Buried Piping Analysis

Discretization

The analysis of a buried piping system requires special modeling consideration. This is
because the (support) restraint provided by soil surrounding a buried pipe is continuous.
Since an AutoPIPE analysis is based on discretely defined points, an accurate model of
the soil's restraint capabilities would require the definition of a large number of closely
spaced piping points. Each soil point would then require a set of support springs which
model the stiffness(es) provided by the soil at that point. The effort required to construct
such a model is logistically impractical. Therefore, a rational method is needed in order to
determine the number and distribution of soil points required to perform an analysis
which produces reasonably accurate results.

When bends (i.e. elbows, miters, & field bends) or tees (including any other multi-pipe
connections) are present in a buried piping system, significant bending moments and
axial forces can be generated in the region of these geometric features under thermal,
pressure, or seismic load applications. At some distance away from the bend (or tee) the
bending moment will become very small leaving the pipeline subjected to axial force
only. In order to obtain a cost/time effective yet accurate stress analysis, the spacing
between soil points near the bend should be small in order to simulate the pipe bending
and transverse soil bearing interaction. However, further away from the bend, the only
effect to be modeled is the transfer of soil friction to axial load in the pipe. The axial
effect can be adequately modeled using longer spacing between soil points.

Guidelines for choosing soil point spacing in an arbitrary buried pipeline have been
developed from considerations of the behavior of a flexible beam on an elastic foundation
and pipe anchorage effects. Figure D-1(A) shows a flexible beam on a continuous elastic
foundation subject to a point load R. The distribution of displacement (and bearing
reaction), shown in Figure D-1(B), is given by

The length of Zone 1 (Lb) is identified as the point where the lateral displacement first
becomes zero (closest to the bend). This length is known as the bearing span, and is given
by:
The length of Zone 1 plus Zone 2 is identified as the point where the axial frictional
resistance provided by the soil decreases to zero (no pipe displacement relative to the
soil). This distance is known as the virtual anchor length (La). The virtual anchor length
cannot be quantified directly since its value depends on the type of soil force-
displacement response. However, it can be estimated fairly well based on another term
known as the maximum slippage length (Lm).

Consider a buried pipe, shown in the above figure, where the soil force vs. displacement
response is rigid-plastic. When the pipe is subjected to a thermal load (causing
elongation), the longitudinal yield strength of the soil (P1) is generated in order to restrain
the pipe growth. The maximum slippage length can be determined by equating the length
of pipe over which the yield strength of the soil must act in order to resist the thermal
pipe load. Thus, in order to achieve a balance of force
The position of the maximum slippage length (Lm) relative to the plastic strain region
depends on the shape of the soil force vs. displacement response curve. However, the
calculated value of Lm will always be less than the virtual anchor length (La) for elastic-
plastic, and elastic response curves. Therefore, the following formula (based on past
studies) should be used to estimate La regardless of the type of soil force-displacement
response.

where

co = Dimensionless constant which lies in the range from 1.0 ("stiff" soils) to about 2
(soft soils).

The magnitude of co can be greater than 2.0 if the longitudinal yield strength (P1 ) is not
achieved by the soil.Comparison of analysis results for similar system models with
various virtual anchor lengths indicate that estimated values of co should be limited to the
following range (unless a larger value can be justified):

1.5 co 2.0

The system model can be terminated and anchored at a distance away from a bend (or
tee) equal to the virtual anchor length (La) without any loss of accuracy to the analyzed
results. Depending on the overall configuration of the piping system and the type of
analysis to be performed, the inclusion of Zone 3 in the system model may or may not be
necessary or desirable. This decision is left to the judgement of the Engineer/modeler.

Defining Soil Points in AutoPIPE

The process of defining a buried piping system is a combination of user defined piping
points, and internally generated (by AutoPIPE) soil points. The user only needs to define
piping points for identifying the following critical parts of a buried piping system:

1. As required by changes in the system geometry.

2. For specification of piping components (e.g. valves, reducers, flanges, anchors, etc.).
3. Where soil properties change.
4. Where the maximum spacing (between the internally generated soil points) defined for the
current soil identifier is to be changed.
Calculation of Soil Restraint Properties
Ideally, a geotechnical engineer should evaluate the specific soils to be modeled in
AutoPIPE, and supply design values for the required input variables (K1, P1, & K2).
However, these variables are not always included in a soil report. Therefore, this section
has been provided in order to offer the piping designer guidelines for estimating (ranges
of) the required input data. Remember, these are guidelines only, soil conditions can vary
greatly and each situation should be reviewed by a geotechnical engineer (e.g. frozen
soils will require additional consideration).

Typically, laboratory tests and field studies can provide the following soil property data:

unit weight ( )
water content
relative compaction
undrained shear strength (Su)
cohesion (c)
angle of internal friction ( )
angle of friction of soil against pipe face ( )

An elastic-plastic response (K2 = 0) will be assumed for the determination of the soil
restraint force vs. displacement relationship. This is an assumption which is consistent
with accepted practice, and provides some conservatism of design. Also, any surcharges
(e.g. a foundation load at ground level) effecting soil properties must be considered in
addition to the methods defined in the following sub-sections.

First, the soil(s) should be identified according to a standard system of classification.

Where applicable, the Unified Soil Classification System (see Table D-2, and Figure D-9)
will be referred to for defining soil types. Once a soil is classified, equations can be
selected for determining the soil properties required by AutoPIPE. Knowledge of the soil
type also helps when property values must be estimated. Table D-3 has been supplied in
order to provide comparative values for , c, and .
NOTES
1. Atterberg limits plotting in the hatched area (see Figure D-9) are borderline classifications requiring use of
dual Group Symbols. The equation of the "A-line" is as follows: Ip = 0.73(WL - 20)
2. For soils having 5 to 12% of fines passing through a No. 200 sieve, use dual Group Symbols (e.g. GW-GC).
3. Variable Definitions:

Cu = uniformity coefficient
Ip = plasticity index (%)
Cz = coefficient of curvature
wL = liquid limit (%)
The properties for soil backfill around a buried pipe can vary greatly depending on the
type of soil, and the degree of backfill compaction. General guidelines for estimating soil
properties based on these factors are summarized in Table D-4.
 Soil Stiffness are defined in the local coordinate system which correctly captures
soil interaction for sloped pipe.

AutoPIPE performs automatic mass lumping at the soil springs, whereas the
Caesar approach is to set the pipe density =0 which can cause problems e.g.
sloped pipe.

Trans Horizontal Resist pipe movement along the local z axis.

Longitudinal Resist axial pipe movement (along the local x axis)

Trans Vertical Up Resist movement in the local +y axis

Trans Vertical Dn Resist movement in the local y axis

Non-Vertical Pipe
 Horizontal
Axial or Longitudinal

Vertical Pipe
Transverse Horizontal Soil Properties

The resistance to transverse deflections of buried pipelines can be estimated using

information derived from studies of the behavior of laterally loaded piles and the passive
resistance of anchor plates. Studies have shown that the value of K1 increases
approximately linearly with depth in both sandy and clayey soils. Thus the value of K1
can be expressed in the form:

where

K1 = initial slope of the transverse resistance-displacement curve (F/L2).

Z = depth from ground surface to center of pipe (L).
ki = soil stiffness parameter (F/L3), see Table D-5 for clays or Table D-6 for
sands.

Typical values of the soil stiffness parameter (ki) are given in Table D-5 for clay, and in
Table D-6 for sand. For both clay and sand, the resistance to transverse deflection
depends strongly on whether the backfill is compacted or not. Uncompacted backfill
offers little resistance to pipe displacement, and would be characterized by values on the
order of 10 lb/in3. Compaction increases the strength of clays and the relative density of
sands. Well compacted backfills of either sandy or clayey soils would be characterized by
values on the order of 100 lb/in3.
The ultimate resistance (P1) of a backfill soil to a transverse pipe displacement is
governed by the strength of the soil. An estimate of P1 may be based on passive earth
pressure theories for anchor blocks. The following sub-sections describe equations for P1
where the backfill soil is clay based, and sand based respectively.

Clay Based Soils

For buried pipes backfilled with clay, the value of P1 can be calculated from the
expression:

where:

P1 = ultimate soil resistance to transverse pipe

displacement, assuming K2 equals zero (F/L).
Rc = coefficient whose value varies with the depth of
embedment (dimensionless). Refer to Table D-
7.
Su = undrained shear strength (F/L2).
d = outside diameter of pipe (L).

Sand Based Soils

For buried pipes backfilled with sand, the value of P1 can be calculated from the
expression
where

P1 = ultimate soil resistance to transverse pipe

displacement, assuming K2 equals zero (F/L).
Rs = coefficient whose value varies with the depth of
embedment and the relative density of sand
(dimensionless). Refer to Table D-8 below.
d = outside diameter of pipe (L).
Z = depth from ground surface to center of pipe (L).
= effective unit weight of the soil (F/L ).
= sat - H2O for depths below the water table.
= dry + w( sat - dry) for depths above the water table.
Longitudinal Soil Properties

Based on studies of pile behavior under conditions of axial load, a relationship between
skin friction (or adhesion) and longitudinal displacement of a pipe has been developed.
Values for K1 and P1 are dependant upon the diameter of the buried pipe, the depth of
burial, and the properties of the backfill soil. The following sub-sections describe
equations for K1 and P1 where the backfill soil is clay based, and sand based
respectively.

Clay Based Soils

For saturated clay soils ( = 0), the ultimate longitudinal soil resistance (P1) can be
calculated from the expression

where:

P1 = ultimate soil resistance to longitudinal pipe

displacement, assuming K2 equals zero
(F/L).
d = outside diameter of pipe (L).
= adhesion factor (dimensionless).
Su = undrained shear strength (F/L2).

Figure D-10 shows values of which have been suggested by various investigators.
Also, see Table D-9 for estimates of the adhesion ( Su) of cohesive soils in contact with
various piping materials.
 Figure:D-10

For piles, the displacement required to mobilize the full adhesion resistance is on the
order of 0.2 to 0.5 inches for 12 inch diameter piles. On this basis, the displacement
which corresponds to P1 can be expressed as a function of the pipe diameter, and it lies in
the following range:

Therefore, the value of K1 can be estimated to be in the range:

Sand Based Soils

In sandy soil, the value of P1 can be expressed in terms of the angle of frictional
resistance between the pipe and the backfill ( ). Values of , suggested by two
investigators, are given in Tables D-9 and D-10.
As shown in Table D-9, the recommended values of depend only on the type of pipe
material in contact with the soil. Table D-10 shows the variation in values of for several
granular soil types

For sandy soils, the ultimate longitudinal soil resistance (P1) can be calculated from the
expression

where:

P1 = ultimate soil resistance to longitudinal pipe

displacement, assuming K2 equals zero
(F/L).
d = outside diameter of pipe (L).
ks = coefficient of lateral earth pressure
(dimensionless).
Z = depth from ground surface to center of pipe
(L).
= effective unit weight of the soil (F/L3).
 = sat - H2O for depths below the water table.
= dry + w( sat - dry) for depths above the
water table.
= angle of friction of soil against pipe face
( ).

Values of ks may vary widely depending upon the relative density or compaction of the
backfill soil. For loose sands, ks may be as low as 0.25; while for compacted backfill, ks
may range from 0.5 - 1.0. The magnitude of the displacement which corresponds to P1 is
about the same as it is for clays (see Equation D-12). Thus, the value of k1 can be
estimated to be in the range defined by Equation D-13. Considering possible variations
and uncertainties in the estimated values of the contributing soil parameters, a range of
the soil resistance variables (K1, and P1) should be determined.

Vertical Soil Properties

Downward Pipe Movement

The vertical support provided by a soil against downward movement of a buried pipe
may be expressed in terms of conventional bearing capacity theory. The pipeline is
assumed to act as a cylindrically shaped continuous strip footing. Thus, the ultimate soil
reaction (P1) for all soil types can be determined from the expression:

where

P1 = ultimate soil resistance to downward pipe

displacement, assuming K2 equals zero
(F/L).
d = outside diameter of pipe (L).
c = cohesion (or Su) of soil below pipe (F/L2).
H = depth from ground surface to bottom of pipe
(L).
= effective unit weight of the soil, not the
backfill (F/L3).

Refer to Table D-11 below for the proper effective unit weight equation relative to the
location of the water table.

Ng,

Nc,
Nq = bearing capacity factors (dimensionless). Refer to Figure D-11.

Figure D-11 was taken from Leonards, Foundation Engineering, McGraw-Hill, 1962
(after Terzaghi and Meyerhof). Either the Meyerhof or the Terzaghi curves can be used
for dense, stiff soils. Whereas the Terzaghi local shear curves should be used for loose,
compressible soils.

If the angle of internal friction ( ) is zero, as for saturated clay in undrained shear, the
first and third terms in Equation D-15 become very small and only the cohesion
contributes materially to the bearing capacity. Thus, for all practical purposes in a
saturated clay:

For this case, values of Nc with respect to the depth of embedment and pipe diameter
have been developed as shown in Figure D-12. In this chart, the breadth (B) is equal to
the outside diameter of the pipe (d). Figure D-12 is after Skempton (1951), and was taken
from Poulos and Davis' Pile Foundation Analysis and Design, Wiley, 1980.

The displacement required to mobilize the full soil resistance is generally considered to
be on the order of 10% to 15% of the outside diameter of the pipe. On this basis, the
displacement which corresponds to P1 can be expressed as a function of the pipe
diameter, and it lies in the following range
Therefore, the value of K1 can be estimated to be in the range:

Upward Pipe Movement

Design approximations for this type of movement are based on studies of earth anchors
subjected to vertical uplift forces. The uplift resistance consists of two parts: the weight
of the soil wedge (Ws) above the anchor; and the shear resistance of the soil wedge being
moved upwards.

The shape of the soil wedge is a major factor in the total uplift resistance. Figure D-13
shows several possible shapes of breakout wedges which have been observed in tests on
various types of soils. Figure D-13(A) indicates the condition for a wedge of uniform
width, which produces the lower bound value of uplift resistance. Figure D-13(B) shows
a tapered wedge which produces higher uplift resistance; the angle of the taper has been
found to be related to the angle of internal friction ( ). The shape shown in Figure D-
13(C) has been observed in tests involving deep embedment in dense sands or clays. For
weak soils or remolded soils, as is the case for un-compacted backfill, the shape of the
breakout wedge will likely correspond to Figure D-13(A). It is recommended that this
shape of wedge be used for design.

Thus, the ultimate soil resistance (P1), for all soil types, can be determined from the
expression:
where:

P1 = ultimate soil resistance to upward pipe displacement, assumes K2=0 (F/L).

Ws = effective weight of soil wedge per unit length (F/L). Use the equations for
(shown below) when considering the water content of the backfill.
d = outside diameter of pipe (L).
c = cohesion (or Su) of soil above pipe (F/L2).
= effective unit weight of the soil (F/L3).
= sat - H2O for depths below the water table.

= dry + w( sat - dry) for depths above the water table.

D = depth from ground surface to top of pipe (L).

Fc,
Fq = breakout factors (dimensionless) see Figure D-14 and Figure D-15.

The curves for Fc and Fq, shown in Figures D-14 and D-15, are expressed as functions of
the ratio of depth of burial to pipe diameter (D/d), and the angle of internal friction ( ).
Clay Based Soils

Tests indicate that relatively large displacements, on the order of 10% to 20% of the
depth to the top of the pipe (D), are required to mobilize the full uplift resistance. On this
basis, the displacement which corresponds to P1 can be expressed as a function of the
depth "D", and it lies in the following range:

Therefore, the value of K1 can be estimated to be in the range:

Sand Based Soils

Tests indicate that relatively small displacements, on the order of 1% to 2% of the depth
to the top of the pipe (D), are required to mobilize the full uplift resistance for dense to
loose sands. On this basis, the displacement which corresponds to P1 can be expressed as
a function of the depth "D", and it lies in the following range:

Therefore, the value of K1 can be estimated to be in the range:

Soil Example:-
Coordinates of A00 : DX = 12’,DY= 12’,DZ = 0
Sp.Gr. contents = 0.85
Pressure= 100 psi
Temperature = 300 deg F
Gap above V-stop = 2”
Friction factor = 0.2

Soil Conditions:-

Gravel-Sand mixture.
Dense compaction of backfill
Water table is well below the level of pipe
Rho_soil = 125 lb/ft3