Chapter 2

TRANSPORTATION PIPELINE AND

PIPING THEORY

2.0. General
In the modern society, products are seldom consumed in the same location where they are
produced. This is true for most items, from basic farm products to very sophisticated high-
technology industrial products. These products are often transported thousands of miles to reach
their consumers. The difference lies in terms of the most suitable means of transporting these
products into the market. For liquid and gaseous products, pipelines have proven to be the safest
and most economical means of transportation onshore, and one of the most reliable methods for
transporting these products across the water, exemplifies the needs of transportation pipelines for
crude oil and its refined products. The crude produced from the oil field has to be transported to
the user, either a refinery or a shipping terminal, located hundreds of miles away. Because an oil
refinery is generally located away from a populated area that will be consuming the refined
products, the refined products will then need to be transported from the oil refinery to the
consuming market. This is the situation with oil and natural gas. The main characteristic of the
transportation pipeline is that it involves a large quantity of pipe, which often requires multiple
suppliers for a single project. A transportation pipeline also generally runs many miles without
any attachment of special components, such as elbows, tees, and other stress risers [5].

Figure (2.1)

10
 Transporting Crude Oil and Refined Products

Pipelines have been used to transport many different things. The most common pipelines are
water distribution systems, which we are all familiar with. Coal, iron ore, and other solid goods,
in water slurry, have also been transported through pipelines occasionally. However, this chapter
deals only with gas and liquid petroleum pipelines. However, this work deals only with the
general overall behaviors of the pipeline system. The local pipe shell deformation due to soil
burden and road crossing is not discussed. The design of transportation pipelines involves close
cooperation among geotechnical, civil, and mechanical engineers. However, not all pipelines are
equal. For instance, the design of pipelines installed in open flat wastelands in warm climate
areas is essentially complete with just the calculation of the pipe wall thickness [5]. On the other
hand, the design of lines situated in frigid areas with sensitive environment and vast areas of
water and mountains to cross would require the most sophisticated engineering skills and
resources. This chapter covers only the basic fundamental mechanical analyses on the general
behaviors of long pipelines and their interfaces with local equipment.

2.0.1 Basic Stresses

Stresses in thin cylindrical shells (Pipes and pipelines)

When a thin cylinder is subjected to internal pressure, stresses are induced on the longitudinal
section XX, Fig. (2.2), due to the force tending to separate the top and bottom halves, and on the
circumferential section YY due to the force tending to separate the right- and left-hand ends of
the cylinder. The stress on the longitudinal section is termed the circumferential stress or most
frequently hoop stress and that on the circumferential section is termed the longitudinal stress;
the type of stress is determined by the direction of the arrows. In determining the stresses
induced, it is assumed that the thickness is small in comparison with the diameter so that the
stress on a cross-section may be taken as uniform and also that the ends give no support to the
sides, an assumption which would be appropriate to a long cylinder such as a pipe or pipeline
[9].

Let the internal diameter and length be d and l respectively, the thickness of metal be t and the
internal pressure be p.

Figure (2.2)

Circumferential stress (Hoop stress)

11
 The force tending to separate the top and bottom halves is the pressure multiplied by the
projected area in a direction perpendicular to the diametric plane of Fig. (2.2) [9].
i.e.
σc = pdl

This is resisted by the stress acting on the longitudinal section, XX, i.e.

pdl pd
σc = = (2.1)
2tl 2t

If the cylinder is made up from welded plates and the efficiency of the longitudinal joints is η
then the average stress in the joint is given by
pd
σc = 2t η (2.2)
Longitudinal Stress
The force tending to separate the right- and left hand halves is the pressure multiplied by the area
of one end, Fig. 2.3,

Figure (2.3)

𝜋
𝑃𝑙 = 𝑝 × 4 𝑑 2 (2.3a)

This is resisted by the stress acting on the circumferential section, YY,

𝜋
𝑝× 𝑑2 𝑝𝑑
4
𝜎𝑙 = = (2.3b)
𝜋𝑑𝑡 4𝑡

If the cylinder is made up from welded plates and the efficiency of the circumferential joints is η
then the average stress in the joint is given by
𝑝𝑑
𝜎𝑙 = (2.4)
4𝑡𝜂

2.0.2 Thermal Stress

12
A change in temperature can cause a body to change its dimensions. Generally, if the
temperature increases, the body will expand, whereas if the temperature decreases, it will
contract. Ordinarily this expansion or contraction is linearly related to the temperature increase
or decrease that occurs. If this is the case, and the material is homogeneous and isotropic, it has
been found from experiment that the displacement of a member having a length L can be
calculated using the formula[9].

𝛿𝑇 = 𝛼 ∆𝑇 𝐿 (2.5)

Where α is a property of the material, referred to as the linear coefficient of thermal expansion.
The units measure strain per degree of temperature. They are (Fahrenheit) in the FPS system, and
(Celsius) or (Kelvin) in the SI system. Typical values are given on the inside back cover the
algebraic change in temperature of the member the original length of the member the algebraic
change in the length of the member

ΔT = the algebraic change in temperature of the member

L= the original length of the member
δT= the algebraic change in the length of the member

The change in length of a statically determinate member can easily be calculated using Eq. 2.5,
since the member is free to expand or contract when it undergoes a temperature change.
However, in a statically indeterminate member, these thermal displacements will be constrained
by the supports, thereby producing thermal stresses that must be considered in design[9].

2.1 Governing codes and General Design Requirements

Because many pipelines run across public spaces, they are generally regulated by government
agencies. In the United States, Title 49 of the Code of Federal Regulations (49 CFR), Parts 192
[5] and 195[5], cover the minimum safety standards for pipelines. 49 CFR emphasizes more of
the administration aspects of pipelines rather than the technical matters. For technical matters,
the American Society of Mechanical Engineers (ASME) Code for Pressure Piping [5] provides
more detailed rules. The discussions that follow are based mainly on the ASME code, commonly
referred to as ASME B31 code since transportation pipelines are quite unique compared with
other plant piping systems, ASMEB31 code has three sections that deal specifically with them:
B31.4 for liquid petroleum pipeline, B31.8 for gas transmission pipeline, and B31.11 for slurry
pipeline [5]. As noted above, a pipeline generally runs for miles without any fittings attached.
Because of this simplicity, the stress in the majority portion of a pipeline is quite predictable.
Taking advantage of this basic characteristic, the code’s allowable stress for a pipeline is greatly
increased, as compared to that for plant piping, to reduce the tonnage of steel required. With the
large amount of piping required, the savings on steel cost translate to a high proportion of

13
savings in the whole project. This increase in allowable stress may appear to have decreased the
apparent safety factor. In reality, the geometrical and shape simplicity in a long pipeline
eliminates a large chunk of uncertainty and increases the predictability of stresses. This certainty
makes the pipeline’s real safety factor comparable to that of other sections of the ASME B31
code. Generally; ASME B31 code’s allowable stress for a pipe is based on overall considerations
of yield strength, ultimate strength, and creep strength of each specific pipe material. However,
the allowable stress of a transportation pipeline is based primarily on yield strength only. In other
words, the allowable stress of a pipeline is mainly to protect the pipe from gross deformation. As
long as the ductility of the material meets the pipe specifications, the margin of the ultimate
strength over the yield strength is considered sufficient to provide the required safety factor,
assuming the pipe is not grossly deformed. As the pipeline operates at a much lower temperature
than the creep temperature, no creep consideration is needed. Taking advantage of this unique
characteristic, higher and higher yield strength pipes are being manufactured for pipeline use.
The American Petroleum Institute (API) SPEC 5L [5] has specifications for pipes with yield
strengths as high as 120 ksi (827 MPa). However, the savings on steel tonnage from using the
ultra-high yield pipe is offset by the price premium of the high yield pipe, higher cost in special
welding and installation procedure, and so forth. The minimum thickness required for handling
and bending also prevents the use of very high yield steel. For these reasons, the use of very high
yield pipe is not always the best choice. Without a detailed cost analysis, most pipelines tend to
use moderately high yield pipes, such as grades X42 or X46. The number following the X is the
yield strength in kilo pounds per square inch (ksi) unit. High yield steel pipes have relatively low
ultimate strength to yield strength ratio. The grade X80 pipe, for instance, has a yield strength of
80 ksi (552 MPa) and an ultimate strength of 90 ksi (620MPa). The ultimate strength to yield
strength ratio is 1.125. This is quite small compared to the ratio for low yield strength steel pipes,
which usually has an ultimate strength to yield strength ratio of about 2.0. The low ultimate to
yield strength ratio also means a low ultimate strength to allowable stress ratio. This low ultimate
strength to allowable stress ratio requires the assurance that the local stress excursion be
controlled. This means that special consideration of local primary stress on certain components,
such as branch connections, may be required. The following discussions are limited to the
general stress aspects of the code. Engineers should consult the full text of the code for details.

2.1.1 B31.4 LIQUID PETROLEUM PIPELINE

ASME B31.4 [5] covers piping systems transporting liquids, such as crude oil, condensate,
natural gasoline, natural gas liquids, liquefied petroleum gas, liquid alcohol, liquid anhydrous
ammonia, and liquid petroleum products, between producers’ lease facilities, such as tank farms,
pump stations, natural gas processing plants, refineries, ammonia plants, terminals (marine, rail,
and truck), and other delivery and receiving points. The following sections provide the code
requirement from the stress analysis point of view. For more details about this subject the reader
can see references [3-5].

14
(a) B31.4 basic allowable stress. The basic allowable stress is used for basic pressure designs,
such as calculating the required wall thickness. In the B31.4 code, it is based on the specified
minimum yield strength (SMYS) and is defined as
S = 0.72 Ej (SMYS) ( 2.6)

where 0.72 is the design factor and Ej is the weld joint factor. In setting the above design factor,
the B31.4 code considered and included the allowance for under-thickness tolerance and
maximum allowable depth of imperfections provided for in the material specifications approved
by the code. In other words, this design factor is based on nominal wall thickness.
The code has a specific weld joint factor for each material. These official numbers have to
be used for all calculations. In general, seamless and electric resistance welded pipe has a weld
joint factor equal to 1.0. The electric fusion welded pipe generally has a weld joint factor of 0.8,
and the furnace butt-welded pipe has a weld joint factor of 0.6. For pipe that has been cold
worked in order to meet the SMYS and subsequently heated to 600°F (330°C) or higher (welding
excepted), the basic allowable stress value shall be 75% of the value determined by Eq. (2.6).
From the basic allowable stress, the required wall thickness is determined as [5]:
𝑃𝐷
tn = + 𝐴 (2.7)
2𝑆

where tn is the nominal wall thickness, P is the internal pressure, and D is the internal diameter.
This is a departure from the basic design of plant piping, which uses the minimum potential net
thickness after subtracting the manufacturing under tolerance. Thickness allowance, A, is the
sum of the allowances for threading, grooving, corrosion, and so forth. The wall thickness
allowance for corrosion is not required if pipe and components are protected against corrosion in
accordance with the requirements and procedures prescribed in Chapter VIII “Corrosion
Control” of the B31.4 code [3].
b) B31.4 allowable stresses for combined loading. Due to the very high allowable stress
allocated to the pressure loading, the allowable stresses for other loadings applied to the pipeline
are considerably less than that of other sections of the B31 code. Generally, the stresses due to
some common load types are calculated and evaluated as follows:

(2) Internal pressure hoop stress, SHP

PD
SHP =2(t ≤ 1.0S (2.8)
n − A)

where tn is the nominal thickness of the pipe and A is the allowance including corrosion
allowance and thread allowance corrosion allowance is not required for pipe and components
that are properly corrosion-controlled according to the code requirements.

15
(2) External pressure. The limitation of hoop stress due to external pressure is also governed by
Eq. (2.8). In addition to the hoop stress limitation, the pipe shall also be evaluated for buckling
stability protection [5].

(3) Thermal expansion stress range, SE. The thermal expansion stress range is an equivalent
stress calculated based on Maximum Shear Failure Theory. It is defined as twice the value of the
maximum shear stress.

For fully restrained lines:

SL = VSHP– E α (T2 – T1)
SE = greater of ( | SL| or | SL – SHP | )
SE ≤ 0.9(SMYS) (2.9)
For unrestrained lines:

SE ≤ 0.7(SMYS) (2.10)

The SE for unrestrained lines is calculated by the standard flexibility procedure applicable to all
sections of the B31code. For above-ground fully restrained lines, the weight bending stress has to
be added to SE in Eq. ( ).

(4) Additive longitudinal stress, SLL. The sum of the longitudinal stresses due to internal
pressure, weight, and other sustained loadings shall not exceed 75% of the allowable stress for
thermal expansion. That is, for unrestrained lines, we have

𝑆𝐻𝑃
SLL = + 𝑆𝐿𝑆 ≤0.525(SMYS) (2.11)
2

where SLS is the longitudinal stress due to weight and other sustained loadings.
(5) Additive circumferential stress, SH. The sum of the circumferential stresses due to internal
pressure and external load in the pipe, installed under railroad or highways without use of casing,
shall not exceed the basic allowable stress, S. That is

SH = SHP + SHC≤ 0.1𝑆 (2.12)

where SHC is the circumferential stress due to external loads on railroad or highway crossings
This stress is not automatically calculated by most pipe stress computer software packages.
However Caesar II take it into consideration. It needs to be calculated separately[5] or others. By
comparing Eqs. (2.11) and (2.12), it is clear that the thickness of the pipe at road crossings needs
to be increased over the thickness of the main pipeline.

16
(6) Occasional stress, Socc. The sum of the longitudinal stresses produced by pressure, weight
load and other sustained loads, hydraulic transient load, and wind or earthquake shall not exceed
80% of the SMYS. That is,
𝑆𝐻𝑃
SOCC = + SLS + SLT + greater of ( SLW or SLQ ) ≤ 0.8(SMYS) ( 2.13)
2

where SLT, SLW, and SLQ are the longitudinal stresses due to hydraulic transient, wind, and
earthquake, respectively. Wind and earthquake are not considered to occur simultaneously.
Moreover, some specifications do not require the combination of wind or earthquake with the
hydraulic transient load. The above stress limitations are given in subparagraphs 402.3.2 and
402.3.3 of the B31.4 code. Items (1), (3), (4), and (6) are generally evaluated automatically by
the pipe stress computer program. Items (2) and (5) are generally evaluated separately either by
manual calculations or with other specialized computer programs.

2.2 Behavior of Long Pipeline

The stress analysis of a pipeline is quite different from that of plant piping [5]. The most
fundamental difference between pipeline and plant piping is the very long length of the pipeline.
A pipeline with miles in length has the potential of producing a very large amount of expansion.
A reasonable estimate of the movement and its interaction with the end resistance force afforded
by connecting piping and equipment are very important aspects in designing a pipeline.

2.2.1 PRESSURE ELONGATION

When estimating the potential movement of the pipeline, it is important to include pressure
elongation in addition to thermal expansion. The pressure elongation per unit length of a pipe is
determined as

𝑆𝐻𝑃
𝑒𝐿 = ( 0.5 – v ) ( 2.14 )
𝐸

Because the pressure hoop stress is generally very high in a pipeline, pressure elongation is very
significant compared to thermal expansion. As shown in Fig. (2.4) a hoop stress of 30 ksi (207
MPa) produces a pressure elongation equivalent to a 35°F (20°C) temperature difference in
thermal expansion. This is very significant in a pipeline that is generally operating at about the
same amount of temperature difference. After including this pressure elongation, the total
expansion per unit length of the pipe is

𝑆𝐻𝑃
e = eT + eL = 𝛼 (T2 – T1) +(0.5 – v) ( 2.15 )
𝐸

2.2.2 ANCHOR FORCE

17
The first step in finding the potential movement of the pipeline is to determine the force required
to stop the movement. This anchor force is also needed for designing the anchors placed at
strategic locations to prevent the pipeline movement from causing damage to the connecting
piping and equipment. From the total expansion given in Eq. (2.15), we can convert it to stress
and force as
F = A×S = A×E×e = A[ Eα(T2 – T1) + (0.5 – v)SHP] ( 2.16 )

where A is the cross-sectional area of the pipe material. The anchor force, F, is also the total
axial driving force that generates the pipeline movement. It is also called the potential pipeline
expansion force.

2.2.3 POTENTIAL MOVEMENT OF FREE ENDS

With the potential expansion force given, the potential free end movement of the long pipeline is
estimated as shown in Fig.(2.4). The expansion will be completely suppressed by an axial force
or resistance force equal to the potential expansion force, which is the anchor force given in Eq.
(2.16). The resistance force of the free end comes mainly from the longitudinal friction force
created by the soil for a buried pipeline. Assuming that the longitudinal resistance force is f, per
unit length of the pipe, the length of pipe needed to generate enough friction force to reach the
anchor force is
𝐹
L=𝑓 ( 2.17 )

L is called the virtual anchor length. It is also called the active length. The point where the
expansion is completely suppressed is called the virtual anchor point. The net expansion rate, in
which the potential expansion rate subtracts the squeezing by the axial force, increases linearly
from nothing at the virtual anchor point to reach the maximum full expansion rate at the end. The
maximum end movement is therefore equal to the average net expansion rate multiplied by the
active length. That is,
e 1 F F F2
y0 = 2 L = (2 AE) f = 2AEf ( 2.18 )

18
 Fig.(2.4)

Because the potential force, F, contains the e and AE terms, y0 is directly proportional to the
square of the potential expansion rate and also to AE. It is inversely proportional to the friction
resistance.

2.2.4 MOVEMENT OF RESTRAINED ENDS

The pipeline end movement will be reduced by the end resistance of the connecting piping and
restraints. Assuming the end resistance is Q as shown in Fig. 2.5, the potential force for
expanding the pipeline will be reduced to (F - Q). The exact value of Q is determined from the
balance of the end movement and system resistance. As the potential expanding force is
reduced, its maximum expansion rate and end displacement are reduced to as follows.

𝐹−𝑄 𝐹−𝑄
e= , L=
𝐴𝐸 𝑓

𝐿 ( 𝐹−𝑄 )2
y = 2𝑒 = (2.19)
2𝐴𝐸𝑓

The end resistance reduces the movement more than just linearly. Equation (2.18) is represented
by the interactive curve. This interactive curve is a very important tool in the
analysis of piping and equipment connected to a long pipeline. Because it is non-linear, the
interaction is included in the piping analysis by using an iterative process. The simulation is
available in some advanced computer software packages [5].

19
2.2.5 STRESSES AT FULLY RESTRAINED SECTION
The major portion of the pipeline is fully restrained either by anchor blocks or virtual anchors.
For the line fixed by virtual anchors, the longitudinal pipe stress distribution is as shown in Fig.
(2.5). At the free end, the longitudinal stress is mainly due to the pressure effect and is equal to
one-half of the pressure hoop stress and is in tension for internal pressure.

Figure (2.5)

The pressure hoop stress is essentially constant throughout the system, but the longitudinal pipe
stress is gradually reduced by the friction force as the pipeline moves. The sign of the stress
eventually reverses to compression when enough movement and friction is produced. The
longitudinal stress maintains a constant value at the fully restrained portion after reaching the
virtual anchor point. Figure 2.4 shows the stress condition at the fully restrained portion.

For a buried pipeline, the weight is continuously supported; thus, no bending stress due to
weight is produced. Stresses in this situation are mainly due to pressure and suppression of
thermal expansion. As the longitudinal pressure stress is cut off by the anchor, the pressure in the
fully restrained section does not create longitudinal elongation. Instead, it produces Poisson
contraction in the longitudinal direction due to stretching in the diametrical direction. Therefore,
at a fully restrained section, the longitudinal strain being suppressed is

SHP
e = α ( T2 – T1 ) - v for fully restrained ( 2.20 )
E

The longitudinal stress required to suppress the above strain is

20
 SL = -Ee = v𝑆𝐻𝑃 – Eα( T2 – T1 ) for fully restrained ( 2.21 )

In addition to longitudinal stress, the pipe always has pressure hoop stress and pressure radial
tress. The pressure radial stress is generally ignored, leaving only the hoop and longitudinal
stresses. Based on the maximum shear failure theory, the combined equivalent stress is the
absolute sum of the two stresses if they have opposite signs; otherwise, it is equal to the greater
of the two. That is,
SE = greater of ( |SL| , |SHP| or |SHP – SL| ) ( 2.22 )
For an internal pressure, the hoop stress is always in tension. However, the sign of the
longitudinal stress depends on the magnitude of the temperature difference. As the operating
temperature rises, the longitudinal stress changes gradually from tension to compression. As the
longitudinal stress becomes compression, it is additive to the hoop stress to become the
combined stress. Because the allowable stress for the combined equivalent stress is 0.9(SMYS)
as given by Eq. (2.3), the longitudinal stress may dictate the wall thickness in installations with
higher operating temperatures. For fully restrained above-ground pipelines, the bending stress
due to weight and other mechanical loads have to be added absolutely to Eq. (2.21) before
comparing with the 0.9(SMYS) allowable. Bending moment has both tensile and compressive
stresses.
Because longitudinal stress is a function of operating and construction temperatures, an increase
in construction temperature will reduce the longitudinal stress. This may call for a special
construction procedure, if practical, to backfill the soil at the time when the ambient temperature
is at its highest.

Figure (2.6)

21
2.3 CAESAR II Overview

CAESAR II is the Pipe Stress Analysis standard against which all others are measured and
compared. The CAESAR II spreadsheet input technique revolutionized the way piping models
are built, modified, and verified [8].

CAESAR II was the first pipe stress program specifically designed for the PC environment. The
interactive capabilities permit rapid evaluation of both input and output, thereby melding
seamlessly into the "design - analyze" iteration cycle.

CAESAR II incorporates a wide range of capabilities, from numerous piping codes, to expansion
joint, valve and flange and structural data bases, to structural and buried pipe modeling, to
equipment and vessel nozzle evaluation, to spectrum and time history analysis. Most of the
features of CAESAR II are available at a keystroke, but at the same time are not imposed upon
the analyst[8].

A menu driven scroll and select interface provides logical options when and where expected.
Context sensitive help provides instant technical assistance for each data field, with expected
units. Data values presented in the help screens are automatically presented in the current set of
units to aid input [8].

The customization options of CAESAR II have been driven by user requests, code changes, and
the need to benchmark older, existing systems and their initial design. Many of these
customization options enable newer analysis techniques appearing in current literature.

As with all ICAS products, CAESAR II is continuously maintained and improved by the
engineering staff, who have worked in industry for engineering and consulting firms. This
experience not only guides program development, but provides users with knowledgeable
support[8]. This allows CAESAR II to work the way a typical engineer thinks and solves a
problem.

Whether you are a one-man consulting company or a large corporation, if you are looking for a
piping analysis program, CAESAR II is the best solution available - period[8].

2.3.1 Data Input

 CAESAR II makes it easy to input and display all the data needed to accurately define a
piping system analysis model. Input can be accessed or modified on an element-by-
element basis, or datasets can be selected to make global changes.

22
2.3.2 Cutting-edge Graphics

 The CAESAR II input graphics module makes quick work of developing analysis models
while clearly indicating areas of concern and providing an excellent idea of the piping
system's flexibility. Color-coded stress models and animated displacements for any stress
load case are available [8].

2.3.3 Design Tools and Wizards

 Tools and wizards for tasks such as creating expansion loops or viewing plant models in
the analysis space help bridge the gap between knowledge and experience. Such tools
take the guess work out of producing accurate analysis and recommending practical
design changes.

2.3.4 Analysis Options

 Besides the evaluation of a piping system's response to thermal, deadweight and pressure
loads, CAESAR II analyzes the effects of wind, support settlement, seismic loads and
wave loads.
 Nonlinear effects such as support lift off, gap closure and friction are also included.
CAESAR II also selects the proper springs for supporting systems with large vertical
deflections. Dynamic analysis capabilities include modal, harmonic, response spectrum
and time history analysis.

2.3.5 Error Checking and Reports

 The CAESAR II program includes an integrated error checker. This error checker
analyzes the user input and checks for consistency from both a "finite element" and
"piping" point of view. Reports are clear, accurate concise and fully user definable.

2.3.6 Material and Assemblies Databases

 CAESAR II incorporates table look-ups for piping materials and components plus
expansion joints, structural steel sections, spring hangers and material properties
including allowable stress. This ensures correct datasets are used for each analysis.
CAESAR II comes complete with major international piping codes.

2.3.7 Bi-directional Interface with Design

23
  CAESAR II incorporates the industry's first and only seamless, bi-directional link
between CAD plant design and engineering analysis. This link to Intergraph CADWorx®
Plant allows the passing of design and analysis data between these packages without data
loss.

2.4 Types of Loads:

 Thermal Loads at Various Operating Conditions

 Sustained Loads
 Seismic Loads
 Wind Loads
 PSV Reaction Force
 Slug Force
 Forces Induced by Surge Pressure

2.4.1 Thermal Loads at Various Operating Conditions

As temperature increases, so does the length of molecular bonds and solids typically expand on
heating and contract on cooling. The dimensional response to temperature change is expressed
by its coefficient in thermal expansion (CTE). Thus, to avoid “fatigue failure” resulting from
cyclic thermal loads, the piping system should be made flexible (and not stiff) [8].

2.4.2 Analysis Temperature

For thermal analysis, the design temperatures (minimum/maximum values as per line list) are
used. If lines have design temperature higher than 80 deg. And the difference between operating
and design temperature is more than 25 deg C then the maximum/minimum operating
temperatures, as derived from Process Group, can be used in stress analysis instead of design
temperature[8].

Additionally, Piping must also be analyzed for the normal operating temperature. Line Design
Pressure defined in the Line List shall be considered as the analysis pressure[8].

2.4.3 Sustained Loads

Sustained loads exist throughout the plant’s operation. These mainly consist of internal pressure
and dead-weight. Dead-weight is generally from weight of pipes, fittings, components such as
valves, operating fluid, test fluid, insulation, cladding, lining etc. These are to be considered as
follows:

 Design pressure

24
  Weight of Pipe and associated components such as Flanges, Valves, Strainer, Sight glass
etc., mounted on the Piping System.
 Weight of Fluid/contents in the piping
 Insulation and cladding weight
 Hydro test loads, if applicable
 Snow load, if applicable

Different international piping codes have set the limits that are also called “allowable stresses for
sustained loads”, on these axial stresses caused by dead-weight and pressure in order to avoid
“failure by collapse”. To achieve the allowable stresses for sustained loads, it is often required to
support the piping system vertically[8].

2.4.4 Seismic Loads

Projects that require Seismic design to be implemented, there needs to be an equivalent static
seismic analysis in the software to evaluate seismic stresses, displacements and loads on the
piping systems[8].

In case the Seismic coefficient has not been provided by the client, the calculation of static
seismic co-efficient in terms of ‘g’ – as per project specific information must be calculated using
relevant codes (UBC/ASCE/SNIP) and others. Here, the Seismic factor calculation document has
to be attached to the project stress analysis document. Normal operating temperatures are used
when considering occasional loads. Seismic loads are considered to be acting along the
horizontal axes – along North, South, East, and West directions but not acting at the same
time[8].

2.4.5 Occasional loads

These type of loads are imposed on piping by occasional events like wind, earthquake etc. Wind
loads are considered for lines with external diameter 14” NB (including insulation) or above and
at elevation 10 meters or higher from the ground level. Wind normally blows in the horizontal
plane and to protect piping from wind, it is standard practice to attach lateral supports. In case of
an earthquake, the earth seems to move vertically and to protect the piping against both
horizontal/vertical movement, some resting supports might be constructed as integral two way
lateral and vertical restraints. Normal operating temperature is used when analyzing occasional
loads[8].

2.4.6 PSV Reaction Force

If the piping system in question includes PSV (pressure safety valve), then the reaction force due
to PSV operation is considered as applicable. While analyzing PSV connected stress systems, the

25
reaction force needs to be calculated. A dynamic load factor equal to 2.0 must be applied on the
valve reaction force value. But in case of high reaction force values, the DLF value calculated as
per appendix-II of ASME B31.1 should be used. This needs to consider valve data provided by
vendor[8].

2.4.7 Slug Force

Slug force has to be considered in stress analysis for lines that have slug flow regime.

2.4.8 Forces Induced by Surge Pressure

The effect of surge on any line is also considered in stress analysis when identified by the
Process group or a specialist agency. However it should be noted that two concurrent occasional
loading shouldn’t be considered in stress analysis[8].

Piping Stress Analysis is a complex discipline – highly interwoven with piping layout and
support design. Sufficient flexibility and proper routing are important for simplicity and
economy of design. In spite of optimal design tools and sophisticated software, the best solutions
are the ones that use clever engineering judgment, experience and foresight[8].

2.5 Finite element basic principle

The finite element (FE) method was developed more by engineers using physical insight than
by mathematicians using abstract methods. It was first applied to problems of stress analysis
and has since been applied to other problems of continua. In all applications the analyst seeks to
calculate a field quantity: in stress analysis it is the displacement field or the stress field; in
thermal analysis it is the temperature field or the heat flux; in fluid flow it is the stream function
or the velocity potential function; and so on. Results of greatest interest are usually peak values
of either the field quantity or its gradients. The (FE) method is a way of netting a numerical
solution to a specific problem. A (FE) analysis does not produce a formula as a solution, nor
does it solve a class of problems. Also, the solution is approximate unless the problem is so
simple that a convenient exact formula is already available.

An unsophisticated description of the FE method is that it involves cutting a structure into

several elements (pieces of the structure), describing the behavior of each element in a simple
way, then reconnecting elements at "nodes" as is if nodes were pins or drops of glue that hold
elements together.

26
This process results in a set of simultaneous algebraic equations. In stress analysis these
equations are equilibrium equations of the nodes. There may be several hundred or several
thousand such equations, which means that computer implementation is mandatory. (FE)
method has been given remarkable attention, and numerous researches are achieved since the
computers have been available for most engineering and academic establishments and
programming languages have become more versatile and less tedious. Furthermore, many
academic centers of research and specialized engineering software companies present packages
of powerful and adaptable programs for many applications and cases of study.

27