. Hence,
settlement prediction and assumed effective depth of piled foundation is fairly reasonable.
The settlement of actual monitroing data is slightly higher than that of prediction. This
may be due to the additional settlement caused by diaphragm wall panel construction
which influence zone might have extended to 15m from the diaphragm wall (equivalent
distance to bottom of soft clay level ).
Limiting values of the building slopes at two critical conditions specified by various
researchers as shown in the table below were also used in determining the potential building
damage of stage 3 assessment.

483

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

Table 5 Summary of Limiting Building Slope Specified by Various Researchers

Condition

Type of Damage

1
2

Limiting Values by Various Researchers

Skempton and
Macdonald

Meyerhof

Polshin and
Tokar

Bjerrum

Cracking in walls and partitions

1/300 to 1/500

1/500

1/500

1/500

Structural Damage

1/150

1/250

1/200

1/150

6.5.3.2 Stage 3 Assessment by structural frame analysis

Some buildings of critical condition were assessed by structural analysis using detailed
structural survey information as basic data. By simulating the differential settlement of the
building predicted from Additional Stage 2 Assessment, development of moment and shear in
the main structural members were determined with the use of structural analysis software.
Computed moment and shear were assessed in comparison with the capacity of the structural
members estimated from collected building condition survey and detailed structural survey
data. The process involved in structural assessment can be summarized below.
Factor of safety of each structural members were determined base on predicted
differential settlement of the buildings.
A series of assessment were performed to determine the critical differential settlement
causing structural instability of the members or differential settlement at limiting
factor of safety.
Trigger values were established based on above results.
Based on the trigger values and the actual building condition and response in
conjunction with monitoring data, the most sensitive structural members were
protected, mainly by strengthening approach.
6.5.4 Requirement of Stage 3 Assessment for bridge structures
Due to the close proximity to the adjacent station, the predicted maximum settlement in the
Stage 1 Assessment of the flyover and elevated expressway structures was greater than
10mm. Therefore, further assessment using Stage 2 approach was required. However, since
flyovers and elevated expressways are bridge structure with piled foundations, the approach
of strain calculations used in Stage 2 Assessment was considered inappropriate and they were
placed under the requirement of the Stage 3 Assessment as listed in Table 6.
Table 6 List of Flyovers / Bridges required for Stage 3 Assessment
Location

Name of the Bridge /

Flyover

Type of
bridge

Foundation Type

Hua Lamphong

Second Stage Expressway

Steel

Deep-seated bored pile

Sam Yan

Thai-Japan Flyover

Steel

Deep-seated bored pile

Si Lom

Thai-Japan Flyover

Steel

Deep-seated bored pile

Lumphini

Thai-Belgium Flyover

Steel

Deep-seated bored pile

Phetchaburi

Asok Flyover

Steel

Pre-cast pile

484

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

Stage 3 Assessments of above listed bridges were carried out by soil-structure interaction
analysis using finite element method. FEM Analyses were mainly carried out by in-house
design team consisted of experience geotechnical engineers of Tokyu Construction Company
Limited. Two-dimensional non-linear soil model applying Duncan-Chang parameters was
employed. The process involved in soil-structure analysis is presented in diagram as shown in
Figure 5.

Assign Data and Conditions for Analysis

Structures
Ground Properties
Construction Sequence
Finite Element Analysis
Non-linear Elastic Analysis

Superstructure assessment

Pile Assessment

Distribution of movement of the

ground surface

Distribution of horizontal ground

movement along piles

Calculation of vertical and horizontal

movement of the piers

Structural analysis

Assessment
based on
differential
Settlement criteria

Assessment
based on
horizontal
movement criteria

Applied response
displacement
method

Verification of RC section of
pile

Figure 5 Diagram showing process of stage 3 assessment for bridge structure

Among the bridges listed in Table 6, impact of excavation work on the structure of
Petchaburi flyover was considered the most critical due to its close proximity to the
diaphragm wall and having relatively less rigid and shallower foundation of pre-cast piles.
Detailed structural assessment of the existing pre-cast piles of the flyover was performed
using results of FEM and verified with structural capacities of the piles calculated from the
available as-built drawings. Predicted differential settlement of two adjacent piers (pier close
to diaphragm wall and pier away from diaphragm wall) caused by subsurface settlement at
pile toe level was used to calculate the angular distortion of the flyover superstructure and
compared with the tolerable angular distortion specified by for simple span bridge. Predicted
angular distortion (0.0003) was much lower than specified tolerable value of ASSHTO
(0.005). Predicted horizontal movement of flyover piers were 10mm and 12mm for left and
right pier (of Figure 6) respectively. Tolerable horizontal movement of flyover structure based
on the observations of actual bridge performance from the works of various researchers
(extracted from Xanthakos, 1995) are shown in Table 7.

485

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

Table 7 Tolerable horizontal movement of bridge recommended by various researchers

Horizontal
Recommendation
Recommended by
Movement (mm)
25
Not harmful
Bozozuk (1978)
38
Tolerable in most cases
Moulton et. al. (1985)
51
Harmful but tolerable
Bozozuk (1978)

Extensive sets of instrumentation were installed including tilt meter, settlement points
and inclinometers. Monitoring was carried out in high frequency during critical stage of
excavation and compared with established trigger values.

Figure 6 Cross-section of Petchaburi Station diaphragm wall showing the position of

adjacent flyover foundations
7. PREDICTION OF GROUND MOVEMENT INDUCED BY TUNNELING AND RISK
ASSESSMENT

7. 1 Prediction of surface settlement

The Gaussian normal distribution curve developed by Peck (1969), OReilly and New (1982)
was used to estimate the ground settlement induced by a single tunnel.
7.2 Stage 1 Assessment
First stage assessment of the risk of damage to buildings and structures in tunneling zones
involved preparation of settlement profile induced by tunnel excavation. Settlement was
computed at every 100m interval along tunnel alignment using the method based on Gaussian
normal distribution curve presented above. Settlement induced by each tunnel (north and

486

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

south bound) was predicted independently and then superimposed to determine the total
settlement induced by the twin tunnels. 10mm and 25mm settlement contours were plotted in
the topographic plan with buildings and other structures including flyover and khlong bridges
in the tunnel alignment. Buildings falling outside 10mm settlement contours were eliminated
from further assessment. Remaining buildings with settlement greater than 10mm were shortlisted for stage 2 assessment.
7. 3 Stage 2 Assessment
The buildings and structures short-listed in stage 1 were assessed in this stage. Stage 2
assessment involves calculation of maximum tensile strain induced in the building and
structures due to tunnel boring and classification of risk category. Risk category of the
buildings was determined applying the method proposed by Mair et. al. (1996), similar to the
building damage assessment made for excavation induced settlement presented in earlier
sections.
7. 4 Additional Stage 2 Assessment in tunnel zones
As most of the buildings and structures in Bangkok along tunnel alignment are supported by
piled foundations, it was considered that Stage 2 Assessment using surface settlement profile
might be insufficient. Hence, Additional Stage 2 Assessment considering the variation of
settlement trough shape and magnitude with depth (subsurface settlement) was carried out for
the critical buildings.
In general, subsurface settlement trough will be narrower and steeper than those of the surface
as shown in the sketch below. A simple approach was used in predicting the subsurface
settlement profile with following assumptions.
(1) The shape of the subsurface settlement profile caused by the tunnel construction is
characterized by a Gaussian distribution similar to surface settlement profile.
(2) The ground settlement influence line for tunnel in clay is the same for subsurface as
shown in the figure below:

Figure 7 Distribution of surface and subsurface settlement induced by tunnel boring

Therefore, substituting the distance above the tunnel axis Z0-Z for Z0, equations derived from
Gaussian normal distribution curve subsurface settlement profile was determined. In addition
to this method following equation can also be used to predict the subsurface settlement.

487

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

K = [0.14+0.26(1-z/z0)] / (1-z/z0)
By substituting K values for K in equation derived from Gaussian normal distribution curve
the subsurface settlement at required depth can be calculated.
7. 5 Stage 3 Assessment
7.5.1 Stage 3 Assessment for the buildings
The method used in Stage 3 Assessment of the buildings in tunnel zone was similar to that of
station excavation zone based on the superimposed predicted subsurface settlement trough
induced by twin tunnels.
7.5.2 Structures Required for Stage 3 Assessment
The most critical structure required for Stage 3 Assessment were MWA main water tunnel as
tunnel alignment crosses the water tunnels at two critical locations the first near Rama IV
and Rachadaphisek road intersection and the second at Rama IV and Siphraya road
intersection near Sam Yan station. The position of water tunnels in relation to MRT tunnels
drawn in the subsoil profile at relevant locations are presented in Figure 8 and 9.

Figure 8 Position of MWA water tunnel in relation to MRT tunnels (clear distance are 3.38m
and 3.18m for North and South bound respectively) at the first crossing
According to the available information, MWA tunnel at the first crossing is of 2.5m
internal diameter steel lining with 150mm segmental ring and at the second crossing MWA
tunnel is of 2.5m diameter segmental ring with lightly reinforced concrete lining.
In preliminary stage 3 assessment, based on the predicted subsurface profile using the
method described in earlier section, structural assessment of water main was carried out and
results were compared with structural capacity of water tunnel calculated from the material
properties indicated in the construction drawings provided by MWA. As bending moment
induced in the water tunnel was considered critical parameter in assessing the damage which
is governed by both maximum settlement and curvature of the water tunnel further assessment
was proposed.

488

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

To carry out back-calculation using monitoring data obtained from the settlement
arrays (settlement points, extensometer data) at first 3 locations before reaching to
MWWA water tunnel near Rama IV Rachadapisek intersection, to determine the
volume loss and relationship between surface and subsurface settlement as well as
best estimation of subsurface K value and subsurface settlement profile at water
tunnel level.
To perform beam-on-elastic foundation analysis using the subsurface settlement
profile obtained from above item (1) to determine bending moment likely to occur in
MWWA water tunnel.
To compare the result obtained from above item with allowable values.
To establish the final trigger values based on above comparison.
To confirm the settlement profile of last array is within the alert level of final trigger
value.

Figure 9 Position of MWA water tunnel in relation to MRT tunnels (clear distance are 1.5m
and 2m for North and South bound respectively) at the second crossing
For the first location of MRT-MWA tunnel crossing, target ground loss from tunnel
operation was set at 0.5%, whereas ground loss based on allowable movement curvature of
water tunnel derived from predicted pattern of subsurface settlement was 0.65%. Backcalculated ground loss derived from the monitoring data of rod-extensometer placed at the
crown of MWA water tunnel was within target value. In addition to the preliminary
assessment based on prediction of subsurface settlement using empirical formulas, a detailed
soil-structure interaction was analyzed by experience geotechnical engineers of Tokyu
Construction Company Limited using 2-dimensional FEM considering behavior of tunnel
boring by earth pressure balance (EPB) method. Stress relief from the effect of cutter-head,
tail void and over-cut from EPB tunnel boring was considered in the analyses. Adding ground
movement due to consolidation caused by disturbance from tunnel operation surface and
subsurface settlement were predicted by finite element analysis. Detailed parameters and
assumption made in analysis was presented in the published paper of Sakai & Sugden (2000).
After the successful passing of both tunnels at the first MWA crossing, another trial area
tunneling was planned to further determine the ground movement associated with various
tunneling parameters well before reaching to more critical crossing after Sam Yan station.
Tunnel section between Silom and Samyan station, in which soil profile similar to that of
second crossing was selected and comprehensive instrumentation program was established to

489

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

monitor surface and subsurface settlement as well as horizontal movement of the ground at
depth.
8. COMPARISON OF PREDICTION AND ACTUAL MONITORING RESULTS

The monitored ground deformation due to station box construction and bored tunneling were
compared with the predicted values for one of the station and tunnel sections to evaluate the
performance of the actual operation.
8.1 Ground Movement Induced by Excavation Works
The monitored wall deflections and ground settlement induced due to station constructions
works were within the design prediction. Adopted top-down method in excavation works of
station boxes utilizing rigid diaphragm wall extensively propped by stiff permanent slabs, was
the key factor contributed in minimizing associated ground movement and disturbance to
adjacent buildings and structures. Figure 10 shows the comparison of monitored wall
deflections with predicted values at final stages of station construction.
0

10

20

30

40

50

60

-10

70

100

95

95
Reduced Level (m)

Reduced Level (m)

-10
100

90
85
80
75
70

10

20

30

40

50

60

70

90
85
80
75
70

Deflection (mm )

Deflection (mm )

A mber_Co ncourse
Red_Co nco urse
M o nitored_Conco urse

A mber_B ase
Red_B ase
M onito red_B ase

Figure 10 Diaphragm wall deflection showing predicted trigger levels and monitored data for
final excavation stages to concourse and base slab level
The comparison of monitored and predicted ground settlement profile resulting from
station excavation is presented in Figure 11.
S urf a c e S e t t le m e nt P ro f ile B e hind D / Wa ll R e s ult ing f ro m S t a t io n
B o x E xc a v a t io n
20
10
0
-10

-1

-4

-9

-20

-4

-4

-2 -1
-2

-4

-15

-17

-22

-30

-30

-40
-50

-32

DW_East

-60

DW_West

-70
-60

-50

-40

-30

-20

-10

10

20

30

40

50

60

D is t a nc e , m
Initial Reading

M easured-End o f Ro o f Exc.

M easured-End o f A ccess Exc

M easured-End o f Co nco urse Exc

M easured-End o f B ase Exc

P redicted-End o f Statio n B o x Excavatio n

Figure 11 The comparison of monitored and predicted ground settlement profile resulting
from station excavation works

490

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

8.2 Ground Movement Induced by Tunneling Works

The observed ground movement induced by tunneling works were within predicted values.
Well organized plan, systematic approach of observational method to achieve control values
backup by extensive instrumentation and close co-operation among relevant personnel were
the key factors contributed to successful completion of EPB tunneling particularly in the most
sensitive points at MWA water tunnel crossings. The monitored ground settlement profile in
the most critical section of parallel twin tunnel between Petchaburi and Sukhumvit station is
presented in Figure 12. As can be seen in the figure monitored settlement profile was within
that of prediction. The volume loss due to tunneling works in this tunnel zone was back
calculated using the monitored profile and found less than 2% and is typically in the range of
1.1% to 1.7%.
T ra ns v e rs e S e t t le m e nt T ro ugh due t o
T unne l B o ring, P e t c ha buri- S uk hum v it T unne l
20
10
0
-10
-20
-30
-40
-50

Monitored

-60
-25

-20

-15

-10

-5

Predicted

SB

NB

-70

10

15

20

25

D is t a nc e , m

Figure 12 Predicted and observed ground settlement profile due to tunnel boring

Distance (m)

Subsurface Settlement (mm)

10

20

30

40

50

60

70

80

-10

Settlement _ NB Tunnel
-30

Settlement_SB Tunnel
Total Settlement

-50

491

90

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

Figure 13 Predicted subsurface settlement trough of assumed ground loss 1% calculated

from empirical formula and actual subsurface settlement (indicated by dark-circle) observed
from rod extensometer
9. CONCLUSION

The M.R.T Chaloem Ratchamongkhon Line, one of the most modern underground mass
transit systems in the region was successfully completed and has been in operation for over 1
year. Ground movement prediction and the detailed stage assessment carried out in the M.R.T
Chaloem Ratchamongkhon Line are presented. The measured ground and associated
movement of the buildings and structures were within predicted values. All deep excavation
works and tunneling were successfully completed without any significant damages to the
adjacent buildings and structures. Well organized plan, systematic approach of observational
method to achieve control values backup by extensive instrumentation and close co-operation
among relevant personnel were the key factors contributed to successful completion of the
project. The purpose of this paper however is not to extol success but to provide a set of
technical guidelines for ground movement prediction and building damage risk assessment of
deep excavation and tunneling works in Bangkok subsoil which may partly serve as a
reference for the future underground mass transit system projects in urban area of Bangkok.
Successful completion of the first MRT project in Bangkok marked the practicality of wellexecuted deep excavation works and tunneling by EPB method in protecting the sensitive
urban environment.
10. ACKNOWLEDGEMENT
The authors wish to express their appreciation to Mass Rapid Transit Authority (MRTA) for
permission to publish this paper. The authors would like to thank Construction Supervision
Consultants (CSC 1 Consortium) and Project Management Consultant (MPMC) for their
cooperation. Valuable advice and support given by Dr. D. J. Maconochie (CSC Consortium)
and Mr. A. J. Polley (former geotechnical advisor, Sindhu Maunsell Consultants) are
gratefully acknowledge.
11. REFERENCES
Boscarding, M.D. and Cording, E.G. (1989), Building Response to Excavation-Induced
Settlement. Journal of Geotechnical Engineering, ASCE, 115;1;1-21
Bowels J.E (1988) Foundation Analysis and Design, 4th Edition, McGRAW-Hill International
Book Company, Singapore, 1988
Burland, J.B., and Wroth, C.P. (1974), Settlement of Buildings and Associated Damage.
SOA Review, Conference on Settlement of Structures, Cambridge, Pentech Press.
London, pp. 611-654
Burland J.B., Broms J.B. and de Mello VFB (1977), Behavior of Foundations and Structures,
SOA Report Session 2, Proceedings of 9th International Conference, SMFE, Tokyo,
2:495-546
Clough, G.W., and ORourke, T.D., (1990), Construction Induced Movements of In-situ
Walls, ASCE Geotechnical Special Publication No. 25, pp. 439-470
Clough C.W. and T.D. ORourke (1990), Deep Excavation and Tunneling, 7th ICSME,
Mexico City

492

International Symposium on Underground Excavation and Tunelling

Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok

JV BCKT, (1997) Geotechnical Interpretation Report (GIR) for MRTA Initial System UGS, Bangkok
Mair, R.J., Taylor, R.N., and Burland, J.B., (1996), Prediction of Ground Movements and
Assessment of Risk of Building Damage. Geotechnical Aspects of Underground
Construction in Soft Ground, pp. 712-718, Balkema, Rotterdam
OReilly M.P and New B.M (1982), Settlements above tunnels in United Kingdom their
magnitude and prediction, Tunneling 82, pp. 173-181, London, IMM
Peck R.B. , (1969) Deep Excavation and Tunneling in Soft Ground, 7th ICSME, Mexico City,
State of the Art 3, 225-290
Petros P. Xanthakos (1995) Bridge Substructure and Foundation Design, Prentice Hall
PTR, USA, 1995

493

Proceedings of the Observational Method and its Application in Hong

Kong Seminar
18 November 2006, Hong Kong

Observational Method and its

Application in Hong Kong
Key Factors in Application of Observation
Method Bangkok Experience
Noppadol Phienwej
Asian Institute of Technology, Pathumthani, Thailand

Zaw Zaw Aye

SEAFCO Public Co., Ltd. Bangkok, Thailand

495

KEY FACTORS IN APPLICATION OF OBSERVATION METHOD

BANGKOK EXPERIENCE
Noppadol Phienwej1 and Zaw Zaw Aye2
Abstract: Application of observational method in two different projects is presented in this
paper. Intensive modification of construction sequence in actual work execution with value
engineering options different from tender stage design along with application of
observational method in Diaphragm-wall-support two level underground car park building
located in the historically and culturally significant area of Bangkok is firstly demonstrated.
Simple approach of observational method in building damage assessment backup by
extensive instrumentation in tunneling and deep excavation works of Contract No. 1 of the
M.R.T Chaloem Ratchamongkhon Line, the first underground mass transit system project of
Bangkok (MRTA) is also presented. Key Factors contributed in successful application of
observational method are discussed.
INTRODUCTION
Observational Method is not a very popular term among geotechnical engineers in
Thailand, though the method has been applied in some major projects. The main reason of
unpopularity of the method in Thailand may be due to the fact that traditionally there is less
option opened for design modification as the project progresses. Geotechnical engineers may
be more concerned on misused and limitation of the method and do not prefer to take risk. It
is important to discuss among geotechnical engineers with regard to overall philosophy, key
requirement, limitation and practical approach in application of observational method. In this
paper, application of observation method in two different projects is presented.

CONSTRUCTION OF UNDERGROUND CAR PARK IN HISTORICAL

AREA OF BANGKOK
The project is a two-level underground car park located in the center of Rattanakosin
Island, the heart of an old established, historically and culturally significant area of Bangkok.
The project owner, Bangkok Metropolitan Authority (BMA) awarded the semi-turnkey basis
construction contract Lam Kon Muang Underground Car Park to SEAFCO Co., Ltd. as a
contractor. The contract consists of 3 major scope of works : (1) Construction of building
foundation and retaining structure - diaphragm wall, barrette and bored piles (2) Excavation
works including temporary bracing design and installation (3) Construction of the entire
two-level underground car park building having car park area of 18,552 m2 and roof-level
park of 10,936m2 plus cut-and-cover tunnel, underpass access to the City Hall. Geotechnical
Associate Professor and Coordinator, Geotechnical and Geo-environmental Filed of Study,
Asian Institute of Technology, Thailand
2
Senior Vice President, Seafco Public Company Limited, Thailand
1

497

aspects highlighting the application of observational method and performance of

buttress-support diaphragm wall of 0.60m width for two level underground car park building
is discussed in this paper.

Project Requirement and Major Constraints

Since there is a limited availability of car parking space in the surrounding congested
neighborhood and the project site was being used as a grade-level car parking space prior to
the award of the contract, the key requirement was to construct the underground car park in
two phases to construct Phase 1 while leaving space for car parking in Phase 2, and to
utilize semi-finished underground car park of Phase 1 during construction of Phase 2. This
requirement posed the need of temporary retaining wall between Phase 1 and 2.
Construction site is surrounded by numbers of sensitive structures as shown in Fig. 1, in the south, Wat Suthat, one of Thailands most important temples and the Historical
Giant-swing, in the north, the City Hall and at South-east corner, the Historical Brahmin
Temple. Rows of old shop-house buildings are closely located in the east and west boundaries
of the project. Location of the project site itself in the vicinity of sensitive structures and
buildings therefore posed some constraints, which called for the need of careful consideration
in establishing the design principles and sequence of construction.
CITY HALL

OPENING
OPENING

OPENING

BAMRUNG MUANG ROAD

GIANT-SWING

WAT SUTHAT (TEMPLE)

Fig. 1. Layout plan of the project showing adjacent buildings

498

SIRIPONG ROAD

OPENING

OPENING
OPENING

SHOP-HOUSES

SHOP-HOUSES
DINGSOR ROAD

OPENING

SHOP-HOUSES

OPENING

MAHUNNOP ROAD

I-5

I-4

N O R T H D -W A L L

I-6
EAST D-WALL

WEST D-WALL

PH A S E 2

S H E E T PIL E

124 m

S ee d etails in (b )

I-3

I-1

PH A S E 1

S O U T H D -W A L L

I-2

78 m

(a)
D IA P H R A G M W A L L

RC BUTTRESS

BARRETTE

W ALE BEAM

0 .3 0

T IE B E A M

P IL E C A P

SLA B

(b)
+0.20 ROOF SLAB

0.00
CAPPING
BEAM
RC BUTTRESS

G.W.L

2
SOFT CLAY

-9m

-12m

-2.70 SLAB (B1)

PILE CAP
-5.60 (B2)

TIE
BEAM

2
WALE
BEAM

SOFT CLAY

DIAPHRAGM
WALL (tip=-16m)

MEDIUM CLAY

STIFF CLAY

-15m
-16m

-20m.

VERY STIFF CLAY

FOUNDATION PILES
DIA. 600mm (tip=-20m)

BARRETTE
(tip=-20m)

-20m.

Fig. 2 (a) Layout of diaphragm wall and (c)

piles (b) plan of diaphragm wall and buttress (c)
sectional view of the structures
Location of the project site itself in the vicinity of sensitive structures and buildings
therefore posed some constraints, which called for the need of careful consideration in

499

establishing the design principles and sequence of construction. The architectural and utility
aspects of the project called for the design of the basement with a number of openings from
the ground surface to the final basement slab level to facilitate the ventilation system as
shown in figure 2.
Hence the roof slab can not be physically utilized as bracing in most of the area where
the diaphragm wall is to be acting as a cantilever retaining wall in the permanent stage. It was
analyzed in the preliminary analyses that the deflection of the diaphragm wall of 0.60 m width
would be large if it was to be fully cantilevered. As the project is located in a sensitive area,
ground movement induced by large deflection was unfavorable. It was therefore decided to
use buttress to minimize the diaphragm wall deflection as shown in figure 2.
Diaphragm Walls Barrettes and Bored Piles Construction
Diaphragm wall having 600mm width founded at 16m below ground level (B.G.L) was
constructed simultaneously with dry-processed bored piles of diameter 600mm with toe depth
20m below ground level. Barrettes having same toe depth as bored piles were installed at 8m
spacing along with diaphragm wall panels. Sheet pile wall (14m deep) was used as a
temporary retaining wall at the boundary of Phase 1 and 2 as shown in Fig. 3.
Subsoil Condition and Design Parameters
Typical subsoil profile at the site is characterized by thick Bangkok soft clay layer at the
top followed by thin layer of medium clay, and stiff clay layers.
Undrained Shear Strength (kN/m2)

Depth from ground surface (m)

10
15
20

Soft clay

35

Stiff clay
Very Stiff clay

50

150

20
40
60
SPT N-Value (blows / ft)

200

5.00

15.00
20.00
25.00

Medium Dense
Sand

30.00
35.00
40.00

40
45

100

Medium stiff clay 10.00

25
30

50

0.00

0
5

Dense sand

45.00
50.00

Su BH-1
Su BH-3
SPT BH-1
SPT BH-3

Fig. 3. Plotted Su and SPT N-value with design lines

500

Su BH-2
Design (Su)
SPT BH-2
Design (SPT)

80

Undrained shear strength (Su) and SPT N-value obtained from 3 SI boreholes were
plotted and design line was derived as shown in Fig. 3. The design soil parameters are
tabulated in Table 1.
Table 1. Design soil parameters
Soil Type

Depth
(m)

Su
(kN/m2)

Soft clay

s
(kN/m3)

Eu
(kN/m2)

0-9

16.50

18

7000

9 12

17.50

30

19250

12 15

19.00

60

45000

15 - 26

19.50

80

60000

Medium clay
Stiff Clay
Very stiff clay

Original Tender Stage Design

The designers involved in the tender stage design made a fairly conservative design with
two levels temporary bracing as shown in Fig. 4. Uncertainty of the performance of a thin
diaphragm wall in soft clay layer was the likely reason to adopt the conservative design in the
tender stage. It is not unreasonable to adopt the conservative design considering
time-dependent consolidation property of soft marine clay and likely long elapsed time of
un-strutted diaphragm wall (relatively long un-strutted span of about 6.0 m between first
strut and final excavation level for 600mm diaphragm wall) due to large volume of excavation
work involved.

HORIZONTAL STRUT

SLAB

D-WALL

RAKING STRUT

D-WALL

BARRETTE

-16m

BERM

BORED PILE

-20m

Fig. 4. Tender stage bracing system diaphragm wall was designed with soil-berm and 2
struts support (horizontal bracing and raker)
Value Engineering Option for Phase 1
Value engineering review of the temporary works was undertaken by the contractors
new in-house design engineering team prior to the commencement of Phase 1 excavation
works. Rigorous attention to detail of the design concept and constructability was made in the
pre-construction discussions between design engineers and construction team.

501

The main objectives for value engineering options were to minimize the material and
construction sequence involved in temporary works so as to accelerate the excavation time
thereby saving overall costs. without compromising the safety aspect.
After conducting a
series of re-analyses with different conditions major modifications were made : (1) To lower
the first strut level to 1.8m from the original tender stage design level 1.0m (2) To use only
1 temporary strut, omitting second level raking strut with the provision of sloping soil berm
against diaphragm walls. Soil berm was to remove after completion of base slab construction
in the majority of area minimizing the elapsed time of partially un-strutted diaphragm wall
between temporary strut and the final excavation level. Modified bracing system of Phase 1 as
an outcome of value engineering review is illustrated in Fig. 5.
HORIZONTAL STRUT

SLAB

HORIZONTAL STRUT

SLAB

BERM
D-WALL

RAKING STRUT
BARRETTE

-16m

SHEET PILE

-20m

-20m

BORED PILE

SOUTH D-WALL vs SHEET PILE

EAST & WEST D-WALL

Fig. 5.

-14m

-16m

Value engineering option - Modified Bracing system for Phase 1

Implementation of the Observational Method

Professor R. B. Peck set out procedures for the observational method (OM) as
applied in soil mechanics in the Ninth Rankine Lecture (Peck, 1969). Peck described the
limitation and drawbacks of observational method. Powderham (1996) reviewed the main
features of the observational method of Peck and summarized the key requirements as
follows;
(1) It must be possible to alter the design during construction
(2) The contractual condition must be compatible and allow design to be directly related
to actual construction method
(3) An acceptable level of risk must be identified and controlled. In particular this requires
a planned course of action for every foreseeable eventuality
(4) Critical observation must be identified and obtained
During the review of tender stage design for Phase 1, it was recognized that
two-phase excavation works in this project was ideally suitable for application of the
observation method. The flexibility of the contractual requirement in temporary design which
allowed the contractor to modify the design and construction method also provided the favor
for the observational method.
In order to assess the most probable condition assumed in value engineering design
and to take necessary actions if monitoring results reveal most unfavorable conditions (i.e.

502

actual deflection of diaphragm wall reaches maximum acceptable limit), the observational
method was implemented on the followings basis.
Reviewed the design parameters together with critical conditions posed on sites as well as
most critical stage in excavation and basement construction work
Predicted the performance of diaphragm wall with most probable as well as most
unfavorableconditions and parameters.
Established the trigger criteria based on predicted diaphragm wall deflection
Predefined the practical contingency plan for most unfavorable conditions where wall
deflection reaches trigger levels
Set out the instrumentation program with the consideration of above factors
Monitored the performance of diaphragm wall. Compared the monitoring results with the
predicted and trigger values and reassessed
Implemented the contingency measures if monitoring result reaches action level of trigger
values
Figure 6 shows the predicted diaphragm wall lateral displacement or deflection of
Phase 1 (east, west and south diaphragm wall) at two conditions together with trigger levels
and tender stage prediction. It should be noted that the diaphragm wall deflection was
predicted to be maximum or most critical after removing the horizontal temporary strut.
Cum. Displacement , mm.
0
0
2

20
NORMAL

40

60
ALARM

80

100

ACTION

MOST
PROBABLE
PREDICTION

Depth, m.

6
MOST
UNFAVORABLE
PREDICTION

8
10
12
14
16
18
20

Tender Stage
Prediction with buttress :
Alarm Trigger Level
Prediction without buttress
Action Trigger Level

Fig. 6. Prediction of diaphragm wall performance in pre-construction stage with trigger

criteria in comparison with tender stage prediction
Most probable condition was established for the predicted deflection of diaphragm
wall with full influence of buttress support assuming buttress effectively supports as
permanent strut in diaphragm wall analysis model. Most unfavorable condition was set out for
the predicted deflection of diaphragm wall without considering influence of buttress
buttress was excluded in the model. Diaphragm wall reinforcement was designed based on the

503

most unfavorable condition.

In establishing the criteria for trigger values, it was necessary to consider the broad
context in which diaphragm wall exists, design assumption and concept, likely behaviour of
the wall itself or its predicted performance and effectiveness of selected temporary bracing
system. Trigger levels were established to provide the design team and construction team, an
opportunity for early review and resetting of the monitoring frequency as well as for
implementing the contingency measure as necessary.
In general, exceeding alert trigger levels must initiate a review of design data,
construction progress and monitoring frequencies with the consideration of possible measures
to limit further deflection. Exceeding action trigger levels must initiate further review of
above mentioned points and if necessary to initiate a planned course of action or contingency
measures.
Effective and good communications between the design team and construction crew
were made all along the excavation stages with clear responsibilities in construction control.
The contingency measures included immediate backing filling of excavated soil and
installing of temporary struts in the critical area.
Monitoring Program
In planning the monitoring system and program, it is important to consider the
parameters to be measured which reflect the actual performance of the diaphragm wall
support excavation. It is also necessary to take account the practical measurement applicable
for established trigger criteria, number and frequency of measurement required to carry out a
meaningful interpretation of wall behaviour which would be integrated in the implementation
of observational method.
Comprehensive and robust monitoring program was set up as a key element in
application of the observational method and to ensure that modified construction sequence
would not have adverse effect in temporary stage and on permanent design. A total of 6
inclinometers (3 in each phase) were installed in diaphragm wall together with some survey
points. In order to make effective use of the established trigger levels, an adequate number of
measurement were carried out at appropriate frequency. Typical monitoring frequencies for
inclinometer set up as guideline for the project is outlined below.
Measured immediately before commencing excavation in the vicinity of instrument
Minimum readings of 2 times a week while excavation in progress
Minimum readings of 1 time a week when no excavation the vicinity of instrument
Minimum readings of 3 times a week when measured deflection values exceeded alert
trigger levels
Minimum reading 1 time a day when measured deflection values reached action trigger
levels
As the most critical stage was predicted at the time horizontal temporary bracings
were removed, a full attention was paid to the inclinometer monitoring with the following
special criteria and frequencies.
First temporary strut removal was to carry out at the diaphragm wall panel where
inclinometer was located
Measured immediately before removal of temporary strut at the closest distance to the
instrument.
Measured every 6-8 hours immediately after removing the first strut
Second strut to be removed must be the one located immediately adjacent to first strut

504

which had removed

Not to remove the second strut until inclinometer measured deflection values had
stabilized
Not to remove more struts unless measured deflection values were stabilized and within
alarm trigger levels

In addition to inclinometer measurement, diaphragm wall movement was also

monitored by the survey points strategically marked on the wall panels. Ground settlement
and surface cracks behind the diaphragm wall were also visually checked by the construction
team as daily basic.
Performance of Phase 1 Diaphragm Wall
After carrying out the comprehensive desk studies and establishment of systematic
monitoring program presented above, Phase 1 excavation work was carefully commenced.
Figure 7 shows the maximum accumulated diaphragm wall deflection at different stages of
excavation monitored by inclinometer No.1 (I-1 at East wall of Phase 1) together with trigger
levels and predicted maximum deflection profile of 3 different conditions - tender stage
design (2 temporary struts), modified design with buttress and modified design without
buttress. It can be observed from figure that measured lateral movement pattern of diaphragm
wall agreed well with that of prediction for modified design with buttress - meaning
buttress-support has significant influence on wall deflection.
Cum. Displacement , mm.
0
0
2

20
NORMAL

40

60

80

ALARM

100

ACTION

MOST PROBABLE PREDICTION

4
MOST
UNFAVORABLE
PREDICTION

Depth, m.

6
8
10
12
14

Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
Tender Stage

16
18

Alarm Trigger Level

Action Trigger Level
Prediction without buttress

20

Fig. 7. Phase 1 East Wall-Monitored diaphragm wall deflection at different stages with
trigger levels

505

Cum. Displacement , mm.

0

20

40

60

80

100

0
NORMAL

ALARM

ACTION

MOST PROBABLE PREDICTION

4
6
MOST UNFAVORABLE
PREDICTION

Depth, m.

Stage 1

10

Stage 2
Stage 3

12

Stage 4
14

Stage 5
Tender Stage

16

Alarm Trigger Level

Action Trigger Level

18

Prediction without buttress

20

Fig. 8. Phase 1 South Wall - Monitored diaphragm wall deflection at different stages with
trigger levels
Deflection profile of South diaphragm wall which braced against temporary sheet pile
wall at Phase 1 and 2 boundaries is presented in Fig. 8. As can be seen in Fig. 8, South
diaphragm wall deflection is significantly higher than that of east wall, which is likely to be
caused by the fact that South diaphragm wall is braced with more flexible sheet pile wall.
description of stages shown in the legend of Fig. 7 and 8 is summarized in Table 2.
Table 2. Description of stages shown in Fig. 7 & 8
Stage
Description
1

Excavate to 2.2m and installed temporary strut

Excavate to 4m

Excavate to 6.6m with berm

Removal of berm

Removal of temporary strut after completion of buttress

506

Fig. 9. Phase 1 excavation work in progress with historical Buddhist temple Wat Suthat in
background (Phase 2 area was being used as car parking space)
Modification of Phase 2 Bracing System
Monitoring results of the Phase 1 excavation work provided an ample opportunity to
review the design assumption, fine tune the parameters used in the analysis of the diaphragm
wall for the Phase 2 and made modification of construction sequence. The major
modifications are : (1) Removal of soil-berm at East and West diaphragm wall in shorter
duration than that of Phase 1, and (2) Using raking struts instead of horizontal strut for North
diaphragm wall a shown in Fig. 10.

HORIZONTAL STRUT

SLAB

COMPLETED PHASE 1
STRUCTURE

RAKING STRUT
SLAB

D-WALL

BERM

D-WALL

BARRETTE

-16m

BERM

-16m

-20m

-20m
EAST & WEST D-WALL

NORTH D-WALL with raker support

Fig. 10. Bracing system of Phase 2 diaphragm walls

507

Performance of Phase 2 Diaphragm Wall

Figure 11 depicts the measured deflection of east diaphragm wall. As can be
observed in Fig. 11 in comparison with Fig. 7, the maximum deflection of east diaphragm
wall in Phase 2 is larger than that of Phase 1.
The likely reasons of this observation are;
Un-strutted elapsed time for first temporary bracing in Phase 2 was longer than that of
Phase 1.
In Phase 1, horizontal struts were installed in north-south direction which temporary
kingpost columns and strut were integrated in crisscross pattern with east-west struts providing complete-support more rigid bracing system. Whereas in Phase 2, horizontal
struts were installed only in east-west direction without having crisscross pattern with
north wall having less rigid bracing system than that of Phase 1.
Cum. Displacement , mm.
0
-2

20
NORMAL

40

60
ALARM

80

100

ACTION
MOST PROBABLE PREDICTION

2
4

Depth, m.

6
MOST UNFAVORABLE
PREDICTION

8
Stage 1

10

Stage 2

12

Stage 3
Stage 4

14

Stage 5

16

Alarm Trigger Level

Action Trigger Level

18

Prediction without buttress

20

Fig. 11 Phase 2 East Wall - Monitored diaphragm wall deflection at different stages with
trigger levels
With assurance of diaphragm wall performance from monitoring results of Phase 1,
original plan of using horizontal struts for North diaphragm wall was modified by using
raking struts instead. As can be seen in Fig. 12, deflection of North diaphragm wall (with
raking strut support) is significantly higher than that of East diaphragm wall (with horizontal

508

strut support). The main reason of larger movement of north diaphragm wall is due to the fact
that it was supported only by the berm for the long period (about 52 days) before completion
of raking struts so that soil-berm became soften during the long elapsed un-strutted period.
Time-dependent deflection pattern due to softening and deformation of soft clay can be
observed in North diaphragm wall as illustrated in Fig. 13. North diaphragm wall moved
progressively toward excavation before completion of raking struts (at 52 days) as can be seen
in Fig. 14.
Table 3. Description of stages shown in Fig. 11 & 12
Stage East Diaphragm wall
North Diaphragm wall
1
Excavate to 2.2m and installed
Excavate to 2.2m at d-wall, and to 6.6m
temporary strut
with sloping berm
2
3
4

Excavate to 4m
Excavate to 6.6m with berm
Removal of berm

Removal of temporary strut after

completion of buttress

Installed raking strut

Removal of berm
Removal of raker after completion of
buttress

Cum. Displacement , mm.

0
-2

20

40

60

80

NORMAL

100
ALARM

120

140

ACTION

MOST PROBABLE PREDICTION

2
4

Depth, m.

6
MOST UNFAVORABLE
PREDICTION

8
10
12
14
16
18

Stage 1
Stage 2
Stage 3
Stage 4
Alarm Trigger Level
Action Trigger Level
Prediction without buttress

20

Fig. 12. Phase 2 North Wall - Monitored diaphragm wall deflection at different stages with
trigger levels

509

Fig. 13. View of buttress-support diaphragm wall prior to temporary bracing removal
Cum. Displacement , mm.
0

20

40

60

80

100 120 140

0
2
4

Depth, m.

6
8
10
12
14
16
18

0 days
30 days
52 day
56 days
74 days
Prediction

20

Fig. 14. Phase 2 Time dependent wall deflection of North diaphragm wall diaphragm wall
was supported only by soil-berm for 52 days before completion of raker installation.
Implementation of Contingency Plan
Since deflection of North diaphragm wall (Phase 2) approached action trigger levels,
monitoring frequency was increased and the following contingency measures were

510

implemented on site.
Poured 15cm thick 1m wide lean concrete on the top of the berm along North diaphragm
wall to provide bearing-effect
Installed additional king-post and diagonal struts attached to the raking struts to provide
more rigid support against diaphragm wall
Soil-berm was removed locally in bays followed by construction of wale beam, tie beam
and buttress as shown in Fig. 15.
TIE BEAM CONSTRUCTED IN BAYS

TEMPORARY WALER

BUTTRESS

DIAGONAL STRUT

DIAPHRAGM WALL

TEMPORARY KING-POST

CAPPING BEAM

RAKER

BERM

TEMPORARY
STEEL BEAM
PILE
CAP

TED
PLE
COM

B
SLA

Fig. 15. Perspective of North diaphragm wall constructed in bay

Movement of diaphragm wall was observed to be decreased and eventually stabilized by the
above actions. No significant ground settlement was observed in the vicinity of the North
diaphragm wall.

Fig. 16. View of raker and soil-berm support soil-berm was removed locally bay by bay

511

Fig. 17. View of buttress-support diaphragm wall after removal of temporary bracing
Time and Cost Saving from Value Engineering Options and the Application of
Observational Method
Significant cost and time saving were achieved from the value engineering option coupled
with observational method implemented for both Phase 1 and 2. The major savings were
achieved by less operation and material utilized in the following elements of temporary works.

Cancellation of 2nd level raking struts against diaphragm wall for both phases
Modification of bracing system using raking struts with soil-berm support instead of
horizontal struts for North diaphragm wall in Phase 2

Key Factors Contributed in Successful Application of Observational Method

Outcome of a through desk study at post-tender stage provided an effective value
engineering option which offered significant cost and time saving for overall construction
program. Effective and good communications between the design team and construction crew
played a key role in successful application of observational method to completion of the
project. Systematic monitoring program with clear defined trigger criteria was also the

512

important element in implementing the observational method. This case study reveals that a
thin permanent diaphragm wall coupled with effective design and construction method
supplemented by the observational method and robust monitoring program could offer a
logistically and financially attractive solution in construction of underground car park without
disturbing the environment in the prominent historical area of Bangkok.
CONTRACT 1 OF MRTA CHALOEM RATCHAMONGKHON LINE PROJECT
Observational method was applied mainly in building and structure damage assessment
in this Bangkok very first mass transit system project. The M.R.T Chaloem Ratchamongkhon
Line is the first underground mass transit system project in Bangkok developed by the Mass
Rapid Transit Authority (MRTA) of Thailand. Total length of the underground structure is
21.5km, comprises of 16 km of twin single-track bored-tunnel, a total of 4 km long18 cutand-cover stations, a twin 1.5km long cut-and-cover approach tunnel to Depot and other
associated structures. The project was commenced in late 1996 and opened for the public in
2004. The Contract No. 1, Underground Structure South - southern portion of this initial
system comprising 9 km of twin 6m outside diameter bored tunnels, nine underground
stations and cut-and-cover depot approach tunnel, was awarded to the Joint Venture BCKT
consisting of Bilfinger + Berger Bauaktiengesellschaft, Ch Karnchang Public Company
Limited, Kumagai Gumi Company Limited and Tokyu Construction Company Limited. The
project was commenced in late 1996, successfully completed and opened for the public in
2004.
One of the major challenges of the Contract No. 1was the need of stacked-alignment for
a major portion of twin bored-tunnel underneath Rama IV road, one of the busiest roads of
Bangkok, which called for a stacked configuration at 3 stations, Lumpini, Si Lom and Sam
Yan. This requirement led the contractor to design and construct the deepest underground
structures of Bangkok at Si Lom Station - four-floor stacked-station of 32m deep right under
the existing flyover and adjacent to a number of sensitive structures. Deep excavation and
bored tunneling works in congested urban environment of Metropolitan Bangkok requires a
systematic process of ground movement prediction, building risk damage assessment and
protection.
Subsoil Condition along Contract 1 Alignment
Similar to other localities in Bangkok, a typical subsoil profile along the tunnel
alignment is characterized by the alternating layers of clay and sand deposits as shown in
Figure 18. Weathered crust of 2 m thick is commonly found as the top layer. In urban areas
of Bangkok, this layer is covered by fill material. Soft to very soft, highly compressible dark
gray marine clay lies beneath weathered crust or fill. Depending on the location, this layer
extends up to 12-16 m. About 2 m thick medium clay layer can be observed between soft
clay and underlying stiff clay. Generally stiff Clay layer occurs directly underneath medium
clay and its depth goes up to 22 m. Below stiff clay layer, first sand layer of 5-8 m thickness
can be found. This first sand layer, however, is absent in some areas. Stiff to hard clay layer
underlies first sand and it is found to be about 5 m thick. Second sand layer generally occurs
at depths between 45 to 65 m.

513

Figure 18. Generalized soil profile along the tunnel alignment and stations

Overview of Ground Movement Prediction and Building Damage Assessment Process

Applied in the Project
Well accepted published literatures were used as basic reference and method in ground
movement prediction and building damage assessment. For the station boxes where
cut-and-cover methods were used for excavation, the work of Peck (1969) and Clough and
ORourke (1990) were used whereas published papers of Burland et. al. (1977) and Boscardin
& Cording (1989) were applied for bored tunnels. Building damage assessment was carried
out in 3 stages; Stage 1 Assessment, Stage 2 Assessment, and Stage 3 Assessment.
Greenfield Site assumption was adopted in Stage 1 and 2 Assessment ignoring any
influence of the stiffness of the structure and type of supporting foundation. Stage 2
Assessment was further extended by Additional Stage 2 Assessment. In Additional Stage 2
assessment, the variation of settlement trough shape and magnitude with depth were
considered in identifying the risk of building damage. In Stage 3 assessment, a detailed
qualitative assessment of structure and vulnerability of each building or structure to damage
were taken into account, coupled with detailed structural analysis in assessing the buildings
falling in risk category 3 classified by Mair et. al. (1996).

514

Summary of Instrumentation Program

Instrumentation played a major role in the construction of this first underground mass
transit system project in congestive urban area with difficult soil condition. The primary
objective of the instrumentation program was to monitor the performance of the deep
excavation and tunneling operation to ensure safe execution of the construction works and
that adjacent structures were not adversely affected. Furthermore, the instrumentation
program was established to provide feedback in application of the observational method.
Extensive sets of instrumentation were installed around the predefined influence zones of
station box excavation, tunneling and associated building, structures and utilities. In
addition to other geotechnical instruments, inclinometers, extensometers, various types of
settlement points, tile-meter, crack meter and crack gauge were installed to monitor the
ground and associated building and structures movement.

Building Condition Survey

Prior to commencement of the major works, all the existing buildings and structures
within the predefined influence zone were systematically surveyed. The primary objectives of
the surveys were;

To record pre-construction status of all buildings and structures within the influence
zone. The records included were age of building, building type, storey height,
expected foundation types, visible defects supported by sketches and photographs as
appropriate

To utilize the information collected from the building / structure condition survey in
the assessment of the buildings and evaluate sensitive structures with respect to
construction aspects and monitoring

To establish a benchmark to monitor the possible effect of construction and feedback

information in applying observational method and if necessary to implement the
building protection measure

Expected foundation types of different buildings were classified based on the height,
number of floors and age of the buildings. Building survey information was also used to
identify the types of instrumentation required.
Building Damage Assessment and Observational Method
Process of Stage-Assessment is shown in figure 20. After carrying out the initial
assessment of potential building damage, it was recognized that without applying the

515

observational method, the result obtained from assessment process could be conservative for
some buildings and could be underestimated for others. For example, the settlement estimated
in Second Stage Assessment shown in above diagram was based on surface movement. For
the buildings located in close proximity to the excavation founded on piles, it was considered
that subsurface settlement may be more critical. Stage 2 Assessment was then further
extended by Additional Stage 2 Assessment in which the variation of settlement trough shape
and magnitude with depth were considered in identifying the risk of building damage.
Settlement of the buildings due to the excavation works were predicted assuming the green
filed condition. Neglecting restraints from foundation and structure, it was assumed that
buildings follow the ground settlement trough at foundation level (estimated pile tip level).
However, if the building is supported by deep foundations, assessment result could be too
conservative.

Stage 1

Prediction and preparation of surface settlement

(Green Field Site Condition)

< 10mm

No further
assessment

>10mm

Stage 2

Determination of building risk category

(limiting tensile strain) based on surface settlement

< slight

No further
assessment

> slight

Additional Stage 2

Determination of building risk category (limiting tensile strain)

based on sub-surface settlement

< slight

No further
assessment

> slight

Stage 3

Detailed Assessment
(building structural check, consider building stiffness, Soil-structure interaction)

Figure 19. Diagram showing process of Staged-Assessment

Available methods were further reviewed to predict the surface and subsurface
settlement with the consideration of two main factors; (1) simple and practical in application
(2) enable to correlate with the predicted and measured diaphragm wall deflection. The
method proposed by Bowels (1988) was selected as it meets the required criteria mentioned
above. Bowels suggested the ground settlement induced by excavation as a function of ground
loss due to the deflection of the retaining wall. Bowel demonstrated the calculation of

516

settlements at specified distance by assuming parabolic variations of settlement within the

influence distance. Using predicted diaphragm wall deflection, surface settlement behind the
wall was computed by empirical formulas proposed by Bowels.
A simplified prediction of subsurface settlement was carried out based on the
calculated surface settlement described above. First, subsurface settlement influence line was
constructed. As shown in Figure 20, settlement influence zone is assumed to decrease with
depth from Do at the surface and zero at the wall toe. With the assumption of linear
relationship between the volume of deflected wall shape and the volume of settlement trough
at any depth within settlement influence zone, subsurface settlement at different depths were
calculated.

D-wall

Do = 2.5 HE

D O = HW + W
Surface settlement trough
volume, V To

Sio

S WO

W
Total deflected
shape volume, V O

Sub-surface settlement
trough volume, VTy
Settlement
influence zone
deflected volumn V y at depth (H w- Y)

Swy
Siy

HW

HE

Dy

S WO

4V O
DO

S io SWO *
DO
D
DY Y * O
HW
VTY VY *

VTO
VO

Figure 20. Demonstration of subsurface settlement prediction from diaphragm wall

deflection values
Based on the polynomial curve developed in Stage 2 Assessment, subsurface
settlement can be predicted by the similar approach described above. With predicted surface
and subsurface ground movement, risk of damage to the buildings were determined by two
methods (i) Risk of damage based on maximum tensile strain, the method proposed by Mair
et. al. (1996) and (ii) Risk of damage specified by angular distortion and horizontal strain
using the chart proposed by Boscarding and Cording (1989). Table 4 shows the summary of
building damage assessment, risk category obtained for different types of foundation
determined by two different methods.

517

Table 4
type

Summary of risk category obtained from two different methods and foundation

Building No. of Building

Estimated

Distance

Risk category by

Risk category by

Block

Pile Length

to edge

maximum tensile

chart of Boscarding

of wall

strain

and Cording (1989)

stories

Type

Masonry /

Without

With

Without

With

pile

pile

pile

pile

7-14m

18m

7-14m

17.5m

7-14m

32m

< 6m

2m

50m

8m

Concrete

Masonry /
Concrete

Masonry /
Concrete

Masonry /
Concrete

37

RC

The predicted degree of damage of the buildings using simplified chart is more critical
than that of maximum tensile strain method as can be seen in above table. According to the
results obtained from of simplified chart, predicted degree of damage was higher if buildings
were assumed to be with piled foundation rather than without piled foundation.
In order to justify the most probable condition assumed in building damage assessment
and to take necessary actions if monitoring results reveal most unfavorable conditions (i.e.
ground movement induced by deflection of diaphragm wall reaches maximum acceptable
limit), observational method was implemented on the followings basis.

Reviewed the building condition survey record paying particular attention to the type of
foundation and existing visible defects

Predicted the ground movement based on the calculated diaphragm wall deflection

Performed structural assessment based on the predicted ground movement (for Stage 3
assessed buildings)

Established the trigger criteria based on the magnitude of ground movement (maximum
settlement and differential settlement between foundations or structural columns)

Predefined the practical contingency plan for most unfavorable conditions where
ground movement reaches trigger levels (ground movement was derived using simple
approach based on diaphragm wall deflection described in earlier section)

Set out the instrumentation program with the consideration of above factors.

Strictly controlled the construction sequence assumed and defined in design and if
necessary changed or modified the sequence

518

Monitored the performance of retaining wall both diaphragm wall (for station boxes)and
sheet pile walls (for station entrance structures) . Compared the monitoring results with
the predicted and trigger values and reassessed

Monitored the building settlement. Compared the monitoring results with the predicted
and trigger values and reassessed

Implemented the contingency measures if monitoring result reaches action level of trigger
values

Monitoring results obtained from early stage of the project were used to reassess the potential
damage of the buildings and structures. Some buildings of critical condition were put in Stage
3 assessment requirement and assessed by structural analysis using detailed structural survey
information as basic data. By simulating the differential settlement of the building predicted
from Additional Stage 2 Assessment, development of moment and shear in the main structural
members were determined with the use of structural analysis software. Computed moment
and shear were assessed in comparison with the capacity of the structural members estimated
from collected building condition survey and detailed structural survey data. The process
involved in structural assessment can be summarized below.

Factor of safety of each structural members were determined base on predicted

differential settlement of the buildings.

A series of assessment were performed to determine the critical differential settlement

causing structural instability of the members or differential settlement at limiting factor of
safety.

Trigger values were established based on above results.

Based on the trigger values and the actual building condition and response in conjunction
with monitoring data, the most sensitive structural members were protected, mainly by
strengthening approach.
The following assumptions were made in the conservative approach of Stage 3
assessment.
Buidling movement is influence by excavation induced settlement of both main station
box and associated underground structures (e.g. Entrance Building)

Building settlement was casued by subsurface settlement at foundation level as buildings

are supported by piled foundation.

No allowance was made for structural stiffness of the building.

No allowance was made for possible effects of soil-structure interaction.

A summary of risk category determination in Stage 3 Assessment based on the predicted

and monitored settlement of the building adjacent to station excavation work is shown in the
Table 4 below.

519

Table 4 Summary of risk category determination based on Boscarding and Cording

(1989) based on prediction and actual monitoring of settlement
Horizontal Strain

Data Taken
From

Sh1

Sh2

(mm) (mm)
Settlement
Prediction
Actual
Monitoring Data

L
(m)

Angular Distortion

31

24

18.6 0.0004

31*

24*

18.6

0.0004
*

Sv1

Sv2

(mm) (mm) (mm)

L
(m)

Risk

Degree of

Category

Damage

48

23

25

18.60

0.0013

slight

53

14

39

18.60

0.0020

slight

*Predicted horizontal movement was used in computing horizontal strain for actual
miniotring data.
From the above table, following conclusions may be made.
Predicted angular distortion broadly agrees with that of actual monitoring data. Hence,
settlement prediction and assumed effective depth of piled foundation is fairly reasonable.

The settlement of actual monitroing data is slightly higher than that of prediction. This
may be due to the additional settlement caused by diaphragm wall panel construction
which influence zone might have extended to 15m from the diaphragm wall (equivalent
distance to bottom of soft clay level ).

Damage Assessment on Sensitive Structures

The most critical structure required for Stage 3 Assessment were MWA main water
tunnel as tunnel alignment crosses the water tunnels at two critical locations the first near
Rama IV and Rachadaphisek road intersection and the second at Rama IV and Siphraya road
intersection near Sam Yan station. The position of water tunnels in relation to MRT tunnels
drawn in the subsoil profile at relevant locations are presented in Figure 21 and 22.
According to the available information, MWA tunnel at the first crossing is of 2.5m
internal diameter steel lining with 150mm segmental ring and at the second crossing MWA
tunnel is of 2.5m diameter segmental ring with lightly reinforced concrete lining.
In preliminary stage 3 assessment, based on the predicted subsurface profile using the
method described in published paper Ground Movement Prediction and Building Damage
Risk-Assessment for the Deep Excavations and Tunneling Works in Bangkok Subsoil, Zaw
et. al (2006) e, structural assessment of water main was carried out and results were compared
with structural capacity of water tunnel calculated from the material properties indicated in
the construction drawings provided by MWA. As bending moment induced in the water tunnel
was considered critical parameter in assessing the damage which is governed by both
maximum settlement and curvature of the water tunnel further assessment was proposed
coupled with observational approach.

520

To carry out back-calculation using monitoring data obtained from the settlement arrays
(settlement points, extensometer data) at first 3 locations before reaching to MWWA
water tunnel near Rama IV Rachadapisek intersection, to determine the volume loss
and relationship between surface and subsurface settlement as well as best estimation of
subsurface K value and subsurface settlement profile at water tunnel level.

To perform beam-on-elastic foundation analysis using the subsurface settlement profile

obtained from predicted subsurface settlement bending moment likely to occur in
MWWA water tunnel.

To compare the result obtained from above item with allowable values.
To establish the final trigger values based on above comparison.
To confirm the settlement profile from actual monitoring of last array is within the alert
level of final trigger value.

Figure 21. Position of MWA water tunnel in relation to MRT tunnels (clear distance are
3.38m and 3.18m for North and South bound respectively) at the first crossing

521

Figure 22. Position of MWA water tunnel in relation to MRT tunnels (clear distance are 1.5m
and 2m for North and South bound respectively) at the second crossing
For the first location of MRT-MWA tunnel crossing, target ground loss from tunnel
operation was set at 0.5%, whereas ground loss based on allowable movement curvature of
water tunnel derived from predicted pattern of subsurface settlement was 0.65%.
Back-calculated ground loss derived from the monitoring data of rod-extensometer placed at
the crown of MWA water tunnel was within target value. In addition to the preliminary
assessment based on prediction of subsurface settlement using simple empirical formulas, a
detailed soil-structure interaction was analyzed using 2-dimensional FEM considering
behavior of tunnel boring by earth pressure balance (EPB) method. Stress relief from the
effect of cutter-head, tail void and over-cut from EPB tunnel boring was considered in the
analyses. Adding ground movement due to consolidation caused by disturbance from tunnel
operation surface and subsurface settlement were predicted by finite element analysis.
Detailed parameters and assumption made in analysis was presented in the published paper of
Sakai & Sugden (2000).
After the successful passing of both tunnels at the first MWA crossing, another trial area
tunneling was planned to further determine the ground movement associated with various
tunneling parameters well before reaching to more critical crossing after Sam Yan station.
Tunnel section between Silom and Samyan station, in which soil profile similar to that of
second crossing was selected and comprehensive instrumentation program was established to
monitor surface and subsurface settlement as well as horizontal movement of the ground at
depth.
The observed ground movement induced by tunneling works were within predicted
values. Well organized plan, systematic approach of observational method to achieve control

522

values backup by extensive instrumentation and close co-operation among relevant personnel
were the key factors contributed to successful completion of EPB tunneling particularly in the
most sensitive points at MWA water tunnel crossings. The monitored subsurface settlement
data from extensometer and predicted settlement trough of assumed ground loss 1% ground
settlement profile of the first MWA tunnel crossing is presented in Figure 23.

Distance (m)

Subsurface Settlement (mm)

10

20

30

40

50

60

70

80

90

-10

Settlement _ NB Tunnel
-30

Settlement_SB Tunnel
Total Settlement

-50

Figure 23 Predicted subsurface settlement trough of assumed ground loss 1% calculated

from empirical formula and actual subsurface settlement (indicated by dark-circle) observed
from rod extensometer first MWA tunnel crossing
Key Factors in Successful Application of Observational Method
The M.R.T Chaloem Ratchamongkhon Line, one of the most modern underground mass
transit systems in the region was successfully completed and has been in operation for over 2
years. All deep excavation works and tunneling were successfully completed without any
significant damages to the adjacent buildings and structures. Well organized plan, systematic
approach of observational method to achieve control values backup by extensive
instrumentation and close co-operation among relevant personnel were the key factors
contributed to successful completion of the project. The purpose of this paper is to provide a
set of technical guidelines for ground movement prediction and building damage risk
assessment of deep excavation and tunneling works in Bangkok subsoil and application of
observational method which may partly serve as a reference for the future underground mass
transit system projects in urban area of Bangkok. Successful completion of the first MRT
project in Bangkok marked the practicality of well-executed deep excavation works and
tunneling by EPB method coupled with observational method in the sensitive urban
environment.

523

REFERENCES
Boscarding, M.D. and Cording, E.G. (1989),

Building Response to Excavation-Induced

Settlement. Journal of Geotechnical Engineering, ASCE, 115;1;1-21

Bowels J.E (1988) Foundation Analysis and Design, 4th Edition, McGRAW-Hill International
Book Company, Singapore, 1988
Burland, J.B., and Wroth, C.P. (1974), Settlement of Buildings and Associated
Damage.
SOA
Review,
Conference on
Settlement
of
Structures,
Cambridge,
Pentech Press. London, pp. 611-654
Burland J.B., Broms J.B. and de Mello VFB (1977), Behavior of Foundations and Structures,
SOA Report Session 2, Proceedings of 9th International Conference, SMFE, Tokyo,
2:495-546
Clough, G.W., and ORourke, T.D., (1990), Construction Induced Movements of
In-situ
Walls, ASCE Geotechnical Special Publication No. 25, pp. 439-470
Clough C.W. and T.D. ORourke (1990), Deep Excavation and Tunneling, 7th ICSME,
Mexico City
JV BCKT, (1997) Geotechnical Interpretation Report (GIR) for MRTA Initial System UGS, Bangkok
Mair, R.J., Taylor, R.N., and Burland, J.B., (1996), Prediction of Ground Movements and
Assessment of Risk of Building Damage. Geotechnical
Aspects of Underground
Construction in Soft Ground, pp. 712-718, Balkema, Rotterdam
OReilly M.P and New B.M (1982), Settlements above tunnels in United Kingdom their
magnitude and prediction, Tunneling 82, pp. 173-181, London, IMM
Peck R.B. , (1969) Deep Excavation and Tunneling in Soft Ground, 7th ICSME, Mexico City,
State of the Art 3, 225-290
Peck, R. B. [1969]. Advantages and limitations of the observational method in applied soil
mechanics, Geotechniques 19, No.2, pp. 171-187
Powderham, A. J. [1996]. An overview of the observational method : development in cut and
cover and bored tunneling projects, The Observational Method in Geotechnical
Engineering, The institute of civil engineers, Thomas Telford Publishing, London, pp.
7-24
Zaw Zaw Aye , Dhiraj Karki and Christian Schulz [2006]. Ground Movement Prediction and
Building Damage Risk-Assessment for the Deep Excavations and Tunneling Works in
Bangkok Subsoil, Proceedings of the International Symposium on Underground
Excavation and Tunneling, Urban Tunnel Construction and Protection of Environment,
EIT, Bangkok, Thailand, , pp. 281-297

524

Proceedings of the 16th Southeast Asian Geotechnical Conference

8-11 May 2007, Kuala Lumpur, Malaysia

16th Southeast Asian

Geotechnical Conference
Edited by:
Ooi Teik Aun, Yee Yee Weng and Ting Wen Hui

Ground Movements Associated

Underground Excavations of the
First Bangkok MRT Line

with

Noppadol Phienwej and Attasit Sirivachiraporn

Asian Institute of Technology, Pathumthani, Thailand

Suchatvee Suwansawat

King Mongkut Institute of Technology-Lad Krabang, Bangkok, Thailand

Zaw Zaw Aye

SEAFCO Public Co., Ltd. Bangkok, Thailand

525

the

Ground Movements Associated with the Underground Excavations of the

First Bangkok MRT Line
N. Phienwej & A. Sirivachiraporn

School of Engineering and Technology, Asian Institute of Technology, Pathumthani, Thailand

[email protected]

S. Suwansawat

Department of Civil Engineering, King Mongkut Institute of Technology-Lad Krabang, Bangkok, Thailand
[email protected]

Z. Z. Aye

Seafco Public Company, Bangkok, Thailand

[email protected]

Abstract: Works of the first Bangkok Underground MRT project involved the largest diameter shield tunneling and the deepest diaphragm wall excavation ever attempted in Bangkok subsoils. The design-built contracts required the design and construction to avoid
damages to existing buildings, thus the Observational Method was adopted and large amount of instrumentation for monitoring of
ground movements was integrated as important component of the works. The ground and building movement data associated with the
shield tunneling and station excavation are interpreted and the findings are briefly presented. Tunnel volume loss in the EPB shield
tunneling was typically in the range of 0.5-2.5% while the maximum ground surface settlement above the tunnel was typically 20-60
mm. The magnitude of ground movement varied with the face pressure used in the shield driving. The settlement trough width parameter was i = 0.5-0.6 Z for tunneling in clay layers and < 0.4Z for tunneling in the underlying sand layer. Maximum lateral deflection in
braced mode of the diaphragm wall in the station excavation was in the range of 0.05-0.4% of the excavation depth. A similar range
was observed for the maximum ground surface settlement adjacent to the wall. There were only limited cases of damages to adjacent
buildings, which occurred only to small sized buildings founded on short piles.

INTRODUCTION

The Bangkok MRT Initial System Project (ISP), the Chaloem

Ratchamongkhol Line, is the first underground MRT of Bangkok.
Total length of the line is 21.5 km, comprising 16 km of twin
single-track tunnels, 18 cut-and-cover stations and twin 1.5-kmlong cut-and-cover approach tunnels to Depot. The project was
commenced in late 1996 and opened for public use in 2004. The
alignment mostly followed congested city roads. Construction
works of the underground structures were carried out in two fast
track design-build contracts, the South and North Sections. The
tunnel tubes were bored at 6.3-m diameter by 8 EPB TBMs while
the stations were excavated with an aid of diaphragm walls following the top down construction method. The depth of the tunnels was between 8 and 25 m below ground level. The general
layout of the twin tunnels is in side by side configuration (mostly
with clearance between tunnel walls of 1.0-1.5 times the excavated tunnel diameter (5.0-9.0 m) but the smallest clearance was
as smaller as 2.6 m). The vertically stacked configuration of the
tunnels was adopted for the first 4.5 km of the alignment from the
terminal station at the main intercity railway station owing to obstruction from foundations of elevated roadways and existence of
water main tunnels. Consequently, the stations in the section
were excavated at depth as deep as 30 m. Two stations are located beneath existing intersection overpasses that necessitated
underpinning works. Other design and construction issues involve consideration of the tunnels passing close to concrete-

527

lined water main tunnels at two locations, and removal and underpinning of piled foundations of four canal bridges.
Prior to the start of the MRT construction, there had been numerous shield tunneling and deep excavation works for both
building construction and utilities development. However, underground works of the MRT project marked the largest tunnel diameter and the deepest excavation ever attempted in Bangkok
subsoils. Thus great caution must be paid to the design and implementation of the works for safety assurance of the works and
success of the project development.
Control of ground movement and preventing damages to adjacent structures from the excavation was the key for the success
for the project. The Observation Method design was adopted for
that purpose. A comprehensive instrumentation program for
monitoring of ground movements and responses of concerned
structures was carried out as an important component of the contracts. The monitoring data served as indices for control of works
and safety assurance during construction as well as allowed for a
better understanding on the ground movement behavior induced
by EPB shield tunneling and deep excavation with diaphragm
wall in the stratified Bangkok soils.
2

SUBSOIL CONDITIONS

Bangkok is situated on the low-lying plain. The subsoil within

the first 20-30 m depth generally consists of layers of soft clay,
stiff to hard clay, and sand. Below are alternating layers of stiff to

hard clay and dense sand existing to a great depth. The subsoil
condition is relatively uniform through out the city area. The soft
clay layer generally extends from the ground surface to a depth of
12 to 15 m. The soft clay which is known as Bangkok soft clay
has high water content (70-120%). The underlying stiff clay layer
has much better engineering properties than the soft clay and is
an ideal medium for tunnelling within. Below the stiff clay layer
is a layer of fine to medium, dense to very dense, silty to clayey
sand. The sand is normally saturated but the piezometric pressure has been considerably drawn down as a result of excessive
deep well pumping in the past 60 years. Presently, the piezometric head in the sand layer is mostly at 21-24 m below the ground
surface. The condition is beneficial to tunnelling and deep excavation works. Fig. 1 shows subsoil stratigraphy along the MRT
alignment. Although most of the tunnel alignment is placed in the
stiff clay layer, some shallow sections have roof in the overlying
soft clay and some of the deep alignment was excavated in the
underlying sand layer.

TUNNEL EXCAVATION

Eight earth-pressure-balanced (EPB) shields were used in the 16

km long twin tunnel excavation- 4 Kawasaki shields in North
Contract and 2 Kawasaki +2 Herrenknecht shields in South Contract. The tunnels were supported with 300-mm-thick bolted concrete segmental lining ring. To achieve effective performance of
the EPB shield operation, conditioning additives were added to
excavated soils in the EPB chamber. The additives were bentonite slurry, polymer, or organic foam. The four shields in each
of the contracts took about 14-15 months to excavate the tunnel
length. The average advance rate of excavation of each shield is
60-70 m/week. Despite some major downtime due to breakdowns
of main parts, all of the shields performed very satisfactory in all
types of soils. The rates of excavation of the EPB shields in the
MRT project in all types of soil were essentially similar except
for excavation in the stiff clay that the rate could be smaller due
to higher cohesion and stickiness.

Subsurface Condition of North Section

DEPTH BELOW GROUND LEVEL (m)

THIAM RUAM MIT

PRACHARAT BAMPHEN

SUTTHISARN

RATCHADA

LAT PHRAO

PHAHONYOTHIN

MO CHIT

KAMPHEANG PHET

BANG SU

10

10

20

20

30

30

40

40

50

50

60

60
20+000

21+000

22+000

23+000

24+000

25+000

26+000

FILL

DENSE CLAYEY SAND

BANGKOK SOFT CLAY

DENSE SAND

STIFF TO HARD CLAY

Fig. 1 Subsoil stratigraphy along the MRT alignment

528

27+000

28+000

29+000

30+000

STATION
TUNNELLING HORIZON

STATION EXCAVATION

Excavation of the 18 stations was made by the top-down construction using 1-m-thick concrete diaphragm walls. Tips of the
wall usually extend to the depth of 35 - 40 m below the ground
surface (Fig. 2) to satisfy load bearing and uplift resistance requirements as well as to provide ground water cutoff in the first
sand layer. The distinct difference in depth of embedment of the
wall below the excavation level between the two contracts was
partly due to the difference in the adopted design criteria. Construction period of each station from utility relocation to architectural finishes was usually around 45 months- 6 months of piling
and diaphragm wall construction and 18 months of excavation
and casting of station floors and walls.
OBSERVATIONAL METHOD

The contract required that the design and construction methods limit ground movements from tunneling and station excavation to avoid damage to adjacent structures and buildings.
The Observational Method was then adopted for the control of
work. It involved the followings.
-

Three stages of assessment on the effects of the ground

movement on structures depending upon the results of
each stage.
In stage one, prediction of the green filed ground surface
settlement using approved empirical prediction methods
was made.

Building condition surveys of all structures within the

zone of influence determined by the first stage. Any structures with the predicted ground settlement less than 10
mm need not be subject to further assessment.

Structures located in the zone of surface settlement were

checked for potential damages according to the limit tensile strain criteria (Boscardin & Cording, 1989, Burland,
1995). Structures falling in damage category within slight
level should be acceptably safe thus need no further assessment.

Stage 3 assessment of structures exceeding slight level of

damage need stage assessment which involved detailed
structure analysis considering soil structure interaction).
Depending on the results of the analysis, pre-excavation
protection or preventive measures as well as contingency
plan to be implemented during construction may be
needed.

Monitoring of settlements arising from works must be carried out. Comprehensive instrumentation program was required as a part of the contract.

Relatively few incidents of damages to adjacent structures

occurred. The most significant ones were damages to small
buildings on short piles. No cases of damages to tall buildings
were experienced.
6

GROUND MOVEMENTS IN TUNNELLING

6.1 Surface Settlement

Monitoring of ground surface settlements over the tunnel alignment was made at 50 m intervals. The data showed that the

529

Fig, 2 Depth of station excavation and tip depth of diaphragm

walls
Station, Km
10.000 12.000 14.000 16.000 18.000 20.000 22.000 24.000 26.000 28.000 30.000
0.0
-20.0
-40.0
Settlement, mm

-60.0
-80.0
-100.0
-120.0
-140.0
-160.0
-180.0

Stacked Tunnels

Side-by-SideTunnels

Fig. 3 Maximum ground surface settlement along the tunnel

alignment

Fig. 4 Maximum surface settlement versus shield face pressure

maximum settlements were in the range of 5 mm to 120 mm,
mostly between 20 mm and 60 mm (Fig. 3). Larger settlement
tended to occur in the break-ins and break-outs areas of the tunneling near the walls of the station boxes. The largest magnitudes
were recorded in the early leg of tunneling of the North Contract
and in the areas of the vertically-stacked tunnels in South Contract where the tunnel were bored in sand layer below drawndown ground water level.
Studies were made to examine influencing factors controlling
the magnitude of the ground movement in the EPB shield tunneling (Suwansawat, 2002, Timpong, 2002, Tavaranun, 2003, etc).
It was found that the most important factor seems to be the applied face pressure in the front chamber of the shield machine.
The face pressure played an important role in maintaining stabil-

ity of the excavation and minimizing ground movement. Fig. 4

shows the plot between the observed maximum surface settlement and the face pressure in the project. The plot clearly shows
the influence of the applied face pressure on the magnitude of
ground settlement. The lower the face pressure used, the larger
the settlement occurred. In case of a very high face pressure,
heave occurred at the ground surface. On the other hand, when
the shields were operated at very low face pressures (<100 kPa),
large ground movement occurred. The effect of the face pressure
was observed in all soil settings except for the case of the full
face in sand layer. No trend was found between the face pressure
and surface settlements when the tunnels were excavated entirely
in sand layers.
Out of the 39 locations of settlement trough monitoring arrays,
19 arrays showed the observed trough shapes reasonably fitted
with the Gaussian curve (Peck, 1969), Figs. 5(a) and 5(b). The
tunnel ground loss was mostly in the range of 0.5-2.5%. In difficult sections, the ground loss was up to 3.5%.

tion, the trough expanded and shifted toward the second tunnel.
However the centerline of the trough was normally skewed toward the first excavated tunnel.
In the Bangkok MRT tunneling, the shifted distance of the
centerline of the trough from centerline of the twin tunnels (a)
observed at 24 monitoring arrays ranged from 0.03-0.58 of the
distance between the two tunnel centerlines (d). The mean a/d is
0.16. The width of the settlement trough observed after the excavation of the twin tunnels in the project was about 6 to 12 m for
the vertically-stacked tunnels and 10 to 17 m for the side-by-side
twin tunnels.
A plot is made between relationship between the settlement
trough width parameter (i) and tunnel depth (z), Fig.7. Most of
the data points fall within the range of i = 0.5Z to 0.6Z for tunnels located in both the stiff clay and soft clay layers and i < 0.4Z
for tunnels in sand layer. This is in agreement with the observation of OReilly & New (1982) for tunneling in UK.
Ground settlement array no. 26-AR-001
Zone 26
Ratchada - Lat Phrao

Second tunnel
excavation (NB)
ground loss 0.61%

NB

SB

-20 m.

Stiff clay

Very Stiff clay

-20 m.

NB

Stiff clay

First tunnel
excavation (SB)
ground loss 0.43%

Soft clay

Soft clay

SB

Sand

10.5 m.

(a) Side-by-side Tunnels

Fig. 6 Effect of presence of piled foundation on shape of settlement trough
Soft clay

40

Second tunnel
excavation (NB)
ground loss 0.80%

NB

SB

Stiff clay

-26 m.

Single tunnel

35

-15 m.

Twin tunnel

i = 0.3Z

i = 0.4Z

i = 0.5Z

i = 0.6Z

30

Very Stiff clay

Depth, Z (m)

First tunnel
excavation (SB)
ground loss 0.69%

Dense sand

25
20
15
10

(b) Vertically stacked

5
0

Fig. 5 Surface settlement trough over the MRT twin tunnels

Settlement troughs having shapes not to follow the Gaussian

function were normally affected by existence of adjacent buildings or flyovers. The existing piled foundations within the trough
zone obstructed the ground movement and caused an unsymmetrical trough shape (Fig. 6).
Maximum surface settlement occurred above the centerline
of first tunnel after the shield passage and normally the Gaussian
shaped trough was observed. For the side-by-side twin tunnel
case, once the boring of the second tunnel was made at the sec-

10

15

20

Trough width parameter, i (m)

Fig. 7 Settlement trough width parameters (i) versus depth

6.2 Subsurface Ground Movement
Monitoring of subsurface ground movements was made by means
of borehole extensometers and inclinometers installed at a number of sections along the tunnel alignment. The layouts of the instruments in the two contracts are shown in Fig. 8.

530

Extensometer installation in
South contract
Extensometer

cal/empirical approach (e.g. Verruijt & Booker, 1996; Loganathan & Poulos, 1998; etc.) and the numerical approach (e.g. FEM
PLAXIS, FDA FLAC, etc.) were investigated (Tavaranum,
2003). For the first approach, it was found that the method of Loganathan & Poulos (1998) method generally can give reasonable
fits to subsurface settlement pattern with field data for both single
and twin tunnels cases (8 out of 12 sections). The gap parameter,
g, from back calculation was in the range of 20-60 mm (Timpong, 2002). Reasonable prediction was also offered by the Verruijt & Booker (1996)s method. However, they could not sufficiently give reasonable prediction of lateral ground movement for
cases of excavation with high face pressures, where outward
ground movement occurred (6 out of 12 sections).

Extensometer installation in
North contract

Inclinometer
casing

Extensometer
Soft Clay

Inclinometer
casing

Soft Clay

Stiff Clay

Tunnel

Stiff Clay

3 - 6 m.

Tunnel

Fig. 8 Typical layouts of subsurface instruments for shield tunneling

6.2.1 Subsurface settlement
For areas above the tunnel centerline, the magnitude of subsurface settlement after the tunnel passage generally increased with
depth toward the tunnel roof (Fig. 9a). For the settlement to the
side of the tunnel wall, it was generally observed that the largest
settlement occurred in the soft clay layer near the interface with
the underlying stiff clay (Fig. 9b). The settlement in the stiff clay
along the instrument tube decreased with depth as the measurement points falls outside the movement zone. At some sections, a
small heave was measured below the tunnel level.
Long term ground settlement behavior for the EPB shield tunnel excavation was examined. The subsurface settlement data
showed that after the shield passage, significant increase in the
settlement in long term was only observed in the soft clay layer.
In some cases, the amount of the increase in the settlement 3-4
months after the shield passage was almost 100 percent of the
short term settlement (Fig. 10). On the other hand, the long term
increase in the settlement in the stiff clay layer near the tunnel
roof was generally not significant.

Fig. 9(a) Subsurface settlement above tunnel centerline

6.2.2 Lateral displacement

Fig. 9(b) Subsurface settlement to the side of tunnel wall

E xte nso me te r Mo nito ring R e sult
N o. 2 5 -IE -00 3 Sutthisan - Ra tc ha da
Ela pse d time (da ys)
0

20

40

60

80

100

120

-50
-40

Settlement (mm)

The lateral displacements of ground beside the tunnel were observed at 12 locations along the alignment. The instruments were
placed at distance of 0.20-2.6 m from the tunnel wall. Most of the
measurements showed a similar pattern of the lateral ground
movement which can be divided into two different depth zones,
i.e. above and below the level of the tunnel arch. Above In the
upper zone zone, the ground displaced towards the tunnel with
the maximum displacement Fig. 6 Effect of presence of piled
foundation on shape of settlement trough occurring in the lower
area of the soft clay layer. The inward displacement within this
depth zone continued to increase after the passage of the shield
tail.
In the lower zone, two patterns of the lateral ground displacement were observed. In most cases,, inward movement toward
the tunnel was observed. The magnitude was smaller than those
occurring above the tunnel roof. At few sections where high face
pressure was used in the shield tunnelling, the ground showed
outward movement (Fig. 11). The movement ceased after the
passage of the shield tail. This clearly indicates the effect of the
applied face pressure in the EPB shield tunneling in Bangkok
soils.
Studies were made to investigate applicable methods for prediction of the ground in shield tunnelling. Both the analyti-

-30
-20
-10
Soft clay

25-M E-3/08

-2.02 m.

25-M E-3/07

-5.15 m.

25-M E-3/06

-8.32 m.

25-ME-3/05 - 11.45 m.

10

-15 m.

20

-18.5 m.

30

Stiff clay

Shield passing

6 .24 m.

25-ME-3/04 - 14.58 m.
25-ME-3/03 - 17.71 m.

Fig. 10 Changes in subsurface settlement with time

531

140

of the Bangkok subsoils for the analysis. The results show that
for the soft/medium clay the modulus, Eu should be around 500
Su, where Su is undrained shear strength. For other soil layers
underlying the soft clay, the values are:
- 1st Stiff Clay
Eu = 700 N
- Clayey Sand & Sandy/Silty Clay Eu = 900 N
- 2nd Hard Clay
Eu = 1600 N
- 3rd Hard Clay
Eu = 2500 N

Inclinometer No. IN-T7-04

(Phetchaburi - Sukhumvit stations
0

Made ground

Avg. face pressure = 100 kPa

Avg. grouting pressure = 2.5 bar
C.L. = 18.8 m
2.35 m from the inclinometer

-5
Soft clay
Zone 1

Zone 2

SOUTH CONTRACT

Zone 3

Maximum Lateral Wall Movement/Excavation Depth, dHmax/H

-20

-25

0.0

Clayey sand
A. 2D before the shield face
B. 1D before the shield face
C. at the shield face
D. 0.5D behide the shield face
E. at the shield tail
F. 2D behide the shield face
G. 3D behide the shield face
H. 4D behide the shield face

Very stiff clay

-4

-2

1.0

1.5

2.0

5
10
15
20
25
30
35
Cantilever mode (H=1.6m-4.0m)
Braced mode-stiff soil (H=12.4m-16.7m)

Lateral Deformation (mm)

Fig 11 Lateral displacements beside tunnel (After Suwansawat,

2002)
7

0.5

Stiff clay
Excavation Depth, H (m)

-15

GROUND MOVEMENTS FROM STATION EXCAVATION

Ground movements associated with the excavation of the 18 stations of the project were closely monitored with systematic instrumentation programs. The monitoring data provided useful information to capture the typical characteristics of lateral ground
movement and settlement in the areas adjacent to the controlled
deep excavation using diaphragm wall in Bangkok subsoil. The
station boxes were generally 230 m long by 25 m wide with excavation depths of 20-30 m. Hooi (2002) conducted comprehensive review and interpretation of the monitoring data. The results
showed the ratio of maximum lateral wall displacement to the excavation depth at various stages of excavation as shown in Fig.
12. For the cantilever mode occurring in the first stage of excavation (H = 1.6 m 4.0 m), the dHmax/H = 0.20 - 1.60 %. For the
braced wall mode, the ratio dHmax/H = 0.05 0.60% while excavating in soft clay layer (H = 6.5 m 10.8 m). The wall deflection in the braced wall mode while excavating in the stiff clay or
sand layers (H = 12.4 m 32.6 m) showed dHmax/H = 0.05
0.40%. The observed ground settlement adjacent to the excavation walls is summarized in Fig. 13. The ratio of the maximum
settlement to excavation depth is high in the first stage cantilever
excavation mode. After the excavation advanced down to the
braced mode in either the soft clay layer or the underlying stiff
clay layer, the ratio of the maximum settlement to excavation
depth (dVmax/H) was in the range of 0-0.3%. The observed
width of settlement zone (D) is large, up to 7 D/H for braced
mode with 4 D/H for braced mode with excavation base soft clay
and reducing to 4 D/H in deeper stiff clay or sand. The influence
zone of the diaphragm wall excavation is smaller than that suggested by Peck (1969) for braced excavation using soldier
piles/sheet piles. Since the FEM design analyses used for all station excavations of the MRT followed linear elasto-plastic soil
property, back-analyses based upon the monitoring data were
made to determine the appropriate corresponding modulus values

Braced mode-soft soil (H=6.5m-10.8m)

Braced mode-stiff soil (H=20.0m-32.6m)

Fig. 12 dHmax/H versus depth of excavation (H)

Settlement/Maximum excavation depth,

dVmax/H

Depth (m)

-10

NORTH CONTRACT
Distance from excavation/Maximum excavation depth, D/H
2
3
4
5
6
7
8

10

0
0.1
0.2

A
B

0.3
0.4

0.5
I

II

III

0.6
Cantilever mode (H=1.6m-4.0m)

Braced mode-soft soil (H=6.5m-10.8m)

Braced mode-stiff soil (H=12.4m-16.7m)

Braced mode-stiff soil (H=20.0m-32.6m)

Modifed Category A, B & C

Peck's Category I, II & III

Fig. 13 dVmax/H versus depth of excavation (H)

8. BUILDING RESPONSES
Along the alignment, there were 339 buildings and structures located in the zone of influence. These structures were subjected to
a detailed building condition survey. Instruments were installed
to monitor movements and cracks that may be induced during
and after the passages of the shields. With the existence of the
soft clay layer, buildings constructed in Bangkok need to be
founded on piled foundations. Different length and size of piles
have been used depending on the size and age of the buildings.
Generally, the depth of pile tips can be categorized as shown in
Fig. 14.
The monitoring data of all the instrumented buildings (mostly
reinforced concrete frame types) were evaluated and the broad
picture of the building response to the tunnel excavation could be
depicted by the summary plots of building settlements versus
number of storey of the building and distance from the tunnel excavation (Figs. 15 and 16).
It clearly shows that buildings on deep piles were less influenced by the tunnel excavation (Types 4 Buildings: settlements <
10 mm). For smaller buildings (Types 2 & 3), the settlement
could be larger (up to 30 mm). For buildings on short piles (Type

532

1), settlement could be excessive. Fig 15 shows that significant

building settlement (> 20 mm) only recorded for buildings located within distance of 30 m from the tunnel centerline.
> 10 storey
building
4 to 10 storey
building
2 to 3 storey
building &
shophouse
Old building
and house
0.00 m.
Made ground
<6m

-1.5 m.

Soft clay

6 to 14 m

-15 m.

1st Stiff clay

14 to 30 m.

-22 m.

15 to 22 m

-40 m.

Fig.14. Categories of building and its foundation in Bangkok

2. Maximum ground surface settlement above the tunnels

ranged from 5 mm to 120 mm, mostly 20-60 mm. The corresponding volume loss in the tunnel excavation was 0.52.50%. Tunneling at in water bearing sand layer tended to
give the largest ground movement.

Building settlements vs distance from tunnel

90

Buildings on foundation
Type 1
Type 2
Type 3
Type 4

building settlement (mm)

80
70
60
50

3. Most of the surface settlement trough could be fitted by

Gaussian function. The trough width parameter (i) is approximately between 0.4Z and 0.6Z for tunnels in clays and
less than 0.4Z for tunnels in sands.

40
30

4. The long-term settlements in soft clay layer above the tunnel

could be significant in some sections but not for the stiff clay
layer.

20
10
0
0

10

20

30

40

50

60

70

5. The observed subsurface settlements and lateral displacements could generally be reasonably fitted with the prediction
by the analytical methods of Verruijt & Booker (1996) and
Loganathan & Poulos (1998). However, the prediction was
poor for lateral ground movements in case where lateral outward ground movement occurred.

80

distance from tunnel (m.)

Fig. 15 Building settlement versus distance from tunnel

Building settlement vs No. of storey

6. Both maximum lateral wall deflection and maximum ground

surface settlement near the station excavation using braced
diaphragm walls showed similar ranges of movement. The
upper bound magnitude is around 0.3-0.4% of the excavation
depth.

90
80

building settlement (mm)

CONCLUSIONS

1. EPB shield tunneling for the first Bangkok MRT was successfully completed with the average advance rate of each
shield of 60-70 m/week. Similarly, the deep excavation using diaphragm wall of the 18 stations were completed without major difficulties.

1st Sand deposit

Pile length > 22 m

ential settlement/distance between the two reference points)., The

observed tilts of these buildings were mostly smaller than 1:1000
(0.001) which was the alert level set forth for the control of work
in the Bangkok MRT project. According to Burland (1995) a
(masonry) building experiencing a maximum tilt of 1:500 (0.002)
and a settlement of less than 10 mm has negligible risk of any
damage. In fact, very few cases of damages of the existing buildings due to shield tunneling occurred in the project. These were
limited to old 1-3 storey buildings founded on short piles (probably with tips within soft clay layers). All of the tall buildings
were not damaged by the tunnel excavation even though the tunnel was excavated right next to the piles. A study is currently
made to evaluate the influence of the shield tunneling on the load
carrying capacity and integrity of the piles.

70
60

7. Settlement of adjacent buildings induced by ground movement from the MRT excavation showed clear correlation with
building heights which reflects depth of piled foundations.

50
40
30
20

REFERENCES

10
0
0

10

20

30

40

50

60

No. of storey

Fig. 16 Building settlement versus number of storey

The measured differential settlements of the buildings were
calculated and expressed in term of settlement ratio or tilt (differ-

533

Boscardin, M.D. & Cording, E.J. (1989), Building Response to

Excavation-Induced Settlement, Journal of Geotechnical Engineering, ASCE, 115;1;1-21.
Burland, J.B., 1995. Assessment of risk of damage to buildings
due to tunnelling and excavation, Proc Int. Conf. Earthquake
Geotech. Engrg., IS-Tokyo 95.

Hooi Kah Yeng 2002. Ground Movements Associated with Station Excavations of the First Bangkok MRT Subway. M.S.
Thesis, AIT, Bangkok.
Loganathan N. & Poulos H.G., 1998. Analytical prediction for
tunneling-induced ground movement in clays. J. Geotech.
Engrg ASCE, 124, no. 9, pp.846-856.
Mair R.J., Gunn M.J. & OReilly M.P., 1981, Ground movement
around Shallow Tunnels in Soft Clay, Proceedings of The
10th Conference on Soil Mechanics and Foundation Engineering, Stockholm, Vol.1, pp. 323-328
OReilly, M.P. & New B.M., 1982. Settlements above Tunnels in
United Kingdom-Their Magnitude and Prediction, Tunnelling82, pp 173-181.
Peck, R.B., 1969. Deep Excavation in Soft Ground, Proc. 7th International Conference on Soil Mechanics and Foundation
Engineering, State of the Art Volume, pp 225-290.

Phienwej, N. 1998. Ground movements associated with soft

ground tunnelling in Bangkok. Proc. World Tunnel Congress 98: Tunnels and Metropolises, Sao Paulo, Brazil.
Suwansawat, S., 2002. Earth Pressure Balance Shield Tunneling
in Bangkok: Ground Response and Prediction of Surface Settlements using Artificial Neural Networks. Doctoral Dissertation, MIT, USA.
Timpong, S., 2002. Analysis of Ground Movement in Bangkok
MRT Blue Line Project. M.S. Thesis, AIT, Bangkok.
Tavaranum, S., 2003. Analysis of Subsurface Ground Movement
Induced by EPB Shield Tunnelling in Bangkok MRT Blue
Line Project. M.S. Thesis, AIT, Bangkok.
Verruijt, A. & Booker, J.R., 1996, Surface settlements due to deformation tunnel in an Elastic Half Plane, Geotechnique, Vol.
46, No. 4, pp. 753-756.

534

Proceedings of International Conference

Tunnelling and Trenchless Technology

and

Exhibition

on

1 3 March 2011, Selangor, Malaysia

International Conference and

Exhibition on Tunnelling and
Trenchless Technology
Geotechnical
Risk
Assessment
and
Management for Tunnelling Project in
Congestive Urban Area Bangkok Experience
Zaw Zaw Aye and Phongpat Kitpayuck
SEAFCO Public Co., Ltd. Bangkok, Thailand

Noppadol Phienwej

Asian Institute of Technology, Pathumthani, Thailand

535

Geotechnical Risk Assessment and Management for Tunnelling Project

in Congestive Urban Area Bangkok Experience
Zaw Zaw Aye & Phongpat Kitpayuck
Seafco Public Company Limited, Bangkok, Thailand
[email protected]
Dr. Noppadol Phienwej
Asian Institute of Technology, Bangkok, Thailand
[email protected]

Abstract: The M.R.T Chaloem Ratchamongkhon Line is the first underground mass transit system project in Bangkok developed by
the Mass Rapid Transit Authority (MRTA) of Thailand. This paper presents the methods adopted in prediction of tunneling induced
ground movement and associated building damage risk-assessment carried out for the Contract No. 1 of the M.R.T Chaloem Ratchamongkhon Line constructed in congestive urban area. Application of observational method in managing risk of damage to the existing
buildings and structures are also reported.

1 INTRODUCTION
Geotechnical risk is one of the key elements of the technical and
financial risk in most of the major construction projects. Though
engineers use adequate factor of safety and taking account of
other risk factors in design and construction process, systematic
approach of risk assessment and management is not commonly
applied in most of the projects.
Tunneling works and deep excavation works pose a high degree of geotechnical risk that needs to be identified, assessed,
minimized and managed. Deep excavation and tunneling works
generally interact with natural ground and incorporating their
characteristics as structural elements of their own stability. This
interaction is more significant in tunneling projects than in other
civil engineering projects particularly in congestive urban area.
In this paper, process of geotechnical risk assessment in Bangkok
first underground mass transit project is presented.
2. GEOTECHNICAL RISK MANAGEMENT
Geotechnical risk is the risk to construction project originated
from the ground condition. In general case, once risk have been
identified and assessed, to manage the risk fall into one or more
of the following major categories.
Eliminate or avoid
Transfer
Mitigate
Tolerate
Unfortunately, ground-related risks are not often practical to
completely avoid or transfer. However, systematic approach,
control and management can minimize the risk if not entirely
eliminate. CIRIA Report 185 outlined the four fundamental
stages of risk management as follows ;
Hazard identification : where hazards are identified and documented in a register via experience of similar projects, brainstorming, structured interviews etc.

537

Risk assessment : where the likelihood and consequence of

each hazards are evaluated and combined to estimate the risk corresponding to each hazard. A risk register is produced at this
stage for systematic assessment of the risk.
Risk reduction : where the hazards are eliminated if possible
and the risks are reduced by a combination of design changes,
procedural changes and additional monitoring
Risk control : where the risks are monitored, controlled and
managed throughout the project. The risks normally diminish as
the construction progresses.
In general, risk management is the process of managing response and policies to reduce and control risks. It is an integral
part of the observational method, which concentrates on technical
risk (CIRIA Report 185). The merits of using the OM to manage
ground-related risks have been discussed by Whitman (1984),
Blockely (1994), Chowdhury (1994) and Godfrey (1996). Roles
of different parties involved in construction project namely the
client, the designer and constructor are described in details by
Clayton (2001).
3. CONTRACT 1 OF MRTA
MONGKHON LINE PROJECT

CHALOEM

RATCHA-

The M.R.T Chaloem Ratchamongkhon Line is the first underground mass transit system project in Bangkok developed by
the Mass Rapid Transit Authority (MRTA) of Thailand. Total
length of the underground structure is 21.5km, comprises of 16
km of twin single-track bored-tunnel, a total of 4 km long18 cutand-cover stations, a twin 1.5km long cut-and-cover approach
tunnel to Depot and other associated structures. The project was
commenced in late 1996 and opened for the public in 2004. The
Contract No. 1, Underground Structure South - southern portion
of this initial system comprising 9 km of twin 6m outside diameter bored tunnels, nine underground stations and cut-and-cover
depot approach tunnel, was awarded to the Joint Venture BCKT

Fig. 1. Tunnel alignment of the Contract 1 (South Contract) of Bangkok MRT

Fig. 2. Configuration of deeper stations with Stack Platform

consisting of Bilfinger + Berger Bauaktiengesellschaft, Ch Karnchang Public Company Limited, Kumagai Gumi Company Li
mited and Tokyu Construction Company Limited. The project
was commenced in late 1996, successfully completed and opened
for the public in 2004.

One of the major challenges of the Contract No. 1, was the

need of stacked-alignment for a major portion of twin boredtunnel underneath Rama IV road, one of the busiest roads of
Bangkok, which called for a stacked configuration at 3 stations,
Lumpini, Si Lom and Sam Yan as shown in figure 2. This requirement led the contractor to design and construct the deepest
underground structures of Bangkok at Si Lom Station - four-floor
stacked-station of 32m deep right under the existing flyover and
adjacent to a number of sensitive structures. Moreover, as
alignment of the lower tunnel had to place in the deeper depth
within first sand layers for the last four sections of the Contract 1,
the process involved in tunnelling works were more complicated.
Regional settlement due to active consolidation of top soft
clay layer and piezomatric drawdown caused by deep well pumping was another key factor for which design and construction of
both tunnels and station boxes had have taken into consideration.
Deep excavation and bored tunneling works in congested urban environment of Metropolitan Bangkok requires a systematic
process of ground movement prediction, building risk damage
assessment and called for the stringent management of the risks.
In this paper focus is limited to prediction of tunnelling induced
settlement and assessment of potential damage to existing adjacent building and structures though rigorous assessment was also
made for ground movement caused by deep excavations by diaphragm wall support cut-and-cover construction method.

538

Fig. 3. Generalized soil profile along the tunnel alignment and stations of Contract No. 1 (Phienwej et. al. 2006)
4. SUBSOIL CONDITION ALONG CONTRACT 1 ALIGHMENT
Similar to other localities in Bangkok, a typical subsoil profile
along the tunnel alignment is characterized by the alternating
layers of clay and sand deposits. Figure 3 shows the generalized
soil profile along the tunnel alignment of the Contract No.1.
Weathered crust of 2 m thick is commonly found as the top layer.
In urban areas of Bangkok, this layer is covered by fill material.
Soft to very soft, highly compressible dark gray marine clay lies
beneath weathered crust or fill. Depending on the location, this
layer extends up to 12-16 m. About 2 m thick medium clay layer
can be observed between soft clay and underlying stiff clay.
Generally stiff Clay layer occurs directly underneath medium
clay and its depth goes up to 22 m. Below stiff clay layer, first
sand layer of 5-8 m thickness can be found. This first sand layer,
however, is absent in some areas. Stiff to hard clay layer underlies first sand and it is found to be about 5 m thick. Second sand
layer generally occurs at depths between 45 to 65 m. As can be
seen in Figure 3, tunnel alignment is located mainly in stiff clay
layer. Tunnels were bored in the first sand layer before reaching
to Queen Sirikit Center station and deeper sections between Bon
Kai (Khlong Toey station) and Hua Lamphone stations.
5. STATION BOX CONSTRUCTION AND TUNNELLING
WORKS
Cast-in-place rigid diaphragm walls were used as permanent retaining structures for all stations and some entrance structures
and ventilation buildings. Most of the entrance and ventilation
buildings were excavated by sheet pile walls. Tip levels of diaphragm wall ranges from 19m (at Hua Lamphong station) to 39m
(at Sam Yan and Silom stations) for the final excavation depths
of 15m to 31m. Excavated down to 31 below existing ground
level at Silom station marked the deepest excavation supported
by diaphragm wall in Bangkok. Top-down construction method
was used for all stations having stanchions for vertical support
and skeleton slabs as bracing during bulk excavation of station
boxes (Figure 4).

Fig. 4 View of diaphragm wall, stanchions and skeleton slabs

Earth-pressure-balanced (EPB) tunneling method using shield
tunnel boring machines (TBM) was applied for tunnel excavation
except at the cut-and-cover section of tunnel approach to the depot. Japan made two sets of Kawasaki TBM were used from Rama 9 station to Queen Sirikit Convention Center (QSCC) station.
Tunnel boring from QSCC station to Hulamphong station was
carried out by two sets of Herrenknecht TBM. Kawasaki TBMs
were designed to drill through the concrete diaphragm wall
having specially manufactured Fiber Reinforcement (FRP) installed at tunnel eyes of diaphragm walls. Pre-opening of diaphragm wall concrete and removal of steel reinforcement were necessary at tunnel eyes for Herrenknecht TBM. Tunnels were
bored with overcut by cutter-head at front (total cut diameter during boring was 6.48m) with backfill grouting. At permanent
stage, tunnel was supported by 300mm thick precast segmental
concrete rings having 6.30m and 5.70m outer and inner diameter
respectively. The loads from existing and future structures within
influence zone were taken into account in tunnel lining design.
Movement joints (Omega Joint) were provided at the connection
at station diaphragm walls and tunnels to accommodate the diffe-

539

rential settlement between two structures. Figure 5 shows the

tunnel configuration at permanent stage.
Tunnel alignment generally follows the existing main Bangkok roads, having single-bored twin tunnels for north and south
bound tunnels at operation stage. From Rama 9 station to initial
section of Khlong Toey station, twin tunnels were bored at same
level (about 21m below existing ground level). Due to the presence of existing flyovers located in the middle, two tunnels were
required to change the alignment having stacked one above the
other.

Fig. 7. View of Bangkok major road where twin tunnels were

bored underneath at approximately 21m below existing ground
level
6. OVERVIEW OF GROUND MOVEMENT PREDICTION
AND BUILDING DAMAGE ASSESSMENT PROCESS APPLIED IN THE PROJECT

Fig. 5. Typical sectional configuration of tunnel at permanent

stage

Well accepted published literatures were used as basic reference

and method in ground movement prediction and building damage
assessment. For the station boxes where cut-and-cover methods
were used for excavation, the work of Peck (1969) and Clough
and ORourke (1990) were used whereas published papers of
Burland et. al. (1977) and Boscardin & Cording (1989) were applied for bored tunnels. Building damage assessment was carried
out in 3 stages; Stage 1 Assessment, Stage 2 Assessment, and
Stage 3 Assessment. Greenfield Site assumption was adopted
in Stage 1 and 2 Assessment ignoring any influence of the stiffness of the structure and type of supporting foundation. Stage 2
Assessment was further extended by Additional Stage 2 Assessment. In Additional Stage 2 assessment, the variation of settlement trough shape and magnitude with depth were considered in
identifying the risk of building damage. In Stage 3 assessment, a
detailed qualitative assessment of structure and vulnerability of
each building or structure to damage were taken into account,
coupled with detailed structural analysis in assessing the buildings falling in risk category 3 classified by Mair et. al. (1996).
7. SUMMARY OF INSTRUMENTATION PROGRAM

Fig. 6. View of EPB Tunnel Boring Machine after cutting

through diaphragm wall

Instrumentation played a major role in the construction of this

first underground mass transit system project in congestive urban
area with difficult soil condition. The primary objective of the instrumentation program was to monitor the performance of the
deep excavation and tunneling operation to ensure safe execution
of the construction works and that adjacent structures were not
adversely affected. Furthermore, the instrumentation program
was established to provide feedback in application of the observational method. Extensive sets of instrumentation were installed
around the predefined influence zones of station box excavation,
tunneling and associated building, structures and utilities. In addition to other geotechnical instruments, inclinometers, extensometers, various types of settlement points, tile-meter, crack meter
and crack gauge were installed to monitor the ground and associated building and structures movement.

540

9. BUILDING DAMAGE ASSESSMENT AND APPLICATIO

OF OBSERVATIONAL METHOD

8. BUILDING CONDITION SURVEY

Prior to commencement of the major works, all the existing
buildings and structures within the predefined influence zone
were systematically surveyed. The primary objectives of the surveys were;

To record pre-construction status of all buildings and structures within the influence zone. The records included were
age of building, building type, storey height, expected foundation types, visible defects supported by sketches and photographs as appropriate

To utilize the information collected from the building /

structure condition survey in the assessment of the buildings
and evaluate sensitive structures with respect to construction
aspects and monitoring

To establish a benchmark to monitor the possible effect of

construction and feedback information in applying observational method and if necessary to implement the building
protection measure
Expected foundation types of different buildings were classified based on the height, number of floors and age of the buildings. Building survey information was also used to identify the
types of instrumentation required.

The main reasons of applying observational method in risk assessment were ;

Uncertainty in Ground Movement Prediction and Staged

Assessment of Buildings

Due to the different types of buildings and structures along

MRTA alignment it was recognized that results obtained
from Building Damage Risk Assessment might be conservative for some buildings / structures and could be underestimated for others

To minimize unnecessary implementation of building protection measures without compromising safety

Figure 8 shows the link between observational method and
geotechnical risk management, particularly risk of damage to existing buildings and structures. As mentioned in earlier sections
of the paper, controlling and managing the risk is one of the most
important factors in successful completion of the tunnelling
works in urban environment and hence all parties involved had to
pay special attention on the issue.

Geotechnical
Risk Management

OM

Identify Risk
Risk of damage to building

Design & Plan

Obtained data from fully instrumented
first few locations, reanalyzed and back-calculated

Analysed Degree
of Risk
(Building Damage Category)

Construction Control
Monitoring & Review
Trigger Criteria
Exceeds

Control

Yes

No
Implement
Planned Action

Fig. 8. Flowchart showing the OM process involved in Building Damage Risk Assessment
Stage 1

Prediction and preparation of surface settlement

(Green Field Site Condition)

< 10mm

No further
assessment

>10mm

Stage 2

Determination of building risk category

(limiting tensile strain) based on surface settlement

No further
assessment

< slight

> slight

Additional Stage 2

Determination of building risk category (limiting tensile strain)

based on sub-surface settlement

< slight

No further
assessment

> slight

Stage 3

Detailed Assessment
(building structural check, consider building stiffness, Soil-structure interaction)

Fig. 9. Diagram showing process of Staged-Assessment

541

Process of Stage-Assessment is shown in figure 9. After carrying out the initial assessment of potential building damage, it
was recognized that without applying the observational method,
the result obtained from assessment process could be conservative for some buildings and could be underestimated for others.
For example, the settlement estimated in Second Stage Assessment shown in above diagram was based on surface movement.
For the buildings located in close proximity to the excavation
founded on piles, it was considered that subsurface settlement
may be more critical. Stage 2 Assessment was then further extended by Additional Stage 2 Assessment in which the variation
of settlement trough shape and magnitude with depth were considered in identifying the risk of building damage. Settlement of
the buildings due to the excavation works were predicted assuming the green filed condition. Neglecting restraints from foundation and structure, it was assumed that buildings follow the
ground settlement trough at foundation level (estimated pile tip
level). However, if the building is supported by deep foundations, assessment result could be too conservative.
10. PREDICTION OF GROUND MOVERMENT INDUCED
BY TUNNELLING WORK AND RISK ASSESSMENT
10.1 Prediction of surface settlement
The Gaussian normal distribution curve developed by Peck
(1969), OReilly and New (1982) was used to estimate the
ground settlement induced by a single tunnel.
10.2 Stage 1 Assessment

total settlement induced by the twin tunnels. 10mm and 25mm

settlement contours were plotted in the topographic plan with
buildings and other structures including flyover and khlong
bridges in the tunnel alignment. Buildings falling outside 10mm
settlement contours were eliminated from further assessment.
Remaining buildings with settlement greater than 10mm were
short-listed for stage 2 assessment.
10.3 Stage 2 Assessment
The buildings and structures short-listed in stage 1 were assessed
in this stage. Stage 2 assessment involves calculation of maximum tensile strain induced in the building and structures due to
tunnel boring and classification of risk category. Risk category
of the buildings was determined applying the method proposed
by Mair et. al. (1996),
10.4 Additional Stage 2 Assessment in tunnel zones
As most of the buildings and structures in Bangkok along tunnel
alignment are supported by piled foundations, it was considered
that Stage 2 Assessment using surface settlement profile might be
insufficient. Hence, Additional Stage 2 Assessment considering
the variation of settlement trough shape and magnitude with
depth (subsurface settlement) was carried out for the critical
buildings.
In general, subsurface settlement trough will be narrower and
steeper than those of the surface as shown in the sketch below. A
simple approach was used in predicting the subsurface settlement
profile with following assumptions.

First stage assessment of the risk of damage to buildings and

structures in tunneling zones involved preparation of settlement
profile induced by tunnel excavation. Settlement was computed
at every 100m interval along tunnel alignment using the method
based on Gaussian normal distribution curve presented above.
Settlement induced by each tunnel (north and south bound) was
predicted independently and then superimposed to determine the

(1) The shape of the subsurface settlement profile caused

by the tunnel construction is characterized by a Gaussian distribution similar to surface settlement profile.
(2) The ground settlement influence line for tunnel in clay
is the same for subsurface as shown in the figure 10 below:

Fig. 10. Distribution of surface and subsurface settlement induced by tunnel boring.
.

542

Therefore, substituting the distance above the tunnel axis Z0-Z

for Z0, equations derived from Gaussian normal distribution
curve subsurface settlement profile was determined. In addition
to this method following equation can also be used to predict the
subsurface settlement.

The risk of damage to the building in terms of magnitude of building slope with reference to the limiting values specified by various researchers.
The following assumptions were made in the conservative
approach of Stage 3 assessment.

Buidling movement is influence by excavation induced

settlement of both main station box and associated
underground structures (e.g. Entrance Building)

Building settlement was casued by subsurface

settlement at foundation level as buildings are supported
by piled foundation.

No allowance was made for structural stiffness of the

building.

No allowance was made for possible effects of soilstructure interaction.

A summary of risk category determination in Stage 3 Assessment based on the predicted and monitored settlement of the
building adjacent to station excavation work is shown in the Table 1 below.

K = [0.14+0.26(1-z/z0)] / (1-z/z0)
By substituting K values for K in equation derived from
Gaussian normal distribution curve the subsurface settlement at
required depth can be calculated.

0
-10
-20

Surface_Settlement_NB Tunnel
Surface_Settlement_SB Tunnel

-30

Surface_Settlement_Total

-40

Settlement @ 12m _ NB Tunnel

Settlement @ 12m_SB Tunnel

-50

Total Settlement @12m

SB-TUNNEL CENTER

-60

Table 1. Summary of risk category determination based on Boscarding and Cording (1989)

NB_TUNNEL CENTER

-70
0

10

20

30

40

50

60

70

80

Distance (m)

Data Taken
From

Fig. 11. Surface and sub-surface troughs

Transverse Settlement Trough due to Tunnel Boring
Settlement Array, CS-7D(SS-7-53A-0 )
20

CL_NB
CL_SB

15

12/11/99 12:00

10

04/01/00 12:00
19/01/00 12:00

25/01/00 21:00

26/01/00 06:00

-5

26/01/00 09:00

-10

26/01/00 15:00

26/01/00 12:00
26/01/00 18:00

Settlement, m

-15

27/01/00 00:00

-20

27/01/00 03:00
27/01/00 06:00

-25

27/01/00 09:00
27/01/00 12:00

-30

Settlement
Prediction
Actual
Monitoring
Data

Sv1
(mm)

Sv2
(mm)

(mm)

L
(m)

Risk category / degree

of damage

48

23

25

18.6

0.0013

2 / slight

53

14

39

18.6

0.0020

2 / slight

Angular Distortion
S

*Predicted horizontal movement was used in computing

horizontal strain for actual miniotring data.

27/01/00 15:00

-35

27/01/00 21:00

-40

28/01/00 03:00

-45

29/01/00 12:00

From the above table, following conclusions can be made.

28/01/00 12:00
31/01/00 12:00

-50

02/02/00 12:00

-55

09/02/00 12:00
02/03/00 12:00

-60

05/04/00 12:00

-65

Predicted

-70
-25

-20

-15

-10

-5

10

15

20

25

Distance, m

Fig. 12. Actual settlement profile monitored at fully instrumented location in comparison with prediction
10.5 Stage 3 Assessment
The method used in Stage 3 Assessment of the buildings in tunnel zone was based on the superimposed predicted subsurface
settlement trough induced by twin tunnels.
Based on the pre-construction state building condition survey
and detailed structural survery, the method for damage
assessment was selected. Two main approaches were applied.

Conservative approach, mainly applied in additional

stage 2 assessment with some refinements
Frame analysis using FEM structural software

Predicted angular distortion broadly agrees with that of

actual monitoring data. Hence, settlement prediction and
assumed effective depth of piled foundation is fairly
reasonable.

The settlement of actual monitroing data is slightly

higher than that of prediction. This may be due to the
additional settlement caused by diaphragm wall panel
construction which influence zone might have extended
to 15m from the diaphragm wall (equivalent distance to
bottom of soft clay level ).
Limiting values of the building slopes at two critical conditions specified by various researchers as shown in the table 2 below were also used in determining the potential building damage
of stage 3 assessment.

Table 2. Summary of Limiting Building Slope Specified by Various Researchers

Limiting Values by
Various Researchers

Type of Damage
Cracking in walls Structural Damage
and partitions
1/300 to 1/500
1/150

10.6 Stage 3 Assessment by conservative approach

1. Skempton and
Macdonald

Conservative approach applied in Stage 3 Assessment involved

consideration of two separate criteria, namely ;

2. Meyerhof

1/500

1/250

3. Polshin and Tokar

4. Bjerrum

1/500
1/500

1/200
1/150

The risk of damage to the building specified by angular

distortion and horizontal strain.

543

10.7 Stage 3 Assessment by structural frame analysis

Some buildings of critical condition were assessed by structural
analysis using detailed structural survey information as basic data. By simulating the differential settlement of the building predicted from Additional Stage 2 Assessment, development of
moment and shear in the main structural members were determined with the use of structural analysis software. Computed
moment and shear were assessed in comparison with the capacity
of the structural members estimated from collected building condition survey and detailed structural survey data. The process involved in structural assessment can be summarized below.

Factor of safety of each structural members were determined base on predicted differential settlement of
the buildings.
A series of assessment were performed to determine
the critical differential settlement causing structural instability of the members or differential settlement at
limiting factor of safety.
Trigger values were established based on above results.
Based on the trigger values and the actual building
condition and response in conjunction with monitoring
data, the most sensitive structural members were protected, mainly by strengthening approach.

10.8 Damage Assessment on Sensitive Buried Structures

Apart from the flyovers (bridges), the most critical structure required for Stage 3 Assessment were MWA main water tunnel as
tunnel alignment crosses the water tunnels at two critical locations the first near Rama IV and Rachadaphisek road intersection and the second at Rama IV and Siphraya road intersection
near Sam Yan station. The position of water tunnels in relation to
MRT tunnels drawn in the subsoil profile at relevant locations
are shown in Figure 13 and 14.
According to the available information, MWA tunnel at the
first crossing is of 2.5m internal diameter steel lining with
150mm segmental ring and at the second crossing MWA tunnel
is of 2.5m diameter segmental ring with lightly reinforced concrete lining.
In preliminary stage 3 assessment, based on the predicted
subsurface profile using the method described in published paper
Ground Movement Prediction and Building Damage RiskAssessment for the Deep Excavations and Tunneling Works in
Bangkok Subsoil, Zaw et. al (2006) e, structural assessment of
water main was carried out and results were compared with structural capacity of water tunnel calculated from the material properties indicated in the construction drawings provided by MWA.
As bending moment induced in the water tunnel was considered
critical parameter in assessing the damage which is governed by
both maximum settlement and curvature of the water tunnel further assessment was proposed coupled with observational approach.

To carry out back-calculation using monitoring data obtained from the settlement arrays (settlement points, extensometer data) at first 3 locations before reaching to
MWWA water tunnel near Rama IV Rachadapisek intersection, to determine the volume loss and relationship between surface and subsurface settlement as well as best estimation of subsurface K value and subsurface settlement
profile at water tunnel level.

To perform beam-on-elastic foundation analysis using the

subsurface settlement profile obtained from predicted subsurface settlement bending moment likely to occur in
MWWA water tunnel.

To compare the result obtained from above item with allowable values.
To establish the final trigger values based on above comparison.
To confirm the settlement profile from actual monitoring of
last array is within the alert level of final trigger value.

Fig. 13. Position of MWA water tunnel in relation to MRT tunnels (clear distance are 3.38m and 3.18m for North and South
bound respectively) at the first crossing.

Fig. 14. Position of MWA water tunnel in relation to MRT tunnels (clear distance are 1.5m and 2m for North and South bound
respectively) at the second crossing
For the first location of MRT-MWA tunnel crossing, target
ground loss from tunnel operation was set at 0.5%, whereas
ground loss based on allowable movement curvature of water
tunnel derived from predicted pattern of subsurface settlement
was 0.65%. Back-calculated ground loss derived from the monitoring data of rod-extensometer placed at the crown of MWA water tunnel was within target value. In addition to the preliminary
assessment based on prediction of subsurface settlement using
simple empirical formulas, a detailed soil-structure interaction
was analyzed using 2-dimensional FEM considering behavior of
tunnel boring by earth pressure balance (EPB) method. Stress relief from the effect of cutter-head, tail void and over-cut from
EPB tunnel boring was considered in the analyses. Adding
ground movement due to consolidation caused by disturbance
from tunnel operation surface and subsurface settlement were
predicted by finite element analysis. Detailed parameters and assumption made in analysis was presented in the published paper
of Sakai & Sugden (2000).

544

After the successful passing of both tunnels at the first MWA

crossing, another trial area tunneling was planned to further determine the ground movement associated with various tunneling
parameters well before reaching to more critical crossing after
Sam Yan station. Tunnel section between Silom and Samyan station, in which soil profile similar to that of second crossing was
selected and comprehensive instrumentation program was established to monitor surface and subsurface settlement as well as horizontal movement of the ground at depth.
Distance (m)

Subsurface Settlement (mm)

10

20

30

40

50

60

70

80

90

-10

Settlement _ NB Tunnel
-30

Settlement_SB Tunnel
Total Settlement

-50

Fig. 15. Predicted subsurface settlement trough of assumed

ground loss 1% calculated from empirical formula and actual
subsurface settlement (indicated by dark-circle) observed from
rod extensometer first MWA tunnel crossing.
The observed ground movement induced by tunneling works
were within predicted values. Well organized plan, systematic
approach of observational method to achieve control values
backup by extensive instrumentation and close co-operation
among relevant personnel were the key factors contributed to
successful completion of EPB tunneling particularly in the most
sensitive points at MWA water tunnel crossings. The monitored
subsurface settlement data from extensometer and predicted settlement trough of assumed ground loss 1% ground settlement
profile of the first MWA tunnel crossing is presented in figure
15.
10. KEY FACTORS IN SUCESSFUL COMPLETION OF THE
PROJECT
The M.R.T Chaloem Ratchamongkhon Line, one of the most
modern underground mass transit systems in the region was successfully completed and has been in operation for over 6 years.
All deep excavation works and tunneling were successfully completed without any significant damages to the adjacent buildings
and structures. Well organized plan, systematic approach of observational method to achieve control values backup by extensive
instrumentation and close co-operation among relevant personnel
were the key factors contributed to successful completion of the
project. The purpose of this paper is to provide a set of technical
guidelines for ground movement prediction and building damage
risk assessment of deep excavation and tunneling works in Bangkok subsoil and application of observational method which may
partly serve as a reference for the future underground mass transit system projects in urban area of Bangkok. Successful completion of the first MRT project in Bangkok marked the practicality
of well-executed deep excavation works and tunneling by EPB
method coupled with observational method in the sensitive urban
environment.

REFERENCES
Boscarding, M.D. and Cording, E.G. (1989), Building Response
to Excavation-Induced Settlement. Journal of Geotechnical
Engineering, ASCE, 115;1;1-21
Burland, J.B., and Wroth, C.P. (1974), Settlement of Buildings
and Associated Damage. SOA Review, Conference on Settlement of Structures, Cambridge, Pentech Press. London,
pp. 611-654
Burland J.B., Broms J.B. and de Mello VFB (1977), Behavior of
Foundations and Structures, SOA Report Session 2, Proceedings of 9th International Conference, SMFE, Tokyo, 2:495546
Clayton C. R. I (2001), Managing Geotechnical Risk, Improving
Productivity in UK Building and Construction, Institute of
Civil Engineers and Thosmas Telford Ltd. UK
Clough, G.W., and ORourke, T.D., (1990), Construction Induced Movements of In-situ Walls, ASCE Geotechnical
Special Publication No. 25, pp. 439-470
Clough C.W. and T.D. ORourke (1990), Deep Excavation and
Tunneling, 7th ICSME, Mexico City
Graftio H. (1936), Some Features in Connection with the Foundation of Svir 3 Hydro-electric Power Development, Proceedings of the first international conference in Soil Mechanics, Cambridge, Mass., 1, pp. 284-290
JV BCKT, (1997) Geotechnical Interpretation Report (GIR) for
MRTA Initial System - UGS, Bangkok
Mair, R.J., Taylor, R.N., and Burland, J.B., (1996), Prediction of
Ground Movements and Assessment of Risk of Building
Damage. Geotechnical Aspects of Underground Construction
in Soft Ground, pp. 712-718, Balkema, Rotterdam
OReilly M.P and New B.M (1982), Settlements above tunnels in
United Kingdom their magnitude and prediction, Tunneling
82, pp. 173-181, London, IMM
Peck R.B. , (1969) Deep Excavation and Tunneling in Soft
Ground, 7th ICSME, Mexico City, State of the Art 3, 225-290
Peck, R. B. [1969]. Advantages and limitations of the observational method in applied soil mechanics, Geotechniques 19,
No.2, pp. 171-187
Phienewej N., Photayanuvat C., Suwansawat S., Mcconochie D.,
(2006) Geotechnical Aspects in Design and Construction of
Chaloemratchamongkhon Underground MRTA, Bangkok,
Proceedings of the International Symposium on Underground
Excavation and Tunneling, Urban Tunnel Construction and
Protection of Environment, EIT, Bangkok, Thailand, , pp.
271-279
Powderham, A. J. [1996]. An overview of the observational method : development in cut and cover and bored tunneling
projects, The Observational Method in Geotechnical Engineering, The institute of civil engineers, Thomas Telford Publishing, London, pp. 7-24
Zaw Zaw Aye , Dhiraj Karki and Christian Schulz [2006].
Ground Movement Prediction and Building Damage RiskAssessment for the Deep Excavations and Tunneling Works
in Bangkok Subsoil, Proceedings of the International Symposium on Underground Excavation and Tunneling, Urban Tunnel Construction and Protection of Environment, EIT, Bangkok, Thailand, , pp. 281-297

545

Geotechnical Special Publication no. 225

25-29 March 2012, Oakland, California, USA

Geocongress 2012 - State of

the Art and Practice in
Geotechnical Engineering

Tunneling effects on pile group response in

Bangkok
Thayanan Boonyarak and Charles W. W. Ng
The Hong Kong University of Science and Technology

547

GeoCongress 2012 ASCE 2012

3119

Tunneling effects on pile group response in Bangkok

Thayanan Boonyarak1 and Charles W. W. Ng2
1

Research student: [email protected]

Chair Professor, Ph.D, CEng, FICE, FASCE, FHKIE, FHKEng: [email protected]
The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon,
HKSAR
2

ABSTRACT
Tunneling adjacent to an existing pile group may induce additional force,
bending moment and settlement of a pile. Moreover, it may cause load redistribution
and differential settlement in a pile group. In this paper, the construction of a water
diversion tunnel with an outside diameter (D) of 5.55 m constructed in Bangkok is
reported. The tunnel was constructed in very stiff clay using an earth pressure
balance shield with a cover-to-diameter ratio of 4.5 and very close to a pile group.
The pile group was located within a distance of 1D from the tunnel springline.
Tunneling induced ground responses were measured by using surface settlement
points, extensometers and inclinometers. To investigate the influence of tunnelling
on the pile group, a three-dimensional, elasto-plastic, coupled-consolidation
numerical analysis was also conducted to simulate the pile group responses.
Observed ground responses and computed changes of horizontal effective stress and
excess pore water pressure around piles are reported. Redistribution of pile shaft
resistance, load transfer in the pile group and pile head settlements are investigated.
INTRODUCTION
Tunneling adjacent to an existing pile group may induce additional bending moment,
cause redistribution of axial force in each pile and load transfer in a pile group. Any
stress change in the ground due to tunnelling is the main concern to engineers in
evaluating capacity and additional settlement of a pile group.
Chen et al. (1999) and Lee & Ng (2005) carried out numerical analyses to
study the influence of tunneling on adjacent pile. They found that negative skin
friction was induced along the pile shaft as a result of stress relief. Tunneling
adjacent to a pile group was also studied by Pang et al. (2005) and Cheng et al.
(2007). It is reported that the maximum bending moment induced in a pile occurs at
the tunnel axis in the transverse direction of tunnel advancement while no significant
bending moment occurs in the longitudinal direction.
In this paper, monitored soil displacements due to a 5.5m diameter shield
tunnelling adjacent to a pile group are reported and compared with computed finite
element results. The capacity and settlement of the pile group due to the tunneling
are also investigated.

549

GeoCongress 2012 ASCE 2012

3120

THE SITE
Figure 1a shows plan of tunnel alignment and adjacent pile groups of elevated
railways. The layout of instrumentation is also shown in the Figure. Various
instruments were installed to observe tunneling induced ground displacement and
structural response. Three surface settlement points (S1, S2 and S3), two
extensometers (Ex1 and Ex2) and one inclinometer tube (I1) were installed to
measure ground response. Two pier settlement points (C1 and C2) were mounted on
pier to measure pier settlement. The monitored data is interpreted from BMA flood
diversion tunnel project: Sansaeb Canal and Ladprao Canal Section (Lim et. al,
2009). The total length of tunnel in this project is about 5 km.
Figure 1b shows a typical soil profile of the study area. The first upper layer
is 12 m thick soft clay. Below it is the succession of medium stiff clay of 3 m, stiff
clay of 5 m, very stiff clay of 14 m and stiff clay of 14 m thick layers. A 12 m very
dense sand layer underlies the clay layer. A water diversion tunnel with an outer
diameter (D) of 5.55 m was constructed in very stiff clay at cover-to-diameter ratio
(C/D) of 4.5 using an earth pressure balance shield machine. The tunnel alignment is
very close to a (22) pile group with center to center of pile spacing 3 m. The outer
diameter and length of the shield are 5.70 m and 8.05 m, respectively. The diameter
of each concrete bored pile is 1.0 m. The length of piles is 60 m and pile toes are
embedded in very dense sand layer. All four piles are rigidly connected with 1.8 m
thick pile cap. The clear distance between the nearest pile and the tunnel springline is
3.65 m. The tunnel axis depth is approximate at the middle of pile length.
Instrumentation

(a)
t 71 R
umvi
Sukh

Surface settlement point

(b)

3.00
0.00 m

Pier settlement point

Extensometer

oad
.6
18

C2

13

8
.3

4.35 I1

S2

S3

Scale:
10

S1

20

C1

Sukhumvit Road

3.65

Ex2

Shield advancing

-34.0 m

Very
stiff
clay
Stiff
clay

-48.0 m
Very
dense
sand

S2
-3.0 m
Ex1&Ex2
-12.0 m
-15.0 m

-12.0 m Medium
-15.0 m stiff clay
Stiff
-20.0 m clay

Inclinometer

Ex1

C1

Soft
clay

-20.0 m

3.65

-27.55 m

5.55 m
I1

-40.50 m

4- 1.0 m
Concrete pile
-60.0 m

SECTION A

Figure 1. (a) Plan of study area and lay out of instrumentation (b) sectional view
A-A and soil profile
NUMERICAL ANALYSIS
To investigate the influence of tunnelling on the pile group, a three-dimensional,
elasto-plastic, coupled-consolidation numerical analysis was carried out using Finite
Element program ABAQUS (Hibbitt et al., 2008).

550

GeoCongress 2012 ASCE 2012

3121

Figure 2a shows the geometry of a 3D finite element mesh. A plane of

symmetry is identified at the tunnel axis, so only half of the domain is modeled. The
mesh is 40 m wide, 50 m long and 70 m high. Eight-node brick elements, four-node
shell elements and two-node beam elements are used to model the soil, pile cap and
piles, respectively. Roller and pin supports are assigned on the four vertical sides and
bottom of the mesh as boundary conditions, respectively.
Figure 2b illustrates tunnel excavation simulation technique adopted in the
finite element analysis (FEA). This technique is known as Displacement Controlled
Method (DCM), modified from Cheng et al. (2007). During shield advancement
over-excavation due to pitching angle is often encountered at the site. Therefore, it is
assumed that soil displacement at the upper part of the shield is larger than the lower
part of the shield and the shields invert is in contact with soil element. To simulate
each tunnel advancing step, displacement boundary is specified to each node around
the excavated perimeter. Consequently, soil element around the shield is pulled to a
final tunnel diameter of 5.55 m. Also, soil inside the excavated section is removed
simultaneously. A volume loss of 2.5% was imposed on tunnel boundary. By using
this technique, tunnel lining is not required. In order to simulate an earth pressure
balance shield, a roller support is assigned to the shield face to restrain horizontal
soil movement in front of the shield. Shield advancing speed of 5 m per day is
modeled in the analysis. Tunnel alignment is simplified from curve in actual
condition to be straight in FEA. The presence of pile foundations of surrounding
buildings are ignored in the analysis.
In Figure 2c piles and pile cap are arranged in rotated square. Total load on
the pile group is assumed to be 18,000 kN. Excessive pore water pressure generated
in response of the applied load on pile group is allowed to dissipate before tunnel
advancement in the numerical analysis.
(a)

50 m
Y=0

40 m

(b)

Excavated
diameter,
VL = 2.5%

(c) 18000 kN
P2

P4

P3
70 m

Tunnel
axis

Tunnel axis
3m

P1

3m

Final tunnel
diameter
5.55 m

Note:

1.0 m
Beam
element
Direction of shield advancement

Plane of symmetry

Figure 2. (a) Geometry of finite element mesh (b) tunnel excavation simulation
(c) detail of piles and pile cap

551

GeoCongress 2012 ASCE 2012

3122

An elasto-plastic soil model with Mohr-Coulomb failure criterion is used to

simulate soil behaviour. Soil parameters are selected based on a soil investigation
report and adopted from previous studies [Tanaka et. al (2001), Seah et. al (2000)].
Permeability of clay layers and sand layer is assumed to be 1x10-9 and 1x10-3 m/s,
respectively. Soil parameters are summarized in Table 1. The concrete piles and pile
cap are modeled as linear elastic with Youngs modulus and Poissons ratio as 30
GPa and 0.3, respectively. The unit weight of concrete is assumed as 24 kN/m3. The
water table is assumed to be located at the ground surface with a hydrostatic initial
pore water pressure profile.
Table 1. Soil parameters in finite element analysis
Su
SPT
c
Material
t
(kN/m3)
(kPa)
(N/ft) (kPa)
Soft clay
16
18-20
N/A
1.5
Stiff clay 1
19
90-120 12-15
2.0
Very stiff clay
20 160-200 25-30
3.0
Stiff clay 2
19 100-130 13-18
2.0
Very dense sand
20
- 50-60
0.1

()
18
19
20
19.5
35

()
6

'
0.15
0.15
0.15
0.15
0.15

E
(kPa)
1,200
5,000
7,000
6,000
40,000

MONITORED DATA AND COMPUTED RESULTS

In this paper, only measured ground horizontal displacement is reported. Reference
point of the shield distance used in this paper is taken at the shield face. A
monitoring section is selected at the center of the pile group (i.e., at Y=0) as
reference point for tunnel advancement.
Ground horizontal displacement. Measured horizontal ground displacement in
transverse direction of tunnel advancement is shown in Figure 3. When the shield
was more than 12 m away from the monitoring section, measured horizontal ground
displacement was insignificant. When the shield reached at 10 m behind the
monitoring section (i.e., Y= -10 m), soil moved outward from the shield. It indicates
that stress redistribution of soil due to soil arching may occur. More discussion and
explanation is given in the later section. Maximum outward movement of 3 mm from
the tunnel axis was observed at this stage.
When the shield reached the monitoring section (i.e., Y=0 m), soil abruptly
moved inward towards the shield and caused maximum inward displacement. Large
displacement occurred from depth 25 to 32 m which is considered about 1D (tunnel
diameter) from the tunnel axis. Measured horizontal displacement at the tunnel axis
depth was about 15 mm inward to the shield. When the shield reached at 25 m ahead
of the monitoring section (i.e., Y=25 m), measured horizontal displacement reduced
by 3 to 5 mm. The reduction may be caused by stress redistribution. No significant
horizontal ground displacement was observed when the shield advanced further.
Computed ground horizontal displacements show a similar trend as compared
with the measured data. When the shield reaches at 10 m away from the monitoring
section (i.e., Y=-10 m), computed outward soil movement from the shield occurs

552

GeoCongress 2012 ASCE 2012

3123

from depth 23 m to 40 m. When the shield reaches the monitoring section (i.e., Y=0
m), computed results show an abrupt soil displacement towards the shield. In
contrary to the measured data, when the shield passed at 25 m (i.e., Y=25 m) beyond
the monitoring section, a slight additional horizontal displacement towards the shield
occurs. The differences between measured and computed results may be caused by
tail void grouting is not simulated in the FEA. Therefore, outward movement in FEA
as the shield advances beyond the monitoring section does not occur. It can be
noticed that, the computed ground horizontal movement from FEA can reasonably
capture the measured ground response at the depth of tunnel axis. However, there is a
significant difference between computed and measured ground response near the
ground surface. Large measured horizontal ground displacement may be triggered
by other activities rather than tunneling such as surcharge induced by heavy traffic
load near the inclinometer tube.

Horizontal displacement (mm)

0

-15

-10

-5

10

15

20

Shield distance, Y
Measured (Y=-10m)

10

Measured (Y=0m)

15

Computed (Y=-10 m)

Measured (Y=25m)
Computed (Y=0m)

20

Computed (Y=25m)

25
30

35

Depth (m)

-20

40

Tunnel
axis depth

I1

-Y

Horizontal displacement sign

convention
+ : inward to tunnel axis
- : outward from tunnel axis

+Y

45

Figure 3 Comparison of measured and computed horizontal ground

displacement of I1 (measured data from Lim et al, 2009)
COMPUTED PILE GROUP RESPONSE
Load transfer in the pile group. Figure 4a shows changes of computed load at pile
head with respect to the shield advancement. The changes of load at the pile head are
noticeable when the shield advances from Y = -5 to -2 m. Load on P2 significantly
increases while load on P1, P3 and P4 decreases. Since P2 is the farthest pile from
the shield, it is the least affected by tunneling. This explanation is verified with
computed mobilized shaft resistance in the following section. Consequently, load at
pile head from P1, P3 and P4 is transferred to P2. Steady state of load transfer in pile
group is achieved when shield passes at Y=10 m ahead of the monitoring section.

553

GeoCongress 2012 ASCE 2012

3124

Maximum reduction and increment in load on pile head of P1 and P2 is 0.8% and
1.2%, respectively.
Pile head settlement and pier tilting. Figure 4b illustrates the computed results of
pile head settlement against shield advancement. In addition to effects of tunneling,
the observed pier settlement was highly influenced by other factors such as vibration
induced by heavy traffic and the elevated train. Furthermore, pier settlement was
barely noticeable by conventional surveying instrument. Therefore, the measured
pier settlement result is not reported in this paper. Maximum computed pile head
settlement occurs at P3 since significant shaft resistance reduction in this pile occurs
when the shield is approaching to the monitoring section. Moreover, load at pile head
on P3 is almost constant (see Figure 4a). Pile head settlement in P1 and P2 is almost
equal although different mechanism is developed. Pile head settlement in P1 would
be mainly caused by reduction of horizontal effective stress and development of
negative shaft resistance which is discussed in following section. Elastic shortening
due to increase of load at pile head would be the major cause of pile head settlement
in P2. The reduction of load at pile head in P4 is higher than in P3 (see Figure 4a).
Therefore, minimum pile head settlement is induced on P4. Tunneling induced pier
settlement is about 1.3 mm. The differential settlement in the pile group causes the
pier to tilt in the longitudinal direction of the tunnel axis toward P3. The maximum
inclination of the pier becomes 1:12,500 when the shield is 5 m away from the
monitoring section (Y = -5 m). In addition, the computed differential settlement
between two piers is 1:27,000 at the end of tunnel driving (the distance between two
piers is 35 m). Those computed results are considered very small for the
serviceability limit state criteria.
Shield distance, Y (m)

(a)

P (%)

1.0
0.5

-20

-10
P2
P3

10

20

30

+Y

0.0

-0.5
-1.0

P1
P3

Shield distance, Y (m)

-20

-10

10

20

30

P4
P1

-Y

0.0

p (mm)

1.5

(b)

P2
P4

0.5
1.0
Avg. 1.3 mm

1.5

Figure 4 (a) Changes of computed load at pile head (b) induce pile head
settlement
Mobilized shaft resistance. Figure 5 shows computed mobilized unit shaft
resistance of P1, P2, P3 and P4, during shield advancement. The significant change
of shaft resistance occurs in P1 as shown in Figure 5a. Mobilized shaft resistance
decreases significantly at depth 13 to 24 m. Negative shaft resistance develops in P1
when the shield reaches at the monitoring section and advances further, indicating
that soil settles more than pile at depth 12 to 25 m. To maintain load equilibrium in a
pile, mobilized shaft resistance in P1 below the tunnel axis increases. The change of
mobilized shaft resistance in P2 is the lowest as shown in Figure 5b. Negative shaft

554

GeoCongress 2012 ASCE 2012

3125

resistance did not develop in P2, which suggests that impact of tunneling on P2 is the
smallest. It is suggested that tunneling influence zone is about 1.5D from the tunnel
axis. P3 and P4 respond to the shield advancement in the same trend as shown in
Figure 5c and 5d, respectively. The only difference that can be noticed is significant
change occurs in P3 when the shield is at 5 m away from the monitoring section. The
significant change in P4 occurs when the shield is at 5 m beyond the monitoring
section. In other words, the maximum reduction of shaft resistance on each pile
occurs when the shield face is slightly beyond the pile.
(a)
0
5
10

Depth (m)

15

-10

-5

P1

fs (kPa)
0

10

(b)

15 -5

Initial
Y=-5m
Y=0m
Y=5m
Y=25m

fs (kPa)

10

15 -5

P2

(c)

P3

fs (kPa)

10

15 -5

(d)

fs (kPa)
5

10

15

P4

20

Tunnel
axis

25
30
35
40
45

P2
P3

-Y

P4
P1

Sign convention for horizontal axis

+ : positive shaft resistance (piles settle more than soil)
- : negative shaft resistance (soil settle more than pile)

+Y

Figure 5 Computed mobilized unit shaft resistance of (a) P1 (b) P2 (c) P3 (d) P4
Induced axial force and bending moment in pile. Figure 6a illustrates computed
tunneling induced axial force in each pile when shield is at 5m ahead of monitoring
section. The maximum induced axial force occurs in P1 at the depth of tunnel axis.
The trend of induced axial force distribution of P2, P3 and P4 is similar to P1. The
differences of magnitude of induced axial force are due to development of negative
shaft resistance in P1 (see Figure 5a). It can be noticed that minimum induced axial
force occurs in P2 as a result of minimum reduction of unit shaft resistance (see
Figure 5b).
Figure 6b shows the distribution of computed bending moment of each pile in
longitudinal direction of the tunnel axis (Mx). Maximum Mx occurs in P1 when the
shield is at 5 m beyond the pile (Y = 5 m). The bending moment (Mx) causes every
pile to bulge in the same direction as the shield advancement at the depth of tunnel
axis. Bending moment in each pile extends to depth 50 m, which is about 4D below
the tunnel axis.
Figure 6c illustrates the distribution of computed bending moment of each
pile in transverse direction of tunnel axis (My). Maximum My occurs in P1 when the
shield is at 5 m beyond the pile. The bending moment (My) causes the pile to bulge
towards the shield due to stress relief around the shield. By comparing bending
moment of a pile in 2 directions, My is significantly larger than Mx due to higher

555

GeoCongress 2012 ASCE 2012

3126

stress reduction around the shield in transverse direction than in the longitudinal
direction.
The ultimate bending moment capacity of a reinforced concrete bored pile
with diameter 1.0 m is approximate between 1,000 and 1,400 kN.m, given the
reinforcement ratio is 0.5% of the pile section. Therefore, maximum computed
induced bending moment in P1 is only about 20% of its capacity.
(a)
-100
0

(b)

Fx (kN)
0

100

200

400 -100

300

Mx (kN.m)
-50

20

Depth (m)

(c)
50 -100

My (kN.m)
0

100

200

P1
P2
P3
P4

10

30
40

P2

P4

P3

Sign
convention
for Mx
+

Tunnel
axis
+

Sign
convention
for My

P1

50
+Y

-Y

60

Figure 6 Computed (a) induced axial load distribution (b) bending moment in
longitudinal direction, Mx (c) bending moment in transverse direction, My in
each pile due to tunneling
Changes of horizontal effective stress and induced excess pore water pressure.
Figure 7a shows changes of computed normalized horizontal effective stress
(h/h0) with respect to shield advancement. In this study, coupled-consolidation
analysis is conducted which allows both volume and pore water pressure to change
during tunnel advancement. Soil elements around piles at the depth of tunnel axis are
selected to be investigated. As the shield reaches between Y = -10 to -5 m, h/h0
of P1, P2 and P4 increases. This is caused by horizontal effective stress relief around
the shield. Subsequently, arching effect or stress redistribution occurs on soil element
in front of the shield and causes outward movement from the shield. The measured
horizontal ground displacement confirmed the explanation of stress redistribution.
When the shield advances further and reaches at Y= -5 to 3 m, h/h0 suddenly
drops due to stress relief around the shield. Horizontal effective stress around piles
slightly increases when the shield advances beyond the monitoring section (Y = 3 to
25 m). This is due to stress redistribution behind the shield causing a slight outward
movement at the piles. The highest change of h/h0 occurs at P1 with the highest
increment h/h0 of 1% and the maximum reduction h/h0 of -3%. The
tunneling impact in terms of changes of horizontal effective stress on P2 is the lowest
as expected.

556

GeoCongress 2012 ASCE 2012

3127

Figure 7b shows changes of excess pore water pressure at the depth of tunnel
axis which has a similar trend to h /h0. When the shield approaches to the
monitoring section (Y = -20 to -10 m), u/u0 around every pile increases. It indicates
contractive behavior of soil around piles due to stress redistribution induced by
tunnel driving. When the shield reaches at Y= 3 m, u/u0 abruptly decreases to
minimum then it gradually rebounds when the shield advances beyond the
monitoring section. The dilative behaviour suggests that the resultant horizontal
displacement of soil around each pile is towards the shield. The maximum change of
u/u0 occurs on P1 whereas the minimum occurs on P2. The difference of h/h0
and u/u0 among P2, P3 and P4 is not significant. It indicates that the influence zone
of tunneling extends to P1 only.
(a)
-20

-10

1
0
-1

P1

10

20

30

-20

-10

10

20

30

1
0

-1
-2

-2

P2

-3

P3

-3

P4

-4

-4

Shield distance, Y (m)

(b)

u/u0 (%)

'h/'h0 (%)

Shield distance, Y (m)

P2

P4

P3

P1

-Y

+Y

Figure 7. Changes of computed (a) horizontal effective stress and (b) excess pore
water pressure of soil element at depth of tunnel axis
CONCLUSIONS
Based on field observations and computed results, the following conclusions may be
drawn. Within the pile group, P1, P2, P3 and P4 denote the closest, the farthest, the
front and the rear pile to the tunnel, respectively.
1. The maximum measured transverse horizontal ground displacement was 0.2%D
(i.e., 15 mm) at depth of the tunnel axis when the shield was at the monitoring
section. When the shield was approaching to the monitoring section, outward soil
movement away from the shield was observed due to stress redistribution. As the
shield reached the monitoring section, an abrupt change of inward horizontal
ground displacement toward the shield was observed as a result of stress relief
around the shield.
2. The influence zone of load transfer and pile head settlement are computed when
the shield advances from 1D (tunnel diameter) away from the center of the pile
group to 1D beyond the center of pile group. As the shield advances towards the
pier, computed pile load at P2 increases because it is the least affected one by
shaft resistance reduction due to tunneling. On the other hand, loads at pile head
of P1, P2 and P3 decrease because of load re-distribution within the pile group.
The maximum pile head settlement occurs on P3 causing the pier to tilt in the
longitudinal direction of tunnel towards P3.

557

GeoCongress 2012 ASCE 2012

3128

3. Based on computed results, tunneling causes a significant reduction of shaft

resistance in P1 at depth above the tunnel axis. To maintain equilibrium, shaft
resistance below the tunnel axis increases. The influence zone of shaft resistance
reduction is estimated to be 1.5D from the tunnel axis.
4. The maximum computed induced axial force, longitudinal and transverse bending
moments occur in P1 at the depth of tunnel axis when the shield advances at 1D
beyond P1. The largest induced axial force occurs due to the development of
negative shaft resistance. The bending moment of the pile in the longitudinal
direction of tunnel advancement (Mx) causes the pile to bulge in the same
direction as the shield advancement. The bending moment of the pile in the
transverse direction of tunnel advancement (My) causes the pile to bulge towards
the shield due to stress relief around the shield.
ACKNOWLEDGEMENTS
The authors would like to acknowledge the financial support provided by the
General Research Fund 617608 from the Research Grants Council of the HKSAR.
REFERENCES
ABAQUS, (2008). Theory Manual, v 6.8. Hibbitt, Karlson and Sorensen Inc. RI
Chen LT, Poulos HG, and Loganathan N. (1999). Pile responses caused by
tunneling. J. Geotech. Geoenviron. Eng. 125(3), 207-215.
Cheng CY., Dasari GR., Chow YK.,and Leung CF. (2007). Finite element analysis
of tunnelsoilpile interaction using displacement controlled model.
Tunnelling and Underground Space Technology 22, 450466
Lee GTK, Ng CWW. (2005). Effects of Advancing Open face tunneling on an
existing loaded pile. J. Geotech. Geoenviron. Eng., 131(2), 193-201.
Lim S., Teparaksa W. and Shibuya S. (2009). Effects of shield tunneling through
the obstructions in Bangkok. J. of the SEA Geot. Soc., 40 (3): 145-159
Loganathan N. and Poulos HG. (1998). Analytical prediction for tunneling-induced
ground movements in clays. J Geotech Geoenviron Eng, 124(9):846856
Pang CH., Yong KY. and Chow YK. (2005). Three-dimensional numerical
simulation of tunnel advancement on adjacent pile foundation. Underground
Space Use: Analysis of the Past and Lessons for the Future , 1141-1147
Peck RB (1969). Deep excavation and tunneling in soft ground. Organizing
committee (eds) Proc. of 7th Int. Conf. in Soil Mechanics and Found. Eng.,
Mexico City. Sociedad Mexicana de Mecanica de Suelos, 225290
Seah, TH. and Lai, KC. (2000). Strength and deformation behavior of soft Bangkok
clay. Geotechnical Testing Journal, Vol. 26, No. 4
Tanaka H., Locat J., Shibuya S., Soon TT, and Shiwakoti DR. (2001).
Characterization of Singapore, Bangkok, and Ariake clays. Can. Geotech.
J. 38: 378400

558