Underground Singapore 2016

Design & Construction of Large Diameter SCL Tunnel in

Soil
I. Hiroaki
Nishimatsu Construction Co.,Ltd.

C.N Ow & R. Wong D. Poh

Land Transport Authority, Singapore Worley Parsons Pte Ltd

L.W. Weng K.H. Lim & H. Takeshi

Singapore Power, Singapore Parsons Brinckerhoff, Singapore

ABSTRACT: Spray Concrete Lining (“SCL”) tunnels with 10.4m to 10.5m diameter and 9m length
were constructed at approximately 60m below ground level for launching and receiving chambers of
EPB TBM ( diameter 6.88m). These SCL tunnels are connected to a 14m diameter Equipment Build-
ing Shaft for East-West Transmission Cable Tunnel Project EW3. The SCL tunnel for receiving
chamber is one of the biggest diameter SCL tunnel constructed in Soil (Old Alluvium) in Singapore.
This paper presents the design of lining, face stability, construction details and monitoring results of
the SCL tunnel. For SCL tunnel excavation, 1) three stages excavation method (Top heading by Ring
Cut, Bench and Invert), 2) short excavation advance, 3) long fore-poling and 4) dome type end wall
with face bolts were adopted. The SCL tunnel was successfully and safely constructed with effective
control on the tunnel convergence and ground movement.

1 INTRODUCTION

Singapore Power has undertaken the task to develop two new cable tunnels to run the power transmis-
sion cables underground. The project commenced in 2012 and is scheduled for completion in 2018.
The development comprises two main tunnel lines, one running North-South and another East-West of
Singapore with a total length of 35km. The development includes 18 access shafts built with an aver-
age depth of 60m underground. EW3 contract is part of the East-West line with a 5.5km long tunnel
and 3 online access shafts: Paya Lebar Shaft, Airport Road Temporary Shaft and Kallang Shaft.

Two Spray Concrete Lining (SCL) tunnels were constructed at the bottom of a 14m diameter, 60m
deep Kallang Shaft. One of them was of 10.4m diameter constructed as a launching chamber for a
6.88m diameter EPB TBM. The other was 10.5m in diameter and served as a receiving chamber for
another TBM drive from Airport Road Shaft. In this paper the latter SCL tunnel is discussed.

This SCL tunnel has the following features:

1) The biggest diameter SCL tunnel constructed in Soil in Singapore
2) The overburden of SCL tunnel is approximately 50m.
3) Constructed underneath the new Kallang Equipment Building with partial building loads acting on
the SCL tunnel through barrette foundations.

Prior to construction of the two SCL tunnels at Kallang Shaft, a 9.5m diameter SCL tunnel was com-
pleted at Airport Road Temporary Shaft. The soil conditions and the size of that SCL tunnel were
similar to those at Kallang Shaft. The experience and knowledge gathered from SCL tunnelling works
at Airport Shaft provides valuable information and references to enable the design and construction of
SCL tunnels at Kallang Shaft to be executed effectively.
This paper presents the face stability analysis method of SCL tunnel and excavation method based on
that analysis. Comparisons are made on the instrumentation and monitoring results gathered from both
SCL tunnels at Airport Shaft and Kallang Shaft, and the differences in the monitoring results are dis-
cussed.
2 OUTLINE OF SCL TUNNEL IN SOIL

Figures 1 and 2 present the longitudinal and cross section of SCL tunnels at Kallang Shaft respectively.
The tunnel diameters are 10.5m and 10.4m respectively. The tunnel depth is 55m from ground level to
the tunnel spring line level. Excavation of SCL tunnel is in OA (A) layer with average overburden of
18m of OA (A). Tunnel primary lining consists of 350mm thick shotcrete (G40) with lattice girders
spaced at 680mm and 2-layers of wire-mesh.
Fore-Poling
L=12m Ø76.3mm

SCL Tunnel SCL Tunnel

(Launching Kallang shaft (Receiving

hd=5.75m
Chamber) 14.0m Chamber)

Ø=10.5m
10.4m

10.5m

Elephant Foot
9.0m 9.0m

Figure 1 Longitudinal Section of SCL tunnel Figure 2 Cross Section of SCL Tunnel
at Kallang Shaft at Kallang Shaft

Based on the face stability analysis, the following construction methodology was adopted:
a) Excavation length was limited to 0.68m.
b) Three stages excavation sequence was adopted.
1) Top Heading by Ring Cut method
2) Bench
3) Invert
c) Installation of long fore-poling Ø 76.3mm in 12m length and 300mm spacing.
d) At tunnel end, Dome shape wall was adopted with Face Bolts installed.

Figure 3 and Photo 1 present the spray lining support with lattice girder and wire mesh.

Figure 3 Cross Section of Spray Lining Support

Photo 1 Lattice Girder, Wire mesh, Shotcrete

Figure 4 and Figure 5 show borehole locations and soil profile around SCL tunnels at Kallang Shaft
respectively. The site at Kallang shaft is covered with FILL materials of approximately 3-7m thick be-
low ground and underlain by Kallang Formation. The thickness of the Kallang Formation ranges from
26m to 30m around the shaft. Beneath the Kallang Formation is Old Alluvium. SCL tunnel were exca-
vated in Old Alluvium with SPT_N> 100 (i.e. OA(A)). The coefficient of permeability in OA(A) was
in the range of 10-7m/s to10-8m/s. Table 1 shows the design parameters of the soil in which SCL tunnel
was excavated.
 Kallang Shaft

SCL Tunnel
(Receiving Chamber)

SCL Tunnel
(Launching Chamber)

Figure 4 Plan of Kallang shaft and

Borehole locations

Figure 5 Soil Profile of Kallang shaft

Table 1 Soil Types and Soil Design Parameter

Unit Undrained Effective Effective
Weight Shear Stength Cohesion Angle of Friction
Soil Type SPT–N Value
𝛾
′
cu (kN/m2) c’ (kN/m2) φ’(°)
(kN/m3)
Fill - 19 30 0 30
Marine Clay - 15 14 0 22
F1 - 20 0 0 30
OA(D) N< 30 20 150 10 33
OA(C) 30< N <50 20 250 15 33
OA(B) 50< N <100 20 300 15 33
OA(A) 100< N 20 350 15 33

3 METHODOLOGY OF FACE STABILITY ANALYSIS

3.1 Stability Analysis Method

For the face stability analysis of SCL tunnel, “Sliding Wedge Method” was adopted as the “Ultimate
Limit State” of sliding wedge at the tunnel face. Figure 6 shows schematic mechanism of Sliding
Wedge Method. An excavation area of Heading was assumed to be equivalent of a rectangular tunnel
with the same area as Heading area. At the tunnel face, the ultimate state of Sliding Wedge was con-
sidered where loosening pressure of the ground was imposed. The sliding force imposing on the Slid-
ing Wedge was derived from loosening pressure using the Terzaghi’s Silo Theory.

( 𝛾1 , 𝑐1 , ∅1 )

( 𝛾2 , 𝑐2 , ∅2 )

Figure 6 Schematic Sliding Wedge Mechanism at SCL Tunnel Face

The safety factor of face stability (FOS) in term of effective stress can be expressed by the following
equation:

FOS = (1)

Where,
′
𝜎𝑣 : Loosening pressure at the crown of the tunnel using Terzaghi’s Silo Theory
′
𝛾1 : Unit weight of H1 layer
′
𝛾2 : Unit weight of H2 layer
hd: height of heading (m)
′
π ∅
𝜃: angle of sliding 4
+ 22
′
∅2 : internal angle at tunnel centre (OA(A))
′
𝑐2 : effective cohesion of H2 layer
′
fs: coefficient of side friction=1-sin∅2
As1: sliding area of wedge
As2: side area of sliding wedge
Gw: self-weight of sliding wedge
Gs: total loosening pressure acted on the sliding wedge

3.2 Face Stability Analysis

The excavation method of SCL tunnel was carefully determined with the consideration of 1) safety
factor of face stability, 2) bearing capacity of elephant foot and 3) construction efficiency.

3.2.1 Safety Factor of Face Stability

′
A new building load and loosening pressure from ground (𝜎𝑣 ) were imposed on the Lining of the
SCL tunnel. The building load was obtained from distribution load from the toe of barrette foundation
of the new Kallang equipment building. The loosening pressure was calculated using Terzaghi’s Silo
theory. The following three cases were studied to determine the excavation method of the SCL tunnel
as shown in Table 2. For each case, the safety factor of face stability is obtained from Eq(1) above. In
this case study, the safety factor required should be more than 1.2.

Table 2 Case Studies of Excavation Method of SCL Tunnel at Kallang shaft

′
hd 𝒍𝒆 𝝈𝒗 Building Total Load Safety Factor
(m) (m) (kN/m2) Load (kN/m2) FOS
(kN/m2)
Case1 5.75 0.68 40 70 110 1.08
Case2 3.0 0.68 12.5 70 82.5 1.09
Case3 5.75 0.68 0 0 0 2.35
Note: hd: height of heading (m)
𝑙𝑒 : excavation advance (m)

For Case1 heading height hd =5.75m (heading area = approximately 50% of full area of the SCL tun-
nel), excavation advance 𝑙𝑒 =0.68m, the total load =110kN/m2, the safety factor of face stability=1.08
which was not satisfactory.

In order to improve the safety factor, Case2 heading height hd =3.0m which was smaller than that of
Case1, excavation advance 𝑙𝑒 =0.68m, the total load =82.5kN/m2, the safety factor=1.09, which was
again not satisfactory.

Case3 study was carried out with no surcharge load to the face whereby a long fore-poling system was
proposed. The total load was supported by the fore-poling. With that, the safety factor was improved to
2.35 as shown in Table 2, which was more than 1.2 and hence satisfactory.
3.2.2 Elephant Foot

SCL Tunnel was excavated in Old Alluvium without rock bolts. The load imposed on the lining of
heading should be supported by the Elephant Foot (shown in Photo 2). The excavation of Elephant
Foot was made under overhang condition, therefore the shape of Elephant Foot has to be as small as
possible from safety point of view when excavating the overhang portion. In this SCL tunnel, the max-
imum effective width of the Elephant foot was 800mm.

In case of the excavation advance 𝑙𝑒 =0.68m, the safety factor for the bearing capacity of the Elephant
Foot was greater than 1.5.

Photo 2 Elephant Foot at Heading Excavation

3.2.3 Fore Poling

In order to improve face stability, a long fore-poling pipe was adopted to support the loosening pres-
sure and building loads. Under conditions of excavation length 𝑙𝑒 =0.68m, loosening pressure of
40kN/m2 and 70kN/m2 of building load, the following details for fore-poling system were adopted.
 diameter of fore-poling 76.3mm
 fore-poling length 12m
 spacing of fore-poling 300mm with arrangement of pipes in the upper periphery of the tunnel
( 120˚ area)

Photo 3 Drilling for Fore-Poling Photo 4 Face Bolt after Heading Excavation

3.2.4 Face Bolt at End of SCL tunnel

At the end of the SCL tunnel, the fore-poling system is not effective due to the limitation of its length.
The SCL tunnel was served as a launching chamber of TBM for long duration, and the end wall must
keep stable during this period. Face bolts were adopted to enhance the face stability during construc-
tion of end wall. A total of 56 bolts (25mm dia rebar, 5m length spaced at 1m) were installed. The face
bolts were arranged outside of TBM envelope. (shown in Photo 4).
Dome shaped wall was adopted in order to maintain the face stability.
4 DESIGN OF LINING AND DOME OF SCL TUNNEL

4.1 Load Combination

The load combinations of loosening pressure, water pressure and surcharge from Diaphragm wall and
Barrette of Kallang Equipment Building are considered for shotcrete lining and End Wall Dome de-
sign.

The following load combinations are considered as shown in Figure 7.

1) Load 1 (after 24hours for both Heading & Full Section ):
Design load = Loosening pressure without water pressure + Surcharge from Diaphragm wall and
Barrette of Kallang Equipment Building.
2) Load 2 (after 3 days for both Heading & Full Section, after 7 days for Heading and 28 days for
Full Section):
Design load = Loosening pressure + water pressure of loosening height + Surcharge from Dia-
phragm wall and Barrette of Kallang Equipment Building.

a) Load 1
for Top Heading
b) Load 1 c) Load 2 d) Load 2
Legend of Load
for Full Section for Top Heading for Full Section
after 3days &7days after 3days&28days

Figure 7 Load Combinations for Temporary Shotcrete Lining Design

The shotcrete strength / Young’s modulus and design load are shown in Table3.

Table 3 Shotcrete Strength / Young’s Modulus and Design Load of SCL Lining
Age of Compressive E Loosening Water Surcharge Coefficient of
shotcrete strength of (N/mm2) pressure pressure from horizontal earth
shotcrete (kN/m2) (kN/m2) Building pressure
fcu(N/mm2) (kN/m2)
24hours 12 16,000 40 0 68 0.5
3days 23 22,000 40 40 68 0.5
7days 28 25,000 40 40 68 0.5
28days 30 28,000 40 40 68 0.5

4.2 Design of Lining

The structure 2-D ring plain strain model was adopted for design of the SCL lining as shown in Figure
8. At the end of SCL Tunnel, the Dome Shape Structure was constructed to keep the face of the tunnel
stable. The design of the Dome Shape was carried out using a 3-D FEM structural model SAP 2000
version 15 to capture the structural behavior and forces as shown in Figure 9. All the design loadings
and combinations of end wall were adopted from primary lining design.

The soil and structural interaction was simulated by application of nonlinear spring, with stiffness val-
ues equivalent to the stiffness modulus of the respective ground condition.
 a) for Top Heading

b) for Full section

Figure 8 Structural Model for Analysis and Figure 9 Structural 3D Model for
Design of SCL Tunnel Lining Design of Dome Shape End Wall

Based on the results of the above design analyses, the following temporary support was adopted for the
SCL tunnel (shown in Figure 10).
1)Shotcrete is Grade C30N/mm2, and thickness is 350mm with 2 layer wire mesh (D10-100×
100mm).
2)Dome End Wall is 350mm thickness with 2 layer reinforcement D16@100mm c/c at the connection
between the Dome and SCL Lining.

Figure 10 Temporary Support of SCL Tunnel from design calculation

5 CONSTRUCTION SEQUENCE AND CONSTRUCTION RATE OF SCL TUNNEL

5.1 Excavation Sequence

As the SCL tunnel diameter was very large (i.e. 10.5m), the face excavation was divided into 3 phas-
es(i.e. Top Heading, Bench and Invert) in order to maintain the face stability, as shown in Photo 5. At
first, Heading excavation was completed, and followed by the excavation of Bench and Invert. Bench
excavation was ahead of invert excavation by 2 sets of lattice girder (2×excavation advance = 2𝑙𝑒 ).
As for the Heading excavation where the height was very large (5.75m), Ring Cut method was adopted
considering the face stability and working efficiency.

The following construction sequence of SCL tunnel was adopted as shown in Figure 11.
Step1: Installation of fore poling
Excavation of Top Heading using ring cut method till the end wall
Installation of face bolts
Step2: Excavation of Bench and Invert
Step3: Excavation of core at the end wall
Step4: Excavation of remaining Bench and Invert
Installation of face bolts
 Excavation
Heading
Excavation
Bench

Excavation
Invert

Step1 ;Forepoling/Excavate Step2 ;Excavate Bench/

Heading/Face bolts Excavate Invert

Excavation
Core

Excavation
Final
Photo 5 SCL Tunnel Excavation with
Heading, Bench and Invert
Step3; Excavate core Step4; Excavate Final/
Face bolts
Figure 11 Construction Sequence of SCL tunnel

5.2 Excavation Speed

Figure 12 shows Number of Excavation advance and Figure 13 shows the construction rate of SCL
tunnel. In Figure 13, during the idling time after Heading excavation, the Heading excavation of the
opposite SCL tunnel which was connected with the same shaft was carried out.

The construction advance rates of the Heading, Bench and Invert of SCL tunnel were 2.0𝑙𝑒 /day,
2.0𝑙𝑒 /day and 2.5𝑙𝑒 /day respectively (𝑙𝑒 =excavation advance 0.68m). In comparison, the construction
rate of Heading, Bench and Invert of SCL tunnel at Airport Shaft were 1.3𝑙𝑒 /day, 1.4𝑙𝑒 /day and
1.7𝑙𝑒 /day respectively(𝑙𝑒 =1.0m). The main reason behind the faster construction rate at Kallang Shaft
SCL tunnel was the adoption of dry shotcrete for the initial 50mm thick sealing layer (shown in Photo
6).

At SCL tunnel construction at Airport Shaft, all shotcrete works was carried out with wet type shot-
crete. The transportation time from the batching plant was usually quite long and hence the SCL dura-
tion as a whole was affected.
Based on that experience, dry shotcrete for the 50mm thick sealing layer was suggested for SCL tunnel
at Kallang Shaft. As a result, the construction time for the sealing layer was reduced significantly and
the total construction cycle time of SCL tunnel was improved.

Kallang shaft

2.0𝑙𝑒 /day 2.0𝑙𝑒 /day 2.5𝑙𝑒 /day

Heading

Bench
Airport shaft

1.3𝑙𝑒 /day
Invert 1.4𝑙𝑒 /day 1.7𝑙𝑒 /day

Figure 12 Excavation Number of SCL Figure 13 Excavation speed at

Tunnel at Kallang Shaft Kallang Shaft and Airport Shaft
Photo 6 Dry shotcrete for the first sealing layer

6 MONITORING

Figure 14 and Table 4 present the convergence survey results of SCL tunnels at both Kallang Shaft and
Airport Shaft. The maximum convergences recorded were 6.0mm and 10.5mm respectively. The result
of SCL tunnel at Kallang Shaft was 40% less than that of Airport Shaft. It was also observed that, as
expected, convergence for SCL tunnels increased considerably during Bench excavation.

Table 4 Convergence Monitoring Result

Convergence Alert level Work Suspension

(mm) (mm) level (mm)

Kallang
6.0 10.0 15.0
shaft

Airport
10.5 10.0 15.0
shaft

Figure 14 Convergence Result of SCL tunnel

The Alert level and Work Suspension level of convergence of both tunnels were set at 10mm and
15mm respectively. During Bench excavation at Airport Shaft, at one location, the tunnel convergence
exceeded the Alert level. As remedy measure, two rock bolts were installed locally and tunnel conver-
gence at that location was successfully controlled and maintained below the Work Suspension level.

The convergence and ground movement around SCL tunnel at Kallang Shaft were different from those
at Airport Shaft. Main reasons behind such differences were the “Construction Rate” and the use of
“Fore-poling”. The construction rate at Kallang Shaft was approximately 1.5 times faster than that of
Airport Shaft, and the convergence of SCL tunnel at Kallang shaft was 40% smaller than that of Air-
port Shaft. The fore-polings provided temporary supports to the loosening pressure and other loads
imposed on the lining prior to full ring closure, which helped to reduce ground movement. It can be
observed from the monitoring results that immediate shotcreting after Top Heading excavation and
measures taken to reduce the loosening pressure were the key factors to control the ground movement
in the SCL tunnel in soil.
7 CONCLUSION

The SCL tunnel at Kallang Shaft was the biggest SCL tunnel (diameter 10.5m) in soil in Singapore
with additional building loads imposed on the lining during construction.

From design and construction of the SCL tunnel at Kallang Shaft, the following were adopted :
1) The Lining and the Dome of SCL tunnel with thickness 350mm(Grade C30) was able to resist de-
sign load such as loosening pressure(40kN/m2), water pressure equivalent of loosening pres-
sure(40kN/m2) and building load(68kN/m2)

2) Face stability and ground movement were able to be effectively controlled with the following con-
struction method and measures:
a) three stages excavation (top heading, bench and invert), b) excavation advance 0.68m, c) long
fore-poling and face bolts at the end wall .

3) The convergence of SCL tunnel and ground movement due to excavation can be considerably in-
fluenced by the construction duration.

Photo 7 Completion of SCL Tunnel at Kallang shaft

8 REFERENCES

1) Weng Liwei, Poh David, Ishii Hiroaki, Lim Kok Hwee and Hosoi Takeshi : Face Stability of Large Diameter
SCL Tunnel in Soil in Singapore, 19th Southeast Asian Geotechnical Conference & 2nd AGSSEA Conference ,
Kuala Lumpur, 31May -3 June 2016

2) Nishimatsu-KTCCE JV, PERSONS BRINCKERHOFF : Calculation Report Kallang Shaft - Design Report
for Shotcrete Lining Tunnel (SCL), Aug 2014

3) SP POWERASSETS LTD : Contract EW3 Geotechnical Interpretative Baseline Report, 2011