5 Pipe Loops Versus Compression: 2785 - C005.fm Page 177 Friday, April 1, 2005 3:30 PM

2785_C005.
fm Page 177 Friday, April 1, 2005 3:30 PM
CHAPTER 5
Pipe Loops versus Compression
In this chapter we will explore the need for installing pipe loops in order to increase
the throughput in a gas pipeline. Looping will be compared to another means of
increasing pipeline capacity, such as installing compressor stations. The advan-
tages and disadvantages of looping pipes vs. adding compressor stations will be
discussed.
5.1 PURPOSE OF A PIPE LOOP
The purpose of a pipe loop that is installed in a segment of a pipeline is to essentially

reduce the amount of pressure drop in that section of pipe. By doing so, the overall
pressure drop in the pipeline will be reduced. This, in turn, will result in an increased
pipeline flow rate at the same inlet pressure. Alternatively, if the flow rate is kept
constant, reduction in total pressure required will cause a reduction in pumping
horsepower. This is illustrated in Figure 5.1.
The pipe loop can be constructed of the same-diameter pipe as the main
pipeline, or in some cases it can be of a different size. As we have seen in the
analysis of parallel pipes in Chapter 3, the same diameter of pipe loop will result
in equal volumes of gas flow in the main pipe as well as the loop. Thus, an NPS
20 pipe looped with an identical NPS 20 pipe segment will reduce the flow to
one-half its original value in each pipe. If the loop is larger or smaller in diameter
compared to the main pipeline, the volume distribution will not be equal. An NPS
20 pipe looped with an NPS 16 pipe will result in approximately 64% of the flow
rate going through the larger-diameter pipe and 36% through the smaller-diameter
pipe.
177
Copyright 2005 by Taylor & Francis Group, LLC

2785_C005.fm Page 178 Friday, April 1, 2005 3:30 PM
178 GAS PIPELINE HYDRAULICS
without loop
with loop
Pressure
loop
A C B
Figure 5.1 Effect of pipe loop.
5.2 PURPOSE OF COMPRESSION
We have seen in Chapter 3 and Chapter 4 that installing intermediate compressor

stations along a pipeline will increase the flow rate and also reduce the operating
pressure in a long gas transmission pipeline. The installation of the intermediate
compressor station will result in additional operational and maintenance issues in
comparison with pipe loops. Sometimes, additional compression is installed to
increase flow rate in preference to looping the pipeline, since looping will involve
additional permitting and right-of-way issues and could cost considerably more than
adding the new compressor station. Installation of an intermediate compressor to
increase flow rate is illustrated in Figure 5.2.
without second compressor station

1440 psig
with
seco
nd co
mpre
ssor
statio
n
800 psig
Flow Q = 100 MMSCFD
Dover NPS 16 pipeline Leeds

Kent
Distance
Figure 5.2 Adding a compressor station.

PIPE LOOPS VERSUS COMPRESSION 179
5.3 INCREASING PIPELINE CAPACITY
Consider an existing pipeline that is currently limited by the operating pressure that
is close to the MAOP of the pipeline. Suppose the capacity of an NPS 16 pipeline
is 100 MMSCFD and the discharge pressure at the originating compressor station is
1440 psig, as shown in Figure 5.2. It can be seen that, at the given flow rate and
discharge pressure, the delivery pressure is 800 psig. If the pipeline flow rate is
increased to 120 MMSCFD without changing the originating pressure of 1440 psig,
the increased flow will cause greater pressure drop and, hence, the delivery pressure
at the pipeline terminus will drop to some value such as 600 psig. The reduced
delivery pressure may or may not be acceptable to the customer receiving the gas.
However, we cannot increase the discharge pressure at the beginning of the pipeline
to compensate for the drop in delivery pressure because the pressure is already at
the MAOP level. How can we increase the flow rate and still provide the same
delivery pressure as before? By installing an intermediate compressor station as
shown in Figure 5.2, we can pump the increased volume approximately halfway and
then boost the pressure at the new compressor station to the same MAOP level for
ultimate delivery to the pipeline terminus at 800 psig, as before. This is illustrated
in Figure 5.2.
Thus, we have been able to achieve the increased pipeline capacity of 120
MMSCFD by installing an additional compressor station at approximately the half-
way point along the pipeline. Suppose we want to increase the flow rate further
without changing the discharge pressure or the delivery pressure. It is clear that we
could install additional intermediate compressor stations as needed to achieve the
increased throughput, while maintaining the same delivery pressure. This is illus-
trated in Figure 5.3, where two additional compressor stations have been installed
to increase the pipeline throughput while maintaining the desired delivery pressure
at the pipeline terminus.
1440 psig
800 psig
Dover Leeds
Kent
NPS 16 pipeline
Distance
Figure 5.3 Multiple compressor stations.

Pha MOP = 1440 psig

se 1
Ph Q=
as 238
.41
e MM
2 SCF
Q D
Pressure
= ini
28 tia
8. lQ
41 =
M 18
M 8.4 800 psig
SC 1M
FD MS
CF
D
Q
800 psig
Windsor Avon Hart Cardiff
m.p. 0 m.p. 37.55 m.p. 57.33 m.p. 100
NPS 16 pipeline 100 mi long
Figure 5.4 Windsor to Cardiff pipeline.
However, there is a limit to the number of compressor stations that can be

installed in a given pipeline system, since the HP required continues to increase
with flow rate and, hence, the capital cost and operating costs increase as well. At
some point, the cost increases at a very high rate compared to the increase in flow
rate. Each pipe size has a particular volume that can be economically transported
based upon cost. An additional factor that must be taken into consideration as the
flow rate is increased is the resulting higher velocity. As indicated in Chapter 2, the
gas velocity must be well below the erosional velocity for the pipe.
Example 1
A natural gas pipeline is 100 mi long and is constructed of NPS 16 and 0.250 in.
wall thickness and runs from Windsor to Cardiff, as shown in Figure 5.4.
1. Neglecting elevation effects, calculate the maximum throughput capability of this

pipeline, based upon an MAOP of 1440 psig and a delivery pressure of 800 psig at
Cardiff. The suction pressure at Windsor is 800 psig.
2. Determine the requirement for two expansion scenarios. The phase 1 expansion
will increase pipeline throughput by 50 MMSCFD and phase 2 will increase through-
put by another 50 MMSCFD. In each case, calculate the number of compressor
stations and HP required. The gas flow velocities must be checked to ensure that they
are within erosional limits.
3. Also estimate the approximate cost for each of these cases, using an overall installed
cost of $2000 per HP for the compressor stations.
4. Compare these expansion cases using pipe loop instead of compression. Thus, for
phase 1, instead of building intermediate compressor stations, calculate the amount

of pipe loop needed to reduce the pressure drop at the higher flow rate. Similarly,
for the phase 2 flow rate, calculate the looping necessary to maintain pressures without
adding compressor stations. Estimate the cost of the expansion scenarios using pipe
loops instead of compressor stations, based upon an overall installed cost of $500,000
per mile of loop.
Assume a transmission factor of 20, gas flow temperature of 80°F, and compressibility
factor of 0.85 throughout. Additional data are as follows: gas gravity = 0.6, ratio of
specific heats = 1.4, base temperature = 60°F, and base pressure = 14.7 psia. The
compressor isentropic efficiency = 0.8, and the mechanical efficiency of the com-
pressor driver is 0.95.
Solution
1. First, determine the initial capacity, considering one compressor station at Windsor
providing the pressure of 1440 psig needed for delivery pressure of 800 psig at Cardiff.
Using General Flow Equation 2.4, we calculate the initial capacity, Q, of the pipeline
as follows:
0.5
 520   1454.72 − 814.72 
Q = 38.77 × 20    (15.5)2.5 = 188,410,280 SCFD
 14.7   0.6 × 540 × 100 × 0.85 
The HP required is calculated from Equation 4.15:
 0.40 
 1.40   1 + 0.85   1   1454.7  1.40 
HP = 0.0857 × 188.41 (540)  − 1 = 6357
 0.40   2   0.8   814.7 
 
6357
BHP required = = 7064
0.95
Checking gas velocities using Equation 2.26, the gas velocity at Windsor is
 188.41 × 10 6   14.7   540 

u1 = 0.002122    520   1454.7  = 17.46 ft/s
 15.52   
The velocity at Cardiff is
 188.41 × 10 6   14.7   540 

u1 = 0.002122    520   814.7  = 31.18 ft/s
 15.52   
The erosion velocity from Equation 2.31 is
0.85 × 10.73 × 540

umax = 100 = 58.94 ft/s
29 × 0.6 × 814.7

2. Next, we will calculate the compressor station requirement for the phase 1 flow
rate of
Q = 188.41 + 50 = 238.41 MMSCFD
Assume that an additional compressor station is needed for this flow rate. This will
be located at Avon at a distance of L miles from Cardiff, such that a discharge pressure
of 1440 psig at Avon will produce a delivery pressure of 800 psig at Cardiff. We will
calculate the value of the pipe length, L , using General Flow Equation 2.4 as follows:
0.5
 520   1454.72 − 814.72 
238.41 × 10 6 = 38.77 × 20    (15.5)2.5
 14.7   0.6 × 540 × L × 0.85 
Solving for L, we get
L = 62.45 mi
Next, calculate the suction pressure at Avon using 1440 psig at Windsor and consid-
ering a pipe length of 37.55 (100 − 62.45) mi between Windsor and Avon.
Using General Flow Equation 2.4, we get

0.5
 520   1454.72 − P22 
238.41 × 10 = 38.77 × 20 
6
   (15.5)2.5
 14 .7  0 .6 × 540 × 37. 55 × 0. 85 
Solving for the suction pressure at Avon, we get
P2 = 1114.85 psia = 1100.15 psig
Therefore, the compression ratio at Avon is
1454.7
r= = 1.30
1114.85
This is a satisfactory compression ratio for a centrifugal compressor.
The HP required at Windsor and Avon for phase 1 will be calculated using
Equation 4.15.
For Windsor, assuming the compressibility factor at suction is 1.0,
 0.40 
 1.40   1 + 0.85   1   1454.7  1.40 
HP = 0.0857 × 238.41 (540)  − 1 = 8044
 0.40   2   0.8   814.7 
 
8044
Therefore, the BHP required at Windsor for phase 1 = = 8468.
0.95
Similarly, the HP required at Avon is
 1.40   1 + 0.85   1   0.40 

HP = 0.0857 × 238.41  (540 )      (1 .30 ) 1.40 − 1 = 3476

 0.40   2   0.8   

3476
Therefore, the BHP required at Avon = = 3659.
0.95
The total compressor HP required at both compressor stations for the phase 1 flow
rate of 238 MMSCFD is
8468 + 3659 = 12,127 HP
Therefore, the incremental HP for phase 1 is
∆HP = 12,127 – 7064 = 5063 HP
3. This represents the additional compressor HP required for phase 1 for the extra
50 MMSCFD flow rate. The cost of this incremental HP, based on $2000 per installed
HP, is
∆Cost = 5063 × 2000 = $10.13 million
Next, check the gas velocity at the increased flow rate in phase 1 from Equation 2.26.
 238.41 × 10 6   14.7   540 

u1 = 0.002122    520   814.7  = 39.45 ft/s
 15.52   
This velocity is acceptable, since it is less than the erosion velocity of 58.94 ft/s
calculated earlier. The velocity at Windsor at the higher pressure of 1440 psig will
be lower and, hence, less than the erosion velocity.
Next, consider the phase 2 flow rate of

Q = 238.41 + 50 = 288.41 MMSCFD
Since phase 2 occurs after phase 1, where the Avon compressor station is already built,
we might have to install one compressor station between Windsor and Avon and another
between Avon and Cardiff. If we consider this phase independent of phase 1, we could
probably install two additional compressor stations between Windsor and Cardiff to
handle the phase 2 flow of 288.41 MMSCFD. For now, we will consider a compressor
station at Jenks between Windsor and Avon and another one at Hart located between
Avon and Cardiff. We will calculate the distance of L miles from Hart to Cardiff, such
that a discharge pressure of 1440 psig at Hart will produce a delivery pressure of
800 psig at Cardiff. The value of L is calculated, using General Flow Equation 2.4, as
we did before for locating the Avon compressor station, as follows:
0.5
 520   1454.72 − 814.72 
288.41 × 10 6 = 38.77 × 20    (15.5)2.5
 14.7   0.6 × 540 × L × 0.85 
Solving for L, we get

L = 42.67 mi
Therefore, the Hart compressor station will be located at a distance of 19.78 (62.45 −
42.67) mi from Avon. The suction pressure at Hart is calculated next.

0.5
 520   1454.72 − P22 
288.41 × 10 6 = 38.77 × 20     (15.5)2.5
 14.7   0.6 × 540 × 19.78 × 0.85 
Solving for the suction pressure at Hart, we get
P2 = 1201.24 psia = 1186.54 psig
Therefore, the compression ratio at Hart is
1454.7
r= = 1.21
1201.24
This is a satisfactory compression ratio.
Before determining the location of the Jenks compressor station between Windsor
and Avon, calculate the suction pressure at Avon, assuming Jenks doesn’t exist and
that the Windsor compressor station pumps directly into Avon, as in phase 1.
The suction pressure at Avon, considering 1440 psig at Windsor, is calculated using
General Flow Equation 2.4:
0.5
 520   1454.72 − P2 2 
288.41 × 10 6 = 38.77 × 20     (15.5)2.5
 14.7   0.6 × 540 × 37.55 × 0.85 
Solving for the suction pressure at Avon, we get
P2 = 915.54 psia = 900.84 psig
Therefore, the compression ratio at Avon is
1454.7
r= = 1.59
915.54
This is a satisfactory compression ratio.
Therefore, for phase 2, we will need only two compressor stations besides Windsor,
Avon at milepost 37.55 and Hart at milepost 57.33.
Next, calculate the total HP required at Windsor, Avon, and Hart at phase 2 flow
rates.
The HP required at Windsor is, using Equation 4.15,
 0.40 
 1.40   1 + 0.85   1   1454.7  1.40 
HP = 0.0857 × 288.41 (540 ) − 1 = 9731
 0.40   2   0.8   814.7 
 

9731
Therefore, the BHP required at Windsor for phase 2 = = 10,243.
0.95
Similarly, the HP required at Avon is
 1.40   1 + 0.85   1   0.40 

HP = 0.0857 × 288.41  (540)    (1.59) 1.40 − 1 = 7652
 0.40   2   0.8   

7652
Therefore, the BHP required at Avon = = 8055.
0.95
The HP required at Hart is
 1.40   1 + 0.85   1   0.40 

HP = 0.0857 × 288.41  (540)    (1.21) 1.40 − 1 = 3023
 0.40   2   0.8   

3023
Therefore, the BHP required at Hart = = 3182.
0.95
The total compressor HP required at all three compressor stations for phase 2 is
10,243 + 8055 + 3182 = 21,480 HP
The incremental HP for phase 2 compared to phase 1 is
∆HP = 21,480 – 12,127 = 9353 HP
This represents the additional compression HP required for phase 2 compared to
phase 1, for the additional 50 MMSCFD flow rate. The cost of this incremental HP,
based on $2000 per installed HP, is
∆Cost = 9353 × 2000 = $18.71 million
Next, check the velocity at increased flow rates in phase 2 from Equation 2.26.
 288.41 × 10 6   14.7   540 

u1 = 0.002122    520   814.7  = 47.72 ft/s
 15.52   
This velocity is acceptable, since it is less than the erosion velocity. The velocity at
higher pressures will be well within the limits.
4. In the preceding analysis, we accomplished the increase in flow rates for phase 1
and phase 2 by adding intermediate compressor stations. The capital cost for phase 1
expansion was $10.13 million and for the phase 2 expansion was an additional
$18.71 million.
Next, we will explore the two expansions by installing pipe loops without additional
intermediate compressor stations.
For phase 1, assume that L miles of the pipe near Cardiff will be looped. The reason
we picked this section is because in Chapter 3, we found that looping close to the

downstream end is more beneficial than looping near the upstream end, as long as
the flowing temperature was constant. Following the methodology of Chapter 3, we
will determine the equivalent diameter of the pipe loop as follows:
Assuming the loop to be of the same diameter as the main piping and L1 = L2, using
Equation 3.18, we get
Const1 = 1.0
Therefore, the equivalent diameter, using Equation 3.17, is
1/ 5
 1 + 1  2 
De = D1   = 1.32 D1 = 1.32 × 15.5 = 20.46 in.
 1  
 
Considering L miles of pipe of inside diameter 20.46 in., calculate the upstream
pressure at the beginning of the loop as shown in Figure 5.5. The downstream pressure
at Cardiff is 800 psig, and the upstream pressure P is unknown. Using General Flow
Equation 2.4,
0.5
 520   P 2 − 814.72 
238.41 × 10 6 = 38.77 × 20     (20.46)2.5 (5.1)
 14.7   0.6 × 540 × L × 0.85 
There are two unknowns, P and L, in Equation 5.1. We need another equation to
solve for both variables. For this, the pipe segment from Windsor to the start of the
loop will be examined.
Considering 1440 psig at Windsor, calculate the downstream pressure P at the begin-
ning of the loop for a pipe length of (100 – L). Using General Flow Equation 2.4,
 520   
0.5
1454.72 − P 2
238.41 × 10 6 = 38.77 × 20     (15.5)2.5 (5.2)
 14.7   0.6 × 540 × (100 − L ) × 0.85 
MOP = 1440 psig

Phas P = 962 psig
e 1Q
= 23
8 MM
SCF
D
Pressure
800 psig
Loop - 50.03 mi
238 MMSCFD
800 psig
Windsor Cardiff
m.p. 0 m.p. 100
Figure 5.5 Pipe loop for phase 1.

Eliminating P from Equation 5.1 and Equation 5.2, we solve for L as follows:
L = 50.03 mi
Substituting this value of L in Equation 5.1 and solving for P,
P = 976.76 psia = 962.06 psig
Therefore, for phase 1, without an intermediate compressor station, flow increase can
be achieved by looping 50.03 mi of pipe upstream of Cardiff.
The installed cost of this pipe loop is

50.03 × $500,000 = $25.02 million
In addition to this cost of pipe loop, we must also include the increased horsepower
requirement at Windsor for the phase 1 flow rate. Since the discharge pressure at
Windsor is still 1440 psig as before, the HP is the same as that calculated earlier.
The incremental HP is (8468 – 7064) = 1404 HP.
At $2000 per installed HP, the extra cost for incremental HP is

2000 × 1404 = $2.81 million
Thus, for phase 1, the cost of looping pipe upstream of Cardiff and increased HP
cost at Windsor compressor station is
$25.02 + $2.81 = $27.83 million
This compares with $10.13 million calculated earlier for phase 1 using a compressor
station at Avon. Even though at first sight the looping appears to be a more expensive
option, we must also consider the increased operating cost when adding a compressor
station. The annual operating cost for the compressor station can be estimated con-
sidering the fuel consumption, operating and maintenance costs, and other costs. In
Chapter 10, we will discuss more details of capital cost, operating cost, and cost of
service. For now, we will only look at capital costs.
For phase 2, at a flow rate of 288.41 MMSCFD, similarly calculate the amount of
pipe loop needed, without adding any intermediate compression.
The length of loop L required is calculated as follows:
Using General Flow Equation 2.4,

0.5
 520   P 2 − 814.72 
288.41 × 10 = 38.77 × 20 
6
   (20.46)2.5 (5.3)
 14 .7  0 .6 × 540 × L × 0 .85 
Considering 1440 psig at Windsor, calculate the downstream pressure P at the begin-
ning of the loop for a pipe length of (100 – L).
Using General Flow Equation 2.4,
 520   
0.5
1454.72 − P 2
288.41 × 10 6 = 38.77 × 20     (15.5)2.5 (5.4)
 14.7   0.6 × 540 × (100 − L ) × 0.85 
Eliminating P from Equation 5.3 and Equation 5.4, we solve for the loop length L as
L = 76.26 mi

MOP = 1440 psig
P = 1130 psig
Pressure
Pha
se 2
Q=
288
MM
SCF 800 psig
D
288 MMSCFD Loop - 76.26 mi
800 psig
Windsor Cardiff
m.p. 0 m.p. 100
Figure 5.6 Pipe loop for phase 2.
This is shown in Figure 5.6.
Substituting this value of L in Equation 5.4 and solving for P,

P = 1144.54 psia = 1129.84 psig
Therefore, the installed cost of this pipe loop is
76.26 × $500,000 = $38.13 million
In addition to the cost of pipe loop, we must include the increased horsepower
requirement at Windsor for phase 2. Since the discharge pressure at Windsor is still
1440 psig as before, the HP is the same as that calculated earlier. The incremental
HP is (10,243 – 8468) = 1775 HP more than that required for phase 1. At $2000 per
installed HP, the incremental cost is
1775 × $2000 = $3.55 million
compared to phase 1.
Thus, for phase 2 the total incremental cost of additional looping over phase 1 and
increased HP at the Windsor compressor station is
($38.13 – $25.02) + $3.55 = $16.66 million
The costs of the initial case and the two expansion scenarios for the compressor
station option and the pipe loop option are summarized in Table 5.1 and Table 5.2.
Table 5.1 Windsor to Cardiff Pipeline Expansion—Compressor Station Option
Flow, Compressor BHP Required Compression Incremental

Phase MMSCFD Windsor Avon Hart Cost, $ million Cost, $ million
Initial 188.41 7,064 — — 14.13 —
1 238.41 8,468 3,659 — 24.25 10.13
2 288.41 10,243 8,055 3,182 42.96 18.71

Table 5.2 Windsor to Cardiff Pipeline Expansion—Pipe Loop Option

Pipe
Compressor Pipe Compression Loop Total Incremental
Flow, BHP Loop, Cost, Cost, Cost, Cost,
Phase MMSCFD Windsor mi $ million $ million $ million $ million
Initial 188.41 7,064 — 14.13 — 14.13 —
1 238.41 8,468 50.03 16.94 25.02 41.96 27.83
2 288.41 10,243 76.26 20.49 38.13 58.62 16.66
5.4 REDUCING POWER REQUIREMENTS
In an existing pipeline for a given flow rate, we can calculate the HP required based
upon the number of compressor stations, their suction and discharge pressures, and
flow rate. Suppose we are interested in reducing the HP required and, hence, the annual
operating cost of the pipeline. If the flow rate is not reduced, the only way power
consumption can be reduced is to reduce the overall pressure drop between compressor
stations. If the pipeline is 100 miles long and at a flow rate of 100 MMSCFD, an
origin compressor station and an intermediate compressor station are required, each
station operating at 900 psia suction and 1400 psia discharge pressures. The HP
required will depend upon the compression ratio of (1400/900) or 1.56. Since the flow
rate is constant, HP can be reduced by increasing the suction pressure or decreasing
the discharge pressure, both of which reduce the compression ratio. Since the objective
is to operate a gas pipeline at the highest possible pressure for efficiency, we will not
reduce the discharge pressure. That leaves us the option of only increasing the suction
pressure. Suction pressure can be increased by reducing the pressure drop in the
pipeline segment upstream of the compressor station. Since the flow rate and pipe
diameter are fixed, the pressure drop in a pipe segment can be decreased by installing
a pipe loop. Therefore, looping a segment of pipeline, thereby reducing the pressure
drop, will result in a decrease in HP and annual operating cost. We will illustrate this
using an example.
Example 2
A natural gas (specific gravity = 0.60) pipeline is 130 mi long and is constructed of
NPS 20, 0.500 in. wall thickness pipeline (MAOP = 1440 psig) that runs from
Anaheim to Ventura. At a flow rate of 300 MMSCFD, an intermediate compressor
at Brentwood (milepost 70) is needed. Calculate the total HP required for both
compressor stations. In order to reduce the power consumption by 30% at the present
flow rate, it is proposed to loop the pipeline. Calculate the extent of looping required.
For simplicity, use the General Flow equation with a transmission factor F = 20 and
compressibility factor of 0.90. The gas flow temperature is 60°F and base pressure
is 14.7 psia. The base temperature is 60°F. The delivery pressure required at Ventura
is 800 psig. The discharge pressure at Anaheim is 1440 psig, and the suction pressure
is 900 psig. Use 80% isentropic efficiency and 95% mechanical efficiency for com-
pressors. The gas specific heat ratio g = 1.4.

Solution
Using General Flow Equation 2.4, calculate the discharge pressure required at the
Brentwood compressor station for 800 psig delivery pressure at Ventura.
0.5
 520   P12 − 814.72 
300 × 10 6 = 38.77 × 20     (19.0)2.5
 14.7   0.6 × 520 × 60 × 0.9 
Solving for the discharge pressure at Brentwood,

P1 = 1216 psia
Next, calculate the suction pressure at Brentwood, applying the General Flow equation
to the pipeline segment 70 mi long between Anaheim and Brentwood.
0.5
 520   1454.72 − P22 
300 × 10 = 38.77 × 20 
6
  (19.0)2.5
 14.7   0.6 × 520 × 70 × 0.9 
Solving for the suction pressure at Brentwood,

P2 = 1079.25 psia
The compression ratio at Anaheim is
1454.7
r= = 1.59
914.7
The HP required at Anaheim is calculated using Equation 4.15:
 0.40 
 1.40   1 + 0.9   1   1454.7  1.40 
HP = 0.0857 × 300  (520 ) − 1 = 7876
 0.40   2   0.8   914.7 
 
where the compressibility factor for suction conditions is assumed to be 1.0.
Considering a mechanical efficiency of 95%, the BHP required at Anaheim is
7876
BHP = = 8291
0.95
Similarly, calculate the compression ratio and BHP for the Brentwood compressor
station.
The compression ratio at Brentwood is
1216
r= = 1.127
1079.25
The HP required at Brentwood is
 1.40   1.9   1   0.40 

HP = 0.0857 × 300   (520)     (1.127) 1.40 − 1 = 1931
 0.40   2   0.8   

and the BHP is
1931
BHP = = 2033
0.95
By looping the pipe segment between Anaheim and Brentwood using NPS 20 pipe,
the flow rate through each pipe will be one-half the inlet flow of 300 MMSCFD
at Anaheim.
The suction pressure at Brentwood is calculated using the General Flow equation as
0.5
 520   1454.72 − P22 
150 × 10 = 38.77 × 20 
6
  0.6 × 520 × 70 × 0.9  (19.0)2.5
 14. 7  
Solving for P2, we get
P2 = 1371 psia
Since this pressure is more than the discharge pressure of 1216 psia calculated earlier
for Brentwood, we conclude that the Brentwood station will not be needed if we loop
the entire 70 mi pipe segment from Anaheim to Brentwood. This would reduce the
total BHP required to 8291 from (8291 + 2033) calculated earlier for the Anaheim
and Brentwood compressor stations.
The reduction in BHP is
2033
∆BHP = = 0.197, or 19.7%
8291 + 2033
Since the objective is to reduce the power consumption by 30%, we must do more
than just loop the pipe segment between Anaheim and Brentwood. We will recal-
culate the discharge pressure at Anaheim without the Brentwood compressor station,
such that the delivery pressure at Ventura is 800 psig. The reduced discharge pressure
at Anaheim due to the 70 mi pipe loop will reduce the compression ratio and,
hence, the HP at Anaheim.
Using the General Flow equation for the pipe segment between Anaheim and Brentwood,
the downstream pressure at Brentwood must equal the 1216 psia calculated earlier to
ensure 800 psig delivery at Ventura.
Therefore, considering half the total flow rate through each NPS 20 pipe section of
the loop,
0.5
 520   P12 − 12162 
150 × 10 = 38.77 × 20 
6
  (19.0)2.5
 14.7   0.6 × 520 × 70 × 0.9 
Solving for the discharge pressure at Anaheim, we get
P1 = 1310 psia

This reduced discharge pressure at Anaheim causes the compression ratio to be
1310
r= = 1.432
914.7
The revised HP at Anaheim is calculated as
 1.40   1.9   1   0.40 

HP = 0.0857 × 300   (520)     (1.432) 1.40 − 1 = 6003
 0.40   2   0.8   

6003
BHP = = 6319
0.95
Therefore, the total reduction in HP is
(8291 + 2033) – 6319 = 4005
Or the percentage reduction in HP is
4005
= 0.39 or 39%
8291 + 2033
This is well above the 30% reduction in power required. If we reduce the loop pipe
length slightly from 70 mi, we will realize the required 30% reduction. This is left
as an exercise for the reader.
5.5 LOOPING IN DISTRIBUTION PIPING
Another example of pipe loops is as follows. Consider a distribution piping system

as shown in Figure 5.7. Gas enters the pipeline at A at a flow rate of 60 MMSCFD,
and after making gas deliveries at B of 20 MMSCFD and at C of 30 MMSCFD, the
remaining 10 MMSCFD of gas proceeds to D, where an additional 10 MMSCFD
enters the pipeline, which is delivered to the terminus at E. The last segment of pipe
has a flow rate of (60 − 20 − 30 + 10) or 20 MMSCFD. Suppose it is desired to
bring in an extra 10 MMSCFD gas at D so that the delivery at E is increased to 30
MMSCFD. If the delivery pressure at E is to remain the same at 600 psig, it is clear
that the pressure at D will need to be increased to handle the extra flow rate in pipe
segment DE. This, in turn, will raise all pressures upstream of D. Thus, the pressures
60 MMSCFD 40 MMSCFD 10 MMSCFD 20 MMSCFD

600 psig
B C D
A E
20 MMSCFD 30 MMSCFD 10 MMSCFD
Figure 5.7 Distribution piping.

at A, B, and C will all increase, resulting in an increased HP requirement at A.

However, by looping the section DE, we can maintain all pressures the same as
before.
Assume that for the initial case, where the injection at D is 10 MMSCFD, the
pressure at D is 900 psig. The delivery pressure at E is to be maintained constant
at 600 psig. If the entire length of pipe DE is looped with an identical pipe size, the
equivalent diameter Deq is such that at 30 MMSCFD, the pressure drop in the
diameter Deq is the same as the pressure drop in the original pipe diameter D at 20
MMSCFD.
From General Flow Equation 2.4, the flow rate is directly proportional to the
square root of (P12 – P22) and also to the pipe diameter raised to the power of 2.5,
keeping everything else the same. P1 and P2 are the upstream and downstream
pressures in a pipe segment.
Since we want the upstream and downstream pressures for the pipe segment DE
to be the same at both flow rates, at 20 MMSCFD,
( )
0.5
20 = C P12 − P22 D 2.5 (A)
and at 30 MMSCFD,
( )
0.5
30 = C P12 − P22 Deq 2.5 (B)
where C is a constant for the pipe segment.
By dividing one equation by the other, we get

2.5
 Deq  30
 D  = = 1.5
  20
The equivalent diameter is
1
Deq = D(1.5) 2.5
If the initial pipe size of DE was 12.00 in. inside diameter, we need an equivalent
diameter of
1
Deq = 12 × (1.5) 2.5 = 14.11 in.
Next, we need to determine the loop diameter required that will produce the
equivalent diameter just calculated. Since the pressure drop in each pipe loop is the
same, if Q1 and Q2 represent the flow rates in the main pipe and loop respectively,
Q1 + Q2 = 30 (5.5)

and
30 Q1 Q2
2.5
= 2.5
= (5.6)
Deq (12) ( D)2.5
where D is the loop diameter to be calculated and the main pipe is 12.00 in. diameter.
Solving for Q1, we get
2.5
 12 
Q1 = 30  = 20.01 MMSCFD
 14.11 
and the flow rate through the loop is
Q2 = 30 − 20.01 = 9.99 MMSCFD
Therefore, from Equation 5.6,
9.99 30
2.5
=
D (14.11)2.5
Solving for D, we get
1
 9.99  2.5
D= × 14.11 = 9.09 in.
 30 
Therefore, by looping the entire length DE of the existing 12 in. diameter pipe
with a pipe having an inside diameter of 9.09 in., we will maintain the same pressure
at all points as before.
A slightly different case of looping is one in which the inlet flow at A needs to
be increased so that the increased volume can be delivered at B, while keeping all
pressures the same as before. Suppose the delivery volume at B needs to be increased
to 30 MMSCFD, without changing other deliveries or receipt. The inlet volume at
A will increase from 60 MMSCFD to 70 MMSCFD, and the delivery volume at B
will increase from 20 MMSCFD to 30 MMSCFD. We will loop the section AB
such that the pressure at A and B remain the same as before, so that the volumes
and pressures at all points downstream of B remain the same. This will be illustrated
by calculating the pressures and the size of the pipe loop required in the next
example.
Example 3
In a gas distribution pipeline, 60 MMSCFD enters the pipeline at A, as shown in

Figure 5.8. If the delivery at B is increased from 20 MMSCFD to 30 MMSCFD by
increasing the inlet flow at A, keeping all downstream flow rates the same, calculate
the looping necessary for section AB to ensure pressures are not changed throughout
the pipeline. Pipe AB is NPS 14, 0.250 in. wall thickness; BC is NPS 12, 0.250 in.

70 MMSCFD 40 MMSCFD 10 MMSCFD 20 MMSCFD

600 psig
B C D
A E
30 MMSCFD 30 MMSCFD 10 MMSCFD
Figure 5.8 Looping a distribution piping.
wall thickness; CD is NPS 10, 0.250 in. wall thickness; and DE is NPS 12, 0.250 in.
wall thickness. The delivery pressure at E is fixed at 600 psig. The pipe lengths are
as follow:
AB = 12 mi
BC = 18 mi
CD = 20 mi
DE = 8 mi
The gas gravity is 0.60, and the flow temperature is 60°F. The compressibility factor
and transmission factor can be assumed to be 0.85 and 20, respectively, through-
out the pipeline. The base pressure and base temperature are 14.7 psia and 60°F,
respectively.
Solution
First, the pressures at A and B must be calculated for the initial flow rates. Starting
at E, for a delivery pressure of 600 psig at E, the pressures at D, C, and B will be
calculated sequentially. Applying the General Flow equation for the 8 mi section DE
of inside diameter 12.25 in. and at a flow rate of 20 MMSCFD,
0.5
 520   PD2 − 614.72 
20 × 10 6 = 38.77 × 20     (12.25)2.5
 14.7   0.6 × 520 × 8 × 0.85 
Solving for PD, we get
PD = 618.02 psia = 603 psig
Next, calculate the pressure at C, considering 10 MMSCFD flow through the 20 mi

section of pipe CD, of inside diameter 10.25 in.:
0.5
 520   PC2 − 618.022 
10 × 10 = 38.77 × 20 
6
  (10.25)2.5
 14.7   0.6 × 520 × 20 × 0.85 
Solving for PC, we get
PC = 623.04 psia = 608.34 psig

Similarly, the pressure at B is calculated considering 40 MMSCFD flow through the

18 mi section of pipe BC, with an inside diameter 12.25 in.:
0.5
 520   PB2 − 623.04 2 
40 × 10 6 = 38.77 × 20     (12.25)2.5
 14.7   0.6 × 520 × 18 × 0.85 
Solving for PB, we get

PB = 651.90 psia = 637.20 psig
Next, calculate the pressure at A considering 60 MMSCFD flow through the 12
mi section of pipe AB, with an inside diameter 13.5 in.:
0.5
 520   PA2 − 651.90 2 
60 × 10 6 = 38.77 × 20     (13.5)2.5
 14.7   0.6 × 520 × 12 × 0.85 
Solving for PA, we get
PA = 677.45 psia = 662.75 psig
Therefore, the pipe section AB, when flowing 60 MMSCFD of gas, has the following
pressures:
PA = 677.45 psia
PB = 651.90 psia
When the delivery rate at B is increased from 20 MMSCFD to 30 MMSCFD, the

flow rate in pipe segment AB increases from 60 MMSCFD to 70 MMSCFD. Since
the pressures at A and B are to remain the same as before, the pipe segment AB must
be looped to reduce the pressure drop at the higher flow rate. We will assume the
entire 12 mi length will be looped. Next, we calculate the equivalent diameter required
for segment AB, using the General Flow equation, so the pressures at A and B are
the same as before.
0.5
 520   677.452 − 651.90 2 
70 × 10 6 = 38.77 × 20    ( D)2.5
 14.7   0.6 × 520 × 12 × 0.85 
Solving for the diameter,
D = 14.36 in.
The equivalent diameter of the looped line AB must be 14.36 in. to keep pressures
the same as calculated. From Equation 3.17 and Equation 3.18, the diameter of the
loop can be calculated, knowing the equivalent diameter just calculated.
From Equation 3.17,
1
 1 + Const1  2  5
14.36 = 13.50  
 Const1  
 

Solving for Const1
Const1 = 5.99
From Equation 3.18, and since L1 = L2 = 12 mi,
5
 13.5 
5.99 = 
 D2 
Solving for the pipe loop diameter D2,
D2 = 6.6 in.
Therefore, the pipe section AB must be looped with a pipe of inside diameter 6.6 in.
for the entire length of 12 mi. We could also increase the loop diameter and reduce
the pipe length that is looped to get the same effect. For example, increasing the loop
diameter to 10 in. will reduce the length of looping needed. Suppose we decide on
an NPS 10, 0.250 in. wall thickness pipe for the loop length of L mi. upstream of B.
The equivalent diameter will be calculated using Equation 3.17 and Equation 3.18.
From Equation 3.18,
5
 13.5 
Const1 =  = 1.9908
 10.25 
and from Equation 3.17, the equivalent diameter is
1
 1 + 1.9908  2  5
De = 13.50   = 15.89 in.
 1.9908  
 
The pressure at the start of the loop will be calculated from General Flow Equation 2.4:
0.5
 520   P 2 − 651.90 2 
70 × 10 6 = 38.77 × 20     (15.89)2.5
 14.7   0.6 × 520 × L × 0.85 
Simplifying,
P2 – 651.90 2 = 1705L (5.7)
Next, consider the unlooped portion of pipe AB from A to the starting point of the loop.
0.5
 520   677.452 − P 2 
70 × 10 6 = 38.77 × 20     (13.5)2.5
 14.7   0.6 × 520 × (12 − L ) × 0.85 

Simplifying,
677.452 – P2 = 3851.92 (12 – L) (5.8)
Eliminating P from Equation 5.7 and Equation 5.8 and solving for L, we get
L = 5.71 mi
Therefore, by looping the existing NPS 14 pipe from A to B with an identical NPS
14 pipe, 5.71 mi long (measured upstream from B), the pressures will be the same
as before the increased delivery volume at B.
5.6 SUMMARY
We discussed two ways to increase the throughput of a gas pipeline: using interme-
diate compressor stations and installing pipe loops. With intermediate compressor
stations, the flow rate can be increased to fully utilize pipe MAOP. However, adding
compressor stations causes increased capital cost as well as annual operating and
maintenance costs. On the other hand, by installing a pipe loop, the effective diameter
of the pipe is increased, resulting in a lower pressure drop. Therefore, additional flow
rate can be realized without installing an intermediate compressor station. Looping
an existing pipeline causes increase in capital but very little increase in operating
and maintenance costs compared to installing intermediate compressor stations. We
also discussed how the HP required can be reduced by installing a pipe loop. On
distribution piping, an example of increasing delivery rate to certain locations using
pipe loops, without changing pipe pressures in the rest of the pipeline, was also
illustrated.
PROBLEMS
1. A natural gas pipeline from Compton to Merced is 100 mi long and is constructed
of NPS 14, 0.250 in. wall thickness. The pipeline elevation profile is essentially
flat. The MAOP of the pipeline is 1280 psig. The gas delivery pressure at Merced
is 600 psig. What is the maximum pipeline throughput with an origin compressor
station at Compton? The gas gravity is 0.6 and gas flowing temperature is 80°F.
Use the Colebrook equation for pressure drop with a friction factor of 0.01. The
compressibility factor can be assumed to be constant at 0.88. If the flow rate
increases by 50 MMSCFD, calculate the increased HP required at Compton and
the HP required at an intermediate compressor station at Vale. Instead of the
intermediate compressor station at Vale, a portion of the pipe is looped. What
length of NPS 14 loop will be needed? The base pressure and base temperature
are 14.7 psia and 60°F, respectively. The compressor isentropic efficiency = 0.8,
and the mechanical efficiency of the compressor driver is 0.95.
2. A natural gas (specific gravity = 0.60) pipeline is 120 mi long and is constructed
of NPS 20, 0.500 in. wall thickness (MAOP = 1000 psig), and runs from Akers
to Coburn. At a flow rate of 250 MMSCFD, an intermediate compressor at

Bradley (milepost 65) is required. Calculate the total HP required. In order to

reduce the power consumption by 20% at the present flow rate, it is proposed to
loop the pipeline. Calculate the length of looping required. Use the Panhandle A
equation with 95% efficiency. The compressibility factor can be assumed constant
at 0.90. The gas flow temperature is 70°F, and the base pressure and base tempe-
rature are 14.7 psia and 60°F, respectively. The delivery pressure required at
Coburn is 750 psig. The discharge pressure at Akers is 1000 psig, and the suction
pressure is 850 psig. Use 80% isentropic efficiency and 95% mechanical effi-
ciency for the compressors.
3. In a gas distribution pipeline, similar to that shown in Figure 5.8, gas enters the
pipeline at A at a flow rate of 50 MMSCFD. At B and C, deliveries of 10 MMSCFD
and 20 MMSCFD are made. At D, an additional volume of gas at 15 MMSCFD
enters the pipeline. Calculate the pressures at the various pipe nodes A, B, C,
and D, considering a delivery pressure of 500 psig at E. If the incoming volume
at D is increased to 25 MMSCFD and all pressures are to remain the same, how
much of the pipe DE should be looped? The pipe lengths are as follows:
AB: NPS 12, 0.250 length = 18 mi

BC: NPS 10, 0.250 length = 24 mi
CD: NPS 8, 0.250 length = 16 mi
DE: NPS 12, 0.250 length = 20 mi
REFERENCES
1. Mohitpour, M., Golshan, H., and Murray, A., Pipeline Design and Construction,
2nd ed., ASME Press, New York, 2003.
2. Engineering Data Book, 10th ed., Gas Processors Suppliers Association, Tulsa, OK,
1994.
3. Pipeline Design for Hydrocarbon Gases and Liquids, American Society of Civil
Engineers, New York, 1975.


5 Pipe Loops Versus Compression: 2785 - C005.fm Page 177 Friday, April 1, 2005 3:30 PM

Uploaded by

Copyright:

Available Formats

5 Pipe Loops Versus Compression: 2785 - C005.fm Page 177 Friday, April 1, 2005 3:30 PM

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

5 Pipe Loops Versus Compression: 2785 - C005.fm Page 177 Friday, April 1, 2005 3:30 PM

Uploaded by

Copyright:

Available Formats

2785_C005.

fm Page 177 Friday, April 1, 2005 3:30 PM

Pipe Loops versus Compression

5.1 PURPOSE OF A PIPE LOOP

The purpose of a pipe loop that is installed in a segment of a pipeline is to essentially

Copyright 2005 by Taylor & Francis Group, LLC

178 GAS PIPELINE HYDRAULICS

Figure 5.1 Effect of pipe loop.

5.2 PURPOSE OF COMPRESSION

We have seen in Chapter 3 and Chapter 4 that installing intermediate compressor

without second compressor station

Flow Q = 100 MMSCFD

Dover NPS 16 pipeline Leeds

Figure 5.2 Adding a compressor station.

Copyright 2005 by Taylor & Francis Group, LLC

PIPE LOOPS VERSUS COMPRESSION 179

5.3 INCREASING PIPELINE CAPACITY

Figure 5.3 Multiple compressor stations.

Copyright 2005 by Taylor & Francis Group, LLC

180 GAS PIPELINE HYDRAULICS

Pha MOP = 1440 psig

Figure 5.4 Windsor to Cardiff pipeline.

However, there is a limit to the number of compressor stations that can be

1. Neglecting elevation effects, calculate the maximum throughput capability of this

Copyright 2005 by Taylor & Francis Group, LLC

PIPE LOOPS VERSUS COMPRESSION 181

The HP required is calculated from Equation 4.15:

 188.41 × 10 6   14.7   540 

The velocity at Cardiff is

 188.41 × 10 6   14.7   540 

The erosion velocity from Equation 2.31 is

0.85 × 10.73 × 540

Copyright 2005 by Taylor & Francis Group, LLC

182 GAS PIPELINE HYDRAULICS

Using General Flow Equation 2.4, we get

P2 = 1114.85 psia = 1100.15 psig

Therefore, the compression ratio at Avon is

For Windsor, assuming the compressibility factor at suction is 1.0,

 1.40   1 + 0.85   1   0.40 

Copyright 2005 by Taylor & Francis Group, LLC

PIPE LOOPS VERSUS COMPRESSION 183

∆Cost = 5063 × 2000 = $10.13 million

The velocity at Cardiff is

 238.41 × 10 6   14.7   540 

Next, consider the phase 2 flow rate of

Solving for L, we get

Copyright 2005 by Taylor & Francis Group, LLC

184 GAS PIPELINE HYDRAULICS

Using General Flow Equation 2.4, we get

Solving for the suction pressure at Hart, we get

P2 = 1201.24 psia = 1186.54 psig

Therefore, the compression ratio at Hart is

This is a satisfactory compression ratio.

Solving for the suction pressure at Avon, we get

P2 = 915.54 psia = 900.84 psig

Therefore, the compression ratio at Avon is

This is a satisfactory compression ratio.

The HP required at Windsor is, using Equation 4.15,

Copyright 2005 by Taylor & Francis Group, LLC