Professional Documents
Culture Documents
OF STELLENBOSCH
PIPELINE DESIGN
FOR WATER ENGINEERS
IS van der Merwe
5/15/2017
Contents
1 INTRODUCTION .............................................................................................................................. 3
2 PIPELINE HYDRAULICS CALCULATIONS .................................................................................... 4
2.1 FUNDAMENTAL PIPE HYDRAULICS EQUATIONS .............................................................. 4
2.2 HYDRAULIC GRADELINE AND ENERGY LINE .................................................................... 4
2.3 TOTAL HEAD LOSS EQUATION ............................................................................................ 6
2.4 FRICTION LOSSES IN PIPES ................................................................................................ 6
2.4.1 Darcy Weisbach formula and Moody Diagram .................................................................... 7
2.4.2 Colebrook-White equation ................................................................................................... 9
2.4.3 Hazen-Williams formula ..................................................................................................... 10
2.4.4 Manning formula ................................................................................................................ 11
2.5 COMPARISON OF PIPE FRICTION FORMULAE ................................................................ 11
2.5.1 Comparative Calculations .................................................................................................. 11
2.6 MINOR LOSSES ................................................................................................................... 15
2.6.1 Minor Head Loss Equations .............................................................................................. 15
2.7 PRACTICAL DESIGN CONSIDERATIONS .......................................................................... 17
3 PIPE WALL THICKNESS DESIGN ................................................................................................ 19
3.1 INTERNAL PRESSURE ........................................................................................................ 19
3.2 EXTERNAL LOADS .............................................................................................................. 19
3.2.1 Soil load ............................................................................................................................. 20
3.2.2 Traffic Load ........................................................................................................................ 20
3.2.3 Combined Vertical Load .................................................................................................... 21
3.3 BUCKLING ............................................................................................................................ 22
3.4 BENDING AND DEFLECTION .............................................................................................. 22
3.4.1 Bending and Deflection without side support .................................................................... 22
3.4.2 Deflection with side support ............................................................................................... 23
3.4.3 Ring Bending and Arching Stress ...................................................................................... 24
3.5 PERMISSIBLE LOADS ON EMPTY PIPE ............................................................................ 26
3.5.1 Permissible Load Limited by Deflection ............................................................................. 26
3.5.2 Permissible Load Limited by Buckling ............................................................................... 27
3.5.3 Permissible Load Limited by Yield Stress ......................................................................... 27
3.5.4 Graphical Presentation of Formulae .................................................................................. 27
3.6 COMBINED STRESSES ....................................................................................................... 29
3.6.1 Method to Calculate Combined Stress .............................................................................. 29
3.7 MODULUS OF ELASTICITY OF SOIL .................................................................................. 30
3.7.1 Importance of Modulus of Elasticity ................................................................................... 31
3.7.2 Calculate Modulus of Elasticity .......................................................................................... 31
3.8 PIPE STIFFNESS .................................................................................................................. 32
3.8.1 Arching vs Bending stress Ratio ........................................................................................ 32
3.8.2 Pipe Stiffness Classification .............................................................................................. 33
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 1
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 2
1 INTRODUCTION
The water pipeline design engineer is responsible for the design of a pipe system, pumping equipment
and/or other associated pipeline equipment as applicable to deliver water from one point to another in
the most economical way taking account of the growth in water demand and the maintenance
requirements.
Driven by market competition pipelines are being constructed to ever-increasing diameters, lengths
and strengths (for both working pressures and structural strength), and from ever-newer materials with
special reference to plastics and improving steel grades. The market competition and new materials
inevitably leads to thinner pipe wall thicknesses and thus increase the risk of pipe failure. This results
in higher design risks for the pipeline design engineer. Pipe failures and their consequences are
mostly extremely costly and very disruptive.
Pumping systems are large consumers of the world’s decreasing energy resources. Life cycle costing
and minimization of the energy cost of pipeline and pumping systems are therefore of paramount
importance.
The primary aspects to be addressed by the water pipeline design engineer include the following:
• Reliable yield of the source(s) of water supply.
• Route selection: length, longitudinal section, pressure ranges, working space, slopes, access,
servitudes, excavation conditions including rock and shoring needs, water table, services, rivers,
railways, roads, traffic diversions, reinstatement requirements, and environmental impacts.
• Pipe selection: steady state hydraulics, diameter, internal pressure including waterhammer,
external loadings, wall thickness, material, corrosion protection, friction factor, external protection,
jointing, anchorage, expansion and contraction of above ground pipelines, and maintenance.
• Pipe Equipment: air valves, scour valves, in-line valves, non-return valves, control valves,
waterhammer, monitoring equipment, and corrosion protection.
• Pump selection: discharge, delivery pressure, suction pressure, rotation speed, power
requirements, efficiency, control of starting and stopping, power failure, corrosion protection,
pumpstation layout, standby equipment, and maintenance.
• System optimization by means of present value discount analysis and/or other economic
modelling.
In the following sections some of the more important aspects of the above-mentioned are addressed
in more detail.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 3
• Continuity equation: For steady, incompressible, one-dimensional flow the flow rate at any section
equates to the flow rate at any other section along the stream tube, i.e.:
Q1 = Q2 thus A1 V1 = A2 V2
• Momentum equation: From Newton’s basic law of motion follows that the change in momentum
flux between two sections equals the sum of the forces on the fluid causing the change, i.e.:
ΔFx = ρ Q Δ Vx
• Energy Equation: Most commonly known as the Bernouli equation the energy equation for steady
state flow of an incompressible fluid is as set out below showing the total energy between two
points on pipeline system. A term is included to account for energy loss due to friction and
turbulence:
𝑃" 𝑉"' 𝑃' 𝑉''
+
+
𝑍"
=
+
+
𝑍' + ℎ-
γ 2𝑔 γ 2𝑔
Where:
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 4
Figure 1: Ideal Fluid
Figure 2: Sudden Contraction of Pipeline diameter
Figure 3: Sudden Enlargement in Pipeline diameter
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 5
Figure 4: Pipeline with varying ground profile
Figure 5: Gradual decreasing and increasing pipeline
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 6
The friction factor (𝑓 ) in the Darcy Weisbach formula depends on the Reynolds Number (𝑁= ) and
absolute wall roughness (e) of the pipe. The Reynolds Number in turn depends on the density (ρ) and
absolute viscosity (µ) of the fluid, as follows:
𝐷
𝑉𝜌 𝐷
𝑉
𝑁=
=
=
µμ 𝑣
Where:
𝑣 =
-6
Kinematic viscosity (equal to 1.003 x 10 m /s for water at 20 ºC)
2
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 7
1 𝑁= 𝑓
=
2 log
𝑓 2.51
• Transition zone: The friction factor for flow in the transition zone between smooth and
rough pipes depends on the Reynolds Number and the surface roughness as described by
the Colebrook-White equation. See Section 2.4.2 for further discussion.
• Complete turbulence rough pipes: In this region the friction factor is constant as shown to
the right of the dotted line on the Moody diagram (Figure 2.6). The friction factor is only
dependent on the relative roughness (e/D) and could be expressed as
1 3.7
𝐷
=
2 log
𝑓 𝑒
Notice that for a smooth pipe, the friction factor (𝑓 ) is independent of the relative roughness (e/D) and
for a rough pipe it is independent of the Reynolds number (𝑁= )
Figure 6: Moody Diagram
Typical absolute roughness values of pipe materials are given in Table 2.1.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 8
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 9
Figure 7: Friction factor results for commonly installed pipe range
0.100
1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 1.E+08
DN 50, e = 0.6
DN 1000, e = 0.6
2.4.3 Hazen-Williams formula
The empirical nature of the Hazen-Williams formula makes direct calculation of flow possible and trial
and error analysis is not necessary.
The Hazen-Williams empirical formula applies to the transition zone of the Moody diagram and is
given by:
𝑉
=
0.849
𝐶Z
𝑅\ S.]^
𝑆 S.P`
Where:
𝐶Z = Hazen-Williams roughness coefficient
𝑅\ = Hydraulic radius ( equal to 𝐷 4 for pipes)
S = Head loss per unit length (ℎ7 /L )
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 10
Table 2.2: Average values of the Hazen-Williams roughness coefficient for various materials
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 11
Table 2.4: Equivalent friction factor/roughness coefficients for different equations, and absolute
roughnesses (e)
Table 2.5 presents approximate average values as calculated from similar calculations to those shown
in Table 2.4 in order to obtain comparative values for the Hazen- Williams and the Manning
coefficients for selected absolute roughness values
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 12
Table 2.5: Approximate Equivalent Roughness Coefficients for selected Absolute Roughness
values
Hazen-Williams
Absolute Roughness Manning Coefficient
Coefficient
e n
C
Smooth Pipe 0.015 150 0.0088
Average Roughness 0.15 135 0.0100
Aged Pipes 0.3 125 0.0106
Very rough Pipes 0.6 115 0.0144
The following set of graphs depicted in Figure 9 directly compares the friction head loss in meters per
100 m for pipeline diameters 200 mm and 600 mm, using the values for Absolute Roughness, and
Hazen-Williams and Manning coefficients as given in Table 2.5.
It follows from Figure 9 that for a 600 mm diameter pipe there is a small difference between the three
formulae. However, for a 200 mm diameter pipe there is a significant difference for pipe flow velocities
higher than 2 m/s.
Figure 9: Friction loss comparison
FRICTION LOSS COMPARISON FOR DN 200
PIPE FRICTION LOSS COMPARISON FOR DN 600
PIPE
5.0 1.2
4.0 1.0
0.8
3.0
0.6
2.0
0.4
1.0 0.2
0.0 0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Velocity (m/s) Velocity (m/s)
Colebrook HW Manning Colebrook HW Manning
6.0
Head Loss (m/100m)
5.0 1.5
4.0
1.0
3.0
2.0 0.5
1.0
0.0 0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Velocity (m/s) Velocity (m/s)
Colebrook HW Manning Colebrook HW Manning
8.0
7.0 2.0
6.0
5.0
4.0
3.0 1.0
2.0
1.0
0.0 0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Velocity (m/s) Velocity (m/s)
Colebrook HW Manning Colebrook HW Manning
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 13
Table 2.6 provides the comparative friction factor values calculated from the Colebrook-White formula,
Barr formula and the Unknown formula, for selected pipe roughness values and velocities.
Table 2.6: Comparative Friction Factors
Diameter 200 200 200 200 600 600 600 600
Velocity 0.5 1.5 2.5 3.5 0.5 1.5 2.5 3.5
Reynolds number 1.0E+05 3.0E+05 5.0E+05 7.0E+05 3.0E+05 9.0E+05 1.5E+06 2.1E+06
e
Colebrook White 0.015 0.0184 0.0152 0.0141 0.0136 0.0147 0.0124 0.0116 0.0111
Unkown Formula 0.0183 0.0152 0.0142 0.0136 0.0147 0.0124 0.0116 0.0112
Barr formula 0.0183 0.0152 0.0142 0.0137 0.0147 0.0124 0.0116 0.0112
Colebrook White 0.15 0.0213 0.0195 0.0191 0.0189 0.0166 0.0153 0.0149 0.0148
Unkown Formula 0.0214 0.0196 0.0192 0.0190 0.0167 0.0154 0.0150 0.0149
Barr formula 0.0214 0.0197 0.0192 0.0190 0.0167 0.0154 0.0150 0.0149
Colebrook White 0.6 0.0275 0.0266 0.0264 0.0264 0.0206 0.0200 0.0198 0.0198
Unkown Formula 0.0277 0.0268 0.0265 0.0264 0.0207 0.0201 0.0199 0.0198
Barr formula 0.0277 0.0268 0.0266 0.0265 0.0208 0.0201 0.0199 0.0198
Table 2.7 provides the error margin if the head loss calculated with the Hazen-Williams, Manning, Barr
and Unknown formulae were compared to that derived using the Colebrook-White formula.
Table 2.7: Head Loss Error Margin Compared to Colebrook-White formula (mm/100m)
Diameter (mm) 200 200 200 200 600 600 600 600
Velocity (m/s) 0.5 1.5 2.5 3.5 0.5 1.5 2.5 3.5
e
Hazen-Williams 0.015 1.8 10.5 19.6 0.8 0.8 8.6 17.2 19.5
Formula 0.15 4.6 45.4 273.4 737.0 3.6 5.7 26.6 109.2
0.6 13.7 82.7 493.4 1297.0 8.6 18.9 21.5 134.4
Manning
Formula 0.015 12.1 73.7 373.9 911.9 7.0 17.8 6.5 33.2
0.15 0.0 103.2 356.6 761.2 3.9 9.7 9.3 2.2
0.6 1.3 60.9 197.9 412.4 3.0 14.9 34.4 61.3
Unknown
Formula 0.015 0.5 1.0 2.4 10.5 0.1 0.0 1.3 4.0
0.15 0.8 7.4 17.9 30.8 0.1 1.7 4.2 7.3
0.6 1.5 7.6 15.1 23.3 0.3 1.6 3.3 5.1
Barr Formula 0.015 0.4 1.3 9.1 23.6 0.0 0.9 3.9 8.9
0.15 0.9 8.7 21.4 37.2 0.2 2.2 5.4 9.5
0.6 1.6 8.6 17.7 28.2 0.3 1.9 4.1 6.5
It is evident from Table 2.7 that for the conditions high-lighted the Hazen-Williams or Manning
formulae are inaccurate by approximately 1m/1km to as high as almost 13m/1km. This obviously may
have significant consequences for the design of long-distance pipelines.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 14
Typical values of minor loss coefficients (k) for different pipe fittings are given in Figure 2.14.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 15
15 15
Foot
V alve
10 10
9.0 9.0
Globe
V alve
8.0 8.0
7.0 7.0
60%
Open
For approximate
6.0 values
use
radius
6.0
designations
or
5.0 5.0
Obtain
f
from
the
Values of
K-‐ Coeffisient
Moody
Diagram
Angle
V alve then
find
K
for
particular
r/D
4.0 4.0
Lower
Value
f =
0.006
1.5 Ɵ 1.5
r/D
Swing
Check
Medium
Radius Short
Radius
Long Radius
d/D 11.5º
-‐
0
1.0 0.8 0.6 0.4 0.2 0.0 0
Sources: Modified from AWWA M11
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 16
From Table 2.7 it follows that the error margin for the Barr and Unknown formulae are very small
compared to the error margins resulting from the Hazen-William or Manning formulae, when all are
compared to the Colebrook-White formula.
Figures 2.12 and 2.13 provide recommended ranges for the use of the Unknown and Barr formulae in
order to ensure an error margin of less than 0.05% for friction head loss calculations compared to the
Colebrook-White formula. This applies for the comparative roughness coefficients as given in
Table 2.5, and for a maximum flow velocity of 6 m/s.
Figure 2.12: Accuracy Range for Unknown Figure 2.13: Accuracy Range for Barr
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 17
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 18
The following sections deal with the commonly used empirical formulae to calculate pipe wall
thickness. Finite element analysis is also a very helpful design method but limits the designer to one
selection of design parameters per setup. The empirical analysis provides the designer with practical
insights that should improve his design. It is recommended that the two design methods are always
used in parallel for complicated installations. In this section the empirical method is discussed.
Pipes can be laid in many different arrangements and conditions. In the following sections the trench
condition is discussed. The section only deals with pipes of homogeneous material and flexible or thin-
walled pipes.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 19
3.2.1 Soil load
Marston, Spangler and Clarke have developed extensive equations for evaluating soil loads on pipes
in various trench and embankment conditions. In this section only trench conditions are considered.
The vertical soil load on the pipe may be considered as the weight of the rectangular prism of soil
directly above the pipe. The soil prism would have a height equal to the depth of the earth cover and a
width equal to the pipe outside diameter.
𝑤 =
𝛾𝐻
or
𝑊 =
𝛾𝐻𝐷
(3.2)
Where:
𝑤 = Vertical soil pressure (kN/m²)
𝑊 = Vertical soil load (kN/m)
𝛾 = Unit weight of overburden (kN/m³)
H = Burial depth from the top of the pipe (m).
D = Diameter of pipe (m)
Where:
L1 = Load width parallel to direction of the travel.
= 0.253 + 1.75*H
L2 = Load width perpendicular to direction of travel
= 0.509 + 1.75*H for 0.6 m < H < 0.75 m, or
= (13.31 + 1.75*H)/8 for H > 0.75 m
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 20
The traffic load that is induce onto the pipe a depth H can be expressed as:
𝑤-
=
𝑃
𝐼7
/
(𝐿1
𝑥
𝐿2)
(3.3)
Where:
𝑤- = Traffic load pressure (kN/m²)
P = Wheel load (e.g. AASHTO HS-20 truck = 80kN per wheel)
𝐼7 = Impact factor
= 1.1 for 0.6 m < H < 0.9 m, or
= 1.0 for > 0.9 m
In certain applications the pipe may be exposed to two wheels simultaneously in which instance the
load as calculated above should be doubled.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 21
Note that for depths less than approximately 1m the traffic load decreases with increasing burial depth
resulting in a minimum total vertical load of approximately 45 kN/m². The water load is discussed in
the next section.
3.3 BUCKLING
Buried pipes are subjected to external loads composed of vertical soil loads, but most likely also
hydrostatic pressures of groundwater and/or internal vacuum. The pipe may collapse or buckle from
elastic instability or deformation even before the stress limits in the pipe wall was reached. The
allowable buckling pressure (𝑤j ) may be determined by the following:
1 𝐸𝐼 "
𝑤j =
(32
𝑅Z
𝐵
𝐸n
^ )'
𝐹𝑆 𝐷
(3.4)
Where:
𝑤j = Allowable buckling pressure (kN/m²)
FS = Design factor
= 2.5 for H/D ≥ 2
= 3.0 for H/D < 2,
Where: H = Burial depth and D = Pipe diameter (m)
𝑅Z = Water buoyancy factor
𝐻
= 1
–
0.33
Z 𝐻 , 0 ≤
𝐻Z
≤
𝐻 ,
Where: 𝐻Z = equal height of water surface above pipe (m)
B = Empirical coefficient of elastic support
"
=
"
s`t (uv.vwx
y
z./{)
The summation of all the appropriate external loads should be less than or equal to the allowable
buckling pressure as follows:
For pipelines under negative pressure:
𝛾Z
𝐻Z + 𝑅Z 𝑤 +
𝑃.
≤
𝑤j
For pipelines with possible traffic load:
𝛾Z
𝐻Z + 𝑅Z 𝑤 +
𝑤-
≤
𝑤j
The possibility that negative pressure and traffic load are experienced simultaneous is therefore
ignored.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 22
N = Bending coefficient (𝑁~ , 𝑁j or 𝑁n for top, bottom and sides respectively)
(See Table 3.1).
Table 3.1 provides the Spangler coefficients for bending and deflections of an elastic ring under
uniformly distributed vertical load.
Table 3.1: Spangler’s Coefficients for Bending Moments and Deflections for Elastic pipe.
Degrees of
Bottom Bending Top Bending Side Bending Vertical Deflection
Bottom Support
𝑁j 𝑁~ 𝑁n 𝑁€
𝜃
0 0.294 0.150 0.153 0.110
30 0.235 0.148 0.152 0.108
60 0.189 0.143 0.147 0.103
90 0.157 0.137 0.140 0.096
120 0.138 0.131 0.133 0.089
150 0.128 0.126 0.127 0.085
180 0.125 0.125 0.125 0.083
The vertical and horizontal deflections are practically equal and opposite and may be expressed as:
𝑁€
𝑊 𝑟 ^
𝛥𝑦 =
𝐸𝐼
(3.6)
Where:
𝑁€ = Vertical deflection coefficient
3
𝐼 = Moment of inertia (equal to t /12 for pipes)
E = Modulus of elasticity of pipe material
The lateral support of sidefill in a trench increase the strength of flexible pipes considerably and
reduces deformations. A pipe in compacted fill will deflected outwards laterally as it is loaded vertically
thereby increasing the pressure of the sidefill against the sides of the pipe. An equilibrium condition
may be established with the vertical load being transferred to the haunches by arch action as well as
by ring action. The vertical load per unit area on a pipe may therefore be expressed as:
𝑤
=
𝑤ƒ
+
𝑤„
Where:
𝑤ƒ = Ring bending load pressure (kN/m²)
𝑤„ = Lateral soil pressure on the side of
the pipe or active soil pressure or
arching effect. (kN/m²)
Barnard expressed the lateral support in terms of the elastic
theory (𝜎 = 𝑒𝐸). The horizontal stress will decrease away
from the pipe assuming a triangular stress distribution at the
pipe wall decreasing linearly to zero at 2.5D away from the
pipe wall. The corresponding lateral deflection of each side
of the pipe is:
𝛥𝑥/2
=
1.25
(𝑤
– 𝑤ƒ )
𝐷
/
𝐸n
(3.7)
Where:
𝐸n = Modulus of Elasticity of the soil (Refer to Section 3.7)
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 23
It is important to note that with the preceding formula it is assumed that the soil reacts within its elastic
limit. This is only true for small displacements of the soil. In the literature typical allowable pipe
deflections range from 2% to maximum 5%.
Stephenson stated that the factor 1.25 should increase as the lateral deflection increases, since the
radial pressure increases as the radius of the curvature decreases. Stephenson suggested that the
factor becomes 1.4 for a deflection of 2% of the diameter and 1.7 for a 5% deflection.
Putting 𝛥𝑌
=
𝛥𝑋
=
𝛥 and then calculating 𝑤ƒ from both equations 3.6 and 3.7, and then solving for
𝛥/𝐷 by equating the two formulae and eliminating 𝑤ƒ , the deflection of the pipe can be expressed as:
∆ 𝑁€
𝑤
𝐷^
=
𝐷 𝑁𝑦
8𝐸𝐼 +
𝐸n 𝐷^
2.5
(3.8)
Spangler allowed for lateral support to the pipe in a slightly different way to Barnard by changing the
term 𝑁€ /2.5
into a constant. He derived the following formula for vertical deflection:
∆ 𝐷-
𝑁€
𝑤
𝐷^
=
𝐷 8𝐸𝐼 +
0.061
𝐸n 𝐷^
(3.9)
Where:
𝐷- = Deflection lag factor to compensate for the time-consolidation of the soil.
(Varies from 1.0 to 1.5)
Ny = Bedding constant (Varies from 0.11 for point support to 0.083 for bedding the
full width of the pipe, normally taken as 0.1)
w = Vertical pressure due to external loads
D = Pipe diameter
The Spangler formula is very widely used throughout the pipeline industry for deflection calculations.
Deflections due to external loads on pipes under pressure are different to those of an empty pipe. The
pressure inside a pipe increases its stiffness due to the fact that the vertical diameter is compressed to
slightly less than the horizontal diameter. By adding a term for pressure to equation 3.9 a more
accurate expression for vertical deflection for pipes under pressure is obtained:
∆ 𝐷-
𝑁€
𝑤
𝐷^
=
𝐷 8𝐸𝐼 +
0.061
𝐸n 𝑑 ^ + 2𝐷-
𝑁€ 𝑝𝐷^
(3.10)
A portion of the vertical load on a pipe is resisted by bending stress due to the ring action of the pipe
wall. In order to determine the bending stress the related vertical load needs to be calculated.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 24
From equation 3.6 and 3.7 follows:
𝛥𝑦 = 𝑁€
𝑤ƒ 𝐷 `
/
8𝐸𝐼
and
𝛥𝑥 = 2.5
(𝑤
– 𝑤ƒ )
𝐷
/
𝐸n
Where:
w = Vertical soil pressure (equal to W/d)
Equating Δy and Δx and solving for vertical pressure (𝑤ƒ ):
𝑤
𝑤ƒ =
𝑁€
𝐷^ 𝐸n
(
1 +
)
2.5 8
𝐸𝐼
(3.11)
The bending moment in the pipe due to the vertical pressure 𝑤ƒ is expressed as:
𝑓ƒ
=
𝑀𝑦/𝐼
=
6𝑀/𝑡 '
(3.12)
Where:
t = Wall thickness
y = t/2
3
I = t /12
For the bottom of the pipe it follows from equation 3.5:
𝑀
=
𝑁j
𝑤ƒ
𝐷' /2
(3.13)
Replacing M in equation 3.12 with equation 3.13 gives:
𝐷'
𝑓ƒ =
6
Nj
𝑤ƒ '
2𝑡
Replacing 𝑤ƒ with equation 3.11, gives:
𝐷' 𝑤
𝑓ƒ =
6
Nj
2𝑡 ' 𝑁€
𝐷^ 𝐸n
(
1 +
)
2.5 8
𝐸𝐼
(3.14)
Or rearranged to
𝐷' 𝑤
8
𝐸𝐼
𝑓ƒ =
6
Nj '
2𝑡 𝑁€ ^
(
8
𝐸𝐼 +
𝐸
𝐷
)
2.5 n
(3.15)
Rearranging equation 3.15 in terms of the deflection formula, it is possible to express the hoop stress
in the pipe wall due to the vertical load as follows:
𝑁j
𝑡 ∆
𝑓ƒ
=
2
𝐸
𝑁€
𝐷 𝐷
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 25
(3.16)
The proportion of the vertical load that is not resisted by the ring bending action is resisted by the arch
action of the pipe. The stress due to the arch action is compressive and in the extreme case, the
lateral stress will be equal to the vertical load stress and the pipe wall will be in pure compression. The
arch action induces a hoop stress into the pipe wall, which is opposite in direction to that of positive
internal pressure, but in the same direction as negative internal pressure.
The hoop stress due to the vertical load is expressed as:
𝐷
𝑓„
=
𝑤„
2𝑡
or
𝐷
𝑓„
=
𝑤 − 𝑤ƒ
2𝑡
Replacing 𝑤ƒ with equation 3.11 gives:
𝑤 𝐷
𝑓„ = (𝑤 −
)
𝑁€ 𝐷^ 𝐸n 2𝑡
1+
2.5 8𝐸𝐼
(3.17)
Or simplified to:
𝑁€ ^
𝐷 𝐸n 𝐷
𝑓„ = 𝑤(
2.5 )
𝑁€ 2𝑡
8𝐸𝐼 +
𝐸 𝐷^
2.5 n
(3.18)
Rearranging equation 3.18 in terms of the deflection formula, it is possible to express the hoop stress
in the pipe wall due to the vertical load as follows:
𝐸n
𝐷 ∆
𝑓„
=
5
𝑡 𝐷
(3.19)
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 26
8 ^ ^
∆ 12 𝐸𝑡 +
0.061
𝐸n 𝐷
𝑤
=
𝐷 𝐷-
𝑁€
𝐷^
8
∆ 𝐸
𝑡 ^ 0.061
𝐸n
𝑤
=
{
12
^
+
}
𝐷 𝐷-
𝑁€
𝐷 𝐷-
𝑁€
12 𝑁€ 𝐸n 𝐷^
1+
8 2.5 𝐸 𝑡 ^
𝑤 =
𝑥
𝑓€‹t-Œ
𝐷' 12 𝑁€ 𝐸n 𝐷`
3
Nj '
+
`
𝑡 8 2.5 𝐸 𝑡
(3.21)
In Figure 14 the permissible loads as calculated from equations 3.20, 3.4 and 3.21 are plotted for a
typical 1000 mm diameter pipe installation for a range of wall thicknesses. The dotted line on the
figure represents the total external load that the pipe is designed for.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 27
It is evident from Figure 3.2 that the pipe will buckle if the wall thickness is too thin even though the
stress limitation is not exceeded.
Deflection is rarely a limiting criterion in designing steel pipes. The pipe wall stresses or the possibility
of buckling usually rule. Pipe wall stresses are discussed in the next section.
Figure 14: Permissible loads
In Figure 15 the permissible loads determined for Buckling (Equation 3.4) and for stress
(Equation 3.21), for the most commonly used range of D/t ratios for three different pipe materials,
namely Steel, uPVC and HDPE, and for an Es-value of 3 MPa, are plotted. It is evident from Figure 15
that buckling is the limiting criterion for uPVC and for HDPE, even if the pipe wall is very thick. For
steel pipes with D/t ratios smaller than 150, stress becomes the limiting criterion. The same more or
less applies for Es-values higher than 3 MPa.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 28
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 29
3.7 MODULUS OF ELASTICITY OF SOIL
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 30
From Figure 17 it follows that the combined stress in the pipe wall of an empty pipe varies from
approximately 450 MPa to 150 MPa for E-values of 3 MPa to 20 MPa respectively. For 3 MPa side
support the combined stresses for the empty pipe scenario determines the pipe wall thickness, while
for the higher Es-values the combined stresses for the pressure pipe scenario determine the wall
thickness.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 31
of Elasticity (Es)
(MPa)
Water logged, marshy or very clayey soils trench
3
conditions e.g. river and stream crossings
Firm and dry natural trench conditions with coarse sand
7
no fines bedding and blanket, well compacted.
Natural soft or harder rock trenching with coarse sand no
13
fines bedding and blanket, well compacted.
Note: The Es-values assumes well compacted bedding and blanket.
It is also highly recommended that the modulus of soil reaction is confirmed from geological laboratory
and field tests prior to pipe design, instead of only relying on using the values given in AWWA M11
and M45.
It follows from Figure 18 that for bad trench conditions (Es-value = 3 MPa) the pipe cannot transfer
stress to the side fill, and the arching stress (𝑓„ ) is very small compared to the ring bending stress. For
low D/t ratios (D/t < 100) the pipe is stiff and bends more difficult and therefore cannot transfer the
load to the surrounding soil even if the Es-value is high and therefore the arching stress (𝑓„ ) is still
small compared to the ring bending stress.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 32
However for pipes with low E-value (e.g. uPVC and HDPE) the opposite is true. The pipe deflects
easy and transfer the load to the surrounding soil. Pipes with a low E-value (low stiffness) are
therefore much more dependent on the soil condition than steel pipes with a high E-value (high
stiffness). Pipes with a low stiffness will therefore probably fail because of deformation (buckling)
rather than over stressing of the pipe wall.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 33
will therefore be invalid to use for design. It is again recommended that the maximum displacement is
determined by laboratory investigations.
In extreme cases it is also possible to reduce costs by installing pipe stiffeners. Stiffeners are not
discussed in this section.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 34
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 35
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 36
Fr F1
F2
The resultant force is the vector sum of the forces in both directions of the pipe axis, that is:
𝐹ƒ = 2𝑝𝐴 sin 𝜃/2
This resultant force will typically consist of a downward, horizontal and upward component. The thrust
block needs to be designed in order to resist all three components. For each force component a
different design approach is required as discussed in the following sections.
In order to prevent overturning or unbalanced stresses, a thrust block should be designed as far as
possible so that the line of action of the resisting forces coincides with the centre line of the pipe or in
other words with the direction of the resultant force.
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 37
Bearing
Type of Soil Geological Pick Test Pressure
kPa
Hard Sound Rock Broken with some difficulty and rig when struck. 10 000
Dry 100
Loose Uniform Sand
Saturated 40
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 38
The active pressure is considerably less than the passive pressure. Typical values for 𝑐 and ∅ is given
in Table 4.2. The corresponding passive and active pressures when using the abovementioned
formula are also shown.
Table 4.2: Typical Bearing Pressure for Horizontal Thrust
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 39
BIBLIOGRAPHY :
Robert L. Daugherty, Joseph B. Franzini, “Fluid Mechanics with Engineering Applications”, Seventh
Edition, McGraw-Hill (1977).
Stephenson D, "Pipeline Design for Water Engineers", Elsevier Scientific Publishing Co., Amsterdam
(1979)
A. Chadwick, J. Morfett, M Borthwick, “Hydraulics for Civil and Environmental Engineering”, Fourth
Edition, Spon Press (2004).
American Water Works Association, “Fiberglass Pipe Design”, AWWA M45, (1996)
American Water Works Association, “Steel Pipe – A guide for Design and Installation”, AWWA M11,
(1989)
South African Institute of Civil Engineers, “Short Course in Pumping Station Design”, September 1993.
Moody LF, "Friction factors for pipe flow", Trans. ASME 66.8, November, 671-84 (1944)
N.W.B Clarke, “Buried Pipelines, A manual of structural design and installation”, Maclaren and Sons,
(1968)
2015.08-29
Prepared by IS van der Merwe 31 August 2015 Revision 3 Page 40