October 2016 • Oil and Gas Facilities 1

A New Method To Detect Partial Blockage in Gas Pipelines

Kegang Ling, University of North Dakota; Xingru Wu, University of Oklahoma; and Zheng Shen, Weatherford

plex pipeline network. Furthermore, existing studies assume only single partial blockage in the pipeline, which limits the applica-tion of available models because the detection will be misleading if there is more than one partial blockage in the pipeline. To fill this gap, we developed a model to differentiate the single-partial-blockage scenario from the multiple-partial-blockage scenario on the basis of multirate tests. The identification is critical because it guides partial-blockage detection in the right direction.

IntroductionNatural gas is a clean energy and has a less negative impact on the environment when compared with other major energy sources such as oil and coal. The production of natural gas has been increasing materially since 1995 to meet the ever-increasing demand of nat-ural gas to better protect the environment [US Energy Informa-tion Administration (EIA) 2015]. The consumption of natural gas is next to that of oil and coal, according to EIA (2015) interna-tional energy statistics. New gas pipelines are constructed, and the existing pipelines are expanded to transport more natural gas to gas plants or end users. A partial blockage, which is a common problem in gas-pipeline operation, is the result of chemical and/or physical deposition caused by the changes of composition, pressure, and/or temperature. It is commonly met in gas-pipeline operation. There-fore, the reliable and timely detection of a partial blockage along the pipeline is one of the critical topics in flow assurance.

The detection of a pipeline partial blockage falls into two cat-egories: physical inspection and mathematical model. Convention-ally, physical-inspection methods include acoustic reflectometry, gamma ray transmission scanning, radio-isotope-tracer injection, tomography measurement, radiographic detection, and pipeline- diameter measurement. Usually, the physical inspections are ac-curate, but with the expense of production shutdown and high cost and long downtime, which may be infeasible in a long pipeline or a complex pipeline network. Mathematical models use data such as flow rate, pressure, temperature, and pressure-wave reflection to locate the partial blockage. The models use mass conservation, momentum conservation, energy-balance equations, pressure-pulse decay, and phase shift to estimate partial-blockage size and loca-tion. Mathematical models have the advantage of quick evaluation at a lower cost and can monitor the pipeline continuously without interrupting pipeline operations. Mathematical models usually re-quire flow parameters, which are not always available; therefore, operators are more confident with physical inspection than math-ematical modeling.

Many methods have been used to detect partial blockage. Rogers (1995) described a remotely-operated-vehicle (ROV) -based in-spection method to locate a partial blockage in an offshore pipeline. The partial blockage was located by measuring the change in hoop strain in the pipe as the internal pressure was raised and lowered. Hasan et al. (1996) used pressure-transient analysis to locate a par-tial blockage in a well. The size and location of the partial blockage was determined by the early-time transient wellhead-pressure re-sponse during a drawdown or shut-in test, while late-time steady-flow wellhead-pressure drop was affected by the partial-blockage size only. Scott and Satterwhite (1998) evaluated partial blockage with a backpressure technique by assuming fully rough flow. Scott

Copyright © 2016 Society of Petroleum Engineers

This paper (SPE 174751) was accepted for presentation at the SPE Annual Technical Conference and Exhibition, Houston, 28–30 September 2015, and revised for publication. Original manuscript received for review 16 February 2016. Revised manuscript received for review 2 August 2016. Paper peer approved 8 August 2016.

SummaryBecause of its efficiency, cleanliness, and reliability, natural gas is an important sector in global energy consumption. It supplies nearly one-fourth of all energy used in the United States and is expected to increase 50% within the next 20 years. More gas-delivery infrastructure is being constructed to meet the transpor-tation requirement of the ever-increasing demand for natural gas, while at the same time, the existing gas infrastructure is aging. Ensuring natural-gas-infrastructure reliability is one of the critical needs for the energy sector. Operators prefer to capitalize on the transportation capacity of these old pipeline systems to reduce the cost for building new pipelines, but they run a high risk of encoun-tering partial blockage in the pipeline, which can cause operating pressure to exceed the safety specification. Therefore, the reliable and timely detection of a partial blockage in a gas pipeline is crit-ical to ensuring the reliability of the natural-gas infrastructure.

To design proper pigging tools, it is important to detect the lo-cation and size of partial blockages. Physical inspection and math-ematical-model simulation are used to identify partial blockage in gas pipelines. Generally, the physical method can result in an ac-curate detection of the location and size of the partial blockage, but at the expense of production shutdown and high cost/long time to run the physical detection, which is a very expensive measure in a long-distance gas pipeline. The mathematical simulation detects partial blockage through numerical modeling, which could provide a quick evaluation at a much lower cost, but with higher uncertain-ties. Our literature review indicates that a simple, practical, and re-liable method to detect partial blockage without a recorded inlet or outlet pressure is in great demand.

In this study, we develop a multirate test method to detect par-tial blockage in a gas pipeline. By conducting multirate tests, the location and size of the partial blockage can be evaluated. The new method can be applied under the conditions of no measured inlet or outlet pressure, which have not been investigated before. It is worth locating a partial blockage under these conditions because as oil and gas exploration and production move to harsh environ-ments, no pressure gauge being installed at the inlet or outlet of the pipeline can be a common circumstance in the fields. Even for on-shore fields or fields with easy access, pressure is not transferred to the central office in real time. In addition, the metering equipment and pressure gauges installed in the pipeline may not be working. Therefore, our method provides a practical, quick, and low-compu-tational-cost approach to estimate partial blockages corresponding to these conditions.

The partial blockages in a single pipeline and in parallel/looped pipelines were evaluated in this project by use of the proposed method. Considering that most of the complicated pipeline sys-tems under operation can be decomposed into basic units, such as single pipeline and parallel/looped pipelines, the proposed model can realistically and feasibly identify partial blockage in a com-

2 Oil and Gas Facilities • October 2016

and Yi (1999) used a flow-testing method to estimate the size of partial blockages. However, this method failed to locate the partial blockage. Kashou et al. (2004) indicated the detection of the partial blockage in the gas pipeline from the Genesis Green Canyon 205 platform to the downstream Ship Shoal 354 platform by a transient simulation model. No detailed information regarding this transient simulation model was available. Wang et al. (2005) located a partial blockage in a pipeline by use of dampings from the fluid-transient technique. The frictional and blockage dampings were identified, and the ratios of damping rate were used to locate and quantify the partial blockage. Ma et al. (2007) used guided torsional waves to detect and characterize the sludge and partial blockages inside pipe-lines. Benson and Robins (2007) illustrated the uses of nonintrusive online diagnostic techniques, including gamma ray transmission scanning, tomography measurements, and radio-isotope-tracer in-jections to determine partial-blockage profiles in pipelines. Chen et al. (2007) proposed use of the pressure-wave propagation tech-nique to detect partial blockage in deepwater pipelines through both numerical and experimental studies. Lee et al. (2008) proposed use of the fluid-transient technique for locating partial blockages in a pipeline by extracting the behavior of the system in the form of a frequency-response diagram. Jassim et al. (2008) located a gas-hy-drate blockage in pipelines by numerical simulation, using a com-putational-fluid-dynamics technique. Vidal et al. (2013) applied acoustic reflectometry to identify and measure the location and size of a partial blockage in a pipe of 4-in. internal diameter and 95-m length. They also performed finite-element analyses to reproduce the numerical experimental data successfully.

Although many detection methods are available to detect partial blockage, they can be used to detect single partial blockage only and do not provide a means to differentiate single partial blockage from multiple partial blockages. To the best of our knowledge, no mathematical model can detect partial blockage without knowing flow rate. In this study, we propose a multirate-test method to dif-ferentiate single partial blockage from multiple partial blockages and to detect single partial blockage in scenarios met in the field.

Gas Flow Through Partial Blockage in a PipelineGas flow through partial blockage in a pipeline is similar to the flow through a restriction, a throated pipe, or a choke such as a nozzle or an orifice. It can be evaluated by use of the choke/per-formance relationship. The flow regime can be classified into sub-sonic and sonic flows that are based on gas velocity. Sonic flow occurs when gas velocity through a restriction reaches the sonic velocity in the fluid under in-situ conditions. Under sonic-flow conditions, the upstream cannot sense the pressure wave propa-gated from downstream upward because the gas is traveling in the opposite direction with the same velocity. Subsonic flow occurs when gas velocity is lower than sonic velocity in the gas at in-situ conditions. Sonic flow takes place when the ratio of downstream pressure to upstream pressure is less than the critical-pressure ratio, which is defined as

p

p kc

k

kdown

up

=

+

−2

1

1, ............................................................(1)

where k = Cp/Cv is the specific-heat ratio of fluid; Cp is the fluid heat capacity at constant pressure; Cv is the fluid heat capacity at constant volume; pdown is the downstream pressure (psia); and pup is the upstream pressure (psia).

By use of oilfield units, single-phase gas-flow rate at sonic-flow condition can be calculated by

q C Apk

T kDg

k

k=

+

+−

8792

1

11

upup�

, .....................................(2)

where A is the cross-sectional area of the choke, in.2; CD is the choke-discharge coefficient; q is the gas-flow rate, Mscf/D; Tup is the upstream temperature, °R; and γg is the gas specific gravity.

When the ratio of downstream pressure to upstream pressure is greater than the critical-pressure ratio, the flow is subsonic flow, and gas-flow rate is calculated by

q C Apk

k T

p

p

p

pDg

k

=−( )

−1248

2

1

2

upup

down

up

down

� uup

+k

k

1

.

......................................................................(3)

Gas Flow Through a Pipeline Several equations are available to calculate gas capacity through pipelines. The most commonly used are the Weymouth (1912) equa-tion and the Panhandle A and B equations (GPSA 1994). If the pipe-line diameter is less than 16 in., gas flow in a nonhorizontal pipeline can be calculated by the Weymouth (1912) equation, which is

qET

p

p p D

f T zLsc

sc

s

g e

=−( )3 23

2 2 5. einlet outlet

�, ........................................(4)

where

LL

se

s

=−( )e 1 ........................................................................(5)

and

sz

T zg=

0 0375. � �, .....................................................................(6)

where D is the pipe diameter, in.; E is the efficiency factor, dimen-sionless; e = 2.718; f is the Moody friction factor; L is the pipeline length, miles; psc is the standard-condition pressure, psia; pinlet is the inlet pressure, psia; poutlet is the outlet pressure, psia; T is the average temperature, = (Tinlet + Toutlet)/2, °R; Tsc is the standard-condition temperature, °R; z is the average gas compressibility, = (zinlet + zoutlet)/2; and ∆z is the outlet elevation minus inlet elevation, ft (∆z is positive when outlet is higher than inlet), where the Moody friction factor can be calculated by the Jain (1976) correlation:

11 14 2

21 250 9f

eND= − +

. log.

Re.

, ............................................(7)

where eD is the relative roughness, which is the ratio of absolute roughness to pipe internal diameter:

eDD = � , ..................................................................................(8)

where ε is the pipe absolute roughness.If the pipeline diameter is equal to or greater than 16 in., gas

flow in the pipeline can be calculated by the Panhandle A equation or the Panhandle B equation (GPSA 1994), which are, respectively,

qED T

p

p

g

b

b

=

435 872 6182

0 4604

1 07881

..

.

.

�

inlet22 2 0 5394

−( )

p

T zLoutlet

.

.................................................................................................(9)

October 2016 • Oil and Gas Facilities 3

and

q EDT

p

p p

T zLb

b

=

−( )737 2 530

1 02 2 2.

.

inlet outlet

��g0 961

0 510

.

.

, ........................(10)

where pb is the base pressure in psia and Tb is the base tempera-ture in °R.

The Weymouth and Panhandle equations (Weymouth 1912; GPSA 1994) were developed for perfectly clean pipelines filled with gas. In actual pipelines, water, condensates, sometimes crude oil, and scales in the pipeline reduce flow capacity. The efficiency factor is included in Eqs. 4, 9, and 10 to tune the theoretical pipeline-flow equations to the actual flow capacity of the pipeline in field operation.

Detecting Partial Blockage for Different CasesFor a partial blockage in the pipeline illustrated in Fig. 1, multiple flow-rate tests are used to obtain the flow parameters needed to solve the governing equations to locate the partial blockage and evaluate the partial-blockage size (blockage diameter, Db, and partial-blockage section length, Lb) for different cases. To model and analyze the gas behavior in the pipeline, the following assumptions are made:

•  There is single gas-phase flow in the pipeline.•  Blockage is annular.•  Temperature profile along the pipeline is known.•  Partial blockage occurs in only one location.The following subsections are the applications of multirate tests

for different scenarios. Partial blockage can occur at several points in a pipeline. In this study, we focus on locating and quantifying a single partial blockage in the pipeline. We will discuss the approach to differentiate a single-partial-blockage scenario from a multiple-partial-blockage scenario in the following section. The detection of multiple partial blockages in the pipeline will be the subject of future study.

When multirate tests are conducted, gas rate at the pipeline inlet will be changed. This leads to the redistribution of the pressures along the pipeline. Therefore, it will take a certain time period for inlet and outlet pressures to stabilize. The time period (or stabiliza-tion time) starts from a shut-in condition or a change in flow rate to a stabilized flow rate. To develop models to locate partial block-ages and estimate partial-blockage diameter and section length, we introduce the following dimensionless variables:

•  Dimensionless stabilization time: This is the ratio of stabili-zation time with partial blockage at location x to stabilization time with partial blockage at the outlet of the pipeline:

tt

tDx

stabilized,stabilized, blockage at

stabiliz

=eed, blockage at outlet

, ............................................(11)

where tstabilized,D is the dimensionless stabilization time; tstabi-lized, blockage at x is the stabilization time with partial blockage at

location x; and tstabilized, blockage at outlet is the stabilization time with partial blockage at the outlet of the pipeline.

•  Dimensionless blockage location: This is the ratio of partial-blockage location (measured from inlet of pipeline to partial-blockage location) to total length of pipeline.

LL

LDblockage location,blockage location= , ............................................(12)

where Lblockage location,D is the dimensionless blockage location and Lblockage location is the partial-blockage location.

•  Dimensionless blockage diameter: This is the ratio of partial-blockage diameter to pipeline internal diameter.

DD

Db Db

, = , .........................................................................(13)

where Db,D is the dimensionless blockage diameter and Db is the partial-blockage diameter.

•  Dimensionless blockage length: This is the ratio of partial-blockage section length to pipeline length.

LL

Lb Db

, = , ..........................................................................(14)

where Lb,D is the dimensionless blockage length and Lb is the partial-blockage section length.

•  Dimensionless pressure drop: This is the ratio of pressure drop through the pipeline without blockage to pressure drop through the pipeline with partial blockage:

��

�p

p

pD = no blockage

blockage

, ..............................................................(15)

where ∆pD is the dimensionless pressure drop, ∆pblockage is the pressure drop through the pipeline with blockage, and ∆pno blockage is the pressure drop through the pipeline without blockage.

Case 1: Single Pipeline With Known Flow Rate and Inlet and Outlet Pressures. Two flow-rate tests are required to evaluate par-tial blockage in this case. Gas rate, inlet and outlet pressures, and temperatures are measured when conducting two flow-rate tests. The following steps show the procedure to detect and quantify par-tial blockage:

1. Run the first flow-rate test, and measure the flow rate, inlet and outlet pressures, and the time to reach stabilization.

2. For the first flow-rate test, calculate the pressure drop in the pipeline assuming no blockage in the pipeline by use of the Weymouth or Panhandle equations (Weymouth 1912; GPSA 1994). One should note that the pressure drop without

Pipeline q

p inlet

L

Blockage section

D

x = Lx = 0

Lb

poutlet

Db

Lblockage location

Fig. 1—Gas flows through a pipeline with an annular blockage.

4 Oil and Gas Facilities • October 2016

blockage is minimum compared with partial-blockage cases. Then, calculate the stabilization time of the first flow-rate test by assuming no blockage in the pipeline. The calcula-tion of stabilization time can be implemented numerically by combining material balance with pipe and choke flows.

3. Now, assuming a partial-blockage section length, calculate pressure drops and dimensionless pressure drops that corre-spond to different partial-blockage locations with different partial-blockage diameters. Also, calculate stabilization times and dimensionless stabilization times that correspond to different partial-blockage locations with different partial-blockage diameters.

4. Plot dimensionless stabilization-time/blockage-location/blockage-diameter type curves on the basis of the data ob-tained in Steps 1, 2, and 3, as shown in Fig. 2. In the plot, the x-axis is the dimensionless blockage location, the y-axis is the dimensionless blockage diameter, and the z-axis is the dimensionless stabilization time.

5. Calculate pressure drop, dimensionless pressure drop, and dimensionless stabilization time for the first flow-rate test.

6. Find and connect the intersection points between the di-mensionless stabilization time (calculated in Step 5) and the type-curve plane obtained in Step 4; Line AB in Fig. 2 is the connection of intersection points.

7. Project line A1B1 onto the x–y plane to obtain Line A′1B′1. 8. Run the second flow-rate test, and measure the flow rate,

inlet and outlet pressures, and the time to reach stabilization.9. For the second flow-rate test, repeat Steps 2 through 7 to ob-

tain lines A2B2 and A′2B′2, as shown in Fig. 3. 10. The intersection point C between lines A′1B′1 and A′2B′2 

gives the dimensionless blockage location (point E), and dimensionless blockage diameter (point D), as shown in Fig. 3. From this, the partial-blockage location and partial-blockage diameter can be calculated.

11. Calculate the pressure drops and dimensionless pressure drops that correspond to different partial-blockage section lengths (or different dimensionless blockage lengths) by use of the partial-blockage locations and partial-blockage diam-eters calculated in Step 10.

12. Plot the dimensionless pressure drops vs. dimensionless blockage length type curve as shown in Fig. 4. The inter-section point between the dimensionless pressure drop and the type curve provides the dimensionless blockage length, which is indicated by the two red arrows in Fig. 4. Then, the partial-blockage section length can be calculated.

13. If the calculated partial-blockage section length in Step 12 is different from the guess in Step 3, use the calculated value as the new guess and repeat Steps 3 through 12 until all calculated variables converge. Then, the converged partial-blockage location, partial-blockage diameter, and partial-blockage section length are the solutions.

One should note that it is very important to prepare the pressure-drop curves (or type curves) needed in these dimensionless-variable plots before blockage because the computational time to prepare type curves is intensive. Preparing type curves before a blockage occurs can significantly reduce the computation time to analyze the multirate tests and expedite the detection when a blockage occurs. In the detection procedure, one can use dimensionless blockage length to replace dimensionless blockage diameter. Then, in Fig. 4, dimensionless blockage length should be replaced by dimension-less blockage diameter. This approach is equivalent to the procedure discussed previously. One should know that dimensionless stabi-lization time is more sensitive to dimensionless blockage location than to dimensionless blockage diameter and length, while dimen-sionless pressure drop is more sensitive to dimensionless blockage diameter and length than to dimensionless blockage location. There-fore, these two parameters should be applied to evaluate blockage accurately. Fig. 4 indicates that dimensionless pressure drop is sen-sitive to change in dimensionless blockage length. This is important in partial-blockage detection. Usually, it is easier to locate a long section of partial blockage by use of other detection methods.

Case 2: Single Pipeline With Known Inlet and Outlet Pressures and Unknown Flow Rate. The detection procedure of Case 2 is similar to that of Case 1. Because flow rate is unknown, a flow rate is assumed to construct the type curves and estimate the partial-blockage location, diameter, and length by following the steps in Case 1. Then, the flow rate is back calculated by use of pressures

0.99 0.95 0.90 0.80 0.60 0.40 0.20 0.10 0.050.03

0.001

0.005

0.050

0.200

0.500

Dimensionless

Blockage Location

(measured fro

m the

inlet of p

ipeline)

0.900

0.0B′1A′1

A1

E

B1

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Plot of Dimensionless Stabilization Time/Blockage Location/Blockage Diameter

Dim

ensi

onle

ss S

tabi

lizat

ion

Tim

e

Dimensionless Blockage Diameter

Fig. 2—Plot of dimensionless stabilization time/blockage location/blockage diameter for the first flow-rate test.

October 2016 • Oil and Gas Facilities 5

and calculated partial-blockage location, diameter, and length. If the calculated flow rate is different from the assumed value, use the calculated flow rate as the newly assumed value, and repeat the procedure until the flow rate converges.

Case 3: Single Pipeline With Known Flow Rate and Inlet or Outlet Pressure and Unknown Outlet or Inlet Pressure. The principle is the same as that in Case 2. Because inlet or outlet pres-sure is unknown, a pressure is assumed to construct the type curves and detect partial blockage. Then, the pressure is calculated. If the calculated pressure is different from the assumed one, use the calcu-lated pressure as the newly assumed value and repeat the calculation.

Case 4: Parallel Pipelines Sharing Upstream and/or Down­stream Junction(s), With Known Flow Rate and Inlet and Outlet Pressures. Three flow-rate tests are run to evaluate partial block-age in this case. The procedure to evaluate partial blockage for each flow-rate test is similar to that of Case 1. The type curves of each pipeline are constructed and used. The combination of any two flow rates gives a partial-blockage location, a partial-blockage diameter, and a partial-blockage section length. Therefore, there are three sets of partial-blockage location, diameter, and length for each pipeline. The pipeline that gives the same partial-blockage location, diameter, and length under different flow rates is the one with partial blockage, while the pipelines that give different solutions are excluded.

Case 5: Parallel Pipelines Sharing Upstream and/or Down­stream Junction(s), With Known Inlet and Outlet Pressures and Unknown Flow Rate. The detection steps are similar to those in Case 4. Because the flow rates are unknown, an assumed flow rate is used and an iterative algorithm is required, which is the same as in Case 2.

Case 6: Parallel Pipelines Sharing Upstream and/or Down­stream Junction(s), With Known Flow Rate and Inlet or Outlet Pressure and Unknown Outlet or Inlet Pressure. The detection steps are similar to those in Case 4. The iteration to solve for un-known outlet or inlet pressure is the same as that used in Case 3.

Field ApplicationTo validate the proposed methods, they were applied to detect a blockage in an offshore gas pipeline. A 46.7-km-long gas pipeline with 10-in. diameter was constructed to transport gas from a sat-ellite field to a central platform. The daily gas-flow rate was 3 to 4 million m3/d under normal operating conditions. Inlet pressure varied from 12 to 14 MPa. After several platform shutdowns, it was found that the pressure drop through the gas pipeline had increased significantly when trying to restart the production operation. Anal-ysis of hydraulic data indicated that the pipeline had been partially blocked because of the formation of gas hydrates resulting from the multiple shutdowns. To restore normal production, it was nec-essary to locate and remove the blockage as quickly as possible. At

0.99 0.95 0.90 0.80 0.60 0.40 0.20 0.10 0.050.03

0.001

0.005

0.050

0.200

0.500

Dimensionless

Blockage Location

(measured fro

m the

inlet of p

ipeline)

0.900

0.0B′1A′1

A′2

A2

B′2E

C

D

B2

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Plot of Dimensionless Stabilization Time/Blockage Location/Blockage Diameter

Dim

ensi

onle

ss S

tabi

lizat

ion

Tim

e

Dimensionless Blockage Diameter

Fig. 3—Plot of dimensionless stabilization time/blockage location/blockage diameter for the second flow-rate test.

Dimensionless Pressure Drop vs. Dimensionless Blockage Length

Dimensionless Blockage Length

Dim

ensi

onle

ss P

ress

ure

Dro

p

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

∆pD

Lb,D0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Fig. 4—Estimation of blockage section length by use of the plot of dimensionless pressure drop vs. dimensionless blockage length.

6 Oil and Gas Facilities • October 2016

the same time, the remediation or removal of the partial blockage had to be implemented in a safe manner. This required a good esti-mation of blockage location, size, and length for the appropriate pigging-operation design. The proposed methods were used to detect the blockage location, and evaluate the blockage diameter and length. The calculation gave a 3-in. blockage diameter with 21.6-m blockage length occurring 5.154 km away from the pipe-line inlet. Considering the uncertainty in the mathematical model (the real field conditions may be different from the assumptions in the proposed model), the operator decided that a pigging opera-tion (or physical inspection) to obtain a detailed blockage profile in the pipeline should be conducted to avoid possible delay in case the pigs became stuck in the pipeline during the cleaning opera-tion. The physical inspection showed that the blockage occurred 5.325 km away from the pipeline inlet. Physical inspection dem-onstrated that the blockage length was 24.3 m. The blockage pro-file obtained by the pigging operation indicated that the blockage diameter varied from 2 to 5.3 in., with an average value of 3.5 in. On the basis of the blockage information provided by the pigging operation and the proposed methods, appropriate pig types and sizes were selected. Pigs were launched according to the remedia-tion procedure, and the blockage was removed. Chemicals were injected during the remediation to disassociate the hydrate and expedite the cleaning. The pipeline-transportation capacity was restored successfully after the cleaning.

The comparison of blockage location, length, and diameter be-tween the proposed methods and physical inspection indicated that the proposed methods are useful to narrow down the ranges of blockage location, length, and diameter before they are confirmed by physical inspection. It is worth analyzing the causes that lead to the differences between the model and physical inspection. It was believed that inaccurate measurements of temperature, pressure, and flow rate; change of pipeline geometry resulting from in-situ stress; corrosion and erosion; possible liquid condensed in pipeline; and inaccurate estimation of gas properties were the causes. The comparison shows that the proposed methods can provide a good reference when designing a pigging operation to inspect a pipeline.

Model Applications and LimitationsThe preceding discussion assumes single-phase gas flow in an ideal pipeline. It should be noted that bends, fittings, or junk will impact the pressure drop. The effects can be factored out by measuring the pressure drops under different flow rates before a blockage occurs. With these measurements as references, the influence of bends, fit-tings, or junk on pressure drop can be differentiated from that of a partial blockage. In gas-pipeline operations it is not uncommon that liquids condense out of gas. The liquids in the pipeline will reduce the pipeline efficiency (or require additional pressure drop). Simi-larly, the impact of liquids can be identified by conducting mul-tirate tests before a blockage occurs. One should note that liquid volume in the pipeline is a function of pressure, temperature, and fluid composition. Therefore, a pressure/volume/temperature (PVT) model is required to estimate the liquid and gas fractions at different conditions. Data obtained from multirate tests before a partial blockage occurs can be used to calibrate or tune the PVT model. The preceding analyses assume single partial blockage in a pipeline or parallel pipelines. The proposed method cannot be used to locate multiple partial-blockage segments and evaluate their diameters and lengths. Therefore, it is important to know the number of partial-blockage segments before applying the proposed model. Multiple flow-rate tests (at least three flow-rate tests) can be used to differentiate a single partial blockage from multiple partial blockages in a pipeline or parallel pipelines. The principle is similar to the method used in Case 4 to exclude pipelines without blockage. For a single pipeline, if different flow-rate tests give different partial-blockage phenomena, such as partial-blockage location, diameter, or length, then multiple partial-blockage segments exist. For parallel pipelines, if no pipeline has the same partial-blockage

location, diameter, and length under different flow-rate tests, then multiple partial blockages exist.

Conclusions and RecommendationsThe following conclusions can be drawn upon the completion of this study: •  Multiple flow-rate tests are performed to determine partial-

blockage location and evaluate the diameter and length of the partial blockage.

•  A dimensionless-variable approach has been presented to detect and estimate partial-blockage size.

•  The multirate test method can be used to differentiate a single-partial-blockage scenario from a multiple-partial-blockage sce-nario, which is critical in guiding the partial-blockage detection in the right direction.

•  The proposed method can be applied to parallel/looped pipelines. Therefore, it makes the detection of partial blockages in compli-cated pipeline systems possible.Considering that pipeline networks in operation can be very

complicated and two or more partial blockages can occur in the same pipeline or different pipelines within the systems, future work should expand the application of the proposed methods to more-complicated scenarios such as multiple partial blockages in the same pipeline and/or different pipelines in pipeline networks. Ex-perimental tests should be conducted to validate its application to multiple partial blockages in complex pipeline networks. Experi-ments should also be used to evaluate the uncertainty caused by liquid, bends, fittings, or junk in pipelines. A software that can pro-cess the acquired flow-rate, pressure, and temperature data in real time should be developed to detect partial-blockage location and evaluate the blockage severity in pipelines in the future. By incor-porating the software into the supervisory control and data acquisi-tion system installed in the pipeline systems, it would be possible to detect the partial blockage in real time and avoid significant loss through reducing operation downtime.

Nomenclature A = cross-sectional area of the choke, in.2 C = constant for unit conversion CD = choke-discharge coefficient Cp = fluid heat capacity at constant pressure Cv = fluid heat capacity at constant volume D = pipe diameter, in. Db = blockage diameter Db,D = dimensionless blockage diameter e = 2.718 eD = relative roughness, which is the ratio

of absolute roughness to pipe internal diameter

E = efficiency factor, dimensionless f = friction factor k = Cp/Cv, which is the specific heat ratio of

fluid L = pipeline length, miles Lb = blockage section length Lb,D = dimensionless blockage length Lblockage location,D = dimensionless blockage location Lblockage location = blockage location NRe = Reynolds number pb = base pressure, psia pdown = downstream pressure, psia pup = upstream pressure, psia psc = standard-condition pressure, psia pinlet = inlet pressure, psia poutlet = outlet pressure, psia q = gas-flow rate, Mscf/D T = average temperature, T = (Tinlet +

Toutlet)/2, °R

October 2016 • Oil and Gas Facilities 7

Tb = base temperature, °R Tsc = standard-condition temperature, °R Tup = upstream temperature, °R tstabilized,D = dimensionless stabilization time tstabilized, blockage at outlet = stabilization time with blockage at outlet

of pipeline tstabilized, blockage at x = stabilization time with blockage at

location x z = average gas compressibility, z = (zinlet

+ zoutlet)/2 γg = gas specific gravity ε = pipe absolute roughness ∆pblockage = pressure drop through pipeline with

blockage, psi ∆pD = dimensionless pressure drop ∆pno blockage = pressure drop through pipeline without

blockage, psi ∆z = outlet elevation minus inlet elevation

(∆z is positive when outlet is higher than inlet), ft

AcknowledgmentsThe authors are grateful to the Petroleum Engineering Department at the University of North Dakota. This research is supported in part by the North Dakota Experimental Program To Stimulate Com-petitive Research Program under award number EPS-0814442.

ReferencesBenson, D. and Robins, L. 2007. Nonintrusive Pipeline-Inspection Tech-

niques for Accurate Measurement of Hydrates and Waxes Within Operational Pipelines. Presented at Offshore Europe, Aberdeen, 4–7 September. SPE-108361-MS. http://dx.doi.org/10.2118/108361-MS.

Chen, X., Tsang, Y., Zhang, H.-Q. et al. 2007. Pressure-Wave Propagation Technique for Blockage Detection in Subsea Flowlines. Presented at the SPE Annual Technical Conference and Exhibition, Anaheim, California, USA, 11–14 November. SPE-110570-MS. http://dx.doi.org/10.2118/110570-MS.

Gas Processors Suppliers Association (GPSA). 1994. Field Engineering Data Book, Vol. 2, 10th edition. Tulsa: GPSA.

Hasan, A. R., Kouba, G. E., and Wang, X. 1996. Transient Analysis To Locate and Characterize Plugs in Gas Wells. Presented at the SPE Annual Technical Conference and Exhibition, Denver, 6–9 October. SPE-36553-MS. http://dx.doi.org/10.2118/36553-MS.

Jain, A. K. 1976. Accurate Explicit Equation for Friction Factor. Journal of Hydraulics Division 102 (5): 674–677.

Jassim, E., Abdi, M. A., and Muzychka, Y. 2008. A CFD-Based Model To Locate Flow-Restriction Induced Hydrate Deposition in Pipelines. Presented at the Offshore Technology Conference, Houston, 5–8 May. OTC-19190-MS. http://dx.doi.org/10.4043/19190-MS.

Kashou, S., Subramanian, S., Matthews, P. et al. 2004. GOM Export Gas Pipeline, Hydrate Plug Detection and Removal. Presented at the Off-shore Technology Conference, Houston, 3–6 May. OTC-16691-MS. http://dx.doi.org/10.4043/16691-MS.

Lee, P., Vítkovský, J., Lambert, M. et al. 2008. Discrete Blockage Detec-tion in Pipelines Using the Frequency Response Diagram: Numerical Study. Journal of Hydraulic Engineering 134 (5): 658–663. http://dx.doi.org/10.1061/(ASCE)0733-9429(2008)134:5(658).

Ma, J., Lowe, M. J. S., and Simonetti, F. 2007. Feasibility Study of Sludge and Blockage Detection Inside Pipes Using Guided Torsional Waves. Measurement Science and Technology 18 (8): 2629. http://dx.doi.org/10.1088/0957-0233/18/8/039.

Rogers, L. M. 1995. Pipeline Blockage Location by Strain Measurement Using an ROV. Presented at the Offshore Technology Conference, Houston, 1–4 May. OTC-7862-MS. http://dx.doi.org/10.4043/7862-MS.

Scott, S. L. and Satterwhite, L. A. 1998. Evaluation of the Backpressure Technique for Blockage Detection in Gas Flowlines. J. Energy Re-sour. Technol. 120 (1): 27–31. http://dx.doi.org/10.1115/1.2795005.

Scott, S. L. and Yi, J. 1999. Flow Testing Methods To Detect and Charac-terize Partial Blockages in Looped Subsea Flowlines. J. Energy Re-sour. Technol. 121 (3): 154–160. http://dx.doi.org/10.1115/1.2795975.

US Energy Information Administration (EIA). 2015. International Energy Statistics, http://www.eia.gov/cfapps/ipdbproject (accessed June 2015).

Vidal, J. L. A. E., Silva, L., Netto, T. et al. 2013. Acoustic Reflectometry for Blockages Detection in Pipeline. Presented at the Offshore Tech-nology Conference Brasil, Rio de Janeiro, 29–31 October. OTC-24294-MS. http://dx.doi.org/10.4043/24294-MS.

Wang, X., Lambert, M., and Simpson, A. 2005. Detection and Loca-tion of a Partial Blockage in a pipeline Using Damping of Fluid Transients. Journal of Water Resources Planning and Manage-ment 131 (3): 244–249. http://dx.doi.org/10.1061/(ASCE)0733-9496(2005)131:3(244).

Weymouth, T. R. 1912. Problems in Natural Gas Engineering. In The Journal of the American Society of Mechanical Engineers, Vol. 34, 725–771. New York City: American Society of Mechanical Engineers.

Kegang Ling is an assistant professor of petroleum engineering at the University of North Dakota. His research interests are in the area of pro-duction optimization. Ling holds a BS degree in geology from the China University of Petroleum; an MS degree in petroleum engineering from the University of Louisiana, Lafayette; and a PhD degree in petroleum engineering from Texas A&M University.

Xingru Wu is an associate professor at the University of Oklahoma. Pre-viously, he worked in for BP and CNOOC as a reservoir/petroleum engi-neer. Wu’s research interests include reservoir engineering, numerical modeling and simulation, enhanced oil recovery, and multiphase flow in pipes. He holds a PhD degree in petroleum engineering from the Uni-versity of Texas at Austin.

Zheng Shen is a petroleum engineer at Weatherford. His research in-terests are production optimization and wellbore modeling. Shen has published more than 18 research papers. He holds a PhD degree in petroleum engineering from Texas A&M University. Shen serves as a technical editor for multiple journals. He is a recipient of the 2013 Out-standing Technical Editor Award for the Journal of Unconventional Oil and Gas Resources and the 2014 SPE Outstanding Technical Editor Award for SPE Production & Operations.