Kiefner Final Report Repair Replace

PHMSA Final Report No.
15-019
Final Report
Repair/Replace Considerations
for Pre-Regulation Pipelines
J.F. Kiefner and M. Van Auker

March 11, 2015
0339-1401
Intentionally blank
Final Report No. 15-019
Final Report
On
REPAIR/REPLACE CONSIDERATIONS FOR PRE-REGULATION PIPELINES
DTPH5614H00006, REPAIR/REPLACE CONSIDERATIONS FOR

PRE-REGULATION PIPELINES
to
U.S. DEPARTMENT OF TRANSPORTATION
PIPELINE AND HAZARDOUS MATERIALS SAFETY ADMINISTRATION
March 11, 2015
by
Kiefner and Associates, Inc.

5115 Parkcenter Avenue, Suite 200
Dublin, OH 43017
0339-1401
DISCLAIMER
This document presents findings and/ or recommendations based on engineering services

performed by employees of Kiefner and Associates, Inc. The work addressed herein has been
performed according to the authors’ knowledge, information, and belief in accordance with
commonly accepted procedures consistent with applicable standards of practice, and is not a
guaranty or warranty, either expressed or implied.
The analysis and conclusions provided in this report are for the sole use and benefit of the
Client. No information or representations contained herein are for the use or benefit of any
party other than the party contracting with Kiefner. The scope of use of the information
presented herein is limited to the facts as presented and examined, as outlined within the body
of this document. No additional representations are made as to matters not specifically
addressed within this report. Any additional facts or circumstances in existence but not
described or considered within this report may change the analysis, outcomes and
representations made in this report.
E XECUTIVE SUMMARY
The objective of this project was to develop a standardized process for making repair/replace
decisions for pre-regulation pipelines to assure that they will be replaced if they deteriorate to
the point that their integrity and safety can no longer be maintained. Pre-regulation pipelines
are those that were installed prior to November 12, 1970 when Federal pipeline safety
regulations were first promulgated. The guidelines provide a standardized method for pipeline
operators to decide which of their pre-regulation pipelines can be maintained safely and which
of them should be replaced because of technical and/or integrity shortcomings that cannot be
satisfactorily mitigated at an acceptable cost. The guidelines are intended to apply to
hazardous liquid pipelines, natural gas transmission pipelines, and natural gas distribution
systems. Pipeline operators would need to apply repair/replace processes in a manner that
meets the operational and integrity standards of all applicable Federal and State pipeline safety
codes.
These guidelines focus strictly on technical and safety related factors that must be addressed to
safely maintain a pre-regulation pipeline, in service, independent of the cost or the
convenience. It is recognized that no realistic repair/replace decision scenario is independent of
considerations for feasibility, cost-benefit analysis, restrictions on cost recovery through rates,
and assessment of risk, which are matters not addressed in this report. However, it is expected
that pipeline operators will apply their own analyses to these latter aspects of a repair/replace
decision.
The guidelines presented herein are intended to assist a pipeline operator in deciding whether
or not to replace a given pre-regulation pipeline or distribution system. Replacement is not
necessary if it is feasible to continue to maintain the integrity of the pipeline, main, or service
through on-going integrity assessments and mitigation of findings. This report provides criteria
for assessing what is required to preserve the integrity of a pipeline to assure that it does not
constitute a substantial risk to public safety. The guidelines also show what should be done to
maintain the integrity of an existing main or service operated under Part 192, Subpart P, “Gas
Distribution Integrity Management”. Examples of hypothetical pipelines, mains, and services
are presented. These examples illustrate how an operator can consider pipeline attributes,
potential threats to pipeline integrity, and mitigative responses that would be required to make
a decision either to do what is necessary to maintain the integrity of the pipeline and continue
to operate it or to replace it. It is up to the operator to decide whether or not pipeline integrity
can or will be preserved via the necessary maintenance and repairs or whether, instead, the
pipeline, main, or service will be replaced.
The process for making repair/replace decisions in each case is based on a series of flow charts.
It will be seen that these flowcharts constitute prescriptive procedures for making repair/replace
decisions. These procedures are not meant to prevent a pipeline operator from conducting an
engineering critical assessment of a pipeline that would allow the decision to be made outside
of these prescriptive procedures.
The process of deciding whether to repair or replace depends on having the essential data on
the pipeline, main, or service. Certain data are essential if the pipeline or distribution system
component is to remain in service. The essential data include but are not necessarily limited to
the type and vintage of pipe; its mechanical properties; the type of coating on the pipe (unless
it is plastic pipe); the operating pressure of the pipeline as a function of time; the leak and
rupture history of the pipeline, main, or service; and the pressure test history. These guidelines
define which data elements are essential for deciding whether or not a given pipeline, main, or
service should remain in service.
Repair/replace decisions also depend on recognizing and addressing threats to pipeline

integrity, in particular, those threats that are of a time-dependent nature and could affect the
pipeline, main, or service anywhere along its length. Time-dependent threats that could affect
pipeline integrity globally are the drivers for repair/replace decisions. These include external
corrosion, internal corrosion, selective seam weld corrosion (SSWC), stress corrosion cracking
(SCC), pressure-cycle-induced fatigue (PCIF), and slow crack growth in the case of plastic pipe.
Replacement is not an appropriate means to deal with threats that tend to be random in nature
or that affect only specific areas such as mechanical damage and outside forces. Replacing a
pipeline, main, or service would be an inefficient reaction to such threats. Replacing a pipeline,
main, or service also is not a solution for threats such as equipment failure and incorrect
operation.
Replacement of a pipeline segment, a main, or a service is not necessary if the operator can
afford to mitigate the active time-dependent threats such that the risk of pipeline failures from
such threats is acceptably small. Replacement becomes necessary whenever one or more of
the “driver” integrity threats cannot be mitigated and the risk of failures is not acceptably small.
Replacement also may become the best choice where addressing the threats and/or assuring
safety is technically feasible but the cost of properly addressing the threat and/or assuring
safety is more than the cost of replacement.
The time-dependent nature of the driver threats implies that mitigative responses must either
be continually or periodically applied. This, in turn, implies that reassessment intervals must be
estimated for each active threat that must be addressed periodically. The adequacy of the re-
assessment interval is validated by the absence of frequent in-service failures from time-
dependent threats over the life of a pipeline or by experience that reflects a declining or steady-
state leak history for a distribution system main or service. Replacement is likely to become the
preferred option to a continuing re-assessment and repair program when the cost of the latter
exceeds the cost of replacement. But, in any case it will be up to the operator to decide. The
criteria presented herein are intended only to give the operator a sound technical basis for
defining a mitigative repair plan, one that would be expected to prevent in-service failures of a
pipeline from time-dependent threats or that avoids an increasing leak rate in a distribution
system main or service.
In some cases the appropriate time for replacement can be deferred if the rate of degradation
is slow enough. Lowering the operating pressure of a pipeline in stages may be one way to
extend the time before replacement is needed because the effect of lowering pressure is to
reduce the likelihood of a failure from time-dependent degradation. In the case of a main or
service, replacement can be deferred if there is a likelihood that the leakage rate will not
increase with time.
Replacement may also be deferrable if a repair or operational change mitigates or eliminates a

specific time-dependent threat even if prior in-service failures have occurred due to that threat.
Examples:
• The pipeline is recoated to prevent external corrosion or to replace a coating that shields
the pipe from cathodic protection to mitigate SCC.
• Cathodic protection is enhanced such that external corrosion is mitigated.
• The maximum operating pressure is reduced to a level that mitigates exposure to SCC or
PCIF.
• An operational change involving increasing flow velocity above the water entrainment
threshold or change of service to a less corrosive product to mitigate internal corrosion.
This document begins with a background section, a list of acronyms used throughout the
document, and a list of definitions applicable to repair/replace decision making. The guidelines
are then presented in three main segments:
1. Factors to Consider in Making Repair/Replace Decisions

2. Mitigative Responses to Avoid or Delay Replacement
3. Criteria for Making Repair/Replace Decisions
The third segment presents numerous examples of the process of making a repair/replace
decision for the three types of pipeline systems (hazardous liquid pipelines, natural gas
transmission pipelines, and natural gas distribution systems).
The user of this document must refer to the governing Federal and/or State pipeline safety
codes and referenced standards to properly implement this guidance document and to maintain
safety.
Intentionally blank
T ABLE OF C ONTENTS
BACKGROUND ....................................................................................................................... 1
ACKNOWLEDGEMENTS ............................................................................................................ 3
ACRONYMS .......................................................................................................................... 4
DEFINITIONS ....................................................................................................................... 6
General Terms Used in These Guidelines ........................................................................ 6
49 CFR 192.3 Definitions Applicable to These Guidelines .................................................. 7
49 CFR 195.2 Definitions Applicable to These Guidelines .................................................. 8
FACTORS TO CONSIDER IN MAKING REPAIR/REPLACE DECISIONS ...................................................... 8
Threats to Pipeline Integrity that Could Influence a Repair/Replace Decision ..................... 9
Hazardous Liquid Pipelines ...................................................................................... 10
Natural Gas Transmission Pipelines .......................................................................... 12
Pipe Body Hard Spots and Hard Heat-Affected Zones ................................................ 15
Natural Gas Distribution Systems ............................................................................. 16
Attributes of Pipelines that Matter in Making a Repair/Replace Decision ............................ 19
Hazardous Liquid Pipelines ...................................................................................... 19
Natural Gas Transmission Pipelines .......................................................................... 21
Natural Gas Distribution Systems ............................................................................. 25
Records that Matter in Making a Repair/Replace Decision ................................................ 31
Material Specifications ............................................................................................ 31
Design Records ...................................................................................................... 31
Construction Records .............................................................................................. 31
Welding Procedure and Welder Qualification ............................................................. 32
Girth Weld Inspection Records................................................................................. 32
Pressure Test Records ............................................................................................ 32
Operating Records.................................................................................................. 35
In-Line Inspection Records ...................................................................................... 36
Direct Assessment Records ..................................................................................... 36
Bell-Hole Examination Records ................................................................................ 36
Repair Records ....................................................................................................... 37
Leak Survey Records .............................................................................................. 37
Conditions that Matter in Repair/Replace Decisions ......................................................... 37
Piggability .............................................................................................................. 37
i
Pressure Testability ................................................................................................ 38
Operating Stress Level (applies to steel pipe only)..................................................... 38
MITIGATIVE RESPONSES TO AVOID OR DELAY REPLACEMENT ...........................................................40
Pressure Testing .......................................................................................................... 40
Why Pressure Tests Work ....................................................................................... 40
Limitations ............................................................................................................. 46
Trade-off between Test Pressure and Retest Interval ................................................ 49
Spike Testing ......................................................................................................... 64
Anticipating Test Failures ........................................................................................ 66
Test Medium .......................................................................................................... 66
In-Line Inspection........................................................................................................ 67
ILI for Corrosion-Caused Metal Loss ......................................................................... 68
ILI for Crack-Like Defects........................................................................................ 68
ILI for Dents .......................................................................................................... 71
ILI for Pipe Body Hard Spots ................................................................................... 71
Influence of Failure Stress Prediction Capabilities on Use of ILI-Crack Detection Tools.. 71
Calculating Re-Assessment Intervals for Integrity Assessment by ILI .......................... 72
Direct Assessment ....................................................................................................... 74
ECDA .................................................................................................................... 74
ICDA ..................................................................................................................... 78
SCCDA .................................................................................................................. 79
Mitigative Response for the HSC Threat to a Hard Heat-Affected Zone ........................ 79
Mitigative Responses for Gas Distribution Systems .......................................................... 80
A/A (Additional or accelerated) Actions (Risk Mitigation Measures) for Specific Threats
from Table 6.1 of ANSI/GPTC Guidance Appendix G-192-8 ........................................ 82
CRITERIA FOR MAKING REPAIR/REPLACE DECISIONS .....................................................................88
Flowcharts for Transmission Pipelines ............................................................................ 89
Pipeline or System Attributes, Operational Data, and History ........................................... 89
Base Case Example No. 1: Hazardous Liquid Pipeline Comprised of LF-ERW Pipe .............. 89
Base Case Example No. 2: A Natural Gas Pipeline Comprised of Flash-Welded Pipe
Operated at a Stress Level of 72% of SMYS ................................................................. 100
Base Case Example No. 3: A Natural Gas Pipeline Operated at Stress Levels Less than 30%
of SMYS .................................................................................................................... 114
Other Examples for Transmission Pipelines .................................................................. 122
ii
A Pipeline with an SSAW Seam .............................................................................. 123
A Pipeline Comprised of Furnace Lap-Welded Pipe, Constructed with Acetylene Girth
Welds .................................................................................................................. 127
A Pipeline Comprised of Pipe with a DSAW Seam .................................................... 133
A Pipeline Comprised of Pipe with an HF-ERW Seam ............................................... 139
A Pipeline Comprised of Seamless Pipe................................................................... 142
A Pipeline Susceptible to Internal Corrosion ............................................................ 145
A Pipeline Containing Legacy Repairs ..................................................................... 147
A Pipeline Susceptible to HSC in Pipe Body Hardspots.............................................. 150
A Pipeline Susceptible to HSC in the HAZ of its DC-ERW Seam ................................. 152
Flowcharts for Gas Distribution Systems ...................................................................... 154
System Attributes, Operational Data, and History ......................................................... 154
Base Case Example No. 4: A Steel Gas Distribution Main ............................................... 154
Base Case Example No. 5: A Plastic Gas Distribution System ......................................... 161
Base Case Example No. 6: A Cast Iron Gas Distribution Main ......................................... 165
APPENDIX A - PIPELINE MILEAGE INSTALLED BY AGE .................................................................. 171
APPENDIX B - DETAILS OF REPORTABLE INCIDENTS BY CAUSE ....................................................... 176
APPENDIX C - FLOWCHARTS .................................................................................................. 178
REFERENCES ..................................................................................................................... 206
L IST OF T ABLES
Table 1. Summary of Reportable Incidents by Category for the Period 2004 through 2013 ....... 9
Table 2. Miles of Mains by Material in 2011 .........................................................................17
Table 3. Numbers of Services by Material in 2011 ................................................................17
Table 4. Example of Attributes of a Hazardous Liquid Pipeline ...............................................19
Table 5. Example of Attributes of a Natural Gas Transmission Pipeline ...................................22
Table 6. Example of Attributes of Steel Main or Service in a Natural Gas Distribution System ...26
Table 7. Example of Attributes of Plastic Main or Service in a Natural Gas Distribution System .27
Table 8. Example of Attributes of Cast Iron Main or Service in a Natural Gas Distribution System
..................................................................................................................................30
Table 9. Differences in Defect Sizes that Cause Failures at Various Pressures for a 30-inch-OD,
0.375-inch-wall, X52 Material Depending on Whether the Material Fails in a Ductile Manner
or a Brittle Manner .......................................................................................................44
Table 10. Retest Intervals for a Pressure Test to 1,170 psig with a Test-Pressure-to-Operating-
Pressure Ratio of 1.25 ..................................................................................................50
iii
Table 11. Retest Intervals for a Pressure Test to 1,300 psig with a Test-Pressure-to-Operating-
Table 12. Retest Intervals for a Pressure Test to 813 psig with a Test-Pressure-to-Operating-
Table 13. Retest Intervals for a Pressure Test to 975 psig with a Test-Pressure-to-Operating-
Pressure Ratio of 1.5....................................................................................................53
Table 14. Effect of Test-Pressure-to-Operating-Pressure Ratio Based on the Original
Fessler/Rapp Model......................................................................................................59
Table A1. Miles of Hazardous Liquid Pipelines by Decade of Installation ............................... 172
Table A2. Miles of Natural Gas Transmission Pipelines by Decade of Installation ................... 173
Table A3. Miles of Natural Gas Distribution Mains by Decade of Installation.......................... 174
Table A4. Number of Natural Gas Distribution Services by Decade of Installation .................. 175
Table B1. Reportable Incidents by Cause 1994-2013 – Part 1 of 2 ....................................... 176
L IST OF F IGURES
Figure 1. Failure Pressure Levels of Defects in the Pipe Body of a 30-inch-OD, 0.375-inch-wall,
X52 Pipe Material that Exhibits a Full-Size-Equivalent Charpy Upper Shelf Energy of 40 ft-lb
(Modified Ln-Sec Model) ...............................................................................................41
Figure 2. Failure Pressure Levels of Defects in a 30-inch-OD, 0.375-inch-wall, X52 Pipe Seam
Material that Exhibits a Fracture Toughness of 25.8 ksi-root-inch (equivalent to 5 ft lb of
Charpy energy according to the Barsom and Rolfe correlation) (API 579, Level II) ............43
Figure 3. Test Intervals Based on Fessler/Rapp....................................................................56
Figure 4. Test Intervals Based on Modified Fessler/Rapp ......................................................57
Figure 5. Test Intervals Based on Modified Fessler/Rapp that Decrease with Time ..................59
Figure 6. The Non-Constant Nature of Crack Growth from PCIF ............................................61
Figure 7. Start of Process to Decide Whether to Repair or Replace Base Case Hazardous Liquid
Pipeline.......................................................................................................................90
Figure 8. Threats to Address for Base Case Hazardous Liquid Pipeline Comprised of LF-ERW
Pipe............................................................................................................................92
Figure 9A. Feasibility of Preventing in-Service Failures from PCIF or SSWC by Hydrostatic
Testing or ILI ..............................................................................................................93
Figure 9B. Feasibility of Preventing in-Service Failures from PCIF or SSWC by Lowering
Operating Pressure ......................................................................................................94
Figure 10. Feasibility of Preventing In-Service Failures from External Corrosion ......................96
Figure 11. Feasibility of Preventing In-Service Failures from SCC ...........................................98
Figure 12. Time to Failure at the MOP after a Test (HTP) to 110% of SMYS ......................... 101
Figure 13. Times for Retests to Control SCC based on the Fessler/Rapp Method ................... 102
iv
Figure 14. Start of Process to Decide Whether to Repair or Replace the Base Case Natural Gas
Pipeline that is Operated at a Stress Level of 72% of SMYS ........................................... 105
Figure 15. Threats to Address for Base Case Natural Gas Pipeline Comprised of Flash-welded
Pipe that is Operated at a Stress Level > 30% of SMYS ................................................ 106
Figure 16A. Feasibility of Preventing in-Service Failures from PCIF or SSWC by Hydrostatic
Testing or ILI ............................................................................................................ 108
Figure 16B. Feasibility of Preventing in-Service Failures from PCIF or SSWC by Lowering
Operating Pressure .................................................................................................... 109
Figure 17. Feasibility of Preventing In-Service Failures from External Corrosion .................... 111
Figure 18. Feasibility of Preventing In-Service Failures from SCC ......................................... 113
Figure 19. Start of Process to Decide Whether to Repair or Replace the Base Case Natural Gas
Pipeline that is Operated at a Stress Level < 30% of SMYS ........................................... 115
Figure 20. Threats to Address for Base Case Natural Gas Pipeline Comprised of Flash-welded
Pipe that is Operated at a Stress Level < 30% of SMYS ................................................ 116
Figure 21A. Feasibility of Preventing in-Service Failures from PCIF ...................................... 118
Figure 21B. Feasibility of Preventing in-Service Failures from SSWC..................................... 120
Figure 22. Feasibility of Preventing In-Service Failures from External Corrosion .................... 121
Figure 23. Consideration of the Miter Bends in the Low-Stress Gas Pipeline.......................... 122
Figure 24. Flowchart for Evaluating a Repair/Replace Decision for SSAW Pipe in a Hazardous
Liquid Pipeline ........................................................................................................... 125
Figure 25. Flowchart for Evaluating a Repair/Replace Decision of Furnace Lap-Welded Pipe in a
Hazardous Liquid Pipeline ........................................................................................... 130
Figure 26. A Process by Which the Operator Can Deal with Acetylene Girth Welds ................ 131
Figure 27. Flowchart for Evaluating a Repair/Replace Decision for DSAW Pipe in a Hazardous
Liquid Pipeline ........................................................................................................... 136
Figure 28. Flowchart for Evaluating a Repair/Replace Decision for HF-ERW Pipe in a Hazardous
Liquid Pipeline ........................................................................................................... 141
Figure 29. Flowchart for Evaluating a Repair/Replace Decision for Seamless Pipe in a Hazardous
Liquid Pipeline ........................................................................................................... 144
Figure 30. Preventing Leaks and Ruptures from Internal Corrosion ...................................... 146
Figure 31. Process for Responding to Legacy Repairs ......................................................... 149
Figure 32. Process for Responding to an HSC Threat in a Pipe Body Hard Spot ..................... 151
Figure 33. Process for Responding to an HSC Threat in a Hard HAZ .................................... 153
Figure 34. Start of Process to Decide Whether to Repair or Replace the Base Case Steel Gas
Distribution Main ....................................................................................................... 155
Figure 35. Feasibility of Preventing in-Service Failures from SSWC (for steel gas distribution
main) ....................................................................................................................... 157
v
Figure 36. Feasibility of Preventing in-Service Failures from External Corrosion (for steel gas
distribution main) ...................................................................................................... 158
Figure 37. Feasibility of Preventing in-Service Failures from Internal Corrosion (for steel gas
distribution main) ...................................................................................................... 160
Figure 38. Start of Process to Decide Whether to Repair or Replace the Base Case involving a
Plastic Distribution System.......................................................................................... 162
Figure 39. Feasibility of Preventing in-Service Failures from Slow Crack Growth .................... 163
Figure 40. Feasibility of Preventing in-Service Failures from Joining or Tap Connection type
Failures in a Plastic Main ............................................................................................ 164
Figure 41. Start of Process to Decide Whether to Repair or Replace the Base Case involving a
Cast Iron Distribution Main ......................................................................................... 166
Figure 42. Feasibility of Preventing in-Service Failures from Graphitization ........................... 168
Figure 43. Feasibility of Preventing in-Service Failures from Breakage due to Soil Movement . 169
Figure A1. Trends in the Amounts of Pre-Regulation Pipe in Hazardous Liquid Pipelines (2005-
2012) ....................................................................................................................... 172
Figure A2. Trends in the Amounts of Pre-Regulation Pipe in Natural Gas Transmission Pipelines
(2005-2012) .............................................................................................................. 173
Figure A3. Trends in the Amounts of Pre-Regulation Natural Gas Distribution Mains (2005-2012)
................................................................................................................................ 174
Figure A4. Trends in the Numbers of Pre-Regulation Natural Gas Distribution Services (2005-
2012) ....................................................................................................................... 175
Figure C1. Flowchart START (for hazardous liquid pipelines, high-stress natural gas transmission
pipelines, and low-stress natural gas transmission pipelines) ......................................... 178
Figure C2. Flowchart LEGACY PIPE ................................................................................... 179
Figure C3A. Flowchart LF ERW (Part A) ............................................................................. 180
Figure C3B. Flowchart LF ERW (Part B) ............................................................................. 181
Figure C4. Flowchart SSAW .............................................................................................. 182
Figure C5. Flowchart LAP-WELDED ................................................................................... 183
Figure C6. Flowchart LEGACY FEATURES........................................................................... 184
Figure C7. Flowchart LEGACY REPAIRS ............................................................................. 185
Figure C8. Flowchart MODERN PIPE ................................................................................. 186
Figure C9. Flowchart HF ERW........................................................................................... 187
Figure C10. Flowchart DSAW............................................................................................ 188
Figure C11. Flowchart SEAMLESS ..................................................................................... 189
Figure C12. Flowchart EC ................................................................................................. 190
Figure C13. Flowchart IC ................................................................................................. 191
Figure C14. Flowchart SCC ............................................................................................... 192
vi
Figure C15. Flowchart HSC............................................................................................... 193
Figure C16. Flowchart START (for steel gas distribution system main) ................................. 194
Figure C17. Flowchart SSWC (for steel gas distribution system main) .................................. 195
Figure C18. Flowchart EC (for steel gas distribution system main) ....................................... 196
Figure C19. Flowchart IC (for steel gas distribution system main)........................................ 197
Figure C20. Flowchart LEGACY FEATURES (for steel gas distribution system main) ............... 198
Figure C21. Flowchart START (for plastic gas distribution system) ....................................... 199
Figure C22. Flowchart SCG (for plastic gas distribution system)........................................... 200
Figure C23. Flowchart PLASTIC CONNECTIONS (for plastic gas distribution system) ............. 201
Figure C24. Flowchart START (for cast iron gas distribution system main)............................ 202
Figure C26. Flowchart GRAPHITIZATION (for cast iron gas distribution system main) ........... 204
Figure C27. Flowchart BREAKAGE (for cast iron gas distribution system main)...................... 205
vii
Intentionally blank
FINAL
15-019
Repair/Replace Considerations
for Pre-Regulation Pipelines
BACKGROUND
Based on data submitted through natural gas transmission 2012 annual reports to the Pipeline
and Hazardous Materials Safety Administration (PHMSA), 176,363 miles of the 298,422 miles of
natural gas transmission and gathering pipelines in the U.S. (59%) are “pre-regulation”
pipelines, that is, pipelines installed prior to 1970, when Federal pipeline safety regulations went
into effect. Based on data submitted through hazardous liquid pipeline 2012 annual reports to
PHMSA, 97,316 miles of the 185,922 miles of hazardous liquid pipelines in the U.S. (52%) are
“pre-regulation” pipelines. As shown in Appendix A, trends of annual mileage show that these
numbers are gradually decreasing due to some older pipelines being taken out of service or
replaced. Older pipelines are not necessarily less safe than more modern pipelines. If their
integrity can be maintained by appropriate and timely inspections followed by remedial
responses to restore their serviceability, they can continue to operate without excessive risk to
the public. The primary reason for conducting this project was to develop guidelines that
pipeline operators can use to determine when replacement makes more sense than continuing
to do the necessary repairs to maintain serviceability of a pre-regulation pipeline
In response to concerns raised by PHMSA and the National Transportation Safety Board (NTSB)
with regard to the safety of these older pipelines, the Interstate Natural Gas Association of
America (INGAA) has developed a fitness-for-service (FFS) process that a pipeline operator may
use to assess the safety of pre-regulation pipelines, expecting that over time operators will
adopt the process in making repair/replace decisions regarding their pre-regulation pipelines.
To aid pipeline operators with making repair/replace decisions, this project was initiated to
create guidelines for implementing and executing a pre-regulation pipeline repair/replace
program. While INGAA’s approach to FFS for natural gas transmission pipelines was considered
in this work, the guidelines described in this document were tailored not only to natural gas
transmission pipelines but to the special concerns associated with natural gas distribution
systems and hazardous liquid pipelines as well. The resulting guidelines may be suitable for
inclusion in consensus pipeline safety standards including ASME B31.4 and ASME B31.8.
These guidelines were developed with the view that all pipelines of all types are “pre-
regulation” pipelines if installed prior to November 12, 1970. The rationale for using one date
Kiefner and Associates, Inc. 1 March 2015

FINAL
15-019
for all types of pipelines is that the materials and construction methods and most of the risks
associated with pipelines are similar irrespective of the type of service (liquid 1 or gas 2).
In general, operators of natural gas transmission pipelines and hazardous liquid pipelines
seldom consider replacing a pipeline because satisfactory tools are available to periodically
assess the integrity of their pipelines, namely, in-line inspection, hydrostatic testing, and direct
assessment. Operators who employ these techniques at timely intervals based on engineering
analysis and actual operating experience and respond in a timely manner to repair anomalies
discovered during the assessments have established remarkably good safety records.
Nevertheless, there may be cases where conducting the necessary periodic integrity
assessments needed to prevent failures is not feasible. For those cases the guidelines
developed on this project can be a useful decision-making tool.
In contrast to the transmission segments of the pipeline industry, the natural gas distribution
industry for more than 40 years has considered and often chosen pipe replacement as another
alternative to continued repair and mitigation activities. There are two reasons for this. First,
some distribution systems contain mains and services comprised of obsolete materials, mainly
cast iron pipe and uncoated steel pipe that are subject to severe degradation resulting in
numbers of leaks out of proportion to the amounts of such materials in the systems. Second,
the nature of distribution piping systems is such that the standard integrity assessment tools,
in-line inspection, hydrostatic testing, and direct assessment are generally not applicable.
Distribution operators with systems containing obsolete materials generally have often
established replacement schedules based on available prioritization models. These
circumstances have led the industry, through the American Gas Association, to assemble a
compendium of papers describing the repair/replace strategies used by various operators,
consultants, and research organizationsi. The facts that about 60,000 miles of cast iron mains
were in service in 1984 and that less than 34,000 miles of cast iron mains were still in service in
2011 show that the amount of replacement has been significant.
With the promulgation of 49 CFR 192, Subpart P, “Gas Distribution Integrity Management” in
2009, operators of gas distribution systems must establish a Distribution Integrity Management
Program (DIMP) that covers all parts of each distribution system. Prior to the DIMP regulation,
gas distribution operators had company specific policies and procedures to safely manage their
distribution systems. While many gas distribution operators already have repair/replace
programs in progress based on proprietary models, the DIMP regulations may change the way
1
Hazardous liquid pipelines are governed by 49 CFR Part 195, which was effective on April 1, 1970.
2
Natural gas pipelines are governed by 49 CFR Part 192, which was effective November 12, 1970.

FINAL
15-019
distribution operators execute repair/replace decisions. The DIMP regulations require a
distribution operator to:
1. Develop and implement a written integrity management plan.

2. Know its infrastructure.
3. Identify threats, both existing and of potential future importance.
4. Assess and prioritize risks.
5. Identify and implement appropriate measures to mitigate risks.
6. Measure performance, monitor results, and evaluate the effectiveness of its programs,
making changes where needed.
7. Periodically report a limited set of performance measures to its regulator.
Accordingly, the guidelines developed on this project must embody the DIMP approach to
identifying and mitigating integrity threats. Among other things, the current work is aimed not
just at cast iron and bare steel where replacement often makes sense. This project is intended
to address other material and construction related threats such as the phenomenon of slow
crack growth in polyethylene (PE) plastic pipe. In such cases, successful mitigation of threats is
likely to become the preferred option to replacement.
ACKNOWLEDGEMENTS
The authors are extremely grateful for the guidance and comments provided by a team of
advisors from the pipeline industry. The team was comprised of the following individuals.
David Johnson Energy Transfer

Phillip DePriest Marathon Pipe Line LLC
Ed Newton Southern California Gas Co.
Steve Koetting ExxonMobil Pipeline Co.
Benny Mummy Koch Pipeline
Bill Taylor Enable Midstream Partners
Bruce Paskett Northwestern Natural Gas
Gary Vervake Spectra Energy
Murali Nimmagadda Baltimore Gas and Electric Co.
The authors also greatly appreciate the assistance of Steve Nanney, the technical
representative from the Pipeline and Hazardous Materials Safety Administration (PHMSA).
Steve provided the focus for the numerous conference calls with the industry advisors as well as
reviews and approvals of documents we prepared.

FINAL
15-019
ACRONYMS
A/A Additional or accelerated (actions that a gas distribution operator may need to
take in order to comply with DIMP)
CFR Code of Federal Regulations
49 CFR 192 That part of the Code of Federal Regulations which applies to natural gas
transmission pipelines and natural gas distribution systems
49 CFR 195 That part of the Code of Federal Regulations which applies to hazardous liquid
pipelines
CMFL Circumferential magnetic flux leakage
CVN Charpy V-notch full-size-equivalent upper shelf energy, ft lb
DIMP Distribution Integrity Management Program
DC-ERW electric resistance welded pipe made with direct current
DSAW double submerged-arc welded pipe
ECDA external corrosion direct assessment
ERW electric resistance welded pipe
FERC Federal Energy Regulatory Commission
FFS Fitness for service
HCA high consequence area
HDB hydrostatic design basis in conjunction with polyethylene pipe
HF-ERW electric resistance welded pipe made with high frequency current (e.g., 450 KHz)
HSC hydrogen stress cracking
HVL highly volatile liquid
Hz hertz (a unit of frequency equivalent to one cycle per second)
ICDA internal corrosion direct assessment

FINAL
15-019
INGAA Interstate Natural Gas Association of America
ILI in-line inspection
KHz kilohertz (a unit of frequency equivalent to 1,000 cycles per second)
LF-ERW electric resistance welded pipe made with low frequency current (by definition in
API Specification 5L, a frequency less than or equal to 70 KHz. Commonly, 120
Hz was used in making older LF-ERW materials.)
MAOP maximum allowable operating pressure (usually associated with a natural gas
transmission pipeline or a natural gas distribution system)
MFL magnetic flux leakage
MOP maximum operating pressure (usually associated with a hazardous liquid

pipeline)
NAEC Narrow axial external corrosion
NDE Non-destructive examination
NTSB National Transportation Safety Board
PCIF pressure-cycle-induced fatigue
PE polyethylene
PHMSA Pipeline and Hazardous Materials Safety Administration
PIR Potential impact radius as defined in 49 CFR 192.903
SCADA supervisory control and data acquisition
SCC stress corrosion cracking
SCCDA stress corrosion cracking direct assessment
SMYS specified minimum yield strength
SSAW single submerged-arc welded pipe
SSWC selective seam weld corrosion

FINAL
15-019
D EFINITIONS
General Terms Used in These Guidelines
Grandfathered
“Grandfathered” is a term associated with natural gas transmission pipelines. It refers to a pipe
segment installed before July 1, 1970 with an MAOP established in accordance with 49 CFR
§ 192.619(c).
High Consequence Area (HCA)
A High Consequence Area (HCA) with respect to a natural gas transmission pipeline is defined in
49 CFR § 192.903. A High Consequence Area (HCA) with respect to a hazardous liquid pipeline
is defined in 49 CFR § 195.452.
Legacy Pipe
Legacy pipe means steel pipe manufactured using any now-obsolete technique including
furnace butt-welded, furnace lap-welded, low-frequency-welded electric resistance welded (LF-
ERW), direct-current-welded electric resistance welded (DC-ERW), single submerged-arc welded
(SSAW), or Flash Welded (A.O. Smith) pipe or any pipe with an unknown type of longitudinal
seam. Legacy pipe also includes wrought iron pipe, pipe made from Bessemer steel, pipe of
unknown specification or pipe for which the yield strength is unknown, and any pipe made prior
to 1942 when the minimum levels of API-required manufacturers’ pressure tests were raised
significantly. Lastly, legacy pipe includes polyethylene (PE) plastic pipe materials that have a
history of premature failures from slow crack growth.
Legacy Construction Techniques and Legacy Repair Methods
Legacy construction techniques refer to obsolete and potentially problematic construction

techniques such as wrinkle bends in pipelines operated at stress levels equal to or exceeding
30% of SMYS, miter bends having included angles greater than 3 degrees in pipelines operated
at stress levels equal to or exceeding 30% of SMYS, Dresser Couplings, threaded collars, non-
standard fittings, oxyacetylene welds in cross-country transmission pipelines, and bell and
spigot joints. Repair methods such as welded patches, half-soles and puddle welds constitute
legacy repair methods that are, for the most part, no longer used.
Modern Pipe
Modern pipe refers to all pipe that is not legacy pipe.

FINAL
15-019
Skelp
Iron or steel rolled or forged into narrow strips and ready to be made into pipe by being bent
and welded.
49 CFR 192.3 Definitions Applicable to These Guidelines

Distribution line means a pipeline other than a gathering or transmission line.
Gathering line means a pipeline that transports gas from a current production facility to a
transmission line or main.
High-pressure distribution system means a distribution system in which the gas pressure in the
main is higher than the pressure provided to the customer.
Low-pressure distribution system means a distribution system in which the gas pressure in the
main is substantially the same as the pressure provided to the customer.
Main means a distribution line that serves as a common source of supply for more than one
service line.
Maximum actual operating pressure means the maximum pressure that occurs during normal
operations over a period of 1 year.
Maximum allowable operating pressure (MAOP) means the maximum pressure at which a
pipeline or segment of a pipeline may be operated under this part.
Operator means a person who engages in the transportation of gas.
Service line means a distribution line that transports gas from a common source of supply to an
individual customer, to two adjacent or adjoining residential or small commercial customers, or
to multiple residential or small commercial customers served through a meter header or
manifold. A service line ends at the outlet of the customer meter or at the connection to a
customer's piping, whichever is further downstream, or at the connection to customer piping if
there is no meter.
Transmission line means a pipeline, other than a gathering line, that: (1) Transports gas from a
gathering line or storage facility to a distribution center, storage facility, or large volume
customer that is not down-stream from a distribution center; (2) operates at a hoop stress of
20 percent or more of SMYS; or (3) transports gas within a storage field.

FINAL
15-019
49 CFR 195.2 Definitions Applicable to These Guidelines
Hazardous liquid means petroleum, petroleum products, or anhydrous ammonia.
Highly volatile liquid or HVL means a hazardous liquid which will form a vapor cloud when
released to the atmosphere and which has a vapor pressure exceeding 276 kPa (40 psia) at
37.8 °C (100 °F).
Low-stress pipeline means a hazardous liquid pipeline that is operated in its entirety at a stress
level of 20 percent or less of the specified minimum yield strength of the line pipe.
Maximum operating pressure (MOP) means the maximum pressure at which a pipeline or
segment of a pipeline may be normally operated under this part.
Operator means a person who owns or operates pipeline facilities.
Petroleum means crude oil, condensate, natural gasoline, natural gas liquids, and liquefied
petroleum gas.
Petroleum product means flammable, toxic, or corrosive products obtained from distilling and
processing of crude oil, unfinished oils, natural gas liquids, blend stocks and other
miscellaneous hydrocarbon compounds.
Stress level means the level of tangential or hoop stress, usually expressed as a percentage of
specified minimum yield strength.
Supervisory Control and Data Acquisition (SCADA) system means a computer-based system or
systems used by a controller in a control room that collects and displays information about a
pipeline facility and may have the ability to send commands back to the pipeline facility.
Surge pressure means pressure produced by a change in velocity of the moving stream that
results from shutting down a pump station or pumping unit, closure of a valve, or any other
blockage of the moving stream.
FACTORS TO C ONSIDER IN M AKING R EPAIR/R EPLACE

D ECISIONS
The technical (non-economic) and integrity factors to consider when contemplating making a
repair/replace decision are discussed in the following paragraphs. It will be seen that these
factors overlap significantly those needed to manage pipeline integrity for hazardous liquid
pipelines and natural gas transmission pipelines in high consequence areas (HCAs). In fact,
replacement may become a rational choice if the continuing cost of managing integrity for an
older pipeline or system becomes too high in comparison to the cost of replacement. The
factors to consider include threats to pipeline integrity, the physical attributes of the pipeline or
system, the operating history (pressure and pressure cycles, leak history, pressure test history,

FINAL
15-019
in-line inspection (ILI) inspection history), and the circumstances (conditions) of a given
pipeline segment (piggability, testabililty, operating stress, existence of legacy pipe, existence of
legacy construction features). These groups of factors are discussed below.
Threats to Pipeline Integrity that Could Influence a

Repair/Replace Decision
Experience has shown that certain threats may lead to pipeline leaks or ruptures. PHMSA
requires pipeline operators to submit data on “reportable” incidents. The incidents are
categorized by cause, and the causes correspond to threats to pipeline integrity. The PHMSA
website http://primis.phmsa.dot.gov/comm/reports/safety/AllPSIDet_2004_2013_US.html
presents a summary of reportable incidents as well as the detailed databases of incidents. In
abbreviated form the PHMSA data for 2004 through 2013 are presented in Table 1.
Table 1. Summary of Reportable Incidents by Category for the Period 2004 through
2013
HAZARDOUS NATURAL GAS NATURAL GAS

CAUSE OR THREAT
LIQUID TRANSMISSION DISTRIBUTION
CORROSION 22% 20% 2%

EXCAVATION DAMAGE 5% 15% 30%
INCORRECT OPERATION 11% 3% 4%
MAT'L/WELD/FAILURE 11% 14% 5%
EQUIPMENT FAILURE 37% 15% 3%
NATURAL FORCE DAMAGE 5% 14% 8%
OTHER OUTSIDE FORCE DAMAGE 2% 9% 33%
ALL OTHER CAUSES 5% 10% 15%
Totals 100% 100% 100%
A more detailed breakdown is given in Appendix B. The main things these data show are that
the percentages of incidents by cause were different for hazardous liquid pipelines, natural gas
transmission pipelines, and natural gas distribution systems. Corrosion was a significant cause
of failure for hazardous liquid pipelines and natural gas transmission pipelines but not for
natural gas distribution pipelines. Equipment failure accounted for the largest percentage of
failures in hazardous liquid pipelines, a moderate percentage of failures for natural gas
transmission pipelines, but only a small percentage of failures in natural gas distribution
systems. The largest percentages of failures in natural gas distribution systems were caused by

FINAL
15-019
other outside force damage and excavation damage whereas only small to moderate
percentages of failures of hazardous liquid pipelines and natural gas transmission pipelines were
attributed to these causes 3. Stakeholders in pipeline safety (i.e., pipeline operators, regulators,
and the general public) expect that operators of these systems will pursue mitigative responses
to minimize or prevent the recurrence of these various types of incidents. As discussed below
pipeline operators are required by federal regulations to address the prevention of incidents by
developing and executing pipeline integrity management plans. It is noted, however, that
pipeline replacement is rarely the most effective response to reducing pipeline incidents.
Instead, a replacement decision is likely to be considered if and only if threat management is
not feasible by other means.
Pipeline operators are required by federal regulations to identify and mitigate pipeline integrity
threats via an integrity management plan. For gas distribution lines, these threats are managed
under the operator’s DIMP program. Integrity management plans embody periodic integrity
assessment activities. In most cases repairs made in response to these integrity assessments
will be an effective means to achieve reductions in incidents. In the development of guidelines
for repair/replace decisions, the essential goal is to be able to determine for a particular pipeline
segment whether or not all relevant threats can be successfully mitigated via periodic integrity
assessments via the “repair” option. If it is not technically feasible to successfully mitigate one
or more integrity threats via periodic integrity assessment, then it may be necessary to exercise
the “replace” option. Threats to pipeline or system integrity are described in industry
consensus standards and/or federal regulations.
Hazardous Liquid Pipelines

As outlined in API RP 1160 Managing Systems Integrity for Hazardous Liquid Pipelines, 2nd
Edition, 2013, Chapter 4, hazardous liquid pipelines are thought to be exposed to the following
integrity threats.
1) External corrosion
2) Internal corrosion
3) Selective seam weld corrosion (external or internal)
3
The relatively low rates of failure from excavation damage attributed to hazardous liquid pipelines and natural gas transmission
pipelines are believed to be the result of a focus over the last 20 years or so by pipeline operators, other utility operators,
government regulators, and excavators on preventing such damage. Near-universal adoption of “one-call” network systems
including the 811 phone number, enhanced communication between pipeline operators and excavators, and the development of
public awareness programs by pipeline operators no doubt have contributed to a significant reduction in excavation damage
accidents. Operators of gas distribution systems, as well, have adopted these improved practices. However, the fact that
distribution piping is located in areas of high exposure to damage tends to explain why these systems exhibited a higher rate of
incidents.

FINAL
15-019
4) Stress corrosion cracking (SCC)
5) Manufacturing defects (defective pipe seams including hard heat-affected zones and
defective pipe including pipe body hard spots4)
6) Construction and fabrication defects (including defective girth welds, defective

fabrication welds, wrinkle bends and buckles, and stripped threads/broken coupling
failures)
7) Equipment failure (including gasket or O-ring failure, control/relief equipment failure,

seal/pump packing failure, and miscellaneous)
8) Mechanical damage (causing an immediate failure or from vandalism)
9) Mechanical damage (previously damaged pipe or vandalism causing a delayed failure)
10) Incorrect operations
11) Weather and outside force (cold weather, lightning, heavy rains or floods, and earth
movement)
12) The growth of an initially non-injurious anomaly arising from any one of several of the
above causes into an injurious defect via pressure-cycle-induced fatigue (including
transit fatigue)
In terms of guidelines for repair/replace decisions, any systematic threat that affects an entire
segment such as bare pipe could make the segment a candidate for replacement. In addition
threats that could affect locations at random along buried segments of a pipeline could also
become a reason for replacement because of the efforts required to detect and repair the
affected locations and to mitigate future occurrences of incidents attributable to certain causes.
External corrosion, internal corrosion, SCC, selective seam weld corrosion, manufacturing
defects (which tend to become threats as the result of pressure-cycle-induced fatigue (PCIF)),
and construction and fabrication defects constitute threats that could lead to a decision to
replace a segment rather than to repair it. At the other extreme, threats arising from the
potential for equipment failure and incorrect operations can be mitigated by means of regular
4
Hard-heat affected zones and pipe body hard spots are confined to specific vintages of pipe made by specific manufacturers.
While these phenomena are not wide spread, the can cause failures and may constitute a reason for replacing a pipeline segment.
One segment of these guidelines is dedicated to discussing the nature of this threat and the means by which it may be mitigated is
discussed in another segment.

FINAL
15-019
maintenance and training programs and, therefore, would not be expected to lead to
replacement of a pipeline segment.
From the standpoint of immediate failure from mechanical damage, incidents seem to be
declining in numbers with time because of increased focus on damage prevention (e.g., better
one-call systems, mandated damage prevention programs, improved state one-call laws).
Therefore, it seems unlikely that a hazardous liquid pipeline segment would require replacement
to mitigate immediate failures from mechanical damage. Moreover, to the extent the threat
exists, the replacement segment could be just as exposed as the original segment although the
new pipe could be made more resistant to damage by increased wall thickness, deeper burial,
and/or installing a passive or active warning system.
It seems unlikely that replacement of a segment would be a practical means of mitigating the
threat of delayed failure from mechanical damage. Such damage is not likely to be as wide-
spread as damage from the threats of corrosion, SCC, and manufacturing and construction
defects. Besides, it is relatively easy to locate and repair such damage using an appropriate ILI
technology.
The threat of weather and outside force incidents falls into a special repair/replace decision
category. The areas along a pipeline most likely to be affected by heavy rains/flood are at river
crossings. In some cases it might be necessary to replace a river crossing to decrease its
exposure to damage from flooding. Similarly, areas of earth movement are generally easy to
identify such that a local replacement of pipe could be a rational mitigative response. There is
no expectation that replacement of a pipe segment would be a rational mitigative response to
the threat of cold weather or lightning.
Natural Gas Transmission Pipelines

As outlined in ASME B31.8S – 2012 Managing System Integrity of Gas Pipelines, Chapter 2,
natural gas transmission pipelines are thought to be exposed to the following integrity threats.
(a) Time-Dependent
1. External corrosion
2. Internal corrosion
3. Stress corrosion cracking (SCC)
(b) Stable
1. Manufacturing-related defects

FINAL
15-019
a. Defective pipe seam
b. Defective pipe
2. Welding/fabrication related
a. Defective pipe girth weld (circumferential) including branch and T joints
b. Defective fabrication weld
c. Stripped threads/broken pipe/coupling failure
3. Equipment
a. Gasket, O-ring failure
b. Control/relief equipment malfunction
c. Seal/pump packing failure
d. miscellaneous
(c) Time-Independent
1. Third party/mechanical damage
a. Damage inflicted by first, second, or third parties (immediate failure)
b. Previously damaged pipe (such as dents and/or gouges)(delayed failure)
c. vandalism
2. Incorrect operational procedure
3. Weather-related and outside force
a. Cold weather
b. Lightning
c. Heavy rains or floods
d. Earth movements

FINAL
15-019
The presumed threats to natural gas transmission pipelines are nearly identical to those that are
thought to affect hazardous liquid pipelines. The main difference is that manufacturing and
welding/fabrication defects in gas pipelines are considered stable defects whereas such defects
are considered to be susceptible to growth from pressure-cycle-induced fatigue in a hazardous
liquid pipeline. In most cases the assumption of stability of such defects in gas service is
justified because the pressure-cycle intensity in a typical gas pipeline is much less than that
associated with a typical hazardous liquid pipeline. It should be noted, however, that the
assumption of stability of manufacturing and welding/fabrication defects in gas pipelines may
not always be valid. Such defects are rendered stable, if and only if, a prior hydrostatic test or
pneumatic pressure test has been conducted to establish a satisfactory margin between test
pressure and operating pressure. Moreover, other factors such as changes in operations and/or
changes in the operating environment could have an adverse effect on stability. It is prudent
for a gas pipeline operator to at least evaluate the effects of pressure cycles for their pipelines
in the context of prior test pressures, possible changes in operating pressures, and changes in
the operating environment to be sure that the assumption of stability is justified.
For natural gas transmission pipelines, the threats of external corrosion, internal corrosion,
selective seam weld corrosion (SSWC), and SCC constitute threats that could lead to a decision
to replace a segment rather than to repair it. In addition, the existence of features such as
miter bends and wrinkle bends in pipelines that are operated at stress levels of 30% of SMYS or
more, non-standard fittings, Dresser couplings, and acetylene girth welds (legacy construction
techniques) need to be considered in the context of making repair/replace decisions.
As discussed previously in conjunction with threats to hazardous liquid pipelines, it is likely that
replacement will not constitute a reasonable response to the threat of immediate mechanical
damage or the threat of delayed failure from prior mechanical damage. Similarly, threats arising
from the potential for equipment failure and incorrect operations can be mitigated by means of
regular maintenance and training programs and, therefore, would not be expected to lead to
replacement of a pipeline segment.
The threat of weather and outside force incidents with respect to making repair/replace
decisions is the same for a gas transmission pipeline as discussed previously in conjunction with
hazardous liquid pipelines. In some cases it might be necessary to replace a river crossing to
decrease its exposure to damage from flooding. Similarly, replacement of pipe could be a
rational mitigative response in areas of earth movement. There is no expectation that
replacement of a pipe segment would be a rational mitigative response to the threat of cold
weather or lightning.
A final point relevant to repair/replace decision-making with respect to natural gas transmission
pipelines is that, unlike in the case of a hazardous liquid pipeline, small leaks in a gas pipeline

FINAL
15-019
can be repaired and are generally not hazardous to people, property, or the environment.
Thus, whereas a hazardous liquid pipeline that is prone to develop small leaks repeatedly from
a cause that is hard to control might be a candidate for replacement, replacement of a gas
pipeline that repeatedly developed small leaks might not be the most appropriate response.
Pipe Body Hard Spots and Hard Heat-Affected Zones

This discussion concerns a threat to the integrity of pipelines, both liquid and gas, that is unique
to a certain vintage of line pipe materials if those materials had been accidentally quenched
from a hot-working temperature to room temperature because of an upset in the manufacturing
process. The threat is hydrogen stress cracking (HSC). It is discussed separately from the
other threats, not because it is a common cause of failure (it is not), but because if the threat
exists in a given pipeline, unique methods must be used to control the phenomenon.
From the late 1940s until about 1960, high-yield-strength line pipe materials (API 5L Grades
X42 through X52) tended to rely heavily on carbon and manganese contents 5 higher than
typically used to produce the lower-yield-strength Grade A and Grade B line pipe materials
commonly used prior to the late 1940s. As a result, the ”X-grade” materials tended to be highly
hardenable if quenched rapidly to room temperature from the hot-rolling temperature range or
if allowed to cool too quickly after being welded. An X-Grade material of this vintage cooled too
quickly could end up having an untempered martensite microstructure, a very hard material (35
Rockwell C or more), that is susceptible to HSC if subjected to a high-enough stress level in the
presence of atomic hydrogen. The operating hoop stress levels in many pipelines constitute
sufficient stress, and the atomic hydrogen can be generated at areas of damaged coating by
cathodic protection current.
In the normal production of line pipe skelp in the period from 1947 through 1960, hot-rolled
plates or coils underwent controlled cooling to room temperature so that the microstructure
achieved (pearlite-ferrite) was not susceptible to HSC. During welding processes, the cooling
rate could also be controlled to prevent rapid quenching by means of preheating, by controlling
inter-pass temperatures, or by post-weld heat treatment. The latter, post-weld heat treatment,
was often employed after the making of line pipe with an ERW longitudinal seam to prevent
excessive hardness. By 1960 most line pipe producers had switched to micro-alloyed skelp
where elements such as vanadium or columbium were used in small quantities in place of
excessive amounts of carbon and manganese to achieve the high strength levels characteristic
5
Carbon contents of some (not all) heats could range as high as 0.29 to 0.34 percent by weight and manganese contents could
exceed 1.0 percent by weight. Materials that had both carbon and manganese contents within these ranges tended to be easily
hardenable.

FINAL
15-019
of X-Grade materials. With advent of the use of micro-alloyed materials, the potential for
encountering high hardness levels from a too-high cooling rate was greatly reduced.
The threat of HSC in pipelines (gas or liquid) made from these vintage materials can arise if
areas of the skelp were accidentally quenched at the end of the hot-rolling process or if an ERW
seam made in such a material was given inadequate post-weld heat treatment. One of the best
references on the HSC phenomenon is a paper by Groeneveld ii in 1974. This paper explains the
phenomenon and indicates that HSC had accounted for several pipeline failures. The paper also
discusses steps that pipeline operators may take to mitigate the threat of failures from HSC.
Some of these mitigating measures turned out to be practical, and they are discussed later in
these guidelines.
Natural Gas Distribution Systems

Repair/replacement decisions as well as integrity management 6 for natural gas distribution
systems involve consideration of threats that differ substantially from those associated with
repair/replacement decisions in a hazardous liquid pipeline or a natural gas transmission
system. The reason for the difference is largely due to the following circumstances: 1) axial
rupture (which is a common mode of failure in hazardous liquid and gas transmission pipelines)
is an unlikely mode of failure in a gas main or service because of the low operating stress levels
of distribution systems (below 20% SMYS), 2) over 50% of the mileage of distribution systems
nationwide is comprised of polyethylene pipe (a material that is not commonly used in a gas or
liquid transmission system), 3) cast iron pipe still comprises more than 2% of the gas mains in
the nationwide distribution network whereas none is found in a gas or liquid transmission
pipeline, 4) the methods often used to assess the integrity of hazardous liquid pipelines and
natural gas transmission pipelines (hydrostatic testing and ILI) are seldom applicable to natural
gas distribution systems (where leak surveying is a more common method.) The distribution
industry also uses trenchless pipe replacement methods such as slip-lining, pipe insertion, or
pipe bursting that allows replacement of pipe segments without requiring full exposure of the
pipe.
In 2011 the data from annual reports required by PHMSA iii showed the following distribution of
materials in the nationwide network of mains and services.
6
As noted previously, with the promulgation of 49 CFR 192, Subpart P, “Gas Distribution Integrity Management” in 2009, operators
of gas distribution systems must establish a Distribution Integrity Management Program (DIMP) that covers all parts of each
distribution system.

FINAL
15-019
Table 2. Miles of Mains by Material in 2011
Steel Main Mileage 552,407 44.80%

Plastic Main Mileage 645,252 52.33%
Cast Iron Main Mileage 33,624 2.73%
Ductile Iron Main Mileage 751 0.06%
Copper Main Mileage 30 0.00%
Other Main Mileage 893 0.07%
Total 1,232,955
Table 3. Numbers of Services by Material in 2011
Number of Steel Services 19,411,684 29.22%

Number of Plastic Services 44,273,373 66.64%
Number of Cast Iron Services 14,908 0.02%
Number of Ductile Iron Services 726 0.00%
Number of Copper Services 1,054,804 1.59%
Number of Other Services 1,677,778 2.53%
Total 66,433,273
According to the Code of Federal Regulation 49 CFR §192.1007(b) distribution system operators
must consider the following integrity threats.
1) Corrosion
2) Natural forces
3) Excavation damage
4) Other outside force damage
5) Material or welds
6) Equipment failure
7) Incorrect operations
8) Other concerns that could threaten the integrity of the pipeline
These threats correspond to the major categories of accidents shown in Table 1, but
determining the bases for repair/replace decisions requires consideration of more detailed
analysis of these threats. As one can see from the detailed incident data in Appendix B, neither
external nor internal corrosion has been a significant cause of incidents affecting natural gas

FINAL
15-019
distribution systems. There are several reasons for this. First, more than 50% of the natural
gas distribution piping is comprised of plastic pipe that is not subject to corrosion. Second,
internal corrosion is unlikely because most of the corrosive components of natural gas have
been removed before the gas is received by a distribution system. Third, about 86% of the
steel mains are coated and cathodically protected (based on the 2011 annual reports mentioned
previously). Nevertheless, there could be some steel mains and services, particularly those that
have no external coating and are not subject to cathodic protection, that should be considered
for replacement due to on-going corrosion that cannot be readily mitigated.
The threat of natural forces, (i.e., heavy rains/floods, earth movement, temperature, and
lightning) can be a significant factor in a repair/replace decision for a gas distribution system,
particularly a system containing cast iron mains and/or services. In some cases it might be
necessary to replace a segment of cast iron pipe if it is susceptible to damage from flooding.
Similarly, replacement of cast iron pipe could be a rational mitigative response in areas of earth
movement. The threat of cast iron, coupled pipe, or pipe joined by threaded connections being
damaged by frost heave or any situation resulting in soil subsidence exists particularly if the
pipe is not properly supported or protected. Therefore, the situation where a segment is
installed in soil that is prone to frost heave or other soil subsidence could influence a
repair/replace decision for a cast iron or threaded-joint segment. There is no expectation that
replacement of a pipe segment would be a rational mitigative response to the threat of
lightning.
There likely could be instances where repair/replace decisions will arise from the threat of
material/weld failures. The relatively low operating stress levels of steel pipe in natural gas
distribution systems (hoop stress levels of less than 20% of SMYS) eliminate the threats of SCC
and PCIF that can affect steel pipe in high-pressure transmission pipelines. However, in spite of
the relatively low operating stress levels, plastic, cast iron, and steel mains and services can be
susceptible to failure, and the threat of systematic failures could be a basis for deciding to
replace pipe instead of repairing it. Specifically, some older grades of plastic pipe may be
susceptible to the phenomenon of “slow crack growth” iv. Cast iron pipe is quite brittle and is
known to be susceptible to breakage from earth movement, washouts, and weakening from
“graphitization” v. Bare steel mains or services without cathodic protection could be candidates
for replacement. Steel pipe with an ERW seam that is susceptible to selective seam weld
corrosion could be subject to rupture in spite of a low operating stress level. Systems affected
by these phenomena could become candidates for a repair/replace decision.
Except for cast iron pipe, replacement of mains and services is generally not an appropriate
response to the threats of excavation damage and other outside force damage. Cast iron is
especially susceptible to any disturbances of nearby soil, so it is not unusual to consider
replacement of cast iron pipe in conjunction with activities such as repaving streets or the

FINAL
15-019
installation of a new underground facility, such as a sewer, nearby. In general, however,
replacement of pipe is unlikely to reduce the threat of future excavation damage.
The threat of equipment failure is best addressed by an effective maintenance program, and
the threat of failure from incorrect operations is best addressed by appropriate training of the
operator’s personnel and of contractors employed by the operator. So, these threats do not
appear appropriate for repair/replace decision-making for a natural gas distribution system.
Attributes of Pipelines that Matter in Making a

Repair/Replace Decision
Hazardous Liquid Pipelines
The 29 attributes which are useful for making repair/replace decisions for a hypothetical
example of a hazardous liquid pipeline are shown in Table 4. Suggested remedies are given in
Column 4 of the table for determining an attribute when that attribute is not initially known.
Table 4. Example of Attributes of a Hazardous Liquid Pipeline
Attribute Example Remedy if Unknown and

Attribute
Number Value Essential
Diameter, inches Measure at selected sites or globally
1 16
via a caliper tool.
Maximum Calculate using SMYS, diameter,
2 Operating 1170 wall thickness, design factor, and
Pressure, psig joint factor.
Wall Thickness, Measure globally using a UT wall
3 inch 0.25 thickness ILI tool or see 49 CFR
195.106(c)
Maximum This quantity is based on the design
4 Operating Stress, 72 factor (see below)
%SMYS
Transported This quantity is fixed by the
5 Crude Oil
Product operator's FERC certificate.
Material
6 APL 5L
Specification
7 Grade X52
SMYS, psi use 24,000 psi or see 49 CFR
8 52,000
195.106(b)
Manufacturing See "History of Line Pipe
9 Process (seam LF-ERW Manufacturing in North America" or
type) determine via metallography
10 Joint Factor 1 See 49 CFR 195.106(e)
Year Pipe
11 1953
Manufactured
12 Fracture 96 Values can be determined by

FINAL
15-019
Attribute
Toughness, ksi√in testing samples
Transition Values can be determined by
13 80
Temperature, ⁰F testing samples
14 Design Factor 0.72 See 49 CFR 195.106(a).
15 Year Installed 1953
Length of This quantity is chosen by the
16 900
Pipeline, miles designer.
Length of HCA This quantity is determined by the
17 Segments, miles 200 operator in compliance with 49 CFR
195.452
Miles of Legacy
18 900
Pipe
Electric-arc Girth Can be determined by pigging if
19 120,000
Welds piggable.
Acetylene Girth Can be determined by pigging if
20 0
Welds piggable.
Miter Bends Can be determined by pigging if
21 0
piggable.
Wrinkle Bends Can be determined by pigging if
22 0
piggable.
Mechanical Can be determined by pigging if
23 0
Couplings piggable.
Threaded Collars Can be determined by pigging if
24 0
piggable.
Bell and Spigot Can be determined by pigging if
25 0
Joints piggable.
Locations of Can be determined by pigging if
Valves, Fittings, Provided piggable.
26
and in a list
Appurtenances
Coating Type Coal tar Can be established through direct
enamel examination
27
with felt
wrap
Type of Cathodic Impressed Can be established through direct
28
Protection current examination
Leak Detection volume This leak detection system is
System balance chosen by the operator
29
plus aerial
patrol
The attributes listed in Table 4 are important for making repair/replace decisions because they
are needed to assess the ability of a pipeline to be operated safely with minimal risk of leaks or
ruptures. Methods for calculating failure stresses and remaining lives of defects depend on the

FINAL
15-019
pipe geometry (diameter and wall thickness), pipe material properties (yield strength, ultimate
tensile strength, fracture toughness, transition temperature 7), and maximum operating stress.
The type of longitudinal seam must be known to determine the amount of legacy pipe in the
system. The mileage of HCAs is helpful in assessing the consequences of failures. The types
and locations of circumferential joints should be known in order to assess the risk of failures
from external forces such as earth movement and heavy rains/floods. The locations of valves,
fittings, and appurtenances should be known to facilitate the use of ILI. The type of coating
and type of cathodic protection are useful in determining the risk of failures from corrosion and
SCC. Knowing the type of fluid in the pipeline and the type of leak detection system can be
helpful in assessing the consequences of leaks and ruptures. It is essential to know when the
pipe was manufactured in order to determine whether or not the pipe is legacy pipe simply
because of its age. Lastly, miter bends (misalignments at circumferential joints with more than
3 degrees of angle change) and wrinkle bends are not permitted in hazardous liquid pipelines
Natural Gas Transmission Pipelines

The 48 attributes which are useful for making repair/replace decisions for a hypothetical
example of a natural gas transmission pipeline are shown in Table 5. Because of the existence
of class locations in gas pipelines, the number of attributes is considerably higher than the
number associated with hazardous liquid pipelines. Suggested remedies are given in Column 4
of the table for determining an essential attribute when that attribute is not initially known.
As was the case for hazardous liquid pipelines, the attributes listed in Table 5 are important for
making repair/replace decisions because they are needed to assess the ability of a pipeline to
be operated safely with minimal risk of leaks or ruptures. Methods for calculating failure
stresses and remaining lives of defects depend on the pipe geometry (diameter and wall
thickness), pipe material properties (yield strength, ultimate tensile strength, fracture
toughness, transition temperature), and maximum operating stress. The type of longitudinal
seam, grade, and era of manufacture must be known to determine the amount of legacy pipe in
the system. The mileage of HCAs is helpful in assessing the consequences of failures. The
types and locations of circumferential joints should be known in order to assess the risk of
failures from external forces such as earth movement and heavy rains/floods. The locations of
7
Whereas §195.106(b) for liquid pipelines and §192.107 for gas pipelines give specific guidance for determining the yield strength
of a pipe material of unknown yield strength, no such guidance exists for determining toughness and transition temperature. If an
operator orders API 5L pipe to a certain toughness and/or transition temperature level, the number of such determinations would
be one per heat or one per 100 joints depending on pipe size. An operator would need to make a similar number of determinations
for an existing pipeline by cutting out and testing samples to ascertain a reliable value of base metal toughness and/or transition
temperature. Because of the extreme variability of toughness in the bond line regions of LF-ERW, DC-ERW, and flash-welded pipe
materials, it is doubtful that this number of determinations would be sufficient to ascertain the bond line toughness. For bond line
toughnesses in such materials it is reasonable to assume that a defect in the bond line (e.g., a cold weld) will fail in a brittle manner
and that the effective toughness will be less than or equal to a full-size-equivalent Charpy V-notch impact energy of 1 ft lb.

FINAL
15-019
valves, fittings, and appurtenances should be known to facilitate the use of ILI. The type of
coating and type of cathodic protection are useful in determining the risk of failures from
corrosion and SCC. It is essential to know when the pipe was manufactured in order to
determine whether or not the pipe is legacy pipe simply because of its age.
Table 5. Example of Attributes of a Natural Gas Transmission Pipeline

Attribute
Diameter, inches Measure at selected sites or globally
1 30
via a caliper tool.
MAOP, psig (Class Set by the requirements for
2 936
1 location) throughput
3 936
4 936
5 936
Wall Thickness, Measure using methods described
6 inch (Class 1 0.375 in 49 CFR 192.109
location)
7 inch (Class 2 0.45 in 49 CFR 192.109
location)
8 inch (Class 3 0.54 in 49 CFR 192.109
location)
9 inch (Class 4 0.675 in 49 CFR 192.109
location)
Operating Stress, factor (see below)
10 72
%SMYS (Class 1
location)
11 60
%SMYS (Class 2
location)
12 50
%SMYS (Class 3
location)
13 40
%SMYS (Class 4
location)

FINAL
15-019
Table 5 (continued). Example of Attributes of a Natural Gas Transmission Pipeline
Attribute
Transported Natural This quantity is fixed by the
14
Product Gas operator's FERC certificate.
Material
15 APL 5L
Specification
Fracture Values can be determined by
16 96
Toughness, ksi√in testing samples
Transition Values can be determined by
17 110
Temperature, ⁰F testing samples
18 Grade X52
SMYS, psi use 24,000 psi or see 49 CFR
19 52,000
192.107
Manufacturing See "History of Line Pipe
Flash-
20 Process (seam Manufacturing in North America" or
welded
type) determine via metallography
21 Joint Factor 1 See 49 CFR 192.113
Temperature See 49 CFR 192.115. This factor is
Derating Factor 1 if the segment is operated at a
22 1
temperature of 250⁰F (121⁰C) or
less.
Year Pipe
25 1960
Manufactured
Design Factor See 49 CFR 192.111.
26 0.72
(Class 1 location)
27 0.6
(Class 2 location)
28 0.5
(Class 3 location)
29 0.4
(Class 4 location)
30 Year Installed 1960
Length of This quantity is established by the
31 1000
Pipeline, miles designer.
Length of Class 1 The locations of each class location
Segments and the mileage is determined by
32 800
the operator in compliance with 49
CFR 192.5
Length of Class 2 The locations of each class location
Segments and the mileage is determined by
33 150
the operator in compliance with 49
CFR 192.6

FINAL
15-019
Table 5 (concluded). Example of Attributes of a Natural Gas Transmission Pipeline
Attribute Example
Attribute Remedy if Unknown and Essential
Number Value
Length of Class The locations of each class location and the
34 3 Segments 50 mileage is determined by the operator in
compliance with 49 CFR 192.7
Length of Class The locations of each class location and the
35 4 Segments 0 mileage is determined by the operator in
compliance with 49 CFR 192.8
Length of HCA This quantity is determined by the operator
36 200
Segments, miles in compliance with 49 CFR 192.903
Length of OCA This quantity is determined by the operator
37 200
Segments, miles
Miles of Legacy
38 0
Pipe
Electric-arc Girth Can be determined by pigging if piggable.
39 132,000
Welds
Acetylene Girth Can be determined by pigging if piggable.
40 Welds 0
41 Miter Bends 0 Can be determined by pigging if piggable.
42 Wrinkle Bends 0 Can be determined by pigging if piggable.
Mechanical Can be determined by pigging if piggable.
43 0
Couplings
44 Threaded Collars 0 Can be determined by pigging if piggable.
Bell and Spigot Can be determined by pigging if piggable.
45 0
Joints
Locations of Can be determined by pigging if piggable.
valves, fittings, Provided in
46
and a list
appurtenances
Coating Type Coal tar Can be established through direct
47 enamel with examination
felt wrap
Type of Cathodic Impressed Can be established through direct
48
Protection current examination
Note that miter bends (misalignments at circumferential joints with more than 3 degrees of
angle change) are not permitted in natural gas pipelines that are operated at hoop stress levels
of 30% of SMYS or more. Miter bends with angle changes not exceeding 12.5 degrees are
permitted if the pipeline is operated at a hoop stress level less than 30% of SMYS, miter bends
with angle changes not exceeding 90 degrees are permitted if the pipeline is operated at a hoop
stress level of 10% of SMYS or less. Wrinkle bends are permitted if the pipeline is operated at
a hoop stress level less than 30% of SMYS.

FINAL
15-019
Natural Gas Distribution Systems
Natural gas distribution systems are comprised of mains and services which are often parts of a
complex interconnected (sometimes described as “reticulated”) system designed to distribute
gas from one or more transmission pipelines to industrial, commercial, and residential
customers. Gas may flow in either direction in a main that is part of a complex network of
mains. Services branch out from taps on the mains to provide gas to end users. While any
new or replaced pipe installed in a distribution system must be subjected to a pre-service
pressure test, the nature of the interconnected system and the inability to interrupt customers
makes it very challenging to conduct subsequent integrity assessments of existing pipe via
pressure tests. Moreover, the layout of typical distribution systems and the relatively low
operating pressures of these systems make it very difficult to use on-stream ILI technologies for
integrity assessments. These aspects of distribution systems and the fact that they are
operated at hoop stress levels less than (in many cases, much less than) 20% of SMYS mean
that the risks associated with their operation are significantly different from the risks that are
associated with the operation of a hazardous liquid or a natural gas transmission pipeline.
Unlike hazardous liquid pipelines and natural gas transmission pipelines, many of which are
operated at relatively high pressure levels, failures from hoop stress resulting in axial crack
propagation in a gas distribution system are generally not possible. Instead incidents
associated with gas distribution systems that create hazards for people and property are
generally related to leakage of gas and accumulation of gas within a closed area leading to fires
and/or explosions. Accordingly, 49 CFR 192, Subpart P, “Gas Distribution Integrity
Management” was developed to address the special circumstances of gas distribution systems.
This regulation was finalized on December 4, 2009. Under this rule, operators of gas
distribution systems must establish a Distribution Integrity Management Program (DIMP) that
covers the entirety of each distribution system. Therefore, repair/replace guidelines for gas
distribution systems must take into account the special circumstances of such systems.
Attributes of Steel Mains and Services
The attributes of steel mains and services that need to be considered in a repair/replace
decision for a gas distribution system are summarized in Table 6. Leaks from steel mains or
services most commonly arise from external corrosion, excavation damage, mechanical
couplings, and damage from earth movement or heavy rains/floods. Therefore, pipe geometry,
material strength, operating pressure, locations of appurtenances, the existence of legacy pipe
or joining methods, and the coating and cathodic protection are the essential attributes.

FINAL
15-019
Table 6. Example of Attributes of Steel Main or Service in a Natural Gas Distribution
System
Attribute Example
Number Value
1 Diameter, inches 4.5 Measure at selected locations.
Wall Thickness, See 49 CFR 192.109
2 0.188
inch
Design Pressure, See 49 CFR 192.105
3 1805
psig
4 MAOP, psig 150 Set by the requirements for throughput.
5 Grade of pipe A
6 SMYS, psi 30,000 use 24,000 psi or see 49 CFR 192.107
Type of Seam Metallurgical examination of selected
7 LF-ERW
pieces
8 Design Factor 0.72 Based on Class 3 (49 CFR 192.111)
9 Joint Factor 1.0 See 49 CFR 192.113
Length of Main
10 10
or Service, miles
11 Year installed 1942
Feet of Legacy
12 52,800
Pipe
Electric-arc Girth
14 2,640
Welds
Acetylene Girth
15 0
Welds
16 Miter Bends 0
17 Wrinkle Bends 0
Mechanical
18 0
Couplings
Locations of
Provided in
19 Valves and
a list
Fittings
Locations of
Provided in
20 Service Taps if a
a list
Main
Coating Type Can be established through direct
21 Uncoated
examination
Type of Cathodic Can be established through direct
22 None
Protection examination
Attributes of Plastic Mains and Services
The attributes of plastic mains and services that need to be considered in a repair/replace
decision are summarized in Table 7.

FINAL
15-019
Table 7. Example of Attributes of Plastic Main or Service in a Natural Gas
Distribution System

Attribute
1 Diameter, inches 2.375 Measure at selected locations.
2 Wall Thickness, inch 0.216 Measure at selected locations.
Design Pressure, psig 80 See 49 CFR 192.121
3 MAOP, psig 60 See 49 CFR 192.619
4 Type of Pipe (PE, PVC, PE May be readable on the pipe 8
ABS,
CAB, PA-11)
5 Specification ASTM D2513 May be readable on the pipe6
6 Material Designation PE 2406 May be readable on the pipe6

Code for PE
7 Design Factor 0.32 See 49 CFR 192.121
8 Hydrostatic Design 1250 Approximately 200 times the last two
Basis (73°F), psi digits in the Material Designation
Code6 (eg. PE3408; HDB = 200 x 08
= 1600
9 Maximum Operating 100-140
Temperature, °F
10 Manufacturer Century May be readable on the pipe6
Utilities
Products, Inc.
11 Length of Main or 15
Service, miles
12 Date installed 1976 May be readable on the pipe6 Date of
manufacture may also be "coded" in
the lot number in the print line.
13 Method of Joining End- Butt heat- Can be determined by examining
to-End fusion for PE selected joints
14 Locations of Valves Provided in a
and Fittings list
15 Locations of Service Provided in a
Taps if a Main list
16 Type of connection Provide
and support at the description
main if a service
17 Mechanical Couplings 0
8
This is required in the print line at the time of installation. Print lines may not be present or legible upon excavation at later dates.

FINAL
15-019
Leaks from plastic mains or services most commonly arise from third-party excavation damage,
and damage from earth movement or heavy rains/floods. Generally, these threats would not be
cause for replacement of an entire main or service. However, some specific older PE plastic
pipe materials such as PE 2306 (particularly low-ductile inner wall Aldyl A piping) and PE3306
may be susceptible to the phenomenon of premature slow crack growth failures when subject
to stress concentrations or excessive internal pressure relative to the long-term failure pressure
of the material. Descriptions of the slow crack growth phenomenon, failures from slow crack
growth and the shortcomings of the early research on the long-term behavior of plastic pipe are
described in Reference IV (a special report by the National Transportation Safety Board (NTSB)
that was published in 1998). In addition, PHMSA has issued the following advisory bulletins on
the subject slow crack growth in plastic pipe:
• ADB-99-01, Potential failures due to cracking of plastic pipe manufactured by Century

Utility Products, Inc, September 1, 1999
• ADB-99-02, Potential failures due to cracking of plastic pipe in natural gas systems,
October 1, 1999
• ADB-02-07, Notification of the Susceptibility to Premature Brittle-Like Cracking of Older
Plastic Pipe; Notice; correction, December 3, 2002
• ADB-07-01, Updated Notification of the Susceptibility to Premature Brittle-like Cracking
of Older Plastic Pipe, September 6, 2007
These advisory bulletins may be downloaded from the following website:

http://phmsa.dot.gov/pipeline/regs/advisory-bulletin
Widespread occurrences of failures from slow crack growth could be a reason for considering
replacement of certain vintage(s) of PE plastic mains and services. From the standpoint of the
occurrence of slow crack growth, it is essential to know the type and grade of PE plastic
(Material Designation Code, e.g. PE2306), the pipe manufacturer, the date of manufacture, the
operating pressure, the maximum environmental operating temperature, the hydrostatic design
basis (HDB), the type of fusion joining method, and the locations of possible stress
concentrations (particularly regions where excessive deformation may be induced by inadequate
support at joints, appurtenances, couplings, rock impingement, excessive bending and
connections of services to mains).
Attributes of Cast Iron Mains and Services
Federal regulations do not permit the installation of new mains or services comprised of cast
iron because of inherent risks associated with the material. However, as of 2011, 33,624 miles
of cast iron mains and 14,908 cast iron services were still in existence. Cast iron pipe is
extremely brittle and has been known to fail circumferentially from stresses associated with
earth movement or loss of support from flooding. Moreover, the typical bell and spigot joints in

FINAL
15-019
cast iron systems have essentially no resistance to longitudinal forces, and they are prone to
leakage if they have not been sealed subsequent to original construction9. Also, gray cast iron,
the most common type, is susceptible to the phenomenon of “graphitization”, a form of
corrosion that removes iron, leaving only a weak matrix of carbon which is porous and easily
damaged, resulting in leakage. It is assumed that most distribution system operators are trying
to replace their cast iron mains and services on a time schedule that is based on risk
assessment and economics considerations. In the meantime continued leak surveys in
accordance with the Part 192 regulations (§ 192.723) and management in accordance with
DIMP regulations accompanied by necessary repairs to cast iron mains and services will be
necessary.
The attributes of cast iron mains and services that need to be considered in a repair/replace
decision are summarized in Table 8. The attributes are based largely on a historical document
produced by American Cast Iron Pipe Company vi. It is noted that this document discusses pipe
that was made for water service because the dimension tables refer to “AWWA” (American
Water Works Association) standards (probably AWWA Standard C101) and the working
pressures are given in feet of head. However, it is assumed that the cast iron pipe described in
the above-mentioned document was similar to or the same as the cast iron pipe that was used
in gas service. Operators must manage the safety and integrity of their cast iron systems in
accordance with the provisions of Part 192 and their respective DIMP Programs. In addition,
PHMSA has issued the following alert notices and an advisory bulletin on cast iron pipe:
• ALN-91-02, (Relates to recommended timely removal from service of cast iron pipe that
is constitutes a risk to public safety), October 11, 1991
• ALN-92-02, (Reminds operators that surveillance plans must include surveillance of cast
iron to identify problems and take appropriate action concerning graphitization), January
26, 1992
• ADB-12-05, (urges owners and operators to conduct a comprehensive review of their
cast iron distribution pipelines and replacement programs and accelerate pipeline repair,
rehabilitation and replacement of high-risk pipelines, requests state agencies to consider
enhancements to cast iron replacement plans and programs, and alerts owners and
operators of the pipeline safety requirements for the investigation of failures), March 23,
2012
These alert notices and the advisory bulletin may be downloaded from the following website:
http://phmsa.dot.gov/pipeline/regs/advisory-bulletin
9
Cast iron systems were typically installed at a time when most distribution systems transported only manufactured gas. The
moisture and oils inherent in the manufacturing of the gas tended to saturate the original joint packing, keeping the joints moist
and leak-tight. With the advent of dry natural gas, the packing was prone to drying out leading to leakage. All joints on most cast
iron systems had to be retrofitted with external or internal sealing devices.

FINAL
15-019
Table 8. Example of Attributes of Cast Iron Main or Service in a Natural Gas
Distribution System
Attribute Example
Number Value
1 Type of Material Class A Assume as a worst case
Nominal Measure at selected locations
2 4
Diameter, inches
Barrel OD, Measure at selected locations
3 4.80
inches
Nominal Wall Measure at selected locations
4 0.48
Thickness, inch
Design pressure, See AMERICAN Pipe Manual, Section 12
5 624
psig
6 MAOP, psig 25 Set by the requirements for throughput.
Length of Can be determined by examining selected
7 12
Pieces, feet joints
Length of Main
8 9.5
or Service, miles
Support On blocks Can be established by excavation
9 condition with tamped
backfill
10 Year installed 1925
Method of Can be determined by examining selected
Bell and
11 Joining End-to- joints
Spigot Joints
End
Type of Can be determined by examining selected
12 Subsequent External joints
Joint Sealing
Mechanical
13 0
Couplings
Bell and Spigot
14 4,180
Joints
Locations of
Provided in a
15 Valves and
list
Fittings
Number and
Locations of Provided in a
16
Service Taps if a list
Main
Corrosivity of Susceptibility Determine from operator records (e.g.,
17 soil for cast iron to exposed pipe reports, repairs records,
graphitization graphitization continuing surveillance reports, etc.)

FINAL
15-019
Records that Matter in Making a Repair/Replace Decision
As pipeline operators who have developed integrity management plans know, the records
associated with the design, construction, operation, and maintenance of a pipeline are useful, if
not essential, to the identification of threats and the assessment of risk. The same types of
records are useful if not essential for making repair/replace decisions.
Material Specifications
The pipeline operator must know the specification to which the pipe material was made (e.g.,
API 5L, ASTM A53, ASTM D2513, etc.). Such specifications certify the type of material (e.g.,
steel, wrought iron, cast iron, plastic), the manufacturing process (e.g., seamless, ERW,
DSAW), the diameter, the wall thickness, and the specified minimum yield strength (SMYS) of
the pipe (or hydrostatic design basis in the case of polyethylene pipe). The design formulas for
determining the wall thickness for various types of pipe materials embody the quantities:
diameter, design pressure, SMYS (or hydrostatic design basis in the case of polyethylene pipe),
and the joint factor for the type of longitudinal seam in steel or wrought iron pipe.
Design Records
Whether or not an operator has the original design records, the following parameters need to
be established for integrity management purposes and/or repair/replace decisions. The
essential records are: the design pressure, the location of the pipeline, the diameter(s), the wall
thickness(es), the grade(s) and type(s) of pipe (including the type(s) of seams), the types of
field joints, and the types and locations of appurtenances and other features (such as miter
bends, wrinkle bends, reducers, elbows, tees, flanges, scraper traps, valves, etc.). Alignment
sheets showing the locations of some or all of these attributes are useful. An elevation profile
of the pipeline is essential for planning hydrostatic tests, for mitigating possible internal
corrosion, and for establishing proper operating conditions in the case of a hazardous liquid
pipeline. It is helpful if the design records for a pipeline list the design code and edition that
governed its design. Unfortunately this was not often the case for legacy pipelines, but it is a
good idea for future pipelines. PHMSA issued an advisory bulletin (Verification of Records -
ADB-12-06, dated May 7, 2012, Docket No. PHMSA–2012–0068) to pipeline operators
concerning material and pressure test records needed to confirm pipeline MAOP/MOP.
Construction Records
Construction records should reveal the contractor involved, the dates of construction, and
descriptions of the progress of the work. However, these documents are not essential to
making a repair/replace decision, and no mitigative action is necessary if they cannot be located
or do not exist.

FINAL
15-019
What is useful and necessary for making a repair/replace decision is the actual as-built locations
of appurtenances and other features that can influence the success of integrity verification and
mitigative actions.
Welding Procedure and Welder Qualification

A welding procedure specification for girth welds and the welder and procedure qualification
records are indicative of the quality of work associated with constructing a pipeline. However,
these documents are not essential to making a repair/replace decision. No mitigative action is
necessary if they cannot be located or do not exist.
Girth Weld Inspection Records

Girth weld inspection records at a minimum should describe the inspection procedure and
present a list of the numbers of welds accepted and the number containing rejectable defects.
The records of welds that did not pass but were repaired and not replaced entirely should also
exist. These documents are useful but not essential for making a repair/replace decision. No
mitigative action is necessary if they cannot be located or do not exist.
Pressure Test Records

Pressure tests are an essential means of establishing the fitness of a pipeline for service 10. The
initial pre-service test, if one was conducted, establishes the initial integrity of a pipeline or
distribution main or service. Where time-dependent threats may have degraded a pipeline with
the passage of time, a subsequent hydrostatic test or pneumatic pressure test will tend to
indicate whether or not the degradation has been significant. Therefore, an operator who may
be contemplating repair versus replacement of a given pipeline segment should review the
results of all pressure tests that have ever been conducted on the segment. Such pressure test
records can be extremely useful in determining whether or not continuing mitigation via
hydrostatic testing is feasible as opposed to replacing the segment.
The following paragraphs list the current (2014) regulatory requirements for record keeping
with respect to pressure testing. It is recognized that keeping pressure test records was not
required prior to the advent of federal regulations (November 12, 1970). It is also noted that
the requirements for record keeping have evolved since the regulatory requirements were first
promulgated. Therefore, while the following listings are what is required currently, pipeline
operators should not be expected to have all of these items for all historic tests. By far the
most important records pertain to the pressure levels achieved in the most recent pressure test.
10
As previously noted, it is recognized that distribution operators will generally not be able to conduct pressure tests of their
infrastructure once the assets have been placed in service.

FINAL
15-019
Requirements for Test Records for a Hazardous Liquid Pipeline
Section 195.310 of the 49 CFR 195 requires that the following records of each pressure test to
be kept as long as the facility is in use.
• Pressure recording charts

• Test instrument calibration data
• The name of the pipeline operator
• The name of the person responsible for conducting the test
• The name of the test company used if not tested by the operator
• The time and date of the test
• The minimum test pressure
• The test medium
• A description of the facility tested and the test apparatus
• An explanation of any pressure discontinuities on the pressure recording charts,
including test failures, that appear on the pressure recording charts
• A profile of the pipeline showing the elevation and test sites over the entire length of the
test section where elevation differences in the section under test exceed 100 feet (30
meters)
• The temperature of the test medium during the test period (usually in the form of a
temperature chart)
The test pressure for a hazardous liquid pipeline must be at least 125% of the MOP of the
pipeline, and it must be held at or above that level for 4 hours. For buried segments where
direct visual inspection for leaks is not possible, the pressure must be held an additional 4 hours
at a level not less than 110% of the MOP. If hydrostatic testing is used periodically to validate
the integrity of an existing pipeline for integrity management purposes or to keep from having
to replace it, the operator may want to consider spike testing for a short time to a level
significantly higher than 125% of MOP before completing the required 4-hour test at 125% of
MOP.
Requirements for Test Records for a Natural Gas Transmission Pipeline or a Natural
Gas Distribution Main or Service
Section 192.517 of the 49 CFR 192 requires that the following records of each pressure test be
kept as long as the facility is in use unless the facility operates at a pressure level below 100

FINAL
15-019
psig, the segment is a service line, or the segment is comprised of plastic pipe. Records for the
latter three are to be kept for a minimum period of 5 years 11.
• The name of the pipeline operator

• The name of the person responsible for the test
• The name of the test company used if not tested by the operator
• The test medium used
• The test pressure
• The test duration
• Pressure recording charts, or other record of pressure readings
• Elevation variations, whenever significant for the particular test
• Leaks and failures and their disposition
Natural gas pipelines that are operated at stress levels of 30% of SMYS or more must be
strength tested for a period of at least 8 hours. With some exceptions12 the test must be a
hydrostatic test to a minimum pressure level of 125% of MAOP for Class 1 and Class 2 locations
and 150% of MAOP for Class 3 and Class 4 locations. If hydrostatic testing is to be used once
or periodically to validate the integrity of an existing pipeline for integrity management
purposes or to keep from having to replace it, the operator may want to consider spike testing
for a short time to a level significantly higher than 125% of MAOP before completing the
required 8-hour test.
Natural gas pipelines that are operated at stress levels of less than 30% of SMYS, steel
services, and plastic pipe can be tested with water or pneumatically with natural gas, air, or
inert gas. Except for the initial pre-service test of plastic pipe which is a minimum of 150% of
MAOP (49 CFR 192.619), these tests are primarily for the purpose of establishing the absence
of leakage and the MAOP. Leak-tightness could be a significant factor in a repair/replace
decision. Therefore, the records on any leak tests could be useful for repair/replace decisions.
Pressure Test Failure Records
Besides the above-described records that are required by federal regulations in conjunction with
a pressure test, the occurrences of test failures, the locations and elevations of test failures, the
11
A prudent operator should keep material mechanical/chemical properties test reports and pressure test reports to document the
pipeline MAOP for the life of the pipeline no matter what the minimum Code criteria say. These records are required for an integrity
management program to comply with Part 192 or Part 195
12
Class 1 locations may be tested with air or natural gas to a pressure level of 1.1 times MAOP except for locations within 300 feet
of a building intended for human occupancy. In the latter locations the pipeline may be tested with air of inert gas, but the test
pressure level must be at least 1.25 times MAOP for a distance of 300 feet upstream and downstream from the building. Air or
Inert gas pressure testing is permitted in other class locations, but the pressure limits are such that the resulting MAOP will be less
than the level one can validate via a hydrostatic test where the test pressure can be significantly higher.

FINAL
15-019
failure pressures, and the causes of the failures should be assessed in making a repair/replace
decision.
Operating Records
From the standpoint of making repair/replace decisions, an essential part of the operational
history of a pipeline is its pressure history with the passage of time. Ideally, for a hazardous
liquid pipeline, the pressure history should consist of a record of pressures measured at regular
intervals of not more than 15 seconds such that the pressure-cycle history can be examined to
assess the exposure to pressure-cycle-induced fatigue (PCIF). The interval for pressure
sampling for a natural gas transmission pipeline could be up to one hour (not an hourly
average, high and low values must be captured) because the rate of change with time is usually
much slower than is the case for a hazardous liquid pipeline. While a natural gas pipeline tends
to have a much lower risk of defect growth from fatigue than a hazardous liquid pipeline, the
pressure-cycle history of a natural gas pipeline still needs to be assessed 13.
Another essential part of the operating history of a pipeline is its safety performance in terms of
in-service leaks and ruptures that may have occurred over its service life. Examining the history
of leaks and ruptures versus time, especially where the causes are known, allows one to assess
the risks to the pipeline of failures from the various time-dependent threats, and it also allows
one to assess whether or not the risk of continuing failures of any one particular type can
become or is becoming excessive. While the term excessive is subjective, it refers to a situation
in which replacement becomes the preferred option to continuing repair and mitigative
responses.
For steel pipelines, the history of cathodic protection is useful when a repair/replace decision is
being considered. 49 CFR 195 requires operators of hazardous liquid pipelines and 49 CFR 192
requires operators of natural gas pipelines to maintain records of cathodically protected
pipelines, rectifiers, galvanic anodes and interference bonds and to conduct annual surveys of
pipe-to-soil potential readings. Occasionally, an operator may conduct a comprehensive pipe-
to-soil potential survey (a “close-interval survey”) by walking the right-of-way and taking
readings at intervals of generally between 3 feet and 6 feet. Operators are required to correct
inadequate levels of protection within a reasonable period of time, and the mitigation may
consist of increasing the current at existing rectifiers or adding new rectifiers or anodes, or
eliminating stray currents to enhance the pipe-to-soil potentials. Reviewing the history of these
types of efforts should be part of any repair/replace decision.
13
This is particularly true for natural gas pipelines that have never been subjected to a pressure test sufficiently high to assure the
absence of potentially large defects, for example, a pipeline with an MAOP of 40% of SMYS that has been tested only to a level of
50% of SMYS.

FINAL
15-019
Records of any monitoring for internal corrosion as well as the use of inhibitors, biocides, and
cleaning pigs should be considered in any repair/replace decision.
Records of earth movements and floods and the associated effects on a pipeline, if any, should
be examined as part of the repair/replace decision process.
In-Line Inspection Records

In-line inspection (ILI) records for any pipeline that has been inspected could bear significantly
on a repair/replace decision. ILI records from metal loss or crack-detection tools will tend to
indicate the types of degradation phenomena that may be affecting the segment, and they may
also help to pinpoint the most susceptible locations. Results of successive inspections may
indicate the rate of deterioration. All such records should be reviewed pursuant to making a
repair/replace decision.
Probability of Exceedance analysis of corrosion anomalies can be useful here. If the pipe
segment is badly deteriorated (for example it is uncoated or the coating exhibits widespread
failure), such an analysis may show that the probability of a leak or rupture in the remaining
uninvestigated anomalies cannot be significantly reduced no matter how many additional digs
are performed. Such a line would be a candidate for possible replacement.
Direct Assessment Records

For steel pipelines, records of any direct assessments (ECDA, ICDA, or SCCDA) could be useful
for making a repair/replace decision. Such data will tend to indicate the types of degradation
phenomena that may be affecting the segment, and they may also help to pinpoint the most
susceptible locations. Results of successive inspections may indicate the rate of deterioration.
All such records should be reviewed pursuant to making a repair/replace decision.
Bell-Hole Examination Records

For a variety of reasons, from time to time, pipeline operators excavate and examine portions of
their pipelines. Whenever such excavations are made, Federal Regulations require that
operators make inspections of the pipe for external corrosion (see, for example 49 CFR 195.569
for hazardous liquid pipelines and 49 CFR 192.459 for natural gas pipelines). Data obtained
from these “bell-hole” examinations can be helpful in making repair/replace decisions. Typical
data that one may obtain from a bell-hole examination are pipe condition, diameter

FINAL
15-019
measurements, wall thickness measurements, seam type, coating condition 14, soil resistivity,
and pipe-to-soil potentials. These data should be considered.
Repair Records
Pipeline operators are required to keep records of times and locations where repairs are made
to their pipelines. The repair records typically indicate where, when, why, and how the pipeline
was repaired. These data should be considered in making a repair/replace decision. Analysis of
repair rate trends over time or by segment or geographic region may be useful.
Leak Survey Records

Operators of hazardous liquid pipelines are required to inspect their rights-of-way at least 26
times per year (at intervals not exceeding 3 weeks) by flying, walking, driving, or other
appropriate means of traversing the right-of-way. While this is done primarily to spot
encroachments by digging equipment, it also affords an opportunity to spot leaks. Leakage
surveys for natural gas transmission pipelines must be conducted at least once each calendar
year (no interval to exceed 15 months.). If the gas is not odorized, the leakage surveys using
leak detector equipment must be done twice a year (no interval to exceed 7-1/2 months) in
Class 3 locations and four times a year (no interval to exceed 4-1/2 month) in Class 4 locations.
Leakage surveys with leak detector equipment are required at least once each calendar year
(intervals not to exceed 15 months) for gas distribution systems in business districts and at
least once every 5 calendar years (at intervals not exceeding 63 months) for distribution
systems outside of business districts. Leak surveys are conducted at a minimum of once every
3 years (no interval to exceed 39 months) for CI and unprotected bare steel distribution lines.
These data may be helpful in making a repair/replace decision.
Conditions that Matter in Repair/Replace Decisions

To make sound repair/replace decisions a pipeline operator should consider conditions and
circumstances that are likely to have an impact on the ability or inability to repair a pipeline or
mitigate integrity threats. These are as follows.
Piggability
For pipelines that are operated at hoop stress levels at or above 30% of SMYS, in-line
inspection is a key method for dealing with time-dependent threats to pipeline integrity such as
external corrosion, internal corrosion, SCC, and the growth of defects via pressure-cycle-
induced fatigue (PCIF). Therefore, the following questions are relevant to making a
repair/replace decision.
14
Areas of disbonded pipe coatings should be inspected for SCC.

FINAL
15-019
• Can the segment accommodate all types of smart pigs?
• Can the segment accommodate some types of smart pigs?
• If only some types of smart pigs can be accommodated, are those smart pigs capable of
locating and sizing the anomalies associated with each relevant time-dependent threat?
• If the segment is not currently piggable, can it be made piggable?
• If it cannot be made piggable, can the threats of external corrosion, internal corrosion,
and SCC be assessed via ECDA, ICDA, and/or SSCDA?
• Can other types of internal inspection (other than on-stream ILI tools) be used to assess
pipeline integrity from the standpoint of all relevant threats?
Pressure Testability
For pipelines that are operated at hoop stress levels at or above 30% of SMYS, pressure testing
is another key method for dealing with time-dependent threats to pipeline integrity such as
external corrosion, internal corrosion, SCC, and the growth of defects via pressure-cycle-
induced fatigue (PCIF). The questions that should be considered in making a repair/replace
decision are:
• Can the segment be taken out of service, filled with water, and subjected to a
hydrostatic test?
• If the segment cannot be tested with water, would a pneumatic pressure test with inert
gas or air to levels permitted by 49 CFR 192 or 49 CFR 195 be sufficient to demonstrate
pipeline integrity?
• As a last resort, can the segment be tested safely with natural gas or a liquid petroleum
product? (Part 192 limits pressure testing with natural gas to a stress level not
exceeding 80% of SMYS in Class 1 locations and to a stress level not exceeding 30% of
SMYS in all other class locations. Part 195 limits pressure testing with a petroleum
product to fluids that do not vaporize rapidly, and testing with such a medium may not
be done in populated areas.)
Operating Stress Level (applies to steel pipe only)

The maximum operating stress level of a pipeline should be considered in making a
repair/replace decision because phenomena such as SCC and PCIF generally are not observed
to occur in pipelines that operate at low stress levels and because at low stress levels,
longitudinal rupture is an unlikely mode of failure. One benchmark level is recognized by both
49 CFR 192 and 49 CFR 195, namely 20% of SMYS. Under 49 CFR 192, gas pipelines that are
operated at or above a stress level of 20% of SMYS must be classified as transmission pipelines.
Under 49 CFR 195, hazardous liquid pipelines that operate at or below a stress level of 20% of
SMYS are considered “low-stress” pipelines. Neither SCC failures nor PCIF failures nor

FINAL
15-019
propagating ductile ruptures are known to have occurred at stress levels below 20% of SMYS.
However, propagating brittle fractures have been observed at stress levels below 20% of SMYS.
Another benchmark is 30% of SMYS. Under 49 CFR 192, miter bends of less than 12.5 degrees
and wrinkle bends are permitted in natural gas pipelines that are operated at hoop stress levels
less than 30% of SMYS. In addition, the pressure testing requirements for natural gas pipelines
that operate at stress levels of 30% of SMYS or more differ from those for pipelines that
operate at less than 30% of SMYS.
Aside from these regulatory benchmarks, an operator, when considering a repair/replace

decision, should carry out an analysis of the possible effects of the maximum operating stress
on the likelihood of leaks and ruptures occurring taking into account the actual geometric
dimensions of the pipe; its material properties including strength, toughness, and ductile-to-
brittle fracture transition temperature; and the composition of the transported product with
respect to its ability to drive a propagating fracture.
Legacy Pipe
Legacy pipe means steel pipe manufactured using any now-obsolete technique including
furnace butt-welded, furnace lap-welded, low-frequency-welded electric resistance welded (LF-
ERW), direct-current-welded electric resistance welded (DC-ERW), single submerged-arc welded
(SSAW), or Flash Welded (A.O. Smith) pipe or any pipe with an unknown type of longitudinal
seam. Legacy pipe also includes wrought iron pipe, pipe made from Bessemer steel, pipe of
unknown specification or pipe for which the yield strength is unknown, and any pipe made prior
to 1942 when the minimum levels of API-required manufacturers’ pressure tests were raised
significantly. Lastly, legacy pipe includes polyethylene (PE) plastic pipe materials that have a
history of premature failures from slow crack growth.
Legacy Construction Techniques and Legacy Repair Methods
Legacy construction techniques refer to obsolete and potentially problematic construction

techniques such as wrinkle bends in pipelines operated at stress levels at or above 30% of
SMYS, miter bends having included angles greater than 3 degrees in pipelines operated at
stress levels at or above 30% of SMYS, Dresser Couplings, threaded collars, non-standard
fittings, oxyacetylene welds in cross-country transmission pipelines, and bell and spigot joints.
Repair methods such as welded patches, half-soles and puddle welds constitute legacy repair
methods that are, for the most part, no longer used.

FINAL
15-019
M ITIGATIVE R ESPONSES TO AVOID OR D ELAY R EPLACEMENT

In this segment of the report, the mitigative responses appropriate to repair/replace decisions
for hazardous liquid transmission pipelines, natural gas transmission pipelines, and natural gas
distribution systems are discussed. The mitigative responses of pressure testing, in-line-
inspection, and direct assessment as applied to hazardous liquid pipelines and natural gas
transmission pipelines are discussed first. These discussions are followed by a discussion of the
mitigative responses appropriate to repair/replace decisions in natural gas distribution systems.
It is assumed that if it is feasible for an operator to satisfactorily mitigate pipeline integrity
threats whenever and wherever such mitigation is required to preserve pipeline integrity,
replacing a pipeline can be delayed or avoided altogether.
Pressure Testing
One way to verify the safety and integrity of a pipeline segment is to pressure test it to a level
sufficiently above the maximum operating pressure to show that the segment contains no
defects that could fail at the operating pressure. Pressure testing can be carried out on any
pipeline segment that can be taken out of service for the time required to conduct the test.
The pressure test demonstrates the current serviceability of the segment, and the higher the
ratio of test pressure to operating pressure is, the more confidence the operator can have that
the segment is fit for service and free of injurious defects. Also, the higher the ratio of test
pressure to operating pressure is, the longer the time interval will be before retesting is needed.
The technical issues that illustrate the effectiveness of pressure testing are presented below. A
good summary of the practical aspects of hydrostatic pressure testing is available in a document
recently published by the INGAA Foundation vii.
Why Pressure Tests Work

It is intuitive that a structure can survive a given level of load without failing if that same
structure has previously survived a higher level of the same kind of load. The field of “fracture
mechanics” provides a more scientific explanation of why pressure testing works. Several
different fracture mechanics models exist for predicting failure stress levels of axially oriented,
part-through-wall defects in pressurized pipe. These include, but are not necessarily limited to
PAFFCviii, CorLas™ ix, API 579 – Level II x, API 579 – Level III, the Modified Ln-Sec Model xi,xii and
the Newman/Raju model xiii. Defects in most line pipe materials at normal ambient temperatures
tend to fail in a ductile manner, and in such cases the Modified Ln-Sec Model which evolved
from an empirical surface flaw equation developed by Maxey xiv,xv gives reasonable estimates of
the failure stress levels of longitudinally oriented defects. An important exception involves
defects in the bond line of a LF-ERW, DC-ERW, or flash-welded pipe material. For defects in
the bond lines of such materials, a model that can account for the brittle fracture and/or elastic-

FINAL
15-019
plastic failure behavior as well as plastic collapse behavior may be needed. Models such as
PAFFC, Corlas™ or API 579-Level II (or Level III) could be employed. The API 579-Level II
approach often uses the Newman/Raju model for assessing the brittle fracture behavior and
any appropriate plastic collapse model to assess the plastic collapse behavior. Together they
can produce solutions for failure pressures involving brittle fracture, intermediate elastic-plastic
fracture, or plastic collapse behavior. As such, API 579-Level II or Level III approaches may be
used to assess failure pressure levels not only for defects in the body of the pipe but for defects
in LF-ERW, DC-ERW, or flash-welded seams as well, provided that the fracture toughness levels
local to the regions of the defects are known. The following examples show how a pressure test
to a hoop stress level above the maximum operating stress level removes the threat of failure
of a defect at the maximum operating pressure. The first example involves defects in a
typically ductile pipe body material, and second involves defects in brittle bond line material.
The failure pressure levels for axial, part-through defects in the body of a 30-inch-OD, 0.375-
inch-wall, X52 line pipe material that exhibits a full-size-equivalent Charpy V-notch upper shelf
energy (CVN) of 40 ft-lb as predicted by the Elliptical C-equivalent Modified Ln-Sec Model are
illustrated in Figure 1.
D = 30.000", t = 0.375", SMYS = 52,000 psi, CVN = 40 ft-lb, CVN Area = 0.124
in^2
1,800
d/t
1,600
0.0
1,400 HTP 2 0.1

HTP 1 0.2
1,200
0.3
Pressure, psig
1,000 MOP
0.4
800 0.5
Rupture
600 0.6
0.7
400 Leak
0.8
200 0.9
0
0.0 5.0 10.0 15.0 20.0 25.0
Total Length, inch
Figure 1. Failure Pressure Levels of Defects in the Pipe Body of a 30-inch-OD, 0.375-
inch-wall, X52 Pipe Material that Exhibits a Full-Size-Equivalent Charpy Upper Shelf
Energy of 40 ft-lb (Modified Ln-Sec Model)

FINAL
15-019
In the model represented by Figure 1, as with any fracture mechanics model, failure stress
(failure pressure in this case) of a defect of a given size is a function of the geometry of the
structure (diameter and wall thickness in this case), the size of the defect (length and depth in
this case) and the inherent fracture toughness of the material (represented in this case by
Charpy impact energy). For this particular model, the failure stress is also a function of the
strength of the material (a function of yield strength in this case) because the model is based
on the assumption that the mode of failure will be ductile fracture initiation.
A leak/rupture dividing line appears in Figure 1, and it is based on an experimentally-validated

theoretic model of the behavior of through-wall defects developed by Folias xvi and later
modified by Maxey xvii. The sizes of defects that leak upon failing lie below and to the left of the
leak-rupture curve, whereas those that rupture upon failing lie above and to the right of the
leak-rupture curve.
For predicting failure stress levels of defects located in a brittle material such as a LF-ERW or
flash-welded seam where one can expect the Charpy energy to be less than 15 ft lb, it is
inappropriate to use the Modified Ln-Sec model. Instead, one could use a model such as
PAFFC, Corlas™, API 579-Level II, or API 579-Level III. An example prepared by using the API
579-Level II approach is presented in Figure 2. In this application the Newman/Raju model is
used to account for the linear elastic (i.e., brittle) fracture behavior, and Maxey’s surface flaw
equation (described in Reference XII) is used to account for the plastic collapse behavior. The
curves in Figure 2 are the result of numerous calculations using the API 579-Level II approach.
The intent of Figure 2 is to illustrate the fact that small defects can have a significant impact on
the failure pressure of a brittle pipe seam. This is demonstrated via Figure 2 below, where the
material is assumed to have a fracture toughness of 25.8 ksi-root-inch (equivalent to 5 ft lb of
Charpy energy (CVN) according to the Barsom and Rolfe correlation: 9.35*CVN0.63).

FINAL
15-019
1400
HTP 2, 1300 psig d/t
1200
HTP 1, 1170 psig 10%
1000
20%
Failure Pressure, psig
MOP, 936 psig
800
30%
600
40%
400
60%
200 80%
0
0 1 2 3 4 5 6
Total Length, Inches
Figure 2. Failure Pressure Levels of Defects in a 30-inch-OD, 0.375-inch-wall, X52
Pipe Seam Material that Exhibits a Fracture Toughness of 25.8 ksi-root-inch
(equivalent to 5 ft lb of Charpy energy according to the Barsom and Rolfe
correlation) (API 579, Level II)
As in Figure 1, Figure 2 gives predicted failure pressures for various combinations of defect
length and depth (actually, depth-to-wall-thickness ratios). The difference is that the failure
pressures given in Figure 2 reflect brittle or quasi-brittle behavior whereas those of Figure 1
represent ductile behavior 15.
Given a defect with a particular length and depth, one may determine its failure pressure using
Figure 1 if the material is expected to behave in a ductile manner (or from Figure 2 if brittle
behavior is expected). For example, the failure pressure of a defect that is 15 inches long and
80% of the way through the wall at its deepest point is 400 psig based on the point in Figure 1
where the curve of d/t = 0.8 intersects the vertical line at 15 inches total length. It is probably
15
Note that unlike the relatively evenly spaced d/t curves of Figure 1, the d/t curves of Figure 2 become more closely spaced as d/t
increases. This is the result of the material behavior trending from quasi-brittle (or elastic-plastic) to brittle over the range of
increasing defect depths for the particular range of defect lengths.

FINAL
15-019
not a good idea to try and extrapolate the d/t = 0.8 curve of Figure 2 to a defect length of 15
inches, but Figure 2 predicts that a 5-inch-long defect with d/t = 0.8 would fail at a pressure
level of just over 200 psig, much less than the 400 psig predicted for the ductile material where
the defect is actually 10 inches longer. Similarly, the failure pressure of a defect that is 5 inches
long and 30% of the way through the wall is 1,400 psig based on Figure 1 whereas its failure
pressure based on Figure 2 is just over 700 psig. The point is that defects large enough to just
barely survive a given pressure level in a ductile pipe body material or ductile pipe seam
material would be likely to fail at significantly lower pressures if the same-size defects resided in
a brittle LF-ERW or flash-welded seam.
To resume demonstrating how pressure testing works, consider two situations, one in which the
material fails in a ductile manner and one in which the material fails in a brittle manner. The
maximum operating pressure (MOP) is 936 psig (72% of SMYS) as represented by the
horizontal line labeled MOP in both Figure 1 and Figure 2. A pressure test to 1,170 psig (90%
of SMYS) is represented by the HTP 1 lines, and a pressure test to 1,300 psig (100% of SMYS)
is represented by the HTP 2 lines in both figures. The following table, Table 9, gives the depth-
to-thickness ratios taken from Figure 1 and Figure 2 that would cause failures at the three
pressure levels (936 psig, 1,170 psig, and 1,300 psig) for defects with lengths of 1, 2, 5, 10, 15,
and 20 inches for the 40-ft-lb ductile material and for lengths of 1, 2, and 5 inches for the 5-ft-
lb brittle material.
Table 9. Differences in Defect Sizes that Cause Failures at Various Pressures for a
30-inch-OD, 0.375-inch-wall, X52 Material Depending on Whether the Material Fails
in a Ductile Manner or a Brittle Manner
Defect d/t, %
Length, Ductile Behavior Using Figure 1 Brittle Behavior Using Figure 2
inches (CVN = 40 ft lb) (CVN = 5 ft lb)
MOP HTP 1 HTP 2 MOP HTP 1 HTP 2
936 psig 1,170 psig 1,300 psig 936 psig 1,170 psig 1,300 psig
1 97 95 92 23 11 5
2 92 85 76 22 10 4
5 71 55 42 21 10 4
10 54 36 24 - - -
15 45 27 18 - - -
20 40 23 15 - - -
First, consider the 40-ft-lb ductile material. Note that each depth-to-thickness ratio at a
particular defect length represents a lower bound value of depth-to-thickness ratio at a
particular pressure level. All defects with the same length but greater depth-to-thickness ratios
would fail at lower pressure levels. Hence, the test represented by HTP 1 (1,170 psig) causes

FINAL
15-019
failure of all 10-inch-long defects with depth-to-thickness ratios exceeding 36%, and the test
represented by HTP 2 (1,300 psig) causes failure of all 10-inch-long defects with depth-to-
thickness ratios exceeding 24%. This means that after the HTP 1 test, any 10-inch-long, 54%-
through or deeper defect that would cause failure at or below the MOP would have been
eliminated by the test. The HTP 2 test eliminates all 10-inch-long defects deeper than 24%
through the wall, so it is more effective than the HTP 1 test for assuring safe operation at the
MOP. Note that similar relationships exist for each of the lengths in Table 9.
Next consider the 5-ft-lb brittle material. As in the case of the 40-ft-lb material, each depth-to-
thickness ratio at a particular defect length represents a lower bound value of depth-to-
thickness ratio at a particular pressure level. All defects with the same length but greater
depth-to-thickness ratios would fail at lower pressure levels. Hence, the test represented by
HTP 1 (1,170 psig) causes failure of all 5-inch-long defects with depth-to-thickness ratios
exceeding 10%, and the test represented by HTP 2 (1,300 psig) causes failure of all 5-inch-long
defects with depth-to-thickness ratios exceeding 4%. This means that after the HTP 1 test, any
5-inch-long, 21%- through or deeper defect that would cause failure at or below the MOP
would have been eliminated by the test. The HTP 2 test eliminates all 5-inch-long defects
deeper than 4% through the wall, so it is more effective than the HTP 1 test for assuring safe
operation at the MOP.
On the basis of Figures 1 and 2 and Table 9, one can see why pressure testing works. It
eliminates or proves the absence of defects that would fail at pressure levels lower than the test
pressure. It should also be clear from Figures 1 and 2 and Table 9 that the higher the ratio of
test pressure to the maximum operating pressure is, the less likely it is that a failure will occur
at the maximum operating pressure. Note that the same is true whether the material exhibits
ductile (i.e., a typical pipe body material) or brittle behavior (i.e., typical of the bond line region
of a LF-ERW seam or a flash-welded seam) it’s just that the defect sizes causing failure at a
given stress level would be much smaller for the brittle material. But, a higher-test-pressure-to-
operating-pressure ratio (1,300 psig to 936 psig) is still more effective at assuring pipeline
integrity than a lower test-pressure-to-operating-pressure ratio (1,170 psig to 936 psig) even
for a brittle material.
The concepts embodied in Figure 1 and Figure 2 demonstrate that if defects can grow after the
test, the margin of safety will be eroded with the passage of time, and retesting will be required
to once again to revalidate the integrity of the pipeline. Note that the time to failure after the
test will be longer for a higher test-pressure-to-operating-pressure than for a lower test-
pressure-to-operating-pressure. That is because defects have to grow to a certain predictable
size in order for a failure to occur at the maximum operating pressure, and the starting sizes of
defects that survive a test decrease with increasing test pressure.

FINAL
15-019
Limitations
Some important limitations on pressure testing need to be considered when a test is
contemplated for integrity assessment of an in-service pipeline. First, the pipeline must be
taken out of service or at least closed to transporting the product. The time out of service
represents lost revenue and possible interruption of service to customers and shippers. In the
case of a single transmission pipeline supplying a gas distribution system or in the case of a gas
main, the disruption of customers usually cannot be tolerated, and hence, pressure testing is
seldom if ever conducted on existing pipe segments in such systems. Pressure testing in such
systems is usually only applied to new or replaced pipe segments prior to their being placed in
service. Even where testing is feasible, the time out of service can still have an impact on
revenue and shippers. Moreover, the time out of service is often hard to predict because of the
possibility of leaks and ruptures occurring during the test. In test scenarios where the product
must be displaced by water before the test and the water must then be displaced by the
product after the test, the time out of service can range of days to weeks. Longer times out of
service are common if leaks or ruptures occur. Each leak or rupture that occurs can be
expected to add at least one day to the time of testing, and the locating of a very small leak
may require several days of effort. To find small leaks during the test it may be necessary to
add a specialized leak detection gas to the water or test medium to help find very small leaks.
For the vast majority of cases where pressure testing is chosen for integrity assessment of high-
pressure hazardous liquid pipelines or high-pressure gas transmission pipelines, the test
medium will be water. Not just any water can necessarily be used. To avoid introducing
bacteria or corrosive compounds into the pipeline, potable water purchased from a utility is
often the preferred test medium. If untreated water must be used, it is prudent to run a
biocide disinfectant slug after the test and prior to the line being placed into service. The water
becomes contaminated by residual liquids and solids remaining after the product is displaced,
and the pipeline operator usually must pay for satisfactory treatment of the water before
discharging it to a sewer system, a river, or a reservoir.
It has been demonstrated by actual operating experience xviii that testing natural gas pipelines
that are operated at 72% of SMYS to pressure levels at or above 90% of SMYS (a 1.25 test-
pressure-to-operating-pressure ratio) eliminates manufacturing defects that otherwise might
cause in-service failures. However, there are circumstances that one should be aware of
wherein service failures have occurred after a pressure test to at least 1.25 times MOP. A
recent review of the track record of pressure testing xix revealed the existence of such cases.
The authors of Reference XVIII (Page 18) documented a case where a pressure surge that
caused pressures to rise significantly above the MOP resulted in an in-service failure. It was
speculated that additional in-service failures may have also involved unknown or unmeasured
pressure upsets where the MOP was significantly exceeded. In other cases, it was evident that

FINAL
15-019
retests might have prevented the in-service failures if conducted in a timely manner to eliminate
defects that were growing by fatigue or another time-dependent phenomenon. The authors
noted that pressure testing is not effective at preventing subsequent leaks from short, deep
defects that survive a given pressure test. In any case this study showed that when pressure
tests were conducted at timely intervals with pressure levels exceeding 90% of SMYS and when
pressure excursions significantly in excess of MOP were prevented, no subsequent in-service
ruptures occurred in the subject pipelines.
As noted in the previous paragraph, not all leaks can be prevented by high-pressure testing.
Leaks might occur a short time after a pressure test from short, deep defects. This
phenomenon is explainable in terms of the previous fracture mechanics discussion on the way
testing works. For example, Table 9 shows that a 1-inch-long, 92% through-the-wall defect
could survive a test to 1,300 psig (HTP1). Yet, this defect needs to grow only to a depth of
97% through the wall to fail at the MOP. A 2-inch-long, 76% through-the-wall defect also could
survive a 1,300-psig test, and it would have to grow to a depth of 92% of the wall thickness to
fail at the MOP. In contrast, the longer defects (5, 10, 15, and 20 inches) have to grow to
depths of 71, 54, 45, and 40% through the wall to fail at the MOP. This indicates that short,
deep defects can survive a high-pressure test but grow to failure a short time after the test. As
indicated by the leak/rupture curve, such defects would tend to fail as leaks because they are
too short to fail as ruptures. However, the fact that a pressure test may not be very effective at
preventing future leaks could be a limitation on the effectiveness of such testing, particularly for
a hazardous liquid pipeline where a small leak occurring over a period of time can create
significant environmental damage. In the case of short, deep corrosion pits such leaks can
usually be prevented by using ILI to find them either before the test is performed or shortly
thereafter, so that they can be repaired. However, ILI has not been particularly successful at
locating short, deep lack-of-fusion defects in ERW seams.
A phenomenon sometimes associated with pressure testing that constitutes more of nuisance
than a real limitation is the likelihood of occurrence of “pressure reversals”. Pressure reversals
are nothing new; they have been observed in pressure testing for a long time xx. A pressure
reversal is said to have occurred when a test failure takes place at a pressure level lower than
the level achieved in a recent pressurization. The potential for such an occurrence tends to be
unsettling because it suggests that reaching a given level of test pressure may not be an
absolute guarantee that the margin of safety is equal to the test-pressure-to-operating-pressure
ratio associated with that pressure level. However, research has shown how pressure reversals
arise and that they are normally not a threat to the integrity of a pipeline after it has been
subjected to a pressure test.
Research xxi has demonstrated that when pressurized to a near-failure level, a defect will tend to
grow, and as its failure level is approached, the growth will continue until failure occurs even if

FINAL
15-019
the pressurization is stopped and the pressure is held constant. If the pressure is suddenly
released when a defect is at this “creep” stage, the defect likely will be incapable of enduring
another pressurization to the same level. Because of the accumulated damage on one or more
prior pressurizations and depressurizations such a defect is likely to fail on a subsequent
pressurization at a level below that of the recent prior pressurization. This difference in failure
pressure from that of the prior pressurization is termed a “pressure reversal”. In effect, a
pressure reversal is an extreme case of low-cycle fatigue. Typically, pressure reversals are
encountered during a test in which multiple defects fail, each failure causing a sudden
depressurization to zero, requiring re-pressurization to continue the test.
Fortunately, experience has shown xxii that the likelihood of a pressure reversal occurring
decreases exponentially with the difference in pressure between its failure pressure and the
highest previous pressure. Thus, for example, while a pressure reversal equal to 1% of the test
pressure might be expected to occur in one of every 10 to every 100 pressurizations, a pressure
reversal exceeding 20% of the test pressure would be expected to occur only once in several
million pressurizations. As a result one need not be concerned that a defect surviving a
pressure test might subsequently fail upon pressurization to the maximum operating pressure.
The chance of such an occurrence is not zero, but it is so small that it constitutes a negligible
risk. Moreover, if a test is carried out successfully with no test failures and depressurized
slowly, then there would seem to be little likelihood of a pressure reversal of any size occurring.
A significant limitation on pressure testing is the fact that a test does not identify the types or
locations of defects that may exist after the test. As Figure 1 shows, no matter what test
pressure level is chosen (short of the level that would cause sound pipe to burst), the possibility
of defects with failure pressures above the test pressure remaining in the pipeline certainly
exists. The pipeline operator will have no knowledge of the natures or locations of any such
defects. The only way to assure that such defects will not grow to failure at the operating
pressure at some time in the future after the test is to find and eliminate them (for example, by
means of ILI) or to subject the pipeline to another pressure test before they have time to grow
to failure in service. The concept of timely retesting is explained in the next section.
Finally, it is noted that hydrostatic testing is not effective for preventing failures from HSC in
pipe body hard spots or hard heat-affected zones of ERW seams. This is because hydrostatic
testing will not cause a pipe body hard spot or a hard heat-affected zone of an ERW seam to
fail unless it is already cracked sufficiently by HSC. It is not possible to predict the rate of
growth for HSC. If HSC starts to occur, the failure can occur rapidly, even instantaneously, so
there is no way to predict when a hydrostatic test would be effective. The practical means for
mitigating an HSC threat (which do not include hydrostatic testing) are discussed separately
later in these guidelines.

FINAL
15-019
Trade-off between Test Pressure and Retest Interval
As indicated previously, pressure testing does not eliminate or prove the absence of all defects,
only those that are large enough to fail at the test pressure. Defects that survive the test may
grow with time after the pipeline is restored to service. Defects that have the potential to grow
with time in service include manufacturing defects that grow from pressure-cycle-induced
fatigue (PCIF), corrosion-caused metal loss, and SCC. Conducting periodic pressure tests is an
acceptable method to assure that none of these can cause an in-service failure, though it may
not always be the best method (e.g., preventing in-service failures from corrosion-caused metal
loss via ILI is usually a more efficient method than pressure testing).
If a pipeline operator chooses to use periodic pressure testing to prevent failures from time-
dependent defect growth, each test must be done in a timely manner to prevent in-service
failures. Higher test-pressure-to-operating-pressure ratios lead to longer intervals between
tests because as test pressure is increased, the maximum size of surviving defects decreases.
The number of test failures tends to grow exponentially with increasing test pressure, and the
number of test failures that can be tolerated generally sets an upper bound on test pressure.
Also, if pipe is tested to pressure levels approaching or exceeding 110% of SMYS, some yielding
of pipe may occur. Most pipeline operators prefer to either limit the amount of yielding or to
avoid it altogether. Described below are methods a pipeline operator can use to estimate an
appropriate combination of test pressure and retest interval.
Predicting Retest Intervals for Defects Having a Constant Rate of Growth
It is usually assumed that corrosion-caused metal loss and SCC grow at constant rates with the
passage of time. For any such growth mechanism it is possible to predict an appropriate retest
interval using a defect failure stress prediction model such as the Modified Ln-Secant Model
described in Figure 1 or the API 579-Level II model described in Figure 2. Because the principle
is the same whether the material is ductile or brittle, the process for predicting retest intervals
described below is based on the assumption that the material exhibits ductile fracture initiation.
Hence, the curves of Figure 1 based on the Modified LnSec model will be used. The same
principle could be applied to a material that exhibits brittle fracture behavior. Note, however,
that because the defects that cause failures at the various test pressure levels and at the MOP
are much smaller for the brittle material, the predicted retest intervals are likely to be much
shorter for the defects in a brittle material for the same defect growth rates.
The data in Table 9 for the ductile (40-ft-lb) material indicates that a pressure test to 1,170 psig
(HTP 1) would be expected to eliminate 1-inch-long defects that are deeper than 95% of the
wall thickness. The data also indicates that if a 1-inch-long defect that is 95-percent-through
survives the test, it must grow (in depth only) to a depth of 97% through-the-wall to fail at the
MOP of 936 psig. Since the wall thickness of the pipe was given as 0.375 inch, this 2%-of-wall-

FINAL
15-019
thickness growth is equal to 0.0075 inch. If the growth rate is 0.001 inch per year, the time to
failure is 7.5 years. It is common to impose a safety of factor of 2 on this calculated time to
failure to get an appropriate retest interval. The retest interval, then, is 3.75 years for a
pressure test to 1,170 psig. The retest intervals for all six of the defects listed in Table 9 are
shown in Table 10 for the 1,170-psig test (HTP 1) based on the type of calculation just
described for the 1-inch-long defect.
Table 10. Retest Intervals for a Pressure Test to 1,170 psig with a Test-Pressure-to-
Operating-Pressure Ratio of 1.25
Length, d/t, d/t, Δ% Δ wall Time to Failure Retest

inch MOP,
% HTP 1, Thickness, At rate 0.001 inch Interval,
%
1 97 95 2 0.0075 8 4
2 92 85 7 0.02625 26 13
5 71 55 16 0.06 60 30
10 54 36 18 0.0675 68 34
15 45 27 18 0.0675 68 34
20 40 23 17 0.06375 64 32
The effect of increasing the test-pressure-to-operating-pressure ratio can be seen by comparing

the results in Table 10 with those in Table 11. Table 11 presents retest intervals for the same
30-inch-OD pipeline with a defect growth rate of 0.001 inch/year but with a test to a pressure
level of 1,300 psig (100% of SMYS) rather than to 1,170 psig (90% of SMYS) as was the case
presented in Table 10.
Table 11. Retest Intervals for a Pressure Test to 1,300 psig with a Test-Pressure-to-

inch MOP,
%
1 97 92 5 0.01875 19 9
2 92 76 16 0.06 60 30
5 71 42 29 0.10875 109 54
10 54 24 30 0.1125 113 56
15 45 18 27 0.10125 101 51
20 40 15 25 0.09375 94 47

FINAL
15-019
The safe retest interval for the long defects is 47 years for the 1,300 psig test (a test-pressure-
to-operating-pressure ratio of 1.39) compared to the safe retest interval of 30 years for the long
defects after a test to 1,170 psig (a test-pressure-to-operating-pressure ratio of 1.25).
Pipeline operators need to be aware of the fact that the benefit of a given test-pressure-to-
operating-pressure ratio decreases as the maximum operating stress level decreases as a
percentage of SMYS. Consider the case where the 30-inch-OD, 0.375-inch-wall, X52 pipeline
with a Charpy upper shelf energy of 40 ft-lb used in the previous examples is operated at a
maximum stress level of 50% of SMYS (650 psig). The same defects in terms of lengths now
have to be deeper to fail at the maximum operating pressure of 650 psig than they would have
to be to fail at the maximum operating pressure of 936 psig. If a test-pressure-to-operating-
pressure of 1.25 times MOP is utilized (referred to as HTP 3), the test pressure would be 813
psig. The results of such a test in terms of retest intervals can be seen in Table 12.
Table 12. Retest Intervals for a Pressure Test to 813 psig with a Test-Pressure-to-

inch MOP, HTP 3, Thickness, At rate 0.001 inch Interval,
% % inch per year, years years
1 99 98 1 0.00375 4 2
2 95 93 2 0.0075 8 4
5 83 77 6 0.0225 23 11
10 72 62 10 0.0375 38 19
15 64 53 11 0.04125 41 21
20 60 48 12 0.045 45 23
The data in Table 12 indicate that the safe retest interval for the longer defects is 19 years (the
5-inch-long defect in this case is predicted to fail as a leak). By comparison the test to 1,170
psig for a maximum operating pressure of 936 psig (with the same test-pressure-to-operating-
pressure ratio of 1.25) assures a retest interval of 30 years for the longer defects. This
situation is created by the fact that the sizes of defects remaining after a test increase with
decreasing test pressure.
If the example pipeline is operated at 650 psig because it is a natural gas transmission pipeline
in a Class 3 area, then the operator would have to conduct a pressure test to 975 psig to
comply with 49 CFR 192.619(a)(2)(ii) (test-pressure-to-operating-pressure ratio of 1.5). Table
13 indicates that such a test would result in a safe retest interval of 39 years for the longer
defects. This retest interval is shorter than the 47-year interval associated with testing a similar
pipeline with a maximum operating pressure of 936 psig to a level of 1,300 psig (a test-

FINAL
15-019
pressure-to-operating-pressure ratio of 1.39). These comparisons illustrate that pipelines that
are operated at relatively low stress levels need to be tested with higher test-pressure-to-
operating-pressure ratios than high-stress pipelines to achieve to same length of retest interval.
Two empirical equations are given for selecting appropriate test-pressure-to-operating-pressure

ratios for a pressure test based on the absolute operating stress level are presented in TTO6xxiii.
“TTO6” was a special report on spike hydrostatic testing that was prepared for the Research
and Special Programs Administration 16 of the U.S. Department of Transportation. One of the
equations is associated with pressure-cycle-induced-fatigue (PCIF), and the other is associated
with SCC. They are as follows.
For PCIF:
𝐻𝐻𝐻𝐻𝐻𝐻⁄𝑀𝑀𝑀𝑀𝑀𝑀 = −0.00736(%𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑎𝑎𝑎𝑎 𝑀𝑀𝑀𝑀𝑀𝑀) + 1.919 Equation 1
For SCC:
𝐻𝐻𝐻𝐻𝐻𝐻⁄𝑀𝑀𝑀𝑀𝑀𝑀 = −0.02136(%𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑎𝑎𝑎𝑎 𝑀𝑀𝑀𝑀𝑀𝑀) + 3.068 Equation 2
where
HTP stands for hydrostatic test pressure, psig
MOP stands of maximum operating pressure, psig
One can determine from Equation 1 that the suggested HTP/MOP ratio for a pressure test for a
pipeline that is operated at 72% of SMYS is 1.39, and that the suggested HTP/MOP ratio for a
pressure test for a pipeline that is operated at 50% of SMYS is 1.55. Similarly, one can
determine from Equation 6 that the suggested HTP/MOP ratio for a pressure test for a pipeline
that is operated at 72% of SMYS is 1.53, and that the suggested HTP/MOP ratio for a pressure
test for a pipeline that is operated at 50% of SMYS is 2. It is emphasized that these are only
guidelines. An operator should evaluate the circumstances of each particular pipeline segment
to be pressure tested to select the appropriate test-pressure-to-operating-pressure ratio.
16
The Research and Special Programs Administration (RSPA) was the predecessor Federal agency to PHMSA.

FINAL
15-019
Table 13. Retest Intervals for a Pressure Test to 975 psig with a Test-Pressure-to-
Length, d/t, d/t, Δ wall Time to Failure Retest

Δ%
inch MOP,
1 %
99 97 2 0.0075 8 4
2 95 91 4 0.015 15 8
5 83 69 14 0.0525 53 26
10 72 51 21 0.07875 79 39
15 64 42 22 0.0825 83 41
20 60 37 23 0.08625 86 43
Predicting Retest Intervals from Prior Test Results for Defects Having a Constant
Rate of Growth
Fessler and Rapp xxiv developed a novel approach to predicting retest intervals for SCC. Though
developed for use with pipelines affected by SCC, it would seem that the method is applicable
to any defect growth mechanism where the rate of growth is constant. This method involves
conducting an initial pressure test after an SCC problem has been identified, and conducting a
second pressure test an appropriate amount of time after the first test. The appropriate
amount of time can be estimated via the type of analysis described above. However, all future
tests after the second can then be determined on the basis of a rational schedule established as
follows.
The pressure level reached in the first pressure test is used to establish the maximum sizes of
remaining defects. The time for the second test can be calculated as shown previously using a
failure pressure prediction model and a rate of defect growth that has either been established
from the history of the pipeline of interest or can be obtained from standards such as ASME
B31.8S or research studies xxv. Alternatively, one can use observed depths of defects and a
distribution of starting times of defect growth via a Monte Carlo simulation to obtain a probable
maximum crack growth rate xxvi.
The second test allows one to calculate the “effective” crack growth rate for establishing the
time of the “next” retest. The time of the first hydrostatic test is again used as the starting
point and the “next” hydrostatic test allows one to calculate the crack growth rate for
establishing the time to conduct the “next after the next” hydrostatic test. Certain assumptions
must be made:
• The pipeline contains SCC (or another type of time-dependent defect with a constant
rate of growth).

FINAL
15-019
• The growth rate for surviving cracks will be less than the growth rate for a crack that
has previously failed.
• Cracks initiating in the future will not fail before an existing crack fails.
• Future operating conditions will be no more severe than past operating conditions.
• The crack depth increases linearly with time.
• There is a direct relationship between failure pressure and defect size (the larger the
defect, the lower its failure pressure).
• Successive hydrostatic tests are carried out at the same pressure.
The Fessler/Rapp approach focuses on an “important” crack that is “zero” depth at the time of
the first test. The important crack is assumed to be the right depth to barely survive the
second test and that it becomes deep enough to fail at the MOP/MAOP some time after the
second test. The projected time required for the important defect to grow to failure at the
MOP/MAOP becomes the maximum time interval before the third test must be performed. The
process for projecting the times for the fourth, fifth, and subsequent tests involves the crack
starting at zero-size after the first test and growing to a size that just survives the current test.
This crack is assumed to grow after the current test until it becomes large enough to fail at the
MOP/MAOP. The time for the next test is less than or equal to the time it takes for the crack to
grow to failure at the MOP/MAOP.
Cracks that are larger than zero-size at the time of the first test will either fail in service before
the next test (in which case their growth rate does not matter and the arbitrarily chosen interval
was too long) or survive the next test (in which case their growth rate is less than that of the
important crack). If a crack that is zero-size at the time of the first test grows to a size that
fails in service before the second test is conducted, the arbitrarily chosen interval was too long.
A crack that is zero-size at the time of the first test and grows to a size that fails during the
second test is not important in establishing the next retest time, but its growth rate is higher
than that of the important crack. It is assumed that the faster growing crack is not important
because it is eliminated by the test. Cracks of any size that survive both the first test and the
second test will have growth rates less than or equal to the important crack growth rate.
If no factor of safety is employed, the important crack will be on the verge of failure at the MOP
at the time of the next test.
The equation for the time of the next test is:
 HTP − MOP 
T[(n−1),(n )] = T(cummulative through n-1 )   Equation 3
 FSP − HTP 

FINAL
15-019
where: HTP is hydrostatic test pressure
MOP/MAOP is maximum (allowable) operating pressure
FSP is the pressure corresponding to flow stress
Tn-1,n is the time interval between tests n-1 and n
n represents the current test
n-1 represents the test prior to the current test
T(cumulative through n-1) represents the time from the first test through the test prior
to current test
Each successive interval will be larger than the previous interval. The recommended value of
flow stress is the average of the actual measured yield and ultimate strengths of the material.
If these values are unknown, an alternative is to use the specified ultimate strength of the
material. The use of a relatively high value of flow stress leads to conservative estimates of
retest intervals because it allows larger starting defect sizes.
An example of the use of the Fessler/Rapp method is shown in Figure 3. The example involves
a 30-inch-OD, 0.375-inch-wall, X52 pipeline. The actual measured yield strength is assumed to
be 60,000 psi, and the actual measured ultimate strength is assumed to be 80,000 psi. It is
assumed that each hydrostatic test will be carried out at a pressure level of 1,430 psig (110%
of SMYS). It is also assumed that the interval between the first and second hydrostatic tests
was 2 years. In a real situation any reasonable interval between tests can be chosen, and the
shorter the interval, the more conservative will be the analysis. There is a risk of a service
failure from SCC occurring after the first test if too long an interval is assumed.

FINAL
15-019
Test time based on Fessler/Rapp
2000
1800
1600
1400 2 5 11 24 55 FSP
MOP
1200 HTP
Presure, psig
1000
800
600
400
First Test
200
0
0 10 20 30 40 50 60
Years until retest
Figure 3. Test Intervals Based on Fessler/Rapp
The calculated times for retesting based on this approach are:
• 2 years after the first test
The horizontal lines on Figure 3 represent the flow stress 17 pressure of 1812.5 psig, the test
pressure of 1430 psig, and the maximum operating pressure of 936 psig. Note that the test
times are based on the sloping lines all radiating from the flow stress pressure and Year Zero.
The maximum retest interval is determined when the failure pressure of the important crack
decreases to the MOP. If a factor of safety is desired, one way to achieve that with this method
is to assume that the first interval was 1 year instead of the actual interval of 2 years (half the
time). If this were done the calculated times for retesting would be:
• 1 year after the first test
17
Flow stress is a material property, It is usually defined as the tensile stress level at which the material will fail. It is above the
actual yield strength of the material but usually less than the ultimate tensile strength of the material.

FINAL
15-019
Either way it is seen that the Fessler/Rapp approach will always predict increasing retest
intervals.
A modified version of the Fessler/Rapp approach has been proposed by Sen and
Kariyawasam xxvii. The modified approach (referred to herein as the Modified Fessler/Rapp)
involves only one major change. Instead of taking the starting point of zero depth for the
important crack as the time of the first test, the modified approach assumes that the starting
point of zero depth for the important crack after each test is the year of the previous test. The
equation for the modified approach is:
 HTP − MOP 
T[(n −1),(n )] = T[(n - 2 ),(n −1)]   Equation 4
 FSP − HTP 
The only term that is different is 𝑇𝑇[(𝑛𝑛−2)(𝑛𝑛−1)]. This is simply the time interval between the last
test and the test before the last test. Based on the same conditions that were used to
construct Figure 3 (Fessler/Rapp), the results of a Modified Fessler/Rapp approach are shown in
Figure 4.
Test Times based on Modified Fessler/Rapp

2000
1800
1600
1400 2 5 8 12 18 FSP
MOP
Pressure, psig
1200 HTP
1000
800
600
400
First Test
200
0
0 2 4 6 8 10 12 14 16 18 20
Years until retest
Figure 4. Test Intervals Based on Modified Fessler/Rapp

FINAL
15-019
The calculated times for retesting based on this approach are:

As in the Fessler/Rapp approach, the maximum retest interval is determined when the failure
pressure of the important crack decreases to the MOP. If a factor of safety is desired, one way
to achieve that with this method is to assume that the first interval was 1 year instead of the
actual interval of 2 years (half the time) as was possible with the original Fessler/Rapp Method.
If this were done the calculated times for retesting would be:

Comparing the Modified Fessler/Rapp approach to the original Fessler/Rapp approach, one finds
that the former gives more conservative (i.e., shorter) retest intervals. Another feature of the
modified approach is that, unlike with the original method, the retest intervals under certain
circumstance may actually decrease with time. That is not possible with the original method.
For example, using the same parameters as before except changing the test pressure from
1,430 psig (110% of SMYS) to 1,300 psig (100% of SMYS), one gets the results shown in
Figure 5.

FINAL
15-019
Test Times based on Modified Fessler/Rapp
2000
1800
1600
1400 FSP
2 3 4 5 6 MOP
Pressure, psig
1200 HTP
1000
800
600
400
First Test
200
0
0 1 2 3 4 5 6
Years until retest
Figure 5. Test Intervals Based on Modified Fessler/Rapp that Decrease with Time
Whether or not this is reasonable, it does point out the importance of test-pressure-to-
operating-pressure ratio. Again, it is clear that the higher that ratio is for a given situation, the
longer will be the minimum required retest interval. It is noted that making the same change in
test pressure using the original Fessler/Rapp approach results in decreases in the retest
intervals as well though the intervals still increase with each test. The effect is shown in Table
14.
Table 14. Effect of Test-Pressure-to-Operating-Pressure Ratio Based on the Original

Fessler/Rapp Model
Test Years to the Test After the Years to the Test After the
First Test With a Test to First Test With a Test to
Number 100% of SMYS 110% of SMYS
1 2 2
2 3 5
3 6 11
4 10 24
5 17 55
Predicting Retest Intervals for Defects Having a Non-Constant Rate of Growth
One form of defect growth that is characterized by a non-constant rate of crack growth is
pressure-cycle-induced fatigue (PCIF). PCIF is a recognized threat to the integrity of a

FINAL
15-019
hazardous liquid pipeline. PCIF failures are not believed to be a near-term threat to natural gas
pipelines because of their less-frequent and lower-amplitude pressure cycles. But, whatever the
timing, the threat of failure from PCIF can be addressed by periodic integrity assessment.
Integrity assessment can be accomplished either by pressure testing or by in-line inspection
(ILI) using a suitable crack-detection tool. In this section the use of pressure testing to
mitigate a PCIF threat is discussed.
Scheduling retesting to mitigate PCIF involves fatigue-crack-growth analysis. The analysis

involves establishing the initial sizes of defects, applying representative operational pressure
cycles to cause the defects to grow, and determining the number of pressure cycles required to
cause the defects to attain (final) sizes that will cause a failure at the maximum operating
pressure (MOP) of the pipeline. The number of pressure cycles required to grow the initial
defects to failure corresponds to a certain period of time, so the output of the analysis is a time
to failure for each defect considered. A factor of safety is then applied to the time to failure so
that a response is made well before any growing defect can reach a size that would cause
failure at the MOP.
A commonly-used basis for fatigue crack growth modeling is the “Paris-law” approach named
after its principal developer xxviii. The Paris-law equation is generally written as follows.
𝑑𝑑𝑑𝑑
= 𝐶𝐶(∆𝐾𝐾)𝑛𝑛 Equation 5
𝑑𝑑𝑑𝑑
where:
da/dN is an increment of crack growth, inch/cycle
C and n are constants for a particular material and environment
ΔK is the change in stress intensity factor at the tip of the fatigue crack during a cycle of
changing applied stress, 𝑝𝑝𝑝𝑝𝑝𝑝√𝑖𝑖𝑖𝑖𝑖𝑖ℎ
ΔK is calculated using the Raju/Newmanxi equation which has the form:
∆𝐾𝐾 = 𝐹𝐹∆𝜎𝜎√𝜋𝜋𝜋𝜋 Equation 6

where:
F is a constant that depends on the shape of the stressed element, the ratio of defect
depth to wall thickness, the ratio of defect depth to defect length, and the position of
the tip of the crack
Δσ is the change in stress during a cycle of loading and unloading, psi
a is the depth of a part-through-wall crack, inch

FINAL
15-019
Note that ΔK is a function of the square root of a. This means that even if the stress cycle, Δσ,
is constant, ΔK will increase with the passage of time (as cycles of pressure are continually
applied). As time passes a increases at an increasing rate. An example of a versus time is
shown in Figure 6 for a particular pipe size, defect length, and applied pressure history. The
crack which has an initial depth of 0.01 inch is shown to grow to a depth of 0.16 inch in 130
years at which point it is deep enough to cause failure at the MOP.
a versus N (same N every year)

0.18
Depth that will
0.16 cause crack to fail
at the MOP
0.14
depth of crack, a, inches
0.12
0.1
0.08
0.06
Initial depth at start
of operations
0.04
0.02
0
0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0
years
Figure 6. The Non-Constant Nature of Crack Growth from PCIF
The fact that the a crack begins to grow rapidly as it approaches the size that will fail at the
MOP, means that mitigation in the form of a pressure test must either be applied as early as
possible during the life of a crack or it will have to be repeated frequently if it is capable of
finding only larger cracks. For example, if a pressure test applied to the crack represented in
Figure 6, could eliminate a defect as shallow as 0.02 inch (the depth after 60 years of
operation), the time to failure for any shallower crack would be at least 70 years, and a retest
interval of 35 years (based on a factor of safety of 2) would be feasible. If, however, the
shallowest crack that can be eliminated is 0.065 inch (the depth after 120 years of operation),
the time to failure is only 10 years, and a retest interval of 5 years is required.

FINAL
15-019
The number of cycles of a given level of stress needed for a crack with an initial depth, ai , to
grow to a final depth, af, is calculated by integrating Equation 5. Because of the complexity
associated with most practical structures, closed-form integration of Equation 5 is usually not
possible, and software designed for the purpose of numerical integration is generally used to
perform the integration and calculate cycles to failure. The software models available for
calculating cycles to failure also must be able to count cycles with variable stress ranges
because most real loading situations involve load applications that vary with time. Various
cycle-counting schemes are presented in ASTM 1049-85 (2011)e1, and the one that is often
used in the analyses PCIF in pipelines is called “rainflow” counting.
Defects that remain after a pressure test can be no larger than the size that would have caused
a test failure, so the maximum test pressure is used to establish the initial sizes of a
representative sample of defects with different length-depth combinations that could have
barely survived the test. The minimum time to failure for the worst-case defect modified by the
factor of safety determines when retesting is needed to assure pipeline integrity.
The times to failure after a pressure test can be calculated via a Paris-law approach, provided
that the user is able to supply the relevant data that includes pipe geometry and strength level,
the relevant operating pressure-cycle spectrum and test pressure history for the segment being
assessed. Other factors that affect the times to failures include material toughness, flow stress,
and the crack growth rate constants associated with the Paris-law equation. These latter
factors may not be known for each and every piece of pipe in a pipeline. However, a sensitivity
analysis xxix shows that the analyst can expect to obtain conservative estimates of times to
failure after a pressure test by assuming a toughness level corresponding to a full-size-
equivalent Charpy upper-shelf energy level of 200 ft lb 18 and a flow stress equal to the
minimum specified ultimate tensile strength of the base metal. Experience shows that the crack
growth rate constants found in the API 579 standard for fitness-for-service are acceptable.
Lastly, a factor of safety of 2 should be applied to the calculated times to failure to account for
uncertainties in the material properties and the calculation process. The rational for the factor
of safety of 2 is explained as follows in Reference XIX. “Typically, for the case of times to
failure for defects that could have barely survived a hydrostatic test, a safety factor of 2 has
been applied. That is, the recommended time for retesting is half the calculated time to failure
for the defect with the shortest predicted time to failure. Our experience suggests that this
time to failure is adequate and appropriate. The factor of safety of 2 in this context was
suggested more than 20 years ago when this type of analysis was first being applied to
scheduling retesting. It was chosen somewhat arbitrarily, but with the knowledge that ASME
18
A high value of Charpy energy is used because the remaining defects after a test will be the larger than if a lower (and more
realistic) value is used. The high value results in a shorter retest interval than would be the case if a low value were to be used.

FINAL
15-019
pressure vessel design utilizes a factor of 2 on stress to avoid fatigue failures in service. The
fact that ASME practice is to apply a factor of 20 on cycles to failure was not seen as a practical
approach for fatigue crack growth in pipelines because the ASME practice applies to crack
initiation not crack propagation. That is, fatigue life in the context of pressure vessel design
envisions the cyclic life of a defect-free structure where most of the fatigue life involves the
initiation of a crack. It is believed that the factor of safety of 20 was instituted to account for
the possibility that an as-built structure might not in fact be defect-free. In contrast, the
fatigue-life prediction method for pipelines is based on a known initial defect size, and the
uncertainty involved in the calculation is believed to be much more limited.”
In using fatigue analysis to calculate the times to failure after a pressure test, it must be
assumed that defects could exist anywhere along the pipeline that are severe enough to have
failure pressures no higher than that of the test pressure. This means that the analyst may
have to calculate times to failure for multiple points along the pipeline taking account of the test
level applied at each location, the wall thickness at each location, the effect of the hydraulic
gradient on the pressure cycles at each location, and the effect of elevation on the static head
at each location.
The study described in Reference XXIV further showed that the calculated times to failure after
a pressure test increase exponentially with increasing test-pressure-to-operating-pressure ratio.
Therefore, the operator can maximize the length of time between retests by utilizing the
highest feasible test pressure that will not cause significant permanent expansion of pipe or an
intolerable number of test failures. In absolute terms, the higher the test stress relative to the
specified minimum yield strength of the pipe, the smaller the remaining defects will be. Smaller
remaining defects mean longer times to failure after the test. For that reason, for a pipeline
that is operated at maximum stress levels below 72% of SMYS, the test-pressure-to-operating-
pressure ratio must be greater than that applied on a pipeline that operates at 72% of SMYS to
achieve the same time to failure as that for the pipeline that operates at 72% of SMYS. This
was illustrated previously for defects that have constant growth rates by the comparisons of
data in Tables 12 and 13. This point is even more critical for defects that have non-constant
growth rates such as PCIF. In particular, pipelines operated at stress levels as low as 30 to
50% of SMYS may be exposed to PCIF if they have only been pressure tested to levels of 1.25
times MOP or less. For this reason, it is advisable to carry out fatigue analysis on all pipelines
that are operated at stress levels above 30% of SMYS, even natural gas transmission pipelines.
The analysis may show that retesting is not required within the useful life of the pipeline, but it
is important to know what the fatigue life of a pipeline is.
Lastly, once a satisfactory test has been achieved and a retest interval has been calculated,
future pressure tests should be made at the same target test pressure level. Otherwise, it will
be necessary to change the retest interval.

FINAL
15-019
Spike Testing
A spike test refers to a pressure test in which the target maximum test pressure exceeds the
minimum test pressure established by the operator for a particular pipeline (such as the test
level required by 49 CFR 192 or 49 CFR 195), but that maximum pressure is held no longer
than about 30 minutes after which the pressure level can be reduced to the level established by
the operator for completion of the required hold times. The spike test, being at a higher
pressure than the established test pressure, serves as the verification of integrity of the
pipeline. The spike test pressure establishes the effective test-pressure-to-operating-pressure
ratio for calculating remaining defect sizes and retest intervals.
The concept of spike testing is not new. Fessler xxx in 1979 was suggesting that a pressure test
of 100 to 110% of SMYS held for only a few minutes followed by leak testing at a lower
pressure could be effective for assessing a pipeline affected by SCC. Leis xxxi in 1992 also
advocated similar testing on the basis that the most critical defects are removed within the first
hour of a pressure test. In 2004 “TTO6” a special report on spike hydrostatic testing was
prepared for the Research and Special Programs Administration19 of the U.S. Department of
Transportation. Most recently, Rosenfeld xxxii has discussed why and when spike testing should
be used and when it may not be appropriate. The following is quoted from his paper. The
phrase “standard test level” as used by Rosenfeld is assumed to mean a regulatory-required
test to a minimum pressure level of 1.25 times MOP or MAOP.
“1. Spike testing is beneficial and therefore recommended in certain specific circumstances,
namely:
a. Where crack-like defects such as SCC, selective corrosion of ERW seams, bond line
defects in older vintage ERW seams, and seam fatigue cracks are expected to exist
based on evidence from inspections, failures; or consideration of pipeline integrity
threats;
b. Where it is desired to increase the retest interval for time-dependent flaws, particularly in
high-stress pipelines; or
c. Where documentation is unable to confirm the attributes of the pipe or is unable to

confirm that a prior hydrostatic test has occurred.
2. Spike testing is neither necessary nor harmful in the following situations:
19
The Research and Special Programs Administration (RSPA) was the predecessor Federal agency to PHMSA.

FINAL
15-019
a. Where the purpose of the test is to demonstrate the strength of the pipe where crack-
like defects are not expected to be present;
b. Where the standard test margin is at least 1.4;
c. Where the operating stress is below 40% of SMYS; or
d. where the pipe being tested is new and of known good quality.
3. Spike testing to higher than the standard test level would be inadvisable in a limited set of
specific circumstances, including:
a. Where the spike test pressure would exceed the mill test pressure or damage the pipe,
particularly pipe susceptible to manufacturing defects in seams;
b. Where the spike test pressure would exceed the recommended maximum test pressure
of components such as flanges or valves;
c. Where the margin above the spike test pressure could be insufficient to prevent
excessive stress due to fluid thermal expansion effects during the test;
d. Where large elevation changes must be accommodated during testing; or
e. Where pneumatic testing is practiced.”
Although Rosenfeld’s suggested criteria for when to apply spike testing are generally sound,
operators may want to consider the following:
• When a spike test cannot be at a sufficient pressure level, lowering the MOP or MOAP is
an option.
• Not everyone agrees that spike testing is harmful when conducted at a pressure level in
excess of the manufacturer’s test level. It is likely that a test pressure above the
manufacturer’s test pressure will cause failures of seam manufacturing defects in older-
vintage line pipe that has never been previously tested to a level in excess of the
pressure reached in the manufacturer’s test. However, such failures will tend to
enhance the integrity of the pipeline because the remaining defects will be smaller than
they otherwise would have been if the pressure in the manufacturer’s test had not been
exceeded.
As an example of a spike test, a pipeline operator might choose to do the following. In addition
to the test required by the regulations for a hazardous liquid pipeline (a minimum level of 1.25
times the MOP for 4 hours followed by a test to 1.1 times the MOP for an additional 4 hours - a
“Subpart E test”), the operator could elect to begin the test by pressurizing the pipeline to 1.39
times MOP, holding that pressure for the first 30 minutes of the first 4-hour hold period. After

FINAL
15-019
this “spike test” the pressure would be lowered to 1.25 times the MOP for the remainder of the
first 4-hour test period. The rationale for the spike test is that it eliminates or proves the
absence of defects that would be expected to remain in the pipeline after a test to only 1.25
times the MOP. The hold period for the spike test is intentionally kept short to minimize the
growth of any defects that are too small to fail. This rationale is explained in more detail in
Reference XXIII (TTO6). The spike test, if conducted at a pressure more than 10% above the
1.25-times-MOP test level, virtually assures that no surviving defects will grow during the hold
period at 1.25 times MOP while leak-tightness is being assessed.
Anticipating Test Failures

One of the drawbacks to pressure testing is the risk of ruptures and leaks occurring during the
test. Such occurrences can significantly lengthen the time that the pipeline must be out of
service, and numerous test failures often involve occurrences of pressure reversals. The
number of test failures is likely to increase as the target test pressure level is increased. While
a high-test-pressure-to-operating-pressure ratio is desirable, attempting to reach a relatively
high target test pressure may come at the price of large numbers of test failures. Once the
failures have occurred, and the target level has been achieved, the operator can confidently use
that test pressure level to predict a retest interval.
Before beginning a pressure test, a pipeline operator should establish a target spike test
pressure based on achieving an appropriate test-pressure-to-operating-pressure ratio to assure
a desired re-assessment interval. As the pressure level in the test section is raised toward the
target test pressure, test failures may begin to take place. Experiencexix has shown that once
test failures start to occur, the number of test failures tends to increase exponentially with
increasing pressure and that some failures will occur as pressure reversals. Reference XIX
discusses actual cases in which the failure pressure levels of the first few failures tended to
establish a trend. The trend could then have been used to predict the likely number of failures
that would occur in attaining the target test pressure level. The cases examined showed that
the method was not highly accurate, but nevertheless, it could be useful in deciding whether or
not to continue increasing the test pressure. If a decision is made to accept a lower final spike
test pressure, the operator should be aware that the decision may result in a much shorter re-
assessment interval.
Test Medium
In most cases water is chosen as the test medium for pressure tests of existing pipelines.
There are two significant advantages to using water instead of a pneumatic medium (air, an
inert gas, or natural gas). First, the use of water is much safer than a pneumatic medium
because of the much lower energy release rate in the event of a failure. Fractures will not
propagate when a failure occurs with water whereas fractures could be lengthy with a

FINAL
15-019
pneumatic medium, and a fire or explosion could result if natural gas is used as the medium. It
should also be noted that pressurized air can be hazardous in a pipeline if a residual film of
liquid petroleum or petroleum product is present and an ignition source exists. Second, the
detection of leaks is much easier with water than with a gas because of the much higher rate of
pressure change with loss of test water compared to that with the loss of the test gas.
Using water as the test medium is also preferable to using liquid petroleum product in a liquid
line. In the event of a test failure, the loss of the liquid petroleum product creates the risk of
fire, explosion, and/or significant environmental damage.
In-Line Inspection
It is often possible to demonstrate the integrity of a pipeline segment via in-line inspection.
Conventional means of in-line inspection involve inserting an appropriate inspection device into
the pipeline while it remains in service, pushing the device through the pipeline with the
shipped product, and recovering the device so that the data acquired as the device traveled
through the pipeline can be analyzed. Depending on the type of device selected, the operator
can inspect the pipeline segment for corrosion-caused metal loss, crack-like defects, geometric
irregularities (e.g., dents, buckles, appurtenances), or changes in alignment due to settlement
or other movement. In the context of repair/replace considerations for pre-regulation pipe,
identifying and characterizing corrosion-caused metal loss and crack-like defects are of primary
interest.
Ideally, anomalies can be identified by ILI and their locations and sizes (length and depth) can
be determined. Also, in an ideal situation, the failure pressure level of each anomaly can be
calculated, the time to failure for each anomaly can be predicted, and a remedial response can
be undertaken in a timely manner for each anomaly based on its predicted time to failure
modified by the factor of safety. In reality, neither of these ideal situations can be fully
achieved because occasionally, defects are incorrectly identified as to type and because
frequently, there are measurement errors. To better understand the significance of such errors,
it is desirable for both ILI vendors and in-the-ditch NDE technicians to define the possible
variations or errors in their measurements. Also, it is desirable for both the ILI vendors and the
in-the-ditch NDE technicians to coordination with one another to optimize the interpretation of
both the types and sizes of the identified defects.
Typically, for corrosion-caused metal loss and crack-like defects, an ILI service vendor analyzes
the data and produces a list of anomalies indicating the location, clock-position, type, length,
and depth of each anomaly. The list may also include a failure pressure level for each anomaly
based on a failure stress prediction model, the relevant material properties (including fracture
toughness if the anomaly is crack-like in nature), and the length and depth of the anomaly.

FINAL
15-019
ILI for Corrosion-Caused Metal Loss
ILI technology for finding and characterizing corrosion-caused metal loss, either from external
corrosion or from internal corrosion, has been proven effective by many years of experience.
Pipeline operators can rely on the results of either magnetic-flux leakage or ultrasonic wall
thickness tools to find and characterize metal loss with a high degree of confidence. 20 Failure
stress prediction models for metal loss anomalies have been thoroughly validated xxxiii. Statistical
analysis techniques are available to rank anomalies by severity and to predict re-assessment
intervals xxxiv. Criteria for responding to metal loss anomalies in a timely manner have been
incorporated into pipeline industry standards such as ASME B31.8S xxxv and API RP 1160 xxxvi.
With these tools, pipeline operators can successfully manage the threat of failure from
corrosion-caused metal loss using ILI. Thus, repair of a pipeline affected by corrosion-caused
metal loss can be carried out on a continuing basis for the long term by relying on ILI.
Replacement of a piggable pipeline because of corrosion-caused metal loss likely would be
considered if the cost of the re-assessments and repair responses became higher than the cost
of replacement.
ILI for Crack-Like Defects

ILI technology for finding and characterizing crack-like defects exists and continues to evolve.
At the current time three ILI crack-detection technologies exist. These are: Circumferentially-
oriented magnetic flux leakage (CMFL), ultrasonic angle-beam crack detection (UTCD), and
Electromagnetic Acoustic Transducer (EMAT). Both CMFL tools and the EMAT tools can be run
in natural gas transmission pipelines and hazardous liquid pipelines. The UTCD tools require
liquid coupling between the pipe wall and the transducers. Such coupling is achieved easily in
pipelines filled with a liquid hydrocarbon (Applies to crude oil and most refined products. Such
tools may not be effective in HVLs.) However, the only way a UTCD tool can be used in a
natural gas pipeline is to introduce a slug of liquid to surround the tool at all times. The latter
requirement is difficult to satisfy in most natural gas pipelines, so the UTCD tools tend to be
used mostly for inspecting hazardous liquid pipelines.
As in the case of ILI for corrosion-caused metal loss, pipeline operators respond to lists of
crack-like anomalies received from tool vendors by excavating a number of anomalies to verify
the accuracy of the results. Unlike in the case of ILI for corrosion-caused metal loss where the
depths and lengths of the exposed anomalies can be measured directly, the operator in the
case of a crack-tool run must rely on non-destructive examination (NDE) to assess the
20
Two possible exceptions are “narrow axial external corrosion” and “selective seam weld corrosion”. Because of their long, narrow
configurations, ILI tools sometimes underestimate the amount of metal loss associated with these two phenomena. A special
configuration of magnetic-flux leakage tool that produces a circumferentially-oriented magnetic field is likely to be a more effective
tool than a normal wall-thickness-measuring tool for characterizing anomalies associated with either of these two phenomena.

FINAL
15-019
anomalies once they are excavated. The NDE usually consists of using magnetic particle
examination to locate all crack-like defects at the OD surface and hand-held ultrasonic angle-
beam inspection or phase-array ultrasonic inspection to assess the lengths and depths of the
cracks. These “in-the-ditch” techniques assist the operator in verifying the accuracy of the ILI
tools, however, these techniques are subject to errors as well. The operator decides on the
basis of the verification digs how many more excavations are needed to assess cracks that
appear to be significant based on the failure pressures calculated using the ILI-predicted defect
dimensions or from the defect dimensions alone. Because the behavior of a crack-like defect
typically depends on the toughness of the material, a representative value of toughness must
be acquired or assumed in order to calculate failure pressure. If crack-like defects in the bond
line of a LF-ERW material are the target, it may be prudent to use a brittle fracture criterion
such as the Newman/Raju equations and assume a toughness corresponding to a Charpy V-
notch energy of 1 ft-lb 21 (A fracture toughness of 9.4 ksi-root-inch via the Barsom and Rolfe
equation). Values this low or lower have been inferred from full-scale burst tests. The anomalies
for the “dig” list can be chosen based on a calculated value of failure pressure or based on
other information provided by the vendor such as the length, depth, or type of anomaly.
Anomalies on the dig list are then excavated, examined, and repaired if necessary.
The track records of the three types of tools mentioned above from the standpoint of their
ability to locate and satisfactorily characterize defects in or adjacent to ERW and flash-welded
seams were recently reviewed xxxvii. Thirteen tool-run cases involving 741 miles of pipe were
reviewed, nine involving UTCD tools, three involving CMFL tools (without the enhanced data
analysis developed by Kinder Morganxxxviii), and one involving an EMAT tool 22. Of the 13 cases
examined, it was reported that none of the runs provided full confidence in the seam integrity
of the assessed segment. In most of the cases the agreement between the ILI results and field
NDE was not particularly good. Where anomalies were cut out and examined, inaccuracies in
both the ILI results and the field NDE were found. Moreover, in some cases anomalies were
missed by the ILI that were larger than the stated detection thresholds of the technologies.
Aside from the thirteen cases it was also pointed out that three in-service pipeline failures
caused by ERW seam anomalies occurred at normal operating pressures within two years after
the pipelines had been inspected by means of ILI crack-detection tools. One of these cases
involved the use of a UTCD tool. The other two cases involved the use of both a UTCD tool and
a CMFL tool. In these cases the tools either altogether missed the defect that caused the
failure or mischaracterized the defect as a non-injurious feature.
21
The minimum value one can obtain from an actual Charpy test is generally 2 to 3 ft lb. That is because energy is required to
move the broken pieces. Lower values are sometimes inferred by back-calculation from the results of a full-scale burst test of a
flawed pipe.
22
It should be noted that the EMAT tool was run for the primary purpose of assessing the pipeline for SCC. The vendor also
reported finding some ERW seam anomalies, but the dimensions of those anomalies were not provided.

FINAL
15-019
The capabilities of UTCD tools as demonstrated in two particular situations were assessed in a
2008 paper by Phillips and Maier xxxix. In the first situation, UTCD data from a seam-integrity
assessment of a pipeline comprised of LF-ERW pipe was analyzed. In the second situation,
UTCD data for several runs of small-diameter pipe were used to assess for SCC. The authors of
Reference XXXV suggest that responding to cracks detected by UTCD ILI is similar to
responding to corrosion-caused metal loss detected by magnetic flux leakage ILI and ultrasonic
wall thickness ILI except that there may be less certainty in the case of crack detection. In the
first place, the behavior of a crack-like defect is affected by material toughness (which is often
unknown) whereas the behavior of a metal loss defect depends only on the strength of the
material which can be approximated based on the grade of the material. Secondly, the extent
of external corrosion-caused metal loss can be seen visually and measured directly and the
extent of internal corrosion-caused metal loss can be determined by ultrasonic thickness
measurements, whereas the extent of a crack can only be assessed non-destructively by angle
beam ultrasonic transducers or eddy-current methods. The latter measurements are prone to
be less accurate than measurements of wall thickness. Thirdly, while the discrete maximum
depth of each metal loss anomaly is given by a metal loss tool, depth ranges rather than
discrete depths are given by the UTCD tool for each anomaly (e.g., less than 0.04 inch, 0.04
inch to 0.08 inch, etc. or less than 12.5% of wall thickness, 12.5% to 25% of the wall
thickness, etc.). Lastly, cracking is generally considered to be more of an integrity threat than
metal loss because of the tendency of sharp cracks to extend under load from ductile tearing or
sudden brittle fracture.
The authors of Reference XXXV assert that the investigations of anomalies found by the UTCD
tool in ERW seams help to build confidence in the use of the tool for managing ERW seam
integrity. It is important to note that they did not claim that the use of the UTCD tool for ERW
seam integrity assessment gives as much confidence as a hydrostatic test. A large amount of
scatter was seen in the comparison of ILI-reported depths and those reported by field-NDE. No
data on the veracity of the field-NDE technique were presented. The authors noted the
existence of “false negatives”, that is, anomalies found upon excavation that were not indicated
by the tool. All of these were said to be relatively shallow, though some had depths exceeding
the vendor’s stated depth detection threshold.
Regarding the pipelines where a UTCD tool was used to detect SCC, the authors of Reference
XXXV found that anomalies indicated by the tool to be “crack fields” most often turned out to be
SCC colonies. False negatives, some of which had NDE-determined depths greater than the
depth detection threshold of the tool were found, and the comparison of ILI-reported depths
and those reported by field-NDE showed considerable scatter. No data on the veracity of the
field-NDE technique were presented.

FINAL
15-019
A recent paper on EMAT technology xl suggests that EMAT technology is improving rapidly with
respect to the detection and characterization of SCC. According to the authors, EMAT ILI can
be used in place of a hydrostatic test for assessment of a pipeline for SCC provided that an
EMAT tool run has been thoroughly validated by means of field NDE and that the pipeline has
survived a confirmatory hydrostatic test with no failures from SCC. They noted, however, that
the same cannot be said for the use of EMAT ILI for ERW seam defects. They assert that more
work is needed on pipelines containing ERW seam defects before it can be considered prudent
to substitute an EMAT ILI for a hydrostatic test to assess seam integrity in an ERW pipeline.
Lastly, they acknowledge that field NDE involves uncertainties, and that errors due to NDE will
add to the errors associated with the EMAT ILI itself with the total error being equal to the
square root of the sum of the squares of the tool error and the NDE error.
ILI for Dents

ILI tools exist for characterizing dents, and operators are required to use such tools to locate
dents in high-consequence areas. However, as mentioned previously, dents and mechanical
damage defects are unlikely to cause an operator to consider replacing a pipeline.
ILI for Pipe Body Hard Spots

An MFL tool which has been specifically designed to identify residual magnetism can be used to
locate pipe body hard spots. This was noted by Groeneveld in Reference 2, and it remains a
viable technique for finding pipe body hard spots. Once hard spots have been located, they can
be excavated and the pipe containing the hard spot can be removed. Alternatively, installing a
steel reinforcing sleeve (“Type A” sleeve) or a pressure containing sleeve (“Type B” sleeve) over
a hard spot has been shown to shield the pipe from cathodic protection current, thus
eliminating the source of atomic hydrogen that might cause the hard spot to crack. Also, either
type of sleeve prevents failure in the event that the hard spot cracks. Unfortunately, no ILI tool
has been found that can reliably identify the presence of a hard heat-affected zone adjacent to
an ERW seam. Mitigation of the HSC threat to the latter is discussed in a subsequent section of
these guidelines.
Influence of Failure Stress Prediction Capabilities on Use of ILI-Crack

Detection Tools
From the standpoint of cracks in the base metal of most line pipe materials, the failure stress
levels of longitudinally-oriented crack-like defects can be predicted by any of the well-known
failure stress prediction models mentioned earlier (References VI through X) if depths and
lengths of the anomalies are accurately characterized and the fracture toughness of the
material is known. The adequacy of failure stress prediction in base metal material follows from
empirical observations that cracks in the base metal tend to fail in a ductile manner. As such,
the level of fracture toughness is relatively unimportant. One can make a conservative

FINAL
15-019
assessment in most such cases by assuming a material toughness equivalent to a full-size-
equivalent Charpy upper shelf energy of 15 ft lb. A toughness level corresponding to a full-size-
equivalent Charpy upper shelf energy level of 15 ft lb could be an appropriate value because
that is about the minimum value one can expect for the base metal of a line pipe material
manufactured prior to 1980.
Unfortunately, the ability to make failure stress predictions for anomalies in and adjacent to LF
ERW, DC-ERW, and flash-welded seams is difficult. A recent evaluation of failure stress
prediction methods xli showed that the Modified LnSec modelx gave unreliable predictions of
failure stress for defects in the seams of these materials because the fracture toughness in
these seams was highly variable and because some of the defects (in particular, defects in the
bond line) failed in a brittle manner. This work showed that conservative estimates of failure
stress can be made using the Newman/Raju modelxi if a very low fracture toughness is assumed
(for example, a Charpy energy of 1 ft-lb corresponding to a fracture toughness of 9.4 ksi-root-
inch via the Barsom and Rolfe equation) 23. However, the use of this model with a low fracture
toughness would likely result in predictions of low failure stress for most if not all of the
anomalies identified by an ILI crack-detection tool. That would necessitate excavating and
examining almost every anomaly even though many of the anomalies would turn out to be non-
injurious. Therefore, the selection of ERW seam anomalies for examination may have to be
based on criteria other than predicted failure stress. The appropriate parameters for anomaly
selection should include length, depth, and location of the anomaly (i.e., in the bond line,
adjacent to but not in the bond line). It is questionable whether or not the current ILI crack-
detection technologies are capable of revealing whether or not a crack near an ERW seam is in
the bond line or the heat-affected zone.
Calculating Re-Assessment Intervals for Integrity Assessment by ILI

As in the case of mitigation of time-dependent threats via pressure testing, mitigation by means
of ILI involves crack-growth analysis to predict re-assessment intervals. The analysis involves
knowing the operating pressure spectrum, the initial sizes of defects, the pipe geometry and
material properties (including fracture toughness if crack-like defects are involved), applying
representative crack growth rates to “grow” the defects, and determining the time required to
cause the defects to attain (final) sizes that will cause a failure at the maximum operating
pressure (MOP) of the pipeline. A factor of safety is then applied to the time to failure so that a
response is made well before any growing defect can reach a size that would cause failure at
23
The use of the Newman/Raju empirical equations by themselves could lead to non-conservative predictions of failure stress if the
toughness is not extremely low and the failure stresses of the defects being assessed are higher than about 40% of the flow stress
of the material. In cases where the material is not extremely brittle and/or the failure stresses are higher than about 40% of the
flow stress, an API 579-Level II or Level III analysis should be conducted to include the possible contribution of elastic-plastic
behavior.

FINAL
15-019
the MOP. Where ILI is used as the sole means of periodic integrity verification, analytical tools
must be available to assess the effects of identified defects on pipeline integrity or the pipeline
operator must be willing and able to directly examine and verify the nature the identified
anomalies.
The initial sizes of defects are established from the ILI data considering tool error. The pipe
geometry and nominal strength levels (e.g., SMYS) are usually known. The rates of growth can
be based on the operator’s experience or taken from standards or technical documents. The
final defect sizes for failure at the MOP must be calculated based on a credible failure stress
prediction model. As noted above fracture toughness may not matter that much if the material
can be expected to behave in a ductile manner. In the case of assessment by ILI (unlike in the
case of a hydrostatic test), it is prudent to assume a low value of toughness because the lower
the toughness used in the analysis is, the lower the failure stress of a given defect will be and
the shorter will be the predicted times to failure.
Also, unlike in the case of a hydrostatic test, it is prudent to assume a low value of flow stress
because the lower the flow stress used in the analysis is, the shorter will be the predicted times
to failure after the test. An appropriate level of flow stress for the vintages of materials that are
likely candidates for replacement would be SMYS+10,000 psi.
As in the case involving predicting times to failure from PCIF after a hydrostatic testing, the
crack growth rate constants found in the API 579 standard for fitness-for-service are acceptable
for use in calculating times to failure after a seam assessment via ILI.
In using fatigue analysis to calculate the time to failure from PCIF after a seam integrity
assessment via ILI, the pipeline operator will know where defects that could grow by fatigue
are located and should also be able to tell within certain bounds, the lengths and depths of the
defects. Since the locations of the anomalies are known in the case of assessment by ILI, it is
simply a matter of adjusting the pressure-cycle spectrum from the upstream and active
downstream stations to account for the distance along the hydraulic gradient. An analysis
should be made for all significant anomalies so that the times to failure will be known. The
operator will then be able to prioritize the anomalies by their times to failure and respond in a
timely manner to remediate them before they grow to a size that would cause an in-service
failure.
Assessment of ERW seam integrity using a reliable ILI crack-detection tool should permit longer
intervals between re-assessments than is the case with repeated hydrostatic testing because an
ILI tool should be able to find much smaller defects than those that can survive a hydrostatic
test to the highest feasible test stress levels.

FINAL
15-019
Direct Assessment
Another means of verifying pipeline integrity from the standpoint of specific threats is “direct
assessment”. There are three types of direct assessment intended to address each of the
following threats:
• External corrosion: External corrosion direct assessment (ECDA)

• Internal corrosion: Internal corrosion direct assessment (ICDA)
• Stress corrosion cracking: Stress corrosion cracking direct assessment (SCCDA)
Direct assessment for one or more of these specific threats can be used in place of hydrostatic
testing or ILI if neither of the latter is applicable to a specific segment of pipe. An example of a
natural gas transmission pipeline where conducting hydrostatic testing is not feasible is a single-
feed pipeline serving a community where taking the line out of service and cutting off
customers cannot be tolerated. In many cases, such pipelines also are operated at relatively
low pressures and flow rates, so that pushing a pig for ILI purposes may not be feasible. It is
noted that even if all three types of direct assessment are applied to a particular pipeline, other
threats such as PCIF, if they existed, would still have to be addressed by hydrostatic testing or
ILI. So, direct assessment as the sole means of integrity assessment is typically applied to
natural gas transmission pipelines where the threat of PCIF occurring is insignificant and where
it is not feasible to carry out ILI or hydrostatic testing.
ECDA
A pipeline operator wishing to use ECDA to address the threat of external corrosion should first
consult ANSI/NACE SP0502-2008, “Standard Practice – Pipeline External Corrosion Direct
Assessment Methodology”, the current version of this standard that is accepted by 49 CFR 192
and 49 CFR 195. This standard explains the ECDA process, shows how to apply it, and points
out its limitations. Basically, the methodology involves four steps. The four steps consist of
gathering and integrating data on the pipeline segment to which ECDA will be applied (pre-
assessment), conducting various types of above-ground electrical surveys (indirect inspections)
to identify areas where external corrosion may be occurring or could occur, examining a
sufficient number of such locations by excavating the pipe (direct examinations) to examine its
condition, and assessing the effectiveness of the process as a means of mitigating external
corrosion (post-assessment). At each excavation relevant measurements (such as dimensions
of metal loss and soil resistivity) are made, observations (such as soil type and coating
condition) are noted, and repairs are made where necessary. The data obtained from the
examinations are used to define and execute mitigative actions, to predict the time interval for
the next ECDA, and to assess the adequacy of the ECDA process with respect to addressing the
threat of external corrosion. Operators seeking additional guidance for applying ECDA to a

FINAL
15-019
pipeline segment may find it advantageous to review a systematic approach utilized by one
operator xlii.
Some of the inherent limitations of ECDA are that it may not be very effective in locating
potentially corroding areas on cased pipe, on pipe located in frozen soil, on bare or poorly
coated pipe, and on pipelines with high-dielectric coatings that become disbonded. Another
limitation is that the criteria for interpreting the indirect measurements and for prioritizing
responses are somewhat subjective and may vary from operator to operator. A recent study
sponsored by PHMSA xliii has resulted in improved standards for interpreting indirect
measurements and for prioritizing responses based on considering the type of soil involved with
a particular measurement location.
No data were found in the public domain that could be used to assess the effectiveness of
ECDA as an integrity assessment method for addressing the threat of external corrosion.
However, a pipeline operator who uses ECDA extensively and repeatedly on a particular
segment of a pipeline with uniform parameters (diameter, wall thickness, type of pipe, type of
coating, type of cathodic protection, type of soil, etc.) can acquire a useful database that will
tend to indicate whether or not ECDA is effective as an integrity assessment tool for the threat
of external corrosion for that segment. Most importantly, ECDA will be proven ineffective if in-
service leaks and ruptures from external corrosion continue to occur. The fact that failures do
not occur after successive ECDAs is encouraging, but that fact alone may not be sufficient to
prove effectiveness. The level of effectiveness depends at least to some extent on the following
types of data:
• The number of excavations where corrosion was found when excavations were made
because of the indirect measurements indicating the probable existence of corrosion
(true positive results)
• The number of excavations where no corrosion was found when excavations were made
because of the indirect measurements indicating the existence of corrosion (false
positive results)
• The number of randomly selected excavations where corrosion was found but where the
indirect measurements gave no indication of corrosion taking place (false negative
results)
• The number of randomly selected excavations where no corrosion was found and where
the indirect measurements gave no indication of corrosion taking place (true negative
results)

FINAL
15-019
These data can be represented symbolically as follows:
True Positives CI/CF CI/CNF False Positives
False Negatives CNI/CF CNI/CNF True Negatives

where:
CI stands for corrosion indicated by indirect measurements
CNI stands for corrosion not indicated by indirect measurement
CF stands for corrosion found upon excavation
CNF stands for corrosion not found upon excavation
It has been suggested previously xliv,xlv that these data can be analyzed by means of Bayes
Theorem to establish the probability that external corrosion could be occurring and not be
detected. The smaller that probability, the more effective is the ECDA program. In terms of
the symbols above, Bayes Theorem says that the probability of corrosion being found upon
excavation given that corrosion was not indicated by indirect measurements is given as follows:
𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶∣𝐶𝐶𝐶𝐶)𝑃𝑃(𝐶𝐶𝐶𝐶)
𝑃𝑃( 𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶𝐶𝐶 ) = Equation 7
𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶∣𝐶𝐶𝐶𝐶)𝑃𝑃(𝐶𝐶𝐶𝐶)+𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶∣𝐶𝐶𝐶𝐶𝐶𝐶)𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶)
where:
𝑃𝑃(𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶𝐶𝐶) is the probability of finding corrosion given that corrosion is not indicated by
indirect measurements
𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶) is the probability of corrosion not being indicated given that corrosion exists
𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶𝐶𝐶) is the probability of corrosion not being indicated given that corrosion does not
exist
𝑃𝑃(𝐶𝐶𝐶𝐶) is the probability of finding corrosion upon random excavation
𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶) is the probability of not finding corrosion upon random excavation
Ideally, to get P (CF) and P (CNF) the operator must randomly select 20 to 30 locations for
excavation without being influenced by any indirect measurements. In practice, the operator
may have historical leak data or bell hole examinations that allow an estimation of P (CF).
Alternatively, the operator may be able to excavate many randomly selected locations where
indirect measurements gave no indications of corrosion during an ECDA. Then the probability

FINAL
15-019
of corrosion, P (CF), can be assessed as the number of cases where corrosion was found
divided by the total number of excavations at locations made where there were no indications
from indirect measurements. P (CNF) is 1 – P (CF).
To see how measuring the effectiveness of ECDA works, consider the following examples. In
the first example assume that the probability of finding corrosion anywhere along a segment, P
(CF), and the probability of not finding corrosion, P (CNF), are to be determined by excavation
at randomly selected locations where there are no indications of corrosion. An ECDA is
conducted after which excavations are made to check the results including 20 excavations
where there were no indications. The excavations where there were no indications reveal 2
locations out of these 20 where corrosion is found even though there was no indication of
corrosion meaning that both P (CF) and 𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶) are 2 in 20 or 0.1 and that P (CNF) and
𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶𝐶𝐶) are both 18 in 20 or 0.9. This example could represent a well-coated and well-
protected pipeline. The effectiveness of the ECDA in terms of 𝑃𝑃(𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶𝐶𝐶) is
0.1 ∗ 0.1
𝑃𝑃(𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶𝐶𝐶) = = 0.012
0.1 ∗ 0.1 + 0.9 ∗ 0.9
In the second example assume that the probability of finding corrosion anywhere along a
segment, P (CF) is 0.5 and that P (CNF) is 0.5 based on historical leak data. This example
could represent a poorly-coated pipeline. An ECDA is conducted after which excavations are
made to check the results including 20 excavations where there were no indications. The
excavations where there were no indications reveal 4 locations out of these 20 where corrosion
is found even though there was no indication of corrosion meaning that 𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶) is 4 in 20
or 0.2 and 𝑃𝑃(𝐶𝐶𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶𝐶𝐶) is 16 in 20 or 0.8. The effectiveness of the ECDA in the second case
in terms of 𝑃𝑃(𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶𝐶𝐶) is
0.2 ∗ 0.5
𝑃𝑃(𝐶𝐶𝐶𝐶 ∣ 𝐶𝐶𝐶𝐶𝐶𝐶) = = 0.2
0.2 ∗ 0.5 + 0.8 ∗ 0.5
The ECDA for the segment in the first case is more effective than the ECDA for the segment in
the second case by a factor of 16.7 because the inherent probability of corrosion is lower and
because there were fewer false negatives.
Calculations such of these can be viewed as aids in assessing the effectiveness of ECDA, but
they should not be regarded as absolute criteria for evaluating an ECDA for a number of
reasons. First, there is bound to be scatter in the data such as the number of false negatives
that would be obtained from different sets of random excavations. Second, the criteria used for
defining a false negative could vary from one operator to another. Some might consider finding
minor metal loss at a location that had no indications from indirect measurements as not being
significant and thus not a false negative. Third, different operators are likely to view indications

FINAL
15-019
from indirect measurements differently because of the vague priority ranking criteria in NACE
SP0502. The main value of pursuing such calculations is that it encourages an operator to
make excavations at randomly selected locations where the indirect measurements show no
indications. Conducting such excavations should be considered as a legitimate part of the
expense associated with using ECDA as an integrity assessment method.
ICDA
Processes for conducting internal corrosion direct assessment (ICDA) are described in several
NACE standard practices. These processes can be used to address the threat of internal
corrosion to pipeline integrity. Mitigation of the threat consists of repairing pipe that is found to
contain integrity-threatening amounts of internal corrosion-caused metal loss, instituting or
enhancing preventative measures, and re-assessing integrity at appropriate intervals.
The ICDA standard applicable to natural gas transmission pipelines carrying “dry” gas is NACE
02087 Internal Corrosion Direct Assessment of Gas Transmission Pipelines. This standard is
applicable to natural gas transmission lines that normally carry dry gas but may incur short term
exposure to wet gas or liquid water as the result of upset conditions. The processes described
in the standard involve detailed examinations of locations along a pipeline where an electrolyte
such as water could accumulate. Fluid flow models and the actual profile of a pipeline are used
to determine the most likely locations for water to accumulate. The condition of the remainder
of the pipeline is then inferred from the nature of the metal loss at areas where water could
accumulate. If the metal loss is severe, the operator may decide to investigate other areas of
the pipeline. Repairs are made where necessary, the observed corrosion rates are used to
determine when re-assessment is necessary, and the results of the ICDA are used to assess
the effectiveness of the process.
The ICDA process applicable to natural gas transmission pipelines carrying “wet” gas is NACE
SP0110-2010, Wet Gas Internal Corrosion Direct Assessment Methodology for Pipelines. This
standard is to be used for natural gas transmission pipelines that carry natural gas containing
water or hydrocarbon condensate.
The following two documents address conducting ICDA for hazardous liquid transmission
pipelines. For liquid pipelines there is the added risk of internal corrosion being facilitated by
sludge or other solids accumulating in low-flow locations in addition to water.
• NACE SP0208-2008, Internal Corrosion Direct Assessment Methodology for Liquid

Petroleum Pipelines
• NACE 07169 Internal Corrosion Direct Assessment Methodology for Liquid Petroleum
Pipelines

FINAL
15-019
As in the case of ECDA, one can use numerical calculation methods such as Bayes Theorem to
assess the effectiveness of ICDA. The main requirement for such calculations is to conduct
significant numbers of direct examinations of a segment at randomly selected locations where
no indications from flow calculations or elevation changes would suggest the occurrence of
internal corrosion.
SCCDA
NACE SP0204-2008 (formerly RP0204), Stress Corrosion Cracking (SCC) Direct Assessment
Methodology presents guidelines for assessing the threat of stress corrosion cracking to pipeline
integrity. Guidance is given for identifying segments of pipelines that may be susceptible to
SCC, for selecting sites for excavating and examining locations along any given segment of a
pipeline, for inspecting pipe after excavating it, and for the type of data that is to be collected
at each site. Guidance is also given for mitigating SCC, for establishing re-assessment intervals,
and for evaluating the effectiveness of the process.
Guidance for dealing with SCC may also be obtained from Integrity Management of Stress
Corrosion Cracking in Gas Pipeline High Consequence Areas STP-PT-011 – 2008” and ASME
B31.8S.
As in the case of ECDA, one can use numerical calculations methods such as Bayes Theorem to
assess the effectiveness of SCCDA. The main requirement for such calculations is to conduct
significant numbers of direct examinations of a segment at randomly selected locations where
no occurrences of SCC would be expected.
Mitigative Response for the HSC Threat to a Hard Heat-Affected Zone

On the basis of the known instances of failure, the threat of HSC developing in hard heat-
affected zones of ERW pipe appears to be confined to X-Grades of ERW pipe made between
1947 and 1960 by one manufacturer, Youngstown Sheet and Tube Company. Three such HSC
failures were documented in Youngstown pipe by Clark, et. al. xlvi, and one of the authors of this
guideline document has been involved in at least three investigations of failures from HSC in
hard heat-affected zones, all involving ERW pipe made by Youngstown during the late 1940s.
It is the experience of one of the authors of this guideline document that some but not all
pieces of ERW pipe made by Youngstown during this period have the requisite high hardness
level (at least Rockwell C35) for HSC to occur. Youngstown ERW pipe was made piece-by-piece
from individual plates formed into cans for welding rather than continually from coil as with
most other ERW manufacturers at the time. That means that each piece of Youngstown ERW
pipe could have had a unique alloy content and post-weld heat treatment. Apparently, some
pieces ended up with the requisite high hardness adjacent to the seam to be susceptible to
HSC, while others ended up with hardness levels not susceptible to HSC. One can only verify

FINAL
15-019
the hardness of the narrow heat-affected zone of any given piece of pipe by means of taking
microhardness measurements on a metallographic section across the seam, an impractical
means of locating the susceptible pieces during manufacturing let alone in an existing pipeline.
As has been noted previously herein, hydrostatic testing is not effective as a means of
mitigating HSC, because neither the tendency for HSC to develop nor the rate of growth of HSC
is predictable. Moreover, there is no known ILI technology that can reliably locate those pieces
of pipe that might have a heat-affected-zone hardness sufficiently high to be susceptible to
HSC. Therefore, mitigation of the HSC threat with regard to hard heat-affected zones consists
of trying to minimize the generation of atomic hydrogen from cathodic protection. Atomic
hydrogen can be generated at the surface of the pipe at an area of disbonded coating at a
potential level of -0.85 volts relative to a copper, copper sulfate half-cell, the level considered
by some to be sufficient to mitigate significant corrosion. The corrosion mitigative effect of
cathodic protection increases as the potential level becomes more negative than -0.85, so it is a
practical necessity to have some areas of a pipeline experiencing levels more negative than -
0.85 to assure that all portions of the pipeline are experiencing the -0.85 level. As shown in
Reference II, the amount of hydrogen generated increases exponentially as the potential level
becomes more negative than -0.85. In some of the cases where HSC failures have occurred, it
is known that the pipe-to-soil potential levels ranged from -1.3 to more negative levels. Thus,
controlling the pipe-to-soil potential levels to values not greatly exceeding the level required for
mitigation of corrosion is a means (and at this point in time, the only known means) of
mitigating the threat of HSC failures in a pipeline comprised of Youngstown ERW pipe made
between 1947 and 1960.
Mitigative Responses for Gas Distribution Systems

The substantial differences between natural gas distribution systems and natural gas
transmission pipelines make it impractical simply to apply the integrity management
requirements for transmission pipelines to distribution systems xlvii. In particular, the
characteristics of most gas distribution systems (i.e., low operating pressures, sharp bends,
numerous branch lines, multiple types of pipe materials and diameters, and direct connectivity
to end users) preclude the use of pressure testing (other than in-service leak testing), ILI, and
direct assessment for assessing pipe integrity. Furthermore, the significant diversity of the
pipeline systems within the gas distribution industry renders the concept of establishing
standard requirements for all integrity management circumstances impractical. Representatives
that comprised the joint work/study groups that performed the research for the Integrity
Management for Gas Distribution, Report of Phase 1 Investigations, concluded in December
2005, that the current pipeline safety regulations (49 CFR Part 192) did not convey the concept
of a risk-based distribution integrity management process, and it would be appropriate to
modify the regulations to include such a process for natural gas distribution systems.

FINAL
15-019
Therefore, the joint work/study groups determined that it would be appropriate to require all
distribution pipeline operators, regardless of size, to implement an integrity management
program. In 2009, The Gas Piping Technology Committee (GPTC) released the Distribution
Integrity Management Program (DIMP) guide xlviii with material intended to assist operators with
development of a DIMP, and guidance on compliance with Federal Regulations contained in 49
CFR Part 192, Subpart P xlix. On August 2, 2011, in order to comply with Subpart P, gas
distribution operators were required to develop and implement an integrity management
program that included a written integrity management plan. This plan was required to contain
procedures for developing and implementing the seven elements identified as the essential
components of a DIMP. Collectively, these elements establish a program for distribution
operators that should reasonably manage the integrity of their pipeline systems into the future.
The seven elements of DIMP 24 are:
1. Knowledge
2. Identify threats
3. Evaluate and rank risk
4. Identify and implement measures to address risks
5. Measure performance, monitor results, and evaluate effectiveness
6. Periodic evaluation and improvement
7. Report results
While the GTPC guide does not imply the order in which these elements should be applied, it
does stipulate that the operator needs to address each element in some way. The guide is
intended to be just that, a guide. The authors of the guide caution that it does not anticipate
all conditions that may be encountered, and it also does not restrict the operator from using
other methods to comply with the DIMP regulations.
This report concentrates on the mitigative responses discussed in Section 6 of the DIMP guide,
“Identify and implement measures to address risks”. It will be clear that the mitigative
responses previously discussed in relation to transmission pipeline systems (hydrostatic testing,
ILI, and direct assessment) are not a part of the mitigative responses in the DIMP guide
because they are impractical for distribution systems. ANSI/GPTC Z380, Guide for Gas
Transmission and Distribution Piping Systems, Appendix G-192-8, Table 6.1 – Examples of
Additional or Accelerated (A/A) Actions (occasionally referred to by PHMSA inspectors as risk
reduction measures) provides guidance on possible A/A Actions, which can be taken to reduce
the risks associated with distribution pipeline facilities, however the ultimate decision of which
24
State regulations may have additional DIMP requirements more stringent than Federal DIMP codes in § 192.1007.

FINAL
15-019
A/A Action(s) is/are most practical, reasonable, and economically feasible is left up to the
operator. The operator may choose whether to implement some, all, or none of these actions.
Some of the more common A/A Actions utilized by distribution system operators to address
specific threats are discussed hereafter.
A/A (Additional or accelerated) Actions (Risk Mitigation Measures) for

Specific Threats from Table 6.1 of ANSI/GPTC Guidance Appendix G-
192-8
As noted earlier in the report, the threats to the integrity of a natural gas distribution system
that could possibly be addressed by replacing a segment of pipe are external corrosion, internal
corrosion, natural forces, and material or weld failure. Accordingly, the accelerated actions that
an operator might take to mitigate these threats in order to delay or avoid replacement are
discussed below.
External Corrosion
The A/A Actions discussed in this section pertain to bare steel pipe with or without cathodic
protection (CP), wrapped steel pipe with or without CP, and Cast Iron pipe with external
corrosion or graphitization issues.
Increase frequency of patrolling and leak surveys
Requirements for patrolling distribution systems and conducting leak surveys are established in
49 CFR 192.721 and 192.723. In accordance with 49 CFR 192.721(b)(1) and 192.721(b)(2),
mains in places or on structures where anticipated physical movement or external loading could
cause failure or leakage must be patrolled, in business districts, at intervals not exceeding 4-1⁄2
months, but at least four times each calendar year; and outside business districts, at intervals
not exceeding 7-1⁄2 months, but at least twice each calendar year. Areas where physical
movement or external loading is known to exist, such as known unstable bridge crossings or
areas prone to landslides, should be patrolled in accordance with these intervals. Similarly,
leakage surveys using detector equipment must be conducted in business districts, at intervals
not exceeding 15 months, but at least once each calendar year to remain in compliance with 49
CFR 192.723(1), while outside business districts, they must be conducted as frequently as
necessary, but at least once every 5 calendar years at intervals not exceeding 63 months to
remain in compliance with 49 CFR 192.723(2). However, for cathodically unprotected
distribution lines (e.g. bare steel) subject to 49 CFR 192.465(e) on which electrical surveys for
corrosion are impractical, a leakage survey must be conducted at least once every 3 calendar
years at intervals not exceeding 39 months.

FINAL
15-019
Distribution operators should maintain records of all leaks experienced on their systems to allow
leak trends to be monitored. This enables the operator to determine whether the leak issue is
improving, worsening, or remaining the same. If the operator continues to experience an
increased number of leaks per mile or leaks per thousand services, then another action should
be implemented.
Replace, insert or rehab
Three other mitigative responses available to distribution pipeline operators are to replace the
line, insert new pipe through the line (i.e. push or pull a new, smaller-diameter pipe through an
existing, larger-diameter pipe), or rehab (i.e. repair) the line. These mitigative responses can
be applied along the entirety of the line or only on specific portions of the line. Plastic pipe
inserted through a steel main is an example of an insert type of replacement where the new,
smaller-diameter pipe becomes the carrier for the gas, and the existing, larger-diameter pipe is
abandoned. An operator may continue to rehab segments of pipe or insert new segments of
pipe in problematic areas, as necessary, recognizing that at some point it might be more
economically feasible to just replace the entire line.
Relocate
In instances along a pipeline, where active corrosion is a major issue, correcting cathodic
protection issues is often the most feasible mitigative response. However, relocation may be
appropriate under certain circumstances, as a last resort, where corrosion persists.
Provide hot spot protection (e.g., install anodes at anodic locations)
In areas where cathodic protection is used to mitigate corrosion, operators should monitor
trends in the number of sacrificial anodes installed and the ability to maintain adequate cathodic
protection levels. A cost-benefit type analysis could aid the operator in deciding whether
continuing to replace anodes, adding additional anodes, relocation or replacement is the most
appropriate mitigative response.
Correct cathodic protection deficiencies
In accordance with 49 CFR 192.465(a), which covers pipelines under cathodic protection, the
levels of protection must generally be assessed at least once each calendar year, but with
intervals not exceeding 15 months, to determine whether cathodic protection levels meet the
requirements of 192.463. Operators shall take prompt remedial action to correct any cathodic
protection deficiencies indicated by monitoring as required by 49 CFR 192.465(d). Other
requirements for controlling external corrosion using cathodic protection are given in 49 CFR
192.463 to 192.473.

FINAL
15-019
Internal Corrosion
Internal corrosion in a distribution system can arise in conjunction with condensation and water
accumulation. When possible, the operator should inspect the interior surfaces of the piping for
signs of internal corrosion and/or water accumulation. Water should be drained from the
system, if found, and the segment should be cleaned out, if possible. If pitting is discovered,
the severity needs to be assessed, and if leakage is imminent, it should be managed in an
acceptable manner. If the product transported is corrosive, the operator would need to take
action to manage the leaks along the line and to monitor the leak trends and institute
appropriate mitigative action to reduce leakage.
Install pipe liner
Tight-fit (cured in-place 25) type pipe liners are capable of being installed in cast iron pipelines as
a mitigative response to leakage due to internal corrosion. When liners are properly installed,
the threat of leakage to the pipeline is significantly reduced. This mitigation is only applicable if
the existing pipe structure is strong enough to handle the stresses experienced by the line at
the maximum operating pressure as the liner will not add any structural integrity to the line. CIP
liners are also used as a repair techniques for external localized graphitization or to address
bell/spigot joint leaks (typically due to internal drying from ultra dry and LNG gas).
Install moisture removal/control equipment and evaluate gas supply inputs / take
corrective action with supplier
Moisture removal or control equipment, such as dehydration units, can be installed at pipeline
gate stations to remove excess water from the gas entering the distribution system. While
dehydration units are a viable option for moisture removal or control, the more common
practice among distribution pipeline operators is to monitor the gas entering the system for
quality issues such as H2S, CO2, and water, and take corrective action with the supplier, as
necessary. Situations such as this are especially relevant in areas receiving natural gas from
local shale areas where gas may not have been treated sufficiently or might be different from
the natural gas the operator received previously from long haul pipelines.
Natural forces
Threats due to natural forces require A/A actions to counteract outside force/weather issues
(e.g., earth movement, lightning, heavy rains/floods, temperature extremes, high winds)
25
Operators should check with the applicable governing pipeline codes (Federal and State) to be sure cured in-place liners meet
pipeline regulations for hazardous liquid or natural gas pipelines.

FINAL
15-019
affecting a pipeline. These threats affect all types of pipe materials including steel, plastic, and
cast iron. It is important to remember that these types of threats are situational (i.e. they all
don’t occur everywhere) and if the threat to the pipeline is local then the mitigative response
will also remain local.
Relocate pipe from high risk locations
Relocation can be an effective form of mitigation to avoid high risk locations. For example, if a
pipeline is currently running along a hillside with a history of soil erosion issues and landslides,
the risk of these possible outside forces causing a failure in the line may outweigh the benefits,
thus leading to the decision to relocate the pipeline out of this area. The economics of the
situation along with the risks involved will drive the decision of whether relocation is a viable
option.
Replace pipe in high risk locations
For many reasons, relocating a pipeline may not be possible, thus leaving the operator with the
task of finding another way to protect the pipeline in the high risk location. Replacement could
prove to be a worthwhile option to ensure the safety and integrity of the line, in which the
operator could replace the segment of line in the high risk location with a pipeline designed to
more suitably withstand the threats posed by natural forces in the area.
Install slip or expansion joints for earth movement
In areas where earth movement is an issue and relocating the pipe is not an option, it may be
necessary to modify the existing pipeline, so as to allow slip or expansion joints to be installed
to help the pipeline accommodate some of the earth movement which is known to occur in the
area. Installing these components will help to mitigate at least some of the risk associated with
this type of threat.
Install strain gauges on pipe
In known high risk locations where soil movement is prevalent, such as with mine subsidence,
earthquakes, soil erosion or washouts, soil slumping (on hill sides), or sink holes, installing
strain gauges along the pipeline allows operators to monitor their risk based on the stress and
strain the pipe is experiencing. Monitoring natural forces in this manner allows operators to set
stress and strain limits as a method of determining when to intervene and physically manipulate
the line so as to reduce the stress/strain it is experiencing.

FINAL
15-019
Expand the use of excess flow valves
Excess flow valves (EFV) are small, relatively inexpensive devices that can be installed on
distribution service lines usually in or near the service tee connection to a main. They are
manufactured to activate based on a desired flow rate (at a given system pressure). For
example, such a valve might be manufactured to close at a flow rate of 500 SCF per hour at a
pressure of 10 psig. At 60 psig operating pressure, the closure rate would be much higher.
The closure rates of EFV’s are not adjustable.
Installing EFV’s on high risk service lines could provide a safeguard to prevent a hazardous
situation should excavation damage, a severe leak, or an outside force incident occur
downstream of the valve. These valves are very effective at shutting off the flow of gas
through a service line if damage should occur to the line, and they won’t allow gas to flow again
until the damaged line is repaired. This prevents a large influx of gas from possibly entering an
end user’s residence or place of business should some type of damage to the distribution line
occur downstream of the valve. Other benefits of installing this type of valve are that it allows
first responder crews more time to evacuate the buildings or areas affected by a damaged
distribution pipeline as well as preventing a large vapor cloud from quickly forming and possibly
finding an ignition source.
Conduct leak survey after every significant earthquake or other event
Natural forces can create risk on a pipeline system, especially lines in unstable soil locations. It
is good practice to perform leak surveys after any natural outside force event, such as a
significant earthquake, a winter storm, repeated freeze/thaw occurrences (especially with
respect to cast iron pipe), or any other type of severe weather event to confirm the event did
not cause an integrity issue with the pipeline.
Material or weld failure
Threats due to material or weld failure may require A/A actions to respond to manufacturing
defects and construction/workmanship defects that could affect the pipeline. These threats
could potentially affect all types of pipe materials including steel, plastic, and associated piping
components. It is important to keep in mind that these types of threats have the potential to
exist anywhere in the pipeline, and if found, then trending the pipe material, manufacturer,
installation date, installation contractor, and other pipe properties should be completed in an
attempt to determine whether the issue is local or systemic.
Some of the A/A actions applicable to material or weld failures were already discussed in
previous sections of this report (i.e. the External or Internal Corrosion, or Natural Forces
sections). To avoid being repetitive, refer to the aforementioned sections for details on the

FINAL
15-019
following A/A actions: increasing the frequency of leak surveys, repairing the line, or replacing
the line. Each of these A/A actions is also acceptable to utilize with material or weld failure type
threats. Other A/A actions, which can be utilized in mitigating material or weld failure threats
are discussed in more detail below.
Revise construction procedures
If the evaluation of a failure indicates that it was caused by construction or workmanship

defects, then it may be prudent to review the construction procedures to which company and
contractor personnel must adhere, in order to determine whether modifications or changes to
these procedures are necessary. If the outcome of this review indicates the existing procedures
are adequate, then the issue might involve the construction oversight or the company or
contractor personnel actually performing the work. If this is the case, site inspectors
overseeing construction should receive further training to confirm their technical understanding
of the company construction procedures, while historical construction records should be
reviewed to investigate whether the company or contractor personnel that completed the work
has any instances of poor performance in the past. Keeping historical records of site inspector
and construction personnel performance will allow the pipeline operator to only utilize high
performing personnel on future projects.
Revise material standards
If a failure is determined to have been caused by a material problem, then it might be

necessary to revisit the material standards to which the pipeline was designed to verify the
standards align with prudent engineering and/or generally accepted industry practice. If the
material standards are verified to align with industry practice and specify appropriate materials,
then some trending may be necessary to determine whether the material problem is
widespread throughout the system or whether it is isolated to only certain locations in the
operator’s system. Trending is discussed further in the next section and can be applied when
investigating issues such as described above.
Trend material failures
Trending material failures in a distribution system can provide useful information regarding
whether or not a facility or group of facilities with common traits, needs to be replaced. Any
and all aspects of the material failures experienced can be trended to gain information or find a
pattern as to what happened, where it happened, if it will occur anywhere else, why it
happened, and problematic areas to monitor for the same type of issue going forward. While
there are many properties to trend, a few examples are: material manufacturer, material type,
installation year, installation personnel, location or region, MAOP or operating stress, etc. When

FINAL
15-019
trending, it is important to keep in mind that multiple types of material failures might be
occurring. So, if no trends are found while comparing all locations where material failures
occurred, then an operator should try categorizing the failures by another property or only
comparing a certain subset of the locations where failures occurred to see if a trend becomes
apparent. The operator can then utilize the results of the trending analysis to determine
whether repair or replacement is the more appropriate response for each specific situation.
C RITERIA FOR M AKING R EPAIR /R EPLACE D ECISIONS

Systematic approaches for making repair/replace decisions are presented below for four classes
of pipelines.
1. Hazardous liquid pipelines

2. Natural gas transmission pipelines operated at stress levels of 30% of SMYS or more.
3. Natural gas transmission pipelines operated at stress levels less than 30% of SMYS
4. Natural gas distribution systems
Hazardous liquids are considered in one class regardless of operating stress level while natural
gas transmission lines are divided into “high” stress and “low” stress classes. There are two
reasons why gas pipelines but not liquid pipelines were segregated by the <30% stress
level. The first has to do with how the gas and liquid regulations differ. Part 192, Subpart J
separates pressure testing requirements for pipelines operated at stress levels below 30% of
SMYS from those for pipelines operated at stress levels of 30% of SMYS or more. Also, miter
bends are permitted in gas pipelines that are operated at stress levels below 30% of SMYS but
not in pipelines operated at stress levels of 30% of SMYS or more. Part 195 recognizes “low
stress” pipelines as those that are operated at stress levels of 20% of SMYS or less, but only
those low stress pipelines with very low risk are exempt from pressure testing or any other
requirements of Part 195.
The second reason why gas pipelines but not liquid pipelines were segregated by the <30%
stress level has to do with fracture propagation risk, service interruption, and piggability. Gas
pipelines are subject to long-running fractures when a failure occurs. This risk is considerably
lower (but not negligible) for gas pipelines operated at stress levels below 30% of SMYS. The
types of gas pipelines generally operated at less than 30% of SMYS are single-feed systems for
a local gas distribution operator. Taking such a pipeline out of service for a pressure test is
usually not possible. And, such pipelines generally are not piggable because of the low
operating pressures and the flow rates of the gas. In contrast, while stress level does influence
whether or not a liquid pipeline will leak or rupture when it fails, the length of any rupture that
may occur (except for HVL pipelines) is generally insensitive to the operating stress level. Also,
a low stress level in a liquid pipeline seems to present no barrier to pigging. Lastly, leaks and

FINAL
15-019
spills from liquid lines can have severe impact on the environment regardless of the operating
stress level of the pipeline.
The process for making repair/replace decisions in each case is based on a series of flow charts.
It will be seen that these flowcharts constitute prescriptive procedures for making repair/replace
decisions. These procedures are not meant to prevent a pipeline operator from conducting an
engineering critical assessment of a pipeline that would allow the decision to be made outside
of these prescriptive procedures.
Flowcharts for Transmission Pipelines

The flowcharts for making repair/replace decisions for transmission pipelines are presented in
Appendix C, Figures C1 through C15. These figures cover all three types of transmission
pipelines: hazardous liquid pipelines, natural gas transmission pipelines operated at stress levels
of 30% of SMYS or more, and, natural gas transmission pipelines operated at stress levels less
than 30% of SMYS.
Pipeline or System Attributes, Operational Data, and History

The pipeline attributes that are useful for making repair/replace decisions for hazardous liquid
pipelines and natural gas transmission pipelines were presented previously in Tables 4 and 5 for
the “base case” pipelines. The various operational data and historical operating experience that
bear on repair/replace decision-making were discussed previously. The decision-making criteria
are demonstrated below by means of examples. For each case discussed below complete sets
of attributes, operational data, and historical operating experience are provided at the
beginning of the discussion of each case. Note that no examples of pressure cycle data are
provided because no actual fatigue analyses are carried out.
Base Case Example No. 1: Hazardous Liquid Pipeline

Comprised of LF-ERW Pipe
Consider the 900-mile-long crude oil pipeline described in Table 4. It is comprised of 16-inch-
OD, 0.250-inch-wall, X52 pipe manufactured with a LF-ERW seam and installed in 1953.
Assume that the operator has records documenting all aspects of the pipeline’s design,
construction, and operation since 1953. The pipeline can accommodate all types of ILI tools.
The failure history indicates one in-service rupture from external corrosion in 1984 and several
in-service leaks and ruptures arising from manufacturing defects in the LF-ERW seam. The
causes of the earlier seam failures are unknown because they were not thoroughly investigated.
However, the later ones were found to have been caused by PCIF.
The operator has begun to employ periodic hydrostatic testing to assess the integrity of the
pipeline from the standpoint of PCIF. The retest interval has been established via the type of

FINAL
15-019
fatigue analysis explained earlier in the report. Records are available that show that early tests
resulted in failures from external corrosion-caused metal loss. However, the more recent tests
have produced test failures caused by seam manufacturing defects some of which had exhibited
fatigue crack growth. Lastly, the operator has begun to utilize ILI periodically both to locate
dents and to find and remediate external corrosion-caused metal loss. The re-inspection
interval for corrosion assessment was established based on an assumed rate of corrosion, and
the absence of corrosion-caused failures since the start of the assessments suggests that the
process is working.
To determine whether or not this pipeline should be replaced, the operator can go through the
process described below starting with Flowchart START (Figure C1 of Appendix C). Flowchart
START as adapted to this particular pipeline is repeated as Figure 7 below.
Are the essential attributes

of the segment (i.e.,
diameter, wall thickness, no Can attributes be established no Schedule replacement
grade of material, seam type, or verified?
design factor, and operating
temperature) known? yes
Establish or verify essential

yes attributes.
Does segment contain no

Proceed to
legacy pipe ?
MODERN PIPE
yes
Proceed to LEGACY PIPE
Figure 7. Start of Process to Decide Whether to Repair or Replace Base Case

Hazardous Liquid Pipeline
The first question to answer in the process is: Are the essential design attributes of segment
known? The answer based on the data in Table 4 is “yes”. If the answer had been no, the
operator would have to either establish or verify the essential design attributes or schedule
replacement. The practical constraints on operating any pipeline safely mean that diameter,
nominal wall thickness, grade of material (specified minimum yield strength or SMYS), type of
longitudinal seam (to establish the “joint factor”), design factor (upper bound on operating
hoop stress as a percent of SMYS) , and operating temperature are essential design attributes.
These should be readily available from manufacturer’s certificates, design calculations, and
alignment sheets if such records have been retained in the operator’s archives.

FINAL
15-019
But, what if some or all of these essential attributes are not known? For a hazardous liquid
pipeline, processes for establishing nominal wall thickness, grade of material, and joint factor
when traceable, verifiable and complete records are not available are given in Part 195,
§195.106. In the process of acquiring the samples to determining an unknown yield strength
as required by §195.106, the operator should be able to establish the type of longitudinal seam.
According to ASME B31.4, the standard design formula (the same as the one mandated by Part
195) is applicable as long as the operating temperature is between -20⁰F and 250⁰F. Diameter
should be observable in above-ground portions of the pipeline, and it can be verified continually
along the pipeline if the pipeline can accommodate running a caliper tool. Nominal wall
thickness can be determined as specified in §195.106. Alternatively, wall thicknesses along the
pipeline can be verified via in-line inspection with currently available ILI metal loss tools. Lastly,
though it is not a part of federal regulations, it would seem that an operator could use the
results of a recent hydrostatic test or could conduct a hydrostatic test to verify the integrity of
the pipeline and establish its minimum pressure-carrying capacity. As in-line tools evolve, it
may become possible to verify yield strengths and seam types along a pipeline as well as
diameters and wall thicknesses.
The second question is: Does segment contain legacy pipe? The segment is comprised of LF-
ERW pipe, so the answer is “yes”. The operator must proceed to LEGACY PIPE. If the answer
had been no, the operator would have been directed to the flowchart shown in Figure C8 in
Appendix C (MODERN PIPE). The LEGACY PIPE flowchart as adapted to this particular pipeline
is shown in Figure 8.

FINAL
15-019
Legacy Pipe
Is it feasible to conduct a pressure test
Does the segment qualify for one of no no
of the segment in accordance with to
the alternatives or exceptions listed establish the operating pressure in
Has the segment been no in Part 192 or Part 195? accordance with Part 192 or Part 195?
subjected to a pressure test
in accordance with Part 192 yes
or Part 195? yes
Conduct a pressure test of the segment or lower the
yes
operating pressure to meet the requirements of Part 192 or
Part 195.
Schedule Replacement
Evaluate the threats to a legacy pipeline
1. Segment contains LF-ERW pipe, DC-ERW Proceed to LF-

pipe or flash-welded pipe and may be ERW
susceptible to PCIF and SSWC.
6. Segment may be susceptible

2. Segment contains Proceed to SSAW Proceed to EC
to external corrosion
SSAW pipe
3. Segment contains Proceed to LAP- 7. Segment may be susceptible Proceed to IC

furnace lap-welded pipe WELDED to internal corrosion
4. Segment contains miter bends,

8. Segment may be Proceed to SCC
wrinkle bends, non-standard Proceed to susceptible to SCC
fittings, Dresser couplings, threaded LEGACY FEATURES
couplings, or acetylene girth welds
9. Segment may be Proceed to HSC
susceptible to HSC
5. Segment contains
Proceed to
puddle welds, patches,
LEGACY REPAIRS
or half soles
Figure 8. Threats to Address for Base Case Hazardous Liquid Pipeline Comprised of
LF-ERW Pipe
According to Figure 8, the operator must review the pressure test history of the pipeline. If the
pipeline has not been tested in accordance with the requirements of Part 195, and it does not
qualify for one of the exceptions in the regulations, the operator must either conduct the
appropriate test or lower the MOP to a level that would bring the pipeline into compliance with
the regulations. If neither of these can be done, the operator should schedule replacement of
the pipeline. This particular pipeline has been tested adequately.
This pipeline is comprised of LF-ERW pipe which can be susceptible to seam defects that may
leak or become enlarged in service from PCIF and may also be susceptible to SSWC. In fact the
history of the pipeline shows that it is susceptible to degradation from PCIF and external
corrosion. Because of its operating stress level, it is potentially susceptible to SCC. Therefore,
Cells 1, 7, and 9 are highlighted in Figure 8. The pipeline contains no SSAW pipe, no furnace
lap-welded pipe, no miter bends, no wrinkle bends, no non-standard fittings, no Dresser
couplings, no threaded couplings, no acetylene girth welds, no puddle welds, no patches, and
no half-soles. Internal corrosion has never been detected via ILI or noticed at any time

FINAL
15-019
segments of the pipe have been removed, and the pipeline carries crude oil with no sour
components. ERW pipe is not prone to pipe body hard spots, and since this pipe was not
manufactured by Youngstown Sheet and Tube Company it is not likely to have heat-affected-
zone hardnesses of Rockwell C35 or more. Therefore, the threats represented by Cells 2-6 and
8 and 10 need not be considered. Based on the highlighted cells, the operator of this pipeline
must pursue the process of making a repair/replace decision via Flowcharts LF-ERW, EC, and
SCC (Figures C3, C12, and C14, respectively, of Appendix C). The first of these, Flowchart LF-
ERW, as adapted to this particular pipeline is repeated as Figures 9A and 9B below.
Segment contains LF-ERW

pipe, DC-ERW pipe, or flash- PART A
welded pipe.
Use pressure cycles to calculate Estimate rate of SSWC or rely on

times to failure from PCIF or rely on service and test failure history to
service and test failure history to establish re-assessment interval to
establish re-assessment interval to prevent service failures from SSWC.
prevent service failures from PCIF.
Go to
Is periodically assessing seam no Part B
integrity by means of hydrostatic
testing or ILI. feasible?
yes
Assess seam integrity periodically Assess seam integrity periodically Go to

based on the established interval based on the established interval to Part B
to prevent service failures from prevent service failures from SSWC.
PCIF.
Is it feasible to continue to
Are service failures from PCIF decrease the re-assessment
or SSWC occurring between yes intervals or to decrease the no
re-assessments? operating stress until service
failures from PCIF or SSWC cease
to occur?
no
yes
Continue to operate and maintain Re-assess seam integrity at

in accordance with Part 192 or decreased interval or lower
Part 195 and continue to perform operating stress until service
seam integrity assessments on failures from PCIF or SSWC cease
the same schedule. to occur.
Figure 9A. Feasibility of Preventing in-Service Failures from PCIF or SSWC by

Hydrostatic Testing or ILI

FINAL
15-019
PART B
yes Rely on ECDA to Are failures from no

is the maximum operating
locate and repair SSWC occurring in
stress <30% of SMYS?
SSWC. spite of ECDA?
no
yes
Is it feasible to lower the
operating stress to a level yes Schedule replacement or
that prevents seam failures in rehabilitation based on
service? half the calculated time
to failure.
no Lower operating stress as
needed to prevent seam
failures in service and Continue to operate in
continue to operate in accordance with Part
accordance with Part 192 or 192 or Part 195
Schedule replacement
of pipeline or Part 195.
rehabilitate the affected
segments.
Figure 9B. Feasibility of Preventing in-Service Failures from PCIF or SSWC by

Lowering Operating Pressure
Because the pipeline in question is comprised of LF-ERW pipe, the operator must assure its
continued seam integrity by taking the necessary steps to prevent in-service leaks or ruptures
from PCIF and selective seam weld corrosion (SSWC). As shown in Figure 9A, periodic seam
integrity assessment via hydrostatic testing is being used to address PCIF, and it appears that
the retest interval is sufficiently short to prevent in-service failures from PCIF. SSWC may be a
threat for this pipeline even though no test failures have been caused by SSWC. The periodic
testing to assess for PCIF also would be expected to reveal whether or not any serious SSWC
has occurred. Therefore, the operator can continue to operate the pipeline employing the
hydrostatic retesting on the same schedule.
Also as indicated in Figure 9A, the operator would have the option of employing ILI crack-
detection technology in place of hydrostatic testing to address both PCIF and SSWC.
If the operator cannot or will not employ either hydrostatic testing or ILI to periodically re-
assess the integrity of the ERW seam, lowering the operating pressure would be an option as
shown in Figure 9B. However, it would be incumbent upon the operator to show that in-service
failures from PCIF or SSWC would not be likely to occur at the lower MOP. If it is not possible

FINAL
15-019
to have confidence that no such failures will occur, then replacement of the segment should be
scheduled. Note that it may be possible to replace only those segments that are affected by a
threat (or threats) that makes unacceptable the cost of continuing to repair. This could involve
rehabilitation or pipe replacement at the problematic locations instead of replacement of the
whole pipeline.
If in-service leaks or ruptures caused by either PCIF or SSWC continue to occur in spite of either
a hydrostatic testing or ILI-based re-assessment program, the operator must either decrease
the interval between assessments to the point where such failures cease to occur or schedule
replacement of the pipeline or rehabilitation of the affected segments.
To keep the pipeline in service, the operator must satisfactorily address the threats of failure
from external corrosion and stress corrosion cracking (SCC) to the point where no in-service
failures are caused by either of these phenomena. Response to the threat of external corrosion
as adapted to this particular pipeline are addressed via Flowchart EC repeated below as Figure
10.

FINAL
15-019
Does segment have an no Is the CP adequate to no Schedule replacement.

external coating? mitigate corrosion?
yes yes no
Does the coating tend Can the pipeline Is it feasible to

to shield the pipe yes no mitigate corrosion yes
accomodate metal ∩
from CP when it loss ILI tools? using periodic
becomes disbonded? hydrostatic testing?
yes
no ∩
Can the pipeline Can metal loss be

accomodate metal no adequately monitored no
loss ILI tools? and mitigated using
ECDA?
yes yes
Use analysis of Is it feasible to conduct

corrosion rate and/or assessments and
responses at an interval no
history of corrosion
failures to establish that prevents leaks and
assessment interval. ruptures from corrosion?
yes
Conduct periodic
assessments or
hydrostatic tests at
appropriate intervals to
prevent leaks and
ruptures from external
corrosion and continue to
operate segment in
accordance with Part 192
or Part 195.
Figure 10. Feasibility of Preventing In-Service Failures from External Corrosion
Note that it is prudent to assume that any pipeline is susceptible to external corrosion, so the
question “Is the segment susceptible?” is not asked. In the case of this particular hazardous
liquid pipeline, the pipeline is coated and cathodically protected, but as its history shows, it has
sustained a previous rupture from external corrosion. The coating on the pipeline is of a type
that does not shield the pipe from cathodic protection if it disbonds. The pipeline can
accommodate ILI metal loss inspection tools. Such tools are used to inspect the pipeline at
intervals determined by analysis. Comparisons of successive inspections have shown the
intervals to be sufficiently short to assure the absence of in-service failures from external
corrosion. Therefore, replacement of the pipeline from the standpoint of the threat of external
corrosion does not have to be considered. Note that unburying the pipeline, lifting it out of the

FINAL
15-019
ditch, blasting off the old coating, repairing or removing corroded pipe, re-coating the pipeline,
and lowering and re-burying it could be done in place of total replacement.
If the pipeline cannot accommodate ILI metal-loss inspection tools, the operator could consider
addressing the corrosion threat by utilizing external corrosion direct assessment (ECDA). This is
possible because the coating on this pipeline does not shield the pipe from cathodic protection if
it becomes disbonded. As long as the ECDA program is successful in preventing in-service
failures from external corrosion, replacement from the standpoint of the threat of external
corrosion does not have to be considered.
If it is not feasible for the operator to use either ILI or ECDA, periodic hydrostatic testing could
be used to control the threat of external corrosion. However, this is generally an inefficient
means of assessing integrity from the standpoint of corrosion because of the inability of testing
to reveal short, deep defects. If hydrostatic testing is chosen as the means of assessment for
external corrosion, a close-interval pipe-to-soil potential survey done in conjunction with the
test may greatly improve the likelihood of finding the short, deep defects that survive the test.
Moreover, in cases where ECDA can be used effectively, ECDA combined with a hydrostatic test
would improve the likelihood of finding severe corrosion regardless of its configuration
The threat of SCC as adapted to this particular pipeline is addressed via Flowchart SCC repeated
below as Figure 11.

FINAL
15-019
no Continue to operate in
Is segment susceptible
accordance with Part 192 or
to SCC?
Part 195.
yes
Is the SCC dormant? yes

Schedule replacement.
no or unknown
no
Can SCC be adequately Is it feasible to prevent SCC

Can the pipeline
monitored and no no failures using periodic yes
accomodate ILI crack hydrostatic testing or by
mitigated using tools and is a crack
SCCDA? lowering the operating
tool appropriate?
pressure?
yes
yes
Is it feasible to conduct
Use analysis of SCC rate inspections and
and/or history of SCC responses at an interval no
failures to establish re- that prevents leaks and
assessment interval. ruptures from SCC?
yes Conduct ILI crack tool

inspections at
Monitor and mitigate SCC Use analysis of SCC rate appropropriate
using SCCDA and continue and/or history of SCC intervals to prevent SCC
to operate in accordance failures to establish re- failures and continue to
with Part 192 and Part assessment interval. operate segment in
195. accordance with Part
192 and Part 195.
Conduct hydrostatic tests at Use analysis of SCC rate

appropriate intervals to prevent SCC and/or history of SCC
failures, or lower the operating failures to establish re-
pressure to a level that prevents assessment interval or
failures from SCC and continue to consider lowering the
operate in accordance with Part 192 operating pressure to
and Part 195. mitigate SCC.
Figure 11. Feasibility of Preventing In-Service Failures from SCC

FINAL
15-019
Not every pipeline is susceptible to SCC, so the process begins with the question “Is the
segment susceptible to SCC?” 26. In the case of this particular pipeline the question cannot be
answered “no” because the operator does not know whether it is or not. The pipeline is
operated at a high enough stress level for SCC to be possible, and the coating is not fusion
bonded epoxy (which usually mitigates the SCC-forming environment). The answer to the
second question “Is the SCC dormant?” is “unknown”. In response to a potential SCC threat,
the operator has three choices:
1. Use SCCDA periodically to locate areas where SCC may be occurring.

2. Periodically conduct ILI with an appropriate crack-detection tool and repair cracks.
3. Periodically conduct hydrostatic tests.
The flowchart in Figure 11 runs through these choices in order, but the operator should make
the choice appropriate to the circumstances of the particular pipeline. For a pipeline where SCC
is known to exist because of past SCC failures, either ILI or hydrostatic testing would seem to
be the preferred options. On the other hand, if no evidence of SCC has surfaced, it could make
sense for the operator to choose SCCDA.
The operator of the Base Case Example No. 1 pipeline has seen no evidence that SCC is present
but prudently decides to use SCCDA to err on the safe side. At least, the operator will likely get
an early warning if SCC should develop. Besides, the hydrostatic testing that is necessary to
address the PCIF threat will likely expose any potential SCC threat in the form of a test failure
before the cracks could become large enough to fail in service. Alternatively, the operator may
be able to show by analysis using the Fessler-Rapp methodxxiii that the testing for PCIF
adequately assesses the integrity of the pipeline from the standpoint of SCC as well. In that
case SCCDA would not be needed.
If the pipeline were known to be affected by SCC, and SCCDA proved to be an unreliable way to
find and mitigate the phenomenon, the operator would have to rely on ILI crack detection
technology or periodic hydrostatic testing to mitigate the problem. Alternatively, it may be
possible to mitigate the SCC threat by lowering the operating pressure. If the operator were
able to control the problem (i.e., prevent leaks and ruptures from SCC), then the pipeline would
not need to be replaced.
26
One can refer to ASME B31.8S or API RP 1160 to access guidance on determining susceptibility of SCC or to Integrity
Management of Stress Corrosion Cracking in Gas Pipeline High Consequence Areas STP-PT-011 – 2008 to access guidance on
integrity management of SCC in general.

FINAL
15-019
The above analysis indicates that the threats to the integrity of this pipeline are being mitigated
by the on-going maintenance and integrity assessment actions being taken by the operator.
The question of whether or not the operator should continue to operate this pipeline versus
replacing it can be made on the basis of the cost to maintain and re-assess its integrity and
safety versus the cost to replace it. If these continuing costs are acceptable to the operator,
then replacement is not necessary.
Base Case Example No. 2: A Natural Gas Pipeline Comprised

of Flash-Welded Pipe Operated at a Stress Level of 72% of
SMYS
Consider the natural gas transmission pipeline described in Table 5. It is comprised of 30-inch-
OD, 0.375-inch-wall, X52 pipe manufactured with a flash-welded seam and installed in 1960.
The pipeline was subjected to a pre-service hydrostatic test to a minimum stress level of 90%
of SMYS. It is operated at a maximum hoop stress level of 72% of SMYS in Class 1 locations.
The operator has records documenting all aspects of the pipeline’s design, construction, and
operation since 1960, and the pipeline can accommodate all types of ILI tools. The failure
history indicates two in-service ruptures from external corrosion, one in 1971 and one in 1984.
In addition, an in-service rupture caused by selective seam weld corrosion (SSWC) occurred in
1995, and an in-service rupture arising from high-pH stress corrosion cracking (SCC) occurred in
1999.
The operator decided to employ periodic hydrostatic testing of the first two valve sections
downstream of the compressor stations to assess integrity from the standpoint of SCC after the
occurrence of an SCC in-service rupture in 1999. The first tests for that purpose were
conducted in 1999 to a maximum hoop stress level of 110% of SMYS. The operator assumed
that the rate of SCC growth could be as high as 0.024 inch per year, the rate suggested in
Reference XXIV. The failure-pressure-defect-size relationship shown in Figure 12 was used to
determine an appropriate interval of time before a second test should be conducted.

FINAL
15-019
D = 30.000", t = 0.375", SMYS = 52,000 psi, CVN = 40 ft-lb, CVN Area = 0.124
in^2
1,800
d/t
1,600
0.0
HTP
1,400 0.1
1,200 0.2
0.3
Pressure, psig
1,000 MOP
0.4
800 0.5
Rupture
600 0.6
0.7
400 Leak
0.8
200 0.9
0
0.0 5.0 10.0 15.0 20.0 25.0
Total Length, inch
Figure 12. Time to Failure at the MOP after a Test (HTP) to 110% of SMYS
The 20-inch-long defect in Figure 12 appears to have the shortest time to failure for long
defects. It would be expected to survive the test to 110% of SMYS if its depth did not exceed
8% of the wall thickness, and it would be expected to fail at the MAOP of 72% of SMYS if its
depth reached 40% of the wall thickness. The corresponding amount of defect growth is 0.32
times the wall thickness of 0.375 inch or 0.120 inch. At a growth rate of 0.024 inch per year,
the expected time to failure would be 5 years. The operator elects to conduct another
hydrostatic test of the same segments to the same test stress levels 2 years after the first test
(a factor of safety of 2.5).
In 2006 the publication of the Fessler/Rapp method for scheduling hydrostatic tests to mitigate
SCC was developed. The operator decides to base the timing of future hydrostatic tests on the
Fessler/Rapp method. Since the method is strongly influenced by the flow stress of the
material with higher flow stress levels giving more conservative results, the operator uses lists
of yield and ultimate tensile test results from the manufacturer of the order to calculate the
upper-95-percentile yield and tensile strengths, 60,000 psi and 85,000 psi, respectively. The
flow stress is the average of these two, or 72,500 psi. The resulting times for future retests are
based on Figure 13.

FINAL
15-019
Test time based on Fessler/Rapp
2000
1800
1600
1400 2 5 11 24 55 FSP
MOP
Presure, psig
1200 HTP
1000
800
600
400
First Test
200
0
0 10 20 30 40 50 60
Years until retest
Figure 13. Times for Retests to Control SCC based on the Fessler/Rapp Method
On the basis of the Fessler/Rapp method, the retests should occur at 2, 5, 11, 24, and 55 years
after the first test. Note that the first test was conducted in 1999 arbitrarily followed by a
second test in 2001. A third test was conducted in 2004. These tests though scheduled at
arbitrary intervals, coincidentally, would have been the ones predicted by the Fessler/Rapp
method, had it existed in 1999. Three test breaks from SCC occurred during the 1999 test, one
at a test stress level of 78% of SMYS. Two test breaks from SCC occurred during the 2001 test.
The lowest failure stress level for an SCC break was 99% of SMYS. In the 2004 test only one
test break from SCC occurred at a stress level of 101% of SMYS. The decreasing numbers of
test breaks and increasing levels of failure stress suggested that the SCC threat was being
adequately addressed by the successive hydrostatic tests. On the basis of the Fessler/Rapp
calculations another hydrostatic test was to have taken place in 2010. Instead, the operator
elected to run an EMAT ILI tool to assess the entire pipeline for SCC.
Hydrostatic test failures from causes other than SCC occurred during the 1999, 2001, and 2004
hydrostatic tests. Two test breaks during the 1999 test were caused by SSWC and 9 test
breaks were caused by seam splits. The lowest failure stress level associated with an SSWC
rupture was 81% of SMYS. Nine seam splits from manufacturing defects (confirmed to be hook
cracks with no evidence of in-service crack growth) occurred during the 1999 test. The splits
occurred at stress levels above the highest previous levels to which the pipe had subjected in
the pre-service hydrostatic test. One SSWC break occurred in the 2001 test at a stress level of
98% of SMYS. No SSWC break occurred in the 2004 test. Seam splits occurred in both the
2001 test and the 2004 test, but they were fewer in number than in the 1999 test and occurred

FINAL
15-019
at higher stress levels. The results of the three successive hydrostatic tests in 1999, 2001, and
2004 suggest that the threats of SSWC and seam manufacturing defects are being adequately
addressed by hydrostatic testing.
The operator of the pipeline has utilized ILI in various situations. Following the in-service
external corrosion rupture in 1985, a low-resolution MFL tool was run through the entire
pipeline to locate and investigate other possible metal loss areas. Two hundred locations of
metal loss were identified and prioritized by severity. Fifty of these were excavated and
examined. Fifteen of the areas of metal loss were found to warrant repairs in the form of full-
encirclement sleeves. The rectifier outputs were enhanced such that pipe-to-soil potential
readings at all test leads were brought up to -850 milivolts (Cu,CuSO4).
Following the occurrence of an in-service failure from SSWC in 1995, the operator used a high-
resolution MFL tool to locate and investigate metal loss. The tool was not specifically designed
to find narrow slot-like defects such as SSWC, but the operator hoped that SSWC would be
located in conjunction with the examinations of other metal loss anomalies. Seven hundred and
twenty metal loss anomalies were graded and prioritized by the vendor. Fifty of those were
examined via excavations. Two were found to contain SSWC, but the dimensions of the SSWC
had not been accurately defined by the MFL technology. After hydrostatic testing began in
1999 to address the SCC threat, it was assumed that any SSWC that was severe would fail in
the hydrostatic tests.
In 2004 the operator began an integrity management program in which a high-resolution metal
loss inspection and a caliper-type dent inspection of the entire pipeline were to be conducted
every 7 years. The two inspections were repeated in 2011, and the next round of such
inspections is scheduled for 2018.
In 2010, the operator elected to address the SCC threat for the entire pipeline with EMAT ILI
technology instead of continuing to test certain valve sections based on the Fessler/Rapp
calculations. The results were satisfactory in terms of locating and characterizing SCC, so it is
likely that the operator will continue to use the technology on a periodic basis to assess the
integrity of the pipeline from the standpoint of SCC. The time for the next assessment was
established based on the apparent crack growth since the last hydrostatic test in 2004.
Because, hydrostatic testing as a means of verifying integrity was discontinued in 2010, the
operator began to utilize CMFL technology in 2014 to assess the integrity of the entire pipeline
from the standpoint of SSWC. The results were sufficiently satisfactory that the operator plans
to conduct such inspections periodically. The time for the next assessment was established
based on the apparent SSWC growth since the last hydrostatic test in 2004.

FINAL
15-019
Another consequence of discontinuing hydrostatic testing is that seam integrity must now be
verified by other means. The threat of failure from seam manufacturing anomalies growing by
PCIF would seem to be quite low. This is because, although seam defects have ruptured during
previous hydrostatic tests, none has exhibited any evidence of in-service crack growth (based
on detailed examinations of the fracture surfaces of seam anomalies that failed in the tests),
and all of the test failures caused by seam anomalies have occurred at pressure levels above
that of the manufacturer’s hydrostatic test (to 90% of SMYS). The operator should perform a
pressure-cycle-fatigue analysis via the technique suggested previously. In all likelihood, such
and analysis would show that a defect that could have survived the highest previous hydrostatic
test stress (minimum of 100% of SMYS) would not be expected to fail for hundreds of years.
Such an outcome would be expected because the ratio of test pressure to operating pressure
and the level of operating stress assure that no large seam defect could exist. Nevertheless, the
operator should alert the ILI service vendor that inspections for cracks should include looking
for seam anomalies as well as SCC and SSWC.
Note that a threat to seam integrity from PCIF in a natural gas transmission pipeline can be
significant if the pipeline was not given a pre-service hydrostatic test, if the pre-service
hydrostatic test was to a level of only 1.1 times the MAOP, or if the pipeline is operated at a
maximum stress level that is much less than 72% of SMYS (i.e., 30 to 50 percent of SMYS) and
has only been tested to a level of 1.25 times the MAOP.
process described below starting with START (Figure C1 of Appendix C). Flowchart START as
adapted to this particular pipeline is repeated as Figure 14 below.

FINAL
15-019


yes attributes.

Proceed to
legacy pipe ?
MODERN PIPE
yes
Figure 14. Start of Process to Decide Whether to Repair or Replace the Base Case
Natural Gas Pipeline that is Operated at a Stress Level of 72% of SMYS
The first question to answer in the process is: Are the essential design attributes of segment
known? The answer based on the data in Table 5 is “yes”. If the answer had been no, the
operator would have to either establish or verify the essential design attributes or schedule
replacement. The practical constraints on operating any pipeline safely mean that diameter,
nominal wall thickness, grade of material (specified minimum yield strength or SMYS), type of
longitudinal seam (to establish the “joint factor”), design factor (upper bound on operating
hoop stress as a percent of SMYS) , and operating temperature are essential design attributes.
These should be readily available from manufacturer’s certificates, design calculations, and
alignment sheets if such records have been retained in the operator’s archives.
But, what if some or all of these essential attributes are not known? For a natural gas pipeline,
processes for establishing nominal wall thickness, grade of material, and joint factor when
traceable, verifiable and complete records are not available are given in Part 192, §192.105. In
the process of acquiring the samples required by §192.107 to determine an unknown yield
strength, the operator should be able to establish the type of longitudinal seam. Diameter
should be observable in above-ground portions of the pipeline, and it can be verified continually
along the pipeline if the pipeline can accommodate running a caliper tool. Nominal wall
thickness can be established as described in §192.109. Alternatively, wall thicknesses along the
pipeline can be verified via in-line inspection with currently available ILI metal loss tools. The
design factor can be determined as specified in §192.111, and the joint factor can be
established from the identified type of seam as specified in §192.113. There will be no need for
temperature derating as long as the operating temperature does not exceed 250⁰F. Lastly,

FINAL
15-019
though it is not a part of federal regulations, it would seem that an operator could use the
results of a recent hydrostatic test or could conduct a hydrostatic test to verify the integrity of
the pipeline and establish its minimum pressure-carrying capacity. As in-line tools evolve, it
may become possible to verify yield strengths and seam types along a pipeline as well as
diameters and wall thicknesses.
The second question is: Does segment contain legacy pipe? The segment is comprised of flash-
welded pipe, so the answer is “yes”. The operator must proceed to LEGACY PIPE. If the
answer had been no, the operator would have been directed to the flowchart shown in Figure
C8 in Appendix C (MODERN PIPE). The LEGACY PIPE flowchart as adapted to this particular
pipeline is shown in Figure 15.
Legacy Pipe
or Part 195? yes
yes
Part 195.


SSAW pipe


susceptible to HSC
5. Segment contains
Proceed to
puddle welds, patches, or
LEGACY REPAIRS
half soles
Figure 15. Threats to Address for Base Case Natural Gas Pipeline Comprised of
Flash-welded Pipe that is Operated at a Stress Level > 30% of SMYS
As shown in Figure 15, the operator must review the pressure test history of the pipeline. If
the pipeline has not been tested in accordance with the requirements of Part 192, and it does

FINAL
15-019
not qualify for one of the exceptions in the regulations, the operator must either conduct the
appropriate test or lower the MAOP to a level that would bring the pipeline into compliance with
the regulations. If neither of these can be done, the operator should schedule replacement of
the pipeline. This particular pipeline has been tested adequately.
This pipeline is comprised of flash-welded pipe which can be susceptible to the seam defects
that may leak or become enlarged in service from PCIF and may also be susceptible to SSWC.
In fact, the history of the pipeline shows that it is susceptible to degradation from SSWC,
external corrosion, and SCC. Therefore, Cells 1, 7, and 9 are highlighted in Figure 15. The
pipeline contains no SSAW pipe, no furnace lap-welded pipe, no miter bends, no wrinkle bends,
no non-standard fittings, no Dresser couplings, no threaded couplings, no acetylene girth welds,
no puddle welds, no patches, and no half-soles. Internal corrosion has never been detected via
ILI or noticed at any time segments of the pipe have been removed, and the pipeline carries
dry natural gas with no sour components. Neither pipe body hard spots nor hard heat-affected-
zones adjacent to the seam exist. Therefore, the threats represented by Cells 2-6 and 8 and 10
need not be considered.
Based on the highlighted cells, the operator of this pipeline must pursue the process of making
a repair/replace decision via Flowcharts LF-ERW, EC, and SCC. The first of these, Flowchart LF-
ERW (Figure C3 of Appendix C) as adapted to this particular pipeline is repeated as Figures 16A
and 16B below.

FINAL
15-019
welded pipe.

Go to
yes

PCIF.
to occur?
no
yes

Figure 16A. Feasibility of Preventing in-Service Failures from PCIF or SSWC by

Hydrostatic Testing or ILI

FINAL
15-019
PART B

no
yes
that prevents seam failures in rehabilitation based on
to failure.
segments.
Figure 16B. Feasibility of Preventing in-Service Failures from PCIF or SSWC by

Lowering Operating Pressure
The operator examines the fatigue effect of the pressure spectrum and determines that it is
unlikely that PCIF will be an integrity threat within the useful life of the pipeline. The periodic
retesting for SCC provides additional assurance that it is not, because no test failures have been
caused by PCIF. However, because the pipeline is comprised of flash-welded pipe, the operator
must assure its continuing seam integrity by taking the necessary steps to prevent in-service
leaks or ruptures from selective seam weld corrosion (SSWC) as shown in Figure 16A. Periodic
seam integrity assessment via hydrostatic testing was being used to address SCC, but the retest
interval, if based on a Fessler-Rapp analysis, could be short enough to prevent in-service
failures from SSWC as well. Therefore, the operator would need to continue to employ the
hydrostatic retesting on the same schedule. In 2010, to avoid the disruptive effect of continued
hydrostatic testing, the operator chose the option of employing ILI crack-detection technology
to address SCC and CMFL technology to address SSWC.
In place of hydrostatic testing and ILI, as indicated in Figure 16B, the operator could also
choose to lower the operating pressure of the pipeline to lessen the likelihood of failure from all
causes. However, it would be incumbent upon the operator to show that in-service failures
from SCC or SSWC would not be likely to occur at the lower MAOP.

FINAL
15-019
If in-service leaks or ruptures caused by SCC or SSWC continue to occur in spite of either a
hydrostatic testing or ILI-based re-assessment program, the operator must either decrease the
interval between assessments to the point where such failures cease to occur or schedule
replacement of the pipeline.
As noted in conjunction with the Base Case No. 1 example, total replacement may not be
necessary. In many cases, it might be possible to identify and rehabilitate only those segments
that are affected by a particular threat (or threats) that are judged to be too costly to address
by continued assessment and repair. Such rehabilitation could involve recoating of the pipe
and/or replacement of some pipe, not necessarily the whole pipeline.
To keep the pipeline in service, the operator must satisfactorily address the threats of failure
from external corrosion, SSWC, and SCC to the point where no in-service failure is caused by
any of these phenomena. The threat of external corrosion as adapted to this particular pipeline
is addressed via Flowchart EC (Figure C12 of Appendix C) repeated below as Figure 17.

FINAL
15-019

yes yes no

yes
no ∩

ECDA?
yes yes

yes
Conduct periodic
assessments or
prevent leaks and
operate segment in
or Part 195.
This particular natural gas pipeline is coated and cathodically protected, but as its history
shows, it has sustained previous ruptures from external corrosion. The coating on the pipeline
is of a type that does not shield the pipe from cathodic protection if it disbonds. The pipeline
can accommodate ILI metal loss inspection tools, and such tools are used to inspect the
pipeline at intervals determined by analysis. Comparisons of successive inspections have shown
the intervals to be sufficiently short to assure the absence of in-service failures from external

FINAL
15-019
corrosion. Therefore, replacement of the pipeline from the standpoint of the threat of external
corrosion does not have to be considered.
If the pipeline could not accommodate ILI metal-loss inspection tools, the operator could
consider addressing the corrosion threat by utilizing external corrosion direct assessment
(ECDA). This is possible because the coating on this pipeline does not shield the pipe from
cathodic protection if it becomes disbonded. As long as the ECDA program is successful in
preventing in-service failures from external corrosion, replacement from the standpoint of the
threat of external corrosion does not have to be considered.
If it is not feasible for the operator to use either ILI or ECDA, periodic hydrostatic testing could
be used to control the threat of external corrosion. However, this is generally an inefficient
means of assessing integrity from the standpoint of corrosion because of the inability of testing
to reveal short, deep defects. If hydrostatic testing is chosen as the means of assessment for
external corrosion, a close-interval pipe-to-soil potential survey done in conjunction with the
test may greatly improve the likelihood of finding the short, deep defects that survive the test.
The threat of SCC as adapted to this particular pipeline is addressed via Flowchart SCC (Figure
C14 of Appendix C) repeated below as Figure 18.

FINAL
15-019
to SCC?
Part 195.
yes

no or unknown
no

Can the pipeline
tool appropriate?
pressure?
yes
yes

inspections at
192 and Part 195.

Figure 18. Feasibility of Preventing In-Service Failures from SCC
Because of the occurrence of one in-service failure from SCC, this pipeline is clearly susceptible
to SCC. The operator has no reason to believe the SCC is dormant, so the answer to the
second question “Is the SCC dormant?” is “unknown”. As was noted in conjunction with the
Base Case No. 1 pipeline, the operator has three choices:

FINAL
15-019
1. Use SCCDA periodically to locate areas where SCC may be occurring.
2. Periodically conduct ILI with an appropriate crack-detection tool and repair cracks.
3. Periodically conduct hydrostatic tests.
The flowchart in Figure 18 runs through these choices in order, but the operator should make
the choice appropriate to the circumstances of the particular pipeline. For a pipeline where SCC
is known to exist because of past SCC failures, either ILI or hydrostatic testing would seem to
be the preferred options.
The operator suspects that SCC is present in other locations downstream from compressor
stations and decides that using SCCDA is not a sufficiently robust response to the threat.
Hydrostatic testing was chosen to assess the integrity of the first two valve sections
downstream from each compressor station. Initially, an arbitrary retest frequency of every two
years was chosen. After 2006 when the Fessler/Rapp method was published, the test intervals
were to be chosen via the Fessler/Rapp analysis. By 2010, however, the operator became
convinced that EMAT ILI technology would be capable of addressing the threat of SCC, and
hydrostatic testing was abandoned in favor of using ILI. Another type of ILI, the CMFL tool, is
chosen to assess the pipeline for SSWC in place of hydrostatic testing. The operator plans to
periodically conduct pressure-cycle analyses to assure that PCIF remains a non-threat.
safety versus the cost to replace it. If these continuing costs are acceptable to the operator,
Base Case Example No. 3: A Natural Gas Pipeline Operated at

Stress Levels Less than 30% of SMYS
Consider the following natural gas transmission pipeline. It is comprised of 24-inch-OD, 0.281-
inch-wall, X42 pipe manufactured with a flash-welded seam and installed in 1949. Each piece
of pipe was hydrostatically tested by the manufacturer to a stress level of 90% of SMYS for a
total time of 10 seconds. The pipeline is operated at a maximum hoop stress level of nearly
30% of SMYS. The MAOP of the pipeline is 290 psig (29.5% of SMYS). The operator has
records documenting all aspects of the pipeline’s design, construction, and operation since
1949. The pipeline cannot accommodate ILI tools. The pipeline has no facilities for launching
or receiving pigs, the mainline valves are not full-opening valves, and the pressure and flow
rate are inadequate for pushing a pig. The pipeline also contains 50 miter bends that may pose
obstructions for ILI tools. The failure history indicates three in-service leaks from external

FINAL
15-019
corrosion, one in 1959, one in 1964, and one in 1984. In addition, an in-service rupture caused
by selective seam weld corrosion (SSWC) occurred in 1995.
The pipeline was gas-tested to 110% of MAOP (320 psig) at the time it was placed in service.
The pipeline is a single feed line to a distribution system and cannot be taken out of service for
the purpose of hydrostatic testing. The option of gas testing the pipeline again for any purpose
was ruled out years ago because of the potential hazard of crack propagation in the event of a
test rupture.
process described below starting with Flowchart START (Figure C1 of Appendix C). Flowchart
START as adapted to this particular pipeline is repeated as Figure 19 below.
Are the essential attributes of

the segment (i.e., diameter,
wall thickness, grade of no Can attributes be established no Schedule replacement
material, seam type, design or verified?
factor, and operating
temperature) known?
yes
Establish or Verify
Essential Attributes
yes

Proceed to
legacy pipe ?
MODERN PIPE
yes
Natural Gas Pipeline that is Operated at a Stress Level < 30% of SMYS
The essential attributes of this pipeline are known. The flash-welded seams mean that the
pipeline is comprised of legacy pipe, so the operator may proceed to Flowchart LEGACY PIPE
(Figure C2 of Appendix C). The LEGACY PIPE flowchart is shown as Figure 20 below.

FINAL
15-019
Legacy Pipe
or Part 195? yes
yes
Part 195.


SSAW pipe


susceptible to HSC
5. Segment contains
puddle welds, patches, or Proceed to
half soles LEGACY REPAIRS
Figure 20. Threats to Address for Base Case Natural Gas Pipeline Comprised of
Flash-welded Pipe that is Operated at a Stress Level < 30% of SMYS
The pipeline was gas-tested at the time of commissioning to a pressure level of 1.1 times
MAOP. If the segment is entirely in a Class 1 location, this test meets the requirements of Part
192, §192.503 and §192.619. If parts of the segment are located in other class locations, this
level of test may or may not be sufficient depending on the interpretation. The operator should
establish whether the test does or does not comply with the regulations.
Note that a threat to seam integrity from PCIF in a natural gas transmission pipeline can be
significant if the pipeline was not given a pre-service hydrostatic test, if the pre-service
hydrostatic test was to a level of only 1.1 times the MAOP, or if the pipeline is operated at a
maximum stress level that is much less than 72% of SMYS (i.e., 30 to 50 percent of SMYS) and
has only been tested to a level of 1.25 times the MAOP. The likelihood that large
manufacturing defects with failure stress levels much less than 90% of SMYS exist in this pipe is
low because each piece was subjected to a test to 90% of SMYS for a period of 10 seconds by
the manufacturer.

FINAL
15-019
This pipeline is comprised of flash-welded pipe which can be susceptible to the seam defects
that may leak or become enlarged in service from PCIF and may also be susceptible to SSWC.
The threat of fatigue is improbable because of the low operating stress (<30% of SMYS)
relative to the stress level of the manufacturer’s hydrostatic test of each piece of pipe (90% of
SMYS). Nevertheless, the operator should conduct a pressure-cycle fatigue analysis to estimate
the minimum time to failure. 27
Any pipeline may be susceptible to degradation from external corrosion. Any pipeline
comprised of LF-ERW, DC-ERW, or flash-welded pipe may be susceptible to SSWC as well.
Therefore, Cells 1 and 7 are highlighted in Figure 20. The pipeline contains no SSAW pipe or
furnace lap-welded pipe and no miter bends, no non-standard fittings, no Dresser couplings, no
threaded couplings, no acetylene girth welds, no puddle welds, no patches, and no half-soles.
It does contain wrinkle bends of less than 12.5 degrees, but such wrinkle bends are permitted
because the operating stress level is less than 30% of SMYS. Internal corrosion has never been
noticed at any time segments of the pipe have been removed, and the pipeline carries dry
natural gas with no sour components. The threat of SCC is essentially non-existent because of
the low operating stress level. Also, no pipe body hard spots have been found. Therefore, the
threats represented by Cells 2-6 and 8-10 need not be considered. Based on the highlighted
cells, the operator of this pipeline must pursue the process of making a repair/replace decision
via Flowcharts LF-ERW and EC. The first of these, Flowchart LF-ERW (Figure C3 of Appendix C)
as adapted to this particular pipeline is repeated as Figures 21A and 21B below.
27
It is possible that a hazardous liquid pipeline operated at a stress level less than 30% of SMYS may not be at risk from PCIF
within its useful life. However, the degree of susceptibility depends strongly on the stress level applied in the most recent pressure
test. Because the pressure cycling of liquid pipelines tends to be more aggressive than that of gas pipelines, a PCIF analysis of a
liquid pipeline operated at a stress level below 30% of SMYS would seem to be even more important than it is for a gas pipeline
operated at a stress level below 30% of SMYS.

FINAL
15-019

welded pipe.

Go to
yes

PCIF.
to occur?
no
yes

Figure 21A. Feasibility of Preventing in-Service Failures from PCIF
Because of the flash-welded seam in the pipe, it is necessary to examine the potential threat of
failure from PCIF even if the threat turns out to be negligible. Figure 21A indicates that the
operator should conduct a pressure-cycle fatigue analysis using actual historical pressure
records from the pipeline. For such an analysis, the initial flaw sizes are usually established on
the basis of the most recent pressure test. This pipeline was gas-tested prior to being placed in
service to a stress level of only 32.4% of SMYS (1.1 times the maximum operating stress of
29.5% of SMYS). However, manufacturing records show that each piece of pipe was tested by
the manufacturer to 90% of SMYS. The operator can rely on the manufacturer’s test to 90% of

FINAL
15-019
SMYS to establish the initial flaw sizes. For a conservative analysis, the use of a toughness level
equivalent to a CVN energy of 200 ft lb and a flow stress equal to the specified ultimate tensile
strength of the pipe is recommended.
Assume that the analysis based on a 90%-of-SMYS manufacturer’s test shows that the
expected minimum time to failure is 3000 years. In such a case, the operator can dismiss the
threat to failure from PCIF for the conceivable life of the pipeline, and continue to operate the
pipeline in accordance with Part 192 subject to the outcome of Part B shown in Figure 21B.
Note that if the operator could not prove that the pipe had been tested to 90% of SMYS, it
would have been necessary to do the fatigue analysis using the 32.4% of SMYS test as the
basis for establishing initial flaw sizes. In such a case, the analysis likely would show a
minimum time to failure well within the expected life of the pipeline. To address that threat,
the operator would either have to conduct periodic seam integrity assessments using
hydrostatic testing or ILI, lower the operating pressure to the point where fatigue is no longer a
threat, or schedule replacement of the pipeline after half of the estimated minimum time to
failure has elapsed. The operator cannot take the pipeline out of service to conduct periodic
hydrostatic testing, and the pipeline cannot accommodate ILI pigs. That would leave only the
options of scheduling replacement or lowering the operating pressure. An acceptable lower
operating pressure can be based on a fatigue analysis using a scaled-down pressure-cycle
spectrum associated with the lower operating pressure that results in an acceptable minimum
time to failure.

FINAL
15-019
PART B

no
yes
that prevents seam failures in
rehabilitation based on
service? half the calculated time to
failure.
Schedulereplacement
Schedule
replacement.
segments.
Figure 21B. Feasibility of Preventing in-Service Failures from SSWC
In the case of this pipeline, failures from both external corrosion and SSWC have occurred. The
operator cannot take the pipeline out of service to conduct periodic hydrostatic testing, and the
pipeline cannot accommodate ILI pigs. According to Figure 21B, the operator has the option to
address the threat of SSWC by means of external corrosion direct assessment (ECDA). ECDA
does not identify the presence of SSWC directly. Instead, the ECDA process identifies areas of
the pipe affected by external corrosion, and direct examination of the pipe would be expected
to reveal any significant SSWC that may have occurred. If by using ECDA, the operator is
successful in preventing failures from SSWC, the pipeline can remain in service. As seen in
Figure 22, the operator must prevent the occurrence of failures from external corrosion as well
by using ECDA.

FINAL
15-019

yes yes no

yes
no ∩

ECDA?
yes yes

yes
Conduct periodic
assessments or
prevent leaks and
operate segment in
or Part 195.
safety versus the cost to replace it. If these continuing costs to maintain safety and pipeline
integrity are acceptable to the operator, then replacement may not be necessary.

FINAL
15-019
It was noted in the data for this pipeline that it contains 50 miter bends. Miter bends are a
legacy feature. However, miter bends with angle changes less than or equal to 12.5 degrees
are permitted in gas pipelines that are operated at stress levels of less than 30% of SMYS 28.
Unless individual miter bends are thought to be threats to pipeline integrity at specific locations
or under specific circumstances, the operator does not have to take any action with respect to
the miter bends. The appropriate repair/replace strategy for the miter bends can be assessed
using Flowchart LEGACY FEATURES (Figure C6 of Appendix C). For this particular pipeline, that
flowchart is shown in Figure 23.
Segment contains miter Do the miter bends, wrinkle bends, Schedule replacement of miter
bends, wrinkle bends, non- non-standard fittings, Dresser bends, wrinkle bends, non-standard
couplings, threaded couplings, or yes fittings, Dresser couplings, or
standard fittings, Dresser
couplings, threaded couplings, acetylene girth welds constitute a threaded couplings and continue to
or acetylene girth welds. systematic threat to the integrity of the operate in accordance with Part 192
pipeline? or Part 195.
no
Do miter bends, wrinkle bends,

Is it feasible to replace all
wrinkle bends, non-standard
miter bends, wrinkle bends,
fittings, Dresser couplings, yes yes
non-standard fittings,
threaded couplings, or acetylene
Dresser couplings, threaded
girth welds occasionally cause a
couplings, or acetylene girth
leak or a rupture due to the
welds?
location or a special circumstance?
no
no Schedule replacement of
Conduct an engineering study individual legacy features
to identify locations or at locations or under
Continue to operate in circumstances where each circumstances where the
accordance with Part 192 or specific type of individual particular feature
Part 195. constitutes an integrity
legacy feature could
constitute an integrity threat. threat and continue to
operate in accordance with
Part 192 or Part 195.
Figure 23. Consideration of the Miter Bends in the Low-Stress Gas Pipeline
As seen in Figure 23, the operator does not need to take any action as long as the miter bends
have not been implicated in any leak or rupture.
Other Examples for Transmission Pipelines

The following examples are intended to show additional cases of considerations for
repair/replace decisions for various situations not covered in the three base cases related to
28
Part 192 permits angle changes at a girth weld of up to 3 degrees regardless of the operating stress level. Angle changes of up
to 90 degrees are permitted in pipelines that operate at stress levels of 10% of SMYS or less.

FINAL
15-019
transmission pipelines. These cases are less detailed than the base case examples, because
many of the aspects of these cases were covered previously in the discussions of the base case
examples. In most of these other examples, the first part of the discussion concerns a
hazardous liquid pipeline. At the end of each case, the differences in decision-making
considerations for high-stress and low-stress natural gas transmission pipelines, if any, are then
discussed.
A Pipeline with an SSAW Seam

The Hazardous Liquid Pipeline Case
Consider a 500-mile-long, crude oil pipeline comprised of 24-inch-OD, 0.312-inch-wall, line pipe
with a minimum specified yield strength of 42,000 psi, fabricated with a single-submerged-arc-
welded (SSAW) seam, a type of legacy pipe. The pipeline was installed in 1948 at a time when
only a tentative standard, APL STD 5LX existed for X grades of pipe. However, the pipe was
purchased under an agreement whereby the manufacturer agreed to produce a material
meeting the requirements of the tentative standard. The material was to have a specified
minimum yield strength (SMYS) of 42,000 psi, a minimum ultimate tensile strength of 60,000
psi, and a minimum elongation of 25% in a 2-inch gage length. The agreement included limits
on the following alloys in maximum percentages by weight based on check analyses: carbon:
0.33, manganese: 1.28, phosphorus: 0.115, and sulfur: 0.065. Moreover, the manufacturer
agreed to meet all other requirements expressed in the 1948 tentative standard including a
pressure test of each piece of pipe to a pressure level of 950 psig (87% of SMYS) for a period
of 10 seconds. The pipeline was coated over-the-ditch with a coal-tar enamel and fiberglass-
reinforced felt wrapping. Initially, there was no cathodic protection system, but an impressed-
current cathodic protection system was installed in 1955. The pipeline is operated at an MOP of
786 psig corresponding to a hoop stress level of 72% of SMYS. The pipeline can accommodate
all types of ILI tools. No legacy features such as acetylene girth welds, miter bends, wrinkle
bends, patches, or half soles exist on this pipeline.
A far as records are concerned, the operator has a report from a third-party pipe mill inspection
company that documents the results of tensile tests required by API Specification 5L showing
that the strengths met or exceeded the agreed-upon minimum yield strength and minimum
ultimate tensile strength values and that the alloy contents were within the prescribed limits.
The report also documents the fact that each piece of pipe was tested to a pressure level of
950 psig corresponding to a hoop stress level of 87% of SMYS for period of 10 seconds. The
operator also has records of the initial pre-service hydrostatic test to a minimum hoop stress
level of 90% of SMYS and subsequent hydrostatic tests, records of all service failures, and
records of ILI tool runs.

FINAL
15-019
The service failure history contains incidents caused by external corrosion and seam failures
from manufacturing defects. There have been no failures from internal corrosion and the
prospect of internal corrosion existing seems unlikely because only non-sour crude oil has ever
been shipped and because no internal corrosion has been found in any removed pieces of pipe.
SSAW pipe is not known to be susceptible to SSWC. Therefore, the threats of PCIF failures at
seam manufacturing defects, SCC, and external corrosion are the primary threats to be
mitigated if a decision is made to keep the pipeline in service. To address these threats the
operator can go through the details of the hazardous liquid pipeline flowcharts in Appendix C as
was done in Base Case Example No. 1.
Figure C1 (START) directs the operator to Figure C2 (LEGACY PIPE) which, in turn directs the
operator to Figure C4 (Flowchart SSAW), Figure C12 (EC) and Figure C14 (SCC). Flowchart
SSAW is shown as Figure 24. Periodic seam integrity assessment is needed to mitigate against
service failures from PCIF. If the pressure cycle history of this pipeline is like those of many
other hazardous liquid pipelines, periodic seam integrity assessment will be needed to mitigate
against service failures from PCIF within the time frame of the useful life of the pipeline. Seam
integrity assessment can be done via either hydrostatic tests or successive ILI crack detection
tool runs at intervals predicted by pressure-cycle fatigue analysis as explained previously.

FINAL
15-019
Segment contains SSAW

pipe.
Use pressure cycles to calculate

times to failure from PCIF or rely on
service and test failure history to
establish re-assessment interval to
prevent service failures from PCIF
Is it feasible to lower the Lower operating stress as needed to

Is periodically assessing seam operating stress to a level prevent seam failures in service and
no yes
integrity by means of hydrostatic that prevents seam failures in continue to operate in accordance
testing or ILI feasible? service? with Part 192 or Part 195
yes
no
Assess seam integrity periodically
based on the established interval
to prevent service failures from
PCIF.
decrease the re-assessment
Are service failures from PCIF
yes intervals or to decrease the no Schedule
occurring between re-
operating stress until service replacement.
assessments?
failures from PCIF cease to
occur?
no
yes

seam integrity assessments on failures from PCIF cease to occur.
the same schedule.
Figure 24. Flowchart for Evaluating a Repair/Replace Decision for SSAW Pipe in a
If the operator goes through the details of the external corrosion flowchart as was done in Base
Case Example No. 1 (see Figure 10), it will be clear that periodic integrity assessment is needed
to mitigate against service failures from external corrosion. Integrity assessment to mitigate an
external corrosion threat can be done most effectively via successive ILI metal-loss detection
tool runs at intervals predicted by the linear growth rate analysis technique explained
previously.
The pipeline could also be exposed to the threat of failure from SCC. Mitigation of the SCC
threat is necessary if the pipeline is to remain in service. The operator can go through the SCC
flowchart as was done for the Base Case in Figure 11 to determine the appropriate means to
mitigate the threat of SCC. Since there have been no incidences of SCC having been found or

FINAL
15-019
having caused a failure, the operator could elect to perform periodic SCCDA inspections. If
actual SCC is found (unless it can be shown to be dormant) or if SCC causes a failure, the
operator should establish a program of periodic hydrostatic testing or ILI crack-tool inspections
to minimize the risk of recurrences of SCC-caused failures.
The operator must assure that safety and pipeline integrity are maintained when comparing the
cost to continue mitigating the threats of PCIF, SCC and external corrosion with the cost of
replacement of the pipeline in order to decide whether to repair or replace.
The High-Stress Natural Gas Pipeline Case
Suppose the pipeline just described is a high-stress natural gas transmission pipeline instead of
a hazardous liquid pipeline. Even if the pipe material and the operating stress level are the
same, one could anticipate that some attributes and some aspects of the operating histories
would differ. For example, one would expect the pressure cycles for the gas pipeline to be far
less aggressive than those of the liquid pipeline. Nevertheless, the operator of the gas pipeline
should go through the process described in Figure 24 to determine whether or not PCIF is a
threat. Most likely, it would not be a threat within the conceivable useful life of the pipeline if
the pipeline had been subjected to a pressure test to 90% of SMYS or more. If the pipeline has
not ever been pressure-tested, the PCIF analysis could be based on the fact that each piece of
pipe was tested by the manufacturer to a hoop stress level of 87% of SMYS.
The high-stress gas pipeline might be susceptible to high-pH SCC especially if it was operated at
high discharge temperatures at some time in the past. In that case, mitigation of the SCC
threat is necessary if the pipeline is to remain in service. The operator can go through the SCC
flowchart as was done for the Base Case in Figure 11 to determine the appropriate means to
mitigate the threat of SCC. If there have been no incidences of SCC having been found or
having caused a failure or if what SCC has been found is known to be dormant, the operator
could elect to perform periodic SCCDA inspections. If actual SCC is or has been found (unless it
can be shown to be dormant) or if SCC causes or has caused a failure, the operator should
establish a program of periodic hydrostatic testing or ILI crack-tool inspections to minimize the
risk of recurrences of SCC-caused failures.
previously.

FINAL
15-019
cost to continue mitigating the threats of SCC and external corrosion with the cost of
The Low-Stress Natural Gas Pipeline Case
If the pipeline is a low-stress natural gas transmission pipeline operated at a hoop stress level
of less than 30% of SMYS, the fact that the pipe was subjected to the manufacturer’s test to
87% of SMYS suggests that the threat from SSAW seam integrity issues is negligible. Also,
because of the relatively low operating stress, the threat of failure from SCC is negligible.
However, the threat of failure from external corrosion must be addressed. In the case of a low-
stress pipeline, it is usually not feasible to run ILI tools because of the low pressure, and
pressure testing is too inefficient to be practical even if it were possible to take the pipeline out
of service. The likely choice for mitigating the threat failure from external corrosion is ECDA.
cost to continue mitigating the threat of external corrosion with the cost of replacement of the
pipeline in order to decide whether to repair or replace.
A Pipeline Comprised of Furnace Lap-Welded Pipe, Constructed with

Acetylene Girth Welds
Consider a 200-mile-long, crude oil pipeline comprised of 12.75-inch-OD, 0.375-inch-wall,

furnace lap-welded pipe that was installed in 1930 by means of girth welds made via oxy-
acetylene welding. Company records show that the pipe was constructed in 1930 using lap-
welded steel pipe joined by acetylene girth welds. Although the operator has no records of the
original specifications for the pipe, the pipe is known to be comprised of steel, not wrought iron.
This and the type of girth welds have been confirmed by metallurgical examinations of test
failures associated with a 1995 hydrostatic test. Therefore, because the minimum strength
level for steel pipe was listed by manufacturers as 25,000 psi even before the first edition of API
Specification 5L (1927), the operator has used 25,000 psi as the minimum specified yield
strength of the pipe for design purposes. The design stress of the pipeline could be as high as
14,400 psi based on a design factor of 72% of SMYS and a joint factor of 0.8 for the lap-welded
seam. This stress level corresponds to a pressure level of 847 psig. However, the MOP has
always been 720 psig corresponding to a hoop stress level of 49% of SMYS, most likely because
in the first edition of API Specification 5L (1927), the standard manufacturer’s hydrostatic test
of each piece of pipe was carried out at 900 psig. The pipeline can accommodate all types of
ILI tools.

FINAL
15-019
The pipeline was installed with no external coating. Cathodic protection was applied to the
pipeline in 1955, and the cathodic protection has been upgraded a number of times by adding
rectifiers in an attempt to maintain a satisfactory level of protection. The pipeline is known to
have been affected by external corrosion. Records of in-service leaks prior to 1970 are
incomplete, but what records exist for that period indicate that the pipeline exhibited numerous
leaks. Most of these were attributed to external corrosion. Since 1970, better records have
been kept. These records show that leaks from external corrosion decreased from 6 per year in
1970 to 1 per year in 1995. The drop in leakage rate is thought to have resulted from increases
in the levels of cathodic protection.
An MFL ILI tool was run through the pipeline in 1995 to locate areas of severe metal loss so
that they could be repaired prior to a proposed hydrostatic test. Hundreds of metal loss
anomalies were located, some of which were considered to be integrity-threatening based on
tool-called dimensions. Full-encirclement sleeves were applied to integrity-threatening metal
loss anomalies in more than 100 locations.
Prior to 1995, in-service leaks and ruptures from causes other than external corrosion had
occurred. During the period from 1970 through 1995, 20 leaks occurred that were attributable
to “cracks” in the pipe body. Metallurgical analysis showed these cracks to be “burned-metal”
defects, a type of manufacturing defect that can arise if the material is heated to too-high a
temperature at the time the lap weld is made. Two in-service ruptures are known to have
occurred between 1970 and 1995, and they were said to have been seam splits resulting from
pressure surges. Also, 3 girth-weld leaks and 2 girth-weld full-circumferential ruptures occurred
between 1970 and 1995. The full-circumferential ruptures occurred near one particular river
crossing where flooding is believed to have undermined the pipeline causing excessive external
stress on the girth welds.
The is no record of a pre-service hydrostatic test, although it is said that at the time, the
practice of the pipeline’s operator was to pressure test pipelines to 1.1 times MOP using the
transported product, in this case, crude oil. In 1995 a hydrostatic test of the pipeline was
carried out at a minimum pressure level of 900 psig to validate the MOP of 720 psig and to
assess the overall integrity of the pipeline. Numerous seam splits and test breaks from burned-
metal defects occurred at test pressure levels approaching 900 psig. This suggests that the
pipe probably had been tested by the manufacturer to a level of 900 psig. Since the minimum
test failure pressure of any defect was 850 psig, it is reasonable to assume that no significant
in-service enlargement of the defects had occurred. The fact that no test ruptures or leaks
from external corrosion occurred suggests that the ILI tool run and follow-up repair work
successfully eliminated the integrity-threatening anomalies. Also, there were no test leaks or
breaks involving acetylene girth welds. Since the 1995 hydrostatic test, no in-service failures
from unbonded lap welded seams or from burned-metal defects have occurred.

FINAL
15-019
The risk of a failure from SCC in this pipeline can be considered negligible because of the fact
that it is operated at 49% of SMYS and because no failure from SCC has been known to have
occurred. The risk of a failure from internal corrosion can be considered negligible because the
pipeline has not carried sour crude oil, because no failure from internal corrosion has been
known to have occurred, and because the 1995 ILI for metal loss revealed no significant
amount of internal metal loss.
It might seem like the knowledge of the attributes of this pipeline are insufficient and that the
pipeline should, therefore, be replaced. For example, the operator does not have mill records
of the pipe manufacturing process or the mechanical test data. However, the operator has
verified the type of pipe (it is lap-welded pipe) and the type of material (it is steel, not wrought
iron). The operator has also conducted a pressure test of the pipeline in 1995 that establishes
the MOP of the pipeline. Knowing these items, the operator can continue to operate the
pipeline as long as continuing integrity assessment is feasible.
In 2002 the operator began an integrity management program for the HCA’s on this pipeline.
Because the HCA’s on the pipeline are numerous and scattered, the operator decided to
conduct metal loss integrity assessments of the whole pipeline even though the total HCA
mileage is 30 miles. Thus, the whole pipeline has been inspected for metal loss via an MFL tool
three times since 1995. The inspections were carried out in 2002, 2007, and 2012. In all three
inspections metal loss areas were found, the most severe indications were excavated and
examined, and full-encirclement sleeves were installed at locations judged to be integrity-
threatening. The trend in the numbers found in each successive inspection is fairly level. This
suggests that the corrosion is not being completely mitigated, but it does appear that the
periodic inspections allow for timely repairs to areas of developing metal loss, and it is expected
that future in-service failures from corrosion can be prevented.
For seam integrity assessment purposes, the operator has carried out hydrostatic tests of only
the HCA portions of the pipeline in 2002, 2007, and 2012. The tests have been carried out at a
minimum pressure of 900 psig. Test failures have occurred each time as pressure levels
approached 900 psig from seam defects and burned-metal defects, but neither the failure
pressure levels nor subsequent metallurgical examinations have reveal any evidence of growth
of the defect in service.
The risk of a failure involving an acetylene girth weld seems to be confined to the particular
river crossing where the two full-encirclement ruptures had occurred in the period between
1970 and 1995. Even though no failures of girth welds at that crossing had occurred since
1995, the operator replaced the entire crossing in 2002 with modern coated pipe with electric-
arc girth welds.

FINAL
15-019
If the operator of this pipeline wants to evaluate the economics of replacing this pipeline as
opposed to continue repairing it, the flowcharts in Appendix C could be used as a guide. The
first chart, Figure C1 (START), would steer the operator to Figure C2 (LEGACY PIPE). Figure C2
then steers the operator to Figure C5 (LAP WELDED) to assess the responses to possible lap-
welded pipe anomalies, Figure C6 (LEGACY FEATURES) to assess the responses to acetylene
girth welds, and Figure C12 (EC) for responses to external corrosion. Flowchart (LAP WELD) for
this pipeline is shown in Figure 25.
Segment contains furnace

lap-welded pipe.
Use service and test failure history

to establish re-assessment interval
lack-of-fusion or burned-metal
defects.
Lower operating stress as

Periodically assessing seam Is it feasible to lower the
no operating stress to a level yes
integrity by means of hydrostatic failures in service and continue
testing or ILI. is feasible that prevents seam failures in
to operate in accordance with
service?
Part 192 or Part 195
yes
no
defects.
Are service failures from lack- intervals or to decrease the
yes no
of-fusion or burned-metal Schedule
operating stress until service
defects occurring between re- replacement.
failures from lack-of-fusion or
assessments?
burned-metal defects cease to
no occur?
yes
Re-assess seam integrity at

Continue to operate and maintain
decreased interval or lower
in accordance with Part 192 or
Part 195 and continue to perform
failures from lack-or-fusion or
seam integrity assessments on
burned metal defects cease to
the same schedule.
occur.
Figure 25. Flowchart for Evaluating a Repair/Replace Decision of Furnace Lap-

Welded Pipe in a Hazardous Liquid Pipeline

FINAL
15-019
If the pressure cycle history of this pipeline is like those of many other hazardous liquid
pipelines, periodic seam integrity assessment will be needed to mitigate against service failures
from PCIF within the time frame of the useful life of the pipeline even though the pipeline is
operated at an MOP corresponding to 49% of SMYS. Seam integrity assessment can be done
via hydrostatic tests even though there is very little if any published experience with the effects
of PCIF on furnace lap-welded pipe. A hydrostatic test establishes the worst case family of
surviving defects that could grow by PCIF. Whether or not successive ILI crack detection tool
runs would reveal the presence and size of the two common lap-welded pipe manufacturing
defects (unbonded seams and burned-metal defects) is not known. It is conceivable that such
tools could be used effectively to assess the seam integrity of lap-welded pipes, but it is
doubtful that ILI crack detection technology has been used to a great enough extent in lap-
welded pipe at this time to establish a track record. In any case, the intervals between
assessments are predicted by pressure-cycle fatigue analysis as explained previously.
From the standpoint of the legacy girth welds, the operator can utilize Flowchart LEGACY
FEATURES which is show below as Figure 26.
no

welds?
no
Figure 26. A Process by Which the Operator Can Deal with Acetylene Girth Welds
The acetylene girth welds, except for a few, have performed adequately. It is not feasible to
change out all of them. The operator has done an engineering study based on the location of

FINAL
15-019
two girth weld ruptures at a particular river crossing. As a result of the study the crossing was
replaced with new pipe using electric-arc-welded girth welds. No other action was needed.
previously.
cost to continue mitigating the threats of PCIF and external corrosion with the cost of
replacement of the pipeline in order to decide whether to repair or replace the pipeline. From
the standpoint of the legacy acetylene girth welds the operator may continue to replace
particular locations where such welds are vulnerable to failure in order to avoid replacing the
whole pipeline.
If the same lap-welded pipeline discussed above transports natural gas at an MAOP of 720 psig
instead of crude oil at an MOP of 720 psig and is regulated under Part 192 instead of Part 195,
the main thing that would change with respect to a repair/replace decision would be the
likelihood that PCIF would not constitute a significant threat within the life of the pipeline. As a
result continuing seam integrity assessment especially in the form of hydrostatic testing would
become unnecessary. The cost of continuing to repair the pipeline instead of replacing it,
therefore, would be less than in the case of the liquid pipeline. The costs of continuing to
inspect for corrosion and the costs associated with monitoring and remediating external forces
that could cause failure of an acetylene girth weld in the gas pipeline likely would be about the
same as those for the same activities in the liquid pipeline.
If the same lap-welded pipeline discussed above transports natural gas at an MAOP of 400 psig
(27% of SMYS) instead of crude oil at an MOP of 720 psig and is regulated under Part 192
instead of Part 195, the main thing that would change with respect to a repair/replace decision
would be the likelihood that PCIF would not constitute a significant threat within the life of the
pipeline. As a result continuing seam integrity assessment especially in the form of hydrostatic
testing would become unnecessary. The cost of continuing to repair the pipeline instead of
replacing it, therefore, would be less than in the case of the liquid pipeline. The costs of
continuing to inspect for corrosion and the costs associated with monitoring and remediating

FINAL
15-019
external forces that could cause failure of an acetylene girth weld in the gas pipeline likely
would be about the same as those for the same activities in the liquid pipeline.
In the case of a low-stress pipeline it is usually not feasible to run ILI tools because of the low
pressure, and pressure testing is too inefficient to be practical even if it were possible to take
the pipeline out of service. A possible choice for mitigating the threat of failure from external
corrosion is ECDA, though ECDA is generally not very effective on a bare pipeline. The expense
of trying to mitigate corrosion on such a pipeline using ECDA might cause the operator to
consider replacement of the pipeline.
A Pipeline Comprised of Pipe with a DSAW Seam

Consider an 1100-mile-long refined products pipeline comprised of 36-inch-OD, 0.281-inch-wall,

API 5LX52 line pipe with a double-submerged-arc seam. The operator has copies of mill
certificates indicating that the material properties and chemistries met or exceeded the
minimum requirements of API Specification 5LX, 21st Edition, March 1965. The pipeline was
installed in 1966. It was coated over-the-ditch with a single-layer coating consisting of a
polyethylene outer wrap with a butyl-rubber mastic backing. An impressed-current cathodic
protection system was installed and activated within the first year of operation of the pipeline.
The pipeline is operated at an MOP of 584 psig corresponding to a hoop stress level of 72% of
SMYS. The pipeline can accommodate all types of ILI tools. No legacy features such as
acetylene girth welds, miter bends, wrinkle bends, non-standard fittings, Dresser couplings,
threated couplings, patches, or half soles exist on this pipeline.
Records indicate that the pipeline was subjected to a pre-service hydrostatic test to a minimum
hoop stress level of 90% of SMYS in 1966. No additional integrity assessments were done until
an in-service rupture occurred in 1982 from an area of external corrosion in the body of the
pipe. Following the rupture a low-resolution MFL ILI tool was used to locate other areas of
external corrosion. Several hundred anomalies were found, and 20 of those deemed to be the
most severe were excavated and evaluated resulting in the installation of 20 full-encirclement
sleeves.

FINAL
15-019
In 1989, another in-service rupture of the pipeline occurred. This rupture was found to have
been caused by PCIF. The origin was a railroad fatigue crack 29 along the toe of the longitudinal
seam. The pipe had been shipped from the pipe mill to the marshalling area by rail. The D/t
ratio of 128 made this pipe quite susceptible to deflection from point loading and, hence, from
railroad fatigue. The operator conducted a fatigue analysis to determine the locations along the
pipeline where PCIF is most likely to occur. In 1990, the operator then subjected those
segments of the pipeline to hydrostatic tests to stress levels ranging from 90% to 100% of
SMYS. During the test two more locations of railroad fatigue cracks were revealed by causing
test failures. In both cases it appeared that the railroad fatigue portions of the cracks were
being extended by PCIF, suggesting that both might have failed in service in the future. No
other failures from time-dependent defects occurred during the test.
In 1995, a high-resolution MFL tool was used to inspect the pipeline for corrosion-caused metal
loss. While hundreds of anomalies were located, only 20 warranted excavation and
examination. All of these anomalies were repaired by means of composite wraps.
In 1999, another in-service rupture of the pipeline occurred. This rupture was caused by
narrow axial external corrosion (NAEC) along the longitudinal seam. When the 1995 ILI data
were reviewed, this anomaly did not standout as being anywhere near as deep as the anomaly
that failed. Its failure stress calculated on the basis of the dimensions from the 1995 ILI data
was over 100% of SMYS. While it is possible that the anomaly could have grown sufficiently
between 1995 and 1999 to fail at the maximum operating stress of 72% of SMYS, the problem
more likely is the result of the ILI technology (longitudinal-field MFL) not being capable of
accurately sizing long, narrow anomalies. In 2000, the pipeline was again subjected to an ILI
metal-loss inspection, but this time a circumferential-field MFL (CMFL) tool was run. The
circumferential-field MFL was a new development, and it possesses capabilities for accurately
sizing narrow axial anomalies. When this tool was run, several locations of NAEC that were
considered to be integrity-threatening were found and remediated, none of which likely could
have been correctly identified without the CMFL tool. It is noted that the NAEC phenomenon is
associated with the fact that the tape coating on the pipeline tended to form a “tent” over the
crown of the DSAW seam. In places, ground water penetrated into these narrow voids along
the toe of the weld, and corrosion took place resulting in these long and narrow anomalies.
The NAEC phenomenon is rarely found on pipelines with other types of coating such as coal tar
29
A railroad fatigue crack arises during rail shipment of pipe. The most common case involves a crack developing at the toe of the
longitudinal seam of a piece of DSAW pipe that has been improperly loaded onto the rail car such that the crown of the seam weld
is the point of contact with one of the wood bearing strips supporting the pipe. See API 5L1: Recommended Practice for Railroad
Transportation of Pipe.

FINAL
15-019
enamel or fusion-bonded epoxy, because voids at the toes of the DSAW seam normally do not
exist in conjunction with these types of coatings.
In 2002, the operator instituted an integrity management plan for HCAs along the pipeline. The
plan called for ILI metal loss inspections, ILI inspections using CMFL technology to monitor for
NAEC, and ILI crack-tool inspections to verify seam integrity from the standpoint of PCIF. The
intervals were set at 5 years for these inspections, but the start times were staggered such that
only one type of tool is used in any given year. The operator also decided to inspect 100% of
the pipeline with each type of tool even though HCAs account for only 30% of the mileage.
The HCAs are scattered along the pipeline making it impractical to inspect only the HCAs.
From the standpoint of a repair/replace decision, one can see that the primary threats to this
pipeline that must be mitigated, if it is to remain in service, are PCIF, external corrosion, and
NAEC. The operating stress level and the type of coating also render it susceptible to SCC even
though no failure from SCC has occurred. The pipeline does not transport a corrosive product,
so there is no significant threat associated with internal corrosion. Besides, if the operator
continues to use high-resolution MFL tools to search for metal loss, any internal corrosion that
may develop will be revealed.
If the time comes when the operator wants to consider replacement of this pipeline, the
decision can be facilitated by means of the hazardous liquid pipeline flowcharts in Appendix C.
Figure C1. Figure C1 (START), steers the operator to Figure C8, MODERN PIPE. It is clear that
the pipeline has met the pressure test requirements of Part 195, and the flowchart tells the
operator to proceed to Figure C10 (DSAW), Figure C12 (EC), and Figure C14 (SCC). Flowchart
DSAW addresses the issues related to the fact that this pipeline is comprised of DSAW pipe.
Flowchart DSAW is shown below as Figure 27.

FINAL
15-019
Is it possible to mitigate narrow

Segment contains DSAW or Is the pipeline coated with axial external corrosion (NAEC)
HSAW pipe made prior to a single or double layer yes no
and/or SCC, if either or both
1971. polyethylene tape coating? occur, using ILI, hydrostatic
testing, or another method?
no yes
times to failure from PCIF or rely on Assess integrity and take
service and test failure history to necessary actions to prevent
establish re-assessment interval to failures from NAEC and SCC

Is periodically assessing seam Is it feasible to lower the needed to prevent seam
testing or ILI feasible? that prevents seam failures in to operate in accordance with
service? Part 192 or Part 195.
yes
no
PCIF
Are service failures from PCIF
operating stress until service replacement.
assessments?
occur?
no
yes

in accordance with Part 192 or Re-assess seam integrity at
Part 195 and continue to perform decreased interval or lower
seam integrity assessments on operating stress until service
the same schedule. failures from PCIF cease to occur.
Figure 27. Flowchart for Evaluating a Repair/Replace Decision for DSAW Pipe in a
The first step for a pipeline comprised of DSAW pipe is to establish whether or not the coating
is of a type that “tents” over the crown of the weld. In this pipeline it does, making the pipe
vulnerable to NAEC (one such failure has already occurred). The single-layer polyethylene
coating also raises the risk of occurrence of SCC because of its tendency to shield the pipe from
cathodic protection if it becomes disbonded. The operator must mitigate the risks of SCC,
NAEC, and external corrosion and perform appropriate periodic integrity assessments in a timely
manner to prevent failures. Note that a CMFL tool needs to be used to assess for NAEC.

FINAL
15-019
A prior in-service failure and two hydrostatic test failures from PCIF mean that periodic seam
integrity assessment will be needed to mitigate against service failures from PCIF. A fatigue
analysis using the pressure cycle history of this pipeline can be used to estimate the length of
interval between integrity assessments. Seam integrity assessment can be done via either
hydrostatic tests or successive ILI crack detection tool runs at intervals predicted by pressure-
cycle fatigue analysis as explained previously. Though the re-assessment interval might not be
the same as for PCIF, the integrity of the pipeline from the standpoint of SCC can also be
assessed periodically via either hydrostatic testing or ILI crack detection tool runs.
cost of performing required integrity assessments over the life of the pipeline with the cost of
replacement.
Suppose the pipeline just described is a high-stress natural gas transmission pipeline instead of
a hazardous liquid pipeline. The comparable 30-inch-OD, X52 DSAW gas pipeline of the 1966
era would likely be comprised of a somewhat thicker material such as 0.375-inch-wall pipe and
would have a correspondingly higher operating pressure (MAOP = 936 psig) for the 72%-of-
SMYS operating stress level. Assume that the gas pipeline also was coated with the same type
of coating, i.e., a polyethylene coating with a butyl rubber mastic backing. One could anticipate
that it will be exposed to same degree to the risks of failures from NAEC, SCC, and external
corrosion.
Because this pipe has a D/t ratio of 80, it is much less likely to have sustained railroad fatigue
than the pipe in the hazardous liquid example (if, indeed, it was shipped by rail). However, if it
was shipped by rail and was loaded improperly on rail cars, it is not immune from railroad
fatigue as one can see by reading API 5L1. If one assumes that the pipeline was given a pre-
service hydrostatic test to a minimum stress level of 90% of SMYS, any surviving railroad
fatigue cracks would be of no more concern than any other potential manufacturing defect.
One would expect that the pressure cycles for the gas pipeline would be far less aggressive
than those of the liquid pipeline. That most likely means that PCIF would not be a threat within
the conceivable useful life of the pipeline if the pipeline had been subjected to a pressure test
to 90% of SMYS or more. If the pipeline has not ever been pressure-tested, the PCIF analysis
could be based on the fact that each piece of pipe was tested by the manufacturer to a hoop
stress level of 90% of SMYS. The likelihood that large manufacturing defects with failure stress
levels much less than 90% of SMYS exist in this pipe is low because each piece was subjected
to a test to 90% of SMYS for a period of 10 seconds by the manufacturer.

FINAL
15-019
Note that the manufacturer’s hydrostatic test does not provide assurance with regard to railroad
fatigue cracks. Therefore, if no hydrostatic test has ever been performed in situ on such a
pipeline, either a test to 90% of SMYS would have to be conducted or a reliable ILI crack-
detection tool would have to be run to assure that no railroad fatigue crack exists that could
cause a failure at the MAOP of the pipeline.
The operator can go through the appropriate flowcharts in Appendix C: Figure C1 (START) and
Figure C8 (MODERN PIPE). The pipeline has been pressure tested in compliance with Part 192,
so the operator needs to address the threats of PCIF, NAEC, EC, and SCC using Figure C10
(DSAW), Figure C12 (EC), and Figure C14 (SCC). Mitigation of PCIF may be unnecessary if
fatigue analysis shows that failure from PCIF is not likely to occur within the useful life of the
pipeline. Mitigation of NAEC requires periodic integrity assessment using an appropriate ILI
technology such as a CMFL tool. If SCC is or has been found (unless it can be shown to be
dormant) or if SCC causes or has caused a failure, the operator should establish a program of
periodic hydrostatic testing or ILI crack-tool inspections to minimize the risk of recurrences of
SCC-caused failures.
previously. However, mitigation of the threat of NAEC requires the periodic use of a CMFL tool
or a hydrostatic test.
cost to continue mitigating the threats of PCIF, NAEC, SCC and external corrosion with the cost
of replacement of the pipeline in order to decide whether to repair or replace.
Consider a hypothetical case in which the pipeline in this example could not accommodate ILI
tools. The mitigative options for the operator in this case would be limited to hydrostatic
testing and SCCDA. ECDA cannot be used on this pipeline because the coating shields the pipe
from cathodic protection when it disbonds. The operator has the following choices and should
make the decision based on risk and consequence factors associated with each integrity
mitigation alternative taking into consideration cost impact.
• Conduct periodic hydrostatic tests to address the threats from NAEC and SCC.
• Recoat the pipeline with a coating that virtually eliminates the risk of NAEC and possibly
SCC as well.
• Make the pipeline capable of handling at least some types of ILI tools.

FINAL
15-019
• Replace the pipeline.
It is likely that a low-stress natural gas transmission pipeline comprised of the same pipe as
discussed previously in conjunction with the high-stress natural gas pipeline will be much
shorter than 1100 miles. Otherwise, however, such a pipeline conceivably could be comprised
of 30-inch-OD, 0.375-inch-wall, X52, DSAW pipe. If such a pipeline is operated at a hoop stress
level of less than 30% of SMYS, the fact that the pipe was subjected to the manufacturer’s test
to 90% of SMYS suggests that the threat from PCIF is negligible. Also, because of the relatively
low operating stress, the threat of failure from SCC is negligible. However, the threat of failure
from NAEC and external corrosion must be addressed. If the pipeline cannot accommodate ILI
tools because of the low operating pressure (<390 psig) or if there is any other reason it is not
piggable, MFL and CMFL tools could not be used to mitigate these threats. Note also, that
ECDA cannot be relied upon because of the external coating that shields the pipe from cathodic
protection when it becomes disbonded. Thus, without conducting periodic hydrostatic tests, the
operator cannot be confident that these threats are being mitigated.
The operator of this pipeline may have to do one of the following:

• Lower the operating pressure to reduce the threat of rupture from corrosion.
• Use alternative technology to assess for corrosion (e.g., a crawler-type tool).
• Take the pipeline out of service periodically for hydrostatic testing.
• Recoat the pipeline so that ECDA can be applied.
• Replace the pipeline.
cost to continue mitigating the threats of NAEC and external corrosion with the cost of
replacement of the pipeline in order to decide whether to repair or replace or exercise one of
the bullet-point options.
A Pipeline Comprised of Pipe with an HF-ERW Seam

Consider a 900-mile-long crude oil pipeline comprised of 16-inch-OD, 0.250-inch-wall, API

5LX52 line pipe with a HF-ERW seam. The pipeline was installed in 1965. It was coated over-
the-ditch with a coal tar enamel coating. An impressed-current cathodic protection system was
installed and activated within the first year of operation of the pipeline. The pipeline is
operated at an MOP of 1170 psig corresponding to a hoop stress level of 72% of SMYS. It can
accommodate all types of ILI. Note that this pipeline is very similar to the pipeline in Base Case

FINAL
15-019
Example 1. The only difference is that it is comprised of high-frequency-welded ERW pipe
made in 1965 instead of low-frequency-welded ERW pipe made in 1953.
In 1965 and up into the 1980’s much of the skelp rolled for ERW pipe was made using
conventional carbon-manganese steel with little or no sulfur control made in open-hearth
furnaces and cast into ingots. Moreover, the coil-leveling, can-forming, and resistance-welding
stands for making both HF-ERW and LF-ERW at the time were similar. Thus, the main (and
possibly the only) difference between HF-ERW and LF-ERW during the period from 1960 until
the complete conversion to basic-oxygen steel making in the 1980s was the nature of the weld
and heat-affected-zone. LF-ERW seams where characterized by large grain size from grain-
coarsening during welding, whereas with HF-ERW seams, the grain size tended to be much like
that in the plate material. As a result the bond line region of an HF-ERW material tends to
behave in a ductile manner when broken whereas that of an LF-ERW seam tends to behave in a
brittle manner when broken. In spite of this clear advantage, HF-ERW materials often were not
necessarily of better quality than LF-ERW materials when it came to hook cracks since the
sulfur contents did not change significantly until the 1980s. Hook cracks arise from
manganese-sulfide inclusions and sulfur contributes to the susceptibility of ERW pipe to SSWC.
Therefore, one must assume that early HF-ERW pipe is susceptible to the threats of PCIF and
SSWC. Once that assumption is made, the process for making a repair/replace decision is
identical to that presented in Base Case Example No. 1. Instead of being directed to Figure C2
(LEGACY PIPE) and Figure C3 (LF-ERW) for legacy pipe, however, the operator is directed to
Figure C8 (MODERN PIPE) and Figure C9 (HF-ERW). Flowchart HF-ERW is shown in Figure 28.

FINAL
15-019
Segment contains HF-ERW
pipe made prior to 1971.


Is periodically assessing seam Is it feasible to lower the needed to prevent seam
no yes
integrity by means of hydrostatic operating stress to a level failures in service and continue
testing or ILI feasible? that prevents seam failures in to operate in accordance with
service? Part 192 or Part 195.
yes
no
Assess seam integrity periodically Assess seam integrity periodically

based on the established interval based on the established interval to
PCIF.
or SSWC occurring between
re-assessments? operating stress until service replacement.
to occur?
no
yes

Figure 28. Flowchart for Evaluating a Repair/Replace Decision for HF-ERW Pipe in a
It is necessary to mitigate the threats of PCIF and SSWC in this pipeline just as it is in a pipeline
comprised of LF-ERW pipe because, the HF-ERW pipe made prior to 1971 was no less
susceptible to these phenomena as explained above.
cost to continue mitigating the threats of PCIF, SCC, and external corrosion with the cost of
Since HF-ERW pipe was rarely ever made in sizes above 24-inch, this case is not exactly the
same as Base Case Example No. 2. However, if one assumes that a 900-mile-long high-stress
natural gas pipeline is comprised of 16-inch-OD, 0.250-inch-wall, X52 HF-ERW pipe operated at
an MAOP of 1170 psig (72% of SMYS), then except for the pipe size and operating pressure
differences, the process for making a repair/replace decision is identical to that presented in
Base Case Example No. 2. Instead of being directed to Figure C2 (LEGACY PIPE) and Figure C3

FINAL
15-019
(LF-ERW) for legacy pipe, however, the operator is directed to Figure C8 (MODERN PIPE) and
Figure C9 (HF-ERW). The likelihood is that a pressure cycle fatigue analysis will show that this
pipeline is not likely to experience a failure from PCIF within the useful life of the pipeline.
However, the pipeline is possibly exposed to the threats of SCC and external corrosion. As
noted previously periodic integrity assessment is needed to mitigate both of these phenomena.
The operator could choose to mitigate SCC via periodic hydrostatic testing, ILI crack detection
tool runs, or possibly by means of SCCDA.
cost to continue mitigating the threats of SCC and external corrosion with the cost of
If one assumes that a 50-mile-long low-stress natural gas pipeline is comprised of 16-inch-OD,
0.250-inch-wall, X52 HF-ERW pipe operated at an MAOP of 485 psig (<30% of SMYS), then
except for the pipe size and operating pressure differences, the process for making a
repair/replace decision is identical to that presented in Base Case Example No. 3. Instead of
being directed to Figure C2 (LEGACY PIPE) and Figure C3 (LF-ERW) for legacy pipe, however,
the operator is directed to Figure C8 (MODERN PIPE) and Figure C9 (HF-ERW).
A Pipeline Comprised of Seamless Pipe

Consider a 200-mile-long refined products pipeline comprised of 12.75-inch-OD, 0.312-inch-

wall, API 5LX42 seamless line pipe. The pipeline was installed in 1955. It was coated over-the-
ditch with a coal tar enamel coating. An impressed-current cathodic protection system was
installed and activated within the first year of operation of the pipeline. The pipeline is
operated at an MOP of 1440 psig corresponding to a hoop stress level of 70% of SMYS. It can
accommodate all types of ILI. It is not exposed to internal corrosion because it transports only
refined products.
The pipeline was subjected to a pre-service hydrostatic test to a minimum pressure level of
1600 psig which is 1.11 times the MOP and corresponds to a hoop stress level of 78% of SMYS.
In 1979, the operator subjected the pipeline to another hydrostatic test to a minimum pressure

FINAL
15-019
level of 1800 psig which is 1.25 times the MOP and corresponds to a hoop stress level of 87.6%
of SMYS. The pipeline has exhibited no leaks or ruptures from any time-dependent cause.
In 2002, as part of the operator’s integrity management plan, the pipeline was subjected to an
integrity assessment for corrosion-caused metal loss using an MFL tool. Areas of external metal
loss were found, and those judged severe enough to be integrity-threatening were excavated
and examined. These areas were repaired if required and recoated. Some internal metal loss
anomalies were found as well. However, these few anomalies were judged to be typical
seamless pipe manufacturing anomalies that can be expected to remain stable. The reason
they were believed to be manufacturing anomalies and not internal corrosion-caused metal loss
is that they tended to be isolated and were oriented at random clock positions. These will be
monitored in subsequent ILI tool runs.
Subsequent ILI tool runs to assess for corrosion-caused metal loss were carried out in 2007 and
2012. Additional external metal loss was found each time and repaired as required. The areas
of internal metal loss seen on the first inspection in 2002 were again identified in 2007 and
2012. Significantly, no new areas of internal metal loss were identified, and as far as could be
determined, the ones identified in 2002 did not appear to have grown. The assumption that
these are non-threatened manufacturing defects appears justified.
Although no evidence of SCC has been discovered on the pipeline, the operator decided that
the threat should be addressed by means of an SCCDA program with periodic examinations of
selected sites.
Concerning the potential threat of failure from PCIF, although no fatigue failures have occurred,
the analysis of pressure cycles on the pipeline carried out in 2012 shows that the predicted
minimum time to failure after the 1979 hydrostatic test is 24 years (forecasting a potential
failure in 2003). The fact that no such failure from PCIF has occurred probably means one or a
combination of three things, 1) the worst-case defect that could have survived the 1979 test did
not exist or was not located in the worst-case location, 2) the pressure cycles have not always
been as aggressive as the spectrum used for the analysis, or 3) the crack growth rate constants
taken from API RP 579 are more conservative than the actual crack growth rate constants. To
address the unknown risk of failure from PCIF, the operator should consider either a hydrostatic
retest or an ILI crack-tool inspection as soon as possible and should also plan to repeat the test
or inspection at an interval of 12 years on the basis of the fatigue analysis showing 24 years as
the minimum time to failure. Assume that the operator elects to carry out an ILI crack-tool
inspection in 2015. The inspection may or may not result in the finding of cracks that may have
been extended by fatigue. Whatever the outcome, assume that the operator plans to repeat
the inspection every 12 years based on the predictions of the 2012 fatigue analysis.

FINAL
15-019
To make a repair/replace decision regarding this pipeline, the operator can utilize the flowcharts
of Appendix C. Figure C1 (Flowchart START), steers the operator to Figure C8 (Flowchart
MODERN PIPE), which in turn steers the operator to Figure C11 (Flowchart SEAMLESS).
Flowchart SEAMLESS addresses the issues related to the fact that this pipeline is comprised of
seamless pipe. Flowchart SEAMLESS is shown below as Figure 29.
Segment contains seamless

Use service and test failure history to

establish re-assessment interval to prevent
service failures from manufacturing defects.
No action is necessary if no such failures have
occurred, if the pipeline has been subjected
to a prior hydrostatic test, and if the
predicted times to failure base on that test
exceed the useful life of the pipeline.

Is periodic integrity assessment Is it feasible to lower the needed to prevent failures in
by means of hydrostatic testing service and continue to operate
or ILI feasible? that prevents failures in in accordance with Part 192 or
service? Part 195
yes
no
Assess integrity periodically based
on the established interval to
prevent service failures from
manufacturing defects
It is feasible to continue to
Service failures from decrease the re-assessment
yes intervals or to decrease the no
manufacturing defects are Schedule
occurring between re- operating stress until service replacement.
assessments failures from manufacturing
no defects cease to occur.
yes
Continue to operate and maintain Re-assess integrity at decreased

in accordance with Part 192 or interval or lower the operating
Part 195 and continue to perform stress until service failures from
integrity assessments on the manufacturing defects cease to
same schedule occur.
Figure 29. Flowchart for Evaluating a Repair/Replace Decision for Seamless Pipe in
a Hazardous Liquid Pipeline
The operator must also review Figure C12 (EC) and Figure C14 (SCC). As suggested in several
of the previous examples periodic ILI tool runs to monitor for metal loss and either periodic

FINAL
15-019
hydrostatic testing, periodic ILI crack-detection tool runs, or SCCDA are needed for mitigating
SCC.
If one assumes that a high-stress natural gas pipeline is comprised of the same seamless pipe
described above for a hazardous liquid pipeline, the required mitigation would be the same for
SCC and external corrosion. The likelihood is the pressure cycle spectrum of the gas pipeline
would not be sufficiently aggressive to cause PCIF to be a threat within the useful life of the
pipeline.
If one assumes that a low-stress natural gas pipeline is comprised of the same seamless pipe
described above for a hazardous liquid pipeline, the required mitigation would be the same for
external corrosion. The likelihood is the pressure cycle spectrum of the gas pipeline would not
be sufficiently aggressive to cause PCIF to be a threat within the useful life of the pipeline. The
operating stress level of the low-stress natural gas pipeline is below the threshold stress for
SCC. Since the low stress pipeline cannot accommodate ILI tools because of the low pressure
level, ECDA will have to be used to mitigate the threat from external corrosion.
A Pipeline Susceptible to Internal Corrosion

Any pipeline whether it is a hazardous liquid pipeline, a high-stress natural gas pipeline, or a
low-stress natural gas pipeline may be exposed to internal corrosion30. For a pipeline that is in
30
In compiling these responses to an internal corrosion threat, only onshore pipelines were considered. It is entirely possible that
the internal corrosion threat to an offshore pipeline could require analysis and solutions beyond the scope of these guidelines. It is
suggested that those interested in repair/replace decisions for offshore pipelines consult the numerous technical papers dealing with
the operation and maintenance of offshore pipelines.

FINAL
15-019
dedicated sour service with water present all of the time, the operator may have in place an
internal corrosion mitigation plan consisting of running cleaning pigs and injecting inhibitors
and/or biocides, and a monitoring plan that may involve periodic ILI tool runs to detect metal
loss. In the case of a pipeline that is normally not in sour service, sour gases or liquids may
enter the pipeline occasionally from upset conditions at gas treatment facilities. In such a case
the operator may not have in place a mitigation and monitoring plan, but it may be prudent to
assess the integrity of the pipeline from the standpoint of internal corrosion. The response
required to prevent leaks and ruptures from internal corrosion can be carried out as portrayed
in Appendix C. The START flowchart (Figure C1) directs the operator to either the LEGACY PIPE
flowchart (Figure C2) or the MODERN PIPE flowchart (Figure C8) either of which, in turn,
directs the operator to the IC flowchart. Flowchart IC is shown below as Figure 30.
Is the transported no Conduct hydrostatic tests at

product corrosive? appropriate intervals to prevent leaks
and ruptures from internal corrosion
yes Continue to operate and continue to operate in
segment in accordance accordance with Part 192 or Part 195.
with Part 192 or Part 195
yes
Is it feasible to prevent
Can the pipeline Can metal loss be leaks and ruptures
no no no
accomodate metal adequately assessed from corrosion using
loss ILI tools? and mitigated using periodic hydrostatic
ICDA? testing?
yes
Schedule
yes replacement.

yes
Conduct metal loss ILI inspections or

ICDA assessments at appropriate
intervals to prevent leaks and
ruptures from internal corrosion and
continue to operate segment in
accordance with Part 192 or Part 195.
Figure 30. Preventing Leaks and Ruptures from Internal Corrosion

FINAL
15-019
The process begins with the question “Is the transported product corrosive?”. The answer is yes
whether the pipeline is in dedicated sour service or is exposed on occasion to sour conditions.
If the pipeline can accommodate ILI metal loss tools, periodic assessments with an ILI tool
should be made using knowledge about the likely rate of metal loss. That case is represented
by the highlighted cells in Figure 30. If the pipeline cannot accommodate the appropriate ILI
tool, the operator can utilize ICDA to periodically monitor for internal metal loss. Alternatively,
the operator could assess the integrity of the pipeline via hydrostatic testing, though as
discussed previously, this approach is an inefficient method to assess integrity from the
standpoint of corrosion. If sour service is known to exist or to have existed and no form of
assessment for internal metal loss can be carried out, the operator should consider replacing
the pipeline. Alternatively it may be possible to increase the flow velocity to a level that is
above water entrainment velocity. The operator might also decide to replace a pipeline if the
cost of mitigating and/or responding to internal corrosion and possibly other threats as well
becomes too high relative to the cost of replacement. The replacement pipeline could be
constructed of a material more resistant to sour conditions, and the operator can plan for
appropriate monitoring and inspection in the design phase.
A Pipeline Containing Legacy Repairs

In the past, defects in pipelines were often repaired by means of puddle welds, patches, or half
soles. Puddle-welding in this context does not include the class of deposited-weld-metal repairs
that follow specific, proven procedures. The old-style puddle weld was usually made in one of
two circumstances. In one circumstance, puddle welding was carried out on in-service pipelines
with little or no procedural documentation. These welds frequently have been accompanied by
crack formation and, in some cases, eventual leakage. In the second circumstance, they were
used to fill pits in corroded pipe which had been removed from service for re-installation in the
same or a different pipeline. In the latter circumstance, the puddle welds are usually of
acceptable quality and have not been known to be a frequent source of leaks.
Patch repairs generally consisted of rectangular, curved pipe segments usually no greater in
size than 6-inches in either direction that were fillet welded over a defect. The longitudinal
welds are acted upon by hoop stress, and thus they can serve as fatigue crack initiators
particularly if under-bead cracks formed from too-rapid cooling and the use of non-low-
hydrogen consumables. In cases where patches were installed on liquid-filled pipelines or
pipelines containing high-pressure flowing gas, the risk of under-bead crack formation was
especially high 31. The longitudinal fillet welds on patches have been shown to be more
31
The same threat could be associated with Type B full-encirclement sleeves where the ends of the sleeves are fillet welded to the
pipe. If leaks or ruptures have occurred in a given pipeline in conjunction with Type B sleeves, the operator should consider
repairing or removing any sleeves made with the same procedures used on sleeves which have caused leaks or ruptures in the past.

FINAL
15-019
vulnerable to failure when installed over leaks where the patches were exposed to operating
pressure fluctuations.
Half soles are nothing more than a large patch, usually made from half of a full-encirclement
sleeve. Half soles have all of the undesirable properties of patches and are now barred from
use in pipelines operated in conformance with ASME B31.4 and ASME B31.8.
When considering whether to replace such legacy repairs, a pipeline operator may use the
flowcharts of Appendix C. The START flowchart (Figure C1) directs the operator to either the
LEGACY PIPE flowchart (Figure C2) or the MODERN PIPE flowchart (Figure C8) which, in turn,
directs the operator to the LEGACY REPAIRS flowchart (Figure C7). Figure C7 is reproduced as
Figure 31 below.

FINAL
15-019
Segment contains
puddle welds*,
patches, or half soles.
Is the segment
Continue to operate in
operated at a stress no
level greater than 20% accordance with Part
192 or Part 195.
of SMYS?
yes
Were the puddle

welds made while the
pipe was in service? no no yes
Were patches or half
Were patches or half no soles installed while the yes
Was a low-hydrogen
yes soles installed over a pipe was either full of
welding procedure used?
leaking defect? liquid or high-pressure,
flowing gas?
yes no
Can the segment be Conduct hydrostatic

Conduct a "one-time" hydrostatic
test in test, make
Can the puddle welds, no subjected to a yes no further legacy repairs after the test, and
and continue to operate in
patches, or half soles continue to with
operate
hydrostatic test in accordance Partin192
accordance
or with Part
be located by ILI? accordance with Part 192 or Part 195 as long as no additional
Part 195.
192 or Part 195? failures due to legacy repairs occur.
yes
no
Repair or replace If failures of legacy repairs continue to
puddle welds, occur after the test, schedule
patches, or half soles Schedule replacement
of pipeline or lower replacement of pipeline or lower
and continue to operating stress to 20% of SMYS or less.
operating stress to
operate in accordance
20% of SMYS or less.
with Part 192 or Part
195.
*Puddle welding refers to welds made with an unknown welding proceddure to repair pits or other defects.
Deposited weld metal repairs made according to a documented and proven procedure are a legitimate
repair method and do not fall under this definition of puddle welding.
Figure 31. Process for Responding to Legacy Repairs
Figure 31 illustrates for any pipeline that can accommodate a MFL or CMFL tool, the procedure
one can follow to determine how to respond to legacy repairs. First, legacy repairs are
considered to be of concern if the pipeline is operated at a stress level in excess of 20% of
SMYS. In the particular pipeline considered in Figure 31, there are no puddle welds, only
patches and half soles. Patches or half soles installed over leaks are slated for mitigative action
which in this case consists of finding the patches or half soles by means of ILI and repairing or
removing them. Note that it may or may not be possible to tell whether a patch or half sole

FINAL
15-019
had been installed over a leak. Also note that any puddle welds that had been made while the
pipeline was in service would be slated for mitigative action as well.
If the legacy repairs are slated for repair or removal and ILI cannot be used to locate them, the
operator has the option of performing a ”one-time” pressure test 32 in conformance with the
federal regulations. If the pipeline cannot be pressure-tested or if failures of patches or half
soles continue to occur after the one-time test, replacement of the whole pipeline or a reduction
to an operating stress level of 20% of SMYS or less is recommended. .
If there are patches or half soles that have not been installed over leaks and were installed
while the pipeline was full of liquid or high-pressure, flowing gas using a low-hydrogen welding
procedure, there is no need to remove them. Similarly, if the patches or half soles had been
installed on an empty pipeline, there is no need to remove them. However, patches and half
soles installed while the pipeline was full of liquid or high-pressure, flowing gas with a non-low-
hydrogen welding procedure should be slated for removal.
If the example pipeline contained puddle welds instead of or in addition to patches and half
soles, no action is required if the puddle welds were fabricated with the pipe out of service. On
the other hand, puddle welds fabricated while the pipeline contained a liquid product or high-
pressure, flowing gas should be located by ILI is feasible and repaired or the affected pipe
should be replaced. Alternatively, if locating the puddle welds by means of ILI is not feasible,
the pipeline should be subjected to a one-time hydrostatic test to locate weak spots associated
with the puddle welds.
As in the cases of other threats discussed previously, where legacy repairs are an issue (in
conjunction with other needs for threat mitigation), the operator is likely to choose between
repair and replacement based on comparing the costs of the two options, taking into account
the cost of maintaining safety and integrity of the pre-regulation pipeline.
A Pipeline Susceptible to HSC in Pipe Body Hardspots

Suppose that the natural gas pipeline described in Table 5 had sustained a failure from a pipe
body hard spot. Such hard spots have been detected in A.O. Smith flash-welded pipe made in
the period between 1947 and 1960. Assume that every other aspect of this case is the same as
described in Base Case 2. The only difference to be considered is the threat of HSC in pipe
32
The idea of a one-time test to address patches, half-soles, or puddle welds is to rid the pipeline of possible weak spots associated
with these types of repairs. If leakage from such repairs continues to occur, the operator should consider replacing the pipeline.
Repeated hydrostatic testing is likely to make leakage of such repairs worse because they are particularly susceptible to pressure
reversals. One of the authors discovered this the hard way by trying to retest pipe specimens with multiple defects using patches
and puddle welds to repair leaks rather than cut them out.

FINAL
15-019
body hard spots. The factors that would influence a repair/replace decision for this pipeline
would be the same as those in Base Case 2 except that operator in this case needs to consider
the threat of failure from HSC as well. That can be done using Appendix C Figure C15 HSC
which is reproduced below as Figure 32 for this particular pipeline.
Is segment comprised of Continue to operator in

line pipe manufactured no
between 1947 and 1960? Part 195. Schedule replacement
yes no
Has a pipe body hard spot ever no
Was pipe manufactured by caused a leak or a rupture or Can segment accomodate MFL
no has a hard spot ever been yes
Youngstown Sheet and tools or, if not, can it be made
Tube Conpany? found due to other inspection to accommodate MFL tools?
actives?
yes
Use special "residual field" MFL

tool to find and repair hard spots
yes and continue to operate in
accordance with Part 192 or Part
195
Has any failure occurred that Can pipe-to-soil potentials be
was attributed to HSC in the yes controlled in a manner that
heat-affected zone of the minimizes hydrogen generation
ERW seam? but still adequately mitigates no
corrosion?
no
Have metallurgical studies yes
and measurements of
alloy contents been made no
that suggest the likelihood
of susceptibility to HSC is
low? Control pipe-to-soil potentials in a
manner that minimizes the generation
yes of atomic hydrogen but still adequately
mitigates corrosion and continue to
Continue to operate in operate in accordance with Part 192 or
accordance with Part 192 Part 195.
or Part 195
Figure 32. Process for Responding to an HSC Threat in a Pipe Body Hard Spot
As can be seen in Figure 32, the fact that a failure of a pipe body hard spot has occurred means
that the operator must address the threat of HSC. Fortunately, the pipeline can accommodate
the type of MFL tool that can be used to locate pipe body hard spots. The operator should use
such a tool to locate the hard spots. Every hard spot should be excavated and examined. Local
hardness measurements can be made to verify the hardness levels. In addition, hard spots are
sometimes visible because the hard region has less curvature than the unaffected pipe due to
its higher yield strength. The pipe containing the hard spots may be removed and replaced
with sound pipe. Alternatively, full-encirclement steel sleeves (not composite wraps) should be

FINAL
15-019
installed over the hard spots to shield them from cathodic protection. This prevents atomic
hydrogen from being generated at the hard surface. This can be a one-time repair program
because it will eliminate the threat. Re-assessment for HSC is not necessary if all of the hard
spots are removed or repaired. In this case the hard spot problem does not lead to pipeline
replacement. However, if the operator is unable to run the appropriate tool to find the hard
spots and eliminate the threat, replacement of the pipeline must be considered.
A Pipeline Susceptible to HSC in the HAZ of its DC-ERW Seam

Suppose that the hazardous pipeline described in Table 4 was comprised of 1953-vintage DC-
ERW pipe manufactured by Youngstown Sheet and Tube Company. Pieces of pipe made by this
manufacturer in the period between 1947 and 1960 have been known to have heat-affected-
zone hardnesses of Rockwell C35 or higher, making them susceptible to HSC. Assume that
every other aspect of this case is the same as described in Base Case 1. The only difference to
be considered is the threat of HSC in heat-affected zones adjacent to the ERW seams. The
factors that would influence a repair/replace decision for this pipeline would be the same as
those in Base Case 1 except that the operator in this case needs to consider the threat of failure
from HSC as well. That can be done using Appendix C Figure C15 HSC which is reproduced
below as Figure 33 for this particular pipeline.

FINAL
15-019

accordance with Part 192 or Schedule replacement
between 1947 and 1960? Part 195.
yes no
actives?
yes
yes
Use special "residual field" MFL
tool to find and repair hard spots
and continue to operate in
accordance with Part 192 or Part
Has any failure occurred that Can pipe-to-soil potentials be
corrosion?
no
and measurements of
or Part 195
Figure 33. Process for Responding to an HSC Threat in a Hard HAZ
Because the pipe was manufactured by Youngstown Sheet and Tube Company during the
critical time period when hard heat-affected zones could have been produced, the operator
must take mitigative action even if no failure from a hard heat-affected-zone crack has
occurred. One way to do this is to examine records of chemical analyses of the heats of steel
(if available) to ascertain whether or not the carbon and manganese contents were sufficiently
high to make the material easily hardenable. Alternatively, the operator may have the
opportunity to remove a significant number of metallurgical samples to verify that high
hardness levels do not exist. If no failure from HSC in a hard heat-affected zone has occurred,
and the material does not seem to be readily hardenable, the operator may be able to avoid
any further remedial action. If the non-susceptibility to HSC cannot be established with a high
level of confidence, then further remedial action is needed.

FINAL
15-019
If a failure from HSC has occurred, or if the non-susceptibility of the pipe cannot be established,
the operator should modify the cathodic protection system by adding ground beds if necessary
to control the range of pipe-to-soil potentials, in order to achieve adequate cathodic protection
without having excessively high pipe-to-soil potential levels that could lead to aggressive
hydrogen charging of the pipe in areas of disbonded coating. Alternately, the operator could
consider using a less conservative cathodic protection criterion like the 100 mv polarization
criteria which uses less polarization than the -0.85-volt criterion. This will help reduce the
potential level for hydrogen charging. If neither of these prove effective at mitigating the HSC
threat, the operator should consider replacing the pipeline.
Flowcharts for Gas Distribution Systems

The flowcharts for making repair/replace decisions for gas distribution systems are presented in
Appendix C, Figures C16 through C27. These figures cover the three types of pipe materials
used in most gas distribution systems: steel, plastic, and cast iron.
System Attributes, Operational Data, and History

The system attributes that are essential for making repair/replace decisions for steel, plastic, or
cast iron pipe systems were presented previously in Tables 6-8 for the “base case” distribution
systems. The various operational data, and historical operating experience that bear on
repair/replace decision-making were discussed previously. The decision-making criteria are
demonstrated below by means of examples.
Base Case Example No. 4: A Steel Gas Distribution Main

Consider the steel natural gas distribution main described in Table 6. It is comprised of 4.5-
inch-OD, 0.188-inch-wall, Grade A pipe manufactured and installed in 1942 with a low-
frequency electric resistance welded seam. Each piece of pipe was hydrostatically tested by the
manufacturer to a stress level of 60% of SMYS for a total time of 10 seconds. The distribution
main is 10 miles long and is operated at a maximum hoop stress level of 5.98% of SMYS. The
MAOP of the pipeline is 150 psig. The main was installed uncoated with no cathodic protection.
operation since 1942. The leak history indicates that leaks from external corrosion have been
generally increasing since 1953. The main also contains 50 drips and three dead-legs that have
the potential to pose issues with internal corrosion, even though the line normally carries dry
natural gas with no sour components, and no leaks due to this threat have occurred.
There is no record of a post construction pressure test for this main. Since the distribution
main is directly connected to the end users through hundreds of service connections, service

FINAL
15-019
cannot be interrupted. The main cannot be taken out of service to perform a pressure test, and
it cannot accommodate an in-line inspection tool.
To determine whether or not this main should be replaced, the operator can go through the
process described below starting with Flowchart START (for steel gas distribution main) (Figure
C16 of Appendix C). Flowchart START (for steel gas distribution main) is repeated as Figure 34
below as adapted to this particular pipeline.
no
Is information (i.e. pipe Can information be
material and properties, no
established or verified?
type of construction,
operating conditions, Identify all unknown
yes
threats, etc.) regarding the information and develop a
steel main known? plan for gaining this
Establish or verify information in accordance
information. with Part 192, Subpart P
yes
(DIMP), section 1007(a)(3)
and as recommended by
GPTC DIMP sections 3.1(d)
to (f).
Continue to operate and
Are any of the following
no maintain segment in
threats causing problems on
accordance with Part 192,
the segment?
Subpart P (DIMP).
yes
Selective seam weld Proceed to Flowchart SSWC

corrosion (SSWC). for steel gas distribution
main.
External corrosion (EC). Proceed to Flowchart EC for

steel gas distribution main.
Internal corrosion (IC). Proceed to Flowchart IC for

Proceed to Flowchart
Threaded underground
LEGACY FEATURES for steel
joints.
gas distribution main.
Steel Gas Distribution Main

FINAL
15-019
The information regarding this main is known. If some or all of the vital information such as
diameter, wall thickness, seam type, or specified minimum yield strength were not known, the
operator would have to identify all unknown information and develop a plan for gaining this
information in accordance with Part 192, Subpart P (DIMP), section 1007(a)(3) and as
recommended by GPTC DIMP 33 sections 3.1(d) to (f).
The segment may be susceptible to selective seam weld corrosion (SSWC), so the next step is
to proceed to Figure 35 which is Flowchart SSWC (for steel gas distribution main) (Figure C17
of Appendix C). Note that types of pipe known to be susceptible to SSWC are Electric Flash
Welded, Low-Frequency Electric Resistance Welded (LF-ERW), and High-Frequency Electric
Resistance Welded (HF-ERW) pipe manufactured prior to 1980, or where the toughness of the
seam region was low or not specified.
Any steel distribution main may be susceptible to degradation from external or internal
corrosion; therefore the operator should assess the segment for these threats accordingly. So,
after assessing the segment for a SSWC threat, the operator of the pipeline must pursue the
process of making a repair/replace decision via Flowcharts EC (for steel gas distribution main)
and IC (for steel gas distribution main) shown in Figure 36 (Figure C18 of Appendix C) and
Figure 37 (Figure C19 of Appendix C), respectively.
This base case pipeline contains no threaded underground joints or other obsolete features, so
the operator of this pipeline need not consider the threats represented in Flowchart LEGACY
FEATURES (for steel gas distribution main) (Figure C20 of Appendix C).
33
GPTC DIMP guidelines are promulgated and maintained by the Gas Piping Technology Committee of the American Gas
Association.

FINAL
15-019
Seam types susceptible to Apply one or a combination

SSWC are: of the following actions:
- Electric Flash Welded -> Increase the frequency of
patrolling and leak surveys
- Low-Frequency Electric and monitor the number of
Resistance Welded (LF-ERW) leaks occurring on the
- High-Frequency Electric segment as well as the leak
Resistance Welded (HF-ERW) trends.
manufactured prior to 1980, -> Manage the leaks
or where the toughness of occurring on the segment
the material was low or not based on risk and in
specified. accordance with the
operator's leak management
Continue to operate and program.
maintain segment in
Is the seam type of the accordance with Part 192, -> Repair leaks that cannot
no
segment susceptible to Subpart P (DIMP). Use new be managed.
SSWC? information gained to -> Schedule replacement of
continue to monitor the entire segment.
yes segment performance.
-> See GPTC DIMP section
Have there been leaks from 6.3, Table 6.1, for additional
no
SSWC? suggestions.
yes
Figure 35. Feasibility of Preventing in-Service Failures from SSWC (for steel gas
distribution main)
According to Figure 35, since the main is comprised of LF-ERW pipe, which is susceptible to
SSWC, the operator should review the history of in-service failures to determine whether any of
the leaks that occurred on the line were due to SSWC. In this particular case, some of the
leaks experienced were found to have occurred in the longitudinal seam of the pipe and thus
were attributed to SSWC. The flowchart suggests that the operator take some type of
additional or accelerated (A/A) action such as those suggested in GPTC DIMP Table 6.1 in order
to appropriately manage the threat and minimize the risk. The A/A actions in this table are not
meant to be restrictive, but instead are meant to serve as suggestions. The operator is free to
take any action that helps reduce the risk of the SSWC threat to the segment. After an action is
applied, the operator should use any new knowledge it gained through this process to monitor
the performance of the main and operate and maintain the segment in accordance with Part
192, Subpart P and the operator’s Distribution Integrity Management Program (DIMP). After
the threat of in-service failures attributed to SSWC has been addressed, the operator should
refer back to Flowchart START (for steel gas distribution main) and assess the other threats to
which the main is susceptible.

FINAL
15-019
Returning to Flowchart START (for steel gas distribution main), the operator of this main should
assess the next applicable threat, which in this case happens to be external corrosion.
Flowchart EC, shown in Figure 36, should be used to analyze this threat.
no
Does segment have an Is the segment experiencing Apply one or a combination
yes
external coating? metal loss and/or leaks of the following actions:
attributed to external
-> Increase the frequency of
yes corrosion?
no no
and monitor the number of
Is the external coating leaks occurring on the
adequate? segment as well as the leak
trends.
yes
-> Manage the leaks
Does the segment have a occurring on the segment
cathodic protection (CP)
no based on risk and in
system? accordance with the
program.
yes
-> Repair leaks that cannot
no
Is the CP system adequate? be managed.
maintain segment in -> Schedule replacement of
accordance with Part 192, the entire segment.
yes
Subpart P (DIMP). Use new
information gained to
6.3, Table 6.1, for additional
continue to monitor
segment performance. suggestions.
yes
Is the trend of leaks due to

external corrosion on the
segment increasing?
no
Increase the frequency of Is the trend of leaks due to

yes
patrolling and leak surveys external corrosion on the
and continue to monitor the segment remaining the
leak trends. same?
no
Patrol the line and complete

The trend of leaks due to
leak surveys at required
intervals and continue to external corrosion on the
segment is decreasing.
monitor the leak trends.
Figure 36. Feasibility of Preventing in-Service Failures from External Corrosion (for
steel gas distribution main)

FINAL
15-019
According to Figure 36, this main does not have an external coating and the operator has
confirmed it is experiencing leaks attributed to external corrosion. The operator should take an
A/A action such as installing sacrificial anodes 34, increasing the frequency of its patrolling and
leak surveys, while also managing the leaks based on risk and in accordance with its leak
management program. After the leaks posing the most risk are addressed, it is noted that the
leak trend along the segment is decreasing. The operator should continue to patrol the line and
complete leak surveys at required intervals, and should continue to operate and maintain the
segment in accordance with the DIMP regulations. Any new information gained throughout the
DIMP process such as updated leak trends should be used to continue to monitor the segment’s
performance. Had the trend of external corrosion leaks in this particular case remained the
same or continued to increase, the operator would need to reassess the situation and take
additional action to manage the leaks occurring along the line. Continually increasing leak
trends despite A/A actions and mitigation efforts would lead the operator to determine that
replacement of the main might be the best option due to the corrosion issue being too severe
to feasibly mitigate.
Again returning to Flowchart START (for steel gas distribution main), the operator of this main
should assess the last applicable threat, which is internal corrosion. Flowchart IC (for steel gas
distribution main), shown in Figure 37, should be used to analyze this threat.
34
When sacrificial anodes are installed, it is important to record their locations and desirable to equip them with a shut-off
mechanism to facilitate on/off pipe-to-soil potential measurements.

FINAL
15-019
Has there ever been metal In the course of normal Apply one or a combination
loss in conjunction with operations, if presented with of the following actions:
yes
drips, dead-legs, and/or the opportunity, inspect the
interior surface of the pipe -> Increase the frequency of
areas of water
for internal corrosion. patrolling and leak surveys
accumulation, such as
downstream of a regulator,
in the system? leaks occurring on the
segment as well as the leak
no trends.
-> Manage the leaks
Is the segment experiencing occurring on the segment
Is the transported product yes metal loss and/or leaks yes based on risk and in
corrosive? attributed to internal accordance with the
corrosion? operator's leak management
no
program.
no
Continue to operate and be managed.
maintain segment in
accordance with Part 192, -> Schedule replacement of
Subpart P (DIMP). Use new the entire segment.
information gained to -> See GPTC DIMP section
continue to monitor 6.3, Table 6.1, for additional
yes

internal corrosion on the
segment increasing?
no

patrolling and leak surveys yes internal corrosion on the
leak trends. same?
no

intervals and continue to internal corrosion on the
Figure 37. Feasibility of Preventing in-Service Failures from Internal Corrosion (for
steel gas distribution main)
According to Figure 37, while this main has both drips and dead-legs that exist in the system,
the operator has confirmed that no metal loss in conjunction with these features has occurred.
During the course of normal operations, if presented with the opportunity such as during a
repair, the operator should inspect the interior surface of the main for signs of internal
corrosion and/or water accumulation. Water should be drained from the system, if found, and
the segment should be cleaned out, if possible. If pitting is discovered, the severity needs to be
assessed, and if leakage is imminent, it should be managed in an acceptable manner. Since the

FINAL
15-019
product transported in this main is not corrosive and none of the leaks occurring along the main
can be attributed to internal corrosion, the operator can continue to operate and maintain the
segment in accordance with the DIMP regulations. Had the transported product been corrosive
and the main been experiencing leaks attributed to internal corrosion, the operator would need
to take an A/A action to manage the leaks along the line as well as monitor the leak trends.
The operator’s capability of addressing internal corrosion issues as well as the costs associated
would determine whether the internal corrosion issues could be managed or if replacement
would be necessary.
The above analysis indicates that the threats to the integrity of this main are being mitigated
through the applied A/A actions being taken by the operator. Judging whether the operator
should continue to apply these A/A actions and continue to operate the main going forward
versus replacing it can be decided based upon the cost to maintain the integrity of the existing
pipeline versus the cost to replace it. The cost to maintain and reassess the main to a level
required for preventing in-service leaks depends on the costs of the various A/A actions
discussed above. If these continuing costs are acceptable to the operator, then replacement is
not necessary.
Base Case Example No. 5: A Plastic Gas Distribution System

Consider the plastic natural gas distribution system piping described in Table 7. It is comprised
of 2.375-inch-OD, 0.216-inch-wall, polyethylene (PE) pipe manufactured by Century Utilities
Products, Inc. and installed in 1976 with butt heat-fusion joints. Each piece of pipe had a
hydrostatic design basis (HDB) at 73°F of 1250 psi. The distribution system consists of 15 miles
of the aforementioned piping. The MAOP of the pipeline is 60 psig.
operation since 1976. The leak history indicates that leaks due to slow crack growth have been
increasing since 1989. The leak history also mentions a number of leaks experienced at joint or
tap connections in 1979, shortly after the system was put into service, however after repairs
were made no further failures of this type have been experienced.
It is assumed that the main was tested to 1.5 times the MAOP. Since the distribution system
piping is directly connected to the end users through hundreds of service connections, service
cannot be interrupted. The main cannot be taken out of service to perform a pressure test, and
it cannot accommodate an in-line inspection tool.
To determine whether or not this piping should be replaced, the operator can go through the
process described below starting with Flowchart START (for plastic gas distribution main)
(Figure C21 of Appendix C). Flowchart START (for plastic gas distribution main) is repeated as
Figure 38 below as adapted to this particular pipeline.

FINAL
15-019
no
material and properties, established or verified?
type of construction, no
yes
threats, etc.)
yes regarding the
information and develop a
plastic distribution system plan for gaining this
known? information. with Part 192, Subpart P
yes and as recommended by
to (f).
the segment?
Subpart P (DIMP).
yes
Slow crack growth (SCG). Proceed to Flowchart SCG.
Fusion/mechanical joining or
service tap connection type
PLASTIC CONNECTIONS.
failures.
involving a Plastic Distribution System
The essential attributes of this main are known. If some or all of the vital information such as
diameter, wall thickness, seam type, or specified minimum yield strength were not known, the
operator would have to identify all unknown information and develop a plan for gaining this
information in accordance with Part 192, Subpart P (DIMP), section 1007(a)(3) and as
recommended by GPTC DIMP sections 3.1(d) to (f).
The segment has been experiencing leaks due to slow crack growth (SCG). So, the next step is
to proceed to Figure 39 which is Flowchart SCG (Figure C22 of Appendix C).
Any plastic distribution system may be susceptible to joining or tap connection failures such as
a bad fusion joint or a bad bolt-on type of tee; therefore, after assessing the segment for an
SCG threat, the operator should proceed to Flowchart PLASTIC CONNECTIONS (Figure C23 of
Appendix C), shown in Figure 40, to assess the segment for a joining or tap connection type
threat accordingly.

FINAL
15-019
Materials with known Is the material of which the
no
susceptibility to premature segment is constructed
brittle-like cracking (i.e. SCG) susceptible to SCG?
under conditions of local
stress intensification are: yes Continue to operate and
- Century Utility Products, maintain segment in
Have there been leaks accordance with Part 192,
Inc. products; no
attributed to SCG? Subpart P (DIMP). Use new
- Pipe extruded by Century information gained to
Utility Products Inc. from continue to monitor
Union Carbide Corporation's yes segment performance.
DHDA 2077 Tan medium
density polyethylene resin; Apply one or a combination
- Low-ductile inner wall of the following actions:
"Aldyl A" piping -> Increase the frequency of Patrol the line and complete
manufactured by DuPont patrolling and leak surveys The trend of leaks due to
Company prior to 1973; and monitor the number of SCG on the segment is intervals and continue to
leaks occurring on the decreasing.
-Polyethylene gas pipe monitor the leak trends.
designated PE 3306; segment as well as the leak
trends. no
- Delrin insert tap tees;
-> Manage the leaks Is the trend of leaks due to Increase the frequency of
- Plexco service tee Celcon occurring on the segment yes
SCG on the segment patrolling and leak surveys
(polyacetal) caps; based on risk and in remaining the same? and continue to monitor the
- Plastic pipe installed in accordance with the leak trends.
natural gas distribution operator's leak management
systems from the 1960s program.
through the early 1980s. no
be managed. Is the trend of leaks due to
SCG on the segment
-> Schedule replacement of
increasing?
the entire segment.
-> See GPTC DIMP section yes
suggestions.
Figure 39. Feasibility of Preventing in-Service Failures from Slow Crack Growth
According to Figure 39, the operator should first determine whether the distribution system was
constructed of a pipe material known to be susceptible to premature brittle-like cracking (i.e.
SCG) under conditions of local stress intensification. An increase in the occurrence of leaks will
typically be the first indication of a SCG problem, and this condition can substantially reduce the
service life of polyethylene piping systems. In this particular case, the operator discovered that
the system piping was manufactured by Century Utility Products, Inc. which has been identified
in a PHMSA advisory bulletin as an older polyethylene pipe material that is susceptible to SCG.
A majority of the system leaks experienced were found to be attributed to SCG. The flowchart
suggests that the operator take some type of additional or accelerated (A/A) action such as
those suggested in GPTC DIMP Table 6.1 in order to appropriately manage the threat and
minimize the risk. The A/A actions in this table are not meant to be restrictive, but instead are
meant to serve as suggestions. The operator is free to take any action that helps reduce the
risk of the SCG threat to the system.
After an action is applied, the operator should monitor the trend of leaks occurring due to SCG
to determine if they are increasing, decreasing, or remaining the same. In this particular case,

FINAL
15-019
after the operator applied an A/A action the trend of leaks began to decrease signifying that the
A/A action was effective. Had the trend stayed the same or continued to increase, the
flowchart would direct the operator to implement another A/A action in a further effort to
manage the leaks still occurring. After a decline in the leak trend is seen, the operator should
continue to implement the A/A action found to be most effective at mitigating the leaks, and
use any new knowledge it gained to monitor the performance of the main. Going forward, the
operator should continue to operate and maintain the segment in accordance with Part 192,
Subpart P and with the operator’s Distribution Integrity Management Program (DIMP).
Additionally, the operator should be resolute in determining the source of the local stress
intensification, which caused the leaks, in order to prevent similar leaks from occurring in the
future. After the threat of in-service failures attributed to SCG has been addressed, the
operator should refer back to Flowchart START (for plastic gas distribution main) and assess the
other threats to which the system is susceptible.
Returning to Flowchart START (for plastic gas distribution main), the operator of this main
should assess the other applicable threat, which is joining or tap connection type failures.
Flowchart PLASTIC CONNECTIONS, shown in Figure 40, should be used to analyze this threat.
Segment contains fusion Apply one or a combination

joints, mechanical joints, of the following actions:
and/or service tap
connections.
leaks occurring on the
Is the segment experiencing trends.
leaks and/or failures
attributed to pipe joining -> Manage the leaks
practices (e.g. a bad fusion yes occurring on the segment
joint) or service tap based on risk and in
connection issues (e.g. a bad accordance with the
bolt on type of tee, or a operator's leak management
broken tee cap)? program.
no
be managed.
the entire segment.
maintain segment in
accordance with Part 192, -> See GPTC DIMP section
Subpart P (DIMP). Use new 6.3, Table 6.1, for additional
information gained to suggestions.
continue to monitor
segment performance.
Figure 40. Feasibility of Preventing in-Service Failures from Joining or Tap

Connection type Failures in a Plastic Main

FINAL
15-019
According to Figure 40, this distribution system contains fusion or mechanical joints and/or
service tap connections; however the system is not currently experiencing any issues with these
types of connections (e.g. a bad fusion joint, or a bad bolt on tee). Since the system is not
experiencing leaks due to this threat, the operator is able to continue to operate and maintain
the segment in accordance with the DIMP regulations. Had the system been experiencing leaks
attributed to joining or connection type issues, the operator would need to take an A/A action
to manage the leaks along the line to maintain integrity and to ensure safety. The operator’s
capability of addressing these connection type issues as well as the costs associated with
remediation would determine whether they could be properly managed or if replacement would
be necessary.
The above analysis indicates that the threats to the integrity of the distribution system
described are being mitigated through the applied A/A actions being taken by the operator.
Judging whether the operator should continue to apply these A/A actions and continue to
operate the system going forward versus replacing it can be decided on a basis that includes
the cost of maintaining pipeline integrity and safety. The cost to maintain and reassess the
system to a level required for mitigating risk to assure safety depends on the costs of the
various A/A actions discussed above. If these continuing costs are acceptable to the operator,
Base Case Example No. 6: A Cast Iron Gas Distribution Main

Consider the cast iron natural gas distribution main described in Table 8. It is comprised of
4.80-inch-OD, 0.48-inch-wall, Class A pipe manufactured and installed in 1925. The distribution
main is approximately 9.5 miles long. The MAOP of the pipeline is 25 psig.
The operator has no records documenting the pipeline’s design and construction in 1925.
However, operating records show that all pipe joints along this main have been sealed, and no
leaks have been experienced after the sealing process was complete. The leak history indicates
that leaks due to graphitization and breakage have been generally increasing since 1967.
There is no record of a post construction pressure test for this main. Since the distribution
main is directly connected to the end users through hundreds of service connections, service
cannot be interrupted. For this reason, the main cannot be taken out of service to perform a
pressure test, and it cannot accommodate an in-line inspection tool.
To determine whether or not this main should be replaced, the operator can go through the
process described below starting with Flowchart START (for cast iron distribution main) (Figure
C24 of Appendix C). Flowchart START (for cast iron distribution main) is repeated as Figure 41
below as adapted to this particular pipeline.

FINAL
15-019
no
yes
cast iron main known? plan for gaining this
yes (DIMP), section 1007(a)(3)
to (f).
the segment?
Subpart P (DIMP).
yes
Joint leakage.
Proceed to Flowchart CAST
IRON CONNECTIONS.
Pipe body leakage due to Proceed to Flowchart

graphitization. GRAPHITIZATION.
Breakage due to external Proceed to Flowchart

forces. BREAKAGE.
involving a Cast Iron Distribution Main
The diameter, thickness, and type of material of this main are known because of the earlier
joint-sealing work. Thus, even though the operator does not have original design and
construction records, the susceptibility to known integrity threats can be assessed. If diameter,
wall thickness, seam type, or specified minimum yield strength were not known, the operator
would have to identify all unknown information and develop a plan for gaining this information
in accordance with Part 192, Subpart P (DIMP), section 1007(a)(3) and as recommended by
GPTC DIMP sections 3.1(d) to (f).
The segment has been experiencing leaks due to graphitization. So, the next step is to proceed
to Figure 42 which is Flowchart GRAPHITIZATION (Figure C26 of Appendix C).

FINAL
15-019
Any cast iron distribution main may be susceptible to breakage from external forces such as soil
movement; therefore, after assessing the segment for a graphitization threat, the operator
should proceed to Flowchart BREAKAGE (Figure C27 of Appendix C), shown in Figure 43, to
assess the segment for a breakage threat accordingly.
This base case pipeline is not experiencing leaks in any of its joint connections, so the operator
of this pipeline need not consider the threats represented in Flowchart CAST IRON
CONNECTIONS (Figure C25 of Appendix C).

FINAL
15-019
Is the segment experiencing Apply one or a combination

yes
leaks and/or failures of the following actions:
attributed to graphitization?
no
trends.
-> Manage the leaks
occurring on the segment
based on risk and in
accordance with the
program.
be managed.
continue to monitor
yes

graphitization on the
segment increasing?
no

yes
patrolling and leak surveys graphitization on the
leak trends. same?
no

intervals and continue to graphitization on the
Figure 42. Feasibility of Preventing in-Service Failures from Graphitization
According to Figure 42, the operator has confirmed that the main is experiencing leaks due to
graphitization. The flowchart suggests that the operator take some type of additional or
accelerated (A/A) action such as those suggested in GPTC DIMP Table 6.1 in order to
appropriately manage the threat and minimize the risk. The A/A actions in this table are not
meant to be restrictive, but instead are meant to serve as suggestions. The operator is free to

FINAL
15-019
take any action that helps reduce the risk of the graphitization threat to the system. After
taking the first A/A action, the operator notes the leak trend is still continuing to increase. The
flowchart directs the operator to take a second A/A action after which the operator again notes
the leak trend is increasing. At this point, the flow chart will continue to recommend that the
operator apply A/A actions until the trend of leaks on the system begins to decrease. Although
the operator of this particular main continues to apply A/A actions, no improvement is seen.
Based on the various locations along the main in which the issue is occurring, along with the
seemingly advanced nature of the graphitization, the operator concludes that the problem is
systemic. Due to its advanced nature, more drastic actions such as a major repair program or
replacement will have to be implemented in order to mitigate the threat and to assure
continued safe operation of the main. Had the trend of graphitization leaks in this particular
case responded to one of the A/A actions applied, the operator may have been able to continue
to operate and maintain the segment going forward in accordance with the DIMP regulations.
Returning to Flowchart START (for cast iron distribution main), the operator of this main should
assess the last applicable threat, which is breakage due to external forces. Flowchart
BREAKAGE, shown in Figure 43, should be used to analyze this threat.

yes
attributed to breakage due
to soil movement or other
outside forces?
no leaks occurring on the
trends.
-> Manage the leaks
accordance with the
program.
maintain segment in
be managed.
Subpart P (DIMP). Use new -> Schedule replacement of
information gained to the entire segment.
continue to monitor
segment performance. -> See GPTC DIMP section
suggestions.
Figure 43. Feasibility of Preventing in-Service Failures from Breakage due to Soil
Movement

FINAL
15-019
According to Figure 43, the operator confirmed that the main is experiencing breakage-type
leaks due to soil movement in isolated locations along the line. The operator should take an
A/A action such as increasing the frequency of patrolling and leak surveys, while managing the
breakage-type leaks based on risk and in accordance with its leak management program. With
the knowledge that this type of failure is only occurring in isolated locations, the operator may
be able to address the threat either by stabilizing the soil movement problem or by replacing
the cast iron only in the locations where soil movement is known to occur. Any new
information gained throughout the DIMP process such as knowledge regarding soil movement
should be used to continue to monitor the segment’s performance. After the threat of the
breakage-type leaks has been addressed, the operator will be able to continue to operate and
maintain the segment in accordance with the DIMP regulations.
The above analysis indicates that the threats to the integrity of this main cannot be
appropriately mitigated through repairs or on-going maintenance and integrity assessment
actions taken by the operator. The severe graphitization threat that the operator has been
unable to mitigate along with the breakage-type leaks occurring in locations with known
susceptibility to soil movement suggest that replacement of this particular main is the most
appropriate mitigative response.

FINAL
15-019
A PPENDIX A - P IPELINE M ILEAGE I NSTALLED B Y AGE

The following tables show the mileage by decade installed of pre-regulation (pre-1970) pipe
reported by pipeline operators in annual reports to PHMSA from 2005 through 2012. These
data may be accessed through the following website.
http://www.phmsa.dot.gov/portal/site/PHMSA/menuitem.6f23687cf7b00b0f22e4c6962d9
c8789/?vgnextoid=78e4f5448a359310VgnVCM1000001ecb7898RCRD&vgnextchannel
=3430fb649a2dc110VgnVCM1000009ed07898RCRD&vgnextfmt=print&vgnextnoice=1
From each table trends are plotted which show (in spite of year-to-year variations that probably
reflect changes in reporting parameters) that the mileages of pre-regulation pipe and mains and
numbers of pre-regulation services have decreased over the 2005-2012 time period. It is
assumed that the downward trends result from older pipelines being replaced.

FINAL
15-019
Table A1. Miles of Hazardous Liquid Pipelines by Decade of Installation
Calendar Pre-20 or 1920- 1930- 1940- 1950- 1960- 1970- 1980- 1990- 2000- 2010- Total
Year Unknown 1929 1939 1949 1959 1969 1979 1989 1999 2009 2019 Miles
2,005 1,377.21 3,034.48 6,343.91 16,819.16 36,036.70 35,586.28 27,468.01 16,990.63 18,095.63 5,013.45 166,765.45
2,006 1,424.03 2,847.21 5,797.98 17,098.78 35,698.46 33,195.64 28,889.86 17,384.29 17,734.80 6,647.81 166,718.88
2,007 1,560.55 2,720.75 5,563.39 15,698.11 36,668.23 33,271.18 28,570.00 18,126.93 18,839.30 8,827.68 169,846.13
2,008 1,569.38 2,896.92 5,686.31 15,500.35 37,400.46 34,527.29 29,302.99 17,921.39 18,360.49 10,623.49 173,789.06
2,009 4,170.42 2,894.73 5,840.64 14,944.67 35,113.40 35,906.24 27,480.69 17,027.30 18,613.62 13,973.78 175,965.48
2,010 5,227.19 2,899.78 4,681.80 14,911.53 32,392.77 35,105.71 30,818.69 18,120.56 18,380.74 17,521.58 1,913.61 181,973.95
2,011 5,953.99 2,894.35 4,681.23 15,490.66 33,967.83 34,316.64 30,315.71 17,183.18 19,234.99 16,832.83 2,510.97 183,382.38
2,012 6,017.79 2,573.77 4,700.37 15,544.60 34,659.09 33,921.74 29,991.11 17,238.21 19,056.42 16,926.58 5,393.98 186,023.64
40,000 Hazardous Liquid Pipelines

Miles of Pre-Regulation Hazardous
35,000
30,000
1920-1929
Liquid Pipelines
1930-1939
25,000
1950-1959
1960-1969
20,000
1940-1949
15,000
10,000
5,000
0
2004 2006 2008 2010 2012 2014
Year
Figure A1. Trends in the Amounts of Pre-Regulation Pipe in Hazardous Liquid

Pipelines (2005-2012)

FINAL
15-019
Table A2. Miles of Natural Gas Transmission Pipelines by Decade of Installation
Pre 1940
Calendar or 1940 - 1950 - 1960 - 1970 - 1980 - 1990 - 2000 - 2010 - Total
Year Unknown 1949 1959 1969 1979 1989 1999 2009 2019 Miles
2,005 21,831.84 24,137.24 72,734.08 74,238.21 32,301.31 28,251.67 31,902.26 15,071.46 300,468.07
2,006 20,869.48 24,095.86 71,008.44 73,975.64 32,313.87 27,624.05 32,316.05 18,120.76 300,324.14
2,007 20,404.81 23,867.73 71,842.22 73,022.16 32,306.28 27,650.68 32,074.42 19,897.87 301,066.17
2,008 18,105.61 22,912.97 72,118.63 73,598.07 32,134.98 27,613.03 32,128.72 24,568.99 303,181.00
2,009 16,775.15 22,657.10 71,391.02 73,187.76 32,375.38 27,383.50 32,340.08 28,462.94 304,572.93
2,010 17,221.68 22,653.36 69,971.38 72,442.01 31,991.01 26,921.15 31,858.39 29,727.34 1,988.78 304,775.10
2,011 14,950.79 22,571.84 69,903.11 72,230.49 32,267.36 26,455.82 31,624.38 30,009.18 5,023.12 305,036.08
2,012 14,517.04 22,041.40 68,813.06 72,027.59 31,686.20 26,034.08 30,856.42 30,463.21 6,893.53 303,332.54
80,000 Natural Gas Transmission Pipelines

Miles of Pre-Regulation Natural Gas
70,000
60,000
Transmission Pipelines
50,000 Pre 1940 or unknown

1940-1949
40,000 1950-1959
1960-1969
30,000
20,000
10,000
0
2004 2006 2008 2010 2012 2014
Year
Figure A2. Trends in the Amounts of Pre-Regulation Pipe in Natural Gas
Transmission Pipelines (2005-2012)

FINAL
15-019
Table A3. Miles of Natural Gas Distribution Mains by Decade of Installation
1940- 1950- 1960- 1970- 1980- 1990- 2000- 2010-
Unknown Pre 1940s 1949 1959 1969 1979 1989 1999 2009 2019 Total Miles
Year
2,005 124,218.36 71,510.58 27,459.02 107,728.93 193,304.76 128,550.42 150,752.22 228,397.21 133,135.48 1,165,057.00
2,006 122,174.91 68,784.34 27,231.90 107,623.25 193,428.56 128,955.48 151,285.90 229,932.65 158,614.85 1,188,031.83
2,007 91,362.21 71,062.39 26,890.36 108,265.70 197,729.42 133,436.08 156,236.61 234,219.58 184,093.89 1,203,296.24
2,008 88,353.78 70,425.24 26,825.59 108,955.49 195,247.18 132,990.80 156,989.26 232,726.22 197,370.50 1,209,884.07
2,009 85,988.93 68,553.48 26,547.58 108,434.19 198,888.09 132,505.21 156,821.45 233,972.06 208,658.22 1,220,369.20
2,010 76,176.44 66,964.24 26,470.76 110,288.11 200,645.10 134,239.19 157,845.21 236,920.05 208,637.52 11,114.16 1,229,300.77
2,011 99,201.84 63,090.22 25,419.02 105,192.35 195,221.46 130,494.70 154,632.62 236,319.96 206,998.64 22,000.53 1,238,571.34
2,012 93,972.79 60,131.37 25,418.14 106,160.99 195,008.60 131,581.18 155,705.38 236,399.66 207,129.95 35,338.18 1,246,846.23
2,013 96,620.57 58,958.33 25,232.27 102,926.39 191,794.87 131,739.17 154,911.51 234,778.48 206,458.30 49,887.84 1,253,307.74
140,000
Gas Distribution Mains
Miles of Pre-Regulation Gas Mains
120,000
100,000
80,000
60,000
Pre-1940 or unknown
40,000 1940-1949
1950-1959
20,000 1960-1969
0
2004 2006 2008 2010 2012 2014
Year
Figure A3. Trends in the Amounts of Pre-Regulation Natural Gas Distribution Mains
(2005-2012)

FINAL
15-019
Table A4. Number of Natural Gas Distribution Services by Decade of Installation
1940- 1950- 1960- 1970- 1980- 1990- 2000- 2010- Total
Unknown Pre-1940s 1949 1959 1969 1979 1989 1999 2009 2019 Services
Year
2005 8,796,749 2,226,803 1,267,203 4,752,717 7,913,840 8,220,076 9,615,676 12,568,061 7,643,953 63,005,078
2006 8,543,258 2,103,602 1,254,899 4,706,143 7,857,407 8,252,816 9,632,179 12,696,178 9,103,616 64,150,098
2007 7,368,330 2,059,634 1,209,080 4,615,020 7,849,059 8,455,123 9,784,113 12,939,711 10,584,043 64,864,113
2008 6,747,807 1,959,971 1,199,514 4,626,829 7,811,142 8,442,645 9,847,813 12,869,539 11,881,499 65,386,759
2009 5,549,515 1,942,976 1,219,358 4,631,658 7,808,206 8,490,444 9,981,694 13,009,388 12,951,278 65,584,517
2010 5,818,364 1,910,390 1,219,165 4,601,660 7,455,020 8,385,022 9,903,986 12,924,407 12,936,848 758,838 65,913,700
2011 5,479,549 1,825,596 1,190,529 4,585,953 7,468,279 8,416,633 9,964,499 13,004,886 12,951,199 1,578,864 66,465,987
2012 5,276,672 1,767,856 1,155,708 4,506,933 7,372,835 8,336,979 9,925,605 12,901,590 12,882,628 2,549,992 66,676,798
2013 5,118,867 1,689,578 1,121,772 4,398,773 7,176,610 8,198,680 9,845,732 12,936,388 12,974,841 3,653,575 67,114,816
10,000,000
Gas Distribution Services
9,000,000
Numbers of Pre-Regulation Gas
8,000,000
7,000,000
6,000,000
Services
5,000,000
4,000,000 unknown 1950-1959

1960-1969 pre-1940
3,000,000
1940-1949
2,000,000
1,000,000
0
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
Figure A4. Trends in the Numbers of Pre-Regulation Natural Gas Distribution
Services (2005-2012)

FINAL
15-019
A PPENDIX B - D ETAILS OF R EPORTABLE I NCIDENTS BY C AUSE

These data may be accessed through the following website.
http://www.phmsa.dot.gov/portal/site/PHMSA/menuitem.6f23687cf7b00b0f22e4c69
62d9c8789/?vgnextoid=78e4f5448a359310VgnVCM1000001ecb7898RCRD&vgnext
channel=3430fb649a2dc110VgnVCM1000009ed07898RCRD&vgnextfmt=print&vgn
extnoice=1
Table B1. Reportable Incidents by Cause 1994-2013 – Part 1 of 2
NUMBER PERCENT
CAUSE OR THREAT HAZARDOUS NATURAL GAS NATURAL GAS HAZARDOUS NATURAL GAS NATURAL GAS
LIQUID TRANSMISSION DISTRIBUTION LIQUID TRANSMISSION DISTRIBUTION
CORROSION
EXTERNAL CORROSION 259 100 30 7.17% 8.57% 2.18%
INTERNAL CORROSION 354 131 1 9.81% 11.23% 0.07%
UNSPECIFIED CORROSION 195 5.40%
Sub Total for CORROSION 808 231 31 22.38% 19.79% 2.25%
EXCAVATION DAMAGE
OPERATOR/CONTRACTOR EXCAVATION DAMAGE 45 29 25 1.25% 2.49% 1.81%
THIRD PARTY EXCAVATION DAMAGE 125 138 389 3.46% 11.83% 28.23%
PREVIOUS DAMAGE DUE TO EXCAVATION 8 9 2 0.22% 0.77% 0.15%
UNSPECIFIED EXCAVATION DAMAGE 12 0.33%
Sub Total EXCAVATION DAMAGE 190 176 416 5.26% 15.08% 30.19%
INCORRECT OPERATION
DAMAGE BY OPERATOR OR OPERATOR'S CONTRACTOR 10 1 5 0.28% 0.36%
OVERFILL/OVERFLOW OF TANK/VESSEL/SUMP 40 1.11%
INCORRECT VALVE POSITION 38 6 4 1.05% 0.51% 0.29%
PIPELINE/EQUIPMENT OVERPRESSURED 24 2 0.66% 0.17%
INCORRECT INSTALLATION 32 1 5 0.89% 0.09% 0.36%
INCORRECT EQUIPMENT 4 0.11%
OTHER INCORRECT OPERATION 36 13 14 1.00% 1.11% 1.02%
UNSPECIFIED INCORRECT OPERATION 230 16 29 6.37% 1.37% 2.10%
Sub Total for INCORRECT OPERATION 414 39 57 11.47% 3.26% 4.14%
MAT'L/WELD/FAILURE
CONSTRUCTION INSTALLATION OR FABRICATION-RELATED 65 30 1.80% 2.57%
MANUFACTURING-RELATED 41 19 1.14% 1.63%
ENVIRONMENTAL CRACKING-RELATED 20 14 0.55% 1.20%
BODY OF PIPE 25 18 18 0.69% 1.54% 1.31%
PIPE SEAM 37 11 1.02% 0.94%
UNSPECIFIED PIPE BODY OR SEAM 0.00%
BUTT WELD 23 30 2 0.64% 2.57% 0.15%
FUSION JOINT 2 0.15%
COMPRESSION FITTING 2 0.15%
MECHANICAL FITTING 7 0.51%
FILLET WELD 14 7 0.39% 0.60%
UNSPECIFIED WELD
JOINT/FITTING/COMPONENT 75 40 27 2.08% 3.43% 1.96%
UNSPECIFIED MAT'L/WELD FAILURE 108 5 2.99% 0.36%
OTHER PIPE/WELD/JOINT FAILURE 5 0.36%
Sub Total for MATERIAL/WELD/FAILURE 408 169 68 11.30% 14.48% 4.93%

FINAL
15-019
Table B1. Reportable Incidents by Cause 1994-2013 – Part 2 of 2
NUMBER PERCENT
CAUSE OR THREAT HAZARDOUS NATURAL GAS NATURAL GAS HAZARDOUS NATURAL GAS NATURAL GAS
LIQUID TRANSMISSION DISTRIBUTION LIQUID TRANSMISSION DISTRIBUTION
EQUIPMENT FAILURE
MALFUNCTION OF CONTROL/RELIEF EQUIPMENT 129 108 19 3.57% 9.25% 1.38%
VALVE 6 0.44%
PUMP/COMPRESSOR-RELATED EQUIPMENT 167 10 4.63% 0.86%
THREADED CONNECTION/COUPLING FAILURE 120 30 4 3.32% 2.57% 0.29%
NON-THREADED CONNECTION FAILURE 227 12 11 6.29% 1.03% 0.80%
DEFECTIVE OR LOOSE TUBING/FITTING 35 1 0.97% 0.09%
FAILURE OF EQUIPMENT BODY 32 4 0.89% 0.34%
OTHER EQUIPMENT FAILURE 112 11 3.10% 0.94%
UNSPECIFIED EQUIPMENT FAILURE 506 14.02%
Sub Total for EQUIPMENT FAILURE 1328 176 40 36.79% 15.08% 2.90%
NATURAL FORCE DAMAGE
EARTH MOVEMENT 19 23 28 0.53% 1.97% 2.03%
HEAVY RAINS/FLOODS 31 90 27 0.86% 7.71% 1.96%
LIGHTNING 20 17 19 0.55% 1.46% 1.38%
TEMPERATURE 53 10 17 1.47% 0.86% 1.23%
HIGH WINDS 30 14 11 0.83% 1.20% 0.80%
OTHER NATURAL FORCE DAMAGE 4 4 5 0.11% 0.34% 0.36%
UNSPECIFIED NATURAL FORCE DAMAGE 35 1 0.97% 0.07%
Sub Total for NATURAL FORCE DAMAGE 192 158 108 5.32% 13.54% 7.84%
OTHER OUTSIDE FORCE DAMAGE
FIRE/EXPLOSION AS PRIMARY CAUSE 5 12 227 0.14% 1.03% 16.47%
VEHICLE NOT ENGAGED IN EXCAVATION 26 63 161 0.72% 5.40% 11.68%
MARITIME EQUIPMENT OR VESSEL ADRIFT 1 1 0.03% 0.09%
FISHING OR MARITIME ACTIVITY 1 7 0.03% 0.60%
ELECTRICAL ARCING FROM OTHER EQUIPMENT/FACILITY 6 2 11 0.17% 0.17% 0.80%
PREVIOUS MECHANICAL DAMAGE 8 7 5 0.22% 0.60% 0.36%
INTENTIONAL DAMAGE 7 3 17 0.19% 0.26% 1.23%
OTHER OUTSIDE FORCE DAMAGE 8 9 24 0.22% 0.77% 1.74%
UNSPECIFIED OUTSIDE FORCE DAMAGE 19 5 0.53% 0.36%
Sub Total for OTHER OUTSIDE FORCE DAMAGE 81 104 450 2.24% 8.91% 32.66%
ALL OTHER CAUSES
MISCELLANEOUS CAUSE 120 88 101 3.32% 7.54% 7.33%
UNKNOWN CAUSE 34 26 107 0.94% 2.23% 7.76%
UNSPECIFIED 35 0.97%
Sub Total for ALL OTHER CAUSES 189 114 208 5.24% 9.77% 15.09%
Totals 3610 1167 1378 100.00% 99.91% 100.00%

FINAL
15-019
A PPENDIX C - F LOWCHARTS


yes attributes.

Proceed to
legacy pipe ?
MODERN PIPE
yes
Figure C1. Flowchart START (for hazardous liquid pipelines, high-stress natural gas
transmission pipelines, and low-stress natural gas transmission pipelines)

FINAL
15-019
Legacy Pipe
or Part 195? yes
yes
Part 195.


SSAW pipe


susceptible to HSC
5. Segment contains
puddle welds, patches, Proceed to
or half soles LEGACY REPAIRS
Figure C2. Flowchart LEGACY PIPE

FINAL
15-019
welded pipe.

Go to
yes

PCIF.
to occur?
no
yes

Figure C3A. Flowchart LF ERW (Part A)

FINAL
15-019
PART B

no
yes
that prevents seam failures in
rehabilitation based on
to failure.
segments.
Figure C3B. Flowchart LF ERW (Part B)

FINAL
15-019
Segment contains SSAW
pipe.

times to failure from PCIF or rely on
service and test failure history to
establish re-assessment interval to
prevent service failures from PCIF
Lower pressure as needed to prevent

Is periodically assessing seam It is feasible to lower the seam failures in service and continue
no yes
integrity by means of hydrostatic MOP to a level that prevents to operate in accordance with Part
testing or ILI feasible? seam failures in service 192 or Part 195
yes
no
PCIF.
yes intervals or to decrease the MOP no Schedule
until service failures from PCIF replacement.
assessments?
cease to occur?
no
yes

Part 195 and continue to perform decreased interval or lower MOP
seam integrity assessments on until service failures from PCIF
the same schedule. cease to occur.
Figure C4. Flowchart SSAW

FINAL
15-019
Segment contains furnace
lap-welded pipe. l
Use service and test failure history

to establish re-assessment interval
defects.

Periodically assessing seam Is it feasible to lower the
testing or ILI. is feasible that prevents seam failures in
service?
Part 192 or Part 195
yes
no
defects.
Are service failures from lack- intervals or to decrease the
yes no
of-fusion or burned-metal Schedule
defects occurring between re- replacement.
failures from lack-of-fusion or
assessments?
burned-metal defects cease to
no occur?
yes
Re-assess seam integrity at

decreased interval or lower
in accordance with Part 192 or
Part 195 and continue to perform
failures from lack-or-fusion or
seam integrity assessments on
burned metal defects cease to
the same schedule.
occur.
Figure C5. Flowchart LAP-WELDED

FINAL
15-019
no

welds?
no
Figure C6. Flowchart LEGACY FEATURES

FINAL
15-019
Segment contains
puddle welds*,
patches, or half soles.
Is the segment
Continue to operate in
operated at a stress no
level greater than 20% accordance with Part
192 or Part 195.
of SMYS?
yes
Were the puddle

welds made while the
pipe was in service? no no yes
Were patches or half
Were patches or half no soles installed while the yes
Was a low-hydrogen
yes soles installed over a pipe was either full of welding procedure used?
leaking defect? liquid or high-pressure,
flowing gas?
yes no
Can the segment be Conduct a "one-time" hydrostatic test, make

Can the puddle welds, no subjected to a yes no further legacy repairs after the test, and
patches, or half soles hydrostatic test in continue to operate in accordance with Part
be located by ILI? accordance with Part 192 or Part 195 as long as no additional
192 or Part 195? failures due to legacy repairs occur.
yes
no
Repair or replace If failures of legacy repairs continue to
puddle welds, occur after the test, schedule
patches, or half soles Schedule replacement
of pipeline or lower replacement of pipeline or lower
and continue to operating stress to 20% of SMYS or less.
operating stress to
operate in accordance
20% of SMYS or less.
with Part 192 or Part
195.
*Puddle welding refers to welds made with an unknown welding proceddure to repair pits or other defects.
Deposited weld metal repairs made according to a documented and proven procedure are a legitimate
repair method and do not fall under this definitaion of puddle welding.
Figure C7. Flowchart LEGACY REPAIRS

FINAL
15-019
Modern Pre-Regulation Pipe
or Part 195? yes
yes
Part 195.
Evaluate the threats to a modern pipeline Schedule Replacement
1. Segment contains HF-ERW pipe made Proceed to HF-

prior to 1971 ERW
2. Segment contains DSAW or Proceed to DSAW

HSAW pipe made prior to 1971 7. Segment may be Proceed to HSC
susceptible to HSC
3. Segment contains seamless Proceed to

pipe made prior to 1971 SEAMLESS
wrinkle bends, non-standard Proceed to
4. Segment may be susceptible fittings, Dresser couplings, threaded LEGACY FEATURES
Proceed to EC couplings, or acetylene girth welds
5. Segment may be susceptible Proceed to IC

9. Segment contains
to internal corrosion Proceed to
puddle welds, patches,
or half soles LEGACY REPAIRS
6. Segment may be susceptible Proceed to SCC

to SCC
Figure C8. Flowchart MODERN PIPE

FINAL
15-019
Segment contains HF-ERW

establish re-assessment interval to prevent service failures from
Lower pressure as needed to

Is periodically assessing seam no Is it feasible to lower the yes prevent seam failures in service
integrity by means of hydrostatic MOP to a level that prevents and continue to operate in
testing or ILI feasible? seam failures in service? accordance with Part 192 or
yes
no
Assess seam integrity periodically Assess seam integrity periodically

based on the established interval based on the established interval to
PCIF.
or SSWC occurring between yes intervals or to decrease the MOP no Schedule
re-assessments? until service failures from PCIF or replacement.
SSWC cease to occur?
no
yes

Part 195 and continue to perform decreased interval or lower MOP
seam integrity assessments on until service failures from PCIF or
the same schedule SSWC cease to occur.
Figure C9. Flowchart HF ERW

FINAL
15-019
Is it possible to mitigate narrow
Segment contains DSAW or Is the pipeline coated with axial external corrosion (NAEC)
HSAW pipe made prior to a single or double layer yes no
and/or SCC, if either or both
1971. polyethylene tape coating? occur, using ILI, hydrostatic
testing, or another method?
no yes
times to failure from PCIF or rely on Assess integrity and take
service and test failure history to necessary actions to prevent
establish re-assessment interval to failures from NAEC and SCC

Is periodically assessing seam Is it feasible to lower the
no yes needed to prevent seam
integrity by means of hydrostatic operating stress to a level
failures in service and continue
testing or ILI feasible? that prevents seam failures in
service?
yes
no
PCIF
Are service failures from PCIF intervals or to decrease the
yes no Schedule
occurring between re- operating stress until service
assessments? replacement.
occur?
no
yes

Part 195 and continue to perform decreased interval or lower
seam integrity assessments on operating stress until service
the same schedule. failures from PCIF cease to occur.
Figure C10. Flowchart DSAW

FINAL
15-019
Segment contains seamless

Use service and test failure history to

establish re-assessment interval to prevent
service failures from manufacturing defects.
No action is necessary if no such failures have
occurred, if the pipeline has been subjected
to a prior hydrostatic test, and if the
predicted times to failure base on that test
exceed the useful life of the pipeline.

Is periodic integrity assessment Is it feasible to lower the needed to prevent failures in
by means of hydrostatic testing service and continue to operate
or ILI feasible? that prevents failures in in accordance with Part 192 or
service? Part 195
yes
no
Assess integrity periodically based
on the established interval to
prevent service failures from
manufacturing defects
It is feasible to continue to
Service failures from decrease the re-assessment
yes intervals or to decrease the no
manufacturing defects are Schedule
occurring between re- operating stress until service replacement.
assessments failures from manufacturing
no defects cease to occur.
yes
Continue to operate and maintain Re-assess integrity at decreased

in accordance with Part 192 or interval or lower the operating
Part 195 and continue to perform stress until service failures from
integrity assessments on the manufacturing defects cease to
same schedule occur.
Figure C11. Flowchart SEAMLESS

FINAL
15-019

yes yes no

yes
no ∩

ECDA?
yes yes

yes
Conduct periodic
assessments or
prevent leaks and
operate segment in
or Part 195.
Figure C12. Flowchart EC

FINAL
15-019
Is the transported no Conduct hydrostatic tests at

product corrosive? appropriate intervals to prevent leaks
and ruptures from internal corrosion
yes Continue to operate and continue to operate in
segment in accordance accordance with Part 192 or Part 195.
with Part 192 or Part 195
yes
Is it feasible to prevent
Can the pipeline Can metal loss be leaks and ruptures
no no no
accomodate metal adequately assessed from corrosion using
loss ILI tools? and mitigated using periodic hydrostatic
ICDA? testing?
yes
Schedule
yes replacement.

yes
Conduct metal loss ILI inspections or

ICDA assessments at appropriate
intervals to prevent leaks and
ruptures from internal corrosion and
continue to operate segment in
accordance with Part 192 or Part 195.
Figure C13. Flowchart IC

FINAL
15-019
to SCC?
Part 195.
yes

no or unknown
no

Can the pipeline
tool appropriate?
pressure?
yes
yes

inspections at
192 and Part 195.

Figure C14. Flowchart SCC

FINAL
15-019

between 1947 and 1960? Part 195. Schedule replacement
yes no
actives?
yes
yes Use special "residual field" MFL tool
to find and repair hard spots and
continue to operate in accordance
Has any failure occurred that Can pipe-to-soil potentials be with Part 192 or Part 195.
corrosion?
no
and measurements of
or Part 195
Figure C15. Flowchart HSC

FINAL
15-019
no
yes
steel main known? plan for gaining this
yes
to (f).
the segment?
Subpart P (DIMP).
yes
Selective seam weld Proceed to Flowchart SSWC

corrosion (SSWC). for steel gas distribution
main.
External corrosion (EC). Proceed to Flowchart EC for

Internal corrosion (IC). Proceed to Flowchart IC for

Threaded underground
LEGACY FEATURES for steel
joints.
gas distribution main.
Figure C16. Flowchart START (for steel gas distribution system main)

FINAL
15-019
Seam types susceptible to Apply one or a combination

SSWC are: of the following actions:
- Electric Flash Welded -> Increase the frequency of
- Low-Frequency Electric and monitor the number of
Resistance Welded (LF-ERW) leaks occurring on the
- High-Frequency Electric segment as well as the leak
Resistance Welded (HF-ERW) trends.
manufactured prior to 1980, -> Manage the leaks
or where the toughness of occurring on the segment
the material was low or not based on risk and in
specified. accordance with the
Continue to operate and program.
maintain segment in
Is the seam type of the accordance with Part 192, -> Repair leaks that cannot
no
segment susceptible to Subpart P (DIMP). Use new be managed.
SSWC? information gained to -> Schedule replacement of
continue to monitor the entire segment.
yes segment performance.
Have there been leaks from 6.3, Table 6.1, for additional
no
SSWC? suggestions.
yes
Figure C17. Flowchart SSWC (for steel gas distribution system main)

FINAL
15-019
no
Does segment have an Is the segment experiencing Apply one or a combination
yes
external coating? metal loss and/or leaks of the following actions:
attributed to external
yes corrosion?
no no
Is the external coating leaks occurring on the
adequate? segment as well as the leak
trends.
yes
-> Manage the leaks
Does the segment have a occurring on the segment
cathodic protection (CP)
no based on risk and in
system? accordance with the
program.
yes
no
Is the CP system adequate? be managed.
yes
continue to monitor
yes

external corrosion on the
segment increasing?
no

yes
patrolling and leak surveys external corrosion on the
leak trends. same?
no

intervals and continue to external corrosion on the
Figure C18. Flowchart EC (for steel gas distribution system main)

FINAL
15-019
Has there ever been metal In the course of normal Apply one or a combination
loss in conjunction with operations, if presented with of the following actions:
yes
drips, dead-legs, and/or the opportunity, inspect the
interior surface of the pipe -> Increase the frequency of
areas of water
for internal corrosion. patrolling and leak surveys
accumulation, such as
downstream of a regulator,
in the system? leaks occurring on the
no trends.
-> Manage the leaks
Is the segment experiencing occurring on the segment
Is the transported product yes metal loss and/or leaks yes based on risk and in
corrosive? attributed to internal accordance with the
corrosion? operator's leak management
no
program.
no
Continue to operate and be managed.
maintain segment in
accordance with Part 192, -> Schedule replacement of
Subpart P (DIMP). Use new the entire segment.
information gained to -> See GPTC DIMP section
continue to monitor 6.3, Table 6.1, for additional
yes

internal corrosion on the
segment increasing?
no

patrolling and leak surveys yes internal corrosion on the
leak trends. same?
no

intervals and continue to internal corrosion on the
Figure C19. Flowchart IC (for steel gas distribution system main)

FINAL
15-019
Segment contains threaded Apply one or a combination

underground joints and/or of the following actions:
other obsolete features.
Is the segment experiencing segment as well as the leak
leaks and/or failures yes trends.
attributed to the threaded
underground joints and/or -> Manage the leaks
obsolete features? occurring on the segment
no accordance with the
program.
maintain segment in
be managed.
continue to monitor
suggestions.
Figure C20. Flowchart LEGACY FEATURES (for steel gas distribution system main)

FINAL
15-019
no
material and properties, established or verified?
type of construction, no
yes
threats, etc.)
yes regarding the
information and develop a
plastic distribution system plan for gaining this
known? information. with Part 192, Subpart P
yes and as recommended by
to (f).
the segment?
Subpart P (DIMP).
yes
Slow crack growth (SCG). Proceed to Flowchart SCG.
Fusion/mechanical joining or
service tap connection type
PLASTIC CONNECTIONS.
failures.
Figure C21. Flowchart START (for plastic gas distribution system)

FINAL
15-019
Materials with known Is the material of which the
no
susceptibility to premature segment is constructed
brittle-like cracking (i.e. SCG) susceptible to SCG?
under conditions of local
stress intensification are: yes Continue to operate and
- Century Utility Products, maintain segment in
Have there been leaks accordance with Part 192,
Inc. products; no
attributed to SCG? Subpart P (DIMP). Use new
- Pipe extruded by Century information gained to
Utility Products Inc. from continue to monitor
Union Carbide Corporation's yes segment performance.
DHDA 2077 Tan medium
density polyethylene resin; Apply one or a combination
- Low-ductile inner wall of the following actions:
"Aldyl A" piping -> Increase the frequency of Patrol the line and complete
manufactured by DuPont patrolling and leak surveys The trend of leaks due to
Company prior to 1973; and monitor the number of SCG on the segment is intervals and continue to
leaks occurring on the decreasing.
-Polyethylene gas pipe monitor the leak trends.
designated PE 3306; segment as well as the leak
trends. no
- Delrin insert tap tees;
-> Manage the leaks Is the trend of leaks due to Increase the frequency of
- Plexco service tee Celcon occurring on the segment yes
SCG on the segment patrolling and leak surveys
(polyacetal) caps; based on risk and in remaining the same? and continue to monitor the
- Plastic pipe installed in accordance with the leak trends.
natural gas distribution operator's leak management
systems from the 1960s program.
through the early 1980s. no
be managed. Is the trend of leaks due to
SCG on the segment
increasing?
the entire segment.
-> See GPTC DIMP section yes
suggestions.
Figure C22. Flowchart SCG (for plastic gas distribution system)

FINAL
15-019
Segment contains fusion Apply one or a combination

joints, mechanical joints, of the following actions:
and/or service tap
connections.
Is the segment experiencing trends.
leaks and/or failures
attributed to pipe joining -> Manage the leaks
practices (e.g. a bad fusion yes occurring on the segment
joint) or service tap based on risk and in
connection issues (e.g. a bad accordance with the
bolt on type of tee, or a operator's leak management
broken tee cap)? program.
no
be managed.
the entire segment.
maintain segment in
accordance with Part 192, -> See GPTC DIMP section
Subpart P (DIMP). Use new 6.3, Table 6.1, for additional
information gained to suggestions.
continue to monitor
segment performance.
Figure C23. Flowchart PLASTIC CONNECTIONS (for plastic gas distribution system)

FINAL
15-019
no
yes
cast iron main known? plan for gaining this
yes (DIMP), section 1007(a)(3)
to (f).
the segment?
Subpart P (DIMP).
yes
Joint leakage.
Proceed to Flowchart CAST
IRON CONNECTIONS.
Pipe body leakage due to Proceed to Flowchart

graphitization. GRAPHITIZATION.
Breakage due to external Proceed to Flowchart

forces. BREAKAGE.
Figure C24. Flowchart START (for cast iron gas distribution system main)

FINAL
15-019
Is the segment experiencing

Are the pipe joints sealed? yes leaks and/or failures at the yes
sealed joints or connections?
Apply one or a combination

no
no
of the following actions:
Are the pipe joints along the no
segment leaking? and monitor the number of
maintain segment in leaks occurring on the
yes accordance with Part 192, segment as well as the leak
Subpart P (DIMP). Use new trends.
information gained to -> Manage the leaks
continue to monitor occurring on the segment
segment performance. based on risk and in
accordance with the
program.
be managed.
the entire segment.
suggestions.
Figure C25. Flowchart CAST IRON CONNECTIONS (for cast iron gas distribution
system main)

FINAL
15-019

yes
attributed to graphitization?
no
trends.
-> Manage the leaks
accordance with the
program.
be managed.
continue to monitor
yes

graphitization on the
segment increasing?
no

yes
patrolling and leak surveys graphitization on the
leak trends. same?
no

intervals and continue to graphitization on the
Figure C26. Flowchart GRAPHITIZATION (for cast iron gas distribution system main)

FINAL
15-019

yes
attributed to breakage due
to soil movement or other
outside forces?
no leaks occurring on the
trends.
-> Manage the leaks
accordance with the
program.
maintain segment in
be managed.
continue to monitor
suggestions.
Figure C27. Flowchart BREAKAGE (for cast iron gas distribution system main)

FINAL
15-019
R EFERENCES
i
Distribution Pipe: Repair and Replacement Decision Manual, A Technical Report, Operating
Section, American Gas Association, Catalog No. XL0702, 2006.
ii
Groeneveld, T.P., “Hydrogen Stress Cracking”, 5th Symposium on Line Pipe Research,
American Gas Association, Catalogue No. L30174, 1974.
iii
2011 Distribution Annual Data, http://www.phmsa.dot.gov/pipeline/library/data-stats
iv
Brittle-Like Cracking in Plastic Pipe for Gas Service, Special Investigative Report, National
Transportation Safety Board, PB98-917001, NTSB/SIR-98/01, April 23, 1998, Notation 6984.
v
Makar, J.M., Desnoyers, R, and McDonald, S.E., Failure Modes and Mechanisms in Gray Cast
Iron Pipe, Institute for Research in Construction, National Research Council Canada, Presented
at Underground Infrastructure Research 2001, Waterloo, Ontario, June 10-13
vi
AMERICAN Pipe Manual, Section 12, AMERICAN Historical Data, American Cast Iron Pipe Co.
vii
“Technical, Operational, Practical, and Safety Considerations of Hydrostatic Pressure Testing
of Existing Pipelines”, Prepared for the INGAA Foundation, Inc., by Jacobs Consultancy, Inc. and
Gas Transmission Systems, Inc., December 5, 2013.
viii
Leis, B. N., Ghadiali, N. D., “Pipe Axial Flaw Failure Criteria – PAFFC”, Version 1.0 Users
Manual and Software, Topical Report to American Gas Association, NG-18, Catalog No. L51720,
1994.
ix
Jaske, C. E., and Beavers, J. A., “Development and Evaluation of Improved Model for
Engineering Critical Assessment of Pipelines”, Proceedings of IPC 2002, 4th International Pipeline
Conference, September 29-October 3, 2002, Calgary, Alberta, Canada.
x
API 579-1/ASME FFS-1, “Fitness for Service” 2007
xi
Kiefner, J.F., Maxey, W.A., Eiber, R.J., and Duffy, A.R., “Failure Stress Levels of Flaws in
Pressurized Cylinders”, Progress in Flaw Growth and Fracture Toughness Testing, ASTM STP
536, American Society for Testing and Materials, pp 461 – 481, 1973.
xii
Kiefner, J.F., “Modified equation helps integrity management”, Oil and Gas Journal, Oct 6,
2008, pp 76-82 and “Modified Ln-Secant equation improves failure prediction”Oct 13, 2008, pp
64-66.
xiii
Newman, J.C. and Raju, I.S., “Stress Intensity Factor Equations for Cracks in Three-
Dimensional Finite Bodies Subjected to Tension and Bending Loads”, NASA Technical
Memorandum 85793, April 11, 1984.

FINAL
15-019
xiv
Eiber, R.J., Hein, G.M., Duffy, A.R., and Atterbury, T.J., “Investigation of the Initiation and
Extent of Ductile Pipe Rupture”, Report No. BMI-1793, Progress Report for May-December,
1966, U.S. Atomic Energy Commission Contract No. w-7405-eng-92, January 1, 1967.
xv
Eiber, R.J., Maxey, W.A.., Duffy, A.R., and Atterbury, T.J., “Investigation of the Initiation and
Extent of Ductile Pipe Rupture”, Report No. BMI-1866, U.S. Atomic Energy Commission Contract
No. w-7405-eng-92, NTIS, July 1969
xvi
Folias, E.S., “The Stresses in a Cylindrical Shell Containing an Axial Crack”, Aerospace
Research Laboratories, ARL 64-174, (October 1964).
xvii
Duffy, A. R., McClure, G. M., Eiber, R. J., and Maxey, W. A., "Fracture Design Practices for
Pressure Piping", Fracture--An Advanced Treatise, Fracture Design of Structures, Edited by H.
Liebowitz, Academic Press, New York, Vol. V., Chapter 3, pp 159-232 (1969).
xviii
Bergman, S. A., “Why Not Higher Operating Pressures for Lines Tested to 90% SMYS?”
Pipeline and Gas Journal (December 1974).
xix
Kiefner, J.F., and Kolovich, K.M., “Effectiveness of Hydrostatic Testing for Assessing the
Integrity of ERW and Flash-Welded Pipe Seams”, the deliverable of Subtask 1.2 on U.S.
Department of Transportation Other Transaction Agreement No.DTPH56-11-T-000003, January
15, 2013. This publication may be found at:
http://primis.phmsa.dot.gov/matrix/PrjHome.rdm?prj=390
xx
Brooks, L. E., “High-Pressure Testing - Pipeline Defect Behavior and Pressure Reversals”
ASME, 68-PET-24, (1968).
xxi
Kiefner, J. F., Maxey, W. A., and Eiber, R. J., "A Study of the Causes of Failure of Defects
That Have Survived a Prior Hydrostatic Test", Pipeline Research Committee, American Gas
Association, NG-18 Report No. 111 (November 3, 1980).
xxii
Kiefner, J.F., Kolovich, K.M., and Kariyawasam, S., “A Study of Cases of Hydrostatic Tests
Where Multiple Test Failures Have Occurred”, IPC2010-31157, Proceedings of the 8th
International Pipeline Conference, IPC 2010, September 27 – October 1, 2010, Calgary, Alberta,
Canada.
xxiii
TTO Number 6, “Spike Hydrostatic Test Evaluation” Integrity Management Program Delivery
Order DTRS56-02-D-70036, Department of Transportation, Research and Special Programs
Administration, Office of Pipeline Safety, Michael Baker, Jr. Inc., July, 2004
xxiv
Fessler, Raymond R., and Rapp, Steve, “Method for Establishing Hydrostatic Re-Test
Intervals for Pipelines with Stress-Corrosion Cracking”, Paper No. IPC2006-10163, Proceedings
of IPC 2006, 6th International Pipeline Conference, September 25-29, 2006, Calgary, Alberta,
Canada

FINAL
15-019
xxv
Public Inquiry Concerning Stress Corrosion Cracking on Canadian Oil and Gas Pipelines, MH-
2-96, November, 1996, National Energy Board, Canada
xxvi
Kiefner, J.F. and Kolovich, K.M., “Calculation of a Corrosion Rate Using Monte Carlo
Simulation”, Paper No. 07120, NACE Corrosion 2007 Conference & Expo, March 2007, Nashville,
TN, USA.
xxvii
Sen, Millan, and Kariyawasam, Shahani, “Analytical Approach to Determine Hydrotest
Intervals”, Paper No. IPC2008-64537, Proceedings of the 7th International Pipeline Conference,
IPC08, September 29 – October 3, 2008, Calgary, Alberta, Canada
xxviii
Paris, P.C., and Erdogan, F., “A Critical Analysis of Crack Propagation Laws”, Transactions of
ASME, Journal of Basis Engineering, Series D, Volume 85, No. 5, (1963) pp 405-09
Kiefner, J.F., and Kolovich, K.M., “Predicting Times to Failure for ERW Seam Defects that
xxix
Grow by Pressure-Cycle-Induced Fatigue”, one of the deliverables of Subtask 2.5 on U.S.

Department of Transportation Other Transaction Agreement No.DTPH56-11-T-000003, January
28, 2013. This publication may be found at:
Fessler, R.R., “Overview of Solutions to the Stress-Corrosion Cracking Problem, 6th

xxx
Symposium on Line Pipe Research, American Gas Association, Catalog No. L30175, 1979
xxxi
Leis, B.N., and Brust, F.W., “Hydrotest Strategies for Gas Transmission Pipelines Based on
Ductile-Flaw Growth Considerations”, American Gas Association, Pipeline Research Committee,
NG-18 Report No. 194, July 27, 1992
xxxii
Rosenfeld, M.J., “Hydrostatic Pressure Spike Testing of Pipelines – Why and When”,
American Gas Association, Operations Conference, Orlando, May 22, 2013
Chauhan, V., Brister, J., and Dafea, M., “A Review of Methods for Assessing the Remaining
xxxiii
Strength of Corroded Pipelines”, Advantica Report No. 6781, to the Pipeline and Hazardous
Materials Safety Administration, U.S. Department of Transportation, October, 2008.
xxxiv
Johnston, D.C., and Kolovich, C.E., “Using a Probability Approach to Rank In-Line Inspection
Anomalies for Excavation and for Setting Re-inspection Intervals”, API’s 51st Annual Pipeline
Conference and Cybernetics Symposium, April 18-20, 2000
xxxv
ASME B31.8S – 2012, Managing System Integrity of Gas Pipelines.
xxxvi
API Recommended Practice 1160, Second Edition, September 2013, Managing System
Integrity for Hazardous Liquid Pipelines.
Kiefner, J.F., Kolovich, K.M., Macrory-Dalton, C.J., Johnston, D.C., Maier, C.J., and Beavers,
xxxvii
J.A., “Track Record of In-Line Inspection as a Means of ERW Seam Integrity Assessment”, the

FINAL
15-019
deliverable of Subtask 1.3 on U.S. Department of Transportation Other Transaction Agreement

No.DTPH56-11-T-000003, November 15, 2012. This publication may be found at:
xxxviii
Duckworth, N., “KMAP™ for Longitudinal Weld Threat Analysis”, PHMSA Pipeline Seam
Weld Workshop, Arlington, VA, July 20, 2011.
xxxix
Phillips, M., and Maier, C., “Responding to Crack Detection ILI”, The Pipeline Pigging and
Integrity Management Conference, Clarion Technical Conferences, February 13-14, 2008
xl
Gao, M., Samarth, T., Krishnamurthy, R, Kania, R., Rosca, G., and Piazza, M., “A
Comprehensive Analysis of EMAT ILI and In-Ditch SCC Inspection Technologies and Its
Application to Pipeline Integrity Management Programs”, Paper 20, APIA, EPRG, PRCI Joint
Technical Meeting, Sydney Australia, 29 April – 3 May, 2013.
xlii
Van Os, M.T., “DIRECT ASSESSMENT-1: Software Module Hones System-Wide Practices”, Oil
and Gas Journal, October 2, 2006.
xliii
“Improvements to the External Corrosion Direct Assessment (ECDA) Process – White Paper
#360”, “Severity Ranking of ECDA Indirect Inspection Indications – Project #242”, for Pipeline
and Hazardous Materials Safety Administration, U.S. Department of Transportation, Contract
No. DTPH56-08-T-000012, Corpro Companies, Inc., June, 2010
xliv
Andrew Francis & Associates Ltd., “A Robust Methodology for External Corrosion Direct
Assessment,” Report AFAA-R0003-04, 2004.
xlv
Andrew Francis & Associates Ltd., “A Robust Methodology for External Corrosion Direct
Assessment-Supplement,” Report AFAA-R0004-04, 2004.
xlvi
Clark, E.B., Leis, B.N., and Eiber, R.J., “Integrity Characteristics of Vintage Pipelines”, Report
No. 2002-50436, The INGAA Foundation, Inc., 2005.
xlvii
Integrity Management for Gas Distribution, Report of Phase 1 Investigations, Prepared by
Stakeholder Public, Gas Distribution Pipeline Industry, State Pipeline Safety Representatives,
Pipeline and Hazardous Materials Safety Administration, December 2005.
xlviii
Distribution Integrity Management Program, Prepared by Gas Piping Technology Committee
Z380, The American Gas Association, 2009.
xlix
Title 49, Transportation, Part 192, Code of Federal Regulations, Office of the Federal
Register, Revised October 1, 2011.

Kiefner Final Report Repair Replace

Uploaded by

Copyright:

Available Formats

Kiefner Final Report Repair Replace

Uploaded by

Document Information

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

Kiefner Final Report Repair Replace

Uploaded by

Copyright:

Available Formats

PHMSA Final Report No.

J.F. Kiefner and M. Van Auker

REPAIR/REPLACE CONSIDERATIONS FOR PRE-REGULATION PIPELINES

DTPH5614H00006, REPAIR/REPLACE CONSIDERATIONS FOR

U.S. DEPARTMENT OF TRANSPORTATION

PIPELINE AND HAZARDOUS MATERIALS SAFETY ADMINISTRATION

March 11, 2015

J.F. Kiefner and M. Van Auker

Kiefner and Associates, Inc.

This document presents findings and/ or recommendations based on engineering services

Repair/replace decisions also depend on recognizing and addressing threats to pipeline

Replacement may also be deferrable if a repair or operational change mitigates or eliminates a

1. Factors to Consider in Making Repair/Replace Decisions

Kiefner and Associates, Inc. 1 March 2015

Kiefner and Associates, Inc. 2 March 2015

1. Develop and implement a written integrity management plan.

David Johnson Energy Transfer

Kiefner and Associates, Inc. 3 March 2015

CFR Code of Federal Regulations

CMFL Circumferential magnetic flux leakage

CVN Charpy V-notch full-size-equivalent upper shelf energy, ft lb

DIMP Distribution Integrity Management Program

DC-ERW electric resistance welded pipe made with direct current

DSAW double submerged-arc welded pipe

ECDA external corrosion direct assessment

ERW electric resistance welded pipe

FERC Federal Energy Regulatory Commission

FFS Fitness for service

HCA high consequence area

HDB hydrostatic design basis in conjunction with polyethylene pipe

HSC hydrogen stress cracking

HVL highly volatile liquid

Hz hertz (a unit of frequency equivalent to one cycle per second)

ICDA internal corrosion direct assessment

Kiefner and Associates, Inc. 4 March 2015

ILI in-line inspection

KHz kilohertz (a unit of frequency equivalent to 1,000 cycles per second)

MFL magnetic flux leakage

MOP maximum operating pressure (usually associated with a hazardous liquid

NAEC Narrow axial external corrosion

NDE Non-destructive examination

NTSB National Transportation Safety Board

PCIF pressure-cycle-induced fatigue

PHMSA Pipeline and Hazardous Materials Safety Administration

PIR Potential impact radius as defined in 49 CFR 192.903

SCADA supervisory control and data acquisition

SCC stress corrosion cracking

SCCDA stress corrosion cracking direct assessment

SMYS specified minimum yield strength

SSAW single submerged-arc welded pipe

SSWC selective seam weld corrosion

Kiefner and Associates, Inc. 5 March 2015

High Consequence Area (HCA)

Legacy Construction Techniques and Legacy Repair Methods

Legacy construction techniques refer to obsolete and potentially problematic construction

Modern pipe refers to all pipe that is not legacy pipe.

Kiefner and Associates, Inc. 6 March 2015

49 CFR 192.3 Definitions Applicable to These Guidelines

Operator means a person who engages in the transportation of gas.

Kiefner and Associates, Inc. 7 March 2015