Design Evaluation Assessment

for Natural Gas Transport

Systems
Module 2
Working Stress Design
Outline

Pipeline materials

Working Stress Design

Where are we?
Geotechnical Considerations
Limit States Design
Regulatory Oversight of Strain Based Design

2
 Basis of Mechanical Design
• The basic design requirement is to ensure the safe
containment of the product within the pipeline by
1. Limiting the magnitude of hoop stress to some
percentage of Specified Minimum Yield Strength (SMYS)
through:
• Maximum Operating Pressure (MOP), its
cycles/fluctuations,Wall thickness and diameter
• Considering Longitudinal/secondary loads (e.g.,
bending, axial tension or compression)
Provide appropriate factors of safety
(i.e. a suitable level of conservatism)
 What Threats Are Addressed When Designing
Pipelines?
ROW
Wall External
Threat Loads Materials Coatings Monitoring /
Thickness Protection
Access

External Corrosion √ √ √ √ √
Internal Corrosion √ √ √ √
√
Third-Party Damage √ √ √ √

Stress Corrosion
Cracking √ √ √
Outside Force /
Ground Movement √ √ √ √ √ √
Manufacturing &
Construction √ √ √
Defects
Purpose of Stress Analysis

The structural design of a pipeline system

needs to take into account:
pressure
temperature
external loads
other loads

Pipelines, unlike many other structures, do not

have a redundant load path to compensate for
a failure at some point in the structure. This
makes stress analysis an important component
of the design phase.
Purpose of Stress Analysis

a) The Canadian
pipeline code Z662
(g)loss of support;
provides a list of other
loadings the a (h)excessive overburden
pipeline designer loads and cyclical traffic
needs to consider: loads;
b) occasional extreme (i)construction and
loads, such as inertial maintenance deformations,
earthquake; including those resulting
c) slope movements; from horizontal directional
d) fault movements; drilling;
e) seismic- (j)mechanical vibrations;
related earth
movements; (k)hydraulic shock; and
f) thaw settlement; (l)vortex shedding.
g) frost heave;
Pipelines vs. Piping
From a stress analysis perspective, there are a
number of factors which make pipelines different
from plant piping:
Long, straight runs of pipe between changes in
direction
Axial compression forces due to soil restraint
Lower temperature differentials
Higher D/t ratios
Reduced overall safety factors
Overburden loads
Subjected to soil movements
Pipelines vs. Piping
Comparison of plant piping code ASME B31.3 with pipeline
code Z662:
 NPS 20, Grade 414, ERW, Cat I

 Design Pressure = 9930 kPa

 Design Temperature = 65 deg Celsius

 Station piping (hoop stress utilization = 50%)

wall thickness Z662 B31.3 thickness comparison

25 mm 80%

difference
20 mm 60%
15 mm 40%
10 mm 20%
5 mm 0%
0 mm 241 359 386 414 448 483 550
241 359 386 414 448 483 550 pipe grade
pipe grade 8
Stress Analysis Approaches

Working Stress Design

Limit States Design
 including strain-based design
Probabilistic Design
 Including RBDA

9
 Stress Analysis Approaches
Working Stress Design
“A method of design in which structures or members are proportioned
for prescribed working loads at stresses which are well below their
ultimate values; linear distribution of flexural stresses is assumed”

Resistance x Safety Factor ≥ Maximum Loads

Example:
Calculation of design pressure given a wall thickness (or w.t. given
a pressure) 10
 Stress Analysis Approaches
Limit States Design
“Limit States design” means a reliability-based design method that uses
factored loads (nominal or specified loads multiplied by a load factor) and
factored resistance (calculated strength, based on nominal dimensions and
specified material properties multiplied by a resistance factor)

Factored load = (αG G + αQ Q + αE E + αA A)

Factored Resistance = y Fy

Limit states design is also utilized for offshore pipelines

 Stress Analysis Approaches
Probabilistic Design, e.g. “Reliability Based Design and
Assessment”
… the reliability-based method is the method
that demonstrates the structural adequacy of a
pipeline by making an explicit estimate of its
reliability and comparing it to a specified reliability target.

(for an introduction to both LSD and RBDA, see Module #4

 Load categories
– Load controlled
• The magnitude and direction of the load (force) is independent of the displacement
• The term load-controlled is intended to describe the nature of the load and does not
imply that the load is controlled within a certain limit of SMYS.
• Limiting the magnitude of a load-controlled load to some percentage of SYMS is a design
function
– e.g. pressure controlled to MOP to limit hoop stress to 0.8 SMYS
• Displacement limited by resistance of pipe

– Displacement controlled
• The magnitude and direction of the displacement imposed on the pipeline (or the
environment interacting with the pipeline) are independent of force
– e.g. ground displacement, and resulting stress and strain of the pipeline in response
to the displacement-controlled event
• Resistance force incapable of limiting displacement

• WE WILL ONLY LOOK AT THELOAD CONTROLLED SCENARIO INTHIS MODULE. IT IS

REFERRED TO AS ALLOWABLE or WORKING STRESS DESIGN

• In MODULE 4 WE WILL EXAMINE THE SECOND SCENARIO DISPLACEMENT CONTROL

• .
Load categories
• The key difference between load-controlled and displacement-
controlled loads is the strength based behaviour and compliance based
behaviour.

– For strength based behavior, the design criteria and failure

condition are primarily based on strength (e.g. stress) and
deformation is secondary.

– For compliance based behavior, the design criteria and failure

condition are primarily based on deformation (e.g. strain) and
stress is secondary.
Codes and Standards
Codes and Standards
• Onshore Pipeline Regulations (NEB - Federal Regulations)

• Title 49 of the Code of Federal Regulations (CFR), Parts 190, 191, 192,
and 199 (US Regulations)

• ASME B31.8 etc;

• National / Provincial Standards

• CSA Z662 (Standard)

General Issues
The goal of all of these regulations is to ensure that pipeline operators:

– deliver product safely and reliably to customers;

– provide training and written instruction for employees;

– establish written procedures to minimize the hazards resulting from pipeline

emergencies; and

– keep records of inspection and testing.

 Mechanical Design
Pipe specifications are selected based on these
criteria:

– code requirements
– standardization to minimize purchase and
inventory costs
– Material weldability /constructability given the
experience of available construction crews.
 PIPE PARAMETERS
(standard pipe sizes)
Strength, specified minimum yield (SMYS) — the minimum yield
strength prescribed by the specification or standard to which a
material is manufactured.

Stress, design operating — the calculated hoop stress developed

in a pipe by the maximum expected operating pressure, based on
nominal dimensions.

Stress, hoop — the stress in the wall of a pipe or component that

is produced by the pressure of the fluid in the piping, any external
hydrostatic pressure, or both, and that acts in the circumferential
 D

t CL of pipe

Force Equilibrium in a Pressurized Thin Pipe

Assumed to be “cut” at its horizontal diameter
Considering Force Equilibrium:

Net Upward Force due to Pressure =P *D* 1 (assuming unit length)

Net Downward Internal Force in Pipe wall = 2*t *Sh *1

From which we obtain the “hoop” stress Sh = P *D

2*t

This is an acceptable approximation in that it conservatively

estimates Sh based on the external diameter and does not
take into account the variation in hoop stress across the wall
thickness . A more accurate representation would be:
Sh = Pi *Di

2*t
Note for thin wall pipe (D/t ratios > 10 ) the radial stress is ignored because it
cannot be greater than Pi for an onshore pipeline & so is quite a bit smaller
than sh = PD/2t. It could be important in offshore pipelines

sr

ro

sh

ri

Pi

Pe

Tangential and Radial Stress in a Thick Wall Pipe

Consider how large these forces are for a 30” diameter
pipeline with an internal pressure of 15MPa
Assuming a section 1 metre long

The tension force in the pipewall must be equal to:

(15MPa) * (30/39.37 m) * 1m
= 11.43 MN
or approximately 1100 tonnes or more than the take off
weight of three 747 airliners.

Which is why we need to be especially careful to prevent

line strikes puncturing the pipe, or a rupture event due to
corrosion or fracture.
 Mechanical Design

Design Pressure

As specified in B31.8 and CAN Z662-99, the maximum design pressure for
steel pipeline systems, or the nominal wall thickness for a given design
pressure, is determined as follows:

2St
P = x 10 3 x F x L x J x T
D

The factors F,L,J and T are various usage factors defined as follows:
 Mechanical Design
Where:
P= design pressure, kPa
S= specified minimum yield strength, as specified in the standard or specification, MPa.
t= design wall thickness, mm
D= outside diameter of pipe, mm
F= design factor (0.8 in CSA and ASME B31.8)
L= location factor
J= joint factor
T= temperature factor
 Basis of Design : Liquid Pipelines
•Oilpipelines; no account
taken of population density
•Possible to design to high
design factors in built-up
areas
•No specified distances of
population from the
pipeline
Class Location
Class location designations are determined
on the basis of class location assessment
areas and on the buildings, dwelling units,
places of public assembly, and industrial
installations contained in such areas.
Class Location
Designing according to class location is intended as a risk
mitigation method. Areas of high consequence (high
population density) have the most stringent design conditions
(Class 4). Areas of lower risk have less stringent design
conditions imposed.
 Class Location
When determining class location designation, due
consideration should be given to the possibility of future
development in the area. If, at the time of planning a new
pipeline, development sufficient to change the class
designation in the future appears likely, this needs be
taken into consideration in design and pressure testing.
Companies are responsible to ensure the pipeline meets
the requirements of changes in class location
Class location determination
Class Location

200 m
200 m
Class 3
Class Locations
Note that the class location definition applies to
pressure testing as well. The highest hoop stress
levels a pipeline may ever experience will occur during
testing.

Designing according to class location is intended as a

risk mitigation method. Areas of high consequence
(high population density) have the most stringent
design conditions (Class 4). Areas of lower risk have
less stringent design conditions imposed.
Classification of Location
Design Factor Onshore Pipelines
Class location
determination
Class Locations
Note that the class location definition applies to
pressure testing as well. The highest hoop stress
levels a pipeline may ever experience will occur during
testing.

Designing according to class location is intended as a

risk mitigation method. Areas of high consequence
(high population density) have the most stringent
design conditions (Class 4). Areas of lower risk have
less stringent design conditions imposed.
 Mechanical Design
Distribution systems operate at pressures which are well below the theoretical design
pressure of steel pipe given by the above formula.

The limiting factor for minimum wall thickness is that which is required to prevent
buckling, flattening and bending during installation. These could result in operational
problems later such as difficulty pigging the line. B31.8 and CSA Z662 both specify
minimum requirements for wall thickness based on that criterion.
 Steel piping systems design requirements

Pressure Design of Straight Pipe

The design pressure for pressure steel gas piping systems or the
nominal wall thickness for a given design pressure is determined by
the following formula
P = 2St FLET
D

Where:
P=design pressure (psig);
S=specified minimum yield strength, psi, stipulated in the
specifications under which the pipe was purchased from the
manufacturer.
Steel piping systems design requirements
• D=nominal outside diameter of pipe (in);
• T=nominal wall thickness (in);
• F= design factor =0.8
• L = Location factor Table 4.1 of CSA Z662.
• E=longitudinal joint factor obtained from Table 4.2;
• T=temperature derating factor obtained from Table 4.3
• There are equivalent tables to these in all pipeline codes
Basic Design factor F

Location Class Location factor L

Location Class 1 0.9
Location Class 2 0.8
Location Class 3 0.625
Location Class 4 0.5
Joint and Temperature derating factors in
CSA
Origin of Class 1 Design factor, F=0.72

• The origin of the 72% SMYS operating stress

is an agreement in 1940 between line pipe
users and A.O. Smith to operate at 80% of the
mill test pressure of 90% of the yield strength
(SMYS) of their improved flash welded cold
expanded pipe
 Mechanical Design

Design Pressure

Repeating :As specified in B31.8 and CAN Z662-99, the maximum design
pressure for steel pipeline systems, or the nominal wall thickness for a
given design pressure, is determined as follows:

2St
P = x 10 3 x F x L x J x T
D

In addition to circumferential stress the pipe can also

experience axial or longitudinal stress
Longitudinal Stress
• An operating Pipeline carries longitudinal as well as hoop
stresses. The first is due to the Poisson effect whereby a metal
bar stretched in tension will contract in the transverse
direction (think of conservation of material). If only
circumferential (hoop ) stress exists the pipe will extend
circumferentially (its diameter would increase) and contract in
the axial direction. If this longitudinal contraction was
prevented by say soil friction or an anchor block then a
longitudinal stress will be induced its magnitude would be :
» n Sh or 0.3 Sh
– Since n is 0.3 for steel

• The second effect is due to temperature . As the temperature

increases the pipeline will want to expand and if it is
restrained from doing so an axial stress is caused
Stress due to temperature
If the pipe is placed in the trench at a
temperature lower than the flowing
temperature of the fluid , then upon start up
the temperature differential will cause an
expansion which if restrained will create a
compressive stress

- Ec α (T2 – T1)
Stress due to temperature
• This axial stress will have a Poisson component
whose effect will be to increase the circumferential
strain and hence stress.
• Suppose the pipe in the last example underwent a
90C temperature increase
• sA = -EaDT= 2.4 *103 (kN/m2/C)*90C
= -216 MPa
and the circumferential component is
64.8MPa
Longitudinal Stresses due to pressure

t
P D

Longitudinal Force Equilibrium due to

Internal Pressure and end cap restraint
Longitudinal Stresses due to pressure

• Suppose the Pipe is capped at its end or is

constrained to rise vertically above ground
• Balancing the forces on the end cap with
those in the wall of the pipe we have:
Pp D2/4 = s PD t
Or s = PD = 0.5 s h
4t
• The pressure within a thin walled pipe that is
either capped, as shown, or rises vertically at
its ends ,creates stresses in both the
circumferential and longitudinal directions as
shown in the following figure.

D = internal diameter
t = wall thickness
P — net internal pressure
• Determine the hoop and longitudinal stresses
in a pipe of 323.8mm outside diameter and
6mm wall thickness if the maximum operating
pressure is 9438kPa and the operating
temperature is 20C?.
PD 9438 x323.8
Sh    254.67 MPa
2t 2 x6

PD 9438 x323.8
SL    127.33MPa
4t 4 x6
Restrained v Unrestrained Conditions

Generally, “unrestrained” means that the pipe is able to strain

along its length AND move laterally.

Pipe that does not meet both of the above requirements is

referred to as restrained. Generally soil friction provides this
constraint, however sufficient length of pipe needs to be buried
for this to occur.
Restrained v Unrestrained Conditions
• Restrained piping is not free to move in response to
stresses (buried piping of sufficient length is restrained);

• Unrestrained piping is free to move in response to forces

(above grade pipe runs are typically un-restrained)
Installation Stresses
 Thermal Stresses
Temperature

The design temperature range for each segment of the

pipeline system needs to be specified by the designer
for the conditions expected during installation,
pressure testing, start-up, and operation. As an
example, for carbon dioxide pipelines, extremely low
temperatures may be encountered during pressure-
relieving or pressure- reducing situations
 Design Temperature

Design temperature is the

lowest possible temperature
which the pipeline will be
exposed to while under
pressure. This will typically
occur during testing or shut in
operation for gas and oil
systems.

Note: Gas pipelines are typically

designed for -5°C for below
grade installations and -45 °C
for above grade.
60
Thermal Stresses

Thermal Expansion and Contraction Loads

 Thermal Stresses
The design must account for the temperature difference
between when a pipeline is first retrained and when it is
operating. This stress differential could well be the governing
stress factor governing wall thickness.
Stress due to temperature
If the pipe is placed in the trench at a temperature
lower than the flowing temperature of the fluid ,
then upon start up the temperature differential will
cause an expansion which if restrained will create a
compressive stress

- Ec α (T2 – T1)
Stress due to temperature
• This axial stress will have a Poisson component
whose effect will be to increase the circumferential
strain and hence stress.
• Suppose a pipe underwent a 90C temperature
increase
• sA = -EaDT= 2.4 *103 (kN/m2/C)*90C
= -216 MPa
and the circumferential component is
64.8MPa
Reducing Soil Loads :Expansion Loops
or Loose backfill

For typical pipeline trench conditions, loose granular backfills (sand or gravel) will offer
less resistance to pipe movement than compacted cohesive backfill materials (clay or silty clay). A
granular material with an angle of internal friction of 35° or less is recommended.

65
Biaxial Membrane Stress
Code Requirements (Z662, 4.7.1):
Unless special design measures are implemented to ensure the stability of the
pipeline, the hoop stress due to design pressure combined with the net longitudinal
stress due to the combined effects of pipe temperature changes and internal fluid
pressure shall be limited in accordance with the following formula:
Sh – SL ≤ 0.90 S × T

Note: This formula does not apply if SL is positive (i.e., tension).

where
Sh = hoop stress due to design pressure, MPa, as determined using the formula
given in Clause 4.6.1.5
SL = longitudinal compression stress, MPa, as determined using the following
formula:
SL = ν Sh – Ec α (T2 – T1 )
T2 = maximum operating temperature, °C
where
Ν = Poisson’s ratio T1 = ambient temperature at time of restraint, °C
E = modulus of elasticity of steel, MPa S = specified minimum yield strength, MPa
α = linear coefficient of thermal expansion, °C– T = temperature factor (see Clause 4.3.3.5.)
1
Biaxial Membrane Stress

Example calculation 1:
High pressure gas pipeline Z662 Wall thickness calc:
NPS 16; Grade 414
DP = 14,400 kPa t = PD/2SFLT
Install temp = -20 deg C = 14.4 x 406/(2 x 414 x 0.8 x 1 x 1)
Design temp = 30 deg C
Location Factor = 1.0
Result t = 8.9 mm
:
Combined hoop and longitudinal stress calc:

SH – SL ≤ 0.9 SMYS SH = 14.4 x 406/2 x 8.9 = 329 MPa

SL = 0.3 x 329 - 207,000 x 12E-6 x (30 - - 20) = -26 MPa

329- -26 = 355 MPa ≤ 0.9x414 = 372 MPa PASS

60
Biaxial Membrane Stress

Example calculation 2:
High pressure gas pipeline Z662 Wall thickness calc:
NPS 16; Grade 414
DP = 14,400 kPa t = PD/2SFLT
Install temp = -20 deg C = 14.4 x 406/(2 x 414 x 0.8 x 1 x 1)
Design temp = 65 deg C
Location Factor = 1.0 Result: t = 8.9 mm

Combined hoop and longitudinal stress calc:

SH = 14.4 x 406/2 x 8.9 = 329 MPa

SL = 0.3 x 329 - 207,000 x 12E-6 x (65 - - 20) = -113 MPa

329- -113 = 442 MPa ≤ 0.9x414 = 372 MPa FAIL

Biaxial Membrane Stress

How much wall thickness do we need for a given ΔT?

•Temperature stress term remains constant

•allowable stress remains constant
=> Only hoop stress changes

372 – 211 = 161 MPa

161 = SH – 0.3 SH SH = 230 MPa
t = 12.7mm
Combined Stresses
nSh EaDT sB

Sh

0.5sh
Or sa
 THEORIES OF FAILURE

• The purpose of failure criteria is to predict or

estimate the failure/yield of a structure.
• A considerable number of theories have been
proposed. However, only the Tresca maximum
shear stress theory used in the Pipeline design
codes is discussed here
 Principal Stresses
Plane State of Stress
• In design we find the stresses in the surfaces of thin-walled pipes
due to combination loading e.g. internal pressure and
temperature . Pipes have one dimension quite a bit smaller than
the other two which means that they have one principal stress*
that is much smaller than the other two. By assuming that this
small principal stress is zero, (it is actually equal to the pressure)
the three-dimensional stress state can be reduced to two
dimensions. Since the remaining two principal stresses lie in a
plane, they are known as a plane stress state.
* Within the pipe surface there are planes on which only normal
stresses exist (i.e. no shear stresses) these are known as principal
stresses
Principal Stresses
The negligible principal stress is oriented in the z-direction.
(thickness) . It is clear that when comparing the hoop and
longitudinal stresses on the pipe surface to the element stresses
that sx = sA and sy = sH ,that they are perpendicular and we can
assume that there are No accompanying shear stresses
 Principal Stresses

• In general the element may be rotated about both x

and y axes into a unique principal orientation in which
all shear stresses vanish. The corresponding normal
stresses in this principal orientation are termed the
principal stresses. A stress state is characterised most
succintly by the principal stresses, say ( σ1, σ2 ), and
failure theories - the next step in failure assessment -
are expressed in terms of principal stresses. It is
therefore necessary to examine how principal stresses
are derived from Cartesian components.
The 2D Stress State
• In general the 2 D stress matrix is given by

where txy = tyx for static equilibrium .

The coordinate directions chosen to analyze any structure
are usually based on the shape of the structure. As a result,
the direct and shear stress components are associated with
these directions. In a pressurised pipe the hoop and
longitudinal directions are perpendicular
 Principal Stresses

• The maximum and minimum principal

stresses are given by:

Or in terms of the hoop and longitudinal stresses

s 1, 2 
s H sL

s H  s L  2  t 2
HL
2 4
 Principal Stresses
• Simplifying when tHL = 0 then the maximum and minimum
principal stress are clearly

s 1,  s H s 2,  s L
The maximum shear stress is equal to one-half the difference
between the two principal stresses i.e.

t max,  s H  s L  / 2
Graphically we can see this using Mohr’s Circle of stress
 Mohr's Circle
• Introduced by Otto Mohr in 1882, Mohr's Circle
illustrates principal stresses and stress transformations
via a graphical format.
Here it is clearly seen
that the diameter of the
circle 2 txy = difference
in the two principal
stresses
 The Limiting Shear Stress
The limiting value of shearing stress (the yield strength in
shear) is:
sA

sH sH

t max  s y 2 t max
sA
P
Or : s1

s h s A  s y
s3 s2 s1

These relationships in fact are the Tresca failure criteria

Worked Example Combined Stresses
• An NPS 12 pipeline has a wall thickness of 11.9mm, a design
pressure of 12,000kPa and is constructed from Grade 359
steel. The design factor may be taken as 0.8, the location
factor 0.625 , while the joint factor J and temperature
derating factor T are both equal to 1.0. The pipe is buried
and may be taken as fully restrained . The temperature at
the time of installation was -20C while the design opeating
temperature is 10C. Young’s modulus may be taken as 207*
103Mpa, Poisson’s ratio is 0.3 and the coefficient of linear
expansion of steel 1.2 * 10-5 / C Determine (i) the required
wall thickness, (ii) the hoop and longitudinal stresses in the
pipe , and (iii) the combined stress .
Worked Example Combined Stresses
The required wall thickness t =
Pd 12000 * 323.9

2S y FJLT 2 * 359000 * 0.8 * .0.625 * .1. *1
t = 10.83mm
The hoop stress
SH = Pd = 12000 * 323.9
2t 2* 11.9
Hence SH = 163.3 Mpa
 Worked Example Combined Stresses
The longitudinal stress SL is found from :
S  n S  Ea (T  T )
l H 2 1

SL = 0.3 * 163.3 – 207 * 103 * 1.2 * 10-5 (10 - (-20))

SL = 48.99 -74.52 = -25.53Mpa
Worked Example Combined Stresses
• The combined stress is SH - SL = 163.3 – (-
25.53) = 188.83Mpa
• The combined stress as a function of the
yield stress is 188.83/359 = 52.6%
• Note the code requirement in CSA Z 662 and
ASME B31.8 is SH - SL < 0.9 Sy * T < 0.9 *359
* 1 hence OK
Pipe Supports

The forces and moments transmitted to connected equipment, such as

valves, separators, tanks, pressure vessels, and compression or pumping
equipment, must be kept within safe limits.

Supports should be designed to support the pipe without causing excessive

local stresses in the pipe and without imposing excessive axial or lateral
friction forces that might preclude the desired freedom of movement.

Where vibration of piping during operation is anticipated, suitable

precautions have to be taken to minimize and control it by installing devices
such as dampeners and braces.
Pipe Supports

Design of Pipe-Supporting Elements

Attachments to the pipe should be designed in such a manner as to
minimize additional stresses in the pipe wall. Nonwelded
attachments, such as pipe clamps and ring girders, are preferred
where they will fulfil the supporting or anchoring function. Where
welded attachments are required for pipelines designed to operate at
stress levels of more than 50% of the specified minimum yield
strength of the pipe, such attachments have to be welded to a
separate cylindrical member that completely encircles the pipe, and
such encircling member shall be welded to the pipe by continuous
circumferential welds.
 Flexibility & Stress
Stress Design of Restrained and Unrestrained Portions of
Pipeline Systems

Due to the fundamental differences in loading conditions

and structural behaviour for restrained portions of
pipeline systems and unrestrained portions not subject to
substantial axial restraint, different limits for allowable
longitudinal expansion stresses and analysis methods are
necessary.
Expansion Loop

u
Sizing an Expansion Loop
Before the widespread use of computer-based methods of
structural analysis, designers had to rely on hand calculations
and a number of simplifying assumptions in order to estimate
pipe flexibility. The ASME B.31 code only deemed such
calculations, or model testing, to
be necessary in those circumstances where some doubt
existed as to the adequate flexibility of the proposed system.
The codes indicate that adequate flexibility may be assumed if
a piping system is:
• of uniform size
• has two anchor points
• has intermediate restraints and satisfies the following
approximate criterion:
Sizing an Expansion Loop
Sizing an Expansion Loop
Sizing an Expansion Loop

Consider the sizing of an expansion loop required for a

NPS 16 pipe of API 5LX60 material supported
between anchor points 60 m apart (Figure 7-4). It is
assumed that the installation temperature is 150C
and the operating temperature is 800C.
For a carbon or low alloy steel the rate of thermal
expansion for the piping between 150C and 800C is
approximately 75 mm/100 m. The restrained anchor
movement y is therefore
 Sizing an Expansion Loop

Substituting D, L, U, and y into the earlier Equation and noting that L = U + 2 h.

 Flexibility & Stress
Note: CSA Z662 nor B31.8 do not define what is meant by retrained pipe. In
these notes, “unrestrained” means that the pipe is able to strain along
its length and move laterally. Pipe that does not meet both of the
above requirements is referred to as restrained.

Typically, long straight lengths of buried pipe and aboveground pipe

on closely spaced rigid supports are classified as restrained, whereas
buried pipelines adjacent to bends or unanchored end caps could be
regarded as restrained or unrestrained, depending on specific
circumstances.
Transition from Full to Unrestrained
. . . .
. . . .
. A . . B
C

. . . . .
.
. .
. . .
Point of Depth . L1 .
.
.
.
Fully Restrained Unrestrained

Max Soil Restraint = FANCH

Soil
Restrained
 S
A

Stress, Strain
s
s

eA

e e
B

Distance
A L B

SB

Stress and Strain Between Points A and B

 CL sLB
Stress, Strain

Anchor
S

L
C
0
Distance
sLA

Distribution at Anchor Location

 Anchor Block

Bearing
Surface

Anchor Flange

. . . . .
. .
. . .. . . 45 . .
. .
. . . .. .. . .

. .. . .
. . .. . . . . Pipe
. . . . .. .
. . . . . .

Backfill
Direction of
Force on Soil
Reinforced
Concrete

Undistrurbed
Soil

Plan View of Anchor Block

 Pipe
Anchor Reinforced Backfill
Flange Concrete
. . . . .
. . . . .
. . .
. . . . . . .
. . Undisturbed
. . . .
. . . Soil
. . .
. . .
Direction of
45 Force on
Block

Anchor
Block

Cross-Section through Anchor Block

Pipeline Anchor Design

Pipeline anchors consists of a flange welded to the pipe and encased

in a concrete block which transfers the anchor load to the soil. The
commonly used design method is based on a Taylor Forge
publication.

Source: Pipeline Rules of Thumb Handbook, pg.

82
Pipeline Anchor Design

The design of the steel flange is the job of the pipeline engineer. The
design of the concrete anchor block is done by the structural engineer.
The steel flange is sometimes a slip-on flange, but it is better to use
purpose-designed pipeline anchor flanges as shown in the pictures.
Pipeline Anchor Design

The Taylor Forge design calculations are based on the ASME

Boiler and Pressure Vessel Code design rules for bolted flange
connections. The process has four steps:

1.Calculation of the anchor force

2.Calculation of the bending moment on the flange
3.Calculation and look-up of shape constants from various
design sheets
4.Calculation of the flange stresses and comparison against
allowables

The following calculations use the Taylor Forge nomenclature.

Anchor Design Example

Calculation Example:

NPS 16, Grade 414, 8.4mm wt DP = 10.2 MPa

Install temp = 10 deg C Design temp = 40 deg C
Assume anchor flange OD = 1.5 pipe OD = 24 in.

The following calculations use the Taylor Forge nomenclature.

Anchor Design Example
Step 1: Calculation of Anchor Force

W = Awall (σ B – σ A)
= Awall [SH(0.5-ν) + EαΔT]
= pDt [0.2 SH + EαΔT]

W = pDt [0.2 SH + EαΔT]

= 1,326 kN
= 300 kips
Anchor Design Example

Step 2: Calculation of design moment on flange

h t 2  A3  C 3 
N   2 
2 
W
3 A C 
N C
g1
g0 hg 
2
W hg
M
B
AD C B N
Anchor Design Example

Step 3: Shape constants

KA => Use Design Sheet I to get T,Z,Y,U

B
g1
; h  Bg ; h Use Design Sheet II to get F, V
g0
0 0
h0  use Design Sheet III to get f

eF
h0

2
2Uh 0g 0
d
V
Anchor Design Example

Step 4: Stress calculations

Assume flange thickness: t = 2 in.

Calculate stress factors: Calculate flange stresses:

α = t e +1 =
SH = f M / (2  g12 ) =
1 = α/T =
SR = α M / ( t2) =
δ = t3/d =
ST = MY / t – Z SR =
2
 = 1 + δ =
SS = W / (p D t) =

Check bearing stress on concrete:

SH = force/bearing area =

70
Anchor Design Example

(cont.) Step 4: Stress calculations – check against allowables

Z662 does not provide stress limits for structural attachments.

 Use another pipe code or BPV code

B31.4 , para.402.3.1 (f):

“Allowable tensile and compressive stress values
for materials used in structural supports and restraints shall not
exceed 66% of the specified minimum yield strength. Allowable stress
values in shear and bearing shall not exceed 45% and 90% of the
specified minimum yield strength, respectively.”
Anchor Design Example

(cont.) Step 4: Stress calculations – check against allowables

Use flange material A 105; SMYS = 36 ksi (248 MPa) per B31.3, Table
K1

Check flange stresses:

Tensile/comp. limit: SH , SR, ST ≤ 66% SMYS = 23.7 ksi
Shear limit:: SS ≤ 45% SMYS = 16.2 ksi
Flexibility & Stress
Hoop Stress
The hoop stress used in stress analysis for any
location on the pipeline system is calculated using
the following formula:
PD -3
Sh = x 10
2tn
Sh= hoop stress, Mpa
tn = pipe nominal wall thickness,
P= design pressure, kPa
D= outside diameter of pipe, mm
 Flexibility & Stress
Steel Properties
Flexibility calculations are based on the modulus of elasticity at the
lowest expected pipe temperature. For carbon and high-strength low-
alloy steels for temperatures up to 230°C, one can use

– 207 000 MPa for the modulus of elasticity (Ec),

– 12 x 10–6/°C-1 as the linear coefficient of thermal expansion
(α), and
– 0.3 for Poisson’s ratio (n).
 Flexibility & Stress
Restrained Spans
Combined Hoop and Longitudinal Stresses
Unless special design measures are implemented to ensure the
stability of the pipeline, the hoop stress due to design pressure
combined with the net longitudinal stress due to the combined
effects of pipe temperature changes and internal fluid pressure is
limited in accordance with the following formula:
Sh – SL < or = 0.90 S × T
where
Sh= hoop stress due to design pressure, MPa,
SL = longitudinal compression stress, MPa, as determined using the following
formula:

Note: This formula does not apply if SL is positive (ie, tension).

 Flexibility & Stress
Restrained Spans
SL = n Sh – Ec α (T2 – T1)
where
n = Poisson’s ratio
Ec = modulus of elasticity of steel, MPa
α = linear coefficient of thermal expansion, °C–1
T2= maximum operating temperature, °C
T1= ambient temperature at time of restraint, °C
S = specified minimum yield strength, MPa
T = temperature factor
 Flexibility & Stress
Restrained Spans
For those portions of restrained pipelines that are
freely spanning or supported above ground, the
combined stress is limited in accordance with the
following formula:

Sh – SL + SB < S × T
 Flexibility & Stress
Restrained Spans
where
Sh = hoop stress due to design pressure, MPa
SL= longitudinal compression stress, MPa.
SB= absolute value of beam bending compression stresses resulting from
live and dead loads, MPa
S= specified minimum yield strength, MPa
T= temperature factor
Flexibility & Stress
Restrained Spans
The maximum beam bending compressive stress
should be calculated using a beam column
analysis, and the stability of such free spans
shall be checked to ensure that the axial
compressive load due to pressure plus positive
temperature differential is less than 0.8 of the
elastic critical buckling load.
Flexibility & Stress
Restrained Spans
Anchors and Restraints
Expansion of a partially restrained buried pipelines
may cause unacceptable movement or stress or
strain at points where such pipelines terminate,
change direction, or change size. Unless such
unacceptable movements are restrained by suitable
anchors or ground restraint, the necessary flexibility
has to be provided.
Flexibility & Stress
Restrained Spans
Unrestrained portions of pipeline systems are designed to have sufficient
flexibility to prevent thermal expansion and contraction from causing
excessive stresses in the piping material, excessive bending and unusual loads
at joints, and undesirable forces and moments at points of connection to
equipment, anchorage points, and guide points.

Where thermal expansion and contraction can occur, flexibility should be

provided by the use of bends, loops, or offsets, or else the thermal strains are
absorbed by expansion joints or couplings of the slip joint or bellows type.
Where expansion joints are used, anchors or ties of sufficient strength and
rigidity have to be installed to withstand end forces due to fluid pressure and
other causes.
 Flexibility & Stress
Restrained Spans
The stresses due to thermal expansion for those
portions of pipeline systems without axial
restraint are combined in accordance with the
following formula:

SE  2
Sb  4 St
21/2
 Flexibility & Stress
Restrained Spans

where
SE = thermal expansion stress range, Mpa

Sb= iMb x 103 resultant bending stress, Mpa

Z
S t= Mt x 103 , torsional stress, MPa
2Z
where i= stress intensification factor
M b= resultant bending moment, Nm
Mt = torsional moment, Nm
Z = section modulus of pipe, mm3
 Flexibility & Stress
Restrained Spans
The thermal expansion stress range, based on 100% of the
expansion, is limited in accordance with the following
formula:

SE </= 0.72 S x T

where

SE = thermal expansion stress range, MPa

S= specified minimum yield strength, MPa
T= temperature factor
 Flexibility & Stress
Restrained Spans
The sum of the longitudinal pressure stress and the
total bending stress due to sustained force and
wind loading are limited in accordance with the
following formula:

0.5 Sh + SB < /= S × F × L × T
 Flexibility & Stress
Restrained Spans
where

Sh = hoop stress due to design pressure, MPa.

SB = absolute value of beam bending compression stresses
resulting from live and dead loads, MPa
S= specified minimum yield strength, MPa
F= design factor
L= location factor
T= temperature factor
 Flexibility & Stress
Restrained Spans
The effects of restraints, such as support friction, branch
connections, and lateral interferences, shall be considered
in the stress calculations. The coefficient of friction of
steel-to-steel contact for support design shall be a
minimum of 0.4 unless special measures are implemented
to reduce friction.

Stress intensification and flexibility of components should be

considered in the stress calculations.
Burial Depth

The depth of burial of gas distribution mains

depends on a number of factors including:

• government regulations and standards;

• loading considerations;

• the depth of existing utilities; and

• assigned depth zones.

 Cover & Clearance

In the US, approximately 40% of pipeline leaks breaks

and ruptures are attributed to unauthorized contact
with the pipeline (outside force, third party damage).
Cover and clearance requirements are aimed at
preventing these incidents.

Designers are cautioned that these requirements often

vary by regulatory authority.
Burial Depth
Government Standards

Municipalities in Alberta, in conjunction with the

utilities, have developed standard burial depths for
the various utilities. The standard depth zone in
Calgary for natural gas is from 0.6 m to 0.9 m.

Pipe is usually buried to the bottom of the depth

zone, or slightly deeper if necessary to ensure a
minimum 0.6 m cover. The gas utility has the
shallowest depth zone, so it is always installed last
in a new development.
Burial Depth

Roads

Typically, road crossings require a minimum 1.4 m

cover beneath the lowest point in the ditch.

Railroads

Railroads require a minimum burial depth of 2.0 m

below the base of the rail, or 0.9 m below the ditch,
whichever gives the greater depth.
Minimum Cover Clearance
• Minimum cover for
mains 24 inches
(600mm). If this cannot
be met or crossing
loads might be
excessive pipe must be
bridged. Care must also
be taken to determine
local ploughing depths
 Burial Depth
Irrigation Canals

Most canal owners require a minimum 2.0 m depth

below the canal
bed. This cover is intended to provide protection for
the pipeline if the canal is dredged to increase it’s
hydraulic efficiency.
Streams and Rivers

These crossings are buried below the 1:50 year flood

scour depth. Hydrologists are routinely hired to
assess stream crossings and provide the most suitable
burial depth. The region of deep installation is
determined by predicting potential changes in the
location and depth of the stream bed.
 Burial Depth
Loading Considerations

Traffic loading is a greater concern for PE than for steel

mains, because PE has a lower strength and is more
susceptible to flattening.

Research has shown that a minimum of 0.6 m of well-

compacted cover of suitable backfill material is required
to limit the deformation of PE mains to acceptable levels.

Loading effects must be considered for crossing

designs. Programs such as PCPISCES can be used for
this purpose.
Burial Depth

Depth of Existing Utilities

The burial depth of gas mains may have to be

altered when crossing other existing utilities. The
recommended minimum clearance from other
structures is 0.3 m (horizontal and vertical).
Burial Depth

Service Line Depths

Service lines are installed at the same depth as the

service tee outlet from the gas main to the property
line when the gas main is located in a public
thoroughfare. Inside property, service lines are
installed with a 0.6 m depth of cover.
Minimum Cover Clearance
• Minimum clearance to adjacent non pipeline
structures:
• 300mm for pipelines,
• 50mm for mains otherwise use insulating
material or casing
Cover & Clearance

Buried pipelines should have minimum depth of

cover such as that shown in the following Table
except where underground structures or adverse
subsurface conditions prevent it. In that case, it is
permissible to install such pipelines with less cover,
provided that they are appropriately protected
against anticipated external loads.
Surface Loading
• The design methods that follow may be used to determine
vertical pressures on the pipe from surface loads. The load is
distributed through the soil such that there is a reduction in
pressure with increasing depth or horizontal distance from
the surcharge load area. The pressure at a point beneath the
surcharge load depends on the magnitude of the load and
on the surface area over which the surcharge is applied.
Usual design practice is to equate the surcharge load on a
buried pipe with downward pressure acting on a plane at the
pipe crown. Once the surcharge load is determined, the total
load acting on the pipe is the sum of the earth load and the
surcharge load.
 Loads on Buried Pipe

• The load applied to a buried pipe consists of

dead load and surcharge load. The dead
load is the permanent load from the weight
of soil and pavement above the pipe.
Surcharge loads are loads applied at the
surface and may or may not be permanent.
Surcharge loads include the loads from
vehicles passing over the pipe. The adopted
basic theories are commonly those credited
to Marston and Spangler. Marston's
equation has been proven to be
conservative.
Live Loads
• Boussinesq's Point Load Formula for Wheel
Load:
cT I
WL  P
L
 Soil Load
• Marston's Formula for Soil Load:
Wd  c d γ B d2

Spangler's Formula for Circumferential Bending Stress:

6 K b We REt
Sb 
Et 3  24K z Pi R 3

Spangler's Formula for Pipe Deflections:

12 K z We R 3
Δ 3
Et  24 K z Pi R 3
 Figure 2 Values of Load Coefficient Cd (Trench Fill)
Coefficient Cd (graph on left)
1.0 1.5 2 3 4 5
40

30
25 B
A C
D
20 2
A

D
B

E
15 E 1.5

Ratio of H/B
Ratio of H/B

10 1.0
9 0.9
8 0.8
7 0.7

6 0.6
5 0.5

4 0.4

3 0.3

0.25

2 0.2

1.5 0.15

0.10 0.15 0.20 0.25 0.3 0.4 0.5 0.6 0.7 1.0 1.5 0.1

A = granular materials with out cohesion Coefficient Cd (graph on right)

B = sand and gravel D = Ordinary soil
C = saturated top soil E = Saturated Clay
Figure 3 Influence Value Is for Concentrated Surcharge or Uniform Surcharge of Limited Extent
Point Load P Uniformly distributed load U

H B
0 H
B
0 Stress at 0 = U x Is
Load on area B x L = P x Is

m = L/2H n = B/2H

m and n are interchangeable m=

0.24 m = 2.5
m = 1.8
m = 1.6
m = 1.4
0.22 m = 1.2
m = 1.0
0.20 m = 0.9
m = 0.8
0.18
m = 0.7
0.16 m = 0.6

0.14 m = 0.5
Values of Is

0.12
m = 0.4
0.10
m = 0.3
0.08

0.06 m = 0.2

0.04
m = 0.1
0.02

0.00
0.01 0.02 0.05 0.1 0.2 0.3 0.5 1.0 2.0 3.0 5.0 10.0
Values of m,n
Nomenclature
• Bd = width of trench at the top of a pipe (i.e. effective width, m)
• Cd = load coefficient for fill load
• CT = Influence coefficient for a single concentrated laod
• E = Young's Modulus, MPa
• Sb = circumferential bending stress in a thin steel pipe wall, MPa
• Sh = tensile hoop stress in a steel pipe, MPa
• I = impact factor for live loads
• Kb = bending parameter
• Kz = deflection parameter
• L = length of pipe (taken as 0.91 m), m
• P = wheel load, kg
• Pi = internal pressure in pipeline, MPa
• R = outside radius of pipe, mm
• t = pipe wall thickness, mm
• Wd = soil dead load on the pipe, N/m
• We = total external load on the pipe, N/m (includes the soil dead load and the vehicle
wheel live load)
• WL = wheel live load on the pipe, N/m
• D = maximum vertical deflection in pipe, mm
• g = unit weight of soil, kg/m3
Deflection of Flexible Pipe

Load
Deflection 
Pipe Stiffness  Soil Stiffness
Deflection of Flexible Pipe

Iowa Formula:
(Spangler)

K W r3
Dx  DL ( )
EI  0.061E ' r 3

Δx = horizontal deflection

Source: Moser, Buried Pipe Design, McGraw Hill

Deflection of Flexible Pipe

Modified Iowa Formula:

Dy DL K P D

EI eq
 0.061E '
3
with: r
Δy = vertical deflection of pipe, in.
DL = deflection lag factor
K = bedding constant
P = pressure on pipe due to soil load and live load, psi D,r = pipe
OD and radius, in.
EIeq =equivalent pipe wall stiffness per inch of pipe length, in./lb.
E’ = modulus of soil reaction, psi
Deflection of Flexible Pipe

Terms of modified Iowa formula

K = bedding constant; takes the support Bedding K

forces under the pipe into account and is a angle
function of bedding angle. In practice, K=0.1 0 0.110
is used. 30 0.108
45 0.105
DL = deflection lag factor; the backfill on the
sides of the pipe will consolidate and 60 0.102
get denser over time, thus transferring 90 0.096
load to the pipe and causing the pipe 120 0.09
to deflect more. For flexible pipe, the
180 0.083
load will not exceed the prism load,
hence DL=1.0 when using the prism
load. If the Marston load is used, DL will
be larger than 1.0. 47
Deflection of Flexible Pipe

Terms of modified
Iowa formula (cont.)
E’ = modulus of soil
reaction; this is a
hybrid modulus, not a
pure material property,
introduced to eliminate
the spring constant in
the original Iowa
formula. It is the
product of the modulus
of passive soil
resistance and the
radius of the pipe.
Source: AWWA, Manual M-11, 3rd edition 48
Through-Wall Bending Stress

Through-Wall Bending:

 Dy  t 
s bw  4E  
 D  D 
with:
Σbw = through-wall bending stress
Δy/D = pipe ovality
D = pipe OD
t = pipe wall thickness
E = elasticity modulus of pipe
Ring Buckling
Ring Buckling:
1 (EI ) eq
Pcrit  32RW B' E'
FS D3
with:
Pcrit = critical buckling pressure
FS = safety factor (2.5 for H/D ≥2; 3.0 for H/D<2) D
= pipe OD
RW = water buoyancy factor = 1-0.33(hw/H)
B’ = empirical coefficient of elastic support
(dimensionless) where:

1
B' (0.0065 H )
D
1 4e
 Transitions at Road and River Crossings
(a)
t2 t1
1.0 mm Max.
(b)

t2 tD
t1


30° Max.

(c)
tD t1
t2

30° Max.

(d)

t2 t1


Notes:
1. t1 & t2 = nominal wall thickness
tD = the design wall thickness, where tD is less than or equal to 1.5 t2
 - 30° max., 14° min.. No minimum where materials being joined have equal specified minimum yield strength.
2. Provided all the requirements of Figure 7-11 are met, and wall thicknesses t2 and tD meet the pressure design
requirements of CSA Standard Z662 - 96 the parts to be joined need not have the same SMYS when using the
above backbevelled details.
 Transitions at Road and River Crossings

60o to 75o

+1.0
t2 t2 t1
-0.2
4
1
+3
1.6  0.8 L
-0

Gap 1.5 to 2.5

Backbevelled Joint Design

for Piping Designed to Operate at a Hoop Stress of 60% SMYS or Less
 Chart Summary of Procedure
for Selection of Transition
t2 t0
+1.0 3 t1
t2 1
t2 -0.2 t0
4 t1
1

+3
Backbevel
L a. Definition of Transition
-0
Counterbore and taper

b. Flow Chart Procedure

NO TRANSITION NOTES:
(t1-t2)<1mm Yes
REQUIRED

No 1. All machining + 1.0mm/-0.2mm

(need only apply over 95% of circumference)
2. Subscript ‘1’ refers to thicker pipe; ‘2’ refers to thinner pipe
3. Sh - Hoop stress, check for both 11 and t2
Sh<0.6 SMYS 4. For backbevel, SMYS1 may be less than SMYS2
Yes No 5. Fittings shall be joined to pipe using backbevel
6. Except for fittings, a counterbore and taper may replace a backbevel
7. D = Pipe outside diameter (mm)
tD>1.5t2 No BACKBEVEL No (t1-t2)>0.3t2 8. SMYS = specified minimum yield strength

COUNTERBORE AND TAPER

Yes (I) L(INSTALLED) = 0.85(Dt2), but not
Yes
less than 50 mm
(ii) SMYS1 > SMYS2
Mechanical Anchor

Attachment
Bracket

Anchor Rod Extension

Swamp Weight

A
Re-Bar Hook

. .. . . . . .
. . . .
. . . . . .
. .
.
. . . .
. . . .
.
. . . .
. .
. . .
. . .
. . .
. .
. .
. .
. .
.

Wire Mesh
Lining

A Section A-A
Elevation
Sack Weights
River Weight
Hook
A (See Detail) Wire Mesh

Bolt c/w Nuts

. . . . . . and Washers
.. . . .
. .
. . . .
. . .
. .
.. . .
.
. .
. . . . .
. . . .
. . .
. .
. . . . . Lifting Hook
. . . .
. . . Detail
. .
Steel Tube Insert
To Accommodate Bolt
Lining

A Elevation Section A-A

Continuous Concrete Coated Pipe
Minimum Concrete
Thickness, T

Flexibility Gap Wire Mesh Compressible Material for Fusion Bonded

or Compressible Material Reinforcing Coating Protection (Where Epoxy Corrosion
(Where Required) Required) Coating or Equivalent

Notes:
Compressible may be required for protection of pipe coating depending on how concrete is applied.
Flexibility gap required for severe bends (e.g. offshore pipeline).
This design to be used mainly at river crossings.

Minimum Concrete
Thickness, T

Flexibility Gap Wire Mesh Compressible Material for Fusion Bonded

or Compressible Material Reinforcing Coating Protection (Where Epoxy Corrosion
(Where Required) Required) Coating or Equivalent

Notes:
Compressible may be required for protection of pipe coating depending on how concrete is applied.
Flexibility gap required for severe bends (e.g. offshore pipeline).
This design to be used mainly at river crossings.
 Fluid Surface

B c= Buoyant Force (Due to Fluid

Displaced By Concrete Weight)
Swamp Weight
Bp = Buoyant Force (Due to Fluid
Displaced By Pipe)

Dp = Outside
Diameter of
Pipe

Corrosion Coating
M p (g) = Force Due to Mass of Pipe
Steel Pipe
M c(g) = Force Due to Mass of Concrete Weight

Free Body Diagram Forces Action on

Submerged Pipeline
Swamp Weight Configuration
Variables in River Weight Spacing Formula

where L = center-to-center weight spacing, m

Mc = mass of concrete weight, kg
Bc = buoyant force due to fluid displaced by weight, N
Bp = buoyant force per unit length of pipe due to fluid displaced by pipe,
N/m
Mp = mass per unit length of steel pipe, kg/m
Mw = mass per unit length of wood lagging, kg/m
Bw = buoyant force per unit length of wood lagging due to fluid displaced by
wood lagging, N/m
g = acceleration due to gravity, m/s2
R = concrete weight length, m.
Wp = weight per unit length of pipe
Ww = weight per unit length of wood
Wc = weight of concrete.
 Variables in Swamp Weight Spacing Formula

Bp

Wp

Bc

Wc

The maximum allowable swamp weight spacing is calculated from the following
design formula:
 Fluid Surface

B c = Buoyant Force (Due to Fluid

Displaced by Concrete Weight)
River Weight
B p = Buoyant Force (Due to Fluid
Displaced by Pipe)

Dp = Outside
Diameter
of Pipe

Corrosion Coating
M p (g) = Force Due to Mass of Pipe
Steel Pipe

M c (g) = Force Due to Mass of Concrete Weight

Free Body Diagram Forces Action on Submerged Pipeline

River Weight Configuration
 Variables in River Weight Spacing Formula

Wp

Ww

Bc

Wc

Bw

Bp

Assuming the use of wood lagging, the maximum allowable river weight spacing is
calculated from the following design formula:
Variables in River Weight Spacing Formula

where L = center-to-center weight spacing, m

Mc = mass of concrete weight, kg
Bc = buoyant force due to fluid displaced by weight, N
Bp = buoyant force per unit length of pipe due to fluid displaced by pipe,
N/m
Mp = mass per unit length of steel pipe, kg/m
Mw = mass per unit length of wood lagging, kg/m
Bw = buoyant force per unit length of wood lagging due to fluid displaced by
wood lagging, N/m
g = acceleration due to gravity, m/s2
R = concrete weight length, m.
Wp = weight per unit length of pipe
Ww = weight per unit length of wood
Wc = weight of concrete.
To derive the formula for the spacing of river
weight consider the system equilibrium as
shown Wp

Ww

Bc

Wc

Bw

Bp

L
 Fluid Surface
Bc =Buoyant Force (Due to Fluid
Displacement by Concrete
Coating)
Continuous
Concrete Bp = Buoyant Force (Due to Fluid
Coating Displacement by Pipe)

Tc = Thickness of
Concrete Coating

Dp= Outside
Diameter of
Pipe

Corrosion
Coating Steel Mp(g) = Force Due to Mass Pipe
Pipe
Mc (g) = Force Due to Mass of Concrete Coating

Free Body Diagram Forces Action on Submerged

Pipeline Continuous Concrete Coating
 Pipe

TC

D Coating
 
1/2

 2  
1  D2 γ c - [D2 - (D - 2t ) ]γ s 
Tc = - D
2   N   
 γ c γ F
- 1 +   
  100   

Variables in
Continuous Concrete Coating Thickness Formula
Negative Buoyancy Concrete Coating
Example
• Determine the continuous concrete coating thickness
required for a pipe with the following parameters. Pipe
diameter 914mm, wall thickness 28.57mm. The density of
concrete gc is 3044kg/m3 , the density of steel gs
7800kg/m3 and the fluid density gf is 1025kg/m3. The
required amount of negative buoyancy is 35 % and the effect
on the external pipe diameter of the epoxy coating present
may be neglected.
• Using the following equation :
 1 /2
 
 2  
1  D2 γ c - [ D2 - (D - 2t ) ] γ 
Tc = s
- D
2   N   
 γ c - γ F 1 +   

  100   

Negative Buoyancy Concrete Coating
Example
• Substituting
 
1/2

 2  
1  0.9142 .3044 - [0.9142 - (0.914 - 2.28.57 ) ] 7800  - 0.914
Tc =
2   35   
 3044 1025 
- 1 +   
  100   

1  2542.95 - 787.8 
1/2

Tc =   - 0.914
2  1660.25  

1

Tc = 1.057 1/2 - 0.914
2

Tc = 57mm.
Stresses in Buried Pipe with
Bends
Sidebends

Sidebends are lateral

changes of direction in the
pipeline. As the pipe is no
longer fully restrained in the
sidebend area, it will push
into the bend, causing lateral
soil reaction forces which in
turn give rise to bending
stress in the pipe. For the
stress calculations, the
pipeline needs to be modeled
with stress analysis software
capable of modeling pipe-soil
interaction.
Source: ALA, Guidelines for the Design of Buried Steel Pipe
Sidebends
In Practice:
•For bend angles below 15 deg , bending stresses are typically not an issue
•For bend angles up to 30 deg AND bend radius large
enough so that SIF=1, bending stresses are not an issue
•For cold bends with bend radius above 37 x pipe OD, bending stresses are not an issue
•Increase the bend radius if the stresses are excessive. This reduces the SIF.
•Above points do not apply to bends in Muskeg
•Bending of pipe and the resulting bending stress is not a big concern at sag bends due to
the fact that the soil at the bottom of the ditch is typically undisturbed and very stiff. Where
the sag bend requires an induction bend, which implies a relatively large bend angle, the
bearing capacity of the soil under the bend should be checked.

BR
Overbends
The principal concern at overbends is the buckling of the pipeline that may
occur due to relatively weak restraining force provided by the overburden.
The pipeline might buckle as a column under the right combination of
eccentricity, axial force, and weak soil restraint against vertical uplift.
If the post-buckling displacement is large enough, local wrinkling may occur
Bending in Two Planes
“Roping” is the term pipeliner’s use to describe the practice of
laying pipe in a ditch that is not perfectly straight in either the
vertical or the horizontal planes. The pipe will elastically
deform as a beam under its own weight when picked up by
the side booms or when placed into the not perfectly straight
ditch. The resulting bending stresses are typically ignored in
the design because good construction specifications limit the
amount of curvature of a ditch bottom.

The allowable curvature may be calculated by using an

appropriate stress limit and the engineer’s theory of simple
beam bending (moment-curvature equation)
 Bending in Two Planes
As noted previously the stress limits, set by Z662, Cl 4.6.2.2, limits the
combination of all principal stresses to be less than or equal to SMYS
modified as necessary by the temperature derating factor:

Sh – SL + SB ≤ SMYS x T
Assuming a straight section of pipeline and noting that the stress limit for
combined hoop and longitudinal stress is 90% SMYS, this leaves a margin
of 10% of SMYS for bending stress:

SB ≤ 10% SMYS (when T = 1)

The stress due to constant curvature of a beam is: SB = E x OD / (2 x

Bending Radius) Hence the minimum allowable roping radius becomes:
BR roping, min = E x OD / 0.2 SMYS
80
 Line Lowering
•Pipelines may have to be relocated for various
reasons. New roads, railroads, utilities, ditches
may require that an existing pipeline be lowered
or moved out of the way. It is desirable that such
a relocation be undertaken while the pipeline
remains in service to avoid loss of revenue and
inconvenience for the owner and users of the
pipeline.
•A procedure to calculate the stresses induced
during a pipeline movement is described in API
RP 1117. Of course, finite element analysis can
also be used as a calculation tool.
Stress Analysis of HDD
HDD
The horizontal directional drilling process is a significant improvement
over cut and cover methods for installing pipelines beneath obstructions,
such as rivers, shorelines, major roads and environmentally sensitive
areas. Tools and techniques used are very similar to those in the oil well
drilling industry. Drill pipe, downhole tools, and drilling fluids are used for
HDD which is why the process is generally referred to as drilling as
opposed to boring.

The process involves three stages:

1.Drilling the pilot hole

2.Reaming

3.Pulling in the carrier pipe

HDD

Source: AGA-PRC, Installation of Pipelines by HDD, 1995

HDD

Source: AGA-PRC, Installation of Pipelines by HDD, 1995 124

HDD

Z662, Clause 4.21 Requirements for pipelines installed by horizontal directional drilling
Pipelines may be installed by directional drilling provided that
(a) A feasibility assessment is made to assess the suitability of subsurface conditions.
(b) The drill path is designed with due consideration given to the location and type of all subsurface
features influencing installation operations.
(c) An assessment is made to determine the risk of accidental release of drilling fluids from the
drilling annulus and an appropriate mitigation plan is prepared.
(d) For steel pipe, longitudinal stresses during installation do not exceed the specified minimum
yield strength of pipe.
Note: For guidance, refer to PCI PR-227-9424 and CAPP Publication 2004- 0022.

125
Source: AGA-PRC, Installation of Pipelines by HDD, 1995 126
HDD
Installation Loads and Stresses

During installation the pipeline is subjected to:

Tension from pulling the pipe into the reamed pilot hole and around the curved
sections of the hole, made up of:
 Frictional drag due to wetted friction between pipe and wall of hole
 Fluidic drag of pipe pulled through the viscous drilling mud trapped in the
hole annulus.
 Unbalanced gravity (weight) effects of pulling the pipe into and out of a hole
at different elevations,
Bending, as the pipe is forced through the curves in the hole
Compressive hoop stress from the external pressure exerted by the hydraulic head
of the drilling mud in the annulus around the pipe

127
HDD
Installation Loads
Key assumptions:
diameter of reamed hole is about 300mm larger than pipe diameter
annulus between pipe and hole is filled with drilling mud of a known density
formation pressure is not considered

The drill path is broken down into straight and curved sections. The curved
sections are assumed to be of constant radius.

Source: AGA-PRC, Installation of Pipelines by HDD, 1995

HDD
Installation Loads - Straight Sections

A straight section is modeled with the variables as shown in the figure below

Source: AGA-PRC, Installation of Pipelines by HDD, 1995

HDD
Installation Loads – Straight Section

T2 = T1 + |frict| + DRAG +/- Ws L sinθ

(-) if T2 tends downhole if T2 tends

(+) uphole
frict = friction between pipe and soil (lbs)
= Ws L cosθ μsoil (μsoil = 0.2 to 0.3 for uncoated steel)
DRAG = fluidic drag between pipe and viscous fluid (lbs)
= 12 p D L μmud (μmud = 0.05 for steel pulled through bentonite mud)
Ws = effective (submerged) weight of pipeline plus internal contents (lbs/ft)
L = length of section

130
HDD
Installation Loads - Curved Section
A curved section is modeled with the variables as shown in the figure below

Source: AGA-PRC, Installation of Pipelines by HDD, 1995

HDD
Installation Loads – Curved Section

R = constant radius of curvature (ft)

α = included angle of the curved section in degrees
θ = (θ1+ θ2) / 2
N1, N, N2 = normal contact forces at right, centre, and left points
frict1, frict, frict2 = friction forces associated with normal forces

Larc = length of curved section; Larc = R θ (p/180), (θ in degrees)

= R (1-cos(α/2))
h
= (12 Tave h – WS/12 cosθ Y) / X
N = 3 Larc –j/2 tanh (U/2)
XY = 18 Larc2 – j2 (1 – 1/cosh(U/2))
jU = sqrt (EI/T)
T = 12 Larc / j (“12” converts Larc from feet to inches)

= average of T1 and T2
HDD

Installation Loads – Curved Section

T2 = T1 + 2|frict| + DRAG +/- Ws Larc sinθ (-) if

T2 tends downhole
(+) if T2 tends uphole

Installation Loads – Total Pulling Load

Ttot = Σi (T2 – T1), for i sections

HDD
Example Pulling Load Calculation
pipe = NPS 12 (OD=12.75 in.), API 5L Grade B, SMYS = 35 ksi;
Emod = 2.9 x107 psi
t = 0.25 in.
WS = - 46.21 lb/ft
μsoil = 0.3
μmud = 0.05 psi
Mud weight = 89.76 lb/ft3
Pipe is installed empty; no pullback as pipe enters the hole

MAOP = 720 psi

Construction temperature = 60°F
Operating temperature = 80°F
Depth of horizontal section = 100 ft below entry/exit datum
Depth of groundwater table = 10 ft
HDD
Example Pulling Load Calculation
HDD
Example Pulling Load Calculation

Section Type Angle Length T2

A to B Straight θ = 20° L = 116.1 ft Down

B to C Curved, R = 1000 ft θ = 20°; α = 20° Larc = 349.1 ft Down

C to D Straight θ = 0° L = 500.3 ft Horiz

D to E Curved, R = 1200 ft θ = 7°; α = 14° Larc = 293.2 ft Up

E to F Straight θ = 14° L = 266.2 ft Up

HDD
Straight Section at Point B:

ΔTBA = TB – TA
= |frict| + DRAG – WS L sinθ
|frict| = WS L cosθ μSOIL
= (- 46.21 lb/ft) (116.1 ft) cos 20° (0.3)
= 1,512 lb
DRAG = 12 p D L μMUD
= 12 p (12.75 in) ( 116.1 ft) ( 0.05 lb/in2)
= 2,790 lb
WS L sinθ = (- 46.21 lb/ft) (116.1 ft) sin 20°
= -1,835 lb
ΔTBA = 1,512 + 2,790 – (-1,835) = 6,137 lb

TB = ΔTBA + TA = 6,137 lb <= pull load at point B

 HDD
Curved Section at Point C:

h = R (1-cos(α/2)) = (1000 ft) (1 – cos 10°) = 15.19 ft

I = p (D – t)3 t / 8 = p (12.75 in – 0.25 in)3 x 0.25in / 8 = 191.75 in4
Assume Tave for section = 10,000 lb to start iterative solution

jUX = (EI / Tave)1/2 = (2.9 E7 x 191.75 / 10,000)1/2 = 745.7 in

Y = 12 Larc / j = 12 x 349.1 ft / 745.7 in = 5.62
= 3 Larc –j/2 tanh (U/2) = 3 x 349.1 ft – 745.7 in/2 tanh(5.62/2) = 677.15 in
= 18 Larc2 – j2 (1 – 1/cosh(U/2))
= 18 (349.1 ft)2 – (745.7 in)2 (1-1/cosh(5.62/2)) = 1,704,320 in2
N = (12 Tave h – WS/12 cosθ Y) / X =
= 12 x 10,000 lb x 15.19 ft – (-46.21 lb/ft)/12 x cos10° x 1,704,320 in2 /677.15 in
12,237 lb
 HDD
(cont.) Curved Section at Point C:
ΔTCB = TC – TB
= 2 |frict| + DRAG – WS Larc sinθ
|frict| = N μSOIL
= 12,237 lb x 0.3 = 3,671 lb
DRAG = 12 p D Larc μMUD = 8,389 lb
= 12 p (12.75 in) ( 349.1 ft) ( 0.05 lb/in2)
WS Larc sinθ = (- 46.21 lb/ft) (349.1 ft) sin 10° = - 2,801 lb

ΔTCB = 2 (3,671 lb) + 8,389 lb – (-2,801) = 18,533 lb

TC = ΔTCB + TB = 24,670 lb <= pull load at point C before Tave assumption

check
Check Tave = (TC + TB)/2 = (24,670+6,137)/2 = 15,404 lb ≠ 10,000 lb

 Redo calculations with 15,404 lb as a starting point

 TC = 25,259 lb <= pull load at point C
 HDD
Straight Section at Point D:

ΔTDC = TD – TC
= |frict| + DRAG – WS L sinθ
|frict| = WS L cosθ μSOIL
= (- 46.21 lb/ft) (500.3 ft) cos 0° x 0.3
= 6,936 lb
DRAG = 12 p D L μMUD
= 12 p (12.75 in) ( 500.3 ft) (0.05 lb/in2)
= 12,024 lb
WS L sinθ = (- 46.21 lb/ft) (500.3 ft) sin 0°
= 0 lb
ΔTBA = 6,936 + 12,024 – 0 = 18,960 lb
= ΔTDC + TC = 44,219 lb <= pull load at point D
TD
HDD

Curved Section at Point E:

h = R (1-cos(α/2)) = (1200 ft) ( 1 – cos 7 deg) = 8.94 ft
I = 191.75 in4

Assume Tave for section = 51,545 lb to start iterative solution

j = (EI / Tave)1/2 = (2.9 E7 x 191.75 / 51,545)1/2 = 328.45 in

U = 12 Larc / j = 12 x 293.2 ft / 328.45 in = 10.71
X = 3 Larc –j/2 tanh (U/2) = 3 x 293.2 ft – 328.45 in/2 tanh(10.71/2) = 715.38 in
Y = 18 Larc2 – j2 (1 – 1/cosh(U/2))
= 18 (293.2 ft)2 – (328.45 in)2 (1-1/cosh(10.71/2)) = 1,440,532 in2
N = (12 Tave h – WS/12 cosθ Y) / X
= 12 x 51,545 lb x 8.94 ft – (-46.21 lb/ft)/12 x cos7° x 1,440,532 in2 /715.38 in
= 15,427 lb
 HDD

(cont.) Curved Section at Point E:

ΔTED = TE – TD
= 2 |frict| + DRAG – WS Larc sinθ
|frict| = N μSOIL
= 15,427 lb x 0.3 = 4,628 lb
DRAG = 12 p D Larc μMUD
= 12 p (12.75 in) ( 293.2 ft) ( 0.05 lb/in2) = 7,047 lb
WS Larc sinθ = (- 46.21 lb/ft) (293.2 ft) sin 7° = -1,651 lb

ΔTED = 2 (4,628 lb) + 7,047 lb + (-1,651) = 14,652 lb

TE = ΔTED + TD = 58,871 lb <= pull load at point E

HDD

Straight Section at Point F:

ΔTDC = TD – TC
= |frict| + DRAG – WS L sinθ
|frict| = WS L cosθ μSOIL
= (- 46.21 lb/ft) (266.2 ft) cos 14° (0.3)
DRAG = 3,581 lb
= 12 p D L μMUD
= 12 p (12.75 in) (266.2 ft) ( 0.05 lb/in2)
= 6,398 lb
WS L sinθ = (- 46.21 lb/ft) (266.2 ft) sin 14°
= -2,976 lb
ΔTFE = 3,581 + 6,398 + (–2,976) = 7,003 lb

TF = ΔTFE + TE = 65,874 lb <= pull load at point F

HDD
Installation Stresses
It is not obvious at which location the combination of tensile,
bending, and hoop stresses produces the maximum stress.
In general, highest stresses will occur at locations with tight
bending radius, high tension, and high hydrostatic head.

The stress analysis approach follows API RP 2A-WSD

“Planning, Designing and Constructing Fixed Offshore
Platforms – Working Stress Design” and consists of the
following steps:

calculate individual stresses – tensile, bending, and hoop

compare individual stresses against limits

calculate combined stresses in equations presented as

unity checks

These calculations are carried out at each point of interest.

HDD
Installation Stresses
Tensile Stress:
ft = T/A
with: T = tension at point of interest [lb]
A = cross-sectional area of pipe wall [in.]
Ft = 60% SMYS

Bending Stress:
fB = E D / 24R
= 75% SMYS for
with: FB
D/t ≤ 1,500,000 / SMYS
= (0.84 - 1.74 SMYS D/Et) x SMYS for
FB 1,500,000 / SMYS < D/t ≤ 3,000,000 / SMYS
= (0.72 – 0.58 SMYS D/Et) x SMYS for
FB 3,000,000 / SMYS < D/t ≤ 300
HDD
Hoop Buckling Stress:

fhc < Fhc / 1.5 allowable limit is critical hoop buckling stress /1.5
Fhe = 0.88 E (t/D)2 elastic hoop buckling stress for long cylinders
The critical hoop buckling stress, Fhc, depends on the value of the elastic hoop
buckling stress relative to the SMYS as follows:
If Fhe is less than 55% SMYS: Fhc = Fhe
For 55% SMYS < Fhe ≤ 160% SMYS: Fhc = 0.45 SMYS + 0.18 Fhe
For 160% SMYS < Fhe ≤ 620% SMYS: Fhc = 1.31 SMYS / (1.15+SMYS/ Fhe)
For Fhe larger than 620% SMYS: Fhc = SMYS
HDD
Combined Loads:

If checking the single load conditions returns satisfactory results, the stress location
must be checked for the combined loading. This is accomplished by conducting two
checks: tensile plus bending, and tensile plus bending plus hoop.
The unity check for tensile plus bending is: ft / 0.6 SMYS + fb/Fb ≤ 1.0

The unity check for combined tensile, bending and external hoop stress is:
A2 + B2 + 2ν |A| B ≤ 1.0
where

A = (ft + fb – 0.5 fh) x 1.25 / SMYS

B = 1.5 fh / Fhc
HDD Pull through Calculation
A review of the pilot hole plot shows that the most likely location of high stress due to combined loading
is at point E. Firstly, this point is closer to the rig side and will have high a pulling force resulting in high
tension. Secondly, there is a tight radius curve (R = 1200 ft) and thirdly, it is near the deepest point with
the highest hydrostatic mud pressure.
Individual Stresses at Point E:

Tensile Stress: ft = TE/A

= 58, 871 lb / [p/4 x (12.752 – 12.252]
= 5,997 psi
Bending Stress: fb = E D /24 R
= (2.9 E7 x 12.75) / 24 x 1200 ft
= 12,839 psi

External Hoop Stress: fh = Δp D / 2 t

= (40.15 psi x 12.75) / (2 x 0.25)
= 1,024 psi
with Δp = 12 ppg x 64.4 ft / 19.25 = 40.15 psi
HDD Pull through Calculation
Compare calculated stresses at point E to allowable stresses:

Allowable Tension: Ft = 0.6 SMYS

= 0.6 x 35,000 psi = 21,000 psi
=> ft < Ft => PASS

Allowable Bending: SMYS x D/t = 35,000 x 12.75/0.25 = 1,785,000

Fb = [0.84 – 1.74 SMYS D/(ET)] x SMYS
= [0.84 – 1.74 x 35,000 x 12.75/ (2.9E7 x 0.25)] x 35,000
= 25,652 psi
=> fb < Fb => PASS

Allowable Hoop Buckling:

Fhe = 0.88 E (t/D)2
= 0.88 x 2.9E7 x (0.25/12.75)2
= 9,812 psi ( = 28% SMYS; ≤ 55% SMYS)
=> Fhc = Fhe = 9,812 psi
=> Fhc/1.5 = 6,541 psi > fh => PASS 149
HDD Pull through Calculation

Unity checks at point E:

Tensile and Bending: ft / 0.6 SMYS + fb/Fb ≤ 1.0

= 5,997 / (0.6 x 35,000) + (12,839/25652)
= 0.79 ≤ 1.0 => PASS

Tensile, Bending, and External Hoop: A2 + B2 + 2ν |A| B ≤ 1.0

A = (ft + fb – 0.5 fh) x 1.25 / SMYS
= (5,997 + 12,839 – 1,024/2) x 1.25 / 35,000
= 0.654
B = 1.5 fh / Fhc
= 1.5 x 1,024 / 9,812
= 0.157

0.6542 + 0.1572 + 2 x 0.3 x 0.654 x 0.157 = 0.51

= 0.51 ≤ 1.0 => PASS
150
HDD
Operating Loads and Stresses
The operating stresses in the HDD installed pipeline are calculated just as those for
a pipeline installed in a ditch. The allowable stress limit is based on Z662, Clause
4.6.2.2:
SH + SL + SB ≤ SMYS x T
Example Operating Loads Stress Analysis

Bending Stress: fB = E D / 24R

= 2.9E7 x 12.75 / 24 x 1000ft
= +/- 15,406 psi

Hoop Stress: Δp = 720 psi – 90 ft x 0.4333 = 681 psi

fh = Δp D / 2t
= 681 x 12.75 / (2 x 0.25)
= 17,366 psi
Thermal Stress: ft = E α (T2 – T1)
= 2.9E7 x 6.5E-6 (60°F – 80°F)
= - 3,770 psi

Total combined stress:

= 17,366 x (1-0.3) +15,406 +3,770
= 31,322 psi
≤ 35,000 psi => PASS 151
Typical Branch Connections
Unreinforced
Fabricated Tee

Reinforced
Fabricated
Tee

Welding
Tee

Half-Coupling
Branch
Connection
 . Concept of Area Reinforcement
Tb
Reinforcement
Zone Excess Wall
tmhd1
Branch in Branch

A2 Excess Wall
L4 A3 in Header
d1
M
AR

tmh Tb
A1

d2

Run

A1 + A2 + A3  AR
d1 = inside diameter of branch
d2 = 1/2 width of reinforcing zone
L4 = height of reinforcement zone
= normally equals 2.5 Tb
AR = required area of reinforcement
= tmhd1
Welded Branch Connection
Reinforcement Area
Example Problem
Example Continued
831.42 Special requirements Branch
connections
Ratio of Branch to Run diameter
Ratio of hoop stress ¼ or less > ¼ but <1/2 >1/2
to SMYS in Run

20% or less Reinforcement Reinforcement not If reinforcement required

not mandatory mandatory it must be full
special cases special cases only encirclement if >half
only circumference

Greater than 20% Calculation not If reinforcement If reinforcement required

but 50% or less req”d for required it must it must be full
openings <2” be full encirclement with special
encirclement with welds
special welds

Greater than 50% Calculation not Welding tee Welding tee preferred
req”d for preferred otherwise full encirclement
openings <2” otherwise full
encirclement