FINAL REPORT

GTI PROJECT NUMBER 21873

Project Title: Statistical Review and Gap

Analysis of LNG Failure Rate Table

Contract Number: DTPH56-15-T-00008

Date Issued:
Jan. 11, 2017

Prepared for:
Mr. Buddy Secor
U.S. Department of Transportation
Pipeline and Hazardous Materials Safety Administration
Office of Pipeline Safety

GTI Project Manager and Technical Contact:

Rich Kooy, P.E., Senior Program Manager
847-768-0512
[email protected]

GTI Team Members:

Ernest Lever, R&D Director, Infrastructure
Saurav Acharya, Ph.D., Data Scientist
Tishun Peng, Engineer

CH-IV International Inc. Team Members:

Phil Suter, LNG Consultant
Jenna Wilson, LNG Consultant

Prepared by:

Gas Technology Institute

1700 S. Mount Prospect Rd.
Des Plaines, Illinois 60018
www.gastechnology.org
Legal Notice

This information was prepared by Gas Technology Institute (“GTI”) for DOT/PHMSA (Contract
Number: DTPH56‐15‐T‐00008).

Neither GTI, the members of GTI, the Sponsor(s), nor any person acting on behalf of any of them:

a. Makes any warranty or representation, express or implied with respect to the accuracy,
completeness, or usefulness of the information contained in this report, or that the use of any
information, apparatus, method, or process disclosed in this report may not infringe privately‐owned
rights. Inasmuch as this project is experimental in nature, the technical information, results, or
conclusions cannot be predicted. Conclusions and analysis of results by GTI represent GTI's opinion
based on inferences from measurements and empirical relationships, which inferences and
assumptions are not infallible, and with respect to which competent specialists may differ.

b. Assumes any liability with respect to the use of, or for any and all damages resulting from the use of,
any information, apparatus, method, or process disclosed in this report; any other use of, or reliance
on, this report by any third party is at the third party's sole risk.

c. The results within this report relate only to the items tested.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page ii
Table of Contents
Legal Notice .................................................................................................................................. ii 
Table of Contents .......................................................................................................................... 1 
Summary of Research and Impact from Research Results .......................................................... 4 
1. Introduction ............................................................................................................................ 12 
1.1 Project Overview .................................................................................................. 12 
1.2 Project Context ..................................................................................................... 12 
1.3 Project Goals ......................................................................................................... 13 
2. Failure Rate Table Overview and Review of 3E-5 to 5E-5 FRT Criterion .............................. 17 
2.1 Derivation of FRT .................................................................................................. 17 
2.2 Summary of FRT’s Criterion Threshold Methodology .................................... 17 
2.3 Review of 3E-5 to 5E-5 FRT Criterion .............................................................. 18 
2.4 Comparison to Other U.S. Risk Criteria ........................................................... 21 
3. Failure Rate Reference Sources: Identification .................................................................... 27 
3.1 Reference Sources Cited in Current FRT ......................................................... 27 
3.2 Additional Primary Reference Sources Identified........................................... 28 
4. Failure Rate Reference Sources: Qualitative Analysis ......................................................... 31 
4.1 Review and Analysis Applicable to Multiple Equipment Categories ........... 31 
4.2 Cryogenic Atmospheric Storage Tanks ............................................................ 42 
4.3 Process Vessels, Distillation Columns, Heat Exchangers and Condensers46 
4.4 Truck Transfer – Arms and Hoses ..................................................................... 50 
4.5 Ship Transfer – Arms (and Hoses) ................................................................... 54 
4.6 Piping – Rupture (and Leak) of Valve .............................................................. 56 
4.7 Piping – Rupture of Expansion Joint ................................................................. 59 
4.8 Piping – Failure (Rupture or Leak) of Gasket ................................................. 61 
4.9 Piping – By Diameter ........................................................................................... 64 
4.10 Summary of Qualitative Review and Analysis .............................................. 73 
5. Failure Rate Reference Sources: Quantitative Analysis Methodology.................................. 75 
5.1 Overall Methodology ............................................................................................ 75 
5.1.1 Framework Layout and Component Protocols ............................................ 75 
5.1.2 Database Development ................................................................................... 76 
5.1.3 Bayesian Network Statistical Analysis .......................................................... 76 
5.1.4 Custom Input/Output Script ........................................................................... 77 
5.2 Sensitivity Analysis .............................................................................................. 77 
5.2.1 Analysis Using Perceived Relevancy Factors ............................................... 77 
5.2.2 Analysis Using Example Wisdom of the Crowd ........................................... 80 
5.2.3 Analysis Using Alternate Bayesian Priors ..................................................... 81 
6. Failure Rate Reference Sources: Quantitative Analysis Results and Recommendations .... 82 
6.1 Overall Summary of Framework, Sources and Results ................................ 82 

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 1
 6.1.1 Results Using Perceived Relevancy Analysis ............................................... 83 
6.1.1 Results Using Example Wisdom of the Crowd Analysis ............................ 84 
6.2 Cryogenic Atmospheric Storage Tanks ............................................................ 84 
6.3 Process Vessels, Distillation Columns, Heat Exchangers and Condensers85 
6.4 Truck Transfer – Arms and Hoses ..................................................................... 90 
6.5 Ship Transfer – Arms (and Hoses) ................................................................... 95 
6.6 Piping – Rupture (and Leak) of Valve .............................................................. 97 
6.7 Piping – Rupture of Expansion Joint ............................................................... 101 
6.8 Piping – Failure (Rupture or Leak) of Gasket ............................................... 102 
6.9 Piping – By Diameter ......................................................................................... 105 
7. Gap Analysis of FRT and Failure Rate Data ........................................................................ 117 
7.1 Age of Available Relevant Failure Rate Data ................................................ 117 
7.2 Inability to See Some Underlying Data.......................................................... 117 
7.3 Non-Uniform Nomenclature.............................................................................. 118 
7.4 Use of Predicted Rather than Observed Failure Rates ................................ 118 
7.5 Very Limited Cryogenic or LNG Equipment Failure Rate Data .................. 118 
7.6 Relevancy of Available Failure Rate Data to LNG ........................................ 119 
7.7 Gaps in Current FRT Categories ...................................................................... 121 
8. Summary and Recommendations ........................................................................................ 122 
8.1 Summary and Recommendations Regarding FRT Criterion ....................... 122 
8.2 Summary and Recommendations Regarding FRT Failure Rates and
Categories ....................................................................................................... 123 
8.3 Recommendations Regarding Subsequent Research and Related Efforts
.......................................................................................................................... 125 
9. Summary of Project Final Financial Contributions ............................................................... 125 
Appendix A: PHMSA LNG Plant Requirements FAQ - Design Spill Determination ................. 126 
Appendix B: Potential Revisions to the FRT Considered in this Research .............................. 130 
Appendix C: Equipment Protocol Framework and Applied References ................................... 132 
Appendix D: “Perceived Relevancy” Analysis Results Using All References .......................... 142 
Appendix E: “Percieved Relevancy” Analysis Results Using All References Except “CCPS ‘89”,
“OREDA ‘15” and “INL CHEM ‘95” ............................................................................................ 150 
Appendix F: “Wisdom of the Crowd” Analysis Results Using All References .......................... 158 
Appendix G: Excerpts from Analysis of LNG Peakshaving Facility Release Prevention Systems
(PNL-4153) ............................................................................................................................... 168 
Appendix H: Excerpts from Analysis of LNG Import Terminal Release Prevention Systems
(PNL-4152) ............................................................................................................................... 183 
Appendix I: Excerpts from LNG Terminal Risk Assessment Study for Oxnard, California (SAI-
75-615-LJ)................................................................................................................................. 191 
Appendix J: Analysis of PHMSA’s Onshore Natural Gas Transmission Pipeline Incident Data
.................................................................................................................................................. 193 
Appendix K: Analysis of PHMSA’s Onshore Hazardous Liquids Pipeline Accident Data ........ 199 

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 2
Appendix L: Visualization of UK HSE Hydrocarbon Releases Database (HCRD) ................... 204 
Appendix M: Cumulative Leak Frequency Curves Developed from UK HSE Hydrocarbon
Releases Database (HCRD) ..................................................................................................... 220 
Appendix N: Subject Matter Experts Contacted in Addition to Technical Advisory Panel........ 227 
Appendix O: Acronyms ............................................................................................................ 230 
Appendix P: Bibliography ......................................................................................................... 232 
Appendix Q: References .......................................................................................................... 239 

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 3
Summary of Research and Impact from Research Results

This report summarizes the results of U.S. Department of Transportation (DOT) Pipeline and
Hazardous Materials Safety Administration (PHMSA) contract #DTPH56-15-T-00008 to review
and consider recommendations for potential refinements of the LNG Failure Rate Table
(“FRT”), which establishes the criteria for Design Spills into impounding areas at LNG facilities
serving only vaporization, process or LNG transfer areas in order to comply with US 49 Code of
Federal Regulation Part 193. US 49 CFR Part 193 incorporates by reference specific sections of
the 2001 and 2006 editions of NFPA 59A Standard for the Production, Storage, and Handling of
Liquefied Natural Gas (LNG).
Unfortunately, neither NFPA 59A (2001 edition) or US 49 CFR Part 193 define Single
Accidental Leakage Sources (“SALS”) for a Design Spill in this impounding area. This resulted
in the U.S. Federal Energy Regulatory Commission (FERC) receiving a wide variety of
interpretations for a Design Spill for these impounding areas in applications proposed for LNG
facilities during the 2010-11 timeframe.
FERC and PHMSA established the FRT and its threshold criteria methodology in 2012 in order
to establish a consistent approach to determine SALS for this Design Spill in LNG facility
applications. FERC and PHMSA established a threshold criteria of 3 x 10-5 failures per year for
equipment and piping segments located in this impounding area, as a level of the risk associated
with the rupture of a LNG storage tank (i.e. container) outlet line (i.e. a Design Spill specified for
containment by NFPA 59A 2001 edition). FERC’s and PHMSA’s determination of this criteria
of 3 x 10-5 failures per year was based in part on a review of five references from 1979-1984.
Accurate estimates of equipment failure rates are important. Specifying “conservative”
equipment failure rates greater than those supported by actual representative field data can have
unintended impacts. The goals of this research effort were to:
 Review the basis for the baseline threshold failure rate criterion of 3E-5 failures per year.
 Develop additional foundational basis for PHMSA’s current LNG FRT, by applying a
rigorous statistical analysis and expanding the data set analyzed.
 Identify gaps in PHMSA’s current LNG FRT, such as equipment items not addressed and
gaps in available failure rate data.
 Propose potential specific revisions to the FRT, as appropriate.
 Identify key follow-on research desirable for further refinements.
The scope of this research did not include generating any new statistical data or estimates of
equipment leak or rupture failure rates.
The project team reviewed more than 150 references sources listed in the References and
Bibliography and considered in detail more than 20 additional data sources beyond those
identified in FERC’s issuance #201203010016. These data sources were identified through the
combined efforts of the project literature review efforts, Technical Advisory Panel (TAP), and
contacts to more than 50 Subject Matter Experts beyond the TAP.
The project team found limited LNG- or cryogenic-specific failure rate data beyond that
previously identified by FERC in its 2012 issuance #20120301-0016. For example:

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 4
  The last survey of LNG equipment failure rates in the U.S. was conducted 35 years ago.
FERC and PHMSA referenced this study in their establishment of the FRT.
 The most recent equipment failure rate data from The Society of International Gas
Tanker and Terminal Operators (SIGTTO) identified in this research issued 20 years ago.
 The International Group of LNG Importers Ltd. (GIIGNL) presented a technical paper at
the LNG17 conference in 2013 that provided failure/leak rate statistics, but it analyzed
LNG terminal subsystems and provided no break-out data for specific components.
 British Compressed Gases Association reviewed failure rates of cryogenic tanks in
Technical Information Sheet 23 issued in 2012, but this yielded a comparative minimum
failure rate from the industrial gas industry (since no failures were observed) and only for
a combination of equipment categories in the FRT.
 The results of two phases of fatigue and crush tests of cryogenic truck transfer hoses
performed in Europe in 2016 were summarized and compared to the truck transfer hose
failure rate specifications in the FRT.
 Three fault tree analysis estimates of predicted membrane tank failure rates were
identified and included for PHMSA’s and FERC’s information.
The project team also identified and included other non-LNG-specific failure rate data sets for
review and consideration. These included: PHMSA’s natural gas transmission and hazardous
liquid pipeline rupture data, and comparable data from European natural gas pipeline operators;
the Handbook Failure Frequencies used by the Flemish Environment, Nature and Energy
Department; two Release Frequency Manuals from the International Oil and Gas Producers
Association (IOGP); API Recommended Practice 581 Risk Based Inspection Methodology;
failure rate data for chemical processing plant equipment and nuclear power plant components
developed by Idaho National Laboratory; and other data as summarized in Section 3.2.
The project team also reviewed key LNG-specific equipment failure rate references produced
during 1979 – 1984. This review indicated that some of the Mean Time Between Failure data
applied as leak or rupture failures was apparently actually failure to operate, and in some cases
minimum rates were apparently applied as actual rates. This understanding further limited the
LNG-specific equipment failure data content in this project’s database, as detailed in Section 4.
The project team also gathered available failure rate data of atmospheric LNG storage tanks (i.e.
containers) for PHMSA’s and FERC’s information. The Design Spill for atmospheric LNG
storage containers is already specified in Table 2.2.3.5 of NFPA 59A 2001 edition specified by
US 49 CFR Part 193.
Some of the available generic equipment failure rate data is quite dated and not necessarily
reflective of modern design practices. Section 7 includes a quote by representatives of the Dutch
National Institute of Public Health and the Environment (RIVM) commenting that the set of
available failure frequencies was not up-to-date in 2006, not only in the Netherlands but all over
Europe and in fact worldwide. Recent efforts by RIVM include a review of its failure rates
specified for double walled LNG storage tanks (with an apparent focus on LNG pressure vessels)
and those results are pending, as noted in Section 4.3.
Some of the underlying data could not be evaluated. European regulatory bodies such as the UK
Health and Safety Executive (HSE) have developed detailed guidelines for evaluating risks and
detailed Quantitative Risk Assessments (QRAs) for significant land planning, such as in the

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 5
HSE’s Failure Rate and Event Data for use within Risk Assessments (FRED) document. While
the FRED document provides very good high-level information about the basis of its failure rate
specifications, many of its citations are confidential.
One modern equipment failure rate database that is publicly available and has significant content
and pedigree is the UK’s Hydrocarbon Release Database (HCRD) System, which contains
failure rate data continuously collected from UK off-shore oil and gas operations since 1992.
This data provides the basis for DNV GL’s LEAK commercial software that estimates leak
frequencies for piping and process equipment, including many of the equipment categories listed
in the FRT. Appendix L illustrates this database and Appendix M provides cumulative
probability failure rate distributions developed by the project team from this database to help
PHMSA and FERC better understand this resource.
Despite the age and limited amount of underlying LNG or cryogenic failure rate data, FERC’s
and PHMSA’s consideration in 2011-12 of the UK HSE FRED and Dutch RIVM equipment
failure rate frequency guidance documents (and additional references cited) is consistent with
other leading global regulatory authorities and appears to be properly included as “best
available” data, including for LNG facilities as highlighted below. Also included below is
information from IOGP and a relevant 2015 ISO standard.
 The UK HSE’s FRED document specifically defines failure rates for “LNG Refrigerated
Vessels” in its Item FR 1.1.2.1. Its specification for failure rates for “Pipework” (in Item
FR 1.3) has 44 citations, which include one confidential analysis of pipework associated
with bulk storage of liquid oxygen and two gas terminal studies.
 Singapore requires using HSE’s FRED document for Fixed Installations in its QRA
Technical Guidance Manual issued August 2016.
 ISO Technical Specification ISO/TS16901:2015 Guidance on performing risk assessment
in the design of onshore LNG installations including the ship/shore interface in its
section A.3 lists many of the same references considered in this analysis and concludes
“It should be noted that there are no publicly available incident databases for LNG plants
that can be available to derive leak frequencies and therefore should rely on the above
more general data.”
 The Dutch government agency RIVM in its Reference Manual Bevi Risk Assessment
2009 document stipulates on page 3 of its Introduction that the “SAFETI-NL” calculation
package by DNV London is “stipulated for carrying out the Quantitative Risk
Assessment (QRA) calculations for establishments that fall under the Bevi”. “SAFETI-
NL” contains equipment failure rate data based apparently in large part on the UK
HCRD.
 The European Industrial Gas Association in its guidance document 60/15 to apply
SEVESO Directives to cryogenic industrial gas facilities in Europe also highlights on p.
21 that the standardized version of DNV’s PHAST/Risk software named “SAFETI-NL”
must be used in the Netherlands.
 IOGP’s comments regarding the applicability of the recommendations in its Risk
Assessment Data Directory: Process Release Frequencies document #434-1 dated March
2010 include: “We therefore recommend use of the same frequencies for LNG
installations as given in Section 2.0” [i.e. for other oil & gas facilities]. IOGP bases its
release frequencies for process equipment and piping on DNV’s analysis of the HCRD.
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 6
 As an additional point of comparison, a representative of Korea Gas Corp. (KOGAS)
indicated that KOGAS primarily uses IOGP data for its QRAs.
Because the consideration of a significant amount of equipment failure data from related but
different (non-LNG or non-cryogenic) service was required, the project team used a Bayesian
Network statistical methodology to assess the data. The project team built a framework and
database, and applied the data using a commercial Bayesian risk analysis software coupled with
a custom input/output script. The project team considered the following two scenarios in its
sensitivity analysis, in addition to considering various Bayesian prior distributions:
1. “Perceived Relevancy” Scenarios:
The project team analyzed failure rate data by its source, its applicability, its site location,
and type (using analyst judgement). Twelve variations of analysis factors were provided
in order to help illustrate various impacts on results for different scenario examples. This
analysis was informative but the lack of specific LNG or cryogenic data limited the
ability to apply new LNG-specific data to the existing generic prior data. Additional
Percieved Relevancy scenarios can also be considered and evaluated in future analysis, if
PHMSA and FERC would like to see future refinements of this analysis.
2. “Wisdom of the Crowd” Scenarios:
The project team analyzed failure rate references by illustrating Wisdom of the Crowd
scenarios, and developed examples that sought to reduce duplication of underlying source
data and put an increased weighted importance on the more well-known or commonly-
used international references or failure rate regulatory guidance documents. Reviewing
the Wisdom of the Crowd example results provided another useful way to quantitatively
compare the overall data sets and develop the results and the recommendations, in
conjunction with the qualitative analysis and conclusions. Additional Wisdom of the
Crowd scenarios can also be considered and evaluated in future analysis, if PHMSA and
FERC would like to see future refinements of this analysis.
Appendix B concisely summarizes the recommendations from this project regarding specific
categories and failure rates specified in the FRT, and which are also intended to support the
strengthening of consensus standards. A summary of key findings and recommendations for
FERC’s and PHMSA’s consideration is:
 The leak and guillotine rupture failure rates specified for piping in the FRT directly
match those in HSE’s FRED guidance document (with only two small exceptions). The
guillotine rupture failure rates specified for piping in the FRT are slightly less than, or
greater than, the guillotine rupture failure rates specified by Dutch regulators (depending
on pipe diameter), and also appear reasonable with compared to the actual rupture rates
observed in PHMSA’s onshore natural gas and hazardous liquid pipeline database (2010-
2015). The single actual data point for LNG piping leak rates identified in this analysis
(from the 1981 survey of LNG peak shaving plants) also matches up favorably with the
FRT’s current specifications for piping. It is recommended to consider retaining the
current piping failure rate specifications in the FRT, but to also consider eliminating the
smallest piping size category (d < 2”) and the current 40”D maximum limit of the largest
piping size category (in order to enhance clarity to end-users of the FRT). Section 6.9
provides more details regarding piping-related findings and recommendations.
 The FRT currently does not address pipe-in-pipe piping such as the vacuum-jacketed

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 7
 piping that is sometimes used in LNG facilities. It is recommended to consider
specifying failure rate reduction factors of 0.01 and 0.1 to modify piping failure rates
when one of either two different types of pipe-in-piping is used. These reduction factors
are based on a review of a 1982 analyses by Pacific Northwest Laboratory (PNL), a 2012
analysis of double containment piping by Idaho National Laboratory, and a relevant
Center for Chemical Process Safety reference.
 The FRT currently does not address potential modification factors in piping failure rates
for long transfer lines or inter-unit piping (as commonly exist in LNG terminals). Some
relevant information was identified for PHMSA’s and FERC’s review and future
consideration on a generalized basis, beyond those engineering analyses submitted for
FERC’s and PHMSA’s consideration on a case-by-case project application basis.
 The leak and guillotine rupture failure rates for LNG transfer hoses and arms specified in
the FRT were compared to other references in this study, and also to results of recent
tests in Europe and other developments. There are significant technology developments
underway on LNG transfer hoses and arms for both ships and trucks (trailers and trailer-
mounted ISO containers). This equipment category was identified as an important watch
area for PHMSA and FERC to seek new failure rate data in the future, because for
example results of crush and fatigue tests on cryogenic hoses in Europe in 2016 indicates
that the truck hose leak rate specified in the FRT may be about 100 times too
conservative, and also that a guillotine rupture of a truck transfer hose may not be a
credible event. However, at the present time it is recommended that: 1) if FERC and
PHMSA want to retain the current basis of “Failures per year of operation”, then they
may consider making no changes to the FRT’s specification; and 2) that PHMSA and
FERC review this new cryogenic test data and consider removing the rupture frequency
specifically for multi-composite hoses. If FERC and PHMSA want to consider changing
the FRT’s basis of specification from “Failures per year of operation” to “Failures per
hour of operation”, then they could consider applying the failure frequency rates
specified by Dutch and Flemish regulators. Sections 6.4 and 6.5 provide more details.
 The failure rate for valves specified in the FRT was compared to other references in this
study. It is recommended to retain the FRT’s currently-specified rupture rate for valves,
but to also add a leak rate consistent with that specified in HSE’s FRED guidance Item
FR 1.2.1 for valves, using the 2 mm hole size considered in Section 4.6 of this report.
 The failure rate for gaskets (i.e. flanges) specified in the FRT was compared to other
references in this study. It is recommended to consider eliminating the “Failure of
Gasket” terminology and specify failure rates consistent with those specified in HSE’s
FRED guidance item FR 1.2.4 for gaskets (and flanges), using the 25 mm and 50 mm
hole sizes considered in Sections 4.8 of this report.
 The failure rate for expansion joints specified in the FRT was compared to other
references in this study and to underlying references. It is recommended to consider
revising the specified failure rate from 4 x 10-3 to 1 x 10-4 failures/year in order to better
align it with the 1975 risk analysis performed by Science Applications Inc. (excerpts
shown in Appendix I) and to also clarify that it applies to single ply expansion joint; both
the existing and recommended rates are greater than the 3 x 10-5 FRT threshold criterion.
In addition, it is recommended to specify a rupture failure rate for double ply expansion
joints.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 8
  The guillotine rupture failure rate specified for “Process Vessels, Distillation Columns,
Heat Exchangers and Condensers” in the FRT aligns well with the rate specified for
Pressure Vessels in HSE’s FRED guidance document and also with the rates specified by
Dutch and Flemish regulators for Process Vessels. The 10 mm hole leak failure rate
specified in the FRT also aligns well with the 10 mm hole leak failure rates specified by
Dutch and Flemish regulators for process vessels, and is more conservative than than the
10 mm hole leak failure rates specified for Pressure Vessels in HSE’s FRED guidance
document. This research also reviewed and summarized the lower failure rates assigned
to Pressure (Storage) Vessels (vs. Process Vessels) by Dutch and Flemish regulators than
the rates specified by the UK HSE’s FRED guidance document for Pressure (Storage)
Vessels, but it was recommended that FRT retain its more conservative basis and not
specify lower failure rates specifically for Pressure (Storage) Vessels as different than
those rates for Process Vessels. This research also explored, accumulated and analyzed
for PHMSA and FERC a number of potential equipment subcategories within this overall
category. In summary it was recommended that the FRT retain this existing category
name and its currently-specified rupture and leak failure rates.
 The FRT currently specifies leak and rupture rates for cryogenic atmospheric storage
tanks. For determination of single accidental leakage sources for process facilities this
information does not appear to be relevant because the Design Spill for LNG tanks (i.e.
containers in NFPA 59A) is already specified in Table 2.2.3.5 of the NFPA 59A (2001
edition) specified by US 49 CFR Part 193. However, if DOT allows for design spills to
be selected using a failure rate or risk based approach (e.g., NFPA 59A 2016 edition)
then this information may be relevant. Therefore, it is recommended that PHMSA
consider eliminating this category or clarify the use of this information for single
accidental leakage sources in impounding areas serving only vaporization, process or
LNG transfer areas.
This research also reviewed the threshold criterion of 3 x 10-5 failures per year used in FRT. The
3 x 10-5 failures per year criterion appears to be based on two key analyses conducted by Pacific
Northwest Laboratories (PNL) in 1982, which are among five references cited by FERC in the
derivation of the FRT Criterion. The two PNL studies used fault tree analysis to estimate a number
of failure rate frequencies for various LNG plant components and subsystems. The project team
confirmed that the 3 x 10-5 to 5 x 10-5 failure rate threshold criterion was contained in the historical
references, and their observations of these and other references in Section 2.3 also included that:
 The 3 x 10-5 failures per year criterion refers to an estimated failure rate for an equipment
component failure, and not the resulting level of risk to society or an individual as is
analyzed in some other risk assessment methodologies.
 The criterion appears to be derived from the failure rate associated with the failure of one
of the most critical plant components, i.e. the LNG container liquid withdrawal line. The
failure rate criteria associated with the failure of line connection penetrations to the
primary container may or may not set an equivalent risk level required for all other
equipment component failure rates.
 The 3 x 10-5 failures per year criterion appears to be associated with the failure rate of an
LNG container liquid withdrawal line isolated with an internal shutoff valve, while US 49
CFR Part 193 also permits evaluation of events with other consequences such as the
nearly complete emptying of the LNG container.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 9
  While for example the failure rates in the FRT for process piping exactly match HSE’s
FRED document (with two small exceptions), the FRT’s threshold criteria methodology
appears to be a somewhat unique risk assessment methodology.
 A risk assessment prepared in 1975 by Science Applications Inc. for the Western LNG
Terminal Company for a LNG terminal proposed for Oxnard, CA also prepared a number
of FTAs and estimated the probability of a “Leak Occurs in Storage Tank or Tank
Outlet” as 1 x 10-6 failures/year and a “Leak Occurs in Outlet of Tank” as 9.8 x 10-11
failures/year (when the isolation systems were considered).
 No direct analysis was made to compare the current FRT criterion of 3 x 10-5 failures per
year to other regulated risk criteria in the U.S., since such an analysis of the risks to
individuals and society of applying the FRT’s methodology and criterion to vaporization,
process, or LNG transfer components versus the risks associated with failure of a primary
LNG container outlet line was beyond the scope of this analysis. This analysis may
include a comparison of a 1-hour spill versus a complete loss of containment.
Recommendations regarding the FRT criterion are:
 No direct basis was identified to propose revising current the FRT Criterion of 3 x 10-5, for
the reason identified in the prior bullet above. Therefore it is recommended to retain the
FRT Criterion of 3 x 10-5.
 FERC and PHMSA should further consider whether methodologies other than the current
FRT criterion methodology would be suitable for defining a SALS for impoundment
areas serving only vaporization, process, or LNG transfer areas.
Recommendations regarding subsequent research and related efforts are:
 PHMSA and FERC should consider funding research to conduct a new survey of LNG
facilites in the US, in order to update and expand upon the most recent survey study
completed in 1981. It was “the opinion of GRI that the failure rate data base should be
periodically updated, at about 5-year intervals” as stated in that survey, but unfortunately
no industry or federal funding was made available for this purpose.
 PHMSA and FERC should consider supporting the coordination of any new industry-
government consortium efforts to create a national database of information related to in-
service performance of LNG piping and components.
This research performed in this project was led by and substantially performed by Gas
Technology Institute (GTI). CH-IV International Inc. was a project subcontractor and
contributed end-user perspective and insights, based on their experience working with many
LNG facility developers to use the LNG FRT and submit design engineering submittals to
FERC. The recommendations provided in this report are solely for PHMSA’s and FERC’s
consideration.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 10
The research project team also expresses their sincere thanks to PHMSA and to the members of
this project’s Technical Advisory Panel for their comments, input and insights:
 Ari Flores and Howard Goldberg, Consolidated Edison, Inc.
 Jon Huddleston, NW Natural
 Brandon Jones, Paiute Pipeline Co./Southwest Gas Corp.
 Andrew Kohout, FERC
 Roy Lucas, PHMSA consultant
 Jeff Marx, Quest Consultants Inc.
 Frank Su and Skip Doucette, National Grid
 Mike Wardman and Jill Willday and their colleagues, UK HSL

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 11
1. Introduction

1.1 Project Overview

This report summarizes the results of U.S. Department of Transportation (DOT) Pipeline and
Hazardous Materials Safety Administration (PHMSA) contract #DTPH56-15-T-00008 to review
and consider potential refinements to PHMSA’s LNG Failure Rate Table1 (“FRT”), which is
referenced by the answers to question #DS1 and #DS2 of PHMSA’s current information resource
webpage “LNG Plant Requirements: Frequently Asked Questions”2 and also in Volume II of
U.S. Federal Energy Regulatory Commission’s (FERC’s) Guidance Manual for Environmental
Report Preparation draft version issued December 2015. The current FRT is shown in Table 1.
The FRT establishes the criteria for Design Spills from vaporization, process or LNG transfer
areas at LNG plants in order to comply with US 49 Code of Federal Regulation Part 193.
Gas Technology Institute (GTI) led and substantially performed the research in this project.
CH-IV International Inc. served as a project subcontractor and contributed valuable end-user
perspective and insights (in addition to the project’s Technical Advisory Panel), based on their
experience working with many LNG facility developers to use the FRT and submit design
engineering submittals to FERC.

1.2 Project Context

FERC’s and PHMSA’s introduction of the LNG Failure Rate Table in 2012 coincided with a
number of subsequent significant changes in the U.S. LNG industry, including:
 A large increase in activity in the U.S. to develop LNG as an export commodity, as a
transportation fuel, for utility peak shaving service (both upgrades to existing facilities
and new facilities), and other purposes.
 The 2013 and 2016 editions of NFPA 59A Standard for the Production, Storage, and
Handling of Liquefied Natural Gas (LNG) incorporate Quantitative Risk Assessments
(QRA) content in new Chapter 15, as an alternative to the prescriptive requirements in
Chapter 5. Table 15.6.1 of NFPA 59A 2013 and 2016 editions specify equipment failure
rates that differ from those in the FRT. This project’s analysis can hopefully help efforts
to harmonize future standards and regulations.
 The 2016 edition of NFPA 59A introduced membrane tanks. This development is not
directly relevant to the FRT since US 49 CFR Part 193 recognizes the 2001 and 2006
editions of NFPA 59A (specific respective sections), but the project team summarized
membrane tank failure rate data for PHMSA’s and FERC’s information since FERC and
PHMSA may receive applications that consider membrane tanks.
Accurate estimates of equipment failure rates are important. Specifying equipment failure rates
that are higher than those supported by actual representative field data can have unintended
impacts. For example:
 If a catastrophic guillotine rupture of a hose used to load or unload a LNG trailer or ISO
shipping container becomes a probable event as determined by the FRT, then designers
may need to propose locating the loading/unloading station near the center of an LNG
facility in order to meet the exclusion zone requirements in US 49 CFR Part 193. This

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 12
 can conflict with a common overall safety (and security) goal by plant designers and
operators to keep most vehicular traffic at the perimeter of a site. This may particularly
impact a number of new or existing sites that are constrained in size and for which LNG-
related improvements have been considered (e.g. for the development of LNG as a
marine vessel bunker fuel as one way for U.S. ship operators to meet low-sulfur Emission
Control Area requirements).
 If a catastrophic guillotine rupture of a long transfer line or other component becomes a
probable event as determined by the FRT, then designers may need to implement
measures to meet exclusion zone requirements that may have unintended impacts.
Examples of design case rupture analyses that have been required using the current FRT
as recorded in public filings on FERC’s website include these related to LNG export
plant applications:
o Freeport LNG – 8.67” diameter hole in LNG rundown line; 26” hole diameter in
marine transfer line
o Trunkline LNG – 6.7” diameter hole and 8” diameter hole in LNG rundown lines
o Magnolia LNG – 10” diameter hole in LNG rundown line
o Corpus Christi LNG – 10” diameter hole in marine transfer line; 10” diameter
hole in LNG rundown line
o Oregon LNG – 10.67” diameter hole in marine transfer line; 8” diameter hole in
LNG rundown line
o Downeast LNG – 12” diameter hole in marine transfer line
o Jordan Cove LNG – 36” diameter hole in marine transfer line; 8” diameter hole in
LNG rundown line
o Golden Pass LNG – 18” diameter hole in LNG rundown line
While these types of failures and events have not been observed at LNG facilities in the
U.S. or at LNG import/export terminals worldwide, LNG-specific piping failure rate data
is known to be limited due in part to the relatively small cumulative operating history of
the LNG industry when compared for example to the chemical process industry or overall
oil and gas industry. This research seeks to consider piping failure rate data from other
industries using a Bayesian analysis and also to review the FRT threshold criterion used
to determine required release scenarios.
This research is also consistent with PHMSA’s continued emphasis to update guidance on
critical components of risk assessment approaches, including appropriate risk metrics and
methods.34

1.3 Project Goals

The goals of this research effort were to:
 Review the basis for the baseline threshold failure rate criterion of 3E-5 failures per year.
 Develop additional foundational basis for PHMSA’s current LNG FRT, by applying a
rigorous statistical analysis and expanding the data set analyzed.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 13
  Identify gaps in PHMSA’s current LNG FRT, such as equipment items not addressed and
gaps in available failure rate data.
 Propose potential specific revisions to the FRT, as appropriate.
 Identify key follow-on research desirable for further refinements.

This research effort identified, reviewed, and analyzed existing leak and rupture equipment
failure rate data or references that may be potentially relevant to the FRT. The scope of this
research did not include generating any new statistical data or estimates of equipment leak or
rupture failure rates.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 14
 Table 1: Nominal Failure Rates Specified in Current FRT

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 15
 Table 1: Nominal Failure Rates Specified in Current FRT (cont.)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 16
2. Failure Rate Table Overview and Review of 3E-5 to 5E-5 FRT Criterion

2.1 Derivation of FRT

PHMSA’s current FRT was derived in partnership with FERC, and its derivation is summarized
in documents such as FERC Issuances 20111115-4001, 20120301-0016 and 20120507-4014. A
more detailed description of its derivation was provided by Andrew Kohout of FERC in a 2012
paper5 which noted that “The selection of the nominal failure rates were determined to provide
conservative and consistent trends failure rates among the various types of failures with
recognition that improved and updated failure rate data may be used pending review and
acceptance of FERC and PHMSA.” As noted both in this 2012 paper and in the title of the FRT,
the equipment failure rates specified in the FRT are nominal failure rates and not mean or
median failure rates. For more details the reader is referred to the 2012 paper which also
summarizes that the derivation of the FRT included consideration of: the fact that outliers in the
data can influence results versus only using mean or median values; that piping and equipment
categorizations differ across failure rate data sets which can make it difficult to derive accurate
mean or median values; and other factors.

2.2 Summary of FRT’s Criterion Threshold Methodology

PHMSA and FERC apply the FRT by utilizing a criterion threshold methodology to determine
the “Single accidental leakage source” (SALS) design spills for individual LNG plant
components or piping segments. Details are specified in Appendix A which contains the
“Design Spill Determination” information from PHMSA’s current “LNG Plant Requirements:
Frequently Asked Questions” webpage, and also in Volume II of FERC’s Guidance Manual for
Environmental Report Preparation issued December 2015. In brief, the FRT is to be applied to
numerous individual piping segments and equipment components in order to determine if the 3 x
10-5 per year failure rate criterion is equaled or exceeded for that particular piping segment and
equipment component; if so, then this design spill must be evaluated. PHMSA and FERC apply
this methodology and the FRT to all hazardous materials in proposed LNG facilities including
LNG, gaseous natural gas, propane, butane, ethylene, acid gases, etc.
The FRT criterion and methodology differs from some other risk assessment methodologies and
appeared to the project team to be somewhat unique, although another example of using a risk
threshold criterion methodology is a Maximum Credible Event scenario analyses6,7. For
example, the FRT criterion and methodology defines potential accident scenarios, estimates
potential accident frequencies, evaluates event consequences and impacts, and then requires the
impacts to not extend beyond the legal control of the operator. This is similar to a Quantitative
Risk Assessment (QRA) which also defines potential accident scenarios, estimates potential
accident frequencies, evaluates event consequences and impacts, but then compares the
cumulative risk to a Frequency-Number (F-N) curve or plot of allowable Societal Risk (SR).
The former essentially does not allow individual events with a certain failure rate from impacting
the public, but may allow certain cumulative risks from extending onto populated areas, while
the latter does not allow certain cumulative risks from extending onto populated areas, but may
allow for individual events with certain failure rates from impacting the public. One example of
a regulatory approach that utilizes a QRA methodology is the “RIVM BEVI ‘09” document
referenced in Section 3.1.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 17
2.3 Review of 3E-5 to 5E-5 FRT Criterion
The primary federal regulation that pertains to LNG facilities is US 49 CFR Part 193 Liquefied
Natural Gas Facilities: Federal Safety Standards; its “Subpart B – Siting Requirements”
incorporates NFPA 59A Table 2.2.3.5 (2001 edition) by reference such as in Section 193.2059.
NFPA 59A Table 2.2.3.5 (2001 edition) specifies that the “Design Spill” for “Impounding areas
serving only vaporization, process, or LNG transfer areas” must be “The flow from any single
accidental leakage source”. NFPA 59A Table 2.2.3.5 (2001 edition) also specifies Design Spill
for containers in three configurations.
The basis of the threshold failure rate criterion of 3 x 10-5 to 5 x 10-5 failures per year in the FRT
is summarized in the 2012 paper by Andrew Kohout of FERC8. The reader is referred to that
paper for details, but a brief summary follows with quotations from that paper:
 Unfortunately neither NFPA 59A (2001 edition) or US 49 CFR Part 193 define a single
accidental leakage source for “impounding areas serving only vaporization, process, or
LNG transfer areas”.
 FERC “received a wide variety of single accidental leakage sources, ranging from
packing and flange leaks to full guillotine ruptures of ship loading lines” in proposed
applications for LNG facilities.
 In order to achieve a consistent approach in LNG facility proposal applications submitted
to FERC, “FERC staff researched the failure rates of the Design Spill specified for
storage tanks” (in Table 2.2.3.5 of NFPA 59A 2001 edition). FERC staff determined that
“the rupture of a storage tank outlet line is on the order of one failure every 20,000 to
30,000 years (6 x 10-5 to 3 x 10-5 failures per year)” (sic) based on the failure rates
contained in five references (these references are referenced in Section 3.1 as “AGA FP
LNG ‘84”, “GRI LNG FRD ’81”, “AGA LNG EXP ‘79”, “PNL PSRP ‘82”, and “PNL
RP ’84”).
 “Because this failure rate applies to a design spill for containment that is specified by
NFPA 59A and adopted into 49 CFR Part 193, FERC staff used this rate to determine
appropriate single accidental leakage sources for impounding areas serving process areas.
PHMSA concurred with this approach.”
 “In addition, FERC staff also researched and compiled past and present internationally
used risk criteria … and determined the failure rate selected is also consistent with risk
criteria and consistent with risk criteria when considering multiple single accidental
leakage events (e.g., 1 – 100) may have overlapping consequences…”
This project team reviewed the five references and agreed with FERC’s findings that there are
estimates that the rupture of a storage tank outlet line will be “on the order of one failure every
20,000 to 30,000 years” (5 x 10-5 to 3 x 10-5 failures per year) within these references. For
example:
 “PNL PSRP ‘82” evaluated release prevention systems for an LNG Peakshaving Facility
that contained one single-containment tank (348,000 barrel capacity) with external pumps
(“LNG is pumped from the storage tank to the vaporizers by three vertical submerged,
pot-mounted pumps” as per p. 3.5) and through a 12” diameter outlet line with an internal

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 18
 valve and with one expansion joint. Section 4.2.6 “Outlet Line Rupture” on Page 4.11 of
“PNL PSRP ‘82” states that:
The outlet line from the storage tank is 12 in. in diameter and exits through the
bottom of the inner tank. LNG is drawn from the tank for vaporization about 20
days per year. The outlet line will rupture about 5 x 10-5 times per year, resulting
in a release of 28,000 gallons of LNG if stopped in one minute. If the release is
isolated in ten minutes, 280,000 equivalent gallons of LNG will be released. The
probability of the release occurring and not being stopped in one minute is 1 x
10-5 times per year. A fault tree analysis for rupture of the outlet line is shown in
Figure B.8. Critical system components are expansion joints, 12 in. internal
valve, and operators.
 “PNL ITRP ‘82” evaluated release prevention systems for an LNG import terminal that
contained two single-containment tanks (each 550,000 barrel capacity) with internal
pumps. “Each storage tank contains two submerged primary sendout pumps which boost
the LNG” and “All piping connections to the inner tank enter through the roof of the
storage tank and are supported by independent structures” as per pp. 3.3-3.4. Page 4.16
of “PNL ITRP ‘82” states that:
…Rupture of an outlet LNG line will occur about 3 x 10-5 times per year. If the
sendout system is shut down and the release isolated in 1 minute, 5,000 gallons
would be spilled. If the system is not shut down for 10 minutes, 41,000 gallons
would be released. This will occur 2 x 10-2 per demand, resulting in a probability
of about 5 x 10-7 per year for this scenario. If an inlet or outlet line ruptures, gas
detectors in the storage tank dike area will warn the operator that an emergency
condition exists and the operator would then have to activate the ESD. Fault trees
for rupture of storage tank inlet and outlet lines are shown in Figures B.10 and
B.11. Supporting calculations are given in Table B.7 and B.8.
Other observations from the review of these and other references from the 1970s-80s included:
 The 3 x 10-5 failures per year criterion refers to an estimated failure rate for an equipment
component failure, and not the resulting level of risk to society or to an individual as is
analyzed in some other risk assessment methods.
 The criterion appears to be derived from the failure rate associated with the failure of one
of the most critical plant components, i.e. the LNG container liquid withdrawal line. The
failure rate criteria associated with the failure of line connection penetrations to the
primary container may or may not set an equivalent risk level required for all other
equipment component failure rates.
 The criterion is derived from the risk level primarily associated with pool boiling
evaporation from a single accidental release from the primary LNG container, in
comparison to the analysis of multiple jetting and flashing release scenarios from piping
and other equipment as typically required to conform to the FRT.
 The Fault Tree Analysis (FTA) of “Large Release from Storage Tank Outlet Line” for
peak shaving plants in “PNL PSRP ‘82” (shown in Table 7 and Table B.7 in Appendix
A):
o included an internal valve.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 19
 o included one (single ply) expansion joint, and the estimated failure rate of the
expansion joint of 1 x 10-7 failure/hour drives the conclusion of this FTA (the Top
Event failure rate reduces from 1 x 10-5 to 1.5 x 10-7 if the expansion joint is
eliminated); the accuracy of the 1 x 10-7 failure/hour of expansion joint is of
limited pedigree, as described in section 4.7.
(While modern single containment tanks typically utilize in-tank pumps without bottom
penetrations and without expansion joints in LNG tank outlet lines upstream of the pump
discharge block valve, there is no prohibition against using the design analyzed in this
FTA.)
 NFPA 59A 2001 edition as enforced by US 49 CFR Part 193 permits LNG containers
with penetrations below the liquid level and without internal shutoff valves. Table
2.2.3.5 defines the duration of the Design Spill (using the formula q=(4/3)*d2*√h) to be:
o “1 hour” for “Containers with penetrations below the liquid level with internal
shutoff valves in accordance with 6.3.3.3”
o “until the differential head acting on the opening is 0 (zero)” for “Containers with
penetrations below the liquid level without internal shutoff valves”; that is, the
duration of the Design Spill is the nearly complete emptying of the primary LNG
container.
In other words, the FRT criterion could be associated with the consequences generated by
a 1 hour spill (of the peak shaving tank in “PNL PSRP ‘82”) while US 49 CFR Part 193
also permits evaluation of other consequences such as the nearly complete emptying of
the primary LNG container.
 An alternate FTA of “Release from Storage Tank Outlet Line with Alternative Design of
In-Tank Pumps” for peak shaving plants was also included in “PNL PSRP ‘82” (shown in
Table 7 and Table B.16 in Appendix A), which is more representative of LNG plant
design practices today but not of those when the SALS language of NFPA 59A was
adopted by US 49 CFR Part 193. As comparison information, the largest failure rate
estimate in this FTA is 1 x 10-6 failures/year of the three sub-scenarios considered in
calculation nos. 18, 19 and 20 on pp. B56-B57.
 An alternate FTA for an import terminal with in-tank pumps in “PNL ITRP ‘82” in its
base case scenario is shown in Table 4.1 and Table B.8 in Appendix B. It estimated the
probability of “Rupture of Storage Tank Outlet Lines” as 5 x 10-7 failures per year, and
not 3 x 10-5 failures per year, when it considered the Emergency Shutdown Device that is
included in all LNG plant designs (including in 1982).
 The estimated failure rates for tank outlet lines were predictions derived by Fault Tree
Analyses (based on a particular number of sendout/vaporization hours/year), and not
from any failure rate data of observed actual failures.
 The estimated failure rates for tank outlet lines were based on estimates of subcomponent
reliability that appear to be taken primarily from the nuclear power industry, including
the failure of a single ply expansion joint for which the pedigree of its estimated failure
rate is limited.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 20
  The SALS language appears to be fully formed in the 1975 (4th) edition of NFPA 59A,
but similar predecessor language appears in the 1971 (2nd) edition in paragraph 2135, and
in the 1972 (3rd) edition in paragraph 2111.
 A 1975 risk assessment by Science Applications Inc. prepared for the Western LNG
Terminal Company for a LNG terminal proposed for Oxnard, CA (referred to in Section
3.2 and this document as “SAI ‘75”) also prepared a number of FTAs and estimated the
probability that a “Leak Occurs in Storage Tank or Tank Outlet Lines” is 1 x 10-6 (for one
tank) and the probability that a “Leak Occurs in Outlet of Tank” as 9.8 x 10-11
failures/year (when the isolation systems are considered). This terminal was proposed to
use four 550,000 barrel capacity tanks with in-tank pumps. Appendix I contains excerpts
from this alternate estimate of failure frequency of a LNG tank outlet line and LNG tank
developed in the 1970s-80s.

2.4 Comparison to Other U.S. Risk Criteria

No direct analysis was made to compare the current FRT criterion of 3 x 10-5 to other regulated
risk criteria in the U.S., since such an analysis of the risks to individuals and society of applying
the FRT’s methodology and criterion to vaporization, process, or LNG transfer components
versus the risks associated with failure of a primary LNG container outlet line was beyond the
scope of this analysis. Such an analysis may also include comparing the impacts of a 1-hour spill
versus total loss of containment of a primary LNG container.
A direct comparative analysis may also involve a comparison to an F-N plot for allowable IR and
SR. PHMSA recently funded a significant body of risk-related research that provides significant
additional relevant information to aid PHMSA and FERC in their review of IR and SR levels;
see for example PHMSA research reports produced in 2016 entitled “Approaches for Preventing
Catastrophic Events” by GTI9, and “Paper Study on Risk Tolerance” by Keifner and Associates
Inc.10 This PHMSA research provides important information to support PHMSA’s future
regulatory decision-making.
NFPA 59A 2013 and 2016 editions illustrate F-N plots of societal risk criteria used by different
jurisdictions in Figure A.15.10.2 of those editions. Comparative F-N risk criteria information
from select U.S. governmental agencies also follows.

U.S. Federal Energy Regulatory Commission

FERC has developed public risk criteria when assessing tolerable individual and societal risk
guidelines for dams in its publication Risk-Informed Decision Making Guideline documents;
Figure 1 illustrates Incremental Risk Guidelines excerpted from Version 4.1 (March 2016) of its
Chapter 3: Risk Assessment.11

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 21
 Figure 1: FERC Risk Criteria F-N Plot and Risk Criteria for Dam Construction

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 22
U.S. Department of Transportation Federal Aviation Administration

The U.S. DOT FAA has established public risk criteria for commercial space launches. The
maximum risk to all members of the public is 1 x 10-4 casualties/launch, and the maximum risk
to an individual member of the public should not exceed a casualty expectation of 1 x 10-6 per
launch for each hazard. An excerpt from the current US 14 CFR Part 417.107 follows:

(b) Public risk criteria. A launch operator may initiate the flight of a launch vehicle
only if flight safety analysis performed under paragraph (f) of this section demonstrates that
any risk to the public satisfies the following public risk criteria:

(1) A launch operator may initiate the flight of a launch vehicle only if the total risk
associated with the launch to all members of the public, excluding persons in water-borne
vessels and aircraft, does not exceed an expected number of 1 × 10−4 casualties. The total
risk consists of risk posed by impacting inert and explosive debris, toxic release, and far
field blast overpressure. The FAA will determine whether to approve public risk due to any
other hazard associated with the proposed flight of a launch vehicle on a case-by-case basis.
The Ec criterion applies to each launch from lift-off through orbital insertion for an orbital
launch, and through final impact for a suborbital launch.

(2) A launch operator may initiate flight only if the risk to any individual member of
the public does not exceed a casualty expectation of 1 × 10−6 per launch for each hazard.

(3) A launch operator must establish any water borne vessel hazard areas necessary to
ensure the probability of impact (Pi) with debris capable of causing a casualty for water
borne vessels does not exceed 1 × 10−5.

(4) A launch operator must establish any aircraft hazard areas necessary to ensure the
probability of impact (Pi) with debris capable of causing a casualty for aircraft does not
exceed 1 × 10−6.

U.S. Department of Defense

The U.S. Department of Defense has established risk criteria standards for national test ranges.
Figure 2 shows an F-N plot and risk criteria established by DoD in its Standard #321-16 (most
recently revised in August of 2016). 12

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 23
 Figure 2: DoD Risk Criteria F-N Plot and Risk Criteria for National Test Ranges

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 24
U.S. Department of the Interior and U.S. Army Corps of Engineers
The U.S. Department of the Interior Bureau of Reclamation and U.S. Army Corps of Engineers
have established risk criteria standards for dam and levee safety, and Figure 3 shows the F-N
plots effective as of July 2015.13 The DOI uses annualized dam failure probability (similar to
Individual Risk criteria) and loss of life (similar to Societal Risk criteria) guidelines to assess
dam safety risk, and calls out the As Low as Reasonably Practicable (ALARP) region.

Figure 3. U.S. Dept. of Interior Bureau of Reclamation and U.S. Army Corp of Engineer F-N Plots

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 25
 Figure 3 (cont.). U.S. Dept. of Interior Bureau of Reclamation and
U.S. Army Corp of Engineer F-N Plots

PHMSA has also recently funded related research that evaluates the extent of vapor dispersion
exclusion zones using the current FRT and different dispersion models, such as the research in
2016 entitled “Comparison of Exclusion Zone Calculations and Vapor Dispersion Modeling
Tools” performed by CH-IV International Inc.14

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 26
3. Failure Rate Reference Sources: Identification

3.1 Reference Sources Cited in Current FRT

FERC cited the following twelve references in its 2012 issuance #20120301-0016 dated March 1,
2012. The abbreviated names for reference sources identified in [brackets] are used in this report,
as listed below as well as in the References and Bibliography.

1. NFPA Standard for the Production, Storage, and Handling of Liquefied Natural Gas
(LNG), NFPA 59A, 2012 DRAFT. [“59A ‘12 DRAFT”]
2. Mniszewski, K.R., IIT Research Institute, “Fire Protection Planning for LNG Facilities”,
AGA Distribution Transmission Conference Proceedings, San Francisco, California, May
7-9, 1984. [“AGA FP LNG ‘84”]
3. Johnson, D.W., and Welker, J.R., Applied Technology Corp. Development of an
Improved LNG Plant Failure Rate Database, Final Report for Gas Research Institute,
GRI-80/0093, 1981. [“GRI LNG FRD ’81”]
4. Welker, J. R., and H. P. Schorr, LNG Plant Experience Database, American Gas
Association, Operating Section Proceedings, United States, 1979. [“AGA LNG EXP
‘79”]
5. Pelto, P.J., Baker, E.G., Holter, G.M, and Powers, T.B., Analysis of LNG Peakshaving
Facility Release Prevention Systems, Pacific Northwest Laboratory, PNL-4153, 1982.
[“PNL PSRP ‘82”]
6. Pelto, P.J. and Baker, E.G., Analysis of Liquefied Natural Gas (LNG) Release Prevention
Systems, Pacific Northwest Laboratory, PNL-SA-12278, AIChE Summer National
Meeting, August 1984. [“PNL RP ‘84”]
7. Mannan, Sam, Lees’ Loss Prevention in the Process Industries – Hazard Identification,
Assessment and Control, Third Edition, Volume 3, Appendix 14, Elsevier, Inc. 2005.
The project team reviewed the Third edition and also the Fourth Edition of this reference
by Mannan, Sam, Elsevier, Inc., 2012. [“LEES ‘12”].
8. TNO Guidelines for Quantitative Risk Assessment (TNO Purple Book), Committee for
the Prevention of Disasters (CPR), National institute of Public Health and the
Environment (RIVM), The Netherlands Organization for Applied Scientific Research
(TNO). First edition 1999/2005. [“TNO PURPLE ‘05”]
9. TNO Methods for the Determination of Possible Damage (TNO Green Book), Committee
for the Prevention of Disasters (CPR), National Institute of Public Health and the
Environment (RIVM), The Netherlands Organization for Applied Scientific Research
(TNO). First edition 1992 [“TNO GREEN’ 92”]
10. TNO Methods for Determining and Processing Probabilities (TNO Red Book),
Committee for the Prevention of Disasters (CPR), National Institute of Public Health and
the Environment (R1VM), The Netherlands Organization for Applied Scientific Research
(TNO). Second edition. 1997/2005. [“TNO RED ‘05”]
11. HSE Failure Rate and Event Data for use within Risk Assessments, UK, June 28, 2012
[“HSE FRED JUN‘12”]

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 27
 12. RIVM Reference Manual Bevi Risk Assessment, Version 3.2, Module C, the Netherlands
National Institute of Public Health and Environment (RIVM), July 1, 2009. [“RIVM
BEVI ‘09”]

3.2 Additional Primary Reference Sources Identified

Available relevant equipment failure rate references were collected in the following manner:
 Literature Review
 Input from this project’s Technical Advisory Panel (TAP)
 Input from more than 50 Subject Matter Experts beyond the TAP (see Appendix L)

The primary additional reference sources considered in this analysis that contain equipment
failure rate data, estimates or specifications were:
1. Alber, T.G., Hunt, R.C., Fogarty, S.P., Wilson, J.R., Idaho Chemical Processing Plant
Failure Rate Database, Idaho National Engineering Laboratory, NEL-95/0422, August
1995. [“INL CHEM ’95”]
2. API, Risked-Based Inspection Methodology, Recommended Practice 581, Third Edition,
American Petroleum Institute, 2016 [“API 581 ‘16”]
3. Baker, E.G, Analysis of LNG Import Terminal Release Prevention Systems, Pacific
Northwest Laboratory, PNL-4152, 1982. [“PNL ITRP ‘82”]
4. Cadwallader, L., Vacuum Bellows, Vacuum Piping, Cryogenic Break, and Copper Joint
Failure Rate Estimates for ITER Design Use, Idaho National Engineering Laboratory,
INL/EXT-10-18973, June 2010. [“INL VJ ’10”]
5. CCPS Guidelines for Process Equipment Reliability Data with Data Tables, American
Institute of Chemical Engineers, Center for Chemical Process Safety, New York, NY,
1989. [“CCPS ‘89”]
6. CONCAWE Performance of European Cross-Country Oil Pipelines: Statistical
Summary of Reported Spillages in 2014 and Since 1971, Report No. 7/16, June 2016.
[“CONCAWE ‘16”]
7. DNV, Failure Frequency Guidance – Process Equipment Leak Frequency Data for Use
in QRA, DNV GL AS03-2013, 2013. [“DNV FFG HCRD ‘13”]
8. DNV, LEAK Software v. 3.3, 2016. [“DNV LEAK 3.3”]
9. EGIG Gas Pipeline Incidents, 9th Report of the European Gas Pipeline Incident Data
Group (1970 – 2013), Doc. Number EGIG 14.R.0402 Feb. 2015 [“EGIG ‘15”].
10. HSE Hydrocarbon Releases Database System (HCRD), 2016 [“HSE HCRD ‘16”].
11. HSE Failure frequencies for major failures of high pressure storage vessels at COMAH
sites: A comparison of data used by HSE and the Netherlands, Dec 2006. [“HSE
COMAH ‘06”]
12. IOGP Risk Assessment Data Directory, Process Release Frequencies, International
Association of Oil and Gas Producers Report No. 434 – 1, March 2010. [“IOGP 434-1”]

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 28
 13. IOGP Risk Assessment Data Directory, Storage Incident Frequencies, International
Association of Oil and Gas Producers Report No. 434 – 3, March 2010. [“IOGP 434-3”]
14. Kim, Hyo, Koh, Jae-Sun, Kim, Youngsoo, University of Seoul Department of Chemical
Engineering, and Theofanous, Theofanius, University of California at Santa Barbara
Center for Risk Studies and Safety, “Risk Assessment of Membrane Type LNG Storage
Tanks in Korea – based on Fault Tree Analysis,” Korean Journal of Chemical
Engineering, Vo. 22., No. 1, 1-8, 2005. [“KJCE ‘05”]
15. Lam, Chio, Statistical Analyses of Historical Pipeline Incident Data with Application to
the Risk Assessment of Onshore Natural Gas Transmission Pipelines, Master of Science
Thesis, The University of Western Ontario, July 2015 [“PHMSA NGT LAM ‘15]
16. Lee, S. R. “Safety comparison of LNG tank designs with fault tree analysis”,
International Gas Union 23rd World Gas Conference proceedings, Amsterdam, Holland.
2006. [“KGSC ‘06”]
17. LNE Handbook of Failure Frequencies, and Appendix, Safety Report, Flemish
Government LNE Department Environment, Nature and Energy Policy Unit Safety
Reporting Division, 2009. [“LNE ’09”]
18. Miyazaki, Sinichi, Tokyo Gas Co. Ltd. and Yamada, Yoshihisa, Tokyo Gas Engineering
Co., Ltd., “Quantitative Risk Assessment of LNG Above Ground Tanks Based on Past
Operating Records of LNG Regasification Terminals and Life Cycle Assessment”,
International Gas Union 22nd World Gas Conference proceedings, Tokyo, June 1-5,
2003. [“TGC ‘03”].
19. NFPA Standard for the Production, Storage, and Handling of Liquefied Natural Gas
(LNG), NFPA 59A, 2013 Edition. [“NFPA 59A ‘13”]
20. NFPA Standard for the Production, Storage, and Handling of Liquefied Natural Gas
(LNG), NFPA 59A, 2016 Edition. [“NFPA 59A ‘16”]
21. O’Donnell, IJ and Phillips, DW, Serco Assurance, and Winter, PW, AEA Technology,
“A New Estimate of the Likelihood of Spontaneous Catastrophic Failure of Pressurized
LPG Storage Vessels”, Hazards XXVIII, Symposium Series No. 150, IChemE, 2004.
[“SERCO AEA ‘04”]
22. OREDA Offshore and Onshore Reliability Data, 6th Edition, Volume 1-3 – Topside
Equipment, Published by the OREDA Participants, Prepared by SINTEF and NTNU, and
Distributed by DNV GL, 2015 [“OREDA ‘15”].
23. Plants, N. P. Industry-average performance for components and initiating events at US
commercial nuclear power plants, Idaho National Engineering Laboratory, NUREG/CR-
6928 (2007) [“INL NUC ‘07”]
24. PHMSA Natural Gas Transmission Pipeline Incident Data15, applied in this analysis as
described in Section 6 and as analyzed in Appendix J [“PHMSA NGT GTI ‘16”]
25. PHMSA Hazardous Liquid Pipeline Accident Data16, applied in this analysis as described
in Section 6 and as analyzed in Appendix K [“PHMSA HL GTI ‘16”]
26. SAI, LNG Terminal Risk Assessment Study for Oxnard, California, Science Applications,
Inc. SAI-75-615-LJ, 1975 [“SAI ‘75”]

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 29
 27. SIGGTO, Accident Prevention – The Use of Hoses & Hard Arm at Marine Terminals
Handling Liquefied Gas, Information Paper No. 4, 1996 [“SIGGTO IP4 ‘96”]
28. Welker, J.R., Brown, L.E., Ice, J.N., Martinsen, W.E., and West, H.H., Fire Safety
Aboard LNG Vessels, Final Report DOT-CG-42, 355A, Task #1 [“WELKER ‘76”]

More than 150 additional reference sources were considered in this analysis as summarized in
the References and the Bibliography. Some of the historical references (e.g. “SAI ‘75” and
“WELKER ‘76”) are incorporated by reference in subsequent references, but were included in
order to capture additional original source content. “SERCO AEA ‘04” was included in part
because its Table 2 provided confidence levels for pressurized storage vessels in its analysis.
Another potential source of data that could be used for input to assess future failure rates are
FERC’s incident reports and semi-annual reports and DOT PHMSA’s incident reports and
annual reports:

 LNG plants under FERC jurisdiction are required to report incidents within 24 hours for
significant non-scheduled events, and safety- and security-related incidents, including,
but not be limited to: leaks and releases of hazardous fluids that constitute an emergency
or exist for 5 minutes or more; explosions; fires; estimated property damage of $50,000
or more; and death or personal injury necessitating in-patient hospitalization. LNG plants
under FERC jurisdiction are also required to report on a semi-annual basis certain
operational activities, changes to facility design and operation, as well as abnormal
operating and maintenance including, but not be limited to: ship arrivals with quantity
and composition of imported and exported LNG; liquefied and vaporized quantities; boil
off/flash gas quantities; unloading/loading/shipping problems; potential hazardous
conditions from offsite vessels; storage tank stratification or rollover; geysering; storage
tank pressure excursions; cold spots on the storage tanks; storage tank vibrations and/or
vibrations in associated cryogenic piping; storage tank settlement; significant equipment
or instrumentation malfunctions or failures; non-scheduled maintenance or repair (and
reasons therefore); relative movement of storage tank inner vessels; hazardous fluids
releases; fires involving hazardous fluids and/or from other sources; negative pressure
(vacuum) within a storage tank; higher than predicted boil off rates; and adverse weather
conditions and the effect on the facility.
 LNG plants under US DOT PHMSA jurisdiction are required to report incidents as soon
as practicable but not more than 30 days after detection, for incidents that meet the
definition of US 49 CFR Section 191.3; information that must be reported regarding
those incidents is contained in PHMSA form F 7100.3. LNG plants under US DOT
PHMSA jurisdiction are also required to report operational activities on an annual basis
in form PHMSA F7100.3-1; this includes leaks and other events that occurred during the
reporting year.

Detailed equipment population data for the facilities that submit these reports would need to be
created in order to generate failure rate frequencies (e.g. number or length, and diameter and type
of valves or piping). Also, the calculated equipment failure rate frequencies would be based on
the reporting threshold requirements listed above, i.e. include only those incidents that are large
enough to be reportable. Refinements in the characterization of leak hole sizes and rupture
definition in future reporting requirements may also be beneficial.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 30
4. Failure Rate Reference Sources: Qualitative Analysis

Comments are provided below about individual references and data sources from which the results
are derived, and are organized by the specific equipment categories currently used in the FRT.
Additional details about the data included in this project’s database is contained in the database
file provided to PHMSA as a deliverable of this contract.

4.1 Review and Analysis Applicable to Multiple Equipment Categories

Comments re: UK “HSE FRED JUN’12”

The Failure Rate and Event Data for use within Risk Assessments (28/06/2012) “HSE FRED
JUN’12” document produced by the UK Health and Safety Executive (HSE) was one of the
documents cited in the derivation of the FRT. It provides a very good informational
understanding of the basis of its failure rate estimates, although many of its citations are
confidential documents.
As a point of comparison, Singapore requires the use of “HSE FRED JUN’12” for Fixed
Installations in its QRA Technical Guidance Manual issued August 2016.17

Comments re: Dutch “RIVM BEVI ‘09”

The Reference Manual Bevi Risk Assessments (“RIVM BEVI ‘09”) produced by the Dutch
National Institute of Public Health and the Environment (RIVM) was one of the documents cited
in the derivation of the FRT. It replaced the TNO Purple Book, and also stipulates that DNV’s
SAFETI-NL software must be used for QRA calculations in the Netherlands, as per p. 3 of its
Introduction.

Comments re: Flemish “LNE ‘09”

The Handbook Failure Frequencies 2009 (“LNE ‘09”) produced by the Flemish Flemish
Environment, Nature and Energy Department (LNE) was not cited in the 2012 derivation of the
FRT, but constitutes another failure rate reference for PHMSA’s/FERC’s consideration.

Comments re: “IOGP 434-1” and “IOGP 434-3”

The International Association of Oil & Gas Producers (now IOGP and formerly known as OGP)
publishes a number of risk assessment documents. Of these, Risk Assessment Data Directory,
Process Release Frequencies document 434-1 and Risk Assessment Data Directory, Storage
Incident Frequencies document 434-3 contain data relevant to FRT; both documents issued in
March 2010. Other IOGP documents reviewed during this project include those listed in the
Bibliography. IOGP documents were not cited in the 2012 derivation of the FRT, but constitute
another failure rate reference set for PHMSA’s/FERC’s consideration.
The basis of process piping and equipment release frequencies in IOGP 434-1 is the UK
Hydrocarbon Release Database System (HCRD), as summarized in this excerpt from p. 27:
The release frequencies for the main process equipment items presented in Section 2.0
are based on an analysis of the HSE hydrocarbon release database (HCRD) for 1992-
2006 [9], according to a methodology described in [4].

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 31
where [9] refers to the publically available UK HCRD dataset (https://www.hse.gov.uk/hcr3/)
and [4] refers to Confidential DNV Report 2004-0869 which appears to be consistent with
subsequent DNV-authored papers such as “DNV FFG HCRD ‘13” based on the methodology
described in Section 4.1.2 of “IOGP 434-1.”

In summary, the “IOGP 434-1” process equipment and piping failure rate frequencies are drawn
from the UK HCRD dataset. Where “IOGP 434-1” data was included in the database for this
project, the release data used was “All Releases” (which included “Full Releases”, “Limited
Releases” and “Zero Pressure” releases such as may occur during equipment maintenance).

While the underlying source for the process release frequencies in IOGP 434-1 is not from LNG
facilities, Page 19 of IOGP 434-1 states that:
The data presented in Section 2.0 can be used for process equipment on the topsides of
offshore installations and for onshore facilities handling hydrocarbons, and could also be
used as appropriate for subsea completions.
DNV [3] have compared failure rate data for LNG facilities with the data presented in
Section 2.0. The comparison indicates that LNG failure frequencies may be around 40%
to 65% of those given here. However, this has not been verified and the data for LNG
installations is relatively sparse. We therefore recommend use of the same frequencies
for LNG installations as given in Section 2.0. A 50% reduction could be considered as a
sensitivity but decisions on this would need to be fully addressed.

As a point of comparison, a representative of Korea Gas Corp. (KOGAS) indicated that KOGAS
primarily uses the IOGP data for its QRAs.18

Comments re: “NFPA 59A ‘16”

NFPA 59A includes “Chapter 15 Performance (Risk Assessment) Based LNG Plant Siting” in its
2013 and 2016 editions, which contain identical equipment failure rate data. This project’s
analysis include the identical “Example Component Failure Database” data contained in:
 Table 15.6.1 of NFPA 59A 2016 edition
 Table 15.6.1 of NFPA 59A 2013 edition, as revised by TIA 13-1

and for simplicity refers to this same data as “NFPA 59A ‘16”.

Comments re: “GRI LNG FRD ’81” and “AGA LNG EXP ‘79”
The “GRI LNG FRD ‘81” reference summarizes work performed by Johnson and Welker of
Applied Technology Corp. in which they analyzed 27 separate LNG base load or satellite
facilities (representing “approximately 1,626,000 hours of plant in-service time”, as per p. 5),
which built upon the prior “AGA LNG EXP ‘79” analysis of 25 LNG peak-shaving plants by
Welker of Applied Technology Corp. and Schorr of Brooklyn Union Gas Co. (representing
“more than 1.3 million manhours of operating time”, as per p. T-263).
As per pp. 6-7 of “GRI LNG FRD ‘81”, Johnson and Welker defined:
Two distinct failure types are presented: major failures and minor failures. Minor failures
are defined as those which cause (or would have caused) an unscheduled shutdown of

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 32
 equipment for a period of less than 24 hours. A major failure is defined as any failure
which results in an unscheduled shutdown for a period of greater than 24 hours. …
The overall failure rate is simply the sum of major and minor failures divided into the
total operating hours of the equipment in question. …
Safety-related failures were defined as failures which resulted either in a fire, injury, loss
of life, or a large leak of liquid or gas. To qualify as a safety-related failure, the liquid or
gas release must have been large enough to: 1) have the potential to injure plant
personnel; 2) actually have injured plant personnel; or 3) been severe enough to
propagate to another area had it not been controlled in the area in which it originated.
Welker and Schorr use similar definitions in “AGA LNG EXP ‘79”, and the following quotation
from page T-263 provide some additional context:
…Minor failures are defined as those that cause an unscheduled shutdown of operating
equipment caused by equipment damage where the shutdown period for repairs is less
than 24 hours. In almost all cases, a minor failure does not result in either a liquid spill or
gas leak. In a few cases, liquid dripping may occur, or small gas losses may occur.
Neither fires nor injury for either public or operating crews results from minor failures.
Major failures are those that result in an unscheduled shutdown of operating equipment
where the repair period is more than 24 hours. In these cases of major failure, injury
occurred to plant operating personnel. In all the equipment failures that were surveyed
and in all the data that were taken, neither public injury nor potential for public injury
was found. As in the case for minor failures, a major failure may or may not have either
gas leak or liquid spill; most do not. However, in any case where a substantial gas leak or
liquid spill occurred, the failure was judged to be a major failure.
The total failure rate is simply the sum of the major and minor failures lumped together
over the period of operation of the facility in question.
A safety related failure is one which resulted either in a fire or a large leak of liquid or
gas. In this case, the liquid or gas release must have been large enough to provide a
potential for injury, or the failure must be severe enough to propagate to another area had
it not been controlled in the area in which it existed. There were no failures found in this
survey that did in fact result in propagation of a failure from one area to another. Safety
related failures generally were small, although some did result in fires or substantial
liquid or gas leaks. Most of the failures reported in this survey could be described more
accurately as malfunctions because they caused no danger and resulted in short term
interruption of normal operations.
Most of the Mean Time Between Failure (MTBF) data reported in these two references refers to
that for Major or Minor failures (i.e. primarily operational failures) and not Safety-related
failures (i.e. primarily release failures); therefore, this MTBF data for Major or Minor failures
was not included in the project database. Where Safety-related failure MTBF data is reported in
“AGA LNG EXP ‘79” by component, it represents minimum failure rates calculated as of the
time of the analysis for all equipment other than for vaporizers. This project’s analysis included
Safety-related MTBF data for vaporizers in the project database (Section 4.3 describes the
calculation by the project team that yielded a vaporizer failure rate of 3.1 x 10-5 per hour
combined from the 1979 and 1981 analyses) but excluded all other Safety-related MTBF data
from “AGA LNG EXP ‘79” since it represents a minimum rate rather than an actual rate.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 33
To elaborate on the above point:
 The 1979 analysis of 25 LNG peak-shaving plants identified zero Safety-related failures
of: LNG (single containment) tanks; cryogenic valves; LNG pipelines and fire water
mains; gas pretreatment and liquefaction systems; and other equipment not considered in
the FRT.
 The 1981 analysis of 27 LNG base load or satellite facilities identified zero Safety-related
failures of: LNG (single containment) tanks; cryogenic valves (“One safety failure was
reported, but this failure was attributed to human error and is not included as a valve
failure.” as per p. 27); cryogenic piping; heat exchangers; vaporizers; gas pretreatment
systems; and other equipment not considered in PHMSA’s FRT. Section 4.1.13 entitled
“Spills and Leaks” on p. 34 identifies that “The major causes of spills and leaks were
electrical seal failures, overfilling of tanks, weld failures, and gasket leaks.” Section
4.1.17 entitled “Safety Related Failures” on p. 45 states that:
Safety related failures occurred infrequently. Only 2 safety-related failures were
reported as such in the survey, although there were 3 other incidents which
qualified as being safety-related. All three were crankcase explosions in
reciprocating compressors and caused equipment shutdowns of several days in
each incident. Twenty-nine leak situations were reported, but not enough
information was given in 18 of them to determine if a leak actually occurred. The
MTBF for safety-related failures ranged from 48,000 to 325,000 hours depending
on whether all potential and actual or only actual occurences were counted. Since
the reported leaks and spills all involved less than 200 gallons of LNG, a
relatively small amount and probably not enough to be considered a substantial
safety problem, the true MTBF for safety-related failures is more than likely in
the area of one failure per 325,000 hours of plant operation. It is important to
note that none of the safety related failures (either reported as such or inferred
from the information on the questionnaires) affected or would have affected the
general public.

Comments re: “PNL PSRP ’82” and “PNL ITRP ‘82”

Pacific Northwest Laboratories assessed LNG plant equipment and system release prevention
systems in several projects funded by the U.S. Department of Energy in the early 1980s. The
analyses most relevant to this PHMSA project are PNL’s release prevention analyses of Peak
Shaving and Import Terminals in report nos. PNL-4153 and PNL-4152, respectively. These
reports both published in 1982 and are referenced in the Bibliography as “PNL PSRP ’82” and
“PNL ITRP ’82”, respectively.
These two analyses assessed failure rate data for key LNG plant components, and share identical
assessments of LNG equipment failure data in the following tables, except for “Vaporizer”:
 “Table 3: Generic Failure Rates for Components of LNG Peakshaving Facilities” in
“PNL PSRP ’82”
 “Table 3.3: Generic Failure Rates for Components of LNG Import Terminals” in “PNL
ITRP ’82”
For simplicity this same data source in Table 3 or 3.3 is referred to as “PNL PSRP ’82” in the
database developed under this PHMSA project. Both reports indicate that:

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 34
 Over a dozen sources were used to obtain the failure rate information included in the
FMEAs. Only one of the sources dealt specifically with land-based LNG facilities
(Welker 1979). For most components, generic failure rate information was used. In most
instances these failure rates came from studies in the nuclear industry (USNRC 1975) and
the chemical processing industry (Anyakora 1971, Lees 1973, Kletz 1973, Kletz 1975).
In addition, some information was obtained from a study of safety on LNG ships (Welker
1976). Most of the failure rate information in this last study is generic and was obtained
from previously mentioned sources.
It was also noted that PNL appears to have used the MTBF rate for Major Failures (i.e. that
which “results in an unscheduled shutdown for a period of 24 hours”, and a failure to operate that
could be due to a number of potential causes including a leak) from the Welker and Schorr 1979
source (i.e. “AGA LNG EXP ‘79”) as equipment leak or rupture failure data in at least one of
PNL’s release prevention analyses (e.g. Table B.9 of “PNL PSRP ‘82”). For example, text on p.
4.12 of “PNL PSRP ‘82” analysis states:
A recent study (Welker 1979) indicated that submerged combustion vaporizer tube
failures occurred at a rate of approximately 1 x 10-4 per hour or 5 x 10-2 times/yr.
and Table 3 of “PNL PSRP ‘82” entitled “Generic Failure Rates for Components of LNG
Peakshaving Facilities” identifies the failure mode as rupture:
Component Failure Mode Faults/Hr Reference
Vaporizer Tube or Panel Rupture 1x10-4 (Welker 1979)
In comparison, Welker and Schorr (“AGA LNG EXP ‘79”) state in the section entitled
“Vaporizers” that “Figure 3 shows that major failures of vaporizers occurred at a mean time of
about 8,000 hours”. This equates to 1.25 x 10-4 per hour; or in other words, it appears that “PNL
PSRP ’82” apparently used “Major Failure” and not “Safety Failure” rate data as its vaporizer
tube or panel rupture data.
And in further comparison, Welker and Schorr (“AGA LNG EXP ‘79”) also state in the section
entitled “Vaporizers” that “Safety-related failures occurred at a mean time of a little over 15,000
operating hours” and “The vaporizers considered in Figure 3 are primarily submerged
combustion and direct fired vaporizers.”
Thus vaporizer rupture data based solely on Walker and Shorr’s 1979 analysis of Safety-related
incidents at peakshaving plants appears to have occurred at a mean time of about 15,000 hours
which equates to 6.7 x 10-5 per hour (not 1 x 10-4 per hour) and for at least two types of
vaporizers (not just submerged combustion). In section 4.3 below, the project team calculated an
estimated vaporizer leak or rupture data at a mean failure time of 3.1 x 10-5 per hour directly
from the 1979 and 1981 original works by Welker and Schorr (“AGA LNG EXP ‘79”) and
Johnson and Welker (“GRI LNG FRD ‘81”).
Additional comments about “PNL PSRP ‘82” are provided in the following sections, but in
summary all data identified as a rupture in Table 3 of “PNL PSRP ‘82” (and thus the
corresponding Table 3.3 in “PNL ITRP ‘82”) was incorporated in this project’s database, other
than for those equipment components not in the analysis framework and with the clarification
that vaporizer rupture data (see section 4.3) and “Pipe Section >3 inch Diameter” (see section
4.9) were excluded; more specifically, rupture data was included for: Storage Tank; Valve;
Expansion Joints; Pipe Fittings (Flanges); and Loading Arm. In addition, rupture data from the

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 35
body of the “PNL PSRP ‘82” report was included for Gross Failure of Storage Tank and failure
rate of Flexible Metal Hose as summarized in Sections 4.2 and 4.4, respectively.
The two PNL analyses “PNL PSRP ’82” and “PNL ITRP ‘82” then used Failure Mode and
Effects Analysis (FMEA) and Fault Tree Analysis (FTA) to predict the likelihood of failure
events for various LNG plant subsystems. This project’s analysis did not include any of these
FMEA or FTA results since they represent integrated system or subsystem failures (and not
individual equipment component failures).

Comment re: “LEE’S ’12”

Lee’s Loss Prevention in the Process Industries – Hazard Identification, Assessment and Control
Fourth Edition contains a significant amount of background information and data, such as in
Chapter section 12.30 and Appendices 7, 8 and 14 (including for example Table A14.7 that
contains failure rate data from the Rasmussen Report AEC WASH-1400 Reactor Safety Study,
1975). The Third Edition (2005) was cited in the 2012 derivation of the FRT.
In some cases the equipment failure mode is not specifically identified; in some cases it is clear
that all failure modes are included, but in a number of cases there is ambiguity. The database for
this project’s results included data from Chapter section 12.30 and Appendices 7, 8 and 14 that
was clearly identified as rupture, catastrophic failure, guillotine break or (in two cases) “serious
leak”, except for that data where units were undefined (e.g. length of piping “section”). Data for
“disruptive failures”, failure to operate, and undefined other “failures” were excluded since the
size of the crack, leak or other defect that prompted the incident or required repair or
replacement was not specified.

Comments re: “WELKER ‘76”

The database for this project’s results only included data that was clearly identified as “rupture”,
and excluded data marked as “External leak or rupture” or “leak” or “external leak” (since no
hole sizes were defined, or the failure mode was not clearly defined).

Comment re: “INL CHEM ’95”

The Idaho National Engineering Laboratory Idaho Chemical Processing Plant Failure Rate
Database 1995 (“INL CHEM ‘95”) was not cited in the 2012 derivation of the FRT, but
constitutes another failure rate data set from the chemical process industry for
PHMSA’s/FERC’s consideration. Appendix D contains “generic equipment failure database
data” which was “taken from a Westinghouse Savannah River Report Savannah River Site
Generic Database Development.” The Savannah River Site Generic Data Base Development
reference document was published in 1982; it considered data sources from multiple industries
but its Category 1 (“sources with actual failure data obtained from a detailed review of failure
events”) utilized the NUCLARR (Nuclear Computerized Library for Assesing Reactor
Reliability) database plus Savannah River Site Reactor data.19 The database for this project
reviewed equipment relevant to the Framework and
 included Chemical Process and Compressed Gas categories
 excluded Water, HVAC/Exhaust, Electric Power, and other categories.
Failure frequency rates were excerpted from the categories of “Heat Exchangers and
Condensors”, “Valves” and “Vessels” in Appendix D and included in the database for this

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 36
project’s results. Unfortunately failure rate data from the Idaho Chemical Process Plant itself (as
contained in Table 1) could not be included in this project’s database since leak sizes are not
characterized nor are rupture rates provided.

Comments re: “INL NUC ‘07”

The Idaho National Laboratory’s Industry-Average Performance for Components and Initiating
Events at U.S. Commercial Nuclear Power Plants 2007 summarizes industry-averaging
performance for components at U.S. commercial nuclear power plants from 1998-2002. It also
was not cited in the 2012 derivation of the FRT but was included in this analysis for
comparative purposes and in support of a Bayesian analysis. The population of components was
large, containing more than 13,000 valves, 700 Heat Exchangers and 700 Pressurized Tanks, but
obviously in much different service than in LNG, CPI, oil and gas, or other hydrocarbon service.
This project’s database included “External Leak Large” from Table 5-1 as rupture data (i.e.
consistent with a relevant historic definition of rupture20 as cited on p. 15) but excluded
“External Leak Small” since the leak size is not characterized.

Comments re: “OREDA ’15”

The Offshore and Onshore Reliability Data (OREDA) data set was not cited in the 2012
derivation of the FRT, but constitutes an additional failure rate reference and data set analysis of
process equipment in the oil and gas industry for PHMSA’s/FERC’s consideration. However its
population of (non-subsea) equipment in categories in the FRT is small, e.g.: 43 Heat
Exchangers; 55 Vessels; and 703 Valves.
Unfortunately the OREDA data set does not differentiate between leak vs. rupture failure modes,
but instead lists one category of “External Leakage”. With respect to other references, the
project team only considered as “ruptures” that data that was clearly identified as “rupture” ,
“catastrophic failure” or “Large Leak” (or related terminology). But because OREDA provides
some additional key useful data (e.g. specifically for valves) for comparison, this difference was
made in order to provide additional comparative information - - OREDA’s “Critical Failure”
mode “External Leakage” reliability data was considered as “rupture” data, even though
“External Leakage” may have been a leak rather than a rupture. For context, OREDA does
define a “Critical Failure” as “a failure which causes immediate and complete loss of an
equipment unit’s capability of providing its output”, so presumably an “External Leakage”
would have been a significant leak in order to be assigned as a “Critical Failure”.
Additional clarifications related to “OREDA ‘15” as used in this project’s database include:
 “Degraded”, “Incipient” or “Unknown” mode “External Leakage” reliability data was
excluded. This data could theoretically be applied to a leak hole category, but no hole
size data is provided or could be inferred. OREDA for example defines a “Degraded
Failure” as a “failure which is not critical, but it prevents an equipment unit from
providing its output within specifications. Such a unit would usually, but not necessarily,
be gradual or partial, and may develop into a critical failure in time.”
 Only “External Leakage” of “Process medium” was considered. “External Leakage” of
“Utility medium” was excluded.
 Only OREDA’s “Volume 1 – Topside Equipment” was considered. Reliability data in
OREDA’s “Volume 2 – Subsea Equipment” was excluded.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 37
  All reliability data used was based on calendar time (and not operational time).
 “Vessels” were considered to be Process Vessels (and not Pressure Vessels, i.e. Pressure
Storage Vessels) and were considered to be single-wall vessels.

Comments re: “API 581 ’16”

API Recommended Practice 581 Risk-Based Inspection Methodology was not cited in the
references identified in the 2012 derivation of the LNG Failure Rate Table, but it provides
another failure rate data set for process equipment in the oil and gas industry for
PHMSA’s/FERC’s consideration. All data in Table 3.1 “Suggested Component Generic Failure
Frequencies” of “API 581 ‘16” was included in the database developed for this project except
that for compressor and pipe. The pipe data was excluded because “API 581 ‘16” specifies the
failure frequency of pipe as “failures/yr” but does not define a piping or section length, and no
assumed average pipe length would necessarily be relevant to the FRT; lengths considered could
range from individual piping elbows to much longer piping sections (e.g. mill-run pipe); others
have also pointed out this issue in API 581.21 Hole size are defined in Table 4.4 as Small (6.4
mm), Medium (25 mm), and Large (102 mm).

Comments re: UK HCRD Database

The UK Hydrocarbon Release Database System (HCRD) was not specifically cited in the 2012
derivation of the LNG Failure Rate Table, but this failure rate data set contains some of the most
detailed hole leak size data available for process equipment and piping to the oil and gas
industry. It was generated from off-shore oil & gas operations and is a primary data basis used
by IOGP, DNV’s LEAK software, and apparently by RIVM in DNV’s SAFETI-NL, a
customized version of DNV’s software that is the sole software program used for performing
QRAs in the Netherlands.22 DNV also recommended using the UK HCRD data set for LNG risk
assessments in its 2014 Liqueified Natural Gas (LNG) Bunkering Study commissioned by the US
DOT Maritime Administration.23
Actual leak hole sizes are recorded by the dutyholders in the HCRD, but the largest leak hole
size recorded is “> 100 mm” (~4”).
Appendix L graphically illustrates some of the data in the HCRD in order to help PHMSA and
FERC better understand this database, and Appendix M provides cumulative probability curves
developed by GTI for PHMSA and FERC from this public database.

Comments re: DNV’s assessments of UK HCRD Reliability Data, including “DNV FFG HCRD
‘13”, and “DNV LEAK 3.3”

DNV has reviewed the HCRD data in detail and developed their commercial LEAK software
product based on their analysis. “DNV FFG HCRD ‘13” summarizes their analysis
methodology, and provides some representative predicted failure rate frequencies.

DNV notes on p. 9 of “DNV FFG HCRD ‘13” that the HCRD data does not directly support a
separate frequency for ruptures:

It is important to be aware that the leak frequency form is imposed on the data and that
this is a mathematical representation of historical data. The data itself does not directly

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 38
 support a separate frequency for ruptures. The historical data related to releases from
large hole sizes is very limited and the uncertainty related to estimation of such leaks is
therefore considerable. The additional rupture frequency Frup and the slope parameter m
are assumed to be constants, i.e. not to be dependent on equipment size, for any equip-
ment type.

DNV does however state on page 9 “The frequency of full-bore ruptures, i.e. holes with diameter
D, is: F(D) = f(D) Dm + Frup” where “D = equipment diameter (mm)”.

A review of Appendixes L and M is useful in this regard, while keeping in mind that the largest
leak hole size recorded is 100 mm (~4”). GTI concurs with DNV that the HCRD data does not
directly support a separate frequency for ruptures. For comparison purposes, however, estimates
of full-bore rupture were developed from the HRCD data consistent with the statement quoted
above from page 9 of “DNV FFG HCRD ‘13”. For example, “DNV FFG HCRD ‘13” indicates
on p. 20 that a total failure frequency rate for a > 150 mm (6”) hole to occur in a 6” (150 mm)
flange is 6.852 x 10-6 failures/year. GTI likewise used DNV’s LEAK software to calculate the
frequency for full-bore leaks of piping, valves, and other components to develop estimates of
ruptures using the “DNV LEAK 3.3” software merely for comparison purposes.
The results provided by GTI to PHMSA in the project database and this report (including the
Appendixes) that use DNV’s LEAK 3.3 software are: 1) based on the leak hole size ranges
defined in Table 2 below; and 2) conservatively use “Total” leak frequencies (i.e. all release
types including “Full pressure leaks” and “Zero pressure leaks”) from the HCRD. Some basic
information is summarized below in order to provide some context regarding the conservatism of
this basis; the reader is referred to those sources if additional information as desired.

 Pages of 9-11 of “DNV FFG HCRD ‘13” summarize the definition of these terms and
Figure 4 calculates that:
o 94% of all leaks in the HCRD are “Full pressure leaks”, which consist of 49%
“Full leak” and 51% “Limited leak” scenarios
o 6% of leaks in the HCRD were “Zero pressure leaks”. Page 10 summarizes in
part regarding “Zero pressure leaks” that “This scenario includes all leaks where
the pressure inside the leaking equipment is virtually zero (0.01 barg or less).
This may be because the equipment has a normal operating pressure of zero (e.g.
open drains), or because the equipment has been depressurized for maintenance,
but not de-inventoried.”
 A summary of IOGP’s recommendation in Section 3.3 of “IOGP 434-1” on pp. 20-21 is:
o “Full Releases” “should always be included in quantified risk assessements”.
o “Limited Releases” “should normally be included” in Coarse QRAs, and “could
be considered” for Detailed QRAs.
o Zero Pressure leaks are “typically excluded from QRA assessments”

For those equipment types for which leak rates are dependent on equipment size (i.e. process
pipe, flanges, and manual and actuated valves) in the LEAK 3.3 software, GTI calculated the
leak frequency results by using the arithmetic mean of the results calculated for diameters of 2”
increments within the FRT’s diameter ranges (e.g. the average of the results calculated for 12”D,
14”D, 16”D and 18”D piping to represent the 12” ≤ d < 20” piping category).
Comments re: Selection of Leak Hole Size Range

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 39
The calculation of frequency of leaks and ruptures from piping, valves, flanges and equipment
requires the definition of hole size categories, which will impact the calculated results, and no
industry-consensus definition of hole size categories was identified during this research. The
following basis was used in this project analysis, and is shown in comparison to
recommendations from the Singapore QRA Technical Guidance manual, three example
Quantitative Risk Assessments24,25,26 , and the definitions used by EGIG and in API RP 581.

Representative Hole Size

Hole Size Range Defined
or Rupture

Basis used in this Project for DNV LEAK software - -

Release from hole with effective diameter of:
2 mm 2 mm 1 - 3 mm
10 mm 10 mm 3 - 15 mm
25 mm 25 mm 15 - 35 mm
50 mm 50 mm 35 - 65 mm
100 mm 100 mm 65 - 135 mm
30 mm 20 - 40 mm
10% diam. of Piping 300 35 mm 25 - 45 mm
mm ≤ d < 499 mm, up to
50 mm 40 mm 25 - 55 mm
45 mm 35 - 60 mm
10% diam. of Piping 500
mm ≤ d < 1000 mm, up 50 mm 35 - 65 mm
to 50 mm
50 mm 35 - 65 mm
1/3 diam. of Average 67 mm 50 - 85 mm
Pipe
150 mm ≤ d < 299 mm 75 mm 65 - 100 mm
83 mm 70 - 110 mm
100 mm 65 - 135 mm
1/3 diam. of Average 117 mm 90 - 150 mm
Pipe
300 mm ≤ d < 499 mm 133 mm 100 - 175 mm
150 mm 120 - 200 mm
167 mm 135 - 205 mm
183 mm 150 - 215 mm
200 mm 155 - 230 mm
217 mm 160 - 260 mm
1/3 diam. of Average 233 mm 170 - 290 mm
Pipe
500 mm ≤ d < 1000 mm 250 mm 175 - 325 mm
267 mm 185 - 345 mm
283 mm 190 - 360 mm
300 mm 195 - 375 mm
317 mm 200 - 400 mm
Catastrophic Rupture
≥ Pipe Diameter
(for comparison only)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 40
 Comparative Examples:
10 mm 0 - 15 mm
QRA Technical 25 mm 16 - 49 mm
Guidance, Singapore
Government 75 mm 50 mm onwards
Catastrophic Failure/Guillotine Full Bore Release
5 mm 3 - 10 mm
QRA Example A 25 mm 10 - 50 mm
Full Bore Rupture 50 mm and larger
2 mm 1 - 2.8 mm
12mm 2.8 - 16.7 mm
QRA Example B 25 mm 16.7 - 31.1 mm
75 mm 31.1 - 100 mm
> 100 mm > 100 mm
10 mm 3 - 25 mm
50 mm 25 - 75 mm
QRA Example C 100 mm 75 - 125 mm
Line Diameter (Full Bore
125 mm - Line Diameter
Rupture)
Pinhole/crack Hole ≤ 20 mm
"EGIG 15" Hole 20 mm < Hole ≤ Pipe Diam.
Full Bore Rupture Hole > Pipe Diam.
6.4 mm 0 - 6.4 mm
25 mm 6.4 - 51 mm
"API 581 '16"
102 mm 51 - 152 mm
Rupture Min (Line Diameter, 406 mm)
Table 2: Leak Hole Size Range Used for “DNV LEAK 3.3” Software in this Research Compared to
Other Leak Hole Size Range Specifications

The specification of the hole size range is an important determinant in the calculated result
and corresponding conclusion, since of course the probability of a release depends directly on
it. Figure 3 illustrates this dependency, using as an example the HCRD data as modeled by
“DNV LEAK 3.3” software for an average pipe diameter in two of the FRT’s piping size
categories.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 41
 Sensitivity of Leak Frequency Rate for 25mm Hole to 
Definition of Hole Size Range

30"D Process Pipe

9"D Process Pipe

0.00E+00 1.00E‐06 2.00E‐06 3.00E‐06 4.00E‐06 5.00E‐06 6.00E‐06 7.00E‐06

10‐40 mm (25 ± 15 mm) 15‐35 mm (25 ± 10 mm) 20‐30 mm (25 ± 5 mm)

Figure 4: Comparison of Leak Frequency Rates as a Function of Hole Size Ranges.

Results Generated Using DNV LEAK 3.3 software.
Differing definitions of “rupture” can also create significant differences in failure frequency
results, e.g. in comparative estimates of rupture using LEAK 3.3 software. The basis defined
above that is used in this project is consistent with this quotation from p. 21 of “IOGP 434-1”:
3.4 Consquence modelling for the largest release size
Where the data tables in Section 2.0 show “>50 mm” or “>150 mm” for the largest hole
diameter range, the consequences of the release should be modelled using the size of the
actual pipe/valve/flange or the largest connection to other equipment types.

4.2 Cryogenic Atmospheric Storage Tanks

US 49 CFR Part 193.2013 recognizes specific respective sections of the 2001 and 2006 editions
of NFPA 59A. Table 2.2.3.5 of the NFPA 59A 2001 edition specifies Design Spills from LNG
containers. The FRT currently also specifies leak and rupture rates for LNG Atmospheric
Storage Tanks, which differs from the Design Spill for containers defined in NFPA 59A 2001
edition.

Comments re: “PNL PSRP ’82”

Both Table 3 on p. 3.13 of the “PNL PSRP ‘82” analysis of LNG peak shaving plants and Table
3.3 of “PNL ITRP ‘82” analysis of LNG import terminals indicates Storage Tank rupture rates of
1 x 10-9 and 1 x 10-10 failures per hour; this equates to 8.8 x 10-6 and 8.8 x 10-7 failures per year
and was included in this project’s database for single containment atmospheric storage tanks.
In addition, Table 7 on p. 4.3 of the “PNL PSRP ‘82” analysis indicates a “Gross Failure of
Storage Tank” at 1 x 10-5 per year. Whereas other release scenarios studied in this analysis
utilized FTA calculations, this particular estimate appears to based solely on this text from p. 4.8:
In their risk assessment of the proposed import terminal at Oxnard, California, SAI
assumed that LNG tank ruptures occur approximately 1 x 10-6 times per tank-year
(Science Applications, Inc. 1975). Based on data for petroleum refinery tanks and our
analysis of some operating scenarios that could lead to failure of the storage tank
(discussed below), it appears that more than 1 x 10-6 failures (maybe 1 x 10-5) per tank-
year can be expected.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 42
In comparison, the first entry of Table 7 on p. 4.3 of PNL’s report lists the “Gross Failure of
Storage Tank” as only the upper limit of 1 x 10-5.
The reference peakshaving system used by PNL for its analysis in “PNL PSRP ’82” was based
on:
 a single containment 350,000 bbl LNG tank (p. 3.3),
 with bottom penetrations -- “The outlet line from the storage tank is 12 inch in diameter
and exits through the bottom of the inner tank” (p.4.11)
 with an expansion joint - - “The reference storage tank has a single ply expansion joint
on the storage tank withdrawal line at the point where the line exits the outer tank shell.”
(p. 5.2)
 and with an internal shutoff valve - - “The reference storage tank has a flapper valve
inside the tank to close off the bottom withdrawal line. There are no other block valves
downstream until after the expansion joint.” (p. 5.2)
While modern single containment tanks typically utilize in-tank pumps without bottom
penetrations and without expansion joints in LNG tank outlet lines upstream of the pump
discharge block valve, this project’s analysis included PNL’s prediction of “Gross Failure of
Storage Tank” from “PNL PSRP ’82” for consideration and applied this data in the following
manner:
 applicable only to a single containment LNG tank.
 included both failure rate estimates i.e. 1 x 10-6 and 1 x 10-5.

Comments re: “LNE ‘09”

The Flemish Government Handbook Failure Frequencies 2009 (“LNE ‘09”) was not included in
the 2012 derivation of the LNG Failure Rate Table, but constitutes another failure rate data set
for PHMSA’s/FERC’s consideration. The tank rupture and deq=10mm leak failure rate
frequency included from Table 3 (p. 14) of “LNE ’09” was applied in this project’s database in
the following manner consistent with Figure 1 on its p. 15:
 “Tank type 4” data was applied to the FRT’s existing “Full Containment Cryogenic
Atmospheric Storage Tank” category. This was consistent with the “Full Containment”
designation in Table 13 on p. 32 of the Appendix.
 “Tank type 3” data was applied to the FRT’s existing “Double Containment Cryogenic
Atmospheric Storage Tank” category. This was consistent with the “Double
containment” designation in Table 13 on p. 32 of the Appendix.
 “Tank type 2” data was applied to the FRT’s existing “Single Containment Cryogenic
Atmospheric Storage Tank” category, since this is representative of LNG Single
Containment Tanks. This was consistent with the “Single Containment With Protective
Outer Shell” designation in Table 13 on p. 32 of the Appendix.
 “Tank type 1” data was applied to an ancillary category in the framework entitled “Single
Containment Refrigerated Atmospheric Storage Tank”, since this is not representative of
LNG tanks but instead is representative of single-shell, externally-insulated tanks such as
used to store refrigerated ammonia or LPG. This was consistent with the “Single

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 43
 Containment Without Protective Outer Shell” designation in Table 13 on p. 32 of the
Appendix. This category is provided merely for comparison and the project team is not
proposing this as a potential revision to the FRT.
It was noted that the “LNE ’09” storage tank data applies to both cryogenic/refrigerated tanks as
well as other service temperatures, due to this comment in Table 13 (p. 32) of Appendix of “LNE
’09” – “No separate frequencies are available for cryogenic tanks.”

Comments re: “HSE FRED JUN’12”

The failure rate data used in this project’s database and taken from Item #FR 1.1.2.1 “LNG
Refrigerated Vessels” on p. 16 of UK “HSE FRED JUN’12” were those specified for Tank
Volumes > 12,000 m3.

As per Note #34 on p. 16 of UK “HSE FRED JUN’12”, the failure rate for single-walled LNG
tanks are the same as those in the generic values in Item FR 1.1.2, where single wall tanks are
defined as “Single wall tanks, where there is no outer containment designed to hold the
cryogenic liquid or vapour.” (i.e. single containment, which would typically also have an outer
tank wall containing the insulation system). The failure rate data for single containment tanks
used in this project’s database was thus taken from Item #FR 1.1.2 “Refrigerated Ambient
Pressure Vessels” on page 13 and were those specified for Tank Volumes > 12,000 m3.

Note #34 on page 16 of UK “HSE FRED JUN’12” states that the failure rate for the release of
vapour from a double walled tank “should be set to zero”. Zero was approximated as 1 x 10-12 in
this project’s database.
The “HSE FRED JUN’12” data source was the only reference identified in this analysis that
specified failure rate of LNG tank roofs (i.e. release of vapour only).

Comments re: “RIVM BEVI ‘09”

The failure rates frequencies excerpted from Netherlands RIVM Reference Manual Bevi Risk
Assessments “RIVM BEVI ’09” and included in this project’s database were based on the inner
tank having no floor (or bottom) penetrations (reference Note #5 on page 40 of 130), consistent
with modern tank designs.

Comments re: “IOGP 434-3”

The Catastrophic Rupture Frequency for new single containment tanks included from IOGP’s
“Storage Incident Frequencies” (“IOGP 434-3”) in the database for this project was based on
“Primary Containment Only”, i.e. the rupture frequency for the “Secondary Containment”
(“corresponds to bund overtopping” as clarified in footnote #2 of Table 2.3 on page 5) was
excluded from this project’s database. The specified frequencies for both existing and new
single containment tanks were included in the database for this project, since FERC and PHMSA
receive applications pertaining to both existing and proposed new LNG facilities.
The Catastrophic Rupture Frequencies for “Double Containment Tanks”, “Full Containment
Tanks” and “Membrane Tank” used the “Secondary Containment” data (and not “Primary
Containment Only”) in Table 2.3 of the “IOGP 434-3” reference.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 44
Comments re: Membrane Tank Failure Rate Estimates, including “TGC ’03”
Three estimates of predicted membrane tank failure rates based on FTA analyses were identified
and included in this project’s database for PHMSA’s and FERC’s information, although
membrane tanks are not permitted in the NFPA 59A editions specified by US 49 CFR Part 193.
These reference were denoted as “TGC ‘03”, “KJCE ‘05” and “KGSC ‘06” and Section 3.2
provides the complete citations.
The 2003 predictive analysis by Miyazaki and Yamada of Tokyo Gas (“TGC ’03”) estimates
failure rates for LNG Membrane tanks and “Steel/PC Double Shell Tank”. The latter data was
applied to the FRT’s “Double Containment Cryogenic Atmospheric Storage Tank” category,
since the authors note that “There is a difference in the vapor control method between a full
containment tank and a Steel/PC double shell tank.” And, “In the Steel/PC double shell tank, …
The PC outer wall operates as a secondary container, preventing liquid spread in an emergency.”

Comments re: Comparative Information from the Industrial Gas Industry (relevant to both
Section 4.1 Cryogenic Atmospheric Storage Tanks and Cryogenic Pressure Vessel-related
content in Section 4.2)
The following comparison point was identified from the industrial gas industry for cryogenic
atmosphere storage tanks and pressure vessels commonly used for storing liquid nitrogen,
oxygen or argon at bulk production sites (i.e. air liquefaction plants) and also at end-user
customer sites. The flat-bottomed bulk cryogenic atmospheric storage tanks at air separation
production sites are akin to LNG single-containment storage tanks, but are of much smaller
volumetric capacity and are built to different code requirements.
 British Compressed Gases Association (BCGA) Technical Information Sheet 2327
entitled “BCGA Policy Regarding Internal Examination and Proof Pressure Testing of
Static Cryogenic Liquid Storage Tanks” states that as of 2012 there had been zero failures
of inner tanks in its survey database:
There are approximately 60,000 cryogenic tanks in service within Europe; some
having been in service since the 1960’s. They have accumulated a very large
number of safe operating hours without failure of the inner vessel during this
time.
Appendix 1 of BCGA Technical Information Sheet 23 summarizes the results of 72
inspections of cryogenic tanks, many of which were performed with an independent
authority present:
 17 inspections were of flat-bottomed bulk storage cryogenic tanks, which were
typically 11 m – 21 m in diameter and ranged in age from 9 – 34 years in service;
and
 3 inspections were of spherical bulk storage tanks; and
 52 inspections were of vacuum-insulated storage tanks, which were typically 1.0
m – 3.9 m in diameter
 European Industrial Gas Association (EIGA) document 119/04/E28 entitled “Periodic
Inspection of Static Cryogenic Vessels” in its section 8 “Incident Statistics” aligns with
BCGA Technical Information Sheet 23. It notes zero internal defects identified in
inspections of 60,000 vessels (apparently the same population as that in BCGA Technical

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 45
 Information Sheet #23):
EIGA member companies collate their own incident data and send it to EIGA for
circulation. There are approximately 60,000 cryogenic vessels in service within
Europe, some of which have been in service since the 1960’s. They have amassed
a large number of safe operating hours during this time. Individual member
groups of EIGA e.g. the BCGA in the UK compile their own data at the request of
the enforcing authority to ensure the continued safe operation of these vessels in
service. This includes the demolition and internal examination of a number of
vessels every year to ensure that there is no unsuspected failure mechanisms at
work. None have ever been found.
Inspections of cryogenic vessel shells that have been carried out by member
companies over a large number of years are also available.
The inspections were generally carried out during equipment modification or
maintenance when the opportunity was taken to examine the vessel shells either
fully, internal of external only, or locally (where the amount of shell able to be
examined was limited by the access available).
On a number of occasions thorough examinations have been made, in conjunction
with Inspecting Authorities, on particular vessels to confirm the industry’s view
that vessels in cryogenic service do not deteriorate. These examinations have
resulted in the Inspecting Authorities giving exemptions from periodic inspection
or test.
It is considered that the large amount of evidence produced, the safe operation of
the vessels, and that no cases have been observed of cryogenic vessels exhibiting
defects reducing the integrity and strength of the vessel shells, provides
considerable support for not requiring their periodic inspection and test during
service.
No minimum failure rates can be developed from these references since the 60,000
population includes undefined numbers and ages of large flat-bottomed atmospheric bulk
storage tanks, spherical bulk storage tanks, and vacuum-insulated cryogenic (pressure)
vessels (presumably the majority of which are the latter based on conventional use in the
industrial gas industry), but this reference does highlight that zero failures of inner tanks
were observed in this population of cryogenic liquid storage tanks (and vessels). But if,
for example the average age of the 60,000 cryogenic tanks and vessels was 20 years, then
the observed consolidated failure rate per tank or vessel would be < 8.3 x10-7
failures/year (< 1/(20x60,000)).

4.3 Process Vessels, Distillation Columns, Heat Exchangers and Condensers

The current FRT contains the category entitled “Process Vessels, Distillation Columns, Heat
Exchangers and Condensers”. This category can contain a wide range of equipment. Potential
refinements to the FRT investigated during this project included evaluating potential
subcategories.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 46
Differentiating process vessels from pressure vessels may be a potential refinement of the FRT
and was evaluated in this analysis. One example definition of a “Pressure Vessel” and “Process
Vessel” is given below:
A Pressure Vessel is a storage vessel in which the pressure is substantially more than 1
bar absolute. Vessels in which only the quantity of substance changes, must be
considered as a Pressure Vessel. A buffer vessel in a process installation can be seen as
an example of this.
In a Process Vessel, a change in the physical properties of the substance occurs, e.g.
temperature or phase. Examples of process vessels are distillation columns, filters and
vessels in which substances are mixed or separated. Vessels where only the level of
liquid changes are considered as Pressure Vessels.
The above definition is a variation of that used in “RIVM BEVI ’09” Reference Manual Module
C (ref. p. 45 of 130). The “RIVM BEVI ’09” definition considers distillation columns in a
separate category (with additional component details) from process vessels.

Overall Comments
Where data was presented in any reference as either Process or Pressure Vessels, the data was
applied to that respective category in this project’s database.
All failure rate data for Process and Pressure Vessels presented in all sources was assumed to be
single-wall vessel data unless it was identified otherwise.

Comments re: Pressure Vessels in “IOGP 434-3”

The leak frequency for pressure vessels used in the database for this project included the data for
“Storage Vessels” but not “Small Containers”, as provided on Table 2.3 on page 6 of “IOGP
434-3” reference. As summarized on page 3 of “IOGP 434-3” reference, storage vessels refers
to fixed tanks whereas “small containers” are “portable cylinders and drums less than
approximately 2 m3 capacity.”

Comments re: Pressure Vessels in “SERCO AEA ‘04”

O’Donnell, Phillips and Winter in “SERCO AEA ‘04” analyzed failure rate data of pressurized
LPG storage vessels for both small and large LPG vessels. This project’s database only
considered the results of their analysis of failure rates for large LPG vessels, which apparently
applies to vessels >6,600 kg capacity (i.e. roughly >3,000 US gallons). Failure rate data for
small vessels was not considered in this analysis.

Comments re: Pressure Vessels and Tank Containers in UK “HSE FRED JUN’12”

Failure rate data used in the database for this project’s results for Pressure Vessels included data
taken from Item #FR 1.1.3 “Pressure Vessels” on page 19 of “HSE FRED JUN’12”. These “are
derived in the Chlorine Siting Policy Colloquium and are applicable to chlorine pressure vessels
in a typical water treatment plant. Although they are not applicable to all types of pressure
vessels the values are a good starting point when trying to derive failure rates for vessels in other
applications.”

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 47
The “HSE FRED JUN’12” reference does not provide any overall estimates of cryogenic
pressure vessels, but implies that cryogenic pressure vessels may have a higher failure rate than
non-cryogenic pressure vessels; page 19 of “HSE FRED JUN’12” refers to “site-specific factors
indicating that a higher rate is appropriate (e.g. semi refrigerated vessels [cryogenic pressure
vessels]).” Several of the 40 advice notes and bibliography references cited by HSE in Item #FR
1.1.3 “Pressure Vessels” of “HSE FRED JUN’12” appear to arise from cryogenic applications,
including a confidential reference that “Estimates the failure rate of pressure vessels for LOX
storage to be in the order of 10-5 per yr.”
Tank Containers (i.e. ISO Containers) are used for various purposes in LNG plants. The only
failure rate data available for ISO containers identified in this project analysis was in the “HSE
FRED JUN’12” document. Although unlikely, it is possible that an applicant may propose to
build a LNG facility that incorporates a stationary ISO container as plant infrastructure (subject
to approval by FERC and PHMSA), so for the sake of completeness data was applied in this
project’s database from Item #FR 3.2.1 on page 60 of “HSE FRED JUN’12” but in stationary
service. The reader is referred to Item #FR 3.2.1 if they seek failure rate data that includes
dynamic movement (e.g. accidental dropping) of ISO containers.

Comments re: Cryogenic-Specific Pressure Vessels

The following excerpt from the Dutch LNG Safety Program website highlights that RIVM (via a
consultant) is currently reviewing failure rate frequencies for double wall LNG storage tanks,
with an apparent focus on vacuum-insulated tanks (i.e. pressure vessels):
The research on the failure frequencies double walled LNG storage tanks was performed
by AVIV commissioned by RIVM.
There are issues about the failure of double walled LNG storage tanks under significant
heat load from external sources, resulting in relative high failure frequencies for this
equipment. During 2015, RIVM has asked consultancy firm AVIV to execute a literature
survey to obtain relevant data for answering the questions.
During their survey, AVIV did not find statistically sufficient evidence to deviate from
the existing failure frequency numbers that have earlier been issued by RIVM. During
2016, AVIV will extend their earlier survey to a more generic basis, allowing potentially
more relevant data to be available for the statistical database.29
In addition, the following comparison points were identified from the industrial gas industry for
cryogenic pressure vessels that are used to store liquid nitrogen, oxygen or argon in vacuum-
insulated vessels at end-user customer sites:
 British Compressed Gases Association (BCGA) Technical Information Sheet 23, and
European Industrial Gas Association (EIGA) document 119/04/E, as summarized in
Section 4.2.
 The Vice President of Engineering for Chart Industries, Inc. estimated that since 1970
there have been about 1,800,000 vessel-years of worldwide experience using stationary
shop-built cryogenic pressure (storage) vessels (utilizing either perlite or multi-layer
insulation in vacuum annular space) containing all types of cryogenic liquids (LIN, LOX,
LAR, LH2, LNG, etc.) and excluding portable containers/dewars; he was unaware of any
ruptures of the inner tanks of this vessel population (before or after 1970) other than one
accident in Japan that occurred due to blatant misoperation.30 This estimate would imply

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 48
 that the rate of instantaneous failure of the inner tank of stationary shop-built cryogenic
vessels (i.e. not portable containers/dewars) is < 5.6 x 10-7 failures/vessel-year (unless the
one gross operating failure in Japan is considered).

The rupture of the liquid nitrogen vessel in Japan occurred at a food factory in 1992 and
resulted from the “shutoff valve for the safety valve had been closed manually” and the
internal pressure in the tank gradually rose over 50-80 days until it exceeded the vessel’s
design pressure.31

Chart’s representative also highlighted that a few fracture incidents of the outer shells of
cryogenic vessel have occurred (e.g. due to contact with cryogenic liquid), but that these
incidents have resulted in minor cracks or frosting of the outer shell that served as a
indicator that the vessel needed repair or replacement.

Comments re: Mounded/Underground Pressure Vessels

“RIVM BEVI ‘09” in Table 15 (p. 34) of Module C specifies the same failure frequencies for
“underground/mounded pressurized storage tanks” that are “mainly used for storage of LPG” as
it does in Table 13 for “pressurized storage tank aboveground”.

Comments re: Vaporizers in “GRI LNG FRD ’81”, “AGA LNG EXP ’79”, “AGA FP LNG ’84”
and “PNL PSRP ’82”
The “GRI LNG FRD ‘81” analysis identified on p. 50 that “No safety-related failures were
reported in this study for vaporizers, although 5 of the reported failures involved leaks or
ruptures of heat exchanger tubes, a potential safety-related failure. Two safety-related failures
were reported in the peakshaving study”. The “peakshaving study” refers to “AGA LNG EXP
‘79”.
Unfortunately the data cannot clearly be differentiated by vaporizer type, since detailed
equipment counts were not reported by type in the 1979 or 1981 reports. However, p. 14 of the
1981 report identifies that “Over 70 percent of the reported operating hours were on submerged
combustion vaporizers with the majority of the remaining operating hours on direct fired units.”
Other vaporizers reported in the 1981 report were running film and plate fin type.
On the above basis, the database for this project included vaporizer rupture and leak data
combined from the 1979 and 1981 sources at a mean failure rate of (2+5 ruptures or leaks) /
(35,000 + 188,000 vaporizer operating hours) which equates to 3.1 x 10-5 ruptures or leaks per
vaporizer operating hour. To put this rate on an annualized rate, a conventional design basis for
an LNG Peak Shaving Plant of 10 days/year of vaporizer sendout was applied (i.e. equivalent to
240 vaporizer operating hours/year of plant operation). This resulted in a mean failure rate of 7.5
x 10-3 vaporizer ruptures or leaks per plant operating year, and this failure rate was
conservatively assigned as a rupture rate to both direct fired heat exchangers and submerged
combustion heat exchangers (since the five reported leak or rupture failures were “potential
safety related failures”).

Mniszewski in “AGA FP LNG ’84” provides a “mean time between failures” of 1.8 x 104
hours/failure (i.e. 5.6 x 10-5 failures/hour) for “Vaporizers” in its Table 6 “Safety Related
Probability Data for LNG Facilities”, which cites Reference 20 (“AGA LNG EXP ‘79”) as its
basis. The failure rate of 1.8 x 104 hours/failures is consistent with the “Safety”-related MTBF

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 49
shown for vaporizers on Figure 3 of “AGA LNG EXP ‘79”. But the failure rate of 5.6 x 10-5
failures/hour was excluded from this project’s database because the rate of 3.1 x 10-5
failures/hour calculated above from the combined results of “AGA LNG EXP ‘79” and “GRI
LNG FRD ‘81” includes the associated data.
As pointed out in Section 4.1, the 1982 analyses by PNL (as reported for example in “PNL PSRP
’82” and “PNL ITRP ’82”) apparently applied MTBF for “Major Failures” as data for leaks or
ruptures of vaporizers; for this reason, the vaporizer failure rate data reported by ”PNL PSRP
’82” was excluded from this project’s database and this analysis. It was also noted that vaporizer
“Tube or Panel Rupture” data differed in these two PNL-authored 1982 reports but were both
attributed to “(Welker 1979)” (i.e. “AGA LNG EXP ‘79”), as shown in Appendixes G and H:
 cited as 1 x 10-5 faults/hour on Table 3.3 on p. 3.9 of “PNL ITRP ‘82” by Baker
 cited as 1 x 10-4 faults/hour on Table 3 on p. 3.13 of “PNL PSRP ‘82” by Pelto etal

Comments re: Heat Exchangers in “GRI LNG FRD ’81”

The “GRI LNG FRD ‘81” analysis reported zero Safety-related failures for heat exchangers (ref.
section 4.1.2 on p. 14) and thus no MTBF leak or rupture data for heat exchangers (other than
vaporizers) was derived from this reference. This same section and page does identify that “The
primary cause of major failures in heat exchangers was tube failure (leaking or cracking).”, but
the Major Failure MTBF data of 177,000 hours (i.e. 5.65 x 10-6 faults/hour) could not be
assigned a hole size to calculate a leak frequency, even if one approximated the “primary cause”
of tube failure as the only cause of this Major failure MTBF rate, and assumed that all failures
were external releases (vs. say tube-to-tubesheet failures or other leaks between fluid streams).

Comments re: Plate Heat Exchangers

The framework in this project considered a subcategory for “plate heat exchangers” because a
number of equipment failure rate references provided this data. However, it is recognized that
plate heat exchangers are not commonly used in LNG facilities in hydrocarbon or hazardous
fluid service. While plate heat exchangers may be used in LNG occasionally in glycol-to-
cooling water service, a rupture there would pose an operational problem but not a safety event.
The “plate heat exchanger” data was included merely for comparison to other heat exchanger
failure rates.

4.4 Truck Transfer – Arms and Hoses

Most generic hose failure rate data has apparently been derived from chlorine, LPG, ammonia
and other non-cryogenic service. For example, only one of the 27 references identified in “HSE
FRED JUN’12” Item FR 1.2.3 was clearly identified to be from cryogenic service (in an air
separation plant).
The FRT currently specifies failure rates for truck loading arms and hoses as “Failures per year
of operation”. Some source references for the failure rates of truck loading hoses and arms
differed in its basis, e.g.:
 “NFPA 59A ‘16” specifies failure rates as ruptures “per year”
 “HSE FRED JUN ‘12” specifies failure rate as “Failure rate per operation”

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 50
  “RIVM BEVI ‘09”, “LNE ‘09”, “PNL PSRP ’82”, “LEES ‘12”, “TNO PURPLE” and
“INL CHEM ‘95” specifies failure rates as “per hour” (“LNE ‘09” draws its specified
failure rates from RIVM BEVI, as per item 8.1 page 60 of its Appendix)
 “CCPS ‘89” specifies as “Aggregated time in service” in hours.
 “INL CHEM ‘95” reports failure rate as “failure rate/hour-foot”
To put the data on a “per year of operation” basis the following basis was used:
 500 transfer operations/year
 1 hour/transfer operation
 20 foot hose length
The typical number of operations at an LNG facility can vary widely and the application of 500
operations/year represents a typical use rate considering the spectrum of LNG facilities. For
example:
 25 operations/year (i.e. equivalent to 25 LNG trailers) in order to completely re-fill a
LNG satellite peak shaving facility that has three 90,000 gallon vacuum-insulated LNG
tanks.
 250 operations/year (i.e. equivalent to 250 LNG trailers, containing in total about
2,700,000 gallons of LNG or about 0.2 BCF natural gas equivalent), during a
summertime season in order to “top-up” of a LNG peak shaving satellite facility that has
one 1 BCF flat-bottomed tank. This would typify one of the largest satellite LNG peak
shaving facilities.
 500 – 1,000+ operations/year/rack at a very large merchant fuel production or import
facility. ENGIE’s Distrigas subsidiary in Boston is the largest LNG truck-loading
company in the U.S., and on average it loads 9,000 LNG trucks per year from four racks,
i.e. 2,350 truck loadings/rack/year.32

If for example a LNG facility hose has a maximum useful life of 8 years, then a basis of 500
operations/year equates to 4,000 uses/lifetime (i.e. desirable failure rate < 2.5 x 10-4
failures/operation), or a basis of 2,500 operations/year of use equates to 20,000 uses/lifetime (i.e.
desirable failure rate < 5 x 10-5 failures/operation). Hoses are also required to have yearly
pressure tests.

Comments re: “RIVM BEVI ‘09”

“RIVM BEVI ‘09” specifies failure rates for loading activities that “take place from a storage
tank to a transport unit (road tanker, tank wagon or ship) or from a transport unit to a storage
tank” and prescribes items that are present at LNG facilities. An excerpt from Module C Section
3.15 of “RIVM BEVI ’09” is provided below.

The “additional scenarios” listed in Table 51 “RIVM BEVI ‘09” were not considered, since LNG
is non-toxic.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 51
 Table 3: Excerpt of “RIVM BEVI ‘09” Module C Section 3.15 “Loading Activities”

Comments re: UK “HSE FRED JUN’12”

Failure rate data used in the project database for hoses included data taken from Item #FR 1.2.3
on page 40 of “HSE FRED JUN’12”, and on the basis that LNG facilities governed by PHMSA
meet the “Average” facility definition below:
Average Two pullaway prevention systems (one of which should be wheel chocks)
as well as inspection and pressure/leak tests to prevent transfer system
leaks and bursts but no effective pullaway mitigation.
The other categories of specified rates (i.e. not incorporated in this project’s database) are
defined in “HSE FRED JUN’12” as:
Basic These have one pullaway prevention systems such as wheel chocks, carry
out inspection and pressure/leak tests to prevent transfer system leaks and
bursts, but have no effective pullaway mitigation.
Multi safety
Systems Two pullaway prevention systems, and also an effective pullaway
mitigation system and inspection and pressure/leak tests to prevent
transfer system leaks and bursts.
The operational basis that HSE used to determine its estimate of hose and coupling failure rates
was not directly discernable. Note #92 on p. 40 of “HSE FRED JUN ‘12” identifies that the
derivation of the failure rate frequency is from “The work was carried out for chlorine transfer
facilities but should be applicable to similar transfer operations.” “HSE FRED JUN ‘12” also
references two internal (confidential) HSL reports as its basis (RAS/000/10 issued in 2000, and
RAS/04/03/1 issued in 2004); thus no details are directly available.
However the failure rates for “Guillotine failure” in Item FR 1.2.3 of “HSE FRED JUN ‘12”
match those for catastrophic failure on page 32 of the HSL’s public document HSL/2000/09
analysis of hose and coupling failure rates published in 200033 (“HSE HOSE ‘00”), although the
15 mm diameter hole results differ. The “HSE HOSE ‘00” analysis identifies that:
 “The reference system is a chlorine unloading facility, i.e., a chlorine user site, dedicated
to the unloading of chlorine. Transfers take place approximately every 6 to 8 weeks,…”
as per p. 4. This equates to 7.4 operations/year (= 52 weeks/year x 1 operation/7 weeks).

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 52
  The overall failure rate derived by HSE appears to be based on a Fault Tree Analysis, as
per this quotation from p. 28: “This report has described the development of a quantified
fault tree analysis of a full bore loss of containment during the transfer of chlorine from a
road tanker to storage for a hypothetical reference site. The generic rate for a
catastrophic failure obtained in the analysis is 4.9 x 10-8 per operation which is lower than
that of the 3 x 10-6 per operation currently used in RISKAT calculations. This reflects the
fact that the RISKAT figure includes all significant releases for a typical facility, where
the analysis here only considered catastrophic failure for a reference system incorporating
numerous safety systems.”
 The type of hoses are described on p. 16: “These hoses are constructed from convoluted
monel tube and are protected by monel braiding, along with armoured outer braiding.
The chlorine supply hose is 1.5 inch NB and the vent line 1 inch NB. Both are
approximately 9 feet in length.”
In summary:
 HSE’s hose failure rate data appears to be based at least in part on hoses of 1.5” and 1.0”
nominal diameter in chlorine service at ambient temperature.
 The generic rate for a catastrophic failure determined by HSE in its Fault Tree Analysis
as reported in “HSE HOSE ‘00” is 4.9 x 10-8 per operation, which is substantially lower
than those values currently specified in “HSE FRED JUN ‘12”, i.e.:
o 4 x 10-6 guillotine failure rate per operation for “Average facilities”
o 2 x 10-7 guillotine failure rate per operation for “Multi safety system facilities”
 The direct basis of operations/year is unknown for Item FR 1.2.3 of “HSE FRED JUN
‘12”, but HSE’s “reference case” appears to be 7.4 operations/year.

Comments re: “PNL PSRP ’82” relevant to Truck Transfer arms and hoses
The “PNL PSRP ‘82” analysis included a failure rate for flexible metal hoses in its narrative
which does not appear in its Table 3. The following failure rate was included in this project’s
database directly from this text in Section 4.4.6 on p. 4.18:
A 3-inch, flexible metal hose is used to connect the LNG trailer to the transport terminal.
The historic failure rate for this type of hose is 1.7 x 10-6/hour.
No reference for this failure rate of 1.7x10-6 faults/hour was identified in “PNL PSRP ‘82”.
Both “Table 3: Generic Failure Rates for Components of LNG Peakshaving Facilities” of “PNL
PSRP ’82” and “Table 3.3: Generic Failure Rates for Components of LNG Import Terminals” in
“PNL ITRP ’82” indicate a rupture rate for loading arms of 3 x 10-7 faults/hour. This failure rate
was not included in the project database for truck loading arms, because LNG trailer truck
loading arms were infrequently used in 1982 and this failure rate was presumably intended for
ship loading arms.

Comments re: “NFPA 59A ‘13” and “NFPA 59A ‘16”

The failure rate data of 3 x 10-8/year provided as an example for “Transfer equipment – rupture
of loading/unloading arm” was included for truck loading/unloading arms in the database for this
project.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 53
Comments re: “INL CHEM ‘95” Relevant to Truck Transfer hoses
The 20 foot hose length basis used to apply the failure rate data specified in “INL CHEM ‘95” to
the FRT was selected as a representative length for hoses used on transport trailers, since as
noted earlier the data sources used from “INL CHEM ‘95” data source were “Chemical Process
System” and “Compressed Gas System”.

4.5 Ship Transfer – Arms (and Hoses)

The FRT currently does not specify a rate for ship transfer loading hoses, which are being used
or being considered for transfer of LNG as maritime vessel fuel during bunkering operations.
The FRT currently specifies failure rates for ship loading arms as “Failures per year of
operation”. Some source data for the failure rates of ship loading arms differed in its basis, e.g.:
 “NFPA 59A ‘16” specifies failure rates as ruptures “per year”
 “RIVM BEVI ‘09” and “LEES ‘12” specifies failure rates as “per hour”
 “PNL PSRP ‘82” (and “PNL ITRP ‘82”) report loading arm rupture failure rate as
“faults/hour”
 “LNE ‘09” draws its specified failure rates from RIVM BEVI (version 3.0, Jan. 1, 2008),
as per item 8.1 page 60 in its Appendix
 “TNO PURPLE ‘05” reports failure rate as “Failure rate per transhipment”
 “HSE FRED JUN’12” reports failure rate as “Failure Frequencies per Transfer
Operation”
To put the data on a “per year of operation” basis the following basis was used:
 50 transhipments/year
 50 transfer operations/year
 12 hours/transfer operation
In comparison, in 2015 there were 4,057 LNG tanker voyages worldwide serving about 130
LNG export and import terminals 34,35, which equates to 62 LNG tanker transfer
operations/terminal-year, on average.
The term “transshipment” is not specifically defined in “TNO PURPLE ‘05”; the project team
understood the failure rate to be arising from cargo shipments (and not necessarily bunkering
operations) since section 3.1 of “TNO PURPLE 2005” identifies that “Loading and unloading
LOCs cover the transshipment of material from transport units to stationary installations and vice
versa” and since LNG bunkering was not occurring in 2005.
The term “transfer operation” is not specifically defined in “HSE FRED JUN’12”; the project
team understood it to be arising from cargo shipments (and not necessarily bunkering operations)
since note #150.1 on p. 71 of “HSE FRED JUN’12” identifies that “a 12-hour transfer time has
been assumed” and since LNG bunkering was not occurring in a significant manner when the
document was issued in 2012.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 54
Comments re: “RIVM BEVI ‘09” (and “LNE ‘09”)

The “additional scenarios” listed in were not considered, since LNG is non-toxic.

Comments re: UK “HSE FRED JUN’12”

Failure rate data used in the database for this project for Ship Hardarms included data taken from
Item #FR 3.3.1 on page 70 of “HSE FRED JUN’12” and on this basis: Transfer of liquefied
gases (vs. liquid cargo); and only one hard arm used (vs. 2 or 3 hard arms used).
The “HSE FRED JUN ‘12” hose rupture data was not applied to ship transfer hoses since
footnote #92 of Item FR 1.2.3 indicates that “The hose and coupling failure rates apply only to
road tanker transfers.”

Comments re: NFPA “59A ‘13” and “NFPA 59A ‘16”

The failure rate data of 3 x 10-8/year provided as an example for “Transfer equipment – rupture
of loading/unloading arm” was included for ship loading arms in this project’s database.

Comments re: “WELKER ‘76”

The median frequency rate was included in this project database for Ship Transfer loading arms
(but not Truck Transfer) since this analysis was for LNG Vessels where loading arms were
commonly employed. This reference also includes failure frequencies for “External Leak” but
specifies no hole size; these leak frequencies were not included in this project’s database due to
lack of hole size definition.

Comments re: “INL CHEM ‘95”

The “INL CHEM ‘95” hose rupture data was not applied to ship transfer hose.

Comments re: “SIGTTO IP4 ‘96”

The “SIGTTO IP4 ‘96” reference summarizes and analyzes accidents arising from failures or
hard-arms, hoses, Emergency Release Couplings (ERC) or other causes that occurred during
about 500,000 port calls over 13 years (1982-95) at marine terminals handling cargo shipments
of liquefied gases such as LNG, LPG and ammonia; of these, the LNG port calls were
approximately 3,000 per year (i.e. about 39,000 over 13 years). SIGTTO considered its analysis
comprehensive and as per p. 35 of “SIGTTO IP4 ‘96” was “virtually certain that all major
accidents for this trade have been recorded.” For this project analysis:
 SIGTTO’s accident data for ERCs was combined with its accident data for hard-arms,
because SIGTTO recommended that an ERC be fitted to all hard-arms for LNG and
refrigerated LPG (and terminals handling ships of over 30,000 m3), and other failure rate
references such as “HSE FRED JUN’12” (as per article 150.3) assume that all hardarms
handling liquefied gases have ERCs. This is noteworthy because including ERCs with
hard-arms tripled the reported failure rate frequency for hard-arms from 1 “Significant
Accident” to 3 “Significant Accidents”.
 SIGTTO’s “Significant Accidents” incident data from Table 9.3 was reviewed as rupture
data for loading arms and hoses for a comparison data point in this analysis, even
though all of the data arose from non-LNG service, and refrigerated LPG hoses can be of
rubber construction:

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 55
 o 3 significant accidents arose from hard-arms and ERCs in LPG and ammonia
service (and zero arose from hard-arms and ERCs in LNG service)
o 4 significant accidents arose from hoses in LPG and ammonia service (and zero
arose from hoses in LNG service)
 “SIGTTO IP4 ‘96” does not identify the number of terminals using hard arms vs. hoses,
but does identify on p. 36 that 18.1% of the annual port calls in its analysis were LPG
ships >30,000 m3 and LNG ships. Pages 3-5 of “SIGTTO IP4 ‘96” recommends that
“there is a strong case for using hard-arms for ships” and that “terminals handling ships
of over 30,000 m3 should use hard-arms”. If one assumed that 50% of the port calls in
SIGTTO’s analysis used hardarms vs. hoses, then:
o When considering non-LNG service accident data, these estimates being entered
into the project database merely for comparison:
 Non-LNG hardarm failure frequency = 3 accidents/(0.5*500,000 port
calls) = 1.2 x 10-5/transfer
 Non-LNG hose failure frequency = 4 accidents/(0.5*500,000 port calls) =
1.6 x 10-5/transfer
o Versus when considering only LNG service accident data:
 LNG hardarm failure frequency = < 1 accident/(39,000 port calls) = < 2.6
x 10-5/transfer
 SIGTTO’s incident data for Quick Disconnects (QCDC) was excluded, since QCDCs are
not always incorporated with a loading arm or hose. If the QCDC incident data from
Table 9.3 were to be included and assigned proportionately to hardarms and hoses, then
the frequencies would increase by roughly 50% since 3 significant accidents arose from
QCDCs in LPG and ammonia service (and zero significant accidents arose from QCDCs
in LNG service).

4.6 Piping – Rupture (and Leak) of Valve

No failure rate data was identified specifically for cryogenic valves beyond those that follow.

Comments re: “GRI LNG FRD ’81” and “AGA LNG EXP ‘79”
The 1979 analysis by Welker and Schorr of 25 peak-shaving plants (“AGA LNG EXP ‘79”)
identified no Safety-related or Major Failures of cryogenic valves in 43,600,000 hours (i.e. 4,980
years) of operation (as per p. 57 of the associated document “GRI LNG FRD ’81”). The “GRI
LNG FRD ’81” survey of 27 base load and satellite LNG plants identified on p. 27 that “One
safety-related failure was reported, but this was attributed to human error and is not included as a
valve failure” in 6,278,000 hours of operation (ref. Table 3). Thus these two key LNG-specific
equipment rate failure analyses in 1979 and 1981 identifed zero cryogenic valve ruptures.
The “AGA LNG EXP ‘79” analysis did report two minor cryogenic valve leaks on p. T-265:
There were no major failures reported for any cryogenic valve, and in more that 40
million operating hours, only two minor failures were reported. One of these was the
failure of a bonnet gasket and one was the leaking of a valve which was repaired by

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 56
 relapping the valve seat. Neither of these failures was considered to be a safety-related
failure.
This equates to an actual observed LNG valve leak rate of 2 leaks / 4,980 years = 4 x 10-4
leaks/year. This data point was not included in the database for this analysis since the size of the
bonnet gasket leak is unknown, but in Section 6.6 this data point is compared to the
recommendations for valve leak rates specified in the FRT.

Comments re: “AGA FP LNG ‘84”

The 1984 work by Mniszewski (“AGA FP LNG ’84”) provides “Safety Related Probability Data
for LNG Facilities” in its Table 6, and indicates that “Table 6 is presented for use in calculating
risks. Probabilities are calculated from Reference 20.” Mniszewski specifies a “mean time
between failures” of 4 x 107 hours/failure (i.e. 2.2 x 10-4 failures/year) for “LNG and cryogenic
valves”. But Reference 20 is “AGA LNG EXP ‘79” which defined:
 Major Failures as “any failure which results in an unscheduled shutdown for a period of
24 hours”
 Safety-related Failures as “failures which resulted either in a fire, injury, loss of life, or a
large leak of liquid or gas. To qualify as a safety-related failure, the liquid or gas release
must have been large enough to: 1) have the potential to injure plant personnel; 2)
actually have injured plant personnel; or 3) been severe enough to propagate to another
area had it not been controlled in the area which it originated.”
Mniszewski in “AGA FP LNG ’84” appears to have the specified MTBF failure data that is
associated with a minimum and not actual MTBF for a Major or Safety-related event (i.e.
calculated as of the time of the 1979 analysis) since 4 x 107 hours/failure is consistent with
Figure 6 of Welker and Schorr’s paper which denotes the minimum rate with an asterisk. Thus
this project’s database did not include LNG and cryogenic valve MTBF data from AGA FP LNG
‘84, since it apparently represents a minimum rather than an actual number associated with either
a Major or Safety-related Failure.

Comments re: “PNL PSRP ‘82” and “PNL ITRP ‘82”

Both of these 1982 analyses by PNL specify a “valve rupture” of 1 x 10-9 faults/hours and cite
“(USNRC 1975 and Welker 1979)”. This can be seen in Table 3 in Appendix G and Table 3.3 of
Appendix H.
In contrast (as summarized above), the 1979 analysis by Welker and Schorr of 25 peak-shaving
plants (“AGA LNG EXP ‘79”) identified zero Safety-related or Major Failures of cryogenic
valves in 43,600,000 hours of operation. It was unclear to the project team why PNL cited 1 x
10-9 faults/hours based on Welker’s and Schorr’s 1979 analysis.
Nevertheless, the valve rupture rate of 1 x 10-9 faults/hours from these two references was
included in this project’s database, since the USNRC 1975 source was also cited by PNL.

Comments re: “HSE FRED JUN‘12”

Item FR 1.2.1 of the UK “HSE FRED JUN‘12” specifies operational failure rates (e.g. failure to
close or operate) for valves but does not specify any rupture rates for valves.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 57
 Table 4: Excerpt of Item FR 1.2.1 Valves in “HSE FRED JUN‘12”

Item FR 1.2.1 does specify a Spray Release Frequency for valves with the “Effective length of
crack” being the ”Shaft circumference”. While the values in FR 1.2.1 are for chlorine duty they
were derived from a review that “included LPG, petrochemical, steam/water, nuclear and other
data” (but apparently no cryogenic data beyond “GRI LNG FRD ‘81”). The equivalent
hydraulic diameter for an annulus leak is dhydraulic = Doutside – Dinside. Valve stem leaks are
typically addressed by tightening the valve packing sleeve nut.
For comparative analysis, the project team applied the Spray Release Frequency of 2 x 10-4 leaks
per valve per year in Item FR 1.2.1 in this project’s database to a “Release from a hole with an
effective diameter of 2 mm”, as an estimate of the maximum gap between a valve shaft and
packing material arising from a loose packing assembly. This should be a very conservative
application of the 2 x 10-4 leaks per valve per year rate, in light of the 0.05 mm definition of
Spray Releases in items 78 and 79 on p. 32 of “HSE FRED JUN’12”:

A spray release is defined as a release where the spray from a hole is broken into droplets
small enough to not rain out, i.e. it is atomised. It could occur in fixed pipework or in a
flexible hose connection (say between a tanker and a storage vessel). Spray releases also
arise from plant such as pumps and valves, particularly around shafts and drives. In order
for a spray release to occur, two conditions are required:
 A very narrow breach in the containment boundary (< 50μm)
 A significant pressure (in excess of 1 barg)

Only crack-like holes, (i.e. with considerable length) need be considered, because point
defects of 50 μm size will have negligible flow rate. Clearly, these small breaches with
specific geometry are a small subset of the range of failures that could occur. No data is
available directly from industry on spray frequencies. Frequencies were estimated by
considering sprays as a subset of all small holes. Data for small holes in the type of plant
that might give rise to sprays were obtained from a variety of sources. The judgements
used in deriving the spray release figures were agreed in an MSDU Panel Paper of 4
February 2004, entitled ‘Spray Releases’ by P J Buckley (Confidential, not in the public
domain). The paper was presented at a panel meeting on 16 February 2004.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 58
Item FR 1.2.1 does not specify any leak rates for valves larger than this Spray Release
Frequency.

Comments re: “RIVM BEVI ‘09”

“RIVM BEVI ‘09” does not separately specify leak or rupture rates for valves, but instead
includes valves with its pipeline failure rates (ref. pp. 42-43 of Module C). “RIVM BEVI ‘09”
does specify failure to operate rates for excess flow and non-return valves (e.g. pp. 63-65 of
Module C) and other failure to operate rates (e.g. pp. 113 of Module C). It was also noted that
catastrophic rupture of a relief valve was “not deemed realistic”, on p. 104 of Module C:
The failure frequencies depend on the construction of the gas container, and are equal to
1 x 10-5 – 2 x 10-5 per annum for the failure of the seal and 1 x 10-5 – 4 x 10-5 per annum
for opening the relief valves. Catastrophic rupture is not included in the safety report
because it is not deemed realistic (< 10-8 per annum). This is not substantiated.
Page 112 of this reference states that “For a manual operated valve (excluding human failure) the
HSL reports give a failure frequency of 1 x 10-4 per operation” and references HSL FRED-
related publications in 1999 and 2000. But this 1 x 10-4 per operation rate was excluded from
this project’s database since the failure mode is unclear and because HSE’s updated document
“HSE FRED JUN’12” does not contain this information.

Comments re: “LNE ‘09”

“LNE ‘09” does specify probable failure rates for non-return (check) valves, excess flow valves,
and blocking valves on blocking systems, but all apparently refer to internal leakage or failure to
close.

Comments re: “WELKER ‘76”, “SAI ‘75” and “LEES ‘12”

Where valve rupture data was specified for chemical process or undefined service applications,
the sources are in some cases 40+ years old and their pedigree is often from the nuclear power
industry. For example:
 “SAI ‘75” used a “Valve Rupture Rate” of 1 x 10-8/hour i.e. ~1 x 10-4/year in its analysis
of an LNG Import Terminal; “SAI ‘75” does not cite a specific source for its valve
rupture data, but the 1975 WASH-1400 Reactor Safety Study (NUREG-75/104) is the
one of few process-related references in its list of 21 References.
 “WELKER ‘76” indicates a median “External leak or rupture” of 1 x 10-8/hour i.e. ~1 x
10-4/year for manual and all automated valves, and cites a draft copy of the 1974 WASH-
1400 Reactor Safety Study (NUREG-75/104) as its sole basis.
 The valve rupture data in Table A14.7 of “LEES ‘12” is data from the Rasmussen report
i.e. the 1975 WASH-1400 Reactor Safety Study.

4.7 Piping – Rupture of Expansion Joint

Comments re: Current FRT Specification
The FRT’s currently-specified expansion joint failure rate of 4 x 10-3 failures/year is greater than
the FRT’s 3 x 10-5 threshhold criteria and thus the failure of any piping segment that contains an

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 59
expansion joint must be analyzed as a SALS. In addition, PHMSA and FERC have used
“rupture” to mean a catastrophic full guillotine failure.
Operational history indicates that this failure rate appears conservative, at least for two-ply
expansion joints. The Cove Point LNG facility received LNG through its marine transfer lines
during its intermittent import operations over about 40 years, and has 100 two-ply expansion
joints in its LNG liquid transfer lines36,37. The current FRT specifies that this facility (if
proposed as new) must be designed for 0.4 full guillotine failures per year in its liquid transfer
lines, i.e. equivalent to 8 full guillotine failures of expansion joints over 20 years, or 16 full
guillotine failures of expansion joints over 40 years, if in continuous service. In comparison, no
catastrophic failures of expansion joints are believed to have occurred to date at the Cove Point
site during the intermittent operation of its marine transfer line.

Comments re: “PNL PSRP ’82” and “PNL ITRP ‘82”

An expansion joint rupture frequency was shown as 1 x 10-7 faults/hour (~1 x 10-3/year) on Table
3 on p. 3.13 of the “PNL PSRP ‘82” reference by Pelto etal (and the parallel Table 3.3 on p. 3.9
of “PNL ITRP ‘82” by Baker). Table 3 of PNL PSRP ‘82” cites two references (“Welker 1976,
SAI 1975”) for this data. Those two references are reviewed as follows.

Comments re: “WELKER ‘76”

This reference specifies a median “External leak or rupture” failure rate for an expansion joint of
2 x 10-6/hour (~2 x 10-2/year) with a Low rate of 1 x 10-8/hour and High rate of 1 x 10-5/hour, but
specifies no hole size. This frequency rate was not included in this project database due to lack
of failure mode or hole size definition (i.e. small or large leak vs. full bore rupture).

Comments re: “SAI ‘75”

This reference specifies a mean “Expansion Joint Ruptures” failure rate of 1 x 10-8/hour (~1 x 10-
4
/year), which was included in this project database. “SAI ‘75” does not cite a specific source
for its expansion joint rupture data, but the 1975 WASH-1400 Reactor Safety Study (NUREG-
75/104) is the one of few process-related references in its list of 21 References.

Comments re: “LEES ‘12”

Table A14.7 of this reference specifies a median “Leak (serious) in post-accident situation”
failure rate of 3 x 10-7/hour (~3 x 10-3/year) for expansion joints based on the Rasmussen Report
(i.e. 1975 WASH-1400 Reactor Safety Study), which was included in this project’s database.

Comments re: “PNL ITRP ‘82” and Possible Basis for Current FRT Specification
The FRT’s current specified failure rate of 4 x 10-3 per year may perhaps be based on the failure
rate shown in Table 3 for “Rupture of Main Transfer Line or Components During Transfer” of
the 1984 paper by Pelto and Baker entitled Analysis of LNG Peakshaving Facility Release
Prevention Systems (“PNL RP ’84”) cited in FERC issuance 20120301-0016. This rate may be
based on an an FTA analysis shown on Table B.3 on p. B.20 in “PNL ITRP ‘82” that a “large
release occurs” of a “LNG Transfer Line” will occur at a rate of 2.5 x 10-6 faults/hour or 4 x 10-3
faults/year based on 13 different equipment or control system failure rates within the transfer line
configuration. The database for this project’s analysis did not include the failure rate of 4 x 10-3
per year since no other source for this failure rate was identified, and it appears to represent a

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 60
FTA fault estimate of the integrated LNG Transfer Line (and not an individual expansion joint
failure).
The “PNL ITRP ‘82” reference analyzes the failure rate of a Double Ply Expansion Joint
assembly (with pressure detection system) in a FTA in its Table B.12 (see Appendix B) which
implies:
 The expansion joint rupture frequency of 1 x 10-7 faults/hour (9 x 10-4/year) represents
single ply construction
 The rupture of double ply construction expansion joint assembly (with leak detection
system) was calculated as 2 x 10-9 ruptures/year.

4.8 Piping – Failure (Rupture or Leak) of Gasket

PHMSA’s FRT currently specifies a failure rate for gaskets but does not identify a failure rate
mechanism (e.g. rupture or effective hole size). “Gaskets” is often understood by users of the
FRT to also mean “Flanges”.

Comments re: “WELKER ‘76”

This reference specifies a median rate of 3 x 10-6/hour (~3 x 10-2/year) for gasket leak, but
specifies no hole size. This frequency rate was not included in this project database due to lack
of failure mode or hole size definition (i.e. small or large leak vs. full bore rupture).

Comments re: “GRI LNG FRD ’81” and “AGA LNG EXP ‘79”
The 1979 analysis of 25 peak-shaving plants (“AGA LNG EXP ‘79) did identify (on p. T-266)
that “One major leak was caused by a flange gasket failure”, but no data is available on the total
gasket count or the quantity of the “major leak” that would allow a frequency calculation and
categorization of the associated failure rate. As also noted in Section 4.9, p. T-265 of AGA LNG
EXP ‘79” states that:
There have been no failures in any of the pipelines or in the piping insulation systems…
There were no reported failures of LNG or cryogenic piping, although in one plant,
gasketing was replaced on the cryogenic lines when the Teflon gaskets cold-flowed and
leaks occurred at the flange joints. The piping failures reported do not include small
leaks from gaskets, which occur more frequently when the system is being cooled down
during startup.
The 1981 analysis of 27 base load or satellite facilites (“GRI LNG FRD ’81”) also did not
provide any detailed analysis of incidents arising from gaskets or flanges. Section 4.1.13 entitled
“Spills and Leaks” on p. 34 identifies that “The major causes of spills and leaks were electrical
seal failures, overfilling of tanks, weld failures, and gasket leaks.”

Comments re: “LEES ‘12”

Page 615 in Chapter 12 of this reference includes:
Frequency of Gasket Failure:
Gaskets 0.6 mm thick = 3 x 10-6 failures/year

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 61
 Gaskets 3 mm thick = 5 x 10-6 failures/year
but specifies no hole size and the source of this data was unclear to the project team. This
frequency rate was excluded from this project database due to lack of failure mode or hole size
definition (i.e. small or large leak vs. full bore rupture), but is provided here as an additional
comparison point.

Comments re: UK HCRD

Flange leak incidents in the UK HCRD database are “selected from Ring Type (RTJ),
Compressed, Spiral Wound, Clamp (Grayloc or similar), or Hammer Union (Chicksan), Other or
Not Known”38. Appendixes L provides additional background information about types of
gaskets represented in this failure rate database and Appendix M provides cumulative probability
distribution for flange leak derived by GTI from the raw data in the HCRD.

Comments re: UK “HSE FRED JUN’12”

Failure rate data provided in Item #FR 1.2.4 on p. 44 of “HSE FRED JUN’12” defines failure
rates for flanges and gaskets, and the following excerpt includes the derivation.

Table 5: Excerpt of Item FR 1.2.4 Flanges and Gaskets in “HSE FRED JUN‘12”

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 62
To apply the equivalent hole size associated with failure of one segment of a gasket, the project
team considered a 3 mm gasket thickness and:

 Considered two alternate methodologies to apply “Hole size calculated as gasket distance
between two bolts and the gasket thickness”:
1. The equivalent hydraulic diameter of this area was calculated to be 6 mm (0.25”)
hole size, for all pipe sizes including the largest piping category in the current
FRT, using the conventional definition of hydraulic equivalent diameter d = 4 x
cross-sectional area of hole / wetted perimeter of hole (as also defined by UK
HSE for the HCRD in its Hydrocarbon Releases System Internet Help File
document).
2. The equivalent hole diameter of a circle of this area was calculated to be about 16
- 24 mm (about 0.6” – 1.0”) for piping of 25 – 1,000 mm (about 1” – 40”)
diameter, respectively (based on 600 lb Weld Neck flanges).
 Applied the failure rate frequency of 5 x 10-6 failure per year per joint as a 25 mm hole in
the framework for this project; this being considered a reasonable basis, especially in
comparison to the equivalent hydraulic diameter of 6 mm.
 Also observed that Note #73 on page 30 of “HSE FRED JUN’12” utilizes a slightly
different (3 x 10-6 vs. 5 x 10-6) failure rate for flanges; it identifies that for chemical
reactors “each flange should be given a failure rate of 3 x 10-6 per year with a hole size
equivalent to assuming a loss of a segment of gasket between two bolts.”

To apply the hole size associated with “Failure of a Spiral Wound Gasket”, the project team
considered a 3 mm spiral wound gasket thickness and:

 Considered two alternate methodologies to apply “Hole size calculated as gasket

thickness multiplied by pipe circumference”:
1. The equivalent hydraulic diameter of this area was calculated to be a 6 mm
(0.25”) hole size, for all pipe sizes including the largest piping category in
PHMSA’s FRT.
2. The equivalent hole diameter of a circle of this area was calculated to be
approximately: 51 mm (2”) for a 8” diameter pipe gasket; 70 mm (2.75”) for a
16” diameter pipe gasket; 85 mm (3.4”) for a 24” diameter pipe gasket; or 105
mm (4.2”) for a 36” diameter pipe gasket (based on outside diameter of piping).
 Reviewed the HCRD database and observed that the largest rupture of a spiral wound
gasket in the HCRD was 4.82 mm; in total, ten ruptures of spiral wound gaskets were
recorded from 1992 through 2015, and they ranged in size from 0.05 mm to 4.82 mm.
Appendix L contains more details.
 Applied the 1 x 10-7 failure rate frequency as a 50 mm hole in the framework for this
project; this being considered a reasonable basis, especially in comparison to the
equivalent hydraulic diameter of 6 mm and the maximum HCRD data of 4.82 mm.

The project team did not directly apply the Spray Release Frequency of 5 x 10-6 failure per
flange in this project’s database, because “HSE FRED JUN’12” does not define crack width.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 63
But for comparison, the Spray Release Frequency of 5 x 10-6 failure per flange matches the
“Failure of one segment of a gasket” if for example a flange crack had the maximum length of
150 mm (6 in) and an approximate width of 3 mm (0.12”); thus yielding an equivalent hydraulic
diameter of about 6 mm or equivalent hole diameter (of circle of this area) of about 24 mm.

The project team also noted that a recent analysis of circumferential cracks in nuclear power
plant piping utilized equivalent hydraulic diameter.39

Comments re: “RIVM BEVI ‘09”

“RIVM BEVI ‘09” does not separately specify leak or rupture rates for flanges, but instead
includes flanges its pipeline failure rates (ref. pp. 42-43 of Module C).

Comments re: “IOGP 434-1”

Table 3.1 on page 27 of “IOGP 434-1” provides release frequency modification factors for
different flange types of ANSI Ring Joints, ANSI Raised Face, Compact and Grayloc. Section
3.5.5 of “IOGP 434-1” clarifies that “the release frequency for each flange type is based on the
release frequency for flanges from HCRD data”. The frequency modification rates in Table 3.1
provide some additional comparative points to PHMSA’s FRT categories, but were not
considered in this project’s database because:
 “IOGP 434-1” cites only two studies that developed these modification factors.
 No explanation was identified in “IOGP 434-1” for the significant step-changes in
modification factors in its Table 3.1, such as increasing from 0.064 to 1.02 when
evaluating 10-50 mm vs. 50-150 mm Grayloc flanges (respectively), or why some
modification factors exceed unity.

4.9 Piping – By Diameter

Comments re: Cryogenic Piping Failure Rate Data and “GRI LNG FRD ’81”
The only directly-observed failure rate data set specifically for cryogenic piping identified in the
literature was one data point derived from the combined results of the “GRI LNG FRD ’81”
analysis and its prior companion “AGA LNG EXP ‘79” reference. The “GRI LNG FRD ’81”
analysis of 27 LNG base load or satellite LNG facilities identified two failures of cryogenic
piping, and on p. 27 states that “Both failures occurred at welded branch connections where a
smaller pipe connected to a larger pipe. These failures were attributed to poor quality welds.
One of the failures resulted in a leakage of about 1 cubic meter of LNG.”
The “GRI LNG FRD ’81” analysis built upon the 1979 (“AGA LNG EXP ‘79”) results that
analyzed 25 LNG peak shaving plants. “AGA LNG EXP ‘79” states on p. T-265 that:
There have been no failures in any of the pipelines or in the piping insulation systems…
There were no reported failures of LNG or cryogenic piping, although in one plant,
gasketing was replaced on the cryogenic lines when the Teflon gaskets cold-flowed and
leaks occurred at the flange joints. The piping failures reported do not include small
leaks from gaskets, which occur more frequently when the system is being cooled down
during startup.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 64
The “GRI LNG FRD ’81” report combined the 1981 and 1979 (“AGA LNG EXP ‘79”) results.
Page 59 of the “GRI LNG FRD ’81” summarizes that “Only 2 failures [of cryogenic piping]
were reported from both the peak-shaving and this study in over two and a half billion foot-hours
of operation.” This single data point was applied in this project’s database in the following
manner:
 Failure rate - - 2 failures in over two and a half billion foot-hours of operation results in a
MTBF of about 1,250,000,000 foot-hours (which is consistent with the Combined Studies
results shown in Figure 24 of “GRI LNG FRD ’81”) and this MTBF (equivalent to 8 x
10-10 failures/foot-hour) is used in this project’s database.
 Classified as leak not rupture - - The “GRI LNG FRD ’81” reference does not indicate
the pipe size or diameter, or provide any details about the amount of time that fluid
leaked, during the two piping failures that it identified in its report. But given the
relatively small amount of leakage identified (“1 cubic meter of LNG”), and given that
both failures occurred at a welded branch connection, it appears unlikely that complete
rupture of the larger pipe occurred in these failures. So this analysis included the 8 x 10-
10
failures/foot-hour rate as representing “Release from a hole with effective diameter of
10% diameter, up to 50mm (2 inches)” for line sizes up to 12” diameter (and not larger
based on consideration of the population of LNG facilities in this survey).
The MTBF data in the “GRI LNG FRD ’81” and “AGA LNG EXP ‘79” reports are based on
operating hours and not calendar hours. However, the cryogenic piping failure data was
converted to be equivalent to PHMSA’s current definition of “Failures per year of operation”,
i.e. the piping failure rate data of “per foot-hour” was directly converted to an annualized rate
equivalent to continuous service (disregarding maintenance and repair periods). In contrast,
factors were applied to put vaporizer heat exchanger MTBF data from these two reports on an
annualized basis, in the manner described in Section 4.3.

Comments re: “NFPA 59A ‘16”

Piping size diameter categories in Table 15.6.1 of NFPA 59A ‘13 and ‘16 editions differ from
the FRT. In addition, the NFPA 59A data specifies its applicability to “aboveground piping”.
To apply and compare the piping failure data in NFPA 59A to the FRT, the following
approximate alignment of pipeline categories was utilized:

NFPA 59A Aboveground Piping

Rupture Diameter Category Est. FRT Equivalent Category
Nominal diameter < 3 in. (75mm) Piping: d < 50mm (2-inch)
Nominal diameter from 3 in. (75mm)
up to and including 6 in. (150mm) Piping: 50mm (2-inch) ≤ d < 149mm (6-inch)
Nominal diameter > 6 in. (150mm) Piping: 150mm (6-inch) ≤ d < 299mm (12-inch)
Piping: 300mm (12-inch) ≤ d
Nominal diameter > 6 in. (150mm) < 499mm (20-inch)
Nominal diameter > 6 in. (150mm) Piping: 300mm (12-inch) ≤ d < 499mm (20-inch)
Nominal diameter > 6 in. (150mm) Piping: 500mm (20-inch) ≤ d < 1000 mm (40-inch)
Table 6: Approximation used to align NFPA 59A data and FRT

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 65
Comments re: UK “HSE FRED JUN’12”
The categories and failure rates for piping in the FRT appear to be primarily based on the
specifications in Item #FR 1.3 “Pipework” of “HSE FRED JUN’12” as shown below. The
project team utilized the same approximations of nominal diameter as currently used in the FRT,
i.e. 50 mm = 2”, 150 mm = 6”, 300 mm = 12”, 500 mm = 20”, and 1,000 mm = 40”.
 All of the pipework diameter categories in the FRT exactly match the categories for
“Pipework” in Item #FR 1.3 on p. 47 of “HSE FRED JUN’12”; and
 All of the data in the FRT exactly matches the failure rates in Item #FR 1.3 of “HSE
FRED JUN’12” with only two small differences:
o 2 x 10-6 failures/meter-year instead of 1 x 10-6 failures/meter-year for “Release from a
hole with effective diameter of 25mm (1-inch)” for Piping size “50 mm (2-inch) ≤ d <
149mm (6-inch)”.
o 2 x 10-8 failures/meter-year instead of 4 x 10-8 failures/meter-year for “Catastrophic
Rupture” for Piping size “500 mm (20-inch) ≤ d < 1000mm (40-inch).

Table 7: Excerpt from FR 1.3 “Pipework” of “HSE FRED JUN’12”

“HSE FRED JUN’12” also specifies failure rates for “Above Ground Pipelines” in Item #FR
3.1.2 (ref. p. 56) that are “applicable to general natural gas aboveground installations where no
site specific information is available” and “The above ground section of pipeline under
assessment to be entirely within a secure compound.” It defines “Rupture (>110 mm)” and
specifies rupture failure rate of 6.5 x 10-9 per meter-year for all pipe sizes. The data in Item FR
3.1.2 was excluded from this project’s database since they are only to be used where no site
specific information is available and cannot be used without contacting HSE’s Topic Specialist.

Comments re: “IOGP 434-1”

The process piping failure rate data in “IOGP 434-1” is derived from DNV’s analysis of the UK
HCRD database (ref. p. 19 of “IOGP 434-1”).

Comments re: “LNE ’09”

The “LNE ‘09” reference specifies failure frequencies for both above ground pipelines and
underground pipelines. The aboveground pipeline frequencies are specified in terms of L/D, and
were excluded from this project’s database because no representative length would necessarily
be correct for the FRT. The underground pipeline rupture frequency of 2.8 x 10-8 per meter-year
was included in this analysis.
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 66
Comments re: PHMSA DOT Natural Gas Transmission Pipeline Incident Data
DOT PHMSA maintains a database of recorded incidents on natural gas transmission and
gathering pipelines, and this data was not cited in the 2012 derivation of the FRT. LNG facilities
and their associated exclusion zones are governed by US 49 CFR Part 193, wheras the
transmission of natural gas and other gas by pipeline is governed by US 49 CFR Part 192.
Nevertheless, PHMSA’s database of recorded incidents on natural gas transmission pipelines
provides for comparison a failure rate data set for high pressure, carbon steel piping in natural
gas transmission service at ambient temperature.

A 2015 review by Lam40 (“PHMSA NGT LAM ’15”) analyzed onshore gas transmission
pipeline rupture incident data in PHMSA’s database from 2002 to 2013, and presented rupture
frequency data by categories of pipe diameter in PHMSA’s database, which are different than
those in the FRT (see Table 8). As summarized by Lam, PHMSA’s data represents
approximately 475,000 km of pipelines, of which “steel pipelines consistently account for over
99% of the total pipeline length” and about 97-98% of the steel pipelines are cathodically
protected coated pipelines.

PHMSA Natural Gas Transmission and

Gathering Failure Rate Data Diameter
Categories PHMSA LNG FRT Category
diameter < 4" Piping: 50mm (2-inch) ≤ d < 149mm (6-inch)
4" ≤ diameter < 10" Piping: 150mm (6-inch) ≤ d < 299mm (12-inch)
10" ≤ diameter < 20" Piping: 300mm (12-inch) ≤ d < 499mm (20-inch)
20" ≤ diameter < 28" Piping: 500mm (20-inch) ≤ d < 1000 mm (40-inch)
diameter ≥ 28" Piping: 500mm (20-inch) ≤ d < 1000 mm (40-inch)
Table 8: Comparison of Piping Size Categories in FRT to those in PHMSA’s Natural Gas
Transmission and Gathering Incident Database (and Lam’s analysis)

GTI separately analyzed this PHMSA incident data set in order to more accurately match this
incident data to the piping size categories in the FRT. Additional details are provided in
Appendix J and the results are shown in Table 9. In summary:
 GTI only used the pipeline rupture data from 2010 - 2015 to calculate the rupture rates
for the diameters given in Table 9, because PHMSA’s data from 2002 – 2009 does not
contain actual pipeline diameters (it only provides a range of diameters such as 0<d≤4,
4<d≤10, 10<d≤20, 20<d≤28 and 28<d).
 GTI compared the rupture rates from 2002 – 2015 and found them to be very similar to
those rupture rates from 2010 – 2015, so the 2010 – 2015 results are also quite
representative of the 2002 – 2015 time period.
 GTI’s calculation of pipeline rupture rates from PHMSA’s database was in close
agreement with Lam’s results, and also considered additional incident data as detailed in
Appendix J.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 67
 Pipeline Rupture Rates from
Pipeline Rupture Rates from
PHMSA’s Natural Gas
Diameter Piping Rupture PHMSA’s Hazardous Liquid
Transmission Pipeline
Category Rate Specified Pipeline Incident Database
Incident Database 2010-2015
(inch) in LNG FRT 2010-2015 "PHMSA HL
"PHMSA NGT GTI '16"
GTI '16" (per m-year)
(per m-year)
0≤d<2 1.00E-06 na na
2≤d<6 5.00E-07 2.78E-08 3.95E-08
6≤d<12 2.00E-07 3.23E-08 2.45E-08
12≤d<20 7.00E-08 4.20E-08 2.01E-08
20≤d<40 2.00E-08 2.96E-08 3.90E-08
Table 9: Comparison of Piping Rupture Rates in FRT vs. those in PHMSA’s NG Transmission and
Hazardous Pipeline Incident Databases (2010-2015) as calculated in Appendix J and K

Comments re: DOT PHMSA Hazardous Liquid Pipeline Incident Data

DOT PHMSA also maintains a database of incidents that occur in the distribution of hazardous
liquids such as crude oil, fuel oil, gasoline, diesel, propane, jet fuel, LPG, carbon dioxide, and
ammonia and this data was not cited in the 2012 derivation of the FRT. GTI analyzed this
database and developed overall rupture rates by the piping size categories in the FRT, to provide
additional data and as a direct comparison to the analysis of PHMSA’s natural gas transmission
incident data. Details are provided in Appendix K and the results are also shown in Table 9.

Comments re: “EGIG’15”

The European Gas Pipeline Incident Data Group (EGIG) maintains a database of recorded
incidents on natural gas transmission lines in Europe. EGIG’s data was not included in the 2012
derivation of the LNG Failure Rate Table, but it provides another failure rate data set for high
pressure, carbon steel piping in natural gas transmission service at ambient temperature for
comparison purposes. Criteria for inclusion in EGIG’s database are that:
 The incident must lead to an unintentional gas release
 The pipeline must fulfill the following conditions:
o To be made of steel
o To be onshore
o To have a Maximum Operating Pressure higher than 15 bar
o To be located outside the fences of the gas installations
EGIG contains some useful hole size data and defines:
 Pinhole/crack: the effective diameter of the hole is smaller than or equal to 2 cm
 Hole: the effective diameter of the hole is larger than 2 cm and smaller than or equal to
the diameter of the pipe
 Rupture: the effective diameter of the hole is larger than the pipeline diameter

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 68
The database for this project:
 used EGIG’s Rupture data as “Catastrophic Rupture” data.
 approximated EGIG’s Pinhole/Crack data for holes ≤ 20 mm as comparable to the FRT’s
“Release from a Hole with Effective Diameter of 25mm”
 disregarded EGIG’s Hole data since it represents a very broad hole size category (i.e. 20
mm < d ≤ pipeline diameter).
EGIG data primarily represents underground and unprotected natural gas transmission service
data “located outside the fences of the gas installations”, and a significant number of incidents
arise from external influence or ground movement. For example, 48% of incidents during 2004-
2013 occurred due to external influence or ground movement (e.g. arising from landslides,
flooding, or swelling rivers) as shown in Figure 5 below.
The project team aligned EGIG’s data to the FRT categories using the following approximation,
and applied data from EGIG’s entire reporting period of 1970-2013:

Pipeline Rupture Piping Rupture

Rate (/m-yr) Rate (/m-yr)
"EGIG '15" actual EGIG Nominal specified in
data (1970-2013) Diameter Category Est. FRT Equivalent Category current FRT
1.33E-07 diameter < 5" 50mm (2") ≤ d < 149mm (6") 5.00E-07
6.40E-08 5" ≤ diameter < 11" 150mm (6") ≤ d < 299mm (12") 2.00E-07
4.10E-08 11" ≤ diameter < 17" 300mm (12") ≤ d < 499mm (20") 7.00E-08
3.40E-08 17" ≤ diameter < 23" 300mm (12") ≤ d < 499mm (20") 7.00E-08
1.20E-08 23" ≤ diameter < 29" 500mm (20") ≤ d < 1000 mm (40") 7.00E-08
1.40E-08 29" ≤ diameter < 35" 500mm (20") ≤ d < 1000 mm (40") 7.00E-08
3.00E-09 35" ≤ diameter < 41" 500mm (20") ≤ d < 1000 mm (40") 7.00E-08
0.00E+00 41" ≤ diameter < 47" NA NA
6.00E-09 diameter ≥ 47" NA NA
Table 10: Approximation used to align EGIG data with FRT Piping Categories

The time span of EGIG’s failure rate data (when analyzed by leak size details) is a potential
consideration and potential weighting factor. Table 8 on Page 59 of “EGIG ‘15” highlights that
the frequency of primary failure incidents reported during 2004-2013 (i.e. primary failure
frequency of 0.157 per 1,000 km-yr) is about 48% of that reported during 1970-2013 (i.e.
primary failure frequency of 0.329 per 1,000 km-yr). The figure below (i.e. Figure 13 from
“EGIG ‘15” reproduced below in Figure 5) illustrates this overall reduction in incident rates.
Unfortunately “EGIG ‘15” does not provide data for primary failure frequencies by pipeline
diameter and size of leak for any time period other than 1970-2013, so the data available from
1970-2013 (i.e. Table 3 from “EGIG ‘15” reproduced below in Figure 5) was used in this
analysis so that EGIG’s Pinhole/crack and Rupture rates could be compared to those in the FRT.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 69
 Figure 5: “EGIG ’15” Figures 13, 16 and 19 and Table 3 reprinted by permission of EGIG:
“Figure 13: Primary (5-year moving average) failure frequency as a function of leak hole size” (top)
“Figure 16: Distribution of incidents (2004-2013)” (upper middle)
“Figure 19: Relation primary failure frequency, cause and size of leak (1970-2013)” (lower middle)
“Table 3: Primary failure frequency, pipeline diameter and size of leak (1970-2013)” (bottom)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 70
Figure 5 (cont.): “EGIG ’15” Figures 13, 16 and 19 and Table 3 reprinted by permission of EGIG:
“Figure 13: Primary (5-year moving average) failure frequency as a function of leak hole size” (top)
“Figure 16: Distribution of incidents (2004-2013)” (upper middle)
“Figure 19: Relation primary failure frequency, cause and size of leak (1970-2013)” (lower middle)
“Table 3: Primary failure frequency, pipeline diameter and size of leak (1970-2013)” (bottom)

Comment re: Potential Use of Discount or other Adjustment Factors to Raw Data
Various weighting, relevancy or discount factors could be applied to EGIG’s and PHMSA’s
pipeline incident rate raw data when considering its applicability to LNG plants, given the above
differences. But this project incorporated EGIG’s and PHMSA’s pipeline incident rate data in a
unadjusted, “as is” manner in this project’s database, except for the influence of the overall
weighting factors considered in sensitivity analysis as described in Section 5.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 71
Comments re: DOT PHMSA Natural Gas Distribution Incident Data
DOT PHMSA also maintains a database of recorded incidents on natural gas distribution. This
data was excluded from this analysis, since it contains a significant amount of polyethylene or
other non-metallic pipe used for the lower-pressure distribution of natural gas.

Comments re: CONCAWE European Cross-Country Oil Pipeline Incident Data (“CONCAWE
’16”)
Concawe is a division of the European Petroleum Refiners Association, whose members are 41
companies that operate petroleum refineries in Europe. Concawe has collected and provides data
on spillages from European cross-country oil pipeline from 1971-2014. Concawe data was not
included in this project’s database, since: hole size definitions include pipe diameter
dependency; hole size data (including rupture) is available for only 55% of the recorded
spillages; spill frequencies by hole size include losses due to theft or attempted theft, which was
significant duing the latest reporting period; and spill frequencies include both hot and cold
pipelines and for example “Whereas 81% of hot oil pipeline spillages are related to corrosion, the
figure is only 19% for cold pipelines, for which third party-related incidents and mechanical
failure are the most prevalent.”41

Comments re: “IOGP 434-1”

No data was included in this project’s database from the IOGP Risk Assessment Data Directory
Report No. 434-4 entitled “Riser & Pipeline Release Frequencies”, since the release frequencies
are based on the EGIG data and CONCAWE data, and this project’s database already included
the most recent data from EGIG.

Comments re: Canadian National Energy Board (NEB) Pipeline Incident Data
Canada’s National Energy Board (NEB) compiles statistics of pipeline incidents, but no NEB
data was included in this project’s analysis. Available incident data from the NEB42 does not
include leak or hole size, or classify ruptures versus leaks. NEB does summarize pipeline
ruptures on NEB-regulated pipelines43, but is “presented for historical perspective only” and
“should be used with caution.”

Comments re: Aboveground vs Underground Pipeline Failure Rate Data

A number of references that provide failure rate data for piping and pipelines do not differentiate
between above ground or below ground.
Pages 42-43 of 130 of Module C of “RIVM BEVI ’09” does provides some comparative data for
rupture rates, but implies that the most significant difference in failure rates applies to
underground piping housed in a pipe bay. The database for this project included “RIVM BEVI
‘09” data for above ground pipelines. Specified pipeline rupture frequencies per meter-year are:
 Above ground pipelines: 1 x 10-7, 3 x 10-7 or 1 x 10-6 (depending on diameter)
 Underground transport pipeline all other pipelines: 5 x 10-7 (any diameter)
 Underground transport pipeline complying with NEN 3650: 1.525 x 10-7 (any diameter)
 Underground transport pipeline in pipe bay: 1 x 10-9 (any diameter)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 72
Failure rates for above ground pipelines are specified in Item #FR 3.1.2 on page 56 of “HSE
FRED JUN’12” but could not be directly compared to failure rate data specified for buried
pipelines in Item #FR 3.1.1, since “HSE FRED JUN’12” directs that buried pipeline failure rate
frequencies be taken from HID C15’s (now HSE CEHMD5) MCPIPIN software.

Comments re: Stainless Piping Failure Rate Data

No failure rate data was identified for stainless steel piping specifically in cryogenic service. For
comparison, the “rupture failure mode” rate of stainless piping used in fission nuclear reactor
power plant as calculated in 2006 EPRI Report EPRI-TR-1013141 revision 1 by K. N. Fleming
and B. Lydell entitled Pipe Rupture Frequencies for Internal Flooding PRAs, and as reported in
Table 4-2 of “INL VJ ’10”, were included in this project’s database. The 2013 revision of this
analysis was not reviewed because its purchase price44 was not within this project’s budget. As
noted in “INL VJ ’10”, nuclear fission power plant feedwater piping can operate at up to ~280°C
(~536°F) and 4.8 to 6.8 MPa (700 to 990 psi) of pressure. The mean failure rupture rate for
fission reactor feedwater stainless steel pipe from 2006 EPRI Report EPRI-TR-1013141 revision
1 as reported in Table 4-2 of “INL VJ ’10” is 3.07 x 10-8 failures/foot length-reactor year (i.e.
1.01 x 10-7 failures/meter length-reactor year).
Comments re: Consideration of Piping Rupture Rate Data when Specified as “Per Section”
Piping rupture failure rate data was exluded from this project’s database when the failure was
stated on a “per section” or undefined length basis, since no appropriate length could be inferred
or assigned. This included: “Pipe Section >3 in dia” in “PNL PSRP ‘82”; “Section Rupture” for
“Pipe < 3” and “ Pipe > 3” ” in “WELKER ‘76”; “Pipe ≤ 3 in. (per section)” and “Pipe > 3 in.
(per section)” in Table A14.7 of “LEES ‘12”; Item #13 “Pipings” on p. 6.57 of “TNO RED ‘05”;
“Above ground pipeline” in “LNE ‘09”; and “Pipe” in Table 3.1 of “API 581 ‘16”.
The following references cite an identical median or nominal piping rupture rate of 1 x 10-10
faults/hour or failures/hour: “Pipe Section >3 in dia. - Rupture” in “PNL PSRP ‘82” (and also in
“PNL ITRP ‘82”); “Section Rupture” for “ Pipe > 3” ” in “WELKER ‘76”; and “Pipe > 3 in.
(per section) – Rupture/plug” in Table A14.7 of “LEES ‘12”. For comparison purposes, this
equates to 4.4 x 10-6 faults/year if one assumes a pipe section length of 5 meters or to 8.8 x 10-6
faults/year if one assumes a pipe section length of 10 meters.

4.10 Summary of Qualitative Review and Analysis

In summary, references contained in this project’s results that expand upon FERC’s 2012
citations are indicated in Table 11 and include “NFPA 59A ‘16” data, “IOGP 434-1” and “IOGP
434-3” data, “LNE ’09” data, “OREDA ‘15” data, PHMSA natural gas transmission and
hazardous liquid pipeline data, EGIG natural gas transmission data, API 581, various reliability
data sets from INL, a SIGTTO reference document, membrane tank reliability estimates in the
“KGSC’06”, “KJCE ’05” and “TGC “03” references, and two earlier source documents.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 73
 Add / Exclude
In FERC's 2012 Data
References Specifically
(Issuance Included in
Reference Abbreviation #20120301-0016) Analysis Comment
AGA FP LNG '84 Yes No - See reasons in Section 4
AGA LNG EXP '79 Yes No - See reasons in Section 4
API 581 No Yes +
CCPS '89 No Yes +
DNV FFG HCRD '13 No Yes * Used more current LEAK 3.3
DNV LEAK 3.3 No Yes +
EGIG '15 No Yes +
GRI LNG FRD '81 Yes Yes *
HSE FRED JUN'12 Yes Yes *
INL CHEM '95 No Yes +
INL NUC '07 No Yes +
INL VJ '10 No Yes +
IOGP 434-1 No Yes +
IOGP 434-3 No Yes +
KGSC '06 No Yes +
KJCE '05 No Yes +
LEES '12 Yes Yes * Reviewed 4th Edition
LNE '09 No Yes +
NFPA 59A 2001 No No - No data found
NFPA 59A 2012 DRAFT Yes No - No data found
NFPA 59A 2016 No Yes + (Data is same as 2013 edition)
OREDA '15 No Yes +
PHMSA NGT GTI '15 No Yes +
PHMSA HL GTI '15 No Yes +
PNL PSRP '82 Yes Yes *
PNL RP '84 Yes No - See reasons in Section 4
RIVM BEVI '09 Yes Yes *
SERCO AEA '04 No Yes +
TGC '03 No Yes +
TNO PURPLE ‘05 Yes Yes *
TNO GREEN ‘92 Yes No - No data found
TNO RED ‘05 Yes Yes *
Notes: "+" = not a FERC 2012 reference; "-" = was a FERC 2012 reference but no data was applied in
database for reason listed; "*" = was a FERC 2012 reference and data was applied in database.
Table 11: Comparison of Data Sources Cited by FERC and Used in This Analysis

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 74
5. Failure Rate Reference Sources: Quantitative Analysis Methodology

5.1 Overall Methodology

The project team developed and explored a framework of equipment component categories that
expanded upon the current FRT, built a database of the relevant failure rate data contained in the
references, and used a Bayesian statistical methodology combined with three different sensitivity
techniques to analyze the results.

5.1.1 Framework Layout and Component Protocols

Appendix C presents the framework layout and component protocols that were developed in this
project as a hierarchical list of the components and their potential sub-components. The
framework maintained the protocol categories currently used in the FRT, and explored potential
changes such as by:
 allowing for the subdivision of the broad category “Process Vessels, Distillation
Columns, Heat Exchangers, Condensers” into more specific components as listed below.
 exploring to see if failure rate data could be applied to both cryogenic and non-cryogenic
categories of Process Vessels, Distillation Columns, Heat Exchangers, Condensers,
Expansion Joints, Gasket Flange, Valves, Truck and Ship transfer arms, and Piping.
 gathering membrane cryogenic atmospheric storage tank data for PHMSA’s information,
as well as attempting to consider available data for cryogenic liquid nitrogen, oxygen and
argon storage.
Column C of Appendix C also identifies the data sources that were input into different
equipment categories (some of which were excluded under different sensitivity and analysis
scenarios described below).
Each equipment protocol in the Framework was assigned a row number. Data obtained from
references was assigned to each row number in the following manner, which was intended to
both help tease out any meaningful differences in the results and to help identify all cryogenic-
derived data for components such as valves or hoses:
 Data was included in non-cryogenic protocol subcategory where it appeared that the data
source was derived entirely from non-cryogenic sources (e.g. “EGIG ‘15”, “INL NUC
’07”, and “PHMSA NGT LAM ‘15”)
 Data was included in the cryogenic protocol subcategory where it appeared that the data
source was derived entirely from LNG or cryogenic references (e.g. “GRI LNG FRD
‘81”)
 Other data was included in the “top level” (i.e. neither cryogenic or non-cryogenic)
protocol subcategory.
One benefit of a Bayesian analysis is that subject matter expertise can be applied. The
judgements made by the project team can be reviewed in the Index (left hand) column of
Appendix C. Assigning the data in this manner does not affect the “roll up” results in the
Bayesian Network analysis, but it provides some additional insights into the reference data.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 75
Some “Super Categories” were also included in the Framework Layout. These “Super
Categories” were included merely to provide some higher-level “roll-up” results or other failure
rate data for comparative information or interest, and are not suggested changes to the FRT:
 Ambient Atmospheric Storage Tanks
 Refrigerated Atmospheric Storage Tanks (typically single shell)
 Cryogenic Atmospheric Storage Tanks (all types)
 Piping: All diameters
The FRT as established by FERC and PHMSA analyzes static and stationary equipment.
Rotating equipment categories (e.g. pumps and compressors) and mobile equipment (e.g. trailers
and mobile ISO containers) were considered during this analysis, but were excluded from the
framework and the FRT for this reason.

5.1.2 Database Development

Data files were developed in Microsoft Excel® software for each individual data source listed in
Table 11 and Appendix C. The relevant equipment protocol row number was then assigned to
each specific data item.

5.1.3 Bayesian Network Statistical Analysis

A Bayesian statistical analysis approach was utilized and leveraged AgenaRisk software45, a
well-proven and recognized commercial software.
Bayesian analysis is a method of statistical inference in which Bayes theorem is used to update
the probability of a hypothesis as more evidence or information becomes available.
Mathematically, Bayes theorem is represented with the equation below.

where, P (A) is the prior distribution which refers to the information known about A before B is
observed.
P (B|A) is the likelihood and refers to the observed fact B, for all values of A
P (A|B) is the posterior distribution which refers to the information known about A after
B is observed. Bayes equation can be further reduced to

To combine multiple data sources into a single database, the information known from one data
source was set as the prior and the information from the second data source was set as the
likelihood. The posterior calculated was then assigned as a prior to the next data source and that
next data source was set as the likelihood. With this repeated process, the probability
distributions of the combined dataset was derived. The final probability distribution fed into the
Bayesian network.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 76
 Figure 6: Bayesian network for failure rate calculation

Figure 6 illustrates the Bayesian network of factors used to derive failure rate using the
“Perceived Relevancy” analysis methodology. Failure rates were calculated by applying the
category weightings discussed in Section 5.2.1or 5.2.2 to the combined database (data sources).
These categories were Applicable Fluid Service, Data Type, and Primary Site Location. The
failure rate also depends on the equipment selected and nature of failure (e.g. rupture or diameter
of leak hole).

5.1.4 Custom Input/Output Script

A custom input/output script was developed in order to read data from the database files, apply
the chosen weighting factors and filters, run the AgenaRisk software, and present the data in the
form of graphs as well as tabular summaries.

5.2 Sensitivity Analysis

Three methods of sensitivity analysis were considered in this analysis.

5.2.1 Analysis Using Perceived Relevancy Factors

The data obtained from the references was evaluated using weightings of perceived relevancy
shown in Table 12. Three example illustrations were developed. Example A weighting provided
one “baseline” analysis, and Examples B and C compared results when increased emphasis was
put on generic chemical process industry references vs. from oil and gas industry references, and
vice versa.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 77
 EXAMPLE A EXAMPLE B EXAMPLE C
DESIGNATION SYMBOL WEIGHTING WEIGHTING WEIGHTING

Source Fluid Service ("SFS")

LNG Facilities LNG 1.0 1.0 1.0
Non-LNG Hydrocarbon (O&G, NGLs,
LPG, etc.) N_LNG_HC 0.8 0.5 0.8
Chemical Process Industry CPI 0.8 0.8 0.5
Combined/Unspecified UNSP 0.8 0.8 0.3
Pipeline - Natural Gas NG_PL 0.8 0.5 0.8
Pipeline - Non-Nat. Gas Hydrocarbon N_NG_PL 0.7 0.4 0.7
LIN/LOX/LAR/LH2 LAIR 0.7 0.8 0.3
Chlorine CHL 0.3 0.5 0.1
Ammonia NH3 0.5 0.7 0.3
Nuclear Industry NUC 0.3 0.2 0.1
Water WAT 0.1 0.0 0.0
Vacuum VAC 0.1 0.1 0.1

Primary Site Location ("PSL")

Off-Shore, Port or Coast OFFSH 1.0 1.0 1.0
On-Shore ONSH 1.0 1.0 1.0
Combined/Unspecified UNSP 1.0 1.0 1.0

Data Type ("DT")

Observed - Primary Source OPS 1.0 1.0 1.0
Observed - Consolidated/Derived OCD 0.9 0.9 0.9
Regulation/Specified REG 0.8 0.8 0.8
Predicted PRED 0.6 0.6 0.6

Applicable Fluid Service ("AFS")

LNG LNG 1.0 1.0 1.0
Non-LNG N_LNG 0 or 1.0 0 or 1.0 0 or 1.0
Table 12: Weighting Factors Applied to Data for “Perceived Relevancy” Analysis

The terminology utilized in this project and in the above table is defined as follows:
 “Source Fluid Service” (or “SFS”) refers to the primary fluid service from which specific
equipment failure rate source data/reference was derived (if identified), in the judgement
of the Project Team. Thus, “Source Fluid Service” does not necessarily refer the fluid
service for which the data is specified as applicable to. For example, hose failure rate
data provided in the “HSE FRED JUN’12” is derived from the chlorine industry, but
HSE FRED failure rate data is applied broadly including to LNG.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 78
  “Applicable Fluid Service” (or “AFS”) refers to whether or not the failure rate source
data is predicted, specified or regulated for LNG, in the judgement of the Project Team.
For example, the “HSE FRED JUN’12” hose failure data is derived from the chlorine
industry, but the FRED failure rate data is applied broadly including to LNG. In contrast,
the EGIG ’15 data of natural gas transmission pipeline failure rate data is not by itself
intended to be directly specified for LNG. This AFS “filter” just provides another way to
easily perform some sensitivity analysis of the data.
Table 13 summarizes the categorization of SFS and AFS to the data sets applied in the
“Perceived Relevancy” analysis:
Primary Site
Reference Source Fluid Applicable Fluid Location Data Type
Abbreviation Service ("SFS") Service ("AFS") ("PSL") ("DT")
API 581 '16 UNSP N_LNG UNSP OCD
CCPS '89 CPI LNG UNSP OCD
DNV LEAK 3.3 N_LNG_HC LNG OFFSH OPS
EGIG '15 NG_PL N_LNG ONSH OPS
GRI LNG FRD '81 LNG LNG UNSP OPS
HSE FRED JUN'12 *M LNG UNSP REG
INL CHEM '95 NUC N_LNG UNSP OCD
INL NUC '07 NUC N_LNG UNSP OCD
INL VJ '10 *M N_LNG UNSP OCD
IOGP 434-1 N_LNG_HC LNG UNSP OPS
IOGP 434-3 UNSP LNG UNSP OCD
KGSC '06 LNG LNG UNSP PRED
KJCE '05 LNG LNG UNSP PRED
LEES '12 *M LNG UNSP OCD
LNE '09 *M LNG UNSP REG
NFPA 59A '16 UNSP LNG UNSP REG
OREDA '15 N_LNG_HC N_LNG UNSP OPS
PHMSA HL GTI '16 N_NG_PL N_LNG ONSH OPS
PHMSA NGT GTI '16 NG_PL N_LNG ONSH OPS
PNL PSRP '82 UNSP LNG UNSP OCD
RIVM BEVI '09 *M LNG UNSP REG
SAI '75 UNSP LNG UNSP OCD
SERCO AEA '04 N_LNG_HC N_LNG UNSP OCD
SIGTTO IP4 '96 UNSP N_LNG OFFSH OPS
TGC '03 LNG LNG UNSP PRED
TNO PURPLE '05 UNSP LNG UNSP REG
TNO RED '05 UNSP LNG UNSP REG
WELKER '76 UNSP LNG UNSP OCD
Notes: *M = Multiple SFS classifications with this reference

Table 13: Categorization of Referenced Data for “Perceived Relevancy” Sensitivity Analysis

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 79
5.2.2 Analysis Using Example Wisdom of the Crowd
A “Wisdom of the Crowd” methodology was also analyzed, by assigning weighting factors to
individual data sets. Eight example illustrations were developed, with the intent to illustrate
potential results as one tries to reduce duplication of underlying source data and increase the
weighted importance of some of the more well-known or more commonly-used international
references/regulatory guidance documents. The examples are shown in Table 14, by moving
from Example 1 to Examples 3 and 4.

Examples 5 through 8 were included to allow PHMSA and FERC to more easily compare the
data in some of the more well-known or more commonly-used international
references/regulatory guidance documents. Example 5 is the “HSE FRED JUN ‘12” reference;
the piping failure rates currently specified in the FRT closely match those in “HSE FRED JUN
‘12”.

Example 5 Wgtg: 100%

Example 6 Wgtg: 100%

Example 7 Wgtg: 100%

Example 8 Wgtg: 100%

"IOGP 434-1 & 434-3"
"HSE FRED JUN '12"

"RIVM BEVI '09"

Example 1 Wgtg:

Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"LNE '09"
Uniform

Reference Abbreviation
API 581 '16 1.0 0.7 0.4 0.1 0.0 0.0 0.0 0.0
59A '16 1.0 0.5 0.3 0.1 0.0 0.0 0.0 0.0
CCPS '89 1.0 0.3 0.0 0.0 0.0 0.0 0.0 0.0
DNV LEAK 3.3 1.0 0.7 0.4 0.0 0.0 0.0 0.0 0.0
EGIG '15 1.0 0.5 0.1 0.0 0.0 0.0 0.0 0.0
GRI LNG FRD '81 1.0 1.0 0.7 0.5 0.0 0.0 0.0 0.0
HSE FRED JUN'12 1.0 1.0 1.0 1.0 1.0 0.0 0.0 0.0
INL CHEM '95 1.0 0.3 0.1 0.0 0.0 0.0 0.0 0.0
INL NUC '07 1.0 0.3 0.1 0.0 0.0 0.0 0.0 0.0
INL VJ '10 1.0 0.3 0.1 0.0 0.0 0.0 0.0 0.0
IOGP 434-1 1.0 0.5 0.3 0.1 0.0 0.0 0.0 1.0
IOGP 434-3 1.0 0.5 0.3 0.1 0.0 0.0 0.0 1.0
KGSC '06 1.0 0.5 0.3 0.1 0.0 0.0 0.0 0.0
KJCE '05 1.0 0.5 0.3 0.1 0.0 0.0 0.0 0.0
LEES '12 1.0 0.3 0.1 0.0 0.0 0.0 0.0 0.0
LNE '09 1.0 1.0 0.7 0.5 0.0 0.0 1.0 0.0
OREDA '15 1.0 0.3 0.1 0.0 0.0 0.0 0.0 0.0
PHMSA HL GTI '16 1.0 0.7 0.5 0.3 0.0 0.0 0.0 0.0
PHMSA NGT GTI '16 1.0 0.7 0.5 0.3 0.0 0.0 0.0 0.0
PNL PSRP '82 1.0 0.5 0.3 0.1 0.0 0.0 0.0 0.0
RIVM BEVI '09 1.0 1.0 0.7 0.5 0.0 1.0 0.0 0.0
SAI '75 1.0 0.3 0.1 0.0 0.0 0.0 0.0 0.0
SERCO AEA '04 1.0 0.5 0.3 0.1 0.0 0.0 0.0 0.0

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 80
 SIGTTO IP4 '96 1.0 0.3 0.1 0.0 0.0 0.0 0.0 0.0
TGC '03 1.0 0.5 0.3 0.1 0.0 0.0 0.0 0.0
TNO PURPLE '05 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
TNO RED '05 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
WELKER '76 1.0 0.3 0.1 0.0 0.0 0.0 0.0 0.0

Table 14: Categorization of References for Example “Wisdom of the Crowd” Analyses

5.2.3 Analysis Using Alternate Bayesian Priors

This project used Beta distribution as Bayesian priors in its analysis. The Beta distribution is a
continuous probability distribution having two parameters α and β that control the shape of the
distribution. The Beta prior describes initial knowledge about probability of equipment failures
before new evidence of failure has been incorporated and models uncertainty related to the
probability of failures of equipment. Alternate Bayesian priors such as binomial distribution and
Bernoulli distributions were considered in the analysis, but no substantive difference in results in
was observed.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 81
6. Failure Rate Reference Sources: Quantitative Analysis Results and
Recommendations

6.1 Overall Summary of Framework, Sources and Results

The project team built a framework and database, and applied data from the sources identified in
FERC’s 2012 Issuance #20120301-0016 as well as other data sources identified during this
project. The project team then used a Bayesian Network statistical analysis approach and two
different methodologies as described in Section 5 to develop this project’s results.
Appendix C illustrates the framework and component protocols used. Column C of Appendix C
shows the data sources that were applied to each of the individual equipment protocol categories.
One observation from Appendix C is that cryogenic-specific or LNG-specific data is even more
limited than may have been previously thought, based on the findings summarized in Section 4
in which the key LNG-specific equipment references from 1979 – 1984 (i.e. “AGA LNG EXP
’79”, “GRI LNG FRD ‘81”, “PNL PSRP ‘82”, and “AGA FP LNG ‘84”) provide only a few
actual leak or rupture failure rate data points generated from actual LNG or cryogenic service.
The Bayesian Network computational results were generated by applying the weighting factors
in the Perceived Relevancy and Wisdom of the Crowd example scenarios and running a large
number of iterative cycles to converge on a calculated result. Some slight differences between
the calculated results and a single dataset can sometimes occur (e.g. as in Wisdom of the Crowd
Examples 5, 6, 7 and 8) since an uninformed prior was used.
Graphical summaries of the raw data and the probably distribution were developed for each
equipment entry in the Framework. Two examples are provided below in Figures 7 and 8 for
framework category #237 – Catastrophic rupture of a piping 6” ≤ d < 12”. Credible intervals
were also developed for each framework category.

Figure 7: Summary of data in project database for Category #237 -

Catastrophic rupture of piping 6” ≤ d < 12”.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 82
 Figure 8: Calculated probability distribution function for Category #237 -
Catastrophic rupture of piping 6” ≤ d < 12” for the “Wisdom of the Crowd” Example
#4 (the FRT’s current specification of 2 x 10-7 is overlaid on right axis).

6.1.1 Results Using Perceived Relevancy Analysis

Appendixes D and E summarize in tabular form the results for estimated mean failure rates using
the “Perceived Relevancy” analysis approach for all of the data sets with the weighting factors
from Table 12 applied to the data in the references in the following manner:

Comparison Comparison Comparison Comparison

Data Set Classification 1 2 3 4
Source Fluid Service (SFS) ALL ALL LNG LNG
Primary Site Location (PSL) ALL ALL ALL ALL
Applicable Fluid Service ("AFS") ALL LNG ALL LNG
Data Type (DT) ALL ALL ALL ALL
Table 15: Comparisons of Results Provided in Appendix D and E

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 83
and also when considering all references or the exclusion of three specific references:

Appendix Appendix
D E
Included "OREDA '15", "CCPS '09", and "INL CHEM '95" Yes No
Applied Percieved Relevancy Weighting Factors Yes Yes
Table 16: Comparisons of Scenarios in Appendix D and E
A comparative review of Appendix D and E indicates that the effect of the perceived analysis
weighting factors was relatively small, and very little LNG-sourced failure rate data was
identified in this analysis. The “Perceived Relevancy” sensitivity analysis effort provided
context for the project and helped identify some data sources that were substantially different
than the bulk of the data for some equipment protocols, but its effectiveness was limited in part
because of the very small amount of available data that had high degree of relevance to LNG or
cryogenics and that could inform the generic prior data.
Additional Percieved Relevancy scenarios can also be considered and evaluated (including
refinement of the database and focusing on specific data sets) as guided by PHMSA and FERC,
if futher refinement of this analysis approach is desired in a follow-up step.

6.1.1 Results Using Example Wisdom of the Crowd Analysis

Appendix F presents the results of a “Wisdom of the Crowd” approach, where individuals can
assign their own opinions about which failure rate references should be heavily emphasized, or
not, when reviewing available references.
The examples shown are mere illustrations of what different hypothetical individuals might
consider as they reduce duplication of underlying source data, but with the intent that they put a
greater weighted importance on the more well-known or commonly used international references
or failure rate regulatory guidance documents moving from Example 1 to Examples 4-8.
Reviewing the Wisdom of the Crowd results in Appendix F (e.g. looking at Examples 3, 4 and 5)
provides another useful way to quantitatively compare the overall data sets and draw the results
and develop the recommendations that follow, in conjunction with the qualitative analysis and
conclusions summarized in Section 4.
Additional Wisdom of the Crowd scenarios can also be considered and evaluated, as guided by
PHMSA and FERC, if futher refinement of this analysis approach is desired in a follow-up step.

6.2 Cryogenic Atmospheric Storage Tanks

The results for cryogenic atmosphere tanks are summarized by tank type in each of the analysis
scenario examples. This comparison is provided only for information and potential future use,
since the failure rates of atmospheric LNG storage tanks (i.e. containers in NFPA 59A) is already
specified in Table 2.2.3.5 of NFPA 59A 2001 edition specified by 49 CFR 193. Also, NFPA
59A 2001 edition does not permit membrane storage tanks (i.e. containers in NFPA 59A). This
information may be relevant for PHMSA and FERC, if for example future regulations by those
agencies allow design spills to be determined using methodologies different from the NFPA 59A
2001 edition.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 84
Recommendations:
 PHMSA and FERC should consider eliminating cryogenic atmospheric storage tanks
from the FRT since NFPA 59A 2001 edition enforced by 49 CFR Part 193 already
specifies Design Spills for LNG tanks.
 If PHMSA and FERC retain the category of cryogenic atmospheric storage tanks in the
FRT, then PHMSA and FERC should consider clarifying the use of this information for
single accidental leakage sources for cryogenic atmospheric tanks in impounding areas
serving only vaporization, process or LNG transfer areas.

6.3 Process Vessels, Distillation Columns, Heat Exchangers and Condensers

Some of the results presented in Appendixes D, E and F show a significant difference in rupture
and leak failure rates for Pressure Vessels vs. Process Vessels. This difference appears in some
of the scenario results in part because “RIVM BEVI ‘09” and “LNE ‘09” specify a 10x
difference. However “HSE FRED JUN’12” does not make this distinction. The differences are
summarized graphically below to aid PHMSA’s and FERC’s understanding. References to
Tables in “RIVM BEVI ‘09” refer to those tables in Module C of “RIVM BEVI ‘09”.

Specified Rates of Failures/Vessel-Year: Pressure Process

"Catastrophic Failure" or "Rupture" or "Instantaneous Release" Vessel Vessel
"LNE '09" Table 1 "Pressure tanks" - "Process installations and other" 3.20E-06
"LNE '09" Table 1 "Pressure tanks" - "Storage tanks" "Underground or
1.00E-07
mounded"
"LNE '09" Table 1 "Pressure tanks" - "Storage Tanks" "Above ground (incl.
3.20E-07
road tankers and tankwagons)"
"RIVM BEVI '09" Table 31 "Reactor vessels and process vessels" 5.00E-06

"RIVM BEVI '09" Table 15 "Underground/mounded pressurized storage tanks" 5.00E-07

"RIVM BEVI '09" Table 13 "Pressurized storage tank, aboveground" 5.00E-07

"HSE FRED JUN'12" Item FR 1.1.4 "Chemical Reactors" 1.00E-05

"HSE FRED JUN'12" Item FR 3.2.1 "Tank Containers (ISO Tankers)" † 3.00E-06

"HSE FRED JUN'12" Item FR 1.1.3.2 "LPG Pressure Vessels" 2.00E-06

"HSE FRED JUN'12" Item FR 1.1.3 "Pressure Vessels" * 4.00E-06

"NFPA 59A '16" "Pressurized storage (Containers)" 5.00E-07

FRT "Process Vessels, Distillation Columns, Heat Exchangers and Condensers" 5.00E-06 5.00E-06
* Median catastrophic rate for Item FR 1.1.3 (Lower = 2.0E-06; Upper = 6.0E-06)
† with a pressure relief system, and zero lifts (i.e. stationary service)      
Table 17: Comparison of Pressure Vessel and Process Vessel Failure Catastrophic Failure, Rupture
or Instantaneous Release Rate Specifications

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 85
 Specified Failure Rates/Vessel‐Year:
Catastrophic Failure, Rupture or Instantaneous Release
FRT "Process Vessels, Distillation Columns, Heat
Exchangers and Condensers"

"NFPA 59A '16" "Pressurized storage (Containers)"

"HSE FRED JUN'12" Item FR 1.1.3 "Pressure Vessels" *

"HSE FRED JUN'12" Item FR 1.1.3.2 "LPG Pressure
Vessels"

"HSE FRED JUN'12" Item FR 3.2.1 "Tank Containers (ISO 
Tankers)" †

"HSE FRED JUN'12" Item FR 1.1.4 "Chemical Reactors"

"RIVM BEVI '09" Table 13 "Pressurized storage tank,
aboveground"

"RIVM BEVI '09" Table 15 "Underground/mounded
pressurized storage tanks"

"RIVM BEVI '09" Table 31 "Reactor vessels and process
vessels"

"LNE '09" Table 1 "Pressure tanks" ‐ "Storage Tanks"
"Above ground (incl. road tankers and tankwagons)"

"LNE '09" Table 1 "Pressure tanks" ‐ "Storage tanks"
"Underground or mounded"

"LNE '09" Table 1 "Pressure tanks" ‐ "Process
installations and other"

0.0E+00 2.0E‐06 4.0E‐06 6.0E‐06 8.0E‐06 1.0E‐05 1.2E‐05

Process Vessel Pressure Vessel

Figure 9:
Comparison of Pressure Vessel and Process Vessel Failure Catastrophic Failure, Rupture or
Instantaneous Release Rate Specifications

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 86
 Specified Rates of Failures/Vessel-Year: Pressure Process
Release from a Hole with Effective Diameter of 10 mm Vessel Vessel
"LNE '09" Table 1 "Pressure tanks" - "Process installations and other" 1.20E-04
"LNE '09" Table 1 "Pressure tanks" - "Storage tanks" "Underground or
1.20E-05
mounded"
"LNE '09" Table 1 "Pressure tanks" - "Storage Tanks" "Above ground
1.20E-05
(incl. road tankers and tankwagons)"
"RIVM BEVI '09" Table 31 "Reactor vessels and process vessels" 1.00E-04
"RIVM BEVI '09" Table 15 "Underground/mounded pressurized storage
1.00E-05
tanks"
"RIVM BEVI '09" Table 13 "Pressurized storage tank, aboveground" 1.00E-05

"HSE FRED JUN'12" Item FR 1.1.4 "Chemical Reactors" * 1.00E-05

"HSE FRED JUN'12" Item FR 3.2.1 "Tank Containers (ISO Tankers)" *† 6.00E-05

"HSE FRED JUN'12" Item FR 1.1.3.2 "LPG Pressure Vessels" * 1.00E-05

"HSE FRED JUN'12" Item FR 1.1.3 "Pressure Vessels" * 1.00E-05

FRT "Process Vessels, Distillation Columns, Heat Exchangers and
1.00E-04 1.00E-04
Condensers"
* approximating 13 mm specification as 10 mm      
† with a pressure relief system, and zero lifts (i.e. stationary service)      

Table 18:
Comparison of Pressure Vessel and Process Vessel Release from 10 mm Hole Rate Specifications

The above tables include data from Item FR 1.1.4 Chemical Reactors from “HSE FRED
JUN’12”. This data is provided only for comparative purposes and was considered by the
project team to be of limited applicability to LNG; two the three advice notes for Item FR 1.1.4
describe glass lined agitator vessels, reactors, filters and centrifuges.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 87
 Specified Failure Rates/Vessel‐Year: 
Release from Hole with Effective Diameter of 10 mm
FRT "Process Vessels, Distillation Columns, Heat
Exchangers and Condensers"

"HSE FRED JUN'12" Item FR 1.1.3 "Pressure Vessels" *

"HSE FRED JUN'12" Item FR 1.1.3.2 "LPG Pressure
Vessels" *
"HSE FRED JUN'12" Item FR 3.2.1 "Tank Containers (ISO 
Tankers)" *†

"HSE FRED JUN'12" Item FR 1.1.4 "Chemical Reactors" *

"RIVM BEVI '09" Table 13 "Pressurized storage tank,
aboveground"
"RIVM BEVI '09" Table 15 "Underground/mounded
pressurized storage tanks"
"RIVM BEVI '09" Table 31 "Reactor vessels and process
vessels"
"LNE '09" Table 1 "Pressure tanks" ‐ "Storage Tanks"
"Above ground (incl. road tankers and tankwagons)"
"LNE '09" Table 1 "Pressure tanks" ‐ "Storage tanks"
"Underground or mounded"
"LNE '09" Table 1 "Pressure tanks" ‐ "Process
installations and other"

0.0E+00 2.0E‐05 4.0E‐05 6.0E‐05 8.0E‐05 1.0E‐04 1.2E‐04 1.4E‐04

Process Vessel Pressure Vessel

Fig 10:
Comparison of Pressure Vessel and Process Vessel Release from 10 mm Hole Rate Specifications

Comments re: Rupture of Pressure Vessels and Process Vessels

An HSE analysis led by Clive Nussey in 200646 compared the estimated failure frequencies of
pressure vessels since “HSE has been criticized for using failure frequencies that are too
pessimistic when compared to values used in the Netherlands”. Nussey’s analysis cited 25
references (including the 2004 detailed analysis “SERCO AEA ‘04”) and his conclusions (ref. p.
13) include that “no convincing evidence or arguments were found to support the claim that the
HSE failure frequencies are too pessimistic”.
As noted in section 4.3, the “HSE FRED JUN’12” in its Item FR 1.1.3 included advice notes and
bibliography references that appear to arise from cryogenic applications, whereas it was unclear
to the project team if the “LNE ‘09” specification includes this type of analysis in its basis.
Pittiglio, Bragatto and Delle Site have also reviewed pressure vessel failure rates and noted the
approximate 10x difference in pressure vessel failure rates when comparing various analyses
worldwide, and concluded that “At the end API an HSE values, as quite conservative, are

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 88
basically confirmed by experimental Italian data, whilst other data are too ‘optimistic” and
inadequate for an application by Italian Authorities”47 (sic).
For the above reasons, the project team recommends that PHMSA and FERC consider retaining
its current failure frequency of 5 x 10-6 failures/year for rupture for “Process Vessels, Distillation
Columns Heat Exchangers, and Condensers” since this is more conservative and is more
consistent with the above findings than the alternative of specifying a separate, lower failure rate
frequency for rupture of pressure (storage) vessels as currently specified in “RIVM BEVI ‘09” or
“LNE ‘09”. In addition, the issue of specifying potentially lower rupture rates for Pressure
Vessels vs. Process Vessels is also moot since the current rupture frequency of 5 x 10-6
failures/year is less than the 3 x 10-5 FRT criterion threshold.
However it should be noted that a significant portion of past analysis of pressure vessel failure
rates appears to have been focused on single wall pressure vessels, such as to properly consider
and safeguard for a Boiling Liquid Expanding Vapor Explosion (BLEVE) event of an LPG
storage vessel. The anecdotal information provided by the Vice President of Engineering for
Chart Industries summarized in Section 4.3 implies that the frequency of instantaneous rupture of
shop-built double-wall cryogenic pressure vessels such as can be used to store LNG appears to
be < 5 x 10-7 failures/vessel-year. The research going on in 2016 being supported by the Dutch
LNG Safety Program and RIVM (see Section 4.3) will hopefully provide additional information
on this topic to PHMSA and FERC, including perhaps both rupture and hole leak frequencies for
double-walled cryogenic pressure vessels.
Comments re: Heat Exchangers (all types)
“HSE FRED JUN’12” does not specify failure rates for Heat Exchangers.
“LNE ’09” specifies failure rates for pipe (shell and tube) heat exchangers shown in Table 19,
and these compare well with the FRT. For example, the 1.3 x 10-5 failures/year rupture rate is
approximately one-half decade different than the FRT’s specification of 5 x 10-6 failures/year,
and both values are below the FRT’s 3 x 10-5 failures/year Criterion. The 10 mm leak failure
rates differ by less than a half decade, i.e. 6 x 10-3 failures/year versus the FRT’s specification of
1 x 10-4 failures/year for “Process Vessels, Distillation Columns Heat Exchangers, and
Condensers”; both values are greater than the FRT’s 3 x 10-5 failures/year Criterion.

Table 19: Excerpt from Section 5.1 “Pipe Heat Exchangers” of “LNE ‘09”

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 89
Comments re: Air-Cooled (Fin Fan) Heat Exchangers
Failure rate data for fin-fan heat exchangers was identified for PHMSA and FERC in this
analysis since these are common in many LNG facilities (e.g. export terminal liquefaction
trains). The data sources identified are shown in Appendix C.
“HSE FRED JUN’12”, “RIVM BEVI ‘09”, and “LNE ’09” do not specify leak rates for fin-fan
heat exchangers. One data source that contains hole size information for fin-fan heat exchangers
is the HCRD database. Our review of the HCRD data set identified that there were six leak
events in fin-fan heat exchangers in the HCRD data set, with hole sizes ranging from 0.1 to 2.6
inches. The FRT’s current specification of 1 x 10-4 per year releases from a 0.01m (0.4”) hole
“Process Vessels, Distillation Columns Heat Exchangers, and Condensers” aligns well with a
prediction by DNV from the HCRD data of 3.802 x 10-4 failures per year for a 3 – 10 mm hole
(ref. p. 22 of “DNV FFG HCRD ‘13”). Appendixes L and M provide additional details.
Comments re: Leak of “Process Vessels, Distillation Columns Heat Exchangers, and
Condensers”
The FRT currently specifies 1 x 10-4 failures/year for a “Release from a hole with effective
diameter of 0.01m (0.4 in)” for “Process Vessels, Distillation Columns Heat Exchangers, and
Condensers”. This rate is consistent with other leading regulatory documents for process vessels
such as “RIVM BEVI ‘09” and “LNE ‘09” as shown in Figure 10.
The FRT currently does not specify any leak hole sizes larger than 0.01m (and smaller than
rupture). In this manner the FRT is consistent with “RIVM BEVI ‘09”. The “LNE ‘09” does
specify 1.1 x 10-5 failures/year for a “Medium Leak 10<d<50 mm deq = 25 mm” and could be
considered in the FRT but the point may be moot since this failure frequency is less than the 3 x
10-5 FRT criterion threshold.
For PHMSA’s and FERC’s information and at their request, the project team also provided
estimates of leak rates for hole sizes of 2, 10, 25, 50 and 100 mm calculated using DNV’s LEAK
3.3 software in Appendixes D, E and F for the categories considered in this framework related to
the general category of Process Vessels, Distillation Columns, Heat Exchangers and Condensers.
Recommendation:
 PHMSA and FERC should consider retaining the existing category name and its
currently-specified failure rate. The FRT’s currently-specified rate for Pressure Vessels
aligns well with those rate specified in “HSE FRED JUN’12”, which also includes
process vessels such as distillation columns; “HSE FRED JUN’12” is silent on Heat
Exchangers. A corollary recommendation is:
o PHMSA and FERC should continue to review and consider future refinements
and potential other categories within this general category for any risk assessment
methodologies other than the current FRT methodology; this may include new
failure rate for cryogenic double-wall pressure vessels.

6.4 Truck Transfer – Arms and Hoses

The failure rate results for truck transfer hose and arm failure rate in the example Wisdom of the
Crowd results for Examples 3 and 4 indicate that the FRT’s specified failure rates for truck

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 90
transfer arms and hoses is about 10x – 100x larger than other references analyzed. However,
these results will vary greatly depending on the reference case selected, since:
 the FRT specifies a fixed rate in “failures per year of operation”; whereas
 most other references specify failure rates on a “per hour of operation” or “per operation”
frequency basis.
Disadvantages of specifying a rate in “failures per year of operation” include determining a
reasonable usage basis for setting the criteria, without creating unnecessary safety-, operational-
or feasibility-related implications for those facilities that have fewer transfer operations.
A summary of some recent developments related to LNG hose failure rates is provided below, to
provide additional context to the conclusions and recommendations.

Relevant Recent Comparison Analysis by RIVM

The following excerpt from the Dutch LNG Safety Program website highlights that in 2015
RIVM and its consultant reviewed but did not identify any statistically-sound existing data to
update LNG delivery hose failure rate frequencies in “RIVM BEVI ‘09”:
Failure of the hoses during the offloading of an LNG tanker truck to the LNG storage
vessel is perceived as one of the main contributors to the risk profile of an LNG
refuelling station. In 2015, RIVM has asked consultancy firm AVIV to execute a
literature survey to obtain relevant data for a possible decision to lower the failure
frequency of typical hoses used for this operation.
During their survey, AVIV did not find statistically sound data to update the existing
failure frequencies for LNG delivery hoses.
Starting in 2016, RIVM will research the availability of incident data for all fuel transfer
systems, to derive update failure frequencies.48

Recent Physical Testing of Cryogenic Truck Transfer Hoses

The Dutch LNG Safety Program49 and its Program Partners recently sponsored the development
of new data for cryogenic transfer hoses conducted by TNO, the Netherlands Organisation for
Applied Scientific Research. TNO with project partners recently concluded physical crush and
fatigue tests in two Phases50,51 on 2” and 4” diameter multi-composite hoses and metal
corrugated hoses. The following excerpt from page 3 of TNO report 2016 R10126 summarizes
that the purpose of this research included:
This project aims to prove that, for small scale LNG transfer systems, the full bore
rupture scenario is much too conservative for existing technology and that much less
LNG or NG outflow can be substantiated. The approach is to prove that there are credible
failure scenarios that may result in leakage but not to full bore rupture. It is hoped that
authorities, based on the project results, can update their QRA calculation procedures
with certain leak scenarios.
TNO provides detailed results in its reports 2015 R10689 (Phase 1 tests) and 2016 R10126
(Phase 2 tests). In summary, crush testing under truck wheels (either 4 or 8 passes) on both
edges and flat surfaces was performed at cryogenic temperatures on:
 Four 2” diameter corrugated metal hoses made by three different manufacturers (two

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 91
 double-braided and one single-braded).
 2” and 4” diameter multi-composite hoses made by one manufacturer
Fatigue testing was performed on 2” and 4” diameter multi-composite hoses from one
manufacturer, and in summary:
 The first observed leak of any type for 2” diameter hose under cryogenic and
atmospheric axial fatigue testing occurred at 0.3 – 4.0 million cycles.
 The first observed leak at a level judged to be detectable in service for 2” diameter hose
under cryogenic and atmospheric axial fatigue testing occurred at 0.9 – 3.0 million
cycles.
 The first observed leak of any type for 4” diameter hose under cryogenic and
atmospheric axial fatigue testing occurred at 0.13 - 0.6 million cycles.
 The first observed leak in service for 2” diameter hose under cryogenic bending fatigue
testing occurred at 900 cycles. The bending fatigue test included a bend restriction
device and flexible length of the hose was only 16 cm (6.3”).
 No full bore ruptures occurred before leakage judged to be detectable by an operator
occurred and testing stopped (such as would occur by a hose being taken out of service).
Conclusions from Phases 1 and 2 of this project include the following excerpt from pages 34 and
36 of TNO report 2016 R10126:
7.2 Project conclusions on “leak before burst”
For one of the credible scenarios – a loaded trailer wheel running over a pressurized hose
- it is demonstrated that none of the five hoses tested (significantly) leaked. The lightest
metal corrugated hose without outer protective wire only showed minor leakage in the
order of 1 l/min after multiple wheels passing. This is very much less than a “full bore
rupture” release. For the composite hoses it is even demonstrated that the crush damage
does not affect the pressure resistance of the hose (>100 barg). It is concluded that “leak
before burst” is proven for the incident “crush by trailer wheel” for five different types of
hoses that could be used in small scale LNG transfer operations. Based on the discussion
in chapter 8, it is expected (that is: it is not proven) that the observed performance will
likely hold for a wider range of crash and impact scenarios, other diameters, and other
brands using similar materials and technologies. This is in particular the case for multi
composite based hoses. Metal hose suppliers can design their product such that it resists
crush loads specified.
Gradual degradation over service due to cyclic loads cannot be avoided and is a potential
cause of inducing burst failure. It is demonstrated that 2” and 4” multi composite hoses of
a particular brand, subjected to cyclic loads, developed leak paths that grow relatively
slowly over time while the residual burst pressure remain (significantly) higher than the
maximum operating pressure (21 barg). Thus, full bore rupture of a fatigue degraded 2”
or 4” composite hose of Gutteling during a LNG transfer operation is not a credible
scenario. The confidence in extrapolation to fatigue test results shifts from “there is
evidence” to “it is not possible” for respectively other loading conditions, other diameters
of the same brand, and different brands. No conclusions can be drawn on the performance
of metal hose with fatigue damage.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 92
 For the scenario crush and fatigue it is clearly demonstrated that for the composite hoses
tested: 1) the damage is (likely) detectable, 2) full bore rupture is not expected and 3) that
the resulting release of product (LNG/NG) will be insignificant compared to full bore
rupture. With respect to metal hoses tested it is demonstrated that a high crush load 1)
will not (always) result in damage, 2) significant damage that may cause leakage is
detectable, 3) the resulting release of product (LNG/NG) will be insignificant compared
to full bore rupture.
It is concluded that there are scenarios, which so far were intuitively labeled as
potentially resulting in full bore rupture, will result in a relatively very small leakage of
NG or LNG that is insignificant in view of the currently assumed full bore rupture.
And also provided below are a few of the question and answers in section 8 of TNO report 2016
R10126:
8 Review of project aims
1 Is it feasible to prove leak before burst for a particular multi composite hose for at
least two potential critical failure scenarios?
Yes, for a multi composite hose it has been demonstrated that crushing and fatigue
likely result in leakage that will be detected by an operator under normal operating
conditions while the residual pressure resistance is higher than the normal operating
pressures (10 to 18 barg).
3 What will be the likely failure modes based on technical judgement (not necessarily
supported by evidence with tests or literature)?
The likely failure mode for multi composite hoses featuring materials, design and
manufacturing technologies similar to that of Gutteling, under incidents referenced
under point 1, is leakage of NG and or limited LNG. Full bore rupture will likely not
occur. Further it has been demonstrated that fatigue damage will grow relatively
slowly which improve the chance on detection.
Metal corrugated hoses have only been tested under crush loads only. From this test
program no further conclusions can be drawn. Based on the hose design and material
behavior it is assumed that, if fatigue cracks or large plastic deformations are present,
leakage can develop faster than in composites.
7 What tests shall be executed to (partially) provide input to RIVM with respect to
reduction of safety distances.
During execution of the project the input RIVM, being a member of the hose working
group, was taken into account in selecting scenarios and tests. The RIVM
representative stated at the final stage of phase 2, that for composite hoses no other
tests are high on their priority list with respect to proving “leak before failure”.
Obviously it is recognized by the working group that for metal corrugated hoses such
evidence is not provided in this test program and is not found in open literature.
If one concluded for example from these recent tests that an experimentally-observed failure rate
for 4” diameter LNG hoses is 0.1 million cycles (1 x 10-5) until any type of leak was observed
(i.e. a very small hole category, and comparable to or no greater than “release from hole with
effective diameter of 10% transfer hose diameter”, or 0.4”), then this would equate to:

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 93
  5 x 10-3 leak failures/year if there are 500 operations/year (= 1 leak failure/((100,000
cycles/500 operations/yr)
which would compare to the FRT’s current specification of 4 x 10-1 leak failures/year for
“release from hole with effective diameter of 10% diameter with maximum of 50 mm (2 in)” in
the current FRT, i.e. a ~100x difference.
In summary, these recent test results:
 Provide data that five different types of hoses (both multi composite and corrugated
metal) that could be used in small scale LNG transfer operation survived truck tire
crushing without rupturing.
 Provide data that a full bore rupture of a fatigue-degraded 2” or 4” diameter multi
composite Multi-LNG White hose made by Gutteling B.V. is not a credible scenario in
both the fatigue and crush tests performed in these tests. In every scenario tested no full
bore ruptures occurred before leakage at a level judged to be detectable by an operator
occurred and testing was stopped.
 Support the possible reduction of up to 100x in the truck transfer hose leak frequency
failure rate specified in the FRT. However, the recent fatigue testing was only done on
composite hoses (and none on any corrugated metal or other hoses) and thus in its
consideration of the potential for fatigue-induced burst failure TNO concluded:
Thus, full bore rupture of a fatigue degraded 2” or 4” composite hose of Gutteling
during a LNG transfer operation is not a credible scenario. The confidence in
extrapolation to fatigue test results shifts from “there is evidence” to “it is not
possible” for respectively other loading conditions, other diameters of the same
brand, and different brands. No conclusions can be drawn on the performance of
metal hose with fatigue damage.
Summary Comments
There are a number of active developments currently underway to develop new types of LNG
truck transfer hoses and loading arms. This equipment category is an important “technology
watch” area, and additional statistical failure frequency estimates or rates will likely be
generated.
 At this time, no direct basis was identified to propose revising the specification in the
FRT for truck hose leak “release from a hole in transfer hose with effective diameter of
10% transfer hose diameter with a maximum of 50 mm (2-inches)”, including any that
would raise the specified leak frequency above the FRT 3 x 10-5 failures/year threshold
criterion.
 However, it appears that credible evidence has been developed for PHMSA and FERC to
consider eliminating the truck hose rupture rate in the FRT specifically for multi
composite hoses such as Multi-LNG White made by Gutteling B.V. PHMSA and FERC
may want to pursue additional discussions with RIVM regulators to assess RIVM’s
conclusions regarding the recent cryogenic hose testing performed by TNO for the Dutch
LNG Safety Program.
Also, PHMSA and FERC could consider revising the FRT so that hoses and arms failure rates
are specified on a “per operation” or “per hour of operation” basis. FERC currently guides

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 94
applicants to describe the number of LNG truck loadings and unloadings per year in their
applications52.
Recommendations
 If FERC and PHMSA want to retain a “failures per year of operation” basis of
specification, then they may consider making no change to the truck transfer hose leak or
rupture frequencies because even a 100x decrease in the leak rate frequency, e.g. from 4 x
10-1 to 5 x 10-3 failures per year, is still greater than the FRT criterion and triggers
analysis of a SALS design spill. Corollary recommendations are:
o PHMSA and FERC should consider if the current rupture rates for truck transfer
hoses are still appropriate based on the results on the recent tests by TNO, and
perhaps after consultation with RIVM and others. For example, PHMSA and
FERC could consider eliminating the truck transfer hose rupture failure rate
frequency if an applicant’s proposed hose is a multi-composite hose such as either
Gutteling B.V.’s Multi-LNG White hose or a multi-composite hose of similar
materials, design and manufacture.
o PHMSA and FERC should continue to review and consider new technical data
generated related to LNG truck transfer hose and loading arm leak and rupture
rates as it considers future revisions to the FRT.
o PHMSA and FERC should consider reviewing this specification again in more
detail if it specifies hose rupture rate for any risk assessment methodologies other
than the current FRT Criterion methodology.
 If FERC and PHMSA want to consider a “time of use” or “frequency per hour of
operation” on an annual basis in the FRT specification, then FERC and PHMSA could
consider applying the “RIVM BEVI ‘09” specifications since these are on a per hour
basis and thus take into account duration of use, and also because they are applicable to
ship transfer (bunkering) operations. Consideration of potentially eliminating the truck
hose rupture failure rate for multi-composite hoses could be included in this manner also.

6.5 Ship Transfer – Arms (and Hoses)

The failure rate results for ship transfer arm failure rate in the example Wisdom of the Crowd
results indicate that the FRT’s specified failure rates for ship transfer arms is approximately
equivalent to the “RIVM BEVI ‘09” in the reference case analyzed, and lesser than “HSE FRED
JUN’12”. However, these results will vary greatly depending on the reference case selected,
since:
 the FRT specifies a fixed rate in “failures per year of operation”; whereas
 most other references (except NFPA 59A 2013 and 2016 editions) specify failure rates on
a “per hour of operation” or “per operation” frequency basis.
Disadvantages of specifying a rate in “failures per year of operation” include determining a
reasonable usage basis for setting the criteria, without creating unnecessary safety-, operational-
or feasibility-related implications for those facilities that have fewer transfer operations.
A summary of some recent events related to LNG ship hose failure rates is provided below, to
provide additional context to the conclusions and recommendations.
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 95
Comments re: GIIGNL (The International Group of Liquefied Natural Gas Importers)
A representative of GIIGNL’s Technical Study Group (TSG) indicated that the TSG currently
has an initiative underway to analyze GIIGNL’s database of incidents that occured at GIIGNL
member’s facilities; it is possible that some data may be publicly released in 2017.53 This
representative also highlighted a GIIGNL-authored technical paper54 delivered at the LNG17
conference in 2013; this paper provides failure/leak rate statistics by plant subsystem (e.g. 1.49
incidents/million hours for “LNG Tanks, In-Tank Pumps & BOG Facilities”, but unfortunately
provides no break-out data specifically for specific components such as transfer arms).

Ship Transfer Hoses in addition to Arms

PHMSA with FERC may want to consider adding a category in the FRT for ship transfer hoses,
since hoses are currently being used for some LNG bunkering operations as well as ship-to-ship
transfer, and this application is expected to grow. Commercial products such as Trelleborg
Industrial Solutions’ LNG Cryoline floating LNG hose are being considered at small-scale LNG
terminals.55 Results from this project’s database for ship transfer hose failure rates are provided.

Physical Testing of Ship Transfer Hoses in 2011

TNO, the Dutch contract research organization, performed physical testing in 2011 of 8”
diameter multi-composite hoses manufactured by Gutteling B.V. for ship-to-ship transfer of
LNG, and reported that it “successfully passed all EN 1474-part 2 requirements.”56 These
requirements include that a minimum of 400,000 bending cycles not result in leakage or damage
under operating conditions (cryogenic and pressure). This implies that a LNG hose leak (or
rupture) failure frequency that meets EN 1474-part 2 requirements should be < 2.5 x 10-6
failures/cycle (i.e. <1 failure/400,000 bending cycles). This equates to a failure rate of:
 < 1.25 x 10-4 leaks/year if 50 cycles/year
 < 1.25 x 10-3 leaks/year if 500 cycles/year

Comments re: relevant analysis by DNV

As an additional point of comparison, DNV applied an estimated failure frequency of 7.6 x 10-5
total failures/visit for LNG bunkering hardarms in a QRA conducted in 2013 for Skangass AS’
LNG Bunkering Terminal at Risavika Harbour in Norway.57
Summary Comments
There are a number of active developments currently underway to develop new types of LNG
ship transfer hoses and loading arms, for applications ranging from bunkering facilities to
import/export terminals. This equipment category is an important “technology watch” area, and
additional statistical failure frequency estimates or rates will likely be generated. However at
this time, no direct basis was identified to propose revising the specification in the FRT for this
category.
Recommendations
 If FERC and PHMSA want to retain a “failures per year of operation” basis of
specification, then they may consider that no change is needed to the ship transfer arm
rupture or leak rates currently specified in the FRT. A corollary recommendation is:

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 96
 o PHMSA and FERC should continue to review and consider new technical data
generated related to LNG ship transfer loading arm and hose leak and rupture
rates as it considers future revisions to the FRT.
 If FERC and PHMSA want to consider a “frequency per hour of operation” for loading
arms on an annual basis in the FRT specification, then FERC and PHMSA could consider
applying the “RIVM BEVI ‘09” specifications since these are on a per hour basis and
thus take into account duration of use.
 PHMSA should consider establishing a category for ship transfer hoses, and if so then
FERC and PHMSA could consider applying the “RIVM BEVI ‘09” specifications since
these are applicable to ship transfer operations.

6.6 Piping – Rupture (and Leak) of Valve

Analysis of Valve Rupture
The source references that contained valve rupture data are shown in Appendix C, and
calculations shown for valve rupture rates in scenario evaluations are shown in Framework Index
#143. A summary of the findings in Section 4.6 and this analysisis:
 The equipment failure rate regulatory references that appear to be frequently cited and
used globally such as “HSE FRED JUN’12”, “RIVM BEVI ‘09” and “LNE ‘09” do not
specify a valve rupture rate. “RIVM BEVI ‘09” includes the failure of valves with its
failure rate for piping, and does not separately define valve failure rate frequencies (ref.
Module C p. 42).
 The data arising from “GRI LNG FRD ‘81”, the most relevant analysis found for LNG
and cryogenic valves, identified zero cryogenic valve ruptures and two leaks in more than
43 million hours of operation. As noted in Section 4.6 of “GRI LNG FRD ‘81”, “One
safety-related failure was reported, but this was attributed to human error and is not
included as a valve failure”. This equates to an observed LNG valve rupture failure rate
< 2 x 10-4 ruptures/year (=1/4,909 years), but does not inform the rate specified in the
FRT since this is merely a minimum rate based on the very short time period examined in
“GRI LNG FRD ‘81”.
 The source of other generic data for valve rupture rates in some of the references are
based in part on nuclear reactor plant valves, where the definition of rupture can be
leakage from piping “greater than 50 gpm”58.
For additional comparison:
 A more recent comparative analysis of nuclear plant valve rupture failure rates is 3.12 x
10-9 failures/year, i.e. 2.2 x 10-5 failures/year as per the mean failure rate for “Manual
Valve External Large Leak” in Table 5-1 of “INL NUC ‘07” (considered a rupture as
noted in Section 4.1); this failure rate database presumably comprises significant carbon
steel as well as some stainless material in valve construction, and the value of 2.2 x 10-5
failures/year is less than the current FRT criterion of 3 x 10-5.
 “DNV FFG HCRD ‘13” indicates 4.859 x 10-5 failures/year and 1.347 x 10-5 failures/year
total failure frequency rates for a >150 mm (>6”) hole to occur in a 6” (150 mm) manual
or actuated valve, respectively (ref. pp. 32 and 34) - - i.e. which some could consider as a

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 97
 full-bore rupture as defined on p. 9 of this reference. However as noted in section 4.1,
GTI concurs with DNV’s comment in “DNV FFG HCRD ‘13” that there is considerable
uncertainty related to using the HRCD data to estimate equipment ruptures.
 The mean failure rate for Catastrophic “Valves-Manual” on p. 199 of “CCPS ‘89” is
more than 100 times greater than that specified in the current FRT. No explanation for
this large difference was identified. “CCPS ‘89” p. 128 cites SAIC’s proprietary data set,
and four of the five sub-references cited appear to pertain prinicipally to nuclear power
plants.
 The valve rupture failure rate data in Appendix D of “INL CHEM ‘95” is also
approximately 100 times greater than that specified in the current FRT. As summarized
in Section 4.1, the analysis underlying Appendix D of “INL CHEM ‘95” considered data
sources from multiple industries but its Category 1 data (“sources with actual failure data
obtained from a detailed review of failure events”) was generated from the nuclear
industry.
 “OREDA ‘15” provides a valve failure data for “External leakage – Process medium” but
as noted in section 4.1 judgement is required to characterize the severity of this leak and
its relevance to valve “rupture”.
In summary, there is a wide range of data associated with valve rupture of differing pedigrees
and industry sources, and some of the higher failure rate frequencies appear to be associated with
data originating principally from the nuclear power industry. The rupture rate of valve of 9 x 10-
6
failures/year currently specified in the FRT appears to be conservative in comparison to an
effective rate of zero in HSE FRED JUN’12”, “RIVM BEVI ‘09”, “LNE ‘09” and zero valve
ruptures identified in “GRI LNG FRD ‘81”; the current rate also exactly matches the value
contained in “PNL PSRP ‘82”. It is recommended that the current valve rupture rate in the FRT
be retained.
Of course one notable valve rupture incident that did occur in the LNG industry was in 1977 at
an LNG plant in Arzew, Algeria. However that valve body that ruptured was constructed of cast
aluminum, whereas the current practice is that valves in LNG service are constructed of stainless
steel.59
Analysis of Valve Leak
Limited quantitative data of valve leak rates by hole size was found.
“RIVM BEVI ‘09” and “LNE ‘09” do not specify valve external leak rates.
“HSE FRED JUN’12” in its Item FR 1.2.1 “Valves” does not specify a valve rupture rate, but it
does specify a 2 x 10-4 Spray Release Frequency failures/year for valve leaks.
 HSE cites its confidential Components Failure Rates publication in its derivation of the
rates in Item FR 1.2.1 “Valves”, which included “a comparison of 12 sources of failure
rates derived elsewhere” and that the “values are for chlorine duty although the review
included LPG, petrochemical, steam/water, nuclear and other data.” It also lists 34
additional references.
 As summarized in Section 4.6, the project team associated 2 x 10-4 Spray Release
Frequency failures/year rate with “Release from a hole with effective diameter of 2 mm”,
which the project team considered a conservative assignment.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 98
  HSE’s specified rate of 2 x 10-4 Spray Release failures/year frequency matches surprising
well with the 4 x 10-4 leaks/year leak rate (of unidentified size) for LNG cryogenic valves
identified in “AGA LNG EXP ’79 and “GRI LNG FRD ’81”, as computed by this
project’s team (see Section 4.6).
The HCRD database provides one resource to evaluate valve leak frequencies for various hole
sizes. As an illustration, Figures 11 and 12 provide some publicly-available estimates of valve
total leak frequencies arising from the HCRD database as presented in “DNV FFG HCRD ‘13”
and “IOGP 434-1”; if considered to represent 25 mm and 100 mm holes, respectively, then these
estimated failure frequencies will tend to be higher than those in Appendixes D, E and F because
of their larger hole size ranges. But these figures illustrate the sensitivity of estimated leak
frequency results to hole size, valve size and valve type when considering information in the
HCRD database.

1.200E‐04
10 ‐ 50 mm Hole Size Leak Frequencies/Year 

1.000E‐04
Actuated Valve "DNV FFG HCRD
8.000E‐05 '13"
Leak Frequency

Actuated Valve "IOGP 434‐1"
6.000E‐05
Manual Valve "DNV FFG HCRD
'13"
4.000E‐05
Manual Valve "IOGP 434‐1"

2.000E‐05
LNG FRT Criterion

0.000E+00
0 10 20 30 40 50
Equipment Size (inch)

Figure 11: 10 – 50 mm Hole Size Valve Leak Rate Frequencies/Year in “DNV FFG HCRD ‘13”
(Total Frequency) and “IOGP 434-1” (All Releases)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 99
 50 ‐ 150 mm Hole Size Leak Frequencies/Year 
3.500E‐05

3.000E‐05

Actuated Valve "DNV FFG HCRD
2.500E‐05
'13"
Leak Frequency

2.000E‐05 Actuated Valve "IOGP 434‐1"

1.500E‐05 Manual Valve "DNV FFG HCRD
'13"
1.000E‐05 Manual Valve "IOGP 434‐1"

5.000E‐06 LNG FRT Criterion

0.000E+00
0 10 20 30 40 50
Equipment Size (inch)

Figure 12: 50 – 150 mm Hole Size Valve Leak Rate Frequencies/Year in “DNV FFG HCRD ‘13”
(Total Frequency) and “IOGP 434-1” (All Releases)

For PHMSA’s and FERC’s information and at their request, the project team provided estimates
of valve leak rates for hole sizes of 2, 10, 25, 50 and 100 mm calculated using DNV’s LEAK 3.3
software in Appendixes D, E and F. Observations from a review of these LEAK 3.3 software
results in these Appendixes include:
 Estimated failure rates vary based on valve diameter and valve type (manual or
automatic). This is consistent with the public information summarized in the above
graphs.
 The failure frequency associated with a 2 mm hole in actuated valves is very close to the
2 x 10-4 Spray Release Frequency failures/year that the project team proposes to associate
with a 2 mm hole size. The failure frequency associated with a 2 mm hole in manual
valves is considerably less than 2 x 10-4. Thus the LEAK 3.3 results for 2 mm hole
results appear consistent with associating the 2 x 10-4 Spray Release Frequency
failures/year rate with “Release from a hole with effective diameter of 2 mm” for the
“worst case” valve type (i.e. actuated valves) in the HCRD.
 All failure frequencies calculated for either a 50 mm or 100 mm hole for any of the
FRT’s pipe diameter ranges, and for either manual or actuated valves, are less that the
FRT criterion of 3 x 10-5 and therefore fall below the FRT’s analysis threshold.
 Failure frequencies calculated for a 25 mm hole vary across the FRT’s pipe diameter
ranges, and for either manual or actuated valves, but for the “worst case” actuated valves
the average is 3 x 10-5 failures/year. If PHMSA and FERC want to associate a failure
rate to a 25 mm valve hole leak, then PHMSA and FERC might consider a 3 x 10-5
failure/year frequency rate if they want to use a rate based solely on the HCRD database
and using the hole size range used in this analysis.
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 100
  Failure frequencies calculated for a 10 mm hole vary across the FRT’s pipe diameter
ranges, and for either manual or actuated valves, but for the “worst case” actuated valves
the average is 2 x 10-4 failures/year. If PHMSA and FERC want to associate a failure
rate to a 10 mm valve hole leak, then PHMSA and FERC might consider a 2 x 10-4
failure/year frequency rate if they want to use a rate based solely on the HCRD database
and using the hole size range used in this analysis.
While the HCRD is very valuable, it does represents one data set. In comparison, HSE includes
more than 35 citations in its References, Advice Note and Bibliogaphy in its determination of its
specifications in Item FR 1.2.2 “Valves”.
The piping failure rate categories used in the FRT are consistent with those in Item FR 1.3
“Pipework” of “HSE FRED JUN’12” (with two small exceptions). For consistency, the project
team recommends that PHMSA and FERC similarly apply for valves the specifications in Item
FR 1.2.1 “Valves” of “HSE FRED JUN’12” and associate the Spray Release Frequency with a 2
mm hole size.
As an additional point of comparison, Table 1 (ref. p. 18) of “INL CHEM ‘95” cites average
“leak” rates ranging from 3.22 x 10-6 failures/year to 3.48 x 10-6 failures/year for gate, ball and
globe valves in the Idaho Chemical Processing Plant Failure Rate Database; the size of the leak
is not characterized, and rupture rates are not provided in this database. But the rupture rate
recommended below (i.e. retaining the FRT’s current specification of 9 x 10-6 failures/year) also
appears conservative in comparison to this reference.
Recommendation:
 PHMSA and FERC should consider retaining its current valve rupture rate of 9 x 10-6
failures/year, which exactly matches the value contained in “PNL PSRP ‘82”. Some
other leading failure rate references effectively assign valve rupture rate to zero, so a
corollary recommendation is:
o PHMSA and FERC should continue to re-assess the specified valve rupture rate
of 9 x 10-6 failures/year if it considers applying this rate to any risk assessment
methodologies other than the current FRT Criterion methodology.
 PHMSA and FERC should consider adding a valve leak rate of 2 x 10-4 failures/year as
“Release from valve with a hole with effective diameter of 2mm (0.08 inch)”.

6.7 Piping – Rupture of Expansion Joint

Section 4.7 summarizes that the FRT’s current specification of 4 x 10-3 failures/year may perhaps
be associated with the failure of an complete transfer line rather than an individual expansion
joint component.

The only reference sources identified for expansion joint failure rates were more than 40 years
old. For example, the failure rate of “Leak (serious) in post-accident situation, λo” for “Elbows,
flanges, expansion joints (containment quality)” listed in Table A14.7 of “LEES’12” is from “the
Rasmussen Report (AEC, 1975)” i.e. the WASH-1400 reactor safety study.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 101
While “PNL PSRP ’82” specifies 1 x 10-7 faults/hour (~1 x 10-3/year) as “Expansion Joint
Rupture”, it refers to two sources:
 “SAI ‘75” specifies a mean “External Joint Ruptures” failure rate of 1 x 10-8/hour (~1 x
10-4/year).
 “WELKER ‘76” specifies a median “External leak or rupture” failure rate for an
expansion joint of 2 x 10-6/hour (~2 x 10-2/year) with a Low rate of 1 x 10-8/hour (~1 x
10-4/year) and High rate of 1 x 10-5/hour (~1 x 10-1/year).”
Based on the above one could reasonably consider that PHMSA and FERC should specify the
“SAI ’75” rate of 1 x 10-4 faults/year rather than the “PNL PSRP ’82” rate of 1 x 10-3/year, since
the underlying reference indicates that the rate used by PNL is for rupture or leak.
The “PNL ITRP ‘82” reference analyzes the failure rate of a Double Ply Expansion Joint
assembly (with pressure detection system) in a FTA in its Table B.12 (see Appendix B), and
calculated a failure rate of 2 x 10-9 faults/year, based on a single ply failure rate of 1 x 10-7
faults/hour and 1,500 hours/yr operation. If a pressure detection system is not present, then the
failure rate based on the methodology in Table B.12 would be (1.5 x 10-4 faults/yr)2 = ~2 x 10-8
faults/year.
If the failure rate for a single ply expansion joint is revised to 1 x 10-4faults/year, then the failure
rate of a double ply expansion joint is estimated as (1 x 10-4 faults/year)2 = 1 x 10-8 faults/year, or
comparable to the failure rates described in the prior paragraph when considering the “PNL ITRP
‘82”. While the pedigree of even the base single ply failure rate of 1 x 10-4faults/year for a single
ply expansion joint is unclear, these calculations support an estimated failure rate of a double ply
expansion joint well below the FRT’s 3 x 10-5 criterion.
Recommendation:
 PHMSA and FERC should consider revising its rupture failure rate specification from 4 x
10-3/year to 1 x 10-4/year and clarify that it applies to Single Ply Expansion Joint. A
corollary recommendation is:
o PHMSA and FERC should consider reviewing this specification again in more
detail if it specifies a Single Ply Expansion Joint rupture rate for any risk
assessment methodologies other than the current FRT Criterion methodology.
 PHMSA and FERC should consider specifying a rupture failure rate specification of 1 x
10-8/year for “Double Ply Expansion Joint” based on the above analysis

6.8 Piping – Failure (Rupture or Leak) of Gasket

The failure rate results identified in this analysis are substantially different than the “failure of
gasket” failure rate that is specified in the current FRT.
The source references applied in this analysis for failure rates associated with flanges and gaskets
are shown in Appendix C. Calculations for rupture rates in different scenario evaluations are
shown for example in Framework Index #173 for all diameters.
Analysis of Gasket Leak
Limited quantitative data of gasket leak rates by hole size was found.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 102
“RIVM BEVI ‘09” and “LNE ‘09” do not specify gasket or flange leak rates.
As summarized in Section 4.8, “HSE FRED JUN’12” in its Item FR 1.2.4 “Flanges and Gaskets”
does specify a gasket failure rate, and hole sizes of 25 mm and 50 mm were associated in this
analysis with those specifications.
In addition, the HCRD database provides one resource to evaluate flange leak frequencies for
various hole sizes. Appendix L illustrates that the majority of the flanges in the HCRD database
are compressed flanges, and many are also of Ring Type (RTJ) or Clamp (e.g. Grayloc) type.
Figures 13 and 14 illustrate some publicly-available estimates of flange total leak frequencies
based on the HCRD database as presented in “DNV FFG HCRD ‘13” and “IOGP 434-1”; if
considered to represent 25 mm and 100 mm holes, respectively, then these failure estimated
failure frequencies will tend to be higher than those in since the Appendixes D, E and F because
of their larger hole size ranges. But these figures provide one indication of the sensitivity of the
results to flange and hole size on leak frequency from information in the HCRD database.

10 ‐ 50 mm Hole Size Leak Frequencies/Year 
4.000E‐05

3.500E‐05

3.000E‐05
Leak Frequency

2.500E‐05

2.000E‐05 Flange "DNV FFG HCRD '13"
Flange "IOGP 434‐1"
1.500E‐05
LNG FRT Criterion
1.000E‐05

5.000E‐06

0.000E+00
0 10 20 30 40 50
Equipment Size (inch)

Figure 13: 10-50 mm Flange Leak Rate Frequencies/Year in “DNV FFG HCRD ‘13” (Total
Frequency) and “IOGP 434-1” (All Releases)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 103
 50 ‐ 150 mm Hole Size Leak Frequencies/Year 
3.500E‐05

3.000E‐05

2.500E‐05
Leak Frequency

2.000E‐05
Flange "DNV FFG HCRD '13"
1.500E‐05
Flange "IOGP 434‐1"
1.000E‐05 LNG FRT Criterion

5.000E‐06

0.000E+00
0 10 20 30 40 50
Equipment Size (inch)

Fig. 14: 50-150 mm Flange Leak Rate Frequencies/Year in “DNV FFG HCRD ‘13” (Total
Frequency) and “IOGP 434-1” (All Releases)
For PHMSA’s and FERC’s information and at their request, the project team provided estimates
of flange leak rates for hole sizes of 2, 10, 25, 50 and 100 mm calculated using DNV’s LEAK
3.3 software in Appendixes D, E and F. Calculated estimates of flange leak rates for the various
analysis scenarios when all reference sources were considered are also provided in Appendixes
D, E and F. See for example the results of index #368 for a 25mm hole and #369 for a 50 mm
hole.
As described in Section 4.8 this analysis also considered the gasket leak rates specified in Item
FR 1.2.4 “Flanges and Gaskets” of “HSE FRED JUN’12” , which lists 25 citations in its
reference, advice note and bibliography. The failure rates specified in “HSE FRED JUN’12” and
associated in this analysis with 25mm and 50mm leak hole sizes align well with the results
shown in Appendixes D, E and F in index #368 (for 25mm hole) and #369 (for 50 mm hole).
Therefore it is recommended that PHMSA and FERC consider incorporating the “HSE FRED
JUN’12” leak rate data and assign the leak rate frequencies to the hole diameters identified in
Section 4.8.
Analysis of Gasket Rupture
The leading failure rate references “HSE FRED JUN’12”, “RIVM BEVI ‘09” and “LNE ‘09” do
not specify a large-scale (i.e. full-bore or complete failure) rupture rate for gaskets or flanges:
 “RIVM BEVI ‘09” includes failure of connections with failure of the pipes, and does not
separately define flange failure rate frequencies (e.g. ref. Module C pp. 30, 42-44).
 “HSE FRED JUN’12” specifically addresses flanges and gaskets in its Item FR 1.2.4 but
does not specify a rupture rate.
Where gasket or flange rupture data was specified, the sources identified were 40+ years old and
their pedigree appears to be from the nuclear power industry. For example, “SAI ‘75” cites a
“Connection Flange Ruptures” rate of 1 x 10-8/hour or 1 x 10-4/year; “SAI ‘75” does not cite a

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 104
specific source for its connection flange rupture data, but the 1975 WASH-1400 Reactor Safety
Study is one of few process-related references in its list of twenty References.
For comparison, “DNV FFG HCRD ‘13” indicates on page 20 that the total failure frequency
rate for a >150 mm (>6”) hole to occur in a 6” (150 mm) flange is 6.852 x 10-6 - - i.e. which
some could consider a full-bore rupture of a gasket/flange as defined on p. 9 of this reference.
As noted in section 4.1, GTI concurs with DNV’s comment in “DNV FFG HCRD ‘13” that there
is considerable uncertainty related to using the HRCD data to estimate ruptures. Appendixes L
and M illustrates the raw data in the HCRD database, and the limited basis for predicting flange
(or other equipment) ruptures from HCRD data.
The “HSE FRED JUN’12” leak rate specification provides useful comparative information. This
project team did not identify any engineering reasoning why a full-bore gasket rupture frequency
(if specified) should exceed leak frequency (i.e. > 1 x 10-7 failures/year) specified in “HSE FRED
JUN’12” that this analysis associated with a 50 mm hole. Therefore it is recommended that the
FRT be clarified that “Failure of Gasket” does not refer to a full-bore or complete guillotine
flange rupture.
Recommendation:
 PHMSA and FERC should consider revising the gasket failure rate specification as
follows:
o Add “Release of one segment of gasket as a hole with effective diameter of 25
mm (1-inch)” failure rate as 5 x 10-6 failures/year
o Add “Release of a Spiral Wound Gasket as a hole with effective diameter of 50
mm (2-inch)” failure rate to 1 x 10-7 failures/year
o Eliminate “Failure of Gasket” terminology.

6.9 Piping – By Diameter

The only LNG-specific (and cryogenic-specific) directly-observed piping failure rate data
identified in this project analysis was that obtained from the combined results of “GRI LNG
FRD ‘81” and “AGA LNG EXP ‘79” analysis of peak shaving plants, as summarized in Section
4.9. The project team’s analysis of that data resulted in a rate of:
 2 x 10-6 failures/meter-year (= 8 x 10-10 failures/foot-hour)
which compares favorably to PHMSA’s current FRT specifications of:
 2 x 10-6, 5 x 10-6 or 7 x 10-7 specified failures/meter-year frequency rate if the hole
represents “Release from a hole with effective diameter of 25mm (1 inch)” in piping size
categories <2”d, 2”≤d<6” or 6”≤ d<12”, respectively
Calculated failure rate results based on the various weighting scenarios considered in this
analysis are presented in Appendixes D, E and F. Results can vary significantly depending on
the weightings given to individual data sets or perceived relevancies. This is to be expected
given the relatively large span of estimated failure rate data.
Table 20 and Figure 15 compare specified or actual rupture or guillotine failure frequencies from
some key references.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 105
 FRT Piping Diameter Category
Reference 2"≤d<6" 6"≤d<12" 12"≤d<20" 20"≤d<40"
FRT Nominal Failure Rates "Piping"
5.00E-07 2.00E-07 7.00E-08 2.00E-08
"Catastrophic Rupture"
"NFPA 59A '16" "Piping - aboveground"
3.00E-07 1.00E-07 1.00E-07 1.00E-07
"Rupture"
"HSE FRED JUN'12" Item FR 1.3 "Pipework"
5.00E-07 2.00E-07 7.00E-08 4.00E-08
"Guillotine"
"HSE FRED JUN'12" Item FR 3.1.2 "Above
Ground Pipelines"
"Rupture (>110mm diameter)" ("...applicable 6.50E-09 6.50E-09 6.50E-09 6.50E-09
to natural gas above ground installations..." but
see Item FR 3.1.2 for limitations)

"RIVM BEVI '09" "pipelines aboveground"

3.00E-07 1.00E-07 1.00E-07 1.00E-07
"Rupture"
"RIVM BEVI '09" "underground transport
pipelines"
5.00E-07 5.00E-07 5.00E-07 5.00E-07
"Other pipelines"
"Rupture"

"RIVM BEVI '09" "underground transport

pipelines"
1.53E-07 1.53E-07 1.53E-07 1.53E-07
"Pipeline complies with NEN 3650"
"Rupture"

"RIVM BEVI '09" "underground transport

pipelines"
7.00E-09 7.00E-09 7.00E-09 7.00E-09
"Pipeline in pipe bay"
"Rupture"
"LNE '09" "Underground pipeline"
2.80E-08 2.80E-08 2.80E-08 2.80E-08
"Rupture"
"EGIG '15" Nat. Gas Trans. Onshore Pipelines
1.33E-07 6.40E-08 3.75E-08 9.67E-09
"Rupture" (actual '70-'13)
"PHMSA NGT GTI '16" Nat. Gas Trans.
Onshore Pipelines
2.78E-08 3.23E-08 4.20E-08 2.96E-08
"Rupture" (actual '10-'15)
as calculated in App. J
"PHMSA HL GTI '16" Haz. Liquid Trans.
Onshore Pipelines
3.95E-08 2.45E-08 2.01E-08 3.90E-08
"Rupture" (actual '10-'15)
as calculated in App. K
Table 20: Comparison of Piping Rupture or Guillotine Failure Rates/m-yr, by FRT Piping
Diameter Category (see text for approximations used to apply data to FRT’s piping diameter
categories, when necessary)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 106
 Comparison of Rupture or Guillotine Failure Rates/m‐yr, by FRT Piping 
Diameter Category

"PHMSA HL GTI '16" Haz. Liquid Trans. Onshore
Pipelines
"Rupture" (actual '10‐'15) as calculated in App. K

"PHMSA NGT GTI '16" Nat. Gas Trans. Onshore
Pipelines
"Rupture" (actual '10‐'15) as calculated in App. J

"EGIG '15" Nat. Gas Trans. Onshore Pipelines
"Rupture" (actual '70‐'13)

"LNE '09" "Underground pipeline"
"Rupture"

"RIVM BEVI '09" "underground transport pipelines"
"Pipeline in pipe bay"
"Rupture"

"RIVM BEVI '09" "underground transport pipelines"
"Pipeline complies with NEN 3650"
"Rupture"

"RIVM BEVI '09" "underground transport pipelines"
"Other pipelines"
"Rupture"

"RIVM BEVI '09" "pipelines aboveground"
"Rupture"

"HSE FRED JUN'12" Item FR 3.1.2 "Above Ground
Pipelines" "Rupture (>110mm diameter)"
("...applicable to natural gas above ground…

"HSE FRED JUN'12" Item FR 1.3 "Pipework"
"Guillotine"

"NFPA 59A '16" "Piping ‐ aboveground"
"Rupture"

FRT Nominal Failure Rates "Piping"
"Catastrophic Rupture"

0.0E+00 1.0E‐07 2.0E‐07 3.0E‐07 4.0E‐07 5.0E‐07 6.0E‐07

20"≤d<40" 12"≤d<20" 6"≤d<12" 2"≤d<6"

Figure 15: Comparison of Piping Rupture or Guillotine Failure Rates/m-yr, by FRT Piping
Diameter Category (see text for approximations used to apply data to FRT’s piping diameter
categories, when necessary)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 107
A review of Table 20 and Figure 15 indicates that the nominal piping rupture rate frequencies
specified by the FRT are greater than the actual mean ruptures rates calculated in Appendix J and
K for onshore natural gas transmission pipeline and hazardous liquid pipeline from PHMSA’s
pipeline incident database for 2010-2015 for all categories except 20”≤d<40”. When
considering this it may be beneficial to recall that external interference and corrosion are the
leading cause of the pipeline ruptures; for example, Figure 5 illustrates that about 70% of
EGIG’s pipeline ruptures are due to external interference, and 59% of all EGIG’s pipeline
incidents are due to either external interference or corrosion (of carbon steel piping). In contrast
LNG facilities are fenced-off and secured areas with minimal construction typically underway,
and all cryogenic piping is typically made of stainless steel. Nevertheless, PHMSA and FERC
could consider revising the failure rate for the 20”≤d<40” category from 2 x 10-8 failures/m-year
to 4 x 10-8 failures/m-year and match the rate specified in “HSE FRED JUN’12”.
Table 21 and Figure 16 compare specified or actual leak rates for holes of effective diameter of
20 or 25 mm (or some may so associate) from some public references, actual rates from
PHMSA’s natural gas and hazardous pipeline incidents database (2010-2015) as calculated in
Appendix L and M, DNV’s LEAK 3.3 software, and information publicly available in “DNV
FFG HCRD ‘13”. The following approximation was made solely in Table 21 and Figure 16 in
order to illustrate this “DNV FFG HCRD ‘13” information (this approximation was not made in
the database for this project) - - Total leak frequencies drawn from “DNV FFG HCRD ‘13” used
these pipe sizes from the available data: 4”D for 2"≤d<6"; 10”D for 6"≤d<12"; and 14”D for
12"≤d<20". The “DNV FFG HCRD ‘15” data shown is based on the period October 1992 to
March 2010, whereas the DNV LEAK 3.3 software includes HCRD data more recent than 2010.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 108
 FRT Piping Diameter Category
Reference 2"≤d<6" 6"≤d<12" 12"≤d<20" 20"≤d<40"
FRT Nominal Failure Rates "Piping"
"Release from hole with effective diameter of 2.00E-06 7.00E-07 5.00E-07 4.00E-07
25 mm"
"HSE FRED JUN'12" Item FR 1.3 "Pipework"
1.00E-06 7.00E-07 5.00E-07 4.00E-07
"25 mm diameter"
"HSE FRED JUN'12" Item FR 3.1.2 "Above
Ground Pipelines"
("...applicable to natural gas above ground
1.60E-07 1.60E-07 1.60E-07 1.60E-07
installations..." but see Item FR 3.1.2 for
limitations)
"Pin Hole (≤ 25 mm diameter)"

"RIVM BEVI '09" "underground transport

pipelines"
1.50E-06 1.50E-06 1.50E-06 1.50E-06
"Other pipelines"
"Leak with an effective diameter of 20 mm"

"RIVM BEVI '09" "underground transport

pipelines"
4.58E-07 4.58E-07 4.58E-07 4.58E-07
"Pipeline complies with NEN 3650"
"Leak with an effective diameter of 20 mm"

"RIVM BEVI '09" "underground transport

pipelines"
6.30E-08 6.30E-08 6.30E-08 6.30E-08
"Pipeline in pipe bay"
"Leak with an effective diameter of 20 mm"

"DNV LEAK 3.3" "Process Pipe" "Total"

Frequency
7.40E-06 3.37E-06 3.00E-06 2.93E-06
(based on HCRD database)
(15 - 35 mm) as calculated in this project

"DNV FFG HCRD '13" "Process Pipe" "Total"

Frequency
(based on HCRD database) 7.46E-06 5.52E-06 5.40E-06
(10 - 50 mm) (using 4"D, 10"D and 14"D
available data)

"EGIG '15" Nat. Gas Trans. Onshore Pipelines

(actual '70-'13) 4.45E-07 2.80E-07 1.15E-07 4.37E-08
"Pinhole/crack" (Eff. Diam. ≤ 20 mm)

Table 21: Comparison of Failure Rates/m-yr for a Release from a Hole of Effective Diameter of 20
or 25 mm (or some may so associate), by FRT Piping Diameter Category (see text for
approximations used to apply data to FRT’s piping diameter categories, when necessary)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 109
 Comparison of Failure Rates/m‐yr for a Release from a Hole of Eff. Diam. of 20 
or 25 mm (or some may so associate), by FRT Piping Diameter Category

"EGIG '15" Nat. Gas Trans. Onshore Pipelines (actual '70‐
'13)
"Pinhole/crack" (Eff. Diam. ≤ 20 mm)

"DNV FFG HCRD '13" "Process Pipe" "Total" Frequency
(based on HCRD database)
(10 ‐ 50 mm) (using 4"D, 10"D and 14"D available data)

"DNV LEAK 3.3" "Process Pipe" "Total" Frequency
(based on HCRD database)
(15 ‐ 35 mm) as calculated in this project

"RIVM BEVI '09" "underground transport pipelines"
"Pipeline in pipe bay"
"Leak with an effective diameter of 20 mm"

"RIVM BEVI '09" "underground transport pipelines"
"Pipeline complies with NEN 3650"
"Leak with an effective diameter of 20 mm"

"RIVM BEVI '09" "underground transport pipelines"
"Other pipelines"
"Leak with an effective diameter of 20 mm"

"HSE FRED JUN'12" Item FR 3.1.2 "Above Ground 
Pipelines"
("...applicable to natural gas above ground 
installations..." but see Item FR 3.1.2 for limitations)
"Pin Hole (≤ 25 mm diameter)"

"HSE FRED JUN'12" Item FR 1.3 "Pipework"
"25 mm diameter"

FRT Nominal Failure Rates "Piping"
"Release from hole with effective diameter of 25 mm"

0.0E+00 2.0E‐06 4.0E‐06 6.0E‐06 8.0E‐06

20"≤d<40" 12"≤d<20" 6"≤d<12" 2"≤d<6"

Figure 16: Comparison of Failure Rates/m-yr for a Release from a Hole of Effective Diameter of 20
or 25 mm (or some may so associate), by FRT Piping Diameter Category (see text for
approximations used to apply data to FRT’s piping diameter categories, when necessary)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 110
Potential Reduction Factors for Transfer Piping and Inter-Unit Piping
PHMSA and FERC could consider specifying reduction factors for piping failure rate frequencies
for transfer piping or for inter-unit piping that spans multiple individual units (examples would be
rundown lines or rundown headers which span from liquefaction units to other liquefaction units).
Previous correspondence between FERC and PHMSA appears to support an assertion by some
that the FRT may over-predict single accidental leakage sources for long lengths of large diameter
solid welded piping.
On April 19, 2005 FERC sent a formal letter to DOT PHMSA Office of Pipeline Safety (OPS)
requesting concurrence on their approach for determining single accidental leakage sources. The
letter states that “the design construction, operation and historical integrity of all-welded large
diameter marine transfer piping does not support a full pipe rupture without ignition as a credible
accident scenario. Marine transfer systems are constructed of a relatively thick-walled seamless
pipe, fully x-ray inspected during construction, and operated at moderate pressures (50 to 80 psi).
Maximum flow rates are limited to the 10 to 12 hour cargo unloading period, a time when extra
staff is on hand to monitor operations and detect abnormal events and quickly activate
emergency shutdown systems. As a result, our determination of a single accidental leakage
source for a marine transfer system is based on facility-specific review of piping and
instrumentation diagrams to identify all small diameter attachments to the transfer piping for
instrumentation, pressure relief, recirculation, etc. and any flanges that may be used at valves or
other equipment, in order to determine the largest spill rate.”60
On May 6, 2005, the OPS responded to FERC staff concurring with FERC’s approach. In
addition, OPS noted that “the OPS agrees that the design and construction of marine cargo
transfer systems is very robust and that failure is unlikely under operational constraints.
Moreover, the extensive security and safety oversights provided by the USGC before, during,
and after transfer operations further reduces the risk that a spill could threaten life and property.
There is no documented evidence of a catastrophic failure ever having occurred in either LNG
operational experience or research. ...”61
In 2005, the majority of projects proposed were import terminals as the energy conditions in the
US supported import rather than export of LNG. While the previous correspondence for marine
transfer lines for import terminals is directly related to marine transfer lines for export terminals,
that correspondence also can be applied to other large bore fully welded stainless steel piping
with minimal connections and all of the other safeguards explicitly referenced.
“IOGP 434-1” specifies on p. 25 these reductions for inter-unit and transfer piping:
The frequencies given in datasheet 1 for steel piping are, for onshore installations, intended
to be applied within process units. For piping linking process units (inter-unit pipe) and
piping to/from storage or loading facilities (transfer pipe), the following release frequency
modification factors can be applied:
 Inter-unit pipe: 0.9
 Transfer pipe: 0.8
These have been derived from detailed analysis of the causes of piping failure [5] and
application to this analysis of judgemental modifications to account for the differences in
inter-unit and transfer pipes [6].

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 111
where datasheet 1 refers to “Equipment Type: (1) Steel process pipes”, and references [5] and [6]
refer to confidential reports prepared by Technica or DNV Technica. The project team considers
the rates specified by IOGP to be conservative reductions based on PHMSA’s and FERC’s
correspondence in 2005 summarized above.
As an additional comparison, DNV comments on page 7 of “DNV FFG HCRD ‘13” that:
A common aspect of uncertainty in QRA is associated with the frequency of inter-unit
pipework / pipeline releases. It is widely accepted that the application of process pipework
failure data will tend to give overly conservative values with respect to longer inter-unit pipe
segments. This can be of particular relevance to LNG facilities, where the loading lines are
often several kilometers long. In the course of conducting a large number of QRA studies,
DNV has had the opportunity to draw on the experience of a range of operators. On the
basis of these discussions, it is considered appropriate to apply a factor of 10 reductions in
the pipe-work failure frequency for inter-unit piping. It should be recognised that this is an
engineering judgement assumption, based on acknowledging operational experience that
inter-unit pipework fails very rarely (in comparison to the process pipework within the main
process areas). This revised basis can be of particular relevance to loading lines, although
should not substitute for consideration of all potential loads (and hence potential frequency
modification factors) that may apply to a particular facility, or particular loading line.
Applicability of Generic Piping Failure Rate Data to LNG or Other Cryogenic Service
The UK “HSE FRED JUN ‘12” does not specify any reductions in its Item #FR1.3 piping failure
rates for LNG facilities.
Likewise, “IOGP 434-1”:
 also bases its failure frequency rates for process piping (and other process equipment
items) on the UK HRCD database (ref. p. 27); and
 states that its process release frequencies “can be used for process equipment on the
topsides of offshore installations and for onshore facilities handling hydrocarbons…” and
also states that “We therefore recommend use of the same frequencies for LNG
installations as given in Section 2.0. A 50% reduction could be considered as a
sensitivity but decisions based on this would need to be fully justified.” (ref. p. 19).

Comments re: Pipe-in-Pipe Piping (all types)

The “HSE FRED JUN ‘12”, “RIVM BEVI ‘09”, “LNE ‘09”, and “IOGP 434-1” references do
not address Pipe-in-Pipe piping or other types of double-containment piping (including vacuum-
jacketed piping).
A 2012 analysis by Cadwallader and Pinna of INL62 reviewed the available operating
experiences of double walled piping used in hazardous chemicals in different industries
(including consideration of pressurized, vacuum and purge configurations) as part of the design
considerations for the International Thermonuclear Experimental Reactor facility in France.
They also identified very little available failure rate information for double containment piping,
and recommended a Beta factor of 0.1 for a conceptual design level review:
Literature searches were performed to identify operating experiences that would indicate
failure rates of double containment piping. Despite the fact that double containment
systems have been widely used in the chemical industry since the 1990s, very little

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 112
 information was found in the literature. There is some qualitative information that
industrial systems overall have performed very well. Two additional cases are described
below. No data sets were found that could serve as a basis for a failure rate calculation.
Therefore, either modeling or analyst judgment is needed to determine the reliability.
Analyst judgment approaches vary in the risk literature from assuming independent
piping to a conservative common-cause type approach of a 0.5 multiplier on the failure
rate of single walled piping to account for the secondary containment pipe. …
A good approach to quantify the reliability of double containment piping where there can
be temperature, vibration, or other effects, is to apply a Beta factor to account for the
outer, non-independent pipe. The carrier pipe is given a multiplier of 0.01 to its leakage
failure rate to account for the second, proximate-location pipe of the same material.
Therefore, the external leak failure rate of the double containment pipe would be the
carrier pipe leak failure rate multiplied by 0.01. If the outer pipe is a different, less strong
material, then a Beta factor of 0.1 is recommended. This Beta approach was put forward
early in the ITER international project and remains a valid approach today. Certainly
some can argue that this approach is also conservative, that the outer pipe could function
better than the Beta factor suggests, especially in view of the opportunity for constant
monitoring of the pipe annulus. For early reliability studies on conceptual designs, this
Beta factor approach is recommended for its simplicity and speed to address the double-
walled piping issue.
For designs advanced past the conceptual design level, there will be enough design
information to support a detailed analysis. The two pipes can either be modeled as a
primary and standby component, as mentioned above, or an engineering assessment can
be performed. A rigorous finite element analysis can be performed to determine if any
common modes (pipe walls touching and transferring forces, vibration through spacers or
centering rings) are affecting both pipes. A corrosion assessment can be performed for
both pipes to determine if there is a high likelihood of corrosion pitting or breaches in
either pipe. The reliability analyst can use these analysis results to estimate the “leak
tightness” of the double containment system.
If the outer pipe is of a different material but provides as a passive independent layer of
protection (IPL) from the external environment, then the minimum applicable Probability of
Failure on Demand (PFD) could be considered as 1 x 10-2, based on Table 6.3 of Layer of
Protection Analysis: Simplified Process Risk Assessment (Edition 1) published by Center for
Chemical Process Safety (CCPS) and Wiley in 2001; if this IPL configuration is considered an
“inherently safe” design, then Table 6.3 of this reference identifies example PFDs from 1 x 10-2
to 1 x 10-6.

Comments re: Pipe-in-Pipe LNG Piping in “PNL ITRP ‘82” (non Vacuum-Jacketed)
PNL in its “PNL ITRP ‘82” analysis of LNG Import Terminals used a FTA to assess the benefit
of double pipe transfer line that “consists of two concentric stainless steel pipes with the inner
line containing the LNG and the annular space being purged with nitrogen” (ref. p. 5.4). PNL
estimated that “it would reduce the number of expected large releases from the transfer line from
1 x 10-3 per year to 2 x 10-7 per year. Failure of the expansion joints in the inner and outer lines
would be the primary cause of failure” (ref. pp. 5.4-5.5). Details are shown in Table B.13 and
Figure B.16 of “PNL ITRP ‘82” in Appendix I. This analysis was based on the outer pipe having
the same integrity as the inner pipe (i.e. 2.5 x 10-6 faults/hour probability).

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 113
In summary, PNL predicted a 5,000 reduction in failure rate (i.e. a 0.0002 multiplier) for double
containment piping, if the outer pipe has the same integrity as the inner pipe and there is a
pressure monitoring system.

Comments re: Vacuum-Jacketed or Vacuum-Insulated Pipe-in-Pipe Piping

No failure rate data was found for vacuum-jacketed (also known as vacuum-insulated, e.g. VJ or
VI) piping that is sometimes used in LNG facilties. The “HSE FRED JUN ‘12”, “RIVM BEVI
‘09”, “LNE ‘09”, and “IOGP 434-1” references do not address vacuum-jacketed piping.
For comparison, a predicted VJ pipe leak rate data of 1.8 x 10-12 failures/meter-hour (i.e. 1.6 x
10-8 failures/meter-year on continuous basis) was identified on p. 33 of the “INL VJ ‘10”
reference, but it was for 250 DN (10”) and 300 DN (12”) diameter Schedule 20 (6.3 mm wall
thickness) vacuum piping service (i.e. thinner than typical ANSI B31.3 process piping schedule
and thickness used at LNG facilities). In addition, this rate was for a leak (not rupture)
associated with small air leaks (to be overcome by the vacuum system) with no hole size
characterization, and so this leak rate prediction was not included in any results calculated from
the project database.
Comments re: Piping <2” and >40” in Diameter
Piping that is less than 2” does not need to be tracked by applicants, as per PHMSA’s answer to
question #DS2 (see Appendix A). In addition, PHMSA requires “a minimum 2-inch hole should
be considered at any location along any piping of 2 inches or larger diameter” as per the answer
to question #DS3. Removing the category for piping <2” from the FRT would help simplify the
FRT and clarify PHMSA’s communication about the FRT’s requirements with end-users.

The FRT currently does not specify failure rates for piping that is greater than 40”, although
some facility designers want to consider piping > 40” in diameter. While the piping size
categories in current FRT matches those in “HSE FRED JUN’12”, the average incident rates for
natural gas transmission piping >41” diameter (to consider one data set available for large
diameter piping in natural gas service) do not exceed and in general are significantly lower the
average incident rates of piping 23” ≤ d < 41”, based on EGIG’s 1970-2013 data (see Fig. 5 in
Section 4.9). Removing the 40” upper limit from the piping category from the FRT would help
clarify PHMSA’s communication about the consideration of large-diameter piping to end-users.
Summary:
 The single LNG-specific and cryogenic-specific piping leak rate data point that is
available (i.e. from “GRI LNG FRD ‘81”) aligns favorably with the piping failure rate
frequencies in the current FRT.
 The piping failure rate frequencies in the current FRT are identical to the failure rates in
Item #FR 1.3 “Pipework” on p. 47 of “HSE FRED JUN’12” (with only two small
exceptions), and the “HSE FRED “JUN’12” specification document is a recognized
international standard for failure rate frequencies. The basis of “HSE FRED JUN’12”
specifications includes consideration of more than 40 references in its underlying analysis
(a significant portion of which are confidential). PHMSA and FERC could consider
revising the two exceptions so that all of the piping failure frequencies in the FRT
directly match the failure rates in Item #FR 1.3 “Pipework” on p. 47 of “HSE FRED
JUN’12”.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 114
  The FRT currently provides a failure rate frequencies for piping < 2” in diameter and
there may be an opportunity to simplify the FRT and enhance its consistency with
PHMSA’s LNG Plant Requirements: Frequently Asked Questions webpage.
 The FRT currently provides no failure rate frequencies for piping > 40” in diameter, and
there may be an opportunity to simplify the FRT and enhance communication with end-
users.
 The FRT currently provides no specification for long transfer or inter-unit piping runs. It
appears reasonable to consider some reductions on an application-specific basis, but no
generalized reduction factors were identified for consideration for the FRT.
 The FRT currently provides no specification for pipe-in-pipe piping.
o INL’s 2012 analysis of all types of double containment systems summarized that:
 “The carrier pipe is given a multiplier of 0.01 to its leakage rate to account
for the second, proximate-location pipe of the same material.”
 “If the outer pipe is a different, less strong material, then a Beta factor of
0.1 is recommended.”
o “PNL ITRP ‘82” calculated a 0.0002 multiplier in its analysis in Table B.13 to
account for the second, proximate-location pipe of the same material using
nitrogen (not vacuum) in the annular space in LNG service and with a pressure
monitoring system.
o The CCPS reference Layer of Protection Analysis: Simplified Process Risk
Assessment in its Table 6.3 indicates a minimum 0.1 multiplier as an example for
any IPL, including “inherently safe” designs.
Recommendations:
PHMSA and FERC should consider retaining their current failure rate frequencies for piping,
with these modifications:
 PHMSA and FERC should consider evaluating if the requirement that “a minimum 2-inch
hole should be considered at any location along any piping of 2 inches or larger diameter”
should be maintained.
 PHMSA and FERC should consider eliminating the current category of d < 2” in the FRT,
if they intend to maintain the requirement that “a minimum 2-inch hole should be
considered at any location along any piping of 2 inches or larger diameter” and that piping
less than 2” in diameter “does not need to be tracked by applicants”.
 PHMSA and FERC should consider revising the current category of (20” ≤ d < 40”) to be
(20” ≤ d).
 PHMSA and FERC should consider applying modification factors to its failure frequency
rates for pipe-in-pipe piping as follows:
o Multiplier of 0.01 if outer and inner pipes are of the same material and schedule.
o Multiplier of 0.10 if outer and inner pipes are not of the same material and schedule,
and outer pipe is made of carbon steel.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 115
In addition:
 PHMSA and FERC may want to consider revising the two small exceptions to the failure
rates in Item #FR 1.3 “Pipework” in “HSE FRED JUN’12” so that all of the piping failure
rate frequencies in the FRT directly match those in HSE FRED JUN’12”.
 PHMSA and FERC may want to consider applying modification factors to inter-unit piping
between process units and to piping to/from storage or loading facilities (transfer pipe).

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 116
7. Gap Analysis of FRT and Failure Rate Data

7.1 Age of Available Relevant Failure Rate Data

Unfortunately much equipment failure rate data is dated, but by necessity continues to be used.
For example, a review by Beerens, Post, and Uijt de Haag of RIVM in 2006 noted that:
“The accuracy of the calculations is determined by the quality of the data used. To use QRA
methods for land-use planning, it is of great importance that the results are standardized
using reliable data. However, the set of available failure frequencies is nowadays not up-to-
date anymore, not only in the Netherlands but all over Europe and in fact worldwide. Thus,
there is a necessity for reviewing and updating the failure rate frequencies defined in the
guidebooks, like the Purple Book, for a number of standard scenarios and installations.”
and illustrated this dependency in Figure 1763 below (modified below in this report in blue font
to reflect that the “RIVM BEVI ‘09” superceded the most recent edition of the TNO Purple Book
“TNO PURPLE ‘05”).

“RIVM BEVI ‘09”

Figure 17: History of the Failure Frequencies used in The Netherlands for Pressure Vessels

7.2 Inability to See Some Underlying Data

The “HSE FRED JUN’12” is one primary failure rate reference document in use globally, but
many of the reference sources cited are confidential or internal reports. For example, 60-70% of
the citations identified in items 1.2.3, 1.2.4 and 1.3 (“Hoses and Couplings”, “Flanges and
Gaskets”, and “Pipework”) are marked as confidential or internal reports.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 117
7.3 Non-Uniform Nomenclature
There can be significant variations in the definitions, regimes and uncertainty of failure rate data,
as is well summarized by Dr. Sam Mannan in Appendix 14 of “LEES ‘12”. But as one
illustration of some of the variations in piping failure rate data, there is:
 Lack of uniformity in piping diameter size categories
 Lack of uniformity regarding hole size category definitions (including definition of
rupture), both in how the data is captured and how the data is sometimes applied (e.g. see
Table 2)
 Lack of uniformity in scope of components included with the pipe failure data, e.g.:
o “HSE FRED JUN’12” excludes valves and flanges, and differentiates process
Pipework from Above Ground Pipeline
o “RIVM BEVI 09” (ref. p. 42 of “Module C) “apply to the pipeline with
connections, such as flanges, welds and valves.” and makes “no distinction is
made between process pipes or transport pipes, the materials from which a
pipeline is made, the presence of cladding, the design pressure of a pipeline or its
location on a pipe bridge.”

7.4 Use of Predicted Rather than Observed Failure Rates

Many past and some recent estimates of LNG equipment failures are predictions based on FTA
or other techniques (e.g. for comparing types of atmospheric storage tanks or new types of LNG
equipment), because relatively few catastrophic failures or major leak incidents of LNG facilities
have occurred, and very limited cryogenic or LNG-specific failure rate data is available or has
been derived. The favorable safety record is of course a very good thing.

7.5 Very Limited Cryogenic or LNG Equipment Failure Rate Data

This review identified very little statistical leak or rupture failure rate data specifically for LNG
and cryogenic equipment. Certainly more publicly-available equipment leak or rupture failure rate
data that directly originates from the LNG and cryogenic industry is needed in order to help inform
regulatory agencies. Sufficient operating hours must obviously also be available to support the
derivation of failure rates, which in some cases are quite low.
The last major survey of equipment failure rate data specific to U.S. LNG plants was completed
in 1981 and a number of LNG plant design practices have changed since (e.g. use of in-tank vs.
external pumps, and limited applications of expansion joints). It was “the opinion of GRI that
the failure rate data base should be periodically updated, at about 5-year intervals” as stated in
“GRI FRD ‘81”, but unfortunately no additional industry or federal funding was made available
for this purpose since the 1981 analysis was completed.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 118
7.6 Relevancy of Available Failure Rate Data to LNG
This analysis confirmed that much of the generic failure rate data in the FRT is derived from
non-cryogenic and non-LNG facilities. Factors sometimes cited by those who favor specifying
failure rate frequencies for LNG equipment and piping different from generic failure rate data
can include that LNG facilities:
 are highly-protected and typically less likely to experience external interference, since all
LNG plant piping is essentially “inside the fence” with well-documented pipe routing
drawings, trained contractors and typically few major construction projects underway.
 are typically located on well-positioned-and-graded sites with above-ground piping, with
design consideration for 100-year and other extreme weather events, so that ground
movement issues are less likely to occur at LNG facilities.
 are serviced by well-trained personnel with established operating procedures who are
unlikely to make hot tap errors or other incidents in comparison to exposed natural gas
pipelines.
 typically contain a significant amount of stainless steel piping and vacuum-jacketed
piping, and thus less susceptible to the general galvanic corrosion that can dominate the
cause of pinhole/crack data in carbon steel piping.
 are sometimes located inland, in comparison to failure rate data derived from off-shore
oil and gas operations in marine salt spray environment (all LNG terminals are of course
located in a marine salt spray environment).
 can contain inert fluids in their piping and equipment, in comparison to more severe
service conditions at some oil and gas or chemical process industry sites.
 may have fewer flanges per unit length of piping than offshore platforms (that can
average 7 m (23 ft) of pipe for each flange64), and thus less susceptibility to failures
arising from from weldment issues, insulation joint corrosion initiation points, etc.
and opposing factors can include:
 may have greater condensation potential than ambient temperature piping, which could
increase galvanic corrosion or Stress Corrosion Cracking.
 may have Stress Corrosion Cracking of stainless steel piping if exposed to sea water.
 may have shorter piping run lengths and more welds/spool length than natural gas or
hazardous liquid transmission pipelines, which could yield higher failure rates/meter-year
in comparison.
 may have less onerous regulatory requirements for UK onshore vs. offshore loactions, as
noted by DNV on p. 6 of “DNV FFG HCRD ‘13”.
Nevertheless FERC’s and PHMSA’s consideration of the failure rate sources referenced in their
derivation of the FRT, along with those considered in this report, appears to be consistent with
other leading global authorities, since these sources appear to comprise the best available data:
 The UK HSE’s “HSE FRED JUN’12” specification document does not make any special
exceptions for LNG or other cryogenic facilities.
o Item FR 1.1.2.1 specifically defines failure rates for “LNG Refrigerated Vessels”.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 119
 o Item FR 1.3 defines failure rates for “Pipework” and cites 44 references, which
include one confidential analysis of pipework associated with bulk storage of
liquid oxygen and two gas terminal studies.
o Item FR 1.1.3 defines failure rates for “Pressure Vessels” and in describing their
derivation note 44 implies that cryogenic pressure vessels should have a higher
failure rate:
The cold catastrophic and hole failure rates are taken from the MHAU
handbook (now archived). These are derived in the Chlorine Siting Policy
Colloquium and are applicable to chlorine pressure vessels in a typical
water treatment plant. Although they are not applicable to all types of
pressure vessels the values are a good starting point when trying to derive
failure rates for vessels in other applications. The value chosen for
catastrophic failure should normally be 2 chances per million (cpm),
assuming that the vessel conforms to BS5500 or an equivalent standard
and that there is good compliance with the HSW etc. act (1974), unless
there are site-specific factors indicating that a higher rate is appropriate
(e.g. semi refrigerated vessels [cryogenic pressure vessels]).
 IOGP’s comments regarding the applicability of its recommendations include this excerpt
from p. 19 of “IOGP 434-1”: “We therefore recommend use of the same frequencies for
LNG installations as given in Section 2.0” [i.e. for other oil & gas facilities]. “A 50%
reduction could be considered as a sensitivity but decisions based on this would need to
be fully justified”.
 ISO Technical Specification ISO/TS1690165 Guidance on performing risk assessment in
the design of onshore LNG installations including the ship/shore interface in its section
A.3 lists the many of the same references considered in this analysis and concludes “It
should be noted that there are no publicly available incident databases for LNG plants
that can be available to derive leak frequencies and therefore should rely on the above
more general data.”
 The Dutch government in “RIVM BEVI ‘09” document stipulates on page 3 of its
Introduction that the “SAFETI-NL” calculation package by DNV London is “stipulated
for carrying out the QRA calculations for establishments that fall under the Bevi”.
“SAFETI-NL” contains equipment failure rate data based apparently in large part on an
analysis of the UK HCRD.
 EIGA in its guidance document 60/1566 to apply SEVESO Directives to cryogenic
industrial gas facilities in Europe highlights on p. 21 that a standardized version of
DNV’s PHAST/Risk software, named “SAFETI-NL” must be used in the Netherlands.
 Singapore requires the use of “HSE FRED JUN’12” for Fixed Installations in its QRA
Technical Guidance Manual issued August 2016.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 120
7.7 Gaps in Current FRT Categories
Potential gaps in the FRT were identified as part of this project. The project team considered a
large number of new potential component categories (as summarized in Appendix C), and
developed the recommendations in Section 6. Gaps that are proposed to be filled by the
recommendations from this research are:
 Definition of valve leak rate hole sizes
 Definition of gasket leak rate hole sizes
 Inclusion of double ply expansion joint and clarification of single ply expansion joint
 Inclusion of pipe-in-pipe piping
Significant gaps in available failure rate data exist in almost every category of the current FRT.
Those categories with some of the largest apparant gaps include:
 Cryogenic/LNG piping
 Cryogenic/LNG valves, expansion joints and gaskets
 Cryogenic/LNG transfer arms and hoses
 Cryogenic/LNG atmospheric storage containers
 Cryogenic/LNG pressure vessels

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 121
8. Summary and Recommendations

8.1 Summary and Recommendations Regarding FRT Criterion

A summary of the review in Section 2 of the FRT Criterion of 3 x 10-5 failures per year is:
 The project team confirmed that failure frequency estimates of 3 x 10-5 to 5 x 10-5
associated with the tank liquid outlet line are contained in the historical references, and
its observations in Section 2.3 also included that:
o The 3 x 10-5 failures per year criterion refers to an estimated failure rate for an
equipment component failure, and not the resulting level of risk to society or an
individual as is analyzed in some other risk assessment methodologies.
o The criterion appears to be derived from the failure rate associated with the
failure of one of the most critical plant components, i.e. the LNG container liquid
withdrawal line. The failure rate criteria associated with the failure of line
connection penetrations to the primary container may or may not set an
equivalent risk level required for all other equipment component failure rates.
o The 3 x 10-5 failures per year criterion appears to be associated with the failure
rate of an LNG container liquid withdrawal line isolated with an internal shutoff
valve, while US 49 CFR Part 193 also permits evaluation of events with other
consequences such as the nearly complete emptying of the LNG container.
o While for example the failure rates in the FRT for process piping exactly match
HSE’s FRED document (with two small exceptions), the FRT’s threshold criteria
methodology appears to be a somewhat unique risk assessment methodology.
o A risk assessment prepared in 1975 by Science Applications Inc. for the Western
LNG Terminal Company for a LNG terminal proposed for Oxnard, CA also
prepared a number of FTAs and estimated the probability of a “Leak Occurs in
Storage Tank or Tank Outlet” as 1 x 10-6 failures/year and a “Leak Occurs in
Outlet of Tank” as 9.8 x 10-11 failures/year (when the isolation systems were
considered).
 No direct analysis was made to compare the current FRT criterion of 3 x 10-5 to other
regulated risk criteria in the U.S., since such an analysis of the risks to individuals and
society of applying the FRT’s methodology and criterion to vaporization, process, or
LNG transfer components versus the risks associated with failure of a primary LNG
container outlet line was beyond the scope of this analysis. This analysis may include a
comparison of a 1-hour spill versus a complete loss of containment.
Recommendations arising from this review pertaining to the FRT Criterion are:
 No direct basis was identified to propose revising current the FRT Criterion of 3 x 10-5, for
the reason identified in the prior bullet above. Therefore it is recommended to retain the
FRT Criterion of 3 x 10-5.
 FERC and PHMSA should further consider whether methodologies other than the current
FRT criterion methodology would be suitable for defining a SALS for impoundment
areas serving only vaporization, process, or LNG transfer areas. Others have also
recommended reviews of risk methodology assessment practices for LNG facilities in the

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 122
 U.S., including at the federal level.67,68

8.2 Summary and Recommendations Regarding FRT Failure Rates and

Categories
Appendix B concisely summarizes the specific recommendations for the FRT developed in this
research for FERC’s and PHMSA’s consideration. Detailed recommendations are provided at the
ends of Sections 6.2 – 6.9. A summary of key findings and recommendations related to the FRT’s
failure rates and its categories is:
 The leak and guillotine rupture failure rates specified for piping in the FRT directly
match those in HSE’s FRED guidance document (with only two small exceptions). The
guillotine rupture failure rates specified for piping in the FRT are slightly less than, or
greater than, the guillotine rupture failure rates specified in “RIVM BEVI ‘09”
(depending on pipe diameter), and also appear reasonable with compared to the actual
rupture rates observed in PHMSA’s onshore natural gas and hazardous liquid pipeline
database (2010-2015). The single actual data point for LNG piping leak rates identified
in this analysis (from the 1981 survey of LNG peak shaving plants) also matches up
favorably with the FRT’s current specifications for piping. It is recommended to
consider retaining the current piping failure rate specifications in the FRT, but to also
consider eliminating the smallest piping size category (d < 2”) and the current 40”D
maximum limit of the largest piping size category (in order to enhance clarity to end-
users of the FRT).
 The FRT currently does not address pipe-in-pipe piping such as the vacuum-jacketed
piping that is sometimes used in LNG facilities. It is recommended to consider
specifying failure rate reduction factors of 0.01 and 0.1 to modify piping failure rates
when one of either two different types of pipe-in-piping is used. These reduction factors
are based on a review of a 1982 analyses by Pacific Northwest Laboratory (PNL), a 2012
analysis of double containment piping by Idaho National Laboratory, and a relevant
Center for Chemical Process Safety reference.
 The FRT currently does not address potential reductions in piping failure rates for long
transfer lines or inter-unit piping as commonly exist in LNG terminals. Some relevant
information was identified for PHMSA’s and FERC’s review and future consideration on
a generalized basis, beyond those engineering analyses submitted for FERC’s and
PHMSA’s consideration on a case-by-case project application basis.
 The leak and guillotine rupture failure rates for LNG transfer hoses and arms specified in
the FRT were compared to other references in this study, and also to results of recent
tests in Europe and other developments. There are significant technology developments
underway on LNG transfer hoses and arms for both ships and trucks (trailers and trailer-
mounted ISO containers). This equipment category was identified as an important watch
area for PHMSA and FERC to seek new failure rate data in the future, because for
example results of crush and fatigue tests on cryogenic hoses in Europe in 2016 indicates
that the truck hose leak rate specified in the FRT may be about 100 times too
conservative, and also that a guillotine rupture of a truck transfer hose may not be a
credible event. However, at the present time it is recommended that: 1) if FERC and
PHMSA want to retain the current basis of “Failures per year of operation”, then they

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 123
 may consider making no changes to the FRT’s specification; and 2) that PHMSA and
FERC review this new cryogenic test data and consider removing the rupture frequency
specifically for multi-composite hoses. If they want to consider changing the FRT’s basis
of specification from “Failures per year of operation” to “Failures per hour of operation”,
then they could consider applying the failure frequency rates specified by Dutch and
Flemish regulators. Sections 6.4 and 6.5 provide more details.
 The failure rate for valves specified in the FRT was compared to other references in this
study. It is recommended to retain the FRT’s currently-specified rupture rate for valves,
but to also add a leak rate consistent with that specified in HSE’s FRED guidance Item
FR 1.2.1 for valves, using the 2 mm hole size considered in Section 4.6 of this report.
 The failure rate for gaskets (i.e. flanges) specified in the FRT was compared to other
references in this study. It is recommended to consider eliminating the “Failure of
Gasket” terminology and specify failure rates consistent with those specified in HSE’s
FRED guidance item FR 1.2.4 for gaskets (and flanges), using the 25 mm and 50 mm
hole sizes considered in Sections 4.8 of this report.
 The failure rate for expansion joints specified in the FRT was compared to other
references in this study and to underlying references. It is recommended to consider
revising the specified failure rate from 4 x 10-3 to 1 x 10-4 failures/year in order to better
align it with the 1975 risk analysis performed by Science Applications Inc. (excerpts
shown in Appendix I) and to also clarify that it applies to single ply expansion joint; both
the existing and recommended rates are greater than the 3 x 10-5 FRT threshold criterion.
In addition, it is recommended to specify a rupture failure rate for double ply expansion
joints.
 The guillotine rupture failure rate specified for “Process Vessels, Distillation Columns,
Heat Exchangers and Condensers” in the FRT aligns well with the rate specified for
Pressure Vessels in HSE’s FRED guidance document and also with the rates specified by
Dutch and Flemish regulators for Process Vessels. The 10 mm hole leak failure rate
specified in the FRT also aligns well with the 10 mm hole leak failure rates specified by
Dutch and Flemish regulators for process vessels, and is more conservative than than the
10 mm hole leak failure rates specified for Pressure Vessels in HSE’s FRED guidance
document. This research also reviewed and summarized the lower failure rates assigned
to Pressure (Storage) Vessels (vs. Process Vessels) by Dutch and Flemish regulators than
the rates specified by the UK HSE’s FRED guidance for document Pressure (Storage)
Vessels, but it was recommended that FRT retain its more conservative basis and not
specify lower failure rates specifically for Pressure (Storage) Vessels as different than
those rates for Process Vessels. This research also explored, accumulated and analyzed
for PHMSA and FERC a number of potential equipment subcategories within this overall
category. In summary it was recommended that the FRT retain this existing category
name and its currently-specified rupture and leak failure rates.
 The FRT currently specifies leak and rupture rates for cryogenic atmospheric storage
tanks. For determination of single accidental leakage sources for process facilities this
information does not appear to be relevant because the Design Spill for LNG tanks (i.e.
containers in NFPA 59A) is already specified in Table 2.2.3.5 of the NFPA 59A (2001
edition) specified by US 49 CFR Part 193. However, if DOT allows for design spills to
be selected using a failure rate or risk based approach (e.g., NFPA 59A 2016 edition)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 124
 then this information may be relevant. Therefore, it is recommended that PHMSA
consider eliminating this category or clarify the use of this information for single
accidental leakage sources in impounding areas serving only vaporization, process or
LNG transfer areas.

8.3 Recommendations Regarding Subsequent Research and Related Efforts

Recommendations regarding subsequent research and related efforts are:
 PHMSA and FERC should consider funding research to conduct a new survey of LNG
facilites in the US, in order to update and expand upon the most recent survey study
completed in 1981 (“GRI FRD ‘81”). The U.S. is in a unique position since many LNG
facilities are older facilities built in the 1960s and 1970s. A new survey can potentially
leverage this historical basis that represents a significant number of operational hours. In
addition, many of the U.S. LNG facilities are in peak shaving service and utilize smaller
piping and components than the many international LNG facilities in export/import
service; the leak and failure rate data that can be generated or estimated from these
facilities would be especially relevant to peak shaving plants (both new or upgrades),
bunkering facilties that may fall under FERC’s and PHMSA’s jurisdictions, in addition to
the many export terminal applications that FERC is currently reviewing.
 PHMSA and FERC should consider supporting the coordination of any new industry-
government consortium efforts to create a national database of information related to the
in-service performance of LNG piping and components. This effort can build upon the
experience of the Plastic Pipe Database Consortium (PPDC), but should incorporate
appropriately-defined nomenclature and should report both incidents and total
populations as well as age, type and other relevant details defined by the consortium.

9. Summary of Project Final Financial Contributions

The financial contributions to the project were consistent with contract #DTPH56-15-T-00008,
and no discrepancies or variances in contributions occurred.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 125
Appendix A: PHMSA LNG Plant Requirements FAQ - Design Spill Determination

The information that PHMSA provides related to (DS) Design Spill Determination on its
Frequently Asked Questions webpage (http://primis.phmsa.dot.gov/lng/faqs.htm) is copied below
(valid as of the date of this report and last revised on 12/7/15). That webpage also provides other
FAQs available in the categories of: (G) General; (D) Design; and (H) Hazards and Hazards
Modeling.

DS1. PHMSA reviews the design criteria for design spills on a case-by-case basis to determine
compliance with Part 193. What information is required to assist PHMSA in its determination of
the design spill criteria acceptable for use?

Applicants must provide a piping and equipment inventory table of LNG plant components in hazardous or
flammable fluid service. The piping and equipment inventory table should be submitted in Excel (*.XL*)
format. Separate tabs or lists should be used for each type of hazardous fluid, as well as a separate tab or
list to present all of the final design spill selections.

The table should include the following information:

a. Line segment or component number to identify potential design spill;

b. Hazardous fluid service (LNG, natural gas, refrigerants (such as ammonia, propane, ethane, mixed refrigerant),
natural gas liquids or gas condensate, hydrogen sulfide, benzene, etc.) for each component;
c. General plant area or service (e.g. liquefaction train, refrigerant storage, marine area, etc.), unless the entire
project is confined to one area;
d. Unit plot plan drawing number reference(s) for each component;
e. Beginning point location (e.g., exchanger outlet flange) for each line;
f. Ending point location (e.g., pump suction nozzle) for each line;
g. P&IDs and drawing number reference(s) for each component;
h. Piping line designation or equipment tag number on P&ID;
i. Pipe diameter or pipe size, volume of container, or size of equipment;
j. Length of piping (feet and meters); or number of components (each);
k. Maximum connection diameter in the piping segment;
l. Failure type or mode selected from the failure rate table;
m. Corresponding nominal failure rates per meter or unit;
n. Calculated failure rate based on pipe length or number of units and failure rates per meter or unit listed in the
failure rate table;
o. Comparison of calculated failure rate to a failure rate criterion of 3x10-5 failures per year;
p. Process or storage conditions (e.g., fluid phase (liquid or vapor); density (lb/ft3); pressure (psig); temperature
(°F); flow rate, (lb/hr); composition of mixed refrigerants, NGL/Condensates, acid gas);
q. Process flow diagram and corresponding heat and material balance stream number;
r. Heat and material design case (e.g., rich, lean, average, etc.);
s. Calculated equivalent hole size based on failure modes listed in the failure rate table; and
t. Calculated design spill flow rates.
u. Design spills selected with release duration, de-inventory duration, height, direction, orientation, rainout
percentage, flashing and jetting vapor mass flow rate, pool vaporization mass flow rate, and total vapor mass flow
rate

DS2. What sorts of pipe, equipment, and containers should be included on the piping and
equipment inventory table?

Components that should be considered to fail in the analysis for determination of the single accidental
leakage source are those containing hazardous or flammable fluids and are listed on the Failure Rate Table.
The table must include pipe of 2 inch diameter and larger size, valves, gaskets, expansion joints, truck
transfer hoses, truck transfer arms, ship transfer arms, pumps, compressors, process vessels, columns, heat
exchangers, condensers, and storage tanks.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 126
DS3. 49 CFR 193 requires that design spills for an LNG plant be selected according to NFPA 59A-
2001 Paragraph 2.2.3.5. NFPA 59A requires the evaluation of accidental flow from “any single
accidental leakage source” (SALS) but does not define this term. How should I select SALS
events in an LNG plant?

For piping and equipment that handle LNG, flammable refrigerants, toxic components, or any other
hazardous fluid, release sources may be chosen using the following guidelines. The SALS selection
methodology is applied to determine the maximum hole sizes of interest for the most significant releases of
each hazardous fluid in each portion of the LNG plant. (The SALS selections are one component of design
spill definition. Please refer to other FAQs for additional design spill definition topics.)

A. For all piping and equipment (including transfer hoses and arms), the failure rate table should be applied to
determine if the 3 x 10-5 per year failure rate criterion is equaled or exceeded; and,
B. For all piping, the failure rate table is applied to a piping segment (i.e., length of pipe) and the hole size is chosen
based on equaling or exceeding the 3 x 10-5 per year failure rate criterion. The following rules should be applied
to piping segments:
1. Piping segments should be selected to begin and end at pieces of equipment and include all tees, loops,
and branches; and
2. A principle to be applied for piping segment begin/end locations is at points where the process conditions
change significantly (typically temperature, pressure, or composition); and
3. Piping segments should not be initiated or terminated at valves (e.g. pressure regulating valves, flow
control valves, etc.), piping spec changes, flange connections, flow meters, reducers, piping fittings, or
other piping appurtenances; and
4. Piping segment length should be based on piping isometrics but may be based on engineering estimates
of the piping path, accounting for vertical distances as well as horizontal distances.
C. For piping connections less than 6 inches in diameter, a full-bore rupture (guillotine failure) is assumed at the
point of connection to the equipment item or piping. Small diameter piping is typically used for connections to
equipment (e.g., storage tanks, vessels, heat exchangers, pumps, etc.) or to piping as drain lines, vent lines,
nitrogen purge lines, PSV connections, valve bypass connections, instrument connections, etc.
D. Regardless of the results obtained by the failure rate table or connections approach above, a minimum 2-inch hole
should be considered at any location along any piping of 2 inches or larger diameter.

DS4. Can a fractional time of use be applied when determining SALS events with a probabilistic
spill selection methodology?

When determining the design spill through the use of a probabilistic spill selection method, a time-of-use
factor may be applied to some piping or equipment groups based on their expected use. For groups that
have a fractional time of use other than 1.0 (100% use), the applicant must be able to demonstrate that the
group can be isolated, purged, instrumented, maintained, and continually documented in such a way that
the fractional time of use is traceable. In groups that operate continually but in more than one mode, each
mode must be considered as a potential design spill source and the sum of the fractional times of use in
each mode must equal 1.0 (100% use). An example of this would be a loading/unloading line that moves a
high mass rate when loading or unloading LNG, but moves a much smaller mass rate when recirculating to
keep the line cooled down.

DS5. Is the largest size hole always used as the hole size in release modeling?

Not necessarily. For any defined maximum hole size, the applicant must demonstrate that that the hole size
selected produces the greatest vapor dispersion distance when accounting for the mechanisms of jetting,
flashing, aerosol formation, and rain-out. If a smaller hole size creates a larger vapor dispersion hazard
distance, that smaller hole size should be used to define the design spill event. This applies to all single
accidental leakage sources, including failures at piping, piping connections, and all other equipment (e.g.,
transfer hoses, vessels, heat exchangers, pumps, valves, flanges, etc.).

DS6. What are the proper release height and orientation to use for a design spill?

For each design spill identified, release height and orientation should be selected to define the largest vapor
dispersion hazard distance while properly characterizing the release scenario. If an applicant can show that
only certain release orientations are possible based on the piping connections and direction of the piping
(e.g., vertically upward for connections to relief valve inlet piping, vertically downward for gravity drain

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 127
connections, and downward for shrouded piping) then a specific orientation may be used in the modeling. All
other piping and equipment failures must consider all horizontal and vertical directional orientations.

DS7. How are release locations defined?

For connection failures, the release location can be identified at the specific point of connection in the LNG
plant. For piping segments to which the failure rate table has been applied, the selected hole can occur at
any location along the piping segment. If vapor barriers, shrouds, or pipe-in-pipe designs are used to reduce
the vapor dispersion distance, locations potentially not impacted by the vapor barrier, shroud, or pipe-in-
pipe should also be selected.

DS8. How do I determine the process conditions to evaluate for hazard modeling?

Process conditions should be based on heat and material balance modes of operation and design cases (e.g.,
rich, lean, average, etc.) that produce the worst case dispersion results from flashing and jetting and liquid
releases. The leakage sources from branch connections should be considered using the potential operational
conditions along the pipe as well as the potential operational conditions that could be experienced at or near
the branch pipe connection to a main process line. In cases that would reduce the back pressure on pump(s)
or compressor(s), the flow rates should consider the potential increased pump or compressor flow
determined by the pump and compressor curve(s) as detailed in D10 and also consider the decrease in
temperature from depressurization.

DS9. Is it permissible to use spill duration of less than 10 minutes for design spill calculations?

For design spills other than those from LNG containers, the event is defined in NFPA 59A to last “for 10
minutes or a shorter time based on demonstrable surveillance and shutdown provisions acceptable to the
authority having jurisdiction.” Demonstrable surveillance and shutdown should include the time required to
detect that the spill is occurring, the time to alert operators to this condition, the time required for operators
to take action, and the time for the system to fully respond to the shutdown action that is initiated, including
any valve closure times. In this case, the applicant should provide a detailed justification or demonstrate
that a maximum or steady state dispersion distance has already been reached by the shorter time.

In cases where a system may deplete its inventory in less than 10 minutes, a release duration of less than
10 minutes may be used. In this case, the applicant should provide a detailed justification or demonstrate
that a maximum or steady state dispersion distance has already been reached by the shorter time.

DS10. What considerations should be given to system inventory in the design spill definition?

The release modeling should account for the available system inventory (including pipework, process
vessels, and other process equipment), the normal flow of fluid into the system, and the demonstrable
surveillance and shutdown provisions that may apply. The release modeling should also continue (even
beyond 10 minutes) until the available system inventory is depleted; available system inventory may be
modified during the event by valve closures. For systems that rely on isolation by emergency shutdown
valves, the valves must be protected from failure, including fire and external impacts.

If the event the duration would potentially be greater than 10 minutes, release and dispersion modeling
should continue after 10 minutes unless a release is demonstrated to reach its furthest vapor dispersion
extent within 10 minutes.

DS11. Do I need to consider pump run-out in release scenario calculations?

Yes. Applicants should use pump run-out (greater flow than in normal pump flow operations) in failure
calculations if the pump design allows increases in flow as the discharge pressure is reduced. Pump run-out
parameters are presented by the pump manufacturer as a pump curve that shows flow increasing as the
discharge pressure decreases. If pump run-out flows are not known at the time of submittal, engineering
estimates may be employed provisionally.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 128
DS12. Should multiple pumps be considered when calculating the greatest flow from a spill to
size impoundments?

Where the greatest flow is potentially fed from multiple pumps, calculate the flow assuming that all pumps
are running at possible pump run-out conditions, unless a mechanical interlock or passive preventive measure
is installed that prevents all pumps from running concurrently.
___.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 129
 Appendix B: Potential Revisions to the FRT Considered in this Research

CURRENT RECOMMENDATIONS
Type of Failure Nominal Failure Rate Nominal Failure Rate
Retain Cryogenic Storage Tanks (General) Failures per year of operation Failures per year of operation
Retain Rupture of Storage Tank Outlet/Withdrawal Line 3E-5 (Failure Rate Criterion) 3E-5 (Failure Rate Criterion)

Delete Single Containment Atmospheric Storage Tanks Failures per year of operation
Delete Catastrophic Failure, Release to Atmosphere 5.0E-06
Delete Catastrophic Failure of Tank Roof 1.0E-04
Delete Release from a hole in inner tank with eff. diam. of 1 m (~3ft) 8.0E-05
Delete Release from a hole in inner tank with eff. diam. of .3 m (~1ft) 2.0E-04
Delete Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in) 1.0E-04
Delete Double Containment Atmospheric Storage Tanks Failures per year of operation
Delete Catastrophic Failure, Release to Atmosphere 5.0E-07
Delete Catastrophic Failure of Tank Roof 1.0E-04
Delete Release from a hole in inner tank with eff. diam. of 1 m (~3ft) 1.0E-05
Delete Release from a hole in inner tank with eff. diam. of .3 m (~1ft) 3.0E-05
Delete Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in) 1.0E-04
Delete Full Containment Atmospheric Storage Tanks Failures per year of operation
Delete Catastrophic Failure, Release to Atmosphere 1.0E-08
Delete Catastrophic Failure of Tank Roof 4.0E-05
Delete Release from a hole in inner tank with eff. diam. of 1 m (~3ft) 1.0E-06
Delete Release from a hole in inner tank with eff. diam. of .3 m (~1ft) 3.0E-06
Delete Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in) 1.0E-04

Retain Condensers Failures per year of operation Failures per year of operation
Retain Catastrophic Failure (Rupture) 5.0E-06 5.0E-06
Retain Release from a hole with eff. diam. of 0.01m (0.4 in) 1.0E-04 1.0E-04

Retain Truck Transfer Failures per year of operation Failures per year of operation
Retain Rupture of transfer arm 3.0E-04 3.0E-04
Retain Release from hole in transfer arm with effective diameter of 10% 3.0E-03 3.0E-03
transfer arm diameter with maximum of 50mm (2-inches)
Retain Rupture of transfer hose 4.0E-02 4.0E-02 *
Retain Release from hole in transfer hose with effective diameter of 10% 4.0E-01 4.0E-01
transfer hose diameter with maximum of 50mm (2-inches)

Retain Ship Transfer Failures per year of operation Failures per year of operation
Retain Rupture of transfer arm 2.0E-05 2.0E-05
Retain Release from hole in transfer arm with effective diameter of 10% 2.0E-04 2.0E-04
transfer arm diameter with maximum of 50mm (2-inches)

* Note: It is recommended that PHMSA and FERC consider if the current rupture rate for truck transfer hoses is still appropriate
based on the results on the recent cryogenic hose tests by TNO, and perhaps after consultation with RIVM and others. For example,
PHMSA and FERC could consider "Rupture of transfer hose unless an applicant’s proposed hose is a multi-composite hose such as
either Gutteling B.V.’s Multi-LNG White hose or another multi-composite hose of similar materials, design and manufacture."

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 130
 Appendix B: Potential Revisions to the FRT Considered in this Research

CURRENT RECOMMENDATIONS
Type of Failure Nominal Failure Rate Nominal Failure Rate
Retain Piping (General) Failures per year of operation Failures per year of operation
Delete Rupture at Valve 9.0E-06
Revise Rupture of Valve 9.0E-06
New Release from valve with a hole with effective diameter of 2mm (0.08 2.0E-04
inch)
Delete Rupture of Expansion Joint 4.0E-03
Revise Rupture of Single Ply Expansion Joint 1.0E-04
Revise Rupture of Double Ply Expansion Joint 1.0E-08
Delete Failure of Gasket 3.0E-02
New Release from hole in a Spiral Wound Gasket with effective diameter of 1.0E-07
50mm (2-inches)
New Release from hole in One Segment of Gasket with effective diameter of 5.0E-06
25mm (1-inch)

Delete Piping: d < 50mm (2-inch) Failures per year of operation

Delete Catastrophic Rupture 1.0E-06
Delete Release from hole with effective diameter of 25mm (1 inch) 5.0E-06

Retain Piping: 50mm (2-inch) ≤ d < 149mm (6-inch) Failures per year of operation Failures per year of operation
Retain Catastrophic Rupture 5.0E-07 5.0E-07
Retain Release from hole with effective diameter of 25mm (1 inch) 2.0E-06 2.0E-06

Retain Piping: 150mm (6-inch) ≤ d < 299mm (12-inch) Failures per year of operation Failures per year of operation
Retain Catastrophic Rupture 2.0E-07 2.0E-07
Retain Release from hole with effective diameter of 1/3 diameter 4.0E-07 4.0E-07
Retain Release from hole with effective diameter of 10% diameter, up to 50mm Failures per year of operation Failures per year of operation
(2-inches)
Retain Release from hole with effective diameter of 25mm (1 inch) 7.0E-07 7.0E-07

Retain Piping: 300mm (12-inch) <≤ d < 499mm (20-inch) Failures per year of operation Failures per year of operation
Retain Catastrophic Rupture 7.0E-08 7.0E-08
Retain Release from hole with effective diameter of 1/3 diameter 2.0E-07 2.0E-07
Retain Release from hole with effective diameter of 10% diameter, up to 50mm 4.0E-07 4.0E-07
(2-inches)
Retain Release from hole with effective diameter of 25mm (1 inch) 5.0E-07 5.0E-07

Revise Piping: 500mm (20-inch) ≤ d < 1000 mm (40-inch) Failures per year of operation
New Piping: 500mm (20-inch) ≤ d Failures per year of operation
Retain Catastrophic Rupture 2.0E-08 2.0E-08
Retain Release from hole with effective diameter of 1/3 diameter 1.0E-07 1.0E-07
Retain Release from hole with effective diameter of 10% diameter, up to 50mm 2.0E-07 2.0E-07
(2-inches)
Retain Release from hole with effective diameter of 25mm (1 inch) 4.0E-07 4.0E-07

New Pipe-in-Pipe Failure Frequency Modification Multipliers Failure Frequency Multiplier

New Piping failure rate frequency multiplier for Pipe-in-Pipe systems if 1.0E-02
outer and inner pipes are of the same material and schedule
New Piping failure rate frequency multiplier for Pipe-in-Pipe systems if 1.0E-01
outer and inner pipes are not of the same material and schedule, and
outer pipe is made of carbon steel

Note: Italicized and struck-through text highlights for the reader those changes to the FRT that were considered in this research.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 131
 Appendix C: Equipment Protocol Framework and Applied References

Potential FRT SubCategory1
Potential FRT SubCategory2

Potential FRT SubCategory4
Super Category Comparison

Potential FRT Subcategory3
Current PHMSA FRT Spec

Current FRT Category
Index Reference Source Applied
1 Ambient Atm. Storage Tanks
2 API_581_16, CCPS_89, INL_CHEM_95, INL_NUC_07, HSE_FRED_JUN_12 Catastrophic Failure, Release to Atmosphere
3 HSE_FRED_JUN_12 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
4 HSE_FRED_JUN_12 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
5 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
6 Refrigerated Atm. Storage Tanks (Typ. Single Shell)
7 CCPS_89, LNE_09, TNO_PURPLE_05 Catastrophic Failure, Release to Atmosphere
8 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
9 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
10 LNE_09, TNO_PURPLE_05 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
11 Cryogenic Atm. Storage Tanks
12 CCPS_89 Catastrophic Failure, Release to Atmosphere
311 Catastrophic Failure of Tank Roof
13 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
14 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
15 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
16 Single Cont. Cryo. Atm. Storage Tanks
59A_16, CCPS_89, IOGP_434_3, LNE_09, PNL_PSRP_82, RIVM_BEVI_09, TGC_03, 
17 Yes TNO_PURPLE_05, HSE_FRED_JUN_12 Catastrophic Failure, Release to Atmosphere
312 Yes HSE_FRED_JUN_12 Catastrophic Failure of Tank Roof
18 Yes HSE_FRED_JUN_12 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
19 Yes HSE_FRED_JUN_12 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
20 Yes LNE_09, RIVM_BEVI_09 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
21 Single Cont. Cryo. Atm. Storage Tanks ‐ HC Only (LNG, Ethane, Ethylene) ‐ Self‐Supp. Inner Tan
22 CCPS_89 Catastrophic Failure, Release to Atmosphere
313 Catastrophic Failure of Tank Roof
23 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
24 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
25 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
26 Single Cont. Cryo. Atm. Storage Tanks ‐ LIN/LOX/LAR ‐ Self‐Supp. Inner Tank
27 HSE_FRED_JUN_12 Catastrophic Failure, Release to Atmosphere
314 Catastrophic Failure of Tank Roof
28 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
29 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
30 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
31 Double Cont. Cryo. Atm. Storage Tanks
32 Yes 59A_16, IOGP_434_3, LNE_09, RIVM_BEVI_09, TGC_03, TNO_PURPLE_05, HSE_FRED_JUN_12 Catastrophic Failure, Release to Atmosphere
315 Yes HSE_FRED_JUN_12 Catastrophic Failure of Tank Roof
33 Yes HSE_FRED_JUN_12 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
34 Yes HSE_FRED_JUN_12 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
35 Yes LNE_09, RIVM_BEVI_09 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
36 Full Cont. Cryo. Atm. Storage Tanks
37 Yes 59A_16, IOGP_434_3, KGSC_06, LNE_09, RIVM_BEVI_09, TNO_PURPLE_05, HSE_FRED_JUN_12 Catastrophic Failure, Release to Atmosphere
316 Yes HSE_FRED_JUN_12 Catastrophic Failure of Tank Roof
38 Yes HSE_FRED_JUN_12 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
39 Yes HSE_FRED_JUN_12 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
40 Yes LNE_09 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
41 Membrane. Cryo. Atm. Storage Tanks
42 59A_16, IOGP_434_3, KGSC_06, KJCE_05, RIVM_BEVI_09, TGC_03 Catastrophic Failure, Release to Atmosphere
317 Catastrophic Failure of Tank Roof
43 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
44 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
45 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
46 Process Vessels, Distillation Columns, Heat Exchangers, and Condensers
47 Yes Catastrophic Failure (Rupture)
48 Yes Release from a hole with eff. diam. of 0.01m (0.4 in)
802 Release from a hole with eff. diam. of 0.025m (1 in)
803 Release from a hole with eff. diam. of 0.05m (2 in)
804 Release from a hole with eff. diam. of 0.10m (4 in)
49 Process Vessels incl Distillation Columns
50 LEES_12, LEES_12, TNO_RED_05 Catastrophic Failure (Rupture)
51 TNO_RED_05 Release from a hole with eff. diam. of 0.01m (0.4 in)
805 Release from a hole with eff. diam. of 0.025m (1 in)
806 Release from a hole with eff. diam. of 0.05m (2 in)
807 Release from a hole with eff. diam. of 0.10m (4 in)
52 PVs incl Dist. Columns ‐ Single Wall
53 OREDA_15 Catastrophic Failure (Rupture)
54 Release from a hole with eff. diam. of 0.01m (0.4 in)
808 Release from a hole with eff. diam. of 0.025m (1 in)
809 Release from a hole with eff. diam. of 0.05m (2 in)
810 Release from a hole with eff. diam. of 0.10m (4 in)
318 Pressure Storage Vessels/Tanks (Single Wall)
59A_16, CCPS_89, INL_CHEM_95, INL_NUC_07, IOGP_434_3, LNE_09, RIVM_BEVI_09, 
319 SERCO_AEA_05, TNO_PURPLE_05, HSE_FRED_JUN_12, HSE_FRED_JUN_12 Catastrophic Failure (Rupture)
320 IOGP_434_3, LNE_09, RIVM_BEVI_09, HSE_FRED_JUN_12, HSE_FRED_JUN_12 Release from a hole with eff. diam. of 0.01m (0.4 in)
380 IOGP_434_3, LNE_09, HSE_FRED_JUN_12, HSE_FRED_JUN_12 Release from a hole with eff. diam. of 0.025m (1 in)
381 HSE_FRED_JUN_12, HSE_FRED_JUN_12 Release from a hole with eff. diam. of 0.05m (2 in)
382 IOGP_434_3 Release from a hole with eff. diam. of 0.10m (4 in)
321 Process Vessels (Single Wall)
322 DNV_LEAK_3.3, LNE_09, RIVM_BEVI_09, TNO_PURPLE_05 Catastrophic Failure (Rupture)
323 DNV_LEAK_3.3, IOGP_434_1, LNE_09, RIVM_BEVI_09 Release from a hole with eff. diam. of 0.01m (0.4 in)
383 DNV_LEAK_3.3, IOGP_434_1, LNE_09 Release from a hole with eff. diam. of 0.025m (1 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 132
 Appendix C: Equipment Protocol Framework and Applied References

Potential FRT SubCategory1
Potential FRT SubCategory2

Potential FRT SubCategory4
Super Category Comparison

Potential FRT Subcategory3
Current PHMSA FRT Spec

Current FRT Category
Index Reference Source Applied
384 DNV_LEAK_3.3 Release from a hole with eff. diam. of 0.05m (2 in)
386 DNV_LEAK_3.3, IOGP_434_1 Release from a hole with eff. diam. of 0.10m (4 in)
55 Mole Sieve Vessel (Single Wall)
56 OREDA_15 Catastrophic Failure (Rupture)
57 Release from a hole with eff. diam. of 0.01m (0.4 in)
387 Release from a hole with eff. diam. of 0.025m (1 in)
388 Release from a hole with eff. diam. of 0.05m (2 in)
389 Release from a hole with eff. diam. of 0.10m (4 in)
58 Distillation Columns (Single Wall)
59 API_581_16, RIVM_BEVI_09 Catastrophic Failure (Rupture)
60 RIVM_BEVI_09 Release from a hole with eff. diam. of 0.01m (0.4 in)
390 API_581_16 Release from a hole with eff. diam. of 0.025m (1 in)
391 Release from a hole with eff. diam. of 0.05m (2 in)
392 API_581_16 Release from a hole with eff. diam. of 0.10m (4 in)
61 Separator (Single Wall)
62 Catastrophic Failure (Rupture)
63 Release from a hole with eff. diam. of 0.01m (0.4 in)
393 Release from a hole with eff. diam. of 0.025m (1 in)
394 Release from a hole with eff. diam. of 0.05m (2 in)
395 Release from a hole with eff. diam. of 0.10m (4 in)
64 PVs incl Dist. Columns ‐ Double Wall (cryogenic)
65 Catastrophic Failure (Rupture)
66 Release from a hole with eff. diam. of 0.01m (0.4 in)
396 Release from a hole with eff. diam. of 0.025m (1 in)
397 Release from a hole with eff. diam. of 0.05m (2 in)
398 Release from a hole with eff. diam. of 0.10m (4 in)
70 ISO Containers
71 HSE_FRED_JUN_12 Catastrophic Failure (Rupture)
72 HSE_FRED_JUN_12 Release from a hole with eff. diam. of 0.01m (0.4 in)
399 Release from a hole with eff. diam. of 0.025m (1 in)
400 Release from a hole with eff. diam. of 0.05m (2 in)
401 Release from a hole with eff. diam. of 0.10m (4 in)
73 Heat Exchangers  incl. Condensers
74 OREDA_15, TNO_RED_05 Catastrophic Failure (Rupture)
75 TNO_RED_05 Release from a hole with eff. diam. of 0.01m (0.4 in)
402 Release from a hole with eff. diam. of 0.025m (1 in)
403 Release from a hole with eff. diam. of 0.05m (2 in)
404 Release from a hole with eff. diam. of 0.10m (4 in)
76 Fired Heat Exchangers
77 GRI_LNG_FRD_81 Catastrophic Failure (Rupture)
78 Release from a hole with eff. diam. of 0.01m (0.4 in)
405 Release from a hole with eff. diam. of 0.025m (1 in)
406 Release from a hole with eff. diam. of 0.05m (2 in)
407 Release from a hole with eff. diam. of 0.10m (4 in)
324 Submerged Combustion Vaporizers
325 GRI_LNG_FRD_81 Catastrophic Failure (Rupture)
326 Release from a hole with eff. diam. of 0.01m (0.4 in)
408 Release from a hole with eff. diam. of 0.025m (1 in)
409 Release from a hole with eff. diam. of 0.05m (2 in)
410 Release from a hole with eff. diam. of 0.10m (4 in)
79 Non‐Fired Heat Exchangers incl. Condensers
80 Catastrophic Failure (Rupture)
81 Release from a hole with eff. diam. of 0.01m (0.4 in)
411 Release from a hole with eff. diam. of 0.025m (1 in)
412 Release from a hole with eff. diam. of 0.05m (2 in)
413 Release from a hole with eff. diam. of 0.10m (4 in)
82 Shell & Tube Heat Exchangers
83 LNE_09, OREDA_15 Catastrophic Failure (Rupture)
84 LNE_09 Release from a hole with eff. diam. of 0.01m (0.4 in)
414 Release from a hole with eff. diam. of 0.025m (1 in)
415 Release from a hole with eff. diam. of 0.05m (2 in)
416 LNE_09 Release from a hole with eff. diam. of 0.10m (4 in)
85 Tube‐side Heat Exchangers (HC in tube)
86 DNV_LEAK_3.3, INL_CHEM_95, INL_NUC_07, TNO_PURPLE_05 Catastrophic Failure (Rupture)
87 DNV_LEAK_3.3, IOGP_434_1, TNO_PURPLE_05 Release from a hole with eff. diam. of 0.01m (0.4 in)
417 DNV_LEAK_3.3, IOGP_434_1 Release from a hole with eff. diam. of 0.025m (1 in)
418 DNV_LEAK_3.3 Release from a hole with eff. diam. of 0.05m (2 in)
419 DNV_LEAK_3.3, IOGP_434_1 Release from a hole with eff. diam. of 0.10m (4 in)
88 Shell‐side Heat Exchangers (HC in shell)
89 API_581_16, DNV_LEAK_3.3, INL_CHEM_95, INL_NUC_07, RIVM_BEVI_09 Catastrophic Failure (Rupture)
90 DNV_LEAK_3.3, IOGP_434_1, RIVM_BEVI_09 Release from a hole with eff. diam. of 0.01m (0.4 in)
420 API_581_16, DNV_LEAK_3.3, IOGP_434_1 Release from a hole with eff. diam. of 0.025m (1 in)
421 DNV_LEAK_3.3 Release from a hole with eff. diam. of 0.05m (2 in)
422 API_581_16, DNV_LEAK_3.3, IOGP_434_1 Release from a hole with eff. diam. of 0.10m (4 in)
91 Plate Heat Exchangers
92 59A_16, DNV_LEAK_3.3, LNE_09, OREDA_15, RIVM_BEVI_09 Catastrophic Failure (Rupture)
93 DNV_LEAK_3.3, IOGP_434_1, LNE_09, RIVM_BEVI_09 Release from a hole with eff. diam. of 0.01m (0.4 in)
423 DNV_LEAK_3.3, IOGP_434_1 Release from a hole with eff. diam. of 0.025m (1 in)
424 DNV_LEAK_3.3 Release from a hole with eff. diam. of 0.05m (2 in)
425 DNV_LEAK_3.3, IOGP_434_1 Release from a hole with eff. diam. of 0.10m (4 in)
94 Air Cooled (Fin Fan) Heat Exchangers
95 API_581_16, DNV_LEAK_3.3 Catastrophic Failure (Rupture)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 133
 Appendix C: Equipment Protocol Framework and Applied References

Potential FRT SubCategory1
Potential FRT SubCategory2

Potential FRT SubCategory4
Super Category Comparison

Potential FRT Subcategory3
Current PHMSA FRT Spec

Current FRT Category
Index Reference Source Applied
96 DNV_LEAK_3.3, IOGP_434_1 Release from a hole with eff. diam. of 0.01m (0.4 in)
426 API_581_16, DNV_LEAK_3.3, IOGP_434_1 Release from a hole with eff. diam. of 0.025m (1 in)
427 DNV_LEAK_3.3 Release from a hole with eff. diam. of 0.05m (2 in)
428 API_581_16, DNV_LEAK_3.3, IOGP_434_1 Release from a hole with eff. diam. of 0.10m (4 in)
97 Printed Heat Exchangers
98 Catastrophic Failure (Rupture)
99 Release from a hole with eff. diam. of 0.01m (0.4 in)
429 Release from a hole with eff. diam. of 0.025m (1 in)
430 Release from a hole with eff. diam. of 0.05m (2 in)
431 Release from a hole with eff. diam. of 0.10m (4 in)
100 Plate‐Fin Heat Exchangers
101 Catastrophic Failure (Rupture)
102 Release from a hole with eff. diam. of 0.01m (0.4 in)
432 Release from a hole with eff. diam. of 0.025m (1 in)
433 Release from a hole with eff. diam. of 0.05m (2 in)
434 Release from a hole with eff. diam. of 0.10m (4 in)
103 Ambient Vaporizer Heat Exchangers
104 Catastrophic Failure (Rupture)
105 Release from a hole with eff. diam. of 0.01m (0.4 in)
435 Release from a hole with eff. diam. of 0.025m (1 in)
436 Release from a hole with eff. diam. of 0.05m (2 in)
437 Release from a hole with eff. diam. of 0.10m (4 in)
106 Truck Transfer Arm
107 Yes 59A_16, LEES_12, LNE_09, RIVM_BEVI_09, TNO_PURPLE_05 Rupture of transfer arm
108 Yes LNE_09, RIVM_BEVI_09, TNO_PURPLE_05 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
109 Truck Transfer Arm ‐ Cryogenic
110 Rupture of transfer arm
111 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
112 Truck Transfer Arm ‐ Non‐cryogenic
113 Rupture of transfer arm
114 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
115 Ship Transfer Arm
116 Yes 59A_16, LEES_12, LNE_09, PNL_PSRP_82, RIVM_BEVI_09, SIGTTO_IP4_96, TNO_PURPLE_05 Rupture of transfer arm
117 Yes LNE_09, RIVM_BEVI_09, TNO_PURPLE_05 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
118 Ship Transfer Arm ‐ Cryogenic
119 Rupture of transfer arm
120 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
121 Ship Transfer Arm ‐ Non‐cryogenic
122 Rupture of transfer arm
123 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
124 Truck Transfer Hose
125 Yes CCPS_89, INL_CHEM_95, LNE_09, PNL_PSRP_82, RIVM_BEVI_09, TNO_PURPLE_05 Rupture of transfer hose
126 Yes LNE_09, RIVM_BEVI_09, TNO_PURPLE_05 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
127 Truck Transfer Hose ‐ Cryogenic
128 Rupture of transfer hose
129 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
130 Truck Transfer Hose ‐ Non‐cryogenic
131 HSE_FRED_JUN_12 Rupture of transfer hose
132 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
133 Ship Transfer Hose
134 LNE_09, RIVM_BEVI_09, SIGTTO_IP4_96 Rupture of transfer hose
135 LNE_09, RIVM_BEVI_09 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
136 Ship Transfer Hose ‐ Cryogenic
137 Rupture of transfer hose
138 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
139 Ship Transfer Hose ‐ Non‐cryogenic
140 Rupture of transfer hose
141 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
142 Valve: All diameters
143 Yes OREDA_15, PNL_PSRP_82 Catastrophic Rupture
327 HSE_FRED_JUN_12 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
144 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
328 Release from hole with eff. diam. of 25 mm (1 in) diam.
329 Release from hole with eff. diam. of 50 mm (2 in) diam.
330 Release from hole with eff. diam. of 100 mm (4 in) diam.
145 Manual Valves: All diameters
146 CCPS_89, INL_CHEM_95, LEES_12 Catastrophic Rupture
331 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
147 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
332 Release from hole with eff. diam. of 25 mm (1 in) diam.
333 Release from hole with eff. diam. of 50 mm (2 in) diam.
334 Release from hole with eff. diam. of 100 mm (4 in) diam.
148 Manual Valves ‐ Cryogenic: All diameters
149 Catastrophic Rupture
335 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
150 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
336 Release from hole with eff. diam. of 25 mm (1 in) diam.
337 Release from hole with eff. diam. of 50 mm (2 in) diam.
338 Release from hole with eff. diam. of 100 mm (4 in) diam.
151 Manual Valves ‐ Non‐cryogenic: All diameters
152 INL_CHEM_95, INL_NUC_07 Catastrophic Rupture
339 Release from hole with eff. diam. of 2 mm (0.08 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 134
 Appendix C: Equipment Protocol Framework and Applied References

Potential FRT SubCategory1
Potential FRT SubCategory2

Potential FRT SubCategory4
Super Category Comparison

Potential FRT Subcategory3
Current PHMSA FRT Spec

Current FRT Category
Index Reference Source Applied
153 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
340 Release from hole with eff. diam. of 25 mm (1 in) diam.
341 Release from hole with eff. diam. of 50 mm (2 in) diam.
342 Release from hole with eff. diam. of 100 mm (4 in) diam.
154 Actuated Valves: All diameters
155 Catastrophic Rupture
343 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
156 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
344 Release from hole with eff. diam. of 25 mm (1 in) diam.
345 Release from hole with eff. diam. of 50 mm (2 in) diam.
346 Release from hole with eff. diam. of 100 mm (4 in) diam.
157 Actuated Valves ‐ Cryogenic: All diameters
158 Catastrophic Rupture
347 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
159 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
348 Release from hole with eff. diam. of 25 mm (1 in) diam.
349 Release from hole with eff. diam. of 50 mm (2 in) diam.
350 Release from hole with eff. diam. of 100 mm (4 in) diam.
160 Actuated Valves ‐ Non‐cryogenic: All diameters
161 INL_CHEM_95, INL_NUC_07, LEES_12 Catastrophic Rupture
351 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
162 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
352 Release from hole with eff. diam. of 25 mm (1 in) diam.
353 Release from hole with eff. diam. of 50 mm (2 in) diam.
354 Release from hole with eff. diam. of 100 mm (4 in) diam.
438 Valve: 2"≤D<6"
439 Catastrophic Rupture
440 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
441 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
442 Release from hole with eff. diam. of 25 mm (1 in) diam.
443 Release from hole with eff. diam. of 50 mm (2 in) diam.
444 Release from hole with eff. diam. of 100 mm (4 in) diam.
445 Manual Valves: 2"≤D<6"
446 Catastrophic Rupture
447 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
448 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
449 Release from hole with eff. diam. of 25 mm (1 in) diam.
450 Release from hole with eff. diam. of 50 mm (2 in) diam.
451 Release from hole with eff. diam. of 100 mm (4 in) diam.
452 Manual Valves ‐ Cryogenic: 2"≤D<6"
453 Catastrophic Rupture
454 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
455 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
456 Release from hole with eff. diam. of 25 mm (1 in) diam.
457 Release from hole with eff. diam. of 50 mm (2 in) diam.
458 Release from hole with eff. diam. of 100 mm (4 in) diam.
459 Manual Valves ‐ Non‐cryogenic: 2"≤D<6"
460 DNV_LEAK_3.3 Catastrophic Rupture
461 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
462 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
463 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
464 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
465 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
466 Actuated Valves: 2"≤D<6"
467 Catastrophic Rupture
468 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
469 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
470 Release from hole with eff. diam. of 25 mm (1 in) diam.
471 Release from hole with eff. diam. of 50 mm (2 in) diam.
472 Release from hole with eff. diam. of 100 mm (4 in) diam.
473 Actuated Valves ‐ Cryogenic: 2"≤D<6"
474 Catastrophic Rupture
475 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
476 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
477 Release from hole with eff. diam. of 25 mm (1 in) diam.
478 Release from hole with eff. diam. of 50 mm (2 in) diam.
479 Release from hole with eff. diam. of 100 mm (4 in) diam.
480 Actuated Valves ‐ Non‐cryogenic: 2"≤D<6"
481 DNV_LEAK_3.3 Catastrophic Rupture
482 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
483 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
484 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
485 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
486 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
487 Valve: 6"≤D<12"
488 Catastrophic Rupture
489 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
490 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
491 Release from hole with eff. diam. of 25 mm (1 in) diam.
492 Release from hole with eff. diam. of 50 mm (2 in) diam.
493 Release from hole with eff. diam. of 100 mm (4 in) diam.
494 Manual Valves: 6"≤D<12"

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 135
 Appendix C: Equipment Protocol Framework and Applied References

Potential FRT SubCategory1
Potential FRT SubCategory2

Potential FRT SubCategory4
Super Category Comparison

Potential FRT Subcategory3
Current PHMSA FRT Spec

Current FRT Category
Index Reference Source Applied
495 Catastrophic Rupture
496 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
497 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
498 Release from hole with eff. diam. of 25 mm (1 in) diam.
499 Release from hole with eff. diam. of 50 mm (2 in) diam.
500 Release from hole with eff. diam. of 100 mm (4 in) diam.
501 Manual Valves ‐ Cryogenic: 6"≤D<12"
502 Catastrophic Rupture
503 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
504 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
505 Release from hole with eff. diam. of 25 mm (1 in) diam.
506 Release from hole with eff. diam. of 50 mm (2 in) diam.
507 Release from hole with eff. diam. of 100 mm (4 in) diam.
508 Manual Valves ‐ Non‐cryogenic: 6"≤D<12"
509 DNV_LEAK_3.3 Catastrophic Rupture
510 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
511 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
512 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
513 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
514 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
515 Actuated Valves: 6"≤D<12"
516 Catastrophic Rupture
517 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
518 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
519 Release from hole with eff. diam. of 25 mm (1 in) diam.
520 Release from hole with eff. diam. of 50 mm (2 in) diam.
521 Release from hole with eff. diam. of 100 mm (4 in) diam.
522 Actuated Valves ‐ Cryogenic: 6"≤D<12"
523 Catastrophic Rupture
524 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
525 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
526 Release from hole with eff. diam. of 25 mm (1 in) diam.
527 Release from hole with eff. diam. of 50 mm (2 in) diam.
528 Release from hole with eff. diam. of 100 mm (4 in) diam.
529 Actuated Valves ‐ Non‐cryogenic: 6"≤D<12"
530 DNV_LEAK_3.3 Catastrophic Rupture
531 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
532 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
533 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
534 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
535 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
536 Valve: 12"≤D<20"
537 Catastrophic Rupture
538 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
539 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
540 Release from hole with eff. diam. of 25 mm (1 in) diam.
541 Release from hole with eff. diam. of 50 mm (2 in) diam.
542 Release from hole with eff. diam. of 100 mm (4 in) diam.
543 Manual Valves: 12"≤D<20"
544 Catastrophic Rupture
545 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
546 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
547 Release from hole with eff. diam. of 25 mm (1 in) diam.
548 Release from hole with eff. diam. of 50 mm (2 in) diam.
549 Release from hole with eff. diam. of 100 mm (4 in) diam.
550 Manual Valves ‐ Cryogenic: 12"≤D<20"
551 Catastrophic Rupture
552 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
553 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
554 Release from hole with eff. diam. of 25 mm (1 in) diam.
555 Release from hole with eff. diam. of 50 mm (2 in) diam.
556 Release from hole with eff. diam. of 100 mm (4 in) diam.
557 Manual Valves ‐ Non‐cryogenic: 12"≤D<20"
558 DNV_LEAK_3.3 Catastrophic Rupture
559 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
560 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
561 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
562 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
563 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
564 Actuated Valves: 12"≤D<20"
565 Catastrophic Rupture
566 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
567 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
568 Release from hole with eff. diam. of 25 mm (1 in) diam.
569 Release from hole with eff. diam. of 50 mm (2 in) diam.
570 Release from hole with eff. diam. of 100 mm (4 in) diam.
571 Actuated Valves ‐ Cryogenic: 12"≤D<20"
572 Catastrophic Rupture
573 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
574 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
575 Release from hole with eff. diam. of 25 mm (1 in) diam.
576 Release from hole with eff. diam. of 50 mm (2 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 136
 Appendix C: Equipment Protocol Framework and Applied References

Potential FRT SubCategory1
Potential FRT SubCategory2

Potential FRT SubCategory4
Super Category Comparison

Potential FRT Subcategory3
Current PHMSA FRT Spec

Current FRT Category
Index Reference Source Applied
577 Release from hole with eff. diam. of 100 mm (4 in) diam.
578 Actuated Valves ‐ Non‐cryogenic: 12"≤D<20"
579 DNV_LEAK_3.3 Catastrophic Rupture
580 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
581 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
582 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
583 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
584 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
585 Valve: 20"≤D<40"
586 Catastrophic Rupture
587 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
588 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
589 Release from hole with eff. diam. of 25 mm (1 in) diam.
590 Release from hole with eff. diam. of 50 mm (2 in) diam.
591 Release from hole with eff. diam. of 100 mm (4 in) diam.
592 Manual Valves: 20"≤D<40"
593 Catastrophic Rupture
594 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
595 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
596 Release from hole with eff. diam. of 25 mm (1 in) diam.
597 Release from hole with eff. diam. of 50 mm (2 in) diam.
598 Release from hole with eff. diam. of 100 mm (4 in) diam.
599 Manual Valves ‐ Cryogenic: 20"≤D<40"
600 Catastrophic Rupture
601 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
602 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
603 Release from hole with eff. diam. of 25 mm (1 in) diam.
604 Release from hole with eff. diam. of 50 mm (2 in) diam.
605 Release from hole with eff. diam. of 100 mm (4 in) diam.
606 Manual Valves ‐ Non‐cryogenic: 20"≤D<40"
607 DNV_LEAK_3.3 Catastrophic Rupture
608 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
609 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
610 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
611 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
612 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
613 Actuated Valves: 20"≤D<40"
614 Catastrophic Rupture
615 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
616 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
617 Release from hole with eff. diam. of 25 mm (1 in) diam.
618 Release from hole with eff. diam. of 50 mm (2 in) diam.
619 Release from hole with eff. diam. of 100 mm (4 in) diam.
620 Actuated Valves ‐ Cryogenic: 20"≤D<40"
621 Catastrophic Rupture
622 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
623 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
624 Release from hole with eff. diam. of 25 mm (1 in) diam.
625 Release from hole with eff. diam. of 50 mm (2 in) diam.
626 Release from hole with eff. diam. of 100 mm (4 in) diam.
627 Actuated Valves ‐ Non‐cryogenic: 20"≤D<40"
628 DNV_LEAK_3.3 Catastrophic Rupture
629 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
630 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
631 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
632 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
633 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
163 Expansion Joint‐ All diameters
164 Yes PNL_PSRP_82 Catastrophic Rupture
355 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
165 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
356 Release from hole with eff. diam. of 25 mm (1 in) diam.
357 Release from hole with eff. diam. of 50 mm (2 in) diam.
358 Release from hole with eff. diam. of 100 mm (4 in) diam.
166 Expansion Joint‐ Cryogenic: All diameters
167 Catastrophic Rupture
359 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
168 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
360 Release from hole with eff. diam. of 25 mm (1 in) diam.
361 Release from hole with eff. diam. of 50 mm (2 in) diam.
362 Release from hole with eff. diam. of 100 mm (4 in) diam.
169 Expansion Joint‐ Non Cryogenic: All diameters
170 LEES_12 Catastrophic Rupture
363 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
171 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
364 Release from hole with eff. diam. of 25 mm (1 in) diam.
365 Release from hole with eff. diam. of 50 mm (2 in) diam.
366 Release from hole with eff. diam. of 100 mm (4 in) diam.
634 Expansion Joint: 2"≤D<6"
635 Catastrophic Rupture
636 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
637 Release from hole with eff. diam. of 10 mm (0.4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 137
 Appendix C: Equipment Protocol Framework and Applied References

Potential FRT SubCategory1
Potential FRT SubCategory2

Potential FRT SubCategory4
Super Category Comparison

Potential FRT Subcategory3
Current PHMSA FRT Spec

Current FRT Category
Index Reference Source Applied
638 Release from hole with eff. diam. of 25 mm (1 in) diam.
639 Release from hole with eff. diam. of 50 mm (2 in) diam.
640 Release from hole with eff. diam. of 100 mm (4 in) diam.
641 Expansion Joint‐ Cryogenic: 2"≤D<6"
642 Catastrophic Rupture
643 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
644 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
645 Release from hole with eff. diam. of 25 mm (1 in) diam.
646 Release from hole with eff. diam. of 50 mm (2 in) diam.
647 Release from hole with eff. diam. of 100 mm (4 in) diam.
648 Expansion Joint‐ Non Cryogenic: 2"≤D<6"
649 Catastrophic Rupture
650 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
651 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
652 Release from hole with eff. diam. of 25 mm (1 in) diam.
653 Release from hole with eff. diam. of 50 mm (2 in) diam.
654 Release from hole with eff. diam. of 100 mm (4 in) diam.
655 Expansion Joint: 6"≤D<12"
656 Catastrophic Rupture
657 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
658 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
659 Release from hole with eff. diam. of 25 mm (1 in) diam.
660 Release from hole with eff. diam. of 50 mm (2 in) diam.
661 Release from hole with eff. diam. of 100 mm (4 in) diam.
662 Expansion Joint‐ Cryogenic: 6"≤D<12"
663 Catastrophic Rupture
664 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
665 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
666 Release from hole with eff. diam. of 25 mm (1 in) diam.
667 Release from hole with eff. diam. of 50 mm (2 in) diam.
668 Release from hole with eff. diam. of 100 mm (4 in) diam.
669 Expansion Joint‐ Non Cryogenic: 6"≤D<12"
670 Catastrophic Rupture
671 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
672 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
673 Release from hole with eff. diam. of 25 mm (1 in) diam.
674 Release from hole with eff. diam. of 50 mm (2 in) diam.
675 Release from hole with eff. diam. of 100 mm (4 in) diam.
676 Expansion Joint: 12"≤D<20"
677 Catastrophic Rupture
678 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
679 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
680 Release from hole with eff. diam. of 25 mm (1 in) diam.
681 Release from hole with eff. diam. of 50 mm (2 in) diam.
682 Release from hole with eff. diam. of 100 mm (4 in) diam.
683 Expansion Joint‐ Cryogenic: 12"≤D<20"
684 Catastrophic Rupture
685 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
686 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
687 Release from hole with eff. diam. of 25 mm (1 in) diam.
688 Release from hole with eff. diam. of 50 mm (2 in) diam.
689 Release from hole with eff. diam. of 100 mm (4 in) diam.
690 Expansion Joint‐ Non Cryogenic: 12"≤D<20"
691 Catastrophic Rupture
692 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
693 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
694 Release from hole with eff. diam. of 25 mm (1 in) diam.
695 Release from hole with eff. diam. of 50 mm (2 in) diam.
696 Release from hole with eff. diam. of 100 mm (4 in) diam.
697 Expansion Joint: 20"≤D<40"
698 Catastrophic Rupture
699 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
700 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
701 Release from hole with eff. diam. of 25 mm (1 in) diam.
702 Release from hole with eff. diam. of 50 mm (2 in) diam.
703 Release from hole with eff. diam. of 100 mm (4 in) diam.
704 Expansion Joint‐ Cryogenic: 20"≤D<40"
705 Catastrophic Rupture
706 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
707 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
708 Release from hole with eff. diam. of 25 mm (1 in) diam.
709 Release from hole with eff. diam. of 50 mm (2 in) diam.
710 Release from hole with eff. diam. of 100 mm (4 in) diam.
711 Expansion Joint‐ Non Cryogenic: 20"≤D<40"
712 Catastrophic Rupture
713 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
714 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
715 Release from hole with eff. diam. of 25 mm (1 in) diam.
716 Release from hole with eff. diam. of 50 mm (2 in) diam.
717 Release from hole with eff. diam. of 100 mm (4 in) diam.
172 Gasket Flange‐ All diameters
173 Yes LEES_12, PNL_PSRP_82 Catastrophic Rupture

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 138
 Appendix C: Equipment Protocol Framework and Applied References

Potential FRT SubCategory1
Potential FRT SubCategory2

Potential FRT SubCategory4
Super Category Comparison

Potential FRT Subcategory3
Current PHMSA FRT Spec

Current FRT Category
Index Reference Source Applied
367 IOGP_434_1 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
174 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
368 HSE_FRED_JUN_12 Release from hole with eff. diam. of 25 mm (1 in) diam.
369 HSE_FRED_JUN_12 Release from hole with eff. diam. of 50 mm (2 in) diam.
370 Release from hole with eff. diam. of 100 mm (4 in) diam.
175 Gasket Flange‐ Cryogenic: All Diameters
176 Catastrophic Rupture
371 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
177 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
372 Release from hole with eff. diam. of 25 mm (1 in) diam.
373 Release from hole with eff. diam. of 50 mm (2 in) diam.
374 Release from hole with eff. diam. of 100 mm (4 in) diam.
178 Gasket Flange‐ Non Cryogenic: All Diameters
179 DNV_LEAK_3.3, INL_CHEM_95 Catastrophic Rupture
375 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
180 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
376 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
377 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
378 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
718 Gasket Flange: 2"≤D<6"
719 Catastrophic Rupture
720 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
721 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
722 Release from hole with eff. diam. of 25 mm (1 in) diam.
723 Release from hole with eff. diam. of 50 mm (2 in) diam.
724 Release from hole with eff. diam. of 100 mm (4 in) diam.
725 Gasket Flange‐ Cryogenic: 2"≤D<6"
726 Catastrophic Rupture
727 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
728 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
729 Release from hole with eff. diam. of 25 mm (1 in) diam.
730 Release from hole with eff. diam. of 50 mm (2 in) diam.
731 Release from hole with eff. diam. of 100 mm (4 in) diam.
732 Gasket Flange‐ Non Cryogenic: 2"≤D<6"
733 DNV_LEAK_3.3 Catastrophic Rupture
734 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
735 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
736 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
737 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
738 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
739 Gasket Flange: 6"≤D<12"
740 Catastrophic Rupture
741 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
742 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
743 Release from hole with eff. diam. of 25 mm (1 in) diam.
744 Release from hole with eff. diam. of 50 mm (2 in) diam.
745 Release from hole with eff. diam. of 100 mm (4 in) diam.
746 Gasket Flange‐ Cryogenic: 6"≤D<12"
747 Catastrophic Rupture
748 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
749 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
750 Release from hole with eff. diam. of 25 mm (1 in) diam.
751 Release from hole with eff. diam. of 50 mm (2 in) diam.
752 Release from hole with eff. diam. of 100 mm (4 in) diam.
753 Gasket Flange‐ Non Cryogenic: 6"≤D<12"
754 DNV_LEAK_3.3 Catastrophic Rupture
755 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
756 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
757 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
758 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
759 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
760 Gasket Flange: 12"≤D<20"
761 Catastrophic Rupture
762 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
763 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
764 Release from hole with eff. diam. of 25 mm (1 in) diam.
765 Release from hole with eff. diam. of 50 mm (2 in) diam.
766 Release from hole with eff. diam. of 100 mm (4 in) diam.
767 Gasket Flange‐ Cryogenic: 12"≤D<20"
768 Catastrophic Rupture
769 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
770 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
771 Release from hole with eff. diam. of 25 mm (1 in) diam.
772 Release from hole with eff. diam. of 50 mm (2 in) diam.
773 Release from hole with eff. diam. of 100 mm (4 in) diam.
774 Gasket Flange‐ Non Cryogenic: 12"≤D<20"
775 DNV_LEAK_3.3 Catastrophic Rupture
776 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
777 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
778 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
779 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
780 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 139
 Appendix C: Equipment Protocol Framework and Applied References

Potential FRT SubCategory1
Potential FRT SubCategory2

Potential FRT SubCategory4
Super Category Comparison

Potential FRT Subcategory3
Current PHMSA FRT Spec

Current FRT Category
Index Reference Source Applied
781 Gasket Flange: 20"≤D<40"
782 Catastrophic Rupture
783 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
784 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
785 Release from hole with eff. diam. of 25 mm (1 in) diam.
786 Release from hole with eff. diam. of 50 mm (2 in) diam.
787 Release from hole with eff. diam. of 100 mm (4 in) diam.
788 Gasket Flange‐ Cryogenic: 20"≤D<40"
789 Catastrophic Rupture
790 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
791 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
792 Release from hole with eff. diam. of 25 mm (1 in) diam.
793 Release from hole with eff. diam. of 50 mm (2 in) diam.
794 Release from hole with eff. diam. of 100 mm (4 in) diam.
795 Gasket Flange‐ Non Cryogenic: 20"≤D<40"
796 DNV_LEAK_3.3 Catastrophic Rupture
797 DNV_LEAK_3.3 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
798 DNV_LEAK_3.3 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
799 DNV_LEAK_3.3 Release from hole with eff. diam. of 25 mm (1 in) diam.
800 DNV_LEAK_3.3 Release from hole with eff. diam. of 50 mm (2 in) diam.
801 DNV_LEAK_3.3 Release from hole with eff. diam. of 100 mm (4 in) diam.
181 Piping:  All diameters
182 Catastrophic Rupture
183 Release from hole with eff. diam. of 1/3 diam.
184 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
185 Release from hole with eff. diam. of 25mm (1 in)
186 Piping:  d < 50mm (2‐inch)
187 Yes 59A_16, CCPS_89, INL_CHEM_95, IOGP_434_1, HSE_FRED_JUN_12 Catastrophic Rupture
188 Release from hole with eff. diam. of 1/3 diam.
189 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
190 Yes HSE_FRED_JUN_12 Release from hole with eff. diam. of 25mm (1 in)
191 Piping:  d < 50mm (2‐inch) ‐ Non‐Cryogenic
192 DNV_LEAK_3.3, LNE_09, RIVM_BEVI_09 Catastrophic Rupture
193 DNV_LEAK_3.3 Release from hole with eff. diam. of 1/3 diam.
194 DNV_LEAK_3.3, RIVM_BEVI_09 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
195 DNV_LEAK_3.3 Release from hole with eff. diam. of 25mm (1 in)
196 Piping:  d < 50mm (2‐inch) ‐ Cryogenic
197 Catastrophic Rupture
198 Release from hole with eff. diam. of 1/3 diam.
199 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
200 GRI_LNG_FRD_81 Release from hole with eff. diam. of 25mm (1 in)
201 Piping:  d < 50mm (2‐inch) ‐ Cryogenic ‐ VJ PIPING
202 Catastrophic Rupture
203 Release from hole with eff. diam. of 1/3 diam.
204 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
205 Release from hole with eff. diam. of 25mm (1 in)
206 Piping:  d < 50mm (2‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
207 Catastrophic Rupture
208 Release from hole with eff. diam. of 1/3 diam.
209 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
210 Release from hole with eff. diam. of 25mm (1 in)
211 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch)
59A_16, CCPS_89, INL_CHEM_95, IOGP_434_1, LEES_12, PNL_PSRP_82, RIVM_BEVI_09, 
212 Yes TNO_PURPLE_05, HSE_FRED_JUN_12 Catastrophic Rupture
213 IOGP_434_1 Release from hole with eff. diam. of 1/3 diam.
214 RIVM_BEVI_09 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
215 Yes HSE_FRED_JUN_12 Release from hole with eff. diam. of 25mm (1 in)
216 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Non‐Cryogenic
217 DNV_LEAK_3.3, EGIG_15, LNE_09, PHMSA_HL_GTI_16, PHMSA_NGT_GTI_16 Catastrophic Rupture
218 DNV_LEAK_3.3 Release from hole with eff. diam. of 1/3 diam.
219 DNV_LEAK_3.3 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
220 DNV_LEAK_3.3, EGIG_15 Release from hole with eff. diam. of 25mm (1 in)
221 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic
222 Catastrophic Rupture
223 Release from hole with eff. diam. of 1/3 diam.
224 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
225 GRI_LNG_FRD_81 Release from hole with eff. diam. of 25mm (1 in)
226 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic ‐ VJ PIPING
227 Catastrophic Rupture
228 Release from hole with eff. diam. of 1/3 diam.
229 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
230 Release from hole with eff. diam. of 25mm (1 in)
231 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
232 Catastrophic Rupture
233 Release from hole with eff. diam. of 1/3 diam.
234 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
235 Release from hole with eff. diam. of 25mm (1 in)
236 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch)
237 Yes 59A_16, CCPS_89, INL_CHEM_95, LEES_12, PNL_PSRP_82, HSE_FRED_JUN_12 Catastrophic Rupture
238 Yes PNL_PSRP_82, HSE_FRED_JUN_12 Release from hole with eff. diam. of 1/3 diam.
239 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
240 Yes HSE_FRED_JUN_12 Release from hole with eff. diam. of 25mm (1 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 140
 Appendix C: Equipment Protocol Framework and Applied References

Potential FRT SubCategory1
Potential FRT SubCategory2

Potential FRT SubCategory4
Super Category Comparison

Potential FRT Subcategory3
Current PHMSA FRT Spec

Current FRT Category
Index Reference Source Applied
241 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Non‐Cryogenic
242 DNV_LEAK_3.3, EGIG_15, INL_VJ_10, LNE_09, PHMSA_HL_GTI_16, PHMSA_NGT_GTI_16 Catastrophic Rupture
243 DNV_LEAK_3.3 Release from hole with eff. diam. of 1/3 diam.
244 DNV_LEAK_3.3 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
245 DNV_LEAK_3.3, EGIG_15 Release from hole with eff. diam. of 25mm (1 in)
246 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic
247 Catastrophic Rupture
248 Release from hole with eff. diam. of 1/3 diam.
249 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
250 GRI_LNG_FRD_81 Release from hole with eff. diam. of 25mm (1 in)
251 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic ‐ VJ PIPING
252 Catastrophic Rupture
253 Release from hole with eff. diam. of 1/3 diam.
254 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
255 Release from hole with eff. diam. of 25mm (1 in)
256 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
257 Catastrophic Rupture
258 Release from hole with eff. diam. of 1/3 diam.
259 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
260 Release from hole with eff. diam. of 25mm (1 in)
261 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch)
262 Yes 59A_16, CCPS_89, INL_CHEM_95, PNL_PSRP_82, HSE_FRED_JUN_12 Catastrophic Rupture
263 Yes HSE_FRED_JUN_12 Release from hole with eff. diam. of 1/3 diam.
264 Yes Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
265 Yes HSE_FRED_JUN_12 Release from hole with eff. diam. of 25mm (1 in)
266 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Non‐Cryogenic
267 DNV_LEAK_3.3, EGIG_15, INL_VJ_10, LNE_09, PHMSA_HL_GTI_16, PHMSA_NGT_GTI_16 Catastrophic Rupture
268 DNV_LEAK_3.3 Release from hole with eff. diam. of 1/3 diam.
269 DNV_LEAK_3.3, EGIG_15 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
270 DNV_LEAK_3.3 Release from hole with eff. diam. of 25mm (1 in)
271 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic
272 Catastrophic Rupture
273 Release from hole with eff. diam. of 1/3 diam.
274 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
275 Release from hole with eff. diam. of 25mm (1 in)
276 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic ‐ VJ PIPING
277 Catastrophic Rupture
278 Release from hole with eff. diam. of 1/3 diam.
279 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
280 Release from hole with eff. diam. of 25mm (1 in)
281 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
282 Catastrophic Rupture
283 Release from hole with eff. diam. of 1/3 diam.
284 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
285 Release from hole with eff. diam. of 25mm (1 in)
286 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch)
287 Yes 59A_16, CCPS_89, INL_CHEM_95, RIVM_BEVI_09, HSE_FRED_JUN_12 Catastrophic Rupture
288 Yes HSE_FRED_JUN_12 Release from hole with eff. diam. of 1/3 diam.
289 Yes RIVM_BEVI_09 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
290 Yes HSE_FRED_JUN_12 Release from hole with eff. diam. of 25mm (1 in)
291 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Non‐Cryogenic
292 DNV_LEAK_3.3, EGIG_15, LNE_09, PHMSA_HL_GTI_16, PHMSA_NGT_GTI_16 Catastrophic Rupture
293 DNV_LEAK_3.3 Release from hole with eff. diam. of 1/3 diam.
294 DNV_LEAK_3.3 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
295 DNV_LEAK_3.3, EGIG_15 Release from hole with eff. diam. of 25mm (1 in)
296 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic
297 Catastrophic Rupture
298 Release from hole with eff. diam. of 1/3 diam.
299 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
300 Release from hole with eff. diam. of 25mm (1 in)
301 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic ‐ VJ PIPING
302 Catastrophic Rupture
303 Release from hole with eff. diam. of 1/3 diam.
304 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
305 Release from hole with eff. diam. of 25mm (1 in)
306 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
307 Catastrophic Rupture
308 Release from hole with eff. diam. of 1/3 diam.
309 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
310 Release from hole with eff. diam. of 25mm (1 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 141
Appendix D: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied
and Using All References "Specifically Included in Analysis" in Table 11

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
1 Ambient Atm. Storage Tanks
2 2.47E‐03 4.95E‐03 2.48E‐03 4.99E‐03 3.16E‐03 5.97E‐03 Catastrophic Failure, Release to Atmosphere
3 9.98E‐05 9.98E‐05 9.98E‐05 9.98E‐05 9.98E‐05 9.98E‐05 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
4 2.50E‐03 2.50E‐03 2.50E‐03 2.50E‐03 2.50E‐03 2.50E‐03 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
5 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
6 Refrigerated Atm. Storage Tanks (Typ. Single Shell)
7 2.30E‐03 2.30E‐03 2.32E‐03 2.32E‐03 3.25E‐03 3.25E‐03 Catastrophic Failure, Release to Atmosphere
8 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
9 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
10 1.26E‐03 1.26E‐03 1.26E‐03 1.26E‐03 1.26E‐03 1.26E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
11 Cryogenic Atm. Storage Tanks
12 7.22E‐04 7.22E‐04 5.25E‐06 5.25E‐06 7.40E‐04 7.40E‐04 5.17E‐06 5.17E‐06 7.82E‐04 7.82E‐04 5.25E‐06 5.25E‐06 Catastrophic Failure, Release to Atmosphere
311 6.86E‐05 6.86E‐05 1.89E‐05 1.89E‐05 5.36E‐05 5.36E‐05 1.89E‐05 1.89E‐05 6.86E‐05 6.86E‐05 1.89E‐05 1.89E‐05 Catastrophic Failure of Tank Roof
13 2.90E‐05 2.90E‐05 1.01E‐06 1.01E‐06 2.06E‐05 2.06E‐05 1.01E‐06 1.01E‐06 2.90E‐05 2.90E‐05 1.01E‐06 1.01E‐06 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
14 2.39E‐05 2.39E‐05 3.06E‐06 3.06E‐06 1.75E‐05 1.75E‐05 3.06E‐06 3.06E‐06 2.39E‐05 2.39E‐05 3.06E‐06 3.06E‐06 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
15 1.44E‐03 1.44E‐03 1.44E‐03 1.44E‐03 1.44E‐03 1.44E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
16 Single Cont. Cryo. Atm. Storage Tanks
17 5.00E‐06 1.14E‐03 1.14E‐03 2.99E‐05 2.99E‐05 1.17E‐03 1.17E‐03 2.99E‐05 2.99E‐05 1.29E‐03 1.29E‐03 2.99E‐05 2.99E‐05 Catastrophic Failure, Release to Atmosphere
312 1.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 Catastrophic Failure of Tank Roof
18 8.00E‐05 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
19 2.00E‐04 8.00E‐05 8.00E‐05 8.00E‐05 8.00E‐05 8.00E‐05 8.00E‐05 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
20 1.00E‐04 1.19E‐03 1.19E‐03 1.19E‐03 1.19E‐03 1.19E‐03 1.19E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
21 Single Cont. Cryo. Atm. Storage Tanks ‐ HC Only (LNG, Ethane, Ethylene) ‐ Self‐Supp. Inner Tank
22 Catastrophic Failure, Release to Atmosphere
313 Catastrophic Failure of Tank Roof
23 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
24 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
25 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
26 Single Cont. Cryo. Atm. Storage Tanks ‐ LIN/LOX/LAR ‐ Self‐Supp. Inner Tank
27 2.20E‐05 2.20E‐05 2.20E‐05 2.20E‐05 2.20E‐05 2.20E‐05 Catastrophic Failure, Release to Atmosphere
314 Catastrophic Failure of Tank Roof
28 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
29 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
30 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
31 Double Cont. Cryo. Atm. Storage Tanks
32 5.00E‐07 3.23E‐08 3.23E‐08 7.44E‐08 7.44E‐08 3.19E‐08 3.19E‐08 7.39E‐08 7.38E‐08 4.64E‐08 4.64E‐08 7.45E‐08 7.44E‐08 Catastrophic Failure, Release to Atmosphere
315 1.00E‐04 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 Catastrophic Failure of Tank Roof
33 1.00E‐05 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
34 3.00E‐05 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
35 1.00E‐04 1.32E‐03 1.32E‐03 1.32E‐03 1.32E‐03 1.32E‐03 1.32E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
36 Full Cont. Cryo. Atm. Storage Tanks
37 1.00E‐08 2.31E‐07 2.31E‐07 7.47E‐07 7.47E‐07 2.22E‐07 2.22E‐07 7.27E‐07 7.27E‐07 4.02E‐07 4.02E‐07 7.47E‐07 7.47E‐07 Catastrophic Failure, Release to Atmosphere
316 4.00E‐05 2.42E‐10 2.46E‐10 2.43E‐10 2.43E‐10 2.41E‐10 2.46E‐10 2.47E‐10 2.47E‐10 2.43E‐10 2.44E‐10 2.41E‐10 2.43E‐10 Catastrophic Failure of Tank Roof
38 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
39 3.00E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
40 1.00E‐04 2.40E‐03 2.40E‐03 2.40E‐03 2.40E‐03 2.40E‐03 2.40E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
41 Membrane. Cryo. Atm. Storage Tanks
42 2.11E‐06 2.11E‐06 4.46E‐06 4.46E‐06 2.05E‐06 2.05E‐06 4.46E‐06 4.46E‐06 3.14E‐06 3.14E‐06 4.46E‐06 4.46E‐06 Catastrophic Failure, Release to Atmosphere
317 Catastrophic Failure of Tank Roof
43 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
44 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
45 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
46 Process Vessels, Distillation Columns, Heat Exchangers, and Condensers
47 5.00E‐06 2.05E‐02 1.89E‐02 2.72E‐01 2.72E‐01 2.32E‐02 2.34E‐02 2.72E‐01 2.72E‐01 3.06E‐02 3.04E‐02 2.72E‐01 2.72E‐01 Catastrophic Failure (Rupture)
48 1.00E‐04 1.22E‐03 1.22E‐03 1.26E‐03 1.26E‐03 1.13E‐03 1.13E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
802 3.26E‐04 3.77E‐04 2.94E‐04 3.56E‐04 4.00E‐04 4.27E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
803 1.16E‐04 1.16E‐04 1.17E‐04 1.17E‐04 1.22E‐04 1.22E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
804 9.97E‐05 1.23E‐04 8.89E‐05 1.16E‐04 1.22E‐04 1.34E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
49 Process Vessels incl Distillation Columns
50 4.16E‐03 3.65E‐05 3.28E‐03 3.36E‐05 7.14E‐03 5.28E‐05 Catastrophic Failure (Rupture)
51 1.94E‐04 1.94E‐04 1.66E‐04 1.66E‐04 2.99E‐04 2.99E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
805 9.47E‐05 1.06E‐04 8.11E‐05 9.24E‐05 1.41E‐04 1.51E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
806 3.73E‐05 3.73E‐05 3.87E‐05 3.87E‐05 4.29E‐05 4.29E‐05 Release from a hole with eff. diam. of 0.05m (2 in)
807 4.59E‐05 6.06E‐05 3.85E‐05 5.37E‐05 6.53E‐05 7.53E‐05 Release from a hole with eff. diam. of 0.10m (4 in)
52 PVs incl Dist. Columns ‐ Single Wall
53 5.18E‐03 3.89E‐05 4.20E‐03 3.51E‐05 8.30E‐03 5.86E‐05 Catastrophic Failure (Rupture)
54 2.22E‐04 2.22E‐04 1.93E‐04 1.93E‐04 3.28E‐04 3.28E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
808 9.47E‐05 1.06E‐04 8.11E‐05 9.24E‐05 1.41E‐04 1.51E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
809 3.73E‐05 3.73E‐05 3.87E‐05 3.87E‐05 4.29E‐05 4.29E‐05 Release from a hole with eff. diam. of 0.05m (2 in)
810 4.59E‐05 6.06E‐05 3.85E‐05 5.37E‐05 6.53E‐05 7.53E‐05 Release from a hole with eff. diam. of 0.10m (4 in)
318 Pressure Storage Vessels/Tanks (Single Wall)
319 1.28E‐05 1.44E‐05 1.44E‐05 1.54E‐05 1.39E‐05 1.80E‐05 Catastrophic Failure (Rupture)
320 1.10E‐05 1.10E‐05 1.11E‐05 1.11E‐05 1.07E‐05 1.07E‐05 Release from a hole with eff. diam. of 0.01m (0.4 in)
380 4.57E‐06 4.57E‐06 4.52E‐06 4.52E‐06 4.69E‐06 4.69E‐06 Release from a hole with eff. diam. of 0.025m (1 in)
381 4.99E‐06 4.99E‐06 4.99E‐06 4.99E‐06 4.99E‐06 4.99E‐06 Release from a hole with eff. diam. of 0.05m (2 in)
382 4.17E‐06 4.17E‐06 4.17E‐06 4.17E‐06 4.17E‐06 4.17E‐06 Release from a hole with eff. diam. of 0.10m (4 in)
321 Process Vessels (Single Wall)
322 8.46E‐05 8.46E‐05 7.47E‐05 7.47E‐05 1.17E‐04 1.17E‐04 Catastrophic Failure (Rupture)
323 4.28E‐04 4.28E‐04 3.86E‐04 3.86E‐04 5.29E‐04 5.29E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
383 Some cells shaded to  Release from a hole with eff. diam. of 0.025m (1 in)
384 protect copyrights or  Release from a hole with eff. diam. of 0.05m (2 in)
386 licensed software Release from a hole with eff. diam. of 0.10m (4 in)
55 Mole Sieve Vessel (Single Wall)
56 Catastrophic Failure (Rupture)
57 Release from a hole with eff. diam. of 0.01m (0.4 in)
387 Release from a hole with eff. diam. of 0.025m (1 in)
388 Release from a hole with eff. diam. of 0.05m (2 in)
389 Release from a hole with eff. diam. of 0.10m (4 in)
58 Distillation Columns (Single Wall)
59 Catastrophic Failure (Rupture)
60 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
390 Release from a hole with eff. diam. of 0.025m (1 in)
391 Release from a hole with eff. diam. of 0.05m (2 in)
392 Release from a hole with eff. diam. of 0.10m (4 in)
61 Separator (Single Wall)
62 Catastrophic Failure (Rupture)
63 Release from a hole with eff. diam. of 0.01m (0.4 in)
393 Release from a hole with eff. diam. of 0.025m (1 in)
394 Release from a hole with eff. diam. of 0.05m (2 in)
395 Release from a hole with eff. diam. of 0.10m (4 in)
64 PVs incl Dist. Columns ‐ Double Wall (cryogenic)
65 Catastrophic Failure (Rupture)
66 Release from a hole with eff. diam. of 0.01m (0.4 in)
396 Release from a hole with eff. diam. of 0.025m (1 in)
397 Release from a hole with eff. diam. of 0.05m (2 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 142
Appendix D: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied
and Using All References "Specifically Included in Analysis" in Table 11

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
398 Release from a hole with eff. diam. of 0.10m (4 in)
70 ISO Containers
71 2.95E‐06 2.95E‐06 2.95E‐06 2.95E‐06 2.95E‐06 2.95E‐06 Catastrophic Failure (Rupture)
72 6.01E‐05 6.01E‐05 6.01E‐05 6.01E‐05 6.01E‐05 6.01E‐05 Release from a hole with eff. diam. of 0.01m (0.4 in)
399 Release from a hole with eff. diam. of 0.025m (1 in)
400 Release from a hole with eff. diam. of 0.05m (2 in)
401 Release from a hole with eff. diam. of 0.10m (4 in)
73 Heat Exchangers  incl. Condensers
74 3.70E‐02 3.84E‐02 2.72E‐01 2.72E‐01 4.39E‐02 4.94E‐02 2.72E‐01 2.72E‐01 4.86E‐02 5.07E‐02 2.72E‐01 2.72E‐01 Catastrophic Failure (Rupture)
75 1.91E‐03 1.91E‐03 2.09E‐03 2.09E‐03 1.52E‐03 1.52E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
402 4.64E‐04 5.44E‐04 4.37E‐04 5.44E‐04 5.11E‐04 5.44E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
403 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
404 1.17E‐04 1.41E‐04 1.07E‐04 1.36E‐04 1.38E‐04 1.49E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
76 Fired Heat Exchangers
77 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 Catastrophic Failure (Rupture)
78 Release from a hole with eff. diam. of 0.01m (0.4 in)
405 Release from a hole with eff. diam. of 0.025m (1 in)
406 Release from a hole with eff. diam. of 0.05m (2 in)
407 Release from a hole with eff. diam. of 0.10m (4 in)
324 Submerged Combustion Vaporizers
325 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 Catastrophic Failure (Rupture)
326 Release from a hole with eff. diam. of 0.01m (0.4 in)
408 Release from a hole with eff. diam. of 0.025m (1 in)
409 Release from a hole with eff. diam. of 0.05m (2 in)
410 Release from a hole with eff. diam. of 0.10m (4 in)
79 Non‐Fired Heat Exchangers incl. Condensers
80 1.02E‐02 1.34E‐04 8.60E‐03 1.20E‐04 1.39E‐02 1.67E‐04 Catastrophic Failure (Rupture)
81 2.02E‐03 2.02E‐03 2.25E‐03 2.25E‐03 1.56E‐03 1.56E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
411 4.64E‐04 5.44E‐04 4.37E‐04 5.44E‐04 5.11E‐04 5.44E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
412 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
413 1.17E‐04 1.41E‐04 1.07E‐04 1.36E‐04 1.38E‐04 1.49E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
82 Shell & Tube Heat Exchangers
83 1.41E‐03 8.30E‐05 1.20E‐03 7.66E‐05 1.89E‐03 9.87E‐05 Catastrophic Failure (Rupture)
84 1.16E‐03 1.16E‐03 1.24E‐03 1.24E‐03 9.97E‐04 9.97E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
414 3.42E‐04 3.94E‐04 3.25E‐04 3.94E‐04 3.73E‐04 3.94E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
415 1.01E‐04 1.01E‐04 1.01E‐04 1.01E‐04 1.01E‐04 1.01E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
416 6.91E‐05 8.13E‐05 6.25E‐05 7.72E‐05 8.29E‐05 8.90E‐05 Release from a hole with eff. diam. of 0.10m (4 in)
85 Tube‐side Heat Exchangers (HC in tube)
86 1.67E‐04 8.85E‐05 1.81E‐04 8.41E‐05 1.92E‐04 9.79E‐05 Catastrophic Failure (Rupture)
87 Release from a hole with eff. diam. of 0.01m (0.4 in)
417 Release from a hole with eff. diam. of 0.025m (1 in)
418 Release from a hole with eff. diam. of 0.05m (2 in)
419 Release from a hole with eff. diam. of 0.10m (4 in)
88 Shell‐side Heat Exchangers (HC in shell)
89 7.81E‐05 1.03E‐04 7.08E‐05 9.75E‐05 9.99E‐05 1.14E‐04 Catastrophic Failure (Rupture)
90 Release from a hole with eff. diam. of 0.01m (0.4 in)
420 Release from a hole with eff. diam. of 0.025m (1 in)
421 Release from a hole with eff. diam. of 0.05m (2 in)
422 Release from a hole with eff. diam. of 0.10m (4 in)
91 Plate Heat Exchangers
92 2.89E‐02 2.36E‐04 2.43E‐02 1.99E‐04 4.11E‐02 3.48E‐04 Catastrophic Failure (Rupture)
93 Release from a hole with eff. diam. of 0.01m (0.4 in)
423 Release from a hole with eff. diam. of 0.025m (1 in)
424 Release from a hole with eff. diam. of 0.05m (2 in)
425 Release from a hole with eff. diam. of 0.10m (4 in)
94 Air Cooled (Fin Fan) Heat Exchangers
95 3.00E‐05 4.87E‐05 2.59E‐05 4.87E‐05 3.95E‐05 4.87E‐05 Catastrophic Failure (Rupture)
96 Release from a hole with eff. diam. of 0.01m (0.4 in)
426 Release from a hole with eff. diam. of 0.025m (1 in)
427 Release from a hole with eff. diam. of 0.05m (2 in)
428 Release from a hole with eff. diam. of 0.10m (4 in)
97 Printed Heat Exchangers
98 Catastrophic Failure (Rupture)
99 Release from a hole with eff. diam. of 0.01m (0.4 in)
429 Release from a hole with eff. diam. of 0.025m (1 in)
430 Release from a hole with eff. diam. of 0.05m (2 in)
431 Release from a hole with eff. diam. of 0.10m (4 in)
100 Plate‐Fin Heat Exchangers
101 Catastrophic Failure (Rupture)
102 Release from a hole with eff. diam. of 0.01m (0.4 in)
432 Release from a hole with eff. diam. of 0.025m (1 in)
433 Release from a hole with eff. diam. of 0.05m (2 in)
434 Release from a hole with eff. diam. of 0.10m (4 in)
103 Ambient Vaporizer Heat Exchangers
104 Catastrophic Failure (Rupture)
105 Release from a hole with eff. diam. of 0.01m (0.4 in)
435 Release from a hole with eff. diam. of 0.025m (1 in)
436 Release from a hole with eff. diam. of 0.05m (2 in)
437 Release from a hole with eff. diam. of 0.10m (4 in)
106 Truck Transfer Arm
107 3.00E‐04 1.51E‐05 1.51E‐05 1.51E‐05 1.51E‐05 1.51E‐05 1.51E‐05 Rupture of transfer arm
108 3.00E‐03 2.09E‐04 2.09E‐04 2.09E‐04 2.09E‐04 2.09E‐04 2.09E‐04 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
109 Truck Transfer Arm ‐ Cryogenic
110 Rupture of transfer arm
111 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
112 Truck Transfer Arm ‐ Non‐cryogenic
113 Rupture of transfer arm
114 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
115 Ship Transfer Arm
116 2.00E‐05 1.03E‐03 1.11E‐03 1.02E‐03 1.11E‐03 1.03E‐03 1.11E‐03 Rupture of transfer arm
117 2.00E‐04 1.89E‐02 1.89E‐02 1.89E‐02 1.89E‐02 1.89E‐02 1.89E‐02 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
118 Ship Transfer Arm ‐ Cryogenic
119 Rupture of transfer arm
120 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
121 Ship Transfer Arm ‐ Non‐cryogenic
122 Rupture of transfer arm
123 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
124 Truck Transfer Hose
125 4.00E‐02 7.25E‐03 4.92E‐03 7.47E‐03 5.06E‐03 1.05E‐02 5.07E‐03 Rupture of transfer hose
126 4.00E‐01 2.82E‐02 2.82E‐02 2.82E‐02 2.82E‐02 2.82E‐02 2.82E‐02 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
127 Truck Transfer Hose ‐ Cryogenic
128 Rupture of transfer hose
129 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
130 Truck Transfer Hose ‐ Non‐cryogenic
131 4.00E‐04 4.00E‐04 4.00E‐04 4.00E‐04 4.00E‐04 4.00E‐04 Rupture of transfer hose

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 143
Appendix D: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied
and Using All References "Specifically Included in Analysis" in Table 11

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
132 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
133 Ship Transfer Hose
134 2.71E‐03 4.00E‐03 2.61E‐03 4.00E‐03 2.71E‐03 4.00E‐03 Rupture of transfer hose
135 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
136 Ship Transfer Hose ‐ Cryogenic
137 Rupture of transfer hose
138 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
139 Ship Transfer Hose ‐ Non‐cryogenic
140 Rupture of transfer hose
141 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
142 Valve: All diameters
143 9.00E‐06 1.98E‐03 1.14E‐04 1.83E‐03 1.46E‐04 2.10E‐03 8.27E‐05 Catastrophic Rupture
327 2.01E‐04 2.01E‐04 2.01E‐04 2.01E‐04 2.01E‐04 2.01E‐04 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
144 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
328 Release from hole with eff. diam. of 25 mm (1 in) diam.
329 Release from hole with eff. diam. of 50 mm (2 in) diam.
330 Release from hole with eff. diam. of 100 mm (4 in) diam.
145 Manual Valves: All diameters
146 2.12E‐04 2.13E‐04 2.73E‐04 2.80E‐04 1.48E‐04 1.44E‐04 Catastrophic Rupture
331 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
147 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
332 Release from hole with eff. diam. of 25 mm (1 in) diam.
333 Release from hole with eff. diam. of 50 mm (2 in) diam.
334 Release from hole with eff. diam. of 100 mm (4 in) diam.
148 Manual Valves ‐ Cryogenic: All diameters
149 Catastrophic Rupture
335 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
150 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
336 Release from hole with eff. diam. of 25 mm (1 in) diam.
337 Release from hole with eff. diam. of 50 mm (2 in) diam.
338 Release from hole with eff. diam. of 100 mm (4 in) diam.
151 Manual Valves ‐ Non‐cryogenic: All diameters
152 1.41E‐05 5.82E‐06 1.73E‐05 5.82E‐06 1.41E‐05 5.83E‐06 Catastrophic Rupture
339 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
153 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
340 Release from hole with eff. diam. of 25 mm (1 in) diam.
341 Release from hole with eff. diam. of 50 mm (2 in) diam.
342 Release from hole with eff. diam. of 100 mm (4 in) diam.
154 Actuated Valves: All diameters
155 2.70E‐05 2.18E‐05 3.01E‐05 2.33E‐05 2.70E‐05 2.18E‐05 Catastrophic Rupture
343 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
156 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
344 Release from hole with eff. diam. of 25 mm (1 in) diam.
345 Release from hole with eff. diam. of 50 mm (2 in) diam.
346 Release from hole with eff. diam. of 100 mm (4 in) diam.
157 Actuated Valves ‐ Cryogenic: All diameters
158 Catastrophic Rupture
347 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
159 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
348 Release from hole with eff. diam. of 25 mm (1 in) diam.
349 Release from hole with eff. diam. of 50 mm (2 in) diam.
350 Release from hole with eff. diam. of 100 mm (4 in) diam.
160 Actuated Valves ‐ Non‐cryogenic: All diameters
161 2.70E‐05 2.18E‐05 3.01E‐05 2.33E‐05 2.70E‐05 2.18E‐05 Catastrophic Rupture
351 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
162 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
352 Release from hole with eff. diam. of 25 mm (1 in) diam.
353 Release from hole with eff. diam. of 50 mm (2 in) diam.
354 Release from hole with eff. diam. of 100 mm (4 in) diam.
438 Valve: 2"≤D<6"
439 Catastrophic Rupture
440 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
441 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
442 Release from hole with eff. diam. of 25 mm (1 in) diam.
443 Release from hole with eff. diam. of 50 mm (2 in) diam.
444 Release from hole with eff. diam. of 100 mm (4 in) diam.
445 Manual Valves: 2"≤D<6"
446 Catastrophic Rupture
447 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
448 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
449 Release from hole with eff. diam. of 25 mm (1 in) diam.
450 Release from hole with eff. diam. of 50 mm (2 in) diam.
451 Release from hole with eff. diam. of 100 mm (4 in) diam.
452 Manual Valves ‐ Cryogenic: 2"≤D<6"
453 Catastrophic Rupture
454 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
455 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
456 Release from hole with eff. diam. of 25 mm (1 in) diam.
457 Release from hole with eff. diam. of 50 mm (2 in) diam.
458 Release from hole with eff. diam. of 100 mm (4 in) diam.
459 Manual Valves ‐ Non‐cryogenic: 2"≤D<6"
460 Catastrophic Rupture
461 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
462 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
463 Release from hole with eff. diam. of 25 mm (1 in) diam.
464 Release from hole with eff. diam. of 50 mm (2 in) diam.
465 Release from hole with eff. diam. of 100 mm (4 in) diam.
466 Actuated Valves: 2"≤D<6"
467 Catastrophic Rupture
468 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
469 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
470 Release from hole with eff. diam. of 25 mm (1 in) diam.
471 Release from hole with eff. diam. of 50 mm (2 in) diam.
472 Release from hole with eff. diam. of 100 mm (4 in) diam.
473 Actuated Valves ‐ Cryogenic: 2"≤D<6"
474 Catastrophic Rupture
475 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
476 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
477 Release from hole with eff. diam. of 25 mm (1 in) diam.
478 Release from hole with eff. diam. of 50 mm (2 in) diam.
479 Release from hole with eff. diam. of 100 mm (4 in) diam.
480 Actuated Valves ‐ Non‐cryogenic: 2"≤D<6"
481 Catastrophic Rupture
482 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
483 Release from hole with eff. diam. of 10 mm (0.4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 144
Appendix D: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied
and Using All References "Specifically Included in Analysis" in Table 11

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
484 Release from hole with eff. diam. of 25 mm (1 in) diam.
485 Release from hole with eff. diam. of 50 mm (2 in) diam.
486 Release from hole with eff. diam. of 100 mm (4 in) diam.
487 Valve: 6"≤D<12"
488 Catastrophic Rupture
489 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
490 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
491 Release from hole with eff. diam. of 25 mm (1 in) diam.
492 Release from hole with eff. diam. of 50 mm (2 in) diam.
493 Release from hole with eff. diam. of 100 mm (4 in) diam.
494 Manual Valves: 6"≤D<12"
495 Catastrophic Rupture
496 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
497 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
498 Release from hole with eff. diam. of 25 mm (1 in) diam.
499 Release from hole with eff. diam. of 50 mm (2 in) diam.
500 Release from hole with eff. diam. of 100 mm (4 in) diam.
501 Manual Valves ‐ Cryogenic: 6"≤D<12"
502 Catastrophic Rupture
503 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
504 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
505 Release from hole with eff. diam. of 25 mm (1 in) diam.
506 Release from hole with eff. diam. of 50 mm (2 in) diam.
507 Release from hole with eff. diam. of 100 mm (4 in) diam.
508 Manual Valves ‐ Non‐cryogenic: 6"≤D<12"
509 Catastrophic Rupture
510 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
511 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
512 Release from hole with eff. diam. of 25 mm (1 in) diam.
513 Release from hole with eff. diam. of 50 mm (2 in) diam.
514 Release from hole with eff. diam. of 100 mm (4 in) diam.
515 Actuated Valves: 6"≤D<12"
516 Catastrophic Rupture
517 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
518 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
519 Release from hole with eff. diam. of 25 mm (1 in) diam.
520 Release from hole with eff. diam. of 50 mm (2 in) diam.
521 Release from hole with eff. diam. of 100 mm (4 in) diam.
522 Actuated Valves ‐ Cryogenic: 6"≤D<12"
523 Catastrophic Rupture
524 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
525 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
526 Release from hole with eff. diam. of 25 mm (1 in) diam.
527 Release from hole with eff. diam. of 50 mm (2 in) diam.
528 Release from hole with eff. diam. of 100 mm (4 in) diam.
529 Actuated Valves ‐ Non‐cryogenic: 6"≤D<12"
530 Catastrophic Rupture
531 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
532 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
533 Release from hole with eff. diam. of 25 mm (1 in) diam.
534 Release from hole with eff. diam. of 50 mm (2 in) diam.
535 Release from hole with eff. diam. of 100 mm (4 in) diam.
536 Valve: 12"≤D<20"
537 Catastrophic Rupture
538 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
539 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
540 Release from hole with eff. diam. of 25 mm (1 in) diam.
541 Release from hole with eff. diam. of 50 mm (2 in) diam.
542 Release from hole with eff. diam. of 100 mm (4 in) diam.
543 Manual Valves: 12"≤D<20"
544 Catastrophic Rupture
545 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
546 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
547 Release from hole with eff. diam. of 25 mm (1 in) diam.
548 Release from hole with eff. diam. of 50 mm (2 in) diam.
549 Release from hole with eff. diam. of 100 mm (4 in) diam.
550 Manual Valves ‐ Cryogenic: 12"≤D<20"
551 Catastrophic Rupture
552 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
553 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
554 Release from hole with eff. diam. of 25 mm (1 in) diam.
555 Release from hole with eff. diam. of 50 mm (2 in) diam.
556 Release from hole with eff. diam. of 100 mm (4 in) diam.
557 Manual Valves ‐ Non‐cryogenic: 12"≤D<20"
558 Catastrophic Rupture
559 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
560 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
561 Release from hole with eff. diam. of 25 mm (1 in) diam.
562 Release from hole with eff. diam. of 50 mm (2 in) diam.
563 Release from hole with eff. diam. of 100 mm (4 in) diam.
564 Actuated Valves: 12"≤D<20"
565 Catastrophic Rupture
566 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
567 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
568 Release from hole with eff. diam. of 25 mm (1 in) diam.
569 Release from hole with eff. diam. of 50 mm (2 in) diam.
570 Release from hole with eff. diam. of 100 mm (4 in) diam.
571 Actuated Valves ‐ Cryogenic: 12"≤D<20"
572 Catastrophic Rupture
573 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
574 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
575 Release from hole with eff. diam. of 25 mm (1 in) diam.
576 Release from hole with eff. diam. of 50 mm (2 in) diam.
577 Release from hole with eff. diam. of 100 mm (4 in) diam.
578 Actuated Valves ‐ Non‐cryogenic: 12"≤D<20"
579 Catastrophic Rupture
580 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
581 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
582 Release from hole with eff. diam. of 25 mm (1 in) diam.
583 Release from hole with eff. diam. of 50 mm (2 in) diam.
584 Release from hole with eff. diam. of 100 mm (4 in) diam.
585 Valve: 20"≤D<40"
586 Catastrophic Rupture
587 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
588 Release from hole with eff. diam. of 10 mm (0.4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 145
Appendix D: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied
and Using All References "Specifically Included in Analysis" in Table 11

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
589 Release from hole with eff. diam. of 25 mm (1 in) diam.
590 Release from hole with eff. diam. of 50 mm (2 in) diam.
591 Release from hole with eff. diam. of 100 mm (4 in) diam.
592 Manual Valves: 20"≤D<40"
593 Catastrophic Rupture
594 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
595 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
596 Release from hole with eff. diam. of 25 mm (1 in) diam.
597 Release from hole with eff. diam. of 50 mm (2 in) diam.
598 Release from hole with eff. diam. of 100 mm (4 in) diam.
599 Manual Valves ‐ Cryogenic: 20"≤D<40"
600 Catastrophic Rupture
601 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
602 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
603 Release from hole with eff. diam. of 25 mm (1 in) diam.
604 Release from hole with eff. diam. of 50 mm (2 in) diam.
605 Release from hole with eff. diam. of 100 mm (4 in) diam.
606 Manual Valves ‐ Non‐cryogenic: 20"≤D<40"
607 Catastrophic Rupture
608 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
609 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
610 Release from hole with eff. diam. of 25 mm (1 in) diam.
611 Release from hole with eff. diam. of 50 mm (2 in) diam.
612 Release from hole with eff. diam. of 100 mm (4 in) diam.
613 Actuated Valves: 20"≤D<40"
614 Catastrophic Rupture
615 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
616 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
617 Release from hole with eff. diam. of 25 mm (1 in) diam.
618 Release from hole with eff. diam. of 50 mm (2 in) diam.
619 Release from hole with eff. diam. of 100 mm (4 in) diam.
620 Actuated Valves ‐ Cryogenic: 20"≤D<40"
621 Catastrophic Rupture
622 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
623 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
624 Release from hole with eff. diam. of 25 mm (1 in) diam.
625 Release from hole with eff. diam. of 50 mm (2 in) diam.
626 Release from hole with eff. diam. of 100 mm (4 in) diam.
627 Actuated Valves ‐ Non‐cryogenic: 20"≤D<40"
628 Catastrophic Rupture
629 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
630 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
631 Release from hole with eff. diam. of 25 mm (1 in) diam.
632 Release from hole with eff. diam. of 50 mm (2 in) diam.
633 Release from hole with eff. diam. of 100 mm (4 in) diam.
163 Expansion Joint‐ All diameters
164 4.00E‐03 1.06E‐03 1.06E‐03 1.06E‐03 1.06E‐03 1.29E‐03 1.29E‐03 Catastrophic Rupture
355 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
165 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
356 Release from hole with eff. diam. of 25 mm (1 in) diam.
357 Release from hole with eff. diam. of 50 mm (2 in) diam.
358 Release from hole with eff. diam. of 100 mm (4 in) diam.
166 Expansion Joint‐ Cryogenic: All diameters
167 Catastrophic Rupture
359 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
168 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
360 Release from hole with eff. diam. of 25 mm (1 in) diam.
361 Release from hole with eff. diam. of 50 mm (2 in) diam.
362 Release from hole with eff. diam. of 100 mm (4 in) diam.
169 Expansion Joint‐ Non Cryogenic: All diameters
170 Catastrophic Rupture
363 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
171 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
364 Release from hole with eff. diam. of 25 mm (1 in) diam.
365 Release from hole with eff. diam. of 50 mm (2 in) diam.
366 Release from hole with eff. diam. of 100 mm (4 in) diam.
634 Expansion Joint: 2"≤D<6"
635 Catastrophic Rupture
636 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
637 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
638 Release from hole with eff. diam. of 25 mm (1 in) diam.
639 Release from hole with eff. diam. of 50 mm (2 in) diam.
640 Release from hole with eff. diam. of 100 mm (4 in) diam.
641 Expansion Joint‐ Cryogenic: 2"≤D<6"
642 Catastrophic Rupture
643 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
644 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
645 Release from hole with eff. diam. of 25 mm (1 in) diam.
646 Release from hole with eff. diam. of 50 mm (2 in) diam.
647 Release from hole with eff. diam. of 100 mm (4 in) diam.
648 Expansion Joint‐ Non Cryogenic: 2"≤D<6"
649 Catastrophic Rupture
650 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
651 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
652 Release from hole with eff. diam. of 25 mm (1 in) diam.
653 Release from hole with eff. diam. of 50 mm (2 in) diam.
654 Release from hole with eff. diam. of 100 mm (4 in) diam.
655 Expansion Joint: 6"≤D<12"
656 Catastrophic Rupture
657 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
658 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
659 Release from hole with eff. diam. of 25 mm (1 in) diam.
660 Release from hole with eff. diam. of 50 mm (2 in) diam.
661 Release from hole with eff. diam. of 100 mm (4 in) diam.
662 Expansion Joint‐ Cryogenic: 6"≤D<12"
663 Catastrophic Rupture
664 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
665 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
666 Release from hole with eff. diam. of 25 mm (1 in) diam.
667 Release from hole with eff. diam. of 50 mm (2 in) diam.
668 Release from hole with eff. diam. of 100 mm (4 in) diam.
669 Expansion Joint‐ Non Cryogenic: 6"≤D<12"
670 Catastrophic Rupture
671 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
672 Release from hole with eff. diam. of 10 mm (0.4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 146
Appendix D: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied
and Using All References "Specifically Included in Analysis" in Table 11

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
673 Release from hole with eff. diam. of 25 mm (1 in) diam.
674 Release from hole with eff. diam. of 50 mm (2 in) diam.
675 Release from hole with eff. diam. of 100 mm (4 in) diam.
676 Expansion Joint: 12"≤D<20"
677 Catastrophic Rupture
678 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
679 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
680 Release from hole with eff. diam. of 25 mm (1 in) diam.
681 Release from hole with eff. diam. of 50 mm (2 in) diam.
682 Release from hole with eff. diam. of 100 mm (4 in) diam.
683 Expansion Joint‐ Cryogenic: 12"≤D<20"
684 Catastrophic Rupture
685 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
686 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
687 Release from hole with eff. diam. of 25 mm (1 in) diam.
688 Release from hole with eff. diam. of 50 mm (2 in) diam.
689 Release from hole with eff. diam. of 100 mm (4 in) diam.
690 Expansion Joint‐ Non Cryogenic: 12"≤D<20"
691 Catastrophic Rupture
692 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
693 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
694 Release from hole with eff. diam. of 25 mm (1 in) diam.
695 Release from hole with eff. diam. of 50 mm (2 in) diam.
696 Release from hole with eff. diam. of 100 mm (4 in) diam.
697 Expansion Joint: 20"≤D<40"
698 Catastrophic Rupture
699 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
700 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
701 Release from hole with eff. diam. of 25 mm (1 in) diam.
702 Release from hole with eff. diam. of 50 mm (2 in) diam.
703 Release from hole with eff. diam. of 100 mm (4 in) diam.
704 Expansion Joint‐ Cryogenic: 20"≤D<40"
705 Catastrophic Rupture
706 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
707 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
708 Release from hole with eff. diam. of 25 mm (1 in) diam.
709 Release from hole with eff. diam. of 50 mm (2 in) diam.
710 Release from hole with eff. diam. of 100 mm (4 in) diam.
711 Expansion Joint‐ Non Cryogenic: 20"≤D<40"
712 Catastrophic Rupture
713 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
714 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
715 Release from hole with eff. diam. of 25 mm (1 in) diam.
716 Release from hole with eff. diam. of 50 mm (2 in) diam.
717 Release from hole with eff. diam. of 100 mm (4 in) diam.
172 Gasket Flange‐ All diameters
173 3.00E‐02 1.95E‐05 1.98E‐05 2.41E‐05 2.46E‐05 1.45E‐05 1.47E‐05 Catastrophic Rupture
367 9.47E‐05 9.47E‐05 9.47E‐05 9.47E‐05 9.47E‐05 9.47E‐05 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
174 5.15E‐05 5.15E‐05 5.15E‐05 5.15E‐05 5.15E‐05 5.15E‐05 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
368 7.49E‐06 7.49E‐06 7.39E‐06 7.39E‐06 7.66E‐06 7.66E‐06 Release from hole with eff. diam. of 25 mm (1 in) diam.
369 6.24E‐06 6.24E‐06 6.00E‐06 5.99E‐06 6.68E‐06 6.68E‐06 Release from hole with eff. diam. of 50 mm (2 in) diam.
370 2.66E‐06 2.66E‐06 2.66E‐06 2.66E‐06 2.66E‐06 2.66E‐06 Release from hole with eff. diam. of 100 mm (4 in) diam.
175 Gasket Flange‐ Cryogenic: All Diameters
176 Catastrophic Rupture
371 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
177 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
372 Release from hole with eff. diam. of 25 mm (1 in) diam.
373 Release from hole with eff. diam. of 50 mm (2 in) diam.
374 Release from hole with eff. diam. of 100 mm (4 in) diam.
178 Gasket Flange‐ Non Cryogenic: All Diameters
179 6.28E‐06 6.21E‐06 6.30E‐06 6.21E‐06 6.28E‐06 6.21E‐06 Catastrophic Rupture
375 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
180 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
376 Release from hole with eff. diam. of 25 mm (1 in) diam.
377 Release from hole with eff. diam. of 50 mm (2 in) diam.
378 Release from hole with eff. diam. of 100 mm (4 in) diam.
718 Gasket Flange: 2"≤D<6"
719 Catastrophic Rupture
720 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
721 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
722 Release from hole with eff. diam. of 25 mm (1 in) diam.
723 Release from hole with eff. diam. of 50 mm (2 in) diam.
724 Release from hole with eff. diam. of 100 mm (4 in) diam.
725 Gasket Flange‐ Cryogenic: 2"≤D<6"
726 Catastrophic Rupture
727 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
728 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
729 Release from hole with eff. diam. of 25 mm (1 in) diam.
730 Release from hole with eff. diam. of 50 mm (2 in) diam.
731 Release from hole with eff. diam. of 100 mm (4 in) diam.
732 Gasket Flange‐ Non Cryogenic: 2"≤D<6"
733 Catastrophic Rupture
734 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
735 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
736 Release from hole with eff. diam. of 25 mm (1 in) diam.
737 Release from hole with eff. diam. of 50 mm (2 in) diam.
738 Release from hole with eff. diam. of 100 mm (4 in) diam.
739 Gasket Flange: 6"≤D<12"
740 Catastrophic Rupture
741 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
742 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
743 Release from hole with eff. diam. of 25 mm (1 in) diam.
744 Release from hole with eff. diam. of 50 mm (2 in) diam.
745 Release from hole with eff. diam. of 100 mm (4 in) diam.
746 Gasket Flange‐ Cryogenic: 6"≤D<12"
747 Catastrophic Rupture
748 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
749 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
750 Release from hole with eff. diam. of 25 mm (1 in) diam.
751 Release from hole with eff. diam. of 50 mm (2 in) diam.
752 Release from hole with eff. diam. of 100 mm (4 in) diam.
753 Gasket Flange‐ Non Cryogenic: 6"≤D<12"
754 Catastrophic Rupture
755 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
756 Release from hole with eff. diam. of 10 mm (0.4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 147
Appendix D: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied
and Using All References "Specifically Included in Analysis" in Table 11

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
757 Release from hole with eff. diam. of 25 mm (1 in) diam.
758 Release from hole with eff. diam. of 50 mm (2 in) diam.
759 Release from hole with eff. diam. of 100 mm (4 in) diam.
760 Gasket Flange: 12"≤D<20"
761 Catastrophic Rupture
762 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
763 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
764 Release from hole with eff. diam. of 25 mm (1 in) diam.
765 Release from hole with eff. diam. of 50 mm (2 in) diam.
766 Release from hole with eff. diam. of 100 mm (4 in) diam.
767 Gasket Flange‐ Cryogenic: 12"≤D<20"
768 Catastrophic Rupture
769 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
770 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
771 Release from hole with eff. diam. of 25 mm (1 in) diam.
772 Release from hole with eff. diam. of 50 mm (2 in) diam.
773 Release from hole with eff. diam. of 100 mm (4 in) diam.
774 Gasket Flange‐ Non Cryogenic: 12"≤D<20"
775 Catastrophic Rupture
776 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
777 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
778 Release from hole with eff. diam. of 25 mm (1 in) diam.
779 Release from hole with eff. diam. of 50 mm (2 in) diam.
780 Release from hole with eff. diam. of 100 mm (4 in) diam.
781 Gasket Flange: 20"≤D<40"
782 Catastrophic Rupture
783 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
784 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
785 Release from hole with eff. diam. of 25 mm (1 in) diam.
786 Release from hole with eff. diam. of 50 mm (2 in) diam.
787 Release from hole with eff. diam. of 100 mm (4 in) diam.
788 Gasket Flange‐ Cryogenic: 20"≤D<40"
789 Catastrophic Rupture
790 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
791 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
792 Release from hole with eff. diam. of 25 mm (1 in) diam.
793 Release from hole with eff. diam. of 50 mm (2 in) diam.
794 Release from hole with eff. diam. of 100 mm (4 in) diam.
795 Gasket Flange‐ Non Cryogenic: 20"≤D<40"
796 Catastrophic Rupture
797 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
798 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
799 Release from hole with eff. diam. of 25 mm (1 in) diam.
800 Release from hole with eff. diam. of 50 mm (2 in) diam.
801 Release from hole with eff. diam. of 100 mm (4 in) diam.
181 Piping:  All diameters
182 2.24E‐06 3.04E‐06 2.01E‐06 2.50E‐06 2.86E‐06 4.65E‐06 Catastrophic Rupture
183 1.75E‐06 1.75E‐06 1.58E‐06 1.58E‐06 2.13E‐06 2.13E‐06 Release from hole with eff. diam. of 1/3 diam.
184 8.44E‐06 1.02E‐05 7.86E‐06 9.29E‐06 9.65E‐06 1.22E‐05 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
185 9.51E‐06 1.23E‐05 2.30E‐05 2.30E‐05 1.05E‐05 1.28E‐05 2.30E‐05 2.30E‐05 1.08E‐05 1.48E‐05 2.30E‐05 2.30E‐05 Release from hole with eff. diam. of 25mm (1 in)
186 Piping:  d < 50mm (2‐inch)
187 1.00E‐06 9.95E‐06 9.90E‐06 8.47E‐06 8.37E‐06 1.37E‐05 1.38E‐05 Catastrophic Rupture
188 Release from hole with eff. diam. of 1/3 diam.
189 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
190 5.00E‐06 Release from hole with eff. diam. of 25mm (1 in)
191 Piping:  d < 50mm (2‐inch) ‐ Non‐Cryogenic
192 1.21E‐05 1.21E‐05 1.12E‐05 1.12E‐05 1.48E‐05 1.48E‐05 Catastrophic Rupture
193 Release from hole with eff. diam. of 1/3 diam.
194 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
195 Release from hole with eff. diam. of 25mm (1 in)
196 Piping:  d < 50mm (2‐inch) ‐ Cryogenic
197 Catastrophic Rupture
198 Release from hole with eff. diam. of 1/3 diam.
199 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
200 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 Release from hole with eff. diam. of 25mm (1 in)
201 Piping:  d < 50mm (2‐inch) ‐ Cryogenic ‐ VJ PIPING
202 Catastrophic Rupture
203 Release from hole with eff. diam. of 1/3 diam.
204 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
205 Release from hole with eff. diam. of 25mm (1 in)
206 Piping:  d < 50mm (2‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
207 Catastrophic Rupture
208 Release from hole with eff. diam. of 1/3 diam.
209 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
210 Release from hole with eff. diam. of 25mm (1 in)
211 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch)
212 5.00E‐07 1.85E‐06 2.22E‐06 1.66E‐06 1.84E‐06 2.49E‐06 3.45E‐06 Catastrophic Rupture
213 Release from hole with eff. diam. of 1/3 diam.
214 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
215 2.00E‐06 Release from hole with eff. diam. of 25mm (1 in)
216 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Non‐Cryogenic
217 2.25E‐06 5.61E‐06 2.34E‐06 5.94E‐06 2.25E‐06 5.61E‐06 Catastrophic Rupture
218 Release from hole with eff. diam. of 1/3 diam.
219 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
220 Release from hole with eff. diam. of 25mm (1 in)
221 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic
222 Catastrophic Rupture
223 Release from hole with eff. diam. of 1/3 diam.
224 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
225 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 Release from hole with eff. diam. of 25mm (1 in)
226 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic ‐ VJ PIPING
227 Catastrophic Rupture
228 Release from hole with eff. diam. of 1/3 diam.
229 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
230 Release from hole with eff. diam. of 25mm (1 in)
231 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
232 Catastrophic Rupture
233 Release from hole with eff. diam. of 1/3 diam.
234 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
235 Release from hole with eff. diam. of 25mm (1 in)
236 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch)
237 2.00E‐07 7.84E‐07 9.85E‐07 7.40E‐07 8.22E‐07 9.68E‐07 1.51E‐06 Catastrophic Rupture
238 4.00E‐07 4.15E‐07 4.15E‐07 4.61E‐07 4.61E‐07 2.73E‐07 2.73E‐07 Release from hole with eff. diam. of 1/3 diam.
239 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
240 7.00E‐07 Release from hole with eff. diam. of 25mm (1 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 148
Appendix D: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied
and Using All References "Specifically Included in Analysis" in Table 11

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
241 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Non‐Cryogenic
242 9.51E‐07 2.98E‐06 9.11E‐07 3.16E‐06 9.51E‐07 2.98E‐06 Catastrophic Rupture
243 Release from hole with eff. diam. of 1/3 diam.
244 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
245 Release from hole with eff. diam. of 25mm (1 in)
246 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic
247 Catastrophic Rupture
248 Release from hole with eff. diam. of 1/3 diam.
249 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
250 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 Release from hole with eff. diam. of 25mm (1 in)
251 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic ‐ VJ PIPING
252 Catastrophic Rupture
253 Release from hole with eff. diam. of 1/3 diam.
254 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
255 Release from hole with eff. diam. of 25mm (1 in)
256 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
257 Catastrophic Rupture
258 Release from hole with eff. diam. of 1/3 diam.
259 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
260 Release from hole with eff. diam. of 25mm (1 in)
261 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch)
262 7.00E‐08 8.65E‐07 1.23E‐06 8.45E‐07 1.05E‐06 9.93E‐07 1.74E‐06 Catastrophic Rupture
263 2.00E‐07 8.91E‐08 8.91E‐08 1.05E‐07 1.05E‐07 4.96E‐08 4.97E‐08 Release from hole with eff. diam. of 1/3 diam.
264 4.00E‐07 1.12E‐07 9.62E‐08 1.12E‐07 9.60E‐08 1.12E‐07 9.61E‐08 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
265 5.00E‐07 1.93E‐06 1.93E‐06 1.72E‐06 1.72E‐06 2.47E‐06 2.47E‐06 Release from hole with eff. diam. of 25mm (1 in)
266 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Non‐Cryogenic
267 8.83E‐07 2.93E‐06 8.50E‐07 3.10E‐06 8.83E‐07 2.93E‐06 Catastrophic Rupture
268 Release from hole with eff. diam. of 1/3 diam.
269 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
270 Release from hole with eff. diam. of 25mm (1 in)
271 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic
272 Catastrophic Rupture
273 Release from hole with eff. diam. of 1/3 diam.
274 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
275 Release from hole with eff. diam. of 25mm (1 in)
276 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic ‐ VJ PIPING
277 Catastrophic Rupture
278 Release from hole with eff. diam. of 1/3 diam.
279 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
280 Release from hole with eff. diam. of 25mm (1 in)
281 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
282 Catastrophic Rupture
283 Release from hole with eff. diam. of 1/3 diam.
284 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
285 Release from hole with eff. diam. of 25mm (1 in)
286 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch)
287 2.00E‐08 7.64E‐07 9.99E‐07 7.61E‐07 8.23E‐07 9.32E‐07 1.51E‐06 Catastrophic Rupture
288 1.00E‐07 4.45E‐08 4.46E‐08 5.20E‐08 5.21E‐08 2.56E‐08 2.56E‐08 Release from hole with eff. diam. of 1/3 diam.
289 2.00E‐07 2.37E‐07 2.37E‐07 2.75E‐07 2.75E‐07 1.41E‐07 1.41E‐07 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
290 4.00E‐07 9.31E‐07 1.87E‐06 8.94E‐07 1.66E‐06 1.00E‐06 2.39E‐06 Release from hole with eff. diam. of 25mm (1 in)
291 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Non‐Cryogenic
292 8.45E‐07 2.68E‐06 8.72E‐07 2.84E‐06 8.45E‐07 2.68E‐06 Catastrophic Rupture
293 Release from hole with eff. diam. of 1/3 diam.
294 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
295 Release from hole with eff. diam. of 25mm (1 in)
296 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic
297 Catastrophic Rupture
298 Release from hole with eff. diam. of 1/3 diam.
299 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
300 Release from hole with eff. diam. of 25mm (1 in)
301 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic ‐ VJ PIPING
302 Catastrophic Rupture
303 Release from hole with eff. diam. of 1/3 diam.
304 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
305 Release from hole with eff. diam. of 25mm (1 in)
306 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
307 Catastrophic Rupture
308 Release from hole with eff. diam. of 1/3 diam.
309 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
310 Release from hole with eff. diam. of 25mm (1 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 149
 Appendix E: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied and Using All References
"Specifically Included in Analysis" in Table 11 Except "OREDA '15", "CCPS '89" and "INL CHEM '95"

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
1 Ambient Atm. Storage Tanks
2 3.45E‐06 4.89E‐06 3.45E‐06 4.90E‐06 4.74E‐06 4.90E‐06 Catastrophic Failure, Release to Atmosphere
3 9.98E‐05 9.98E‐05 9.98E‐05 9.98E‐05 9.98E‐05 9.98E‐05 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
4 2.50E‐03 2.50E‐03 2.50E‐03 2.50E‐03 2.50E‐03 2.50E‐03 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
5 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
6 Refrigerated Atm. Storage Tanks (Typ. Single Shell)
7 5.00E‐06 5.00E‐06 5.00E‐06 5.00E‐06 5.00E‐06 5.00E‐06 Catastrophic Failure, Release to Atmosphere
8 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
9 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
10 1.26E‐03 1.26E‐03 1.26E‐03 1.26E‐03 1.26E‐03 1.26E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
11 Cryogenic Atm. Storage Tanks
12 4.60E‐06 4.60E‐06 5.25E‐06 5.25E‐06 4.22E‐06 4.22E‐06 5.17E‐06 5.17E‐06 6.65E‐06 6.65E‐06 5.25E‐06 5.25E‐06 Catastrophic Failure, Release to Atmosphere
311 6.86E‐05 6.86E‐05 1.89E‐05 1.89E‐05 5.36E‐05 5.36E‐05 1.89E‐05 1.89E‐05 6.86E‐05 6.86E‐05 1.89E‐05 1.89E‐05 Catastrophic Failure of Tank Roof
13 2.90E‐05 2.90E‐05 1.01E‐06 1.01E‐06 2.06E‐05 2.06E‐05 1.01E‐06 1.01E‐06 2.90E‐05 2.90E‐05 1.01E‐06 1.01E‐06 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
14 2.39E‐05 2.39E‐05 3.06E‐06 3.06E‐06 1.75E‐05 1.75E‐05 3.06E‐06 3.06E‐06 2.39E‐05 2.39E‐05 3.06E‐06 3.06E‐06 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
15 1.44E‐03 1.44E‐03 1.44E‐03 1.44E‐03 1.44E‐03 1.44E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
16 Single Cont. Cryo. Atm. Storage Tanks
17 5.00E‐06 1.15E‐05 1.15E‐05 2.99E‐05 2.99E‐05 1.06E‐05 1.06E‐05 2.99E‐05 2.99E‐05 1.73E‐05 1.73E‐05 2.99E‐05 2.99E‐05 Catastrophic Failure, Release to Atmosphere
312 1.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 Catastrophic Failure of Tank Roof
18 8.00E‐05 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
19 2.00E‐04 8.00E‐05 8.00E‐05 8.00E‐05 8.00E‐05 8.00E‐05 8.00E‐05 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
20 1.00E‐04 1.19E‐03 1.19E‐03 1.19E‐03 1.19E‐03 1.19E‐03 1.19E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
21 Single Cont. Cryo. Atm. Storage Tanks ‐ HC Only (LNG, Ethane, Ethylene) ‐ Self‐Supp. Inner Tank
22 Catastrophic Failure, Release to Atmosphere
313 Catastrophic Failure of Tank Roof
23 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
24 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
25 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
26 Single Cont. Cryo. Atm. Storage Tanks ‐ LIN/LOX/LAR ‐ Self‐Supp. Inner Tank
27 2.20E‐05 2.20E‐05 2.20E‐05 2.20E‐05 2.20E‐05 2.20E‐05 Catastrophic Failure, Release to Atmosphere
314 Catastrophic Failure of Tank Roof
28 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
29 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
30 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
31 Double Cont. Cryo. Atm. Storage Tanks
32 5.00E‐07 3.23E‐08 3.23E‐08 7.44E‐08 7.44E‐08 3.19E‐08 3.18E‐08 7.39E‐08 7.39E‐08 4.64E‐08 4.64E‐08 7.44E‐08 7.44E‐08 Catastrophic Failure, Release to Atmosphere
315 1.00E‐04 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 Catastrophic Failure of Tank Roof
33 1.00E‐05 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
34 3.00E‐05 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
35 1.00E‐04 1.32E‐03 1.32E‐03 1.32E‐03 1.32E‐03 1.32E‐03 1.32E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
36 Full Cont. Cryo. Atm. Storage Tanks
37 1.00E‐08 2.31E‐07 2.31E‐07 7.47E‐07 7.47E‐07 2.22E‐07 2.22E‐07 7.27E‐07 7.28E‐07 4.02E‐07 4.02E‐07 7.47E‐07 7.47E‐07 Catastrophic Failure, Release to Atmosphere
316 4.00E‐05 2.44E‐10 2.40E‐10 2.43E‐10 2.39E‐10 2.53E‐10 2.43E‐10 2.41E‐10 2.41E‐10 2.41E‐10 2.48E‐10 2.43E‐10 2.44E‐10 Catastrophic Failure of Tank Roof
38 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
39 3.00E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
40 1.00E‐04 2.40E‐03 2.40E‐03 2.40E‐03 2.40E‐03 2.40E‐03 2.40E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
41 Membrane. Cryo. Atm. Storage Tanks
42 2.11E‐06 2.11E‐06 4.46E‐06 4.46E‐06 2.05E‐06 2.05E‐06 4.46E‐06 4.46E‐06 3.14E‐06 3.14E‐06 4.46E‐06 4.46E‐06 Catastrophic Failure, Release to Atmosphere
317 Catastrophic Failure of Tank Roof
43 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
44 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
45 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
46 Process Vessels, Distillation Columns, Heat Exchangers, and Condensers
47 5.00E‐06 1.67E‐02 1.94E‐02 2.72E‐01 2.72E‐01 2.09E‐02 2.42E‐02 2.72E‐01 2.72E‐01 2.70E‐02 3.13E‐02 2.72E‐01 2.72E‐01 Catastrophic Failure (Rupture)
48 1.00E‐04 1.22E‐03 1.22E‐03 1.26E‐03 1.26E‐03 1.13E‐03 1.13E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
802 3.26E‐04 3.77E‐04 2.94E‐04 3.56E‐04 4.00E‐04 4.27E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
803 1.16E‐04 1.16E‐04 1.17E‐04 1.17E‐04 1.22E‐04 1.22E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
804 9.97E‐05 1.23E‐04 8.89E‐05 1.16E‐04 1.22E‐04 1.34E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
49 Process Vessels incl Distillation Columns
50 2.85E‐05 3.31E‐05 2.63E‐05 2.98E‐05 3.93E‐05 4.96E‐05 Catastrophic Failure (Rupture)
51 1.94E‐04 1.94E‐04 1.66E‐04 1.66E‐04 2.99E‐04 2.99E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
805 9.47E‐05 1.06E‐04 8.11E‐05 9.24E‐05 1.41E‐04 1.51E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
806 3.73E‐05 3.73E‐05 3.87E‐05 3.87E‐05 4.29E‐05 4.28E‐05 Release from a hole with eff. diam. of 0.05m (2 in)
807 4.59E‐05 6.06E‐05 3.85E‐05 5.37E‐05 6.53E‐05 7.53E‐05 Release from a hole with eff. diam. of 0.10m (4 in)
52 PVs incl Dist. Columns ‐ Single Wall
53 2.81E‐05 3.44E‐05 2.50E‐05 2.98E‐05 4.09E‐05 5.49E‐05 Catastrophic Failure (Rupture)
54 2.22E‐04 2.22E‐04 1.93E‐04 1.93E‐04 3.28E‐04 3.28E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
808 9.47E‐05 1.06E‐04 8.11E‐05 9.24E‐05 1.41E‐04 1.51E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
809 3.73E‐05 3.73E‐05 3.87E‐05 3.87E‐05 4.29E‐05 4.29E‐05 Release from a hole with eff. diam. of 0.05m (2 in)
810 4.59E‐05 6.06E‐05 3.85E‐05 5.37E‐05 6.53E‐05 7.53E‐05 Release from a hole with eff. diam. of 0.10m (4 in)
318 Pressure Storage Vessels/Tanks (Single Wall)
319 2.41E‐06 2.38E‐06 2.55E‐06 2.37E‐06 2.37E‐06 2.28E‐06 Catastrophic Failure (Rupture)
320 1.10E‐05 1.10E‐05 1.11E‐05 1.11E‐05 1.07E‐05 1.07E‐05 Release from a hole with eff. diam. of 0.01m (0.4 in)
380 4.57E‐06 4.57E‐06 4.52E‐06 4.52E‐06 4.69E‐06 4.69E‐06 Release from a hole with eff. diam. of 0.025m (1 in)
381 4.99E‐06 4.99E‐06 4.99E‐06 4.99E‐06 4.99E‐06 4.99E‐06 Release from a hole with eff. diam. of 0.05m (2 in)
382 4.17E‐06 4.17E‐06 4.17E‐06 4.17E‐06 4.17E‐06 4.17E‐06 Release from a hole with eff. diam. of 0.10m (4 in)
321 Process Vessels (Single Wall)
322 8.46E‐05 8.46E‐05 7.47E‐05 7.47E‐05 1.17E‐04 1.17E‐04 Catastrophic Failure (Rupture)
323 4.28E‐04 4.28E‐04 3.86E‐04 3.86E‐04 5.29E‐04 5.29E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
383 Some cells shaded to  Release from a hole with eff. diam. of 0.025m (1 in)
384 protect copyrights or  Release from a hole with eff. diam. of 0.05m (2 in)
386 licensed software Release from a hole with eff. diam. of 0.10m (4 in)
55 Mole Sieve Vessel (Single Wall)
56 Catastrophic Failure (Rupture)
57 Release from a hole with eff. diam. of 0.01m (0.4 in)
387 Release from a hole with eff. diam. of 0.025m (1 in)
388 Release from a hole with eff. diam. of 0.05m (2 in)
389 Release from a hole with eff. diam. of 0.10m (4 in)
58 Distillation Columns (Single Wall)
59 Catastrophic Failure (Rupture)
60 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
390 Release from a hole with eff. diam. of 0.025m (1 in)
391 Release from a hole with eff. diam. of 0.05m (2 in)
392 Release from a hole with eff. diam. of 0.10m (4 in)
61 Separator (Single Wall)
62 Catastrophic Failure (Rupture)
63 Release from a hole with eff. diam. of 0.01m (0.4 in)
393 Release from a hole with eff. diam. of 0.025m (1 in)
394 Release from a hole with eff. diam. of 0.05m (2 in)
395 Release from a hole with eff. diam. of 0.10m (4 in)
64 PVs incl Dist. Columns ‐ Double Wall (cryogenic)
65 Catastrophic Failure (Rupture)
66 Release from a hole with eff. diam. of 0.01m (0.4 in)
396 Release from a hole with eff. diam. of 0.025m (1 in)
397 Release from a hole with eff. diam. of 0.05m (2 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 150
 Appendix E: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied and Using All References
"Specifically Included in Analysis" in Table 11 Except "OREDA '15", "CCPS '89" and "INL CHEM '95"

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
398 Release from a hole with eff. diam. of 0.10m (4 in)
70 ISO Containers
71 2.95E‐06 2.95E‐06 2.95E‐06 2.95E‐06 2.95E‐06 2.95E‐06 Catastrophic Failure (Rupture)
72 6.01E‐05 6.01E‐05 6.01E‐05 6.01E‐05 6.01E‐05 6.01E‐05 Release from a hole with eff. diam. of 0.01m (0.4 in)
399 Release from a hole with eff. diam. of 0.025m (1 in)
400 Release from a hole with eff. diam. of 0.05m (2 in)
401 Release from a hole with eff. diam. of 0.10m (4 in)
73 Heat Exchangers  incl. Condensers
74 3.35E‐02 3.84E‐02 2.72E‐01 2.72E‐01 4.24E‐02 4.94E‐02 2.72E‐01 2.72E‐01 4.67E‐02 5.07E‐02 2.72E‐01 2.72E‐01 Catastrophic Failure (Rupture)
75 1.91E‐03 1.91E‐03 2.09E‐03 2.09E‐03 1.52E‐03 1.52E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
402 4.64E‐04 5.44E‐04 4.37E‐04 5.44E‐04 5.11E‐04 5.44E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
403 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
404 1.17E‐04 1.41E‐04 1.07E‐04 1.36E‐04 1.38E‐04 1.49E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
76 Fired Heat Exchangers
77 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 Catastrophic Failure (Rupture)
78 Release from a hole with eff. diam. of 0.01m (0.4 in)
405 Release from a hole with eff. diam. of 0.025m (1 in)
406 Release from a hole with eff. diam. of 0.05m (2 in)
407 Release from a hole with eff. diam. of 0.10m (4 in)
324 Submerged Combustion Vaporizers
325 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 Catastrophic Failure (Rupture)
326 Release from a hole with eff. diam. of 0.01m (0.4 in)
408 Release from a hole with eff. diam. of 0.025m (1 in)
409 Release from a hole with eff. diam. of 0.05m (2 in)
410 Release from a hole with eff. diam. of 0.10m (4 in)
79 Non‐Fired Heat Exchangers incl. Condensers
80 1.16E‐04 1.34E‐04 1.02E‐04 1.20E‐04 1.55E‐04 1.67E‐04 Catastrophic Failure (Rupture)
81 2.02E‐03 2.02E‐03 2.25E‐03 2.25E‐03 1.56E‐03 1.56E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
411 4.64E‐04 5.44E‐04 4.37E‐04 5.44E‐04 5.11E‐04 5.44E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
412 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 1.56E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
413 1.17E‐04 1.41E‐04 1.07E‐04 1.36E‐04 1.38E‐04 1.49E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
82 Shell & Tube Heat Exchangers
83 7.50E‐05 8.30E‐05 6.83E‐05 7.66E‐05 9.51E‐05 9.87E‐05 Catastrophic Failure (Rupture)
84 1.16E‐03 1.16E‐03 1.24E‐03 1.24E‐03 9.97E‐04 9.97E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
414 3.42E‐04 3.94E‐04 3.25E‐04 3.94E‐04 3.73E‐04 3.94E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
415 1.01E‐04 1.01E‐04 1.01E‐04 1.01E‐04 1.01E‐04 1.01E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
416 6.91E‐05 8.13E‐05 6.25E‐05 7.72E‐05 8.29E‐05 8.90E‐05 Release from a hole with eff. diam. of 0.10m (4 in)
85 Tube‐side Heat Exchangers (HC in tube)
86 9.74E‐05 8.85E‐05 9.55E‐05 8.42E‐05 1.08E‐04 9.79E‐05 Catastrophic Failure (Rupture)
87 Release from a hole with eff. diam. of 0.01m (0.4 in)
417 Release from a hole with eff. diam. of 0.025m (1 in)
418 Release from a hole with eff. diam. of 0.05m (2 in)
419 Release from a hole with eff. diam. of 0.10m (4 in)
88 Shell‐side Heat Exchangers (HC in shell)
89 7.25E‐05 1.03E‐04 6.37E‐05 9.75E‐05 9.37E‐05 1.14E‐04 Catastrophic Failure (Rupture)
90 Release from a hole with eff. diam. of 0.01m (0.4 in)
420 Release from a hole with eff. diam. of 0.025m (1 in)
421 Release from a hole with eff. diam. of 0.05m (2 in)
422 Release from a hole with eff. diam. of 0.10m (4 in)
91 Plate Heat Exchangers
92 2.36E‐04 2.36E‐04 1.99E‐04 1.99E‐04 3.48E‐04 3.48E‐04 Catastrophic Failure (Rupture)
93 Release from a hole with eff. diam. of 0.01m (0.4 in)
423 Release from a hole with eff. diam. of 0.025m (1 in)
424 Release from a hole with eff. diam. of 0.05m (2 in)
425 Release from a hole with eff. diam. of 0.10m (4 in)
94 Air Cooled (Fin Fan) Heat Exchangers
95 3.00E‐05 4.87E‐05 2.59E‐05 4.87E‐05 3.95E‐05 4.87E‐05 Catastrophic Failure (Rupture)
96 Release from a hole with eff. diam. of 0.01m (0.4 in)
426 Release from a hole with eff. diam. of 0.025m (1 in)
427 Release from a hole with eff. diam. of 0.05m (2 in)
428 Release from a hole with eff. diam. of 0.10m (4 in)
97 Printed Heat Exchangers
98 Catastrophic Failure (Rupture)
99 Release from a hole with eff. diam. of 0.01m (0.4 in)
429 Release from a hole with eff. diam. of 0.025m (1 in)
430 Release from a hole with eff. diam. of 0.05m (2 in)
431 Release from a hole with eff. diam. of 0.10m (4 in)
100 Plate‐Fin Heat Exchangers
101 Catastrophic Failure (Rupture)
102 Release from a hole with eff. diam. of 0.01m (0.4 in)
432 Release from a hole with eff. diam. of 0.025m (1 in)
433 Release from a hole with eff. diam. of 0.05m (2 in)
434 Release from a hole with eff. diam. of 0.10m (4 in)
103 Ambient Vaporizer Heat Exchangers
104 Catastrophic Failure (Rupture)
105 Release from a hole with eff. diam. of 0.01m (0.4 in)
435 Release from a hole with eff. diam. of 0.025m (1 in)
436 Release from a hole with eff. diam. of 0.05m (2 in)
437 Release from a hole with eff. diam. of 0.10m (4 in)
106 Truck Transfer Arm
107 3.00E‐04 1.51E‐05 1.51E‐05 1.51E‐05 1.51E‐05 1.51E‐05 1.51E‐05 Rupture of transfer arm
108 3.00E‐03 2.09E‐04 2.09E‐04 2.09E‐04 2.09E‐04 2.09E‐04 2.09E‐04 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
109 Truck Transfer Arm ‐ Cryogenic
110 Rupture of transfer arm
111 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
112 Truck Transfer Arm ‐ Non‐cryogenic
113 Rupture of transfer arm
114 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
115 Ship Transfer Arm
116 2.00E‐05 1.03E‐03 1.11E‐03 1.02E‐03 1.11E‐03 1.03E‐03 1.11E‐03 Rupture of transfer arm
117 2.00E‐04 1.89E‐02 1.89E‐02 1.89E‐02 1.89E‐02 1.89E‐02 1.89E‐02 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
118 Ship Transfer Arm ‐ Cryogenic
119 Rupture of transfer arm
120 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
121 Ship Transfer Arm ‐ Non‐cryogenic
122 Rupture of transfer arm
123 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
124 Truck Transfer Hose
125 4.00E‐02 4.90E‐03 4.90E‐03 5.09E‐03 5.09E‐03 5.10E‐03 5.10E‐03 Rupture of transfer hose
126 4.00E‐01 2.82E‐02 2.82E‐02 2.82E‐02 2.82E‐02 2.82E‐02 2.82E‐02 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
127 Truck Transfer Hose ‐ Cryogenic
128 Rupture of transfer hose
129 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
130 Truck Transfer Hose ‐ Non‐cryogenic
131 4.00E‐04 4.00E‐04 4.00E‐04 4.00E‐04 4.00E‐04 4.00E‐04 Rupture of transfer hose

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 151
 Appendix E: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied and Using All References
"Specifically Included in Analysis" in Table 11 Except "OREDA '15", "CCPS '89" and "INL CHEM '95"

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
132 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
133 Ship Transfer Hose
134 2.71E‐03 4.00E‐03 2.61E‐03 4.00E‐03 2.71E‐03 4.00E‐03 Rupture of transfer hose
135 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
136 Ship Transfer Hose ‐ Cryogenic
137 Rupture of transfer hose
138 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
139 Ship Transfer Hose ‐ Non‐cryogenic
140 Rupture of transfer hose
141 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
142 Valve: All diameters
143 9.00E‐06 1.44E‐05 1.44E‐05 1.52E‐05 1.53E‐05 1.47E‐05 1.47E‐05 Catastrophic Rupture
327 2.01E‐04 2.01E‐04 2.01E‐04 2.01E‐04 2.01E‐04 2.01E‐04 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
144 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
328 Release from hole with eff. diam. of 25 mm (1 in) diam.
329 Release from hole with eff. diam. of 50 mm (2 in) diam.
330 Release from hole with eff. diam. of 100 mm (4 in) diam.
145 Manual Valves: All diameters
146 8.54E‐06 8.05E‐06 9.60E‐06 8.96E‐06 8.54E‐06 8.05E‐06 Catastrophic Rupture
331 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
147 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
332 Release from hole with eff. diam. of 25 mm (1 in) diam.
333 Release from hole with eff. diam. of 50 mm (2 in) diam.
334 Release from hole with eff. diam. of 100 mm (4 in) diam.
148 Manual Valves ‐ Cryogenic: All diameters
149 Catastrophic Rupture
335 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
150 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
336 Release from hole with eff. diam. of 25 mm (1 in) diam.
337 Release from hole with eff. diam. of 50 mm (2 in) diam.
338 Release from hole with eff. diam. of 100 mm (4 in) diam.
151 Manual Valves ‐ Non‐cryogenic: All diameters
152 6.38E‐06 5.82E‐06 6.61E‐06 5.83E‐06 6.38E‐06 5.82E‐06 Catastrophic Rupture
339 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
153 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
340 Release from hole with eff. diam. of 25 mm (1 in) diam.
341 Release from hole with eff. diam. of 50 mm (2 in) diam.
342 Release from hole with eff. diam. of 100 mm (4 in) diam.
154 Actuated Valves: All diameters
155 2.11E‐05 2.18E‐05 2.23E‐05 2.33E‐05 2.11E‐05 2.18E‐05 Catastrophic Rupture
343 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
156 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
344 Release from hole with eff. diam. of 25 mm (1 in) diam.
345 Release from hole with eff. diam. of 50 mm (2 in) diam.
346 Release from hole with eff. diam. of 100 mm (4 in) diam.
157 Actuated Valves ‐ Cryogenic: All diameters
158 Catastrophic Rupture
347 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
159 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
348 Release from hole with eff. diam. of 25 mm (1 in) diam.
349 Release from hole with eff. diam. of 50 mm (2 in) diam.
350 Release from hole with eff. diam. of 100 mm (4 in) diam.
160 Actuated Valves ‐ Non‐cryogenic: All diameters
161 2.11E‐05 2.18E‐05 2.23E‐05 2.33E‐05 2.11E‐05 2.18E‐05 Catastrophic Rupture
351 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
162 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
352 Release from hole with eff. diam. of 25 mm (1 in) diam.
353 Release from hole with eff. diam. of 50 mm (2 in) diam.
354 Release from hole with eff. diam. of 100 mm (4 in) diam.
438 Valve: 2"≤D<6"
439 Catastrophic Rupture
440 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
441 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
442 Release from hole with eff. diam. of 25 mm (1 in) diam.
443 Release from hole with eff. diam. of 50 mm (2 in) diam.
444 Release from hole with eff. diam. of 100 mm (4 in) diam.
445 Manual Valves: 2"≤D<6"
446 Catastrophic Rupture
447 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
448 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
449 Release from hole with eff. diam. of 25 mm (1 in) diam.
450 Release from hole with eff. diam. of 50 mm (2 in) diam.
451 Release from hole with eff. diam. of 100 mm (4 in) diam.
452 Manual Valves ‐ Cryogenic: 2"≤D<6"
453 Catastrophic Rupture
454 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
455 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
456 Release from hole with eff. diam. of 25 mm (1 in) diam.
457 Release from hole with eff. diam. of 50 mm (2 in) diam.
458 Release from hole with eff. diam. of 100 mm (4 in) diam.
459 Manual Valves ‐ Non‐cryogenic: 2"≤D<6"
460 Catastrophic Rupture
461 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
462 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
463 Release from hole with eff. diam. of 25 mm (1 in) diam.
464 Release from hole with eff. diam. of 50 mm (2 in) diam.
465 Release from hole with eff. diam. of 100 mm (4 in) diam.
466 Actuated Valves: 2"≤D<6"
467 Catastrophic Rupture
468 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
469 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
470 Release from hole with eff. diam. of 25 mm (1 in) diam.
471 Release from hole with eff. diam. of 50 mm (2 in) diam.
472 Release from hole with eff. diam. of 100 mm (4 in) diam.
473 Actuated Valves ‐ Cryogenic: 2"≤D<6"
474 Catastrophic Rupture
475 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
476 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
477 Release from hole with eff. diam. of 25 mm (1 in) diam.
478 Release from hole with eff. diam. of 50 mm (2 in) diam.
479 Release from hole with eff. diam. of 100 mm (4 in) diam.
480 Actuated Valves ‐ Non‐cryogenic: 2"≤D<6"
481 Catastrophic Rupture
482 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
483 Release from hole with eff. diam. of 10 mm (0.4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 152
 Appendix E: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied and Using All References
"Specifically Included in Analysis" in Table 11 Except "OREDA '15", "CCPS '89" and "INL CHEM '95"

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
484 Release from hole with eff. diam. of 25 mm (1 in) diam.
485 Release from hole with eff. diam. of 50 mm (2 in) diam.
486 Release from hole with eff. diam. of 100 mm (4 in) diam.
487 Valve: 6"≤D<12"
488 Catastrophic Rupture
489 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
490 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
491 Release from hole with eff. diam. of 25 mm (1 in) diam.
492 Release from hole with eff. diam. of 50 mm (2 in) diam.
493 Release from hole with eff. diam. of 100 mm (4 in) diam.
494 Manual Valves: 6"≤D<12"
495 Catastrophic Rupture
496 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
497 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
498 Release from hole with eff. diam. of 25 mm (1 in) diam.
499 Release from hole with eff. diam. of 50 mm (2 in) diam.
500 Release from hole with eff. diam. of 100 mm (4 in) diam.
501 Manual Valves ‐ Cryogenic: 6"≤D<12"
502 Catastrophic Rupture
503 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
504 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
505 Release from hole with eff. diam. of 25 mm (1 in) diam.
506 Release from hole with eff. diam. of 50 mm (2 in) diam.
507 Release from hole with eff. diam. of 100 mm (4 in) diam.
508 Manual Valves ‐ Non‐cryogenic: 6"≤D<12"
509 Catastrophic Rupture
510 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
511 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
512 Release from hole with eff. diam. of 25 mm (1 in) diam.
513 Release from hole with eff. diam. of 50 mm (2 in) diam.
514 Release from hole with eff. diam. of 100 mm (4 in) diam.
515 Actuated Valves: 6"≤D<12"
516 Catastrophic Rupture
517 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
518 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
519 Release from hole with eff. diam. of 25 mm (1 in) diam.
520 Release from hole with eff. diam. of 50 mm (2 in) diam.
521 Release from hole with eff. diam. of 100 mm (4 in) diam.
522 Actuated Valves ‐ Cryogenic: 6"≤D<12"
523 Catastrophic Rupture
524 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
525 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
526 Release from hole with eff. diam. of 25 mm (1 in) diam.
527 Release from hole with eff. diam. of 50 mm (2 in) diam.
528 Release from hole with eff. diam. of 100 mm (4 in) diam.
529 Actuated Valves ‐ Non‐cryogenic: 6"≤D<12"
530 Catastrophic Rupture
531 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
532 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
533 Release from hole with eff. diam. of 25 mm (1 in) diam.
534 Release from hole with eff. diam. of 50 mm (2 in) diam.
535 Release from hole with eff. diam. of 100 mm (4 in) diam.
536 Valve: 12"≤D<20"
537 Catastrophic Rupture
538 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
539 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
540 Release from hole with eff. diam. of 25 mm (1 in) diam.
541 Release from hole with eff. diam. of 50 mm (2 in) diam.
542 Release from hole with eff. diam. of 100 mm (4 in) diam.
543 Manual Valves: 12"≤D<20"
544 Catastrophic Rupture
545 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
546 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
547 Release from hole with eff. diam. of 25 mm (1 in) diam.
548 Release from hole with eff. diam. of 50 mm (2 in) diam.
549 Release from hole with eff. diam. of 100 mm (4 in) diam.
550 Manual Valves ‐ Cryogenic: 12"≤D<20"
551 Catastrophic Rupture
552 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
553 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
554 Release from hole with eff. diam. of 25 mm (1 in) diam.
555 Release from hole with eff. diam. of 50 mm (2 in) diam.
556 Release from hole with eff. diam. of 100 mm (4 in) diam.
557 Manual Valves ‐ Non‐cryogenic: 12"≤D<20"
558 Catastrophic Rupture
559 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
560 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
561 Release from hole with eff. diam. of 25 mm (1 in) diam.
562 Release from hole with eff. diam. of 50 mm (2 in) diam.
563 Release from hole with eff. diam. of 100 mm (4 in) diam.
564 Actuated Valves: 12"≤D<20"
565 Catastrophic Rupture
566 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
567 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
568 Release from hole with eff. diam. of 25 mm (1 in) diam.
569 Release from hole with eff. diam. of 50 mm (2 in) diam.
570 Release from hole with eff. diam. of 100 mm (4 in) diam.
571 Actuated Valves ‐ Cryogenic: 12"≤D<20"
572 Catastrophic Rupture
573 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
574 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
575 Release from hole with eff. diam. of 25 mm (1 in) diam.
576 Release from hole with eff. diam. of 50 mm (2 in) diam.
577 Release from hole with eff. diam. of 100 mm (4 in) diam.
578 Actuated Valves ‐ Non‐cryogenic: 12"≤D<20"
579 Catastrophic Rupture
580 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
581 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
582 Release from hole with eff. diam. of 25 mm (1 in) diam.
583 Release from hole with eff. diam. of 50 mm (2 in) diam.
584 Release from hole with eff. diam. of 100 mm (4 in) diam.
585 Valve: 20"≤D<40"
586 Catastrophic Rupture
587 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
588 Release from hole with eff. diam. of 10 mm (0.4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 153
 Appendix E: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied and Using All References
"Specifically Included in Analysis" in Table 11 Except "OREDA '15", "CCPS '89" and "INL CHEM '95"

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
589 Release from hole with eff. diam. of 25 mm (1 in) diam.
590 Release from hole with eff. diam. of 50 mm (2 in) diam.
591 Release from hole with eff. diam. of 100 mm (4 in) diam.
592 Manual Valves: 20"≤D<40"
593 Catastrophic Rupture
594 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
595 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
596 Release from hole with eff. diam. of 25 mm (1 in) diam.
597 Release from hole with eff. diam. of 50 mm (2 in) diam.
598 Release from hole with eff. diam. of 100 mm (4 in) diam.
599 Manual Valves ‐ Cryogenic: 20"≤D<40"
600 Catastrophic Rupture
601 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
602 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
603 Release from hole with eff. diam. of 25 mm (1 in) diam.
604 Release from hole with eff. diam. of 50 mm (2 in) diam.
605 Release from hole with eff. diam. of 100 mm (4 in) diam.
606 Manual Valves ‐ Non‐cryogenic: 20"≤D<40"
607 Catastrophic Rupture
608 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
609 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
610 Release from hole with eff. diam. of 25 mm (1 in) diam.
611 Release from hole with eff. diam. of 50 mm (2 in) diam.
612 Release from hole with eff. diam. of 100 mm (4 in) diam.
613 Actuated Valves: 20"≤D<40"
614 Catastrophic Rupture
615 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
616 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
617 Release from hole with eff. diam. of 25 mm (1 in) diam.
618 Release from hole with eff. diam. of 50 mm (2 in) diam.
619 Release from hole with eff. diam. of 100 mm (4 in) diam.
620 Actuated Valves ‐ Cryogenic: 20"≤D<40"
621 Catastrophic Rupture
622 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
623 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
624 Release from hole with eff. diam. of 25 mm (1 in) diam.
625 Release from hole with eff. diam. of 50 mm (2 in) diam.
626 Release from hole with eff. diam. of 100 mm (4 in) diam.
627 Actuated Valves ‐ Non‐cryogenic: 20"≤D<40"
628 Catastrophic Rupture
629 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
630 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
631 Release from hole with eff. diam. of 25 mm (1 in) diam.
632 Release from hole with eff. diam. of 50 mm (2 in) diam.
633 Release from hole with eff. diam. of 100 mm (4 in) diam.
163 Expansion Joint‐ All diameters
164 4.00E‐03 1.06E‐03 1.06E‐03 1.06E‐03 1.06E‐03 1.29E‐03 1.29E‐03 Catastrophic Rupture
355 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
165 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
356 Release from hole with eff. diam. of 25 mm (1 in) diam.
357 Release from hole with eff. diam. of 50 mm (2 in) diam.
358 Release from hole with eff. diam. of 100 mm (4 in) diam.
166 Expansion Joint‐ Cryogenic: All diameters
167 Catastrophic Rupture
359 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
168 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
360 Release from hole with eff. diam. of 25 mm (1 in) diam.
361 Release from hole with eff. diam. of 50 mm (2 in) diam.
362 Release from hole with eff. diam. of 100 mm (4 in) diam.
169 Expansion Joint‐ Non Cryogenic: All diameters
170 Catastrophic Rupture
363 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
171 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
364 Release from hole with eff. diam. of 25 mm (1 in) diam.
365 Release from hole with eff. diam. of 50 mm (2 in) diam.
366 Release from hole with eff. diam. of 100 mm (4 in) diam.
634 Expansion Joint: 2"≤D<6"
635 Catastrophic Rupture
636 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
637 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
638 Release from hole with eff. diam. of 25 mm (1 in) diam.
639 Release from hole with eff. diam. of 50 mm (2 in) diam.
640 Release from hole with eff. diam. of 100 mm (4 in) diam.
641 Expansion Joint‐ Cryogenic: 2"≤D<6"
642 Catastrophic Rupture
643 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
644 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
645 Release from hole with eff. diam. of 25 mm (1 in) diam.
646 Release from hole with eff. diam. of 50 mm (2 in) diam.
647 Release from hole with eff. diam. of 100 mm (4 in) diam.
648 Expansion Joint‐ Non Cryogenic: 2"≤D<6"
649 Catastrophic Rupture
650 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
651 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
652 Release from hole with eff. diam. of 25 mm (1 in) diam.
653 Release from hole with eff. diam. of 50 mm (2 in) diam.
654 Release from hole with eff. diam. of 100 mm (4 in) diam.
655 Expansion Joint: 6"≤D<12"
656 Catastrophic Rupture
657 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
658 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
659 Release from hole with eff. diam. of 25 mm (1 in) diam.
660 Release from hole with eff. diam. of 50 mm (2 in) diam.
661 Release from hole with eff. diam. of 100 mm (4 in) diam.
662 Expansion Joint‐ Cryogenic: 6"≤D<12"
663 Catastrophic Rupture
664 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
665 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
666 Release from hole with eff. diam. of 25 mm (1 in) diam.
667 Release from hole with eff. diam. of 50 mm (2 in) diam.
668 Release from hole with eff. diam. of 100 mm (4 in) diam.
669 Expansion Joint‐ Non Cryogenic: 6"≤D<12"
670 Catastrophic Rupture
671 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
672 Release from hole with eff. diam. of 10 mm (0.4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 154
 Appendix E: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied and Using All References
"Specifically Included in Analysis" in Table 11 Except "OREDA '15", "CCPS '89" and "INL CHEM '95"

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
673 Release from hole with eff. diam. of 25 mm (1 in) diam.
674 Release from hole with eff. diam. of 50 mm (2 in) diam.
675 Release from hole with eff. diam. of 100 mm (4 in) diam.
676 Expansion Joint: 12"≤D<20"
677 Catastrophic Rupture
678 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
679 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
680 Release from hole with eff. diam. of 25 mm (1 in) diam.
681 Release from hole with eff. diam. of 50 mm (2 in) diam.
682 Release from hole with eff. diam. of 100 mm (4 in) diam.
683 Expansion Joint‐ Cryogenic: 12"≤D<20"
684 Catastrophic Rupture
685 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
686 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
687 Release from hole with eff. diam. of 25 mm (1 in) diam.
688 Release from hole with eff. diam. of 50 mm (2 in) diam.
689 Release from hole with eff. diam. of 100 mm (4 in) diam.
690 Expansion Joint‐ Non Cryogenic: 12"≤D<20"
691 Catastrophic Rupture
692 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
693 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
694 Release from hole with eff. diam. of 25 mm (1 in) diam.
695 Release from hole with eff. diam. of 50 mm (2 in) diam.
696 Release from hole with eff. diam. of 100 mm (4 in) diam.
697 Expansion Joint: 20"≤D<40"
698 Catastrophic Rupture
699 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
700 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
701 Release from hole with eff. diam. of 25 mm (1 in) diam.
702 Release from hole with eff. diam. of 50 mm (2 in) diam.
703 Release from hole with eff. diam. of 100 mm (4 in) diam.
704 Expansion Joint‐ Cryogenic: 20"≤D<40"
705 Catastrophic Rupture
706 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
707 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
708 Release from hole with eff. diam. of 25 mm (1 in) diam.
709 Release from hole with eff. diam. of 50 mm (2 in) diam.
710 Release from hole with eff. diam. of 100 mm (4 in) diam.
711 Expansion Joint‐ Non Cryogenic: 20"≤D<40"
712 Catastrophic Rupture
713 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
714 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
715 Release from hole with eff. diam. of 25 mm (1 in) diam.
716 Release from hole with eff. diam. of 50 mm (2 in) diam.
717 Release from hole with eff. diam. of 100 mm (4 in) diam.
172 Gasket Flange‐ All diameters
173 3.00E‐02 1.98E‐05 1.98E‐05 2.46E‐05 2.46E‐05 1.47E‐05 1.47E‐05 Catastrophic Rupture
367 9.47E‐05 9.47E‐05 9.47E‐05 9.47E‐05 9.47E‐05 9.47E‐05 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
174 5.15E‐05 5.15E‐05 5.15E‐05 5.15E‐05 5.15E‐05 5.15E‐05 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
368 7.49E‐06 7.49E‐06 7.39E‐06 7.39E‐06 7.66E‐06 7.66E‐06 Release from hole with eff. diam. of 25 mm (1 in) diam.
369 6.24E‐06 6.24E‐06 6.00E‐06 6.00E‐06 6.68E‐06 6.68E‐06 Release from hole with eff. diam. of 50 mm (2 in) diam.
370 2.66E‐06 2.66E‐06 2.66E‐06 2.66E‐06 2.66E‐06 2.66E‐06 Release from hole with eff. diam. of 100 mm (4 in) diam.
175 Gasket Flange‐ Cryogenic: All Diameters
176 Catastrophic Rupture
371 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
177 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
372 Release from hole with eff. diam. of 25 mm (1 in) diam.
373 Release from hole with eff. diam. of 50 mm (2 in) diam.
374 Release from hole with eff. diam. of 100 mm (4 in) diam.
178 Gasket Flange‐ Non Cryogenic: All Diameters
179 6.21E‐06 6.21E‐06 6.21E‐06 6.21E‐06 6.21E‐06 6.21E‐06 Catastrophic Rupture
375 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
180 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
376 Release from hole with eff. diam. of 25 mm (1 in) diam.
377 Release from hole with eff. diam. of 50 mm (2 in) diam.
378 Release from hole with eff. diam. of 100 mm (4 in) diam.
718 Gasket Flange: 2"≤D<6"
719 Catastrophic Rupture
720 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
721 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
722 Release from hole with eff. diam. of 25 mm (1 in) diam.
723 Release from hole with eff. diam. of 50 mm (2 in) diam.
724 Release from hole with eff. diam. of 100 mm (4 in) diam.
725 Gasket Flange‐ Cryogenic: 2"≤D<6"
726 Catastrophic Rupture
727 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
728 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
729 Release from hole with eff. diam. of 25 mm (1 in) diam.
730 Release from hole with eff. diam. of 50 mm (2 in) diam.
731 Release from hole with eff. diam. of 100 mm (4 in) diam.
732 Gasket Flange‐ Non Cryogenic: 2"≤D<6"
733 Catastrophic Rupture
734 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
735 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
736 Release from hole with eff. diam. of 25 mm (1 in) diam.
737 Release from hole with eff. diam. of 50 mm (2 in) diam.
738 Release from hole with eff. diam. of 100 mm (4 in) diam.
739 Gasket Flange: 6"≤D<12"
740 Catastrophic Rupture
741 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
742 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
743 Release from hole with eff. diam. of 25 mm (1 in) diam.
744 Release from hole with eff. diam. of 50 mm (2 in) diam.
745 Release from hole with eff. diam. of 100 mm (4 in) diam.
746 Gasket Flange‐ Cryogenic: 6"≤D<12"
747 Catastrophic Rupture
748 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
749 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
750 Release from hole with eff. diam. of 25 mm (1 in) diam.
751 Release from hole with eff. diam. of 50 mm (2 in) diam.
752 Release from hole with eff. diam. of 100 mm (4 in) diam.
753 Gasket Flange‐ Non Cryogenic: 6"≤D<12"
754 Catastrophic Rupture
755 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
756 Release from hole with eff. diam. of 10 mm (0.4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 155
 Appendix E: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied and Using All References
"Specifically Included in Analysis" in Table 11 Except "OREDA '15", "CCPS '89" and "INL CHEM '95"

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
757 Release from hole with eff. diam. of 25 mm (1 in) diam.
758 Release from hole with eff. diam. of 50 mm (2 in) diam.
759 Release from hole with eff. diam. of 100 mm (4 in) diam.
760 Gasket Flange: 12"≤D<20"
761 Catastrophic Rupture
762 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
763 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
764 Release from hole with eff. diam. of 25 mm (1 in) diam.
765 Release from hole with eff. diam. of 50 mm (2 in) diam.
766 Release from hole with eff. diam. of 100 mm (4 in) diam.
767 Gasket Flange‐ Cryogenic: 12"≤D<20"
768 Catastrophic Rupture
769 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
770 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
771 Release from hole with eff. diam. of 25 mm (1 in) diam.
772 Release from hole with eff. diam. of 50 mm (2 in) diam.
773 Release from hole with eff. diam. of 100 mm (4 in) diam.
774 Gasket Flange‐ Non Cryogenic: 12"≤D<20"
775 Catastrophic Rupture
776 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
777 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
778 Release from hole with eff. diam. of 25 mm (1 in) diam.
779 Release from hole with eff. diam. of 50 mm (2 in) diam.
780 Release from hole with eff. diam. of 100 mm (4 in) diam.
781 Gasket Flange: 20"≤D<40"
782 Catastrophic Rupture
783 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
784 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
785 Release from hole with eff. diam. of 25 mm (1 in) diam.
786 Release from hole with eff. diam. of 50 mm (2 in) diam.
787 Release from hole with eff. diam. of 100 mm (4 in) diam.
788 Gasket Flange‐ Cryogenic: 20"≤D<40"
789 Catastrophic Rupture
790 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
791 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
792 Release from hole with eff. diam. of 25 mm (1 in) diam.
793 Release from hole with eff. diam. of 50 mm (2 in) diam.
794 Release from hole with eff. diam. of 100 mm (4 in) diam.
795 Gasket Flange‐ Non Cryogenic: 20"≤D<40"
796 Catastrophic Rupture
797 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
798 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
799 Release from hole with eff. diam. of 25 mm (1 in) diam.
800 Release from hole with eff. diam. of 50 mm (2 in) diam.
801 Release from hole with eff. diam. of 100 mm (4 in) diam.
181 Piping:  All diameters
182 2.29E‐06 3.51E‐06 2.03E‐06 2.95E‐06 2.88E‐06 5.39E‐06 Catastrophic Rupture
183 1.75E‐06 1.75E‐06 1.58E‐06 1.58E‐06 2.13E‐06 2.13E‐06 Release from hole with eff. diam. of 1/3 diam.
184 8.44E‐06 1.02E‐05 7.86E‐06 9.29E‐06 9.65E‐06 1.22E‐05 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
185 9.51E‐06 1.23E‐05 2.30E‐05 2.30E‐05 1.05E‐05 1.28E‐05 2.30E‐05 2.30E‐05 1.08E‐05 1.48E‐05 2.30E‐05 2.30E‐05 Release from hole with eff. diam. of 25mm (1 in)
186 Piping:  d < 50mm (2‐inch)
187 1.00E‐06 1.16E‐05 1.16E‐05 1.01E‐05 1.01E‐05 1.59E‐05 1.59E‐05 Catastrophic Rupture
188 Release from hole with eff. diam. of 1/3 diam.
189 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
190 5.00E‐06 Release from hole with eff. diam. of 25mm (1 in)
191 Piping:  d < 50mm (2‐inch) ‐ Non‐Cryogenic
192 1.21E‐05 1.21E‐05 1.12E‐05 1.12E‐05 1.48E‐05 1.48E‐05 Catastrophic Rupture
193 Release from hole with eff. diam. of 1/3 diam.
194 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
195 Release from hole with eff. diam. of 25mm (1 in)
196 Piping:  d < 50mm (2‐inch) ‐ Cryogenic
197 Catastrophic Rupture
198 Release from hole with eff. diam. of 1/3 diam.
199 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
200 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 Release from hole with eff. diam. of 25mm (1 in)
201 Piping:  d < 50mm (2‐inch) ‐ Cryogenic ‐ VJ PIPING
202 Catastrophic Rupture
203 Release from hole with eff. diam. of 1/3 diam.
204 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
205 Release from hole with eff. diam. of 25mm (1 in)
206 Piping:  d < 50mm (2‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
207 Catastrophic Rupture
208 Release from hole with eff. diam. of 1/3 diam.
209 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
210 Release from hole with eff. diam. of 25mm (1 in)
211 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch)
212 5.00E‐07 1.85E‐06 2.43E‐06 1.64E‐06 2.03E‐06 2.47E‐06 3.82E‐06 Catastrophic Rupture
213 Release from hole with eff. diam. of 1/3 diam.
214 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
215 2.00E‐06 Release from hole with eff. diam. of 25mm (1 in)
216 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Non‐Cryogenic
217 2.25E‐06 5.61E‐06 2.34E‐06 5.94E‐06 2.25E‐06 5.61E‐06 Catastrophic Rupture
218 Release from hole with eff. diam. of 1/3 diam.
219 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
220 Release from hole with eff. diam. of 25mm (1 in)
221 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic
222 Catastrophic Rupture
223 Release from hole with eff. diam. of 1/3 diam.
224 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
225 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 Release from hole with eff. diam. of 25mm (1 in)
226 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic ‐ VJ PIPING
227 Catastrophic Rupture
228 Release from hole with eff. diam. of 1/3 diam.
229 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
230 Release from hole with eff. diam. of 25mm (1 in)
231 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
232 Catastrophic Rupture
233 Release from hole with eff. diam. of 1/3 diam.
234 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
235 Release from hole with eff. diam. of 25mm (1 in)
236 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch)
237 2.00E‐07 6.81E‐07 1.13E‐06 6.11E‐07 9.55E‐07 8.16E‐07 1.77E‐06 Catastrophic Rupture
238 4.00E‐07 4.15E‐07 4.15E‐07 4.61E‐07 4.61E‐07 2.73E‐07 2.73E‐07 Release from hole with eff. diam. of 1/3 diam.
239 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
240 7.00E‐07 Release from hole with eff. diam. of 25mm (1 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 156
 Appendix E: "Perceived Relevancy" Results with Table 12 Weighting Factors Applied and Using All References
"Specifically Included in Analysis" in Table 11 Except "OREDA '15", "CCPS '89" and "INL CHEM '95"

Super Category Comparison
Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example  Example 
A: Mean  A: Mean  A: Mean  A: Mean  B: Mean  B: Mean  B: Mean  B: Mean  C: Mean  C: Mean  C: Mean  C: Mean 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
Current FRT Category
Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure  Failure 
Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year  Rate/Year 
for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case:  for Case: 
SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG;  SFS=ALL;  SFS=ALL;  SFS=LNG;  SFS=LNG; 
CURRENT  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL;  PSL=ALL; 
PHMSA  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG;  AFS=ALL;  AFS=LNG; 
Index FRT SPEC DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL DT=ALL
241 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Non‐Cryogenic
242 9.51E‐07 2.98E‐06 9.11E‐07 3.16E‐06 9.51E‐07 2.98E‐06 Catastrophic Rupture
243 Release from hole with eff. diam. of 1/3 diam.
244 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
245 Release from hole with eff. diam. of 25mm (1 in)
246 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic
247 Catastrophic Rupture
248 Release from hole with eff. diam. of 1/3 diam.
249 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
250 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 Release from hole with eff. diam. of 25mm (1 in)
251 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic ‐ VJ PIPING
252 Catastrophic Rupture
253 Release from hole with eff. diam. of 1/3 diam.
254 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
255 Release from hole with eff. diam. of 25mm (1 in)
256 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
257 Catastrophic Rupture
258 Release from hole with eff. diam. of 1/3 diam.
259 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
260 Release from hole with eff. diam. of 25mm (1 in)
261 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch)
262 7.00E‐08 7.36E‐07 1.45E‐06 6.82E‐07 1.27E‐06 8.16E‐07 2.06E‐06 Catastrophic Rupture
263 2.00E‐07 8.91E‐08 8.91E‐08 1.05E‐07 1.05E‐07 4.96E‐08 4.96E‐08 Release from hole with eff. diam. of 1/3 diam.
264 4.00E‐07 1.12E‐07 9.61E‐08 1.12E‐07 9.62E‐08 1.12E‐07 9.62E‐08 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
265 5.00E‐07 1.93E‐06 1.93E‐06 1.72E‐06 1.72E‐06 2.47E‐06 2.47E‐06 Release from hole with eff. diam. of 25mm (1 in)
266 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Non‐Cryogenic
267 8.83E‐07 2.93E‐06 8.50E‐07 3.10E‐06 8.83E‐07 2.93E‐06 Catastrophic Rupture
268 Release from hole with eff. diam. of 1/3 diam.
269 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
270 Release from hole with eff. diam. of 25mm (1 in)
271 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic
272 Catastrophic Rupture
273 Release from hole with eff. diam. of 1/3 diam.
274 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
275 Release from hole with eff. diam. of 25mm (1 in)
276 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic ‐ VJ PIPING
277 Catastrophic Rupture
278 Release from hole with eff. diam. of 1/3 diam.
279 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
280 Release from hole with eff. diam. of 25mm (1 in)
281 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
282 Catastrophic Rupture
283 Release from hole with eff. diam. of 1/3 diam.
284 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
285 Release from hole with eff. diam. of 25mm (1 in)
286 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch)
287 2.00E‐08 6.19E‐07 1.19E‐06 5.73E‐07 9.98E‐07 7.41E‐07 1.81E‐06 Catastrophic Rupture
288 1.00E‐07 4.45E‐08 4.45E‐08 5.20E‐08 5.20E‐08 2.56E‐08 2.56E‐08 Release from hole with eff. diam. of 1/3 diam.
289 2.00E‐07 2.37E‐07 2.37E‐07 2.75E‐07 2.75E‐07 1.41E‐07 1.41E‐07 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
290 4.00E‐07 9.31E‐07 1.87E‐06 8.94E‐07 1.66E‐06 1.00E‐06 2.39E‐06 Release from hole with eff. diam. of 25mm (1 in)
291 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Non‐Cryogenic
292 8.45E‐07 2.68E‐06 8.72E‐07 2.84E‐06 8.46E‐07 2.68E‐06 Catastrophic Rupture
293 Release from hole with eff. diam. of 1/3 diam.
294 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
295 Release from hole with eff. diam. of 25mm (1 in)
296 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic
297 Catastrophic Rupture
298 Release from hole with eff. diam. of 1/3 diam.
299 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
300 Release from hole with eff. diam. of 25mm (1 in)
301 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic ‐ VJ PIPING
302 Catastrophic Rupture
303 Release from hole with eff. diam. of 1/3 diam.
304 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
305 Release from hole with eff. diam. of 25mm (1 in)
306 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
307 Catastrophic Rupture
308 Release from hole with eff. diam. of 1/3 diam.
309 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
310 Release from hole with eff. diam. of 25mm (1 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 157
Appendix F: "Wisdom of the Crowd" Results with Table 14 Weighting Factors Applied
and Using All Refernces "Specifically Included in Analysis" on Table 11

Example 5 Wgtg:  100% "HSE 

Example 7 Wgtg:  100% "LNE 

Super Category Comparison
Example 1 Wgtg:  Uniform

Example 6 Wgtg:  100% 

Example 8 Wgtg:  100% 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
"IOGP 434‐1 & 434‐3"

Current FRT Category
Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"RIVM BEVI '09"
FRED JUN '12"
CURRENT 
PHMSA 

'09"
Index FRT SPEC
1 Ambient Atm. Storage Tanks
2 1.66E‐03 9.44E‐04 6.68E‐06 4.36E‐06 4.89E‐06 Catastrophic Failure, Release to Atmosphere
3 9.98E‐05 9.98E‐05 9.98E‐05 9.98E‐05 9.98E‐05 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
4 2.50E‐03 2.50E‐03 2.50E‐03 2.50E‐03 2.50E‐03 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
5 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
6 Refrigerated Atm. Storage Tanks (Typ. Single Shell)
7 2.10E‐03 1.36E‐03 4.98E‐06 4.98E‐06 4.98E‐06 Catastrophic Failure, Release to Atmosphere
8 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
9 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
10 1.26E‐03 2.40E‐03 2.40E‐03 2.40E‐03 2.40E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
11 Cryogenic Atm. Storage Tanks
12 6.66E‐04 3.33E‐04 6.45E‐06 7.98E‐06 1.56E‐05 1.25E‐06 1.72E‐07 5.32E‐06 Catastrophic Failure, Release to Atmosphere
311 7.71E‐05 7.71E‐05 7.71E‐05 7.71E‐05 7.71E‐05 Catastrophic Failure of Tank Roof
13 3.37E‐05 3.37E‐05 3.37E‐05 3.37E‐05 3.37E‐05 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
14 2.74E‐05 2.74E‐05 2.74E‐05 2.74E‐05 2.74E‐05 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
15 1.44E‐03 1.44E‐03 1.44E‐03 1.44E‐03 1.00E‐04 2.40E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
16 Single Cont. Cryo. Atm. Storage Tanks
17 5.00E‐06 1.05E‐03 5.36E‐04 1.62E‐05 1.97E‐05 3.08E‐05 5.09E‐06 4.84E‐07 1.59E‐05 Catastrophic Failure, Release to Atmosphere
312 1.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 2.00E‐04 Catastrophic Failure of Tank Roof
18 8.00E‐05 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
19 2.00E‐04 8.00E‐05 8.00E‐05 8.00E‐05 8.00E‐05 8.00E‐05 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
20 1.00E‐04 1.19E‐03 1.19E‐03 1.19E‐03 1.19E‐03 1.00E‐04 2.40E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
21 Single Cont. Cryo. Atm. Storage Tanks ‐ HC Only (LNG, Ethane, Ethylene) ‐ Self‐Supp. Inner Tank
22 8.63E‐03 8.63E‐03 Catastrophic Failure, Release to Atmosphere
313 Catastrophic Failure of Tank Roof
23 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
24 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
25 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
26 Single Cont. Cryo. Atm. Storage Tanks ‐ LIN/LOX/LAR ‐ Self‐Supp. Inner Tank
27 2.20E‐05 2.20E‐05 2.20E‐05 2.20E‐05 2.20E‐05 Catastrophic Failure, Release to Atmosphere
314 Catastrophic Failure of Tank Roof
28 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
29 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
30 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
31 Double Cont. Cryo. Atm. Storage Tanks
32 5.00E‐07 3.23E‐08 3.21E‐08 3.36E‐08 3.44E‐08 5.49E‐08 1.21E‐08 9.47E‐09 2.29E‐08 Catastrophic Failure, Release to Atmosphere
315 1.00E‐04 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 4.03E‐05 Catastrophic Failure of Tank Roof
33 1.00E‐05 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 1.02E‐06 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
34 3.00E‐05 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 3.09E‐06 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
35 1.00E‐04 1.32E‐03 1.32E‐03 1.32E‐03 1.32E‐03 1.00E‐04 2.40E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
36 Full Cont. Cryo. Atm. Storage Tanks
37 1.00E‐08 2.54E‐07 2.06E‐07 1.75E‐07 1.02E‐07 5.10E‐08 1.66E‐08 1.10E‐08 1.15E‐08 Catastrophic Failure, Release to Atmosphere
316 4.00E‐05 2.42E‐10 2.45E‐10 2.49E‐10 2.42E‐10 2.47E‐10 Catastrophic Failure of Tank Roof
38 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 1.00E‐06 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
39 3.00E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 3.03E‐06 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
40 1.00E‐04 2.40E‐03 2.40E‐03 2.40E‐03 2.40E‐03 2.40E‐03 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
41 Membrane. Cryo. Atm. Storage Tanks
42 2.23E‐06 1.90E‐06 1.82E‐06 1.33E‐06 7.46E‐09 1.15E‐08 Catastrophic Failure, Release to Atmosphere
317 Catastrophic Failure of Tank Roof
43 Release from a hole in inner tank with eff. diam. of 1 m (~3ft)
44 Release from a hole in inner tank with eff. diam. of .3 m (~1ft)
45 Release from a hole in inner tank with eff. diam. of 0.01 m (0.4 in)
46 Process Vessels, Distillation Columns, Heat Exchangers, and Condensers
47 5.00E‐06 1.34E‐02 1.77E‐02 1.85E‐02 2.40E‐02 2.98E‐06 2.12E‐05 8.96E‐06 4.93E‐07 Catastrophic Failure (Rupture)
48 1.00E‐04 1.22E‐03 1.41E‐03 1.36E‐03 1.35E‐03 2.57E‐05 4.58E‐04 5.33E‐03 1.05E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
802 3.00E‐04 2.47E‐04 2.21E‐04 1.09E‐04 4.98E‐06 5.94E‐06 5.28E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
803 1.03E‐04 9.20E‐05 7.29E‐05 4.99E‐06 4.99E‐06 Release from a hole with eff. diam. of 0.05m (2 in)
804 9.59E‐05 8.69E‐05 8.61E‐05 6.77E‐05 1.60E‐05 1.47E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
49 Process Vessels incl Distillation Columns
50 3.01E‐03 1.68E‐03 8.58E‐04 2.72E‐06 2.98E‐06 3.50E‐06 1.76E‐06 4.94E‐07 Catastrophic Failure (Rupture)
51 1.66E‐04 1.45E‐04 1.20E‐04 5.62E‐05 2.57E‐05 7.30E‐05 6.39E‐05 2.99E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
805 7.98E‐05 5.83E‐05 4.71E‐05 1.84E‐05 4.98E‐06 5.94E‐06 1.95E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
806 2.82E‐05 2.28E‐05 1.62E‐05 4.99E‐06 4.99E‐06 Release from a hole with eff. diam. of 0.05m (2 in)
807 4.37E‐05 4.22E‐05 4.24E‐05 3.47E‐05 5.28E‐05 Release from a hole with eff. diam. of 0.10m (4 in)
52 PVs incl Dist. Columns ‐ Single Wall
53 3.91E‐03 1.99E‐03 1.01E‐03 2.67E‐06 2.99E‐06 3.50E‐06 1.76E‐06 4.93E‐07 Catastrophic Failure (Rupture)
54 1.90E‐04 1.54E‐04 1.28E‐04 5.54E‐05 1.00E‐05 7.30E‐05 6.39E‐05 2.99E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
808 7.98E‐05 5.83E‐05 4.71E‐05 1.84E‐05 4.98E‐06 5.94E‐06 1.95E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
809 2.82E‐05 2.28E‐05 1.62E‐05 4.99E‐06 4.99E‐06 Release from a hole with eff. diam. of 0.05m (2 in)
810 4.37E‐05 4.22E‐05 4.24E‐05 3.47E‐05 5.28E‐05 Release from a hole with eff. diam. of 0.10m (4 in)
318 Pressure Storage Vessels/Tanks (Single Wall)
319 1.63E‐05 8.92E‐06 3.45E‐06 1.94E‐06 2.99E‐06 4.86E‐07 3.57E‐07 4.93E‐07 Catastrophic Failure (Rupture)
320 1.08E‐05 1.07E‐05 1.06E‐05 1.04E‐05 1.00E‐05 9.98E‐06 1.20E‐05 1.20E‐05 Release from a hole with eff. diam. of 0.01m (0.4 in)
380 4.54E‐06 4.16E‐06 4.27E‐06 4.30E‐06 4.98E‐06 1.08E‐06 7.20E‐06 Release from a hole with eff. diam. of 0.025m (1 in)
381 4.99E‐06 4.99E‐06 4.99E‐06 4.99E‐06 4.99E‐06 Release from a hole with eff. diam. of 0.05m (2 in)
382 4.17E‐06 4.17E‐06 4.17E‐06 4.17E‐06 4.17E‐06 Release from a hole with eff. diam. of 0.10m (4 in)
321 Process Vessels (Single Wall)
322 7.79E‐05 5.75E‐05 5.04E‐05 4.11E‐06 4.99E‐06 3.18E‐06 Catastrophic Failure (Rupture)
323 4.00E‐04 3.41E‐04 3.17E‐04 1.54E‐04 9.99E‐05 1.20E‐04 5.60E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
383 Some cells shaded to  1.10E‐05 3.50E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
384 protect copyrights or  Release from a hole with eff. diam. of 0.05m (2 in)
386 licensed software 1.10E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
55 Mole Sieve Vessel (Single Wall)
56 Catastrophic Failure (Rupture)
57 Release from a hole with eff. diam. of 0.01m (0.4 in)
387 Release from a hole with eff. diam. of 0.025m (1 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 158
Appendix F: "Wisdom of the Crowd" Results with Table 14 Weighting Factors Applied
and Using All Refernces "Specifically Included in Analysis" on Table 11

Example 5 Wgtg:  100% "HSE 

Example 7 Wgtg:  100% "LNE 

Super Category Comparison
Example 1 Wgtg:  Uniform

Example 6 Wgtg:  100% 

Example 8 Wgtg:  100% 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
"IOGP 434‐1 & 434‐3"

Current FRT Category
Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"RIVM BEVI '09"
FRED JUN '12"
CURRENT 
PHMSA 

'09"
Index FRT SPEC
388 Release from a hole with eff. diam. of 0.05m (2 in)
389 Release from a hole with eff. diam. of 0.10m (4 in)
58 Distillation Columns (Single Wall)
59 4.98E‐06 Catastrophic Failure (Rupture)
60 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 1.00E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
390 Release from a hole with eff. diam. of 0.025m (1 in)
391 Release from a hole with eff. diam. of 0.05m (2 in)
392 Release from a hole with eff. diam. of 0.10m (4 in)
61 Separator (Single Wall)
62 Catastrophic Failure (Rupture)
63 Release from a hole with eff. diam. of 0.01m (0.4 in)
393 Release from a hole with eff. diam. of 0.025m (1 in)
394 Release from a hole with eff. diam. of 0.05m (2 in)
395 Release from a hole with eff. diam. of 0.10m (4 in)
64 PVs incl Dist. Columns ‐ Double Wall (cryogenic)
65 Catastrophic Failure (Rupture)
66 Release from a hole with eff. diam. of 0.01m (0.4 in)
396 Release from a hole with eff. diam. of 0.025m (1 in)
397 Release from a hole with eff. diam. of 0.05m (2 in)
398 Release from a hole with eff. diam. of 0.10m (4 in)
70 ISO Containers
71 2.95E‐06 2.95E‐06 2.95E‐06 2.95E‐06 2.95E‐06 Catastrophic Failure (Rupture)
72 6.01E‐05 6.01E‐05 6.01E‐05 6.01E‐05 6.01E‐05 Release from a hole with eff. diam. of 0.01m (0.4 in)
399 Release from a hole with eff. diam. of 0.025m (1 in)
400 Release from a hole with eff. diam. of 0.05m (2 in)
401 Release from a hole with eff. diam. of 0.10m (4 in)
73 Heat Exchangers  incl. Condensers
74 2.47E‐02 3.40E‐02 3.84E‐02 7.10E‐02 5.01E‐05 1.63E‐05 Catastrophic Failure (Rupture)
75 2.03E‐03 2.70E‐03 2.89E‐03 4.47E‐03 1.00E‐03 1.33E‐02 1.33E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
402 4.56E‐04 4.23E‐04 4.28E‐04 4.95E‐04 6.54E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
403 1.56E‐04 1.56E‐04 1.56E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
404 1.13E‐04 1.01E‐04 9.92E‐05 7.55E‐05 1.60E‐05 1.87E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
76 Fired Heat Exchangers
77 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 Catastrophic Failure (Rupture)
78 Release from a hole with eff. diam. of 0.01m (0.4 in)
405 Release from a hole with eff. diam. of 0.025m (1 in)
406 Release from a hole with eff. diam. of 0.05m (2 in)
407 Release from a hole with eff. diam. of 0.10m (4 in)
324 Submerged Combustion Vaporizers
325 2.72E‐01 2.72E‐01 2.72E‐01 2.72E‐01 Catastrophic Failure (Rupture)
326 Release from a hole with eff. diam. of 0.01m (0.4 in)
408 Release from a hole with eff. diam. of 0.025m (1 in)
409 Release from a hole with eff. diam. of 0.05m (2 in)
410 Release from a hole with eff. diam. of 0.10m (4 in)
79 Non‐Fired Heat Exchangers incl. Condensers
80 7.64E‐03 3.91E‐03 2.30E‐03 3.04E‐05 5.01E‐05 1.63E‐05 Catastrophic Failure (Rupture)
81 2.17E‐03 2.70E‐03 2.89E‐03 4.47E‐03 1.00E‐03 1.33E‐02 1.33E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
411 4.56E‐04 4.23E‐04 4.28E‐04 4.95E‐04 6.54E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
412 1.56E‐04 1.56E‐04 1.56E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
413 1.13E‐04 1.01E‐04 9.92E‐05 7.56E‐05 1.60E‐05 1.87E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
82 Shell & Tube Heat Exchangers
83 1.06E‐03 6.37E‐04 4.09E‐04 2.61E‐05 5.00E‐05 1.29E‐05 Catastrophic Failure (Rupture)
84 1.21E‐03 1.40E‐03 1.47E‐03 2.07E‐03 1.00E‐03 6.00E‐03 9.97E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
414 3.37E‐04 3.11E‐04 3.15E‐04 3.80E‐04 4.79E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
415 1.01E‐04 1.01E‐04 1.01E‐04 Release from a hole with eff. diam. of 0.05m (2 in)
416 6.62E‐05 5.75E‐05 5.60E‐05 3.95E‐05 1.60E‐05 1.16E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
85 Tube‐side Heat Exchangers (HC in tube)
86 4.92E‐04 4.18E‐04 3.17E‐04 Catastrophic Failure (Rupture)
87 8.81E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
417 4.00E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
418 Release from a hole with eff. diam. of 0.05m (2 in)
419 9.10E‐05 Release from a hole with eff. diam. of 0.10m (4 in)
88 Shell‐side Heat Exchangers (HC in shell)
89 9.38E‐05 8.14E‐05 7.54E‐05 4.01E‐05 5.00E‐05 Catastrophic Failure (Rupture)
90 1.00E‐03 1.10E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
420 5.61E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
421 Release from a hole with eff. diam. of 0.05m (2 in)
422 1.40E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
91 Plate Heat Exchangers
92 2.58E‐02 1.10E‐02 5.95E‐03 3.73E‐05 5.01E‐05 2.01E‐05 Catastrophic Failure (Rupture)
93 1.00E‐03 1.80E‐02 2.80E‐03 Release from a hole with eff. diam. of 0.01m (0.4 in)
423 1.60E‐03 Release from a hole with eff. diam. of 0.025m (1 in)
424 Release from a hole with eff. diam. of 0.05m (2 in)
425 4.80E‐04 Release from a hole with eff. diam. of 0.10m (4 in)
94 Air Cooled (Fin Fan) Heat Exchangers
95 Catastrophic Failure (Rupture)
96 4.90E‐04 Release from a hole with eff. diam. of 0.01m (0.4 in)
426 2.40E‐04 Release from a hole with eff. diam. of 0.025m (1 in)
427 Release from a hole with eff. diam. of 0.05m (2 in)
428 3.34E‐05 3.04E‐05 3.08E‐05 3.15E‐05 5.99E‐05 Release from a hole with eff. diam. of 0.10m (4 in)
97 Printed Heat Exchangers
98 Catastrophic Failure (Rupture)
99 Release from a hole with eff. diam. of 0.01m (0.4 in)
429 Release from a hole with eff. diam. of 0.025m (1 in)
430 Release from a hole with eff. diam. of 0.05m (2 in)
431 Release from a hole with eff. diam. of 0.10m (4 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 159
Appendix F: "Wisdom of the Crowd" Results with Table 14 Weighting Factors Applied
and Using All Refernces "Specifically Included in Analysis" on Table 11

Example 5 Wgtg:  100% "HSE 

Example 7 Wgtg:  100% "LNE 

Super Category Comparison
Example 1 Wgtg:  Uniform

Example 6 Wgtg:  100% 

Example 8 Wgtg:  100% 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
"IOGP 434‐1 & 434‐3"

Current FRT Category
Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"RIVM BEVI '09"
FRED JUN '12"
CURRENT 
PHMSA 

'09"
Index FRT SPEC
100 Plate‐Fin Heat Exchangers
101 Catastrophic Failure (Rupture)
102 Release from a hole with eff. diam. of 0.01m (0.4 in)
432 Release from a hole with eff. diam. of 0.025m (1 in)
433 Release from a hole with eff. diam. of 0.05m (2 in)
434 Release from a hole with eff. diam. of 0.10m (4 in)
103 Ambient Vaporizer Heat Exchangers
104 Catastrophic Failure (Rupture)
105 Release from a hole with eff. diam. of 0.01m (0.4 in)
435 Release from a hole with eff. diam. of 0.025m (1 in)
436 Release from a hole with eff. diam. of 0.05m (2 in)
437 Release from a hole with eff. diam. of 0.10m (4 in)
106 Truck Transfer Arm
107 3.00E‐04 1.51E‐05 2.26E‐05 2.38E‐05 2.71E‐05 3.01E‐05 3.01E‐05 Rupture of transfer arm
108 3.00E‐03 2.09E‐04 3.00E‐04 3.00E‐04 3.00E‐04 3.00E‐04 2.99E‐04 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
109 Truck Transfer Arm ‐ Cryogenic
110 Rupture of transfer arm
111 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
112 Truck Transfer Arm ‐ Non‐cryogenic
113 Rupture of transfer arm
114 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
115 Ship Transfer Arm
116 2.00E‐05 1.07E‐03 1.16E‐04 9.42E‐05 5.41E‐05 2.99E‐05 3.00E‐05 Rupture of transfer arm
117 2.00E‐04 1.89E‐02 3.00E‐04 3.00E‐04 3.00E‐04 3.00E‐04 3.00E‐04 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
118 Ship Transfer Arm ‐ Cryogenic
119 Rupture of transfer arm
120 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
121 Ship Transfer Arm ‐ Non‐cryogenic
122 Rupture of transfer arm
123 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
124 Truck Transfer Hose
125 4.00E‐02 1.86E‐02 1.17E‐02 7.06E‐03 2.43E‐03 4.00E‐04 4.00E‐03 4.00E‐03 Rupture of transfer hose
126 4.00E‐01 2.82E‐02 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
127 Truck Transfer Hose ‐ Cryogenic
128 Rupture of transfer hose
129 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
130 Truck Transfer Hose ‐ Non‐cryogenic
131 4.00E‐04 4.00E‐04 4.00E‐04 4.00E‐04 4.00E‐04 Rupture of transfer hose
132 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
133 Ship Transfer Hose
134 2.88E‐03 3.56E‐03 3.77E‐03 4.00E‐03 4.00E‐03 4.00E‐03 Rupture of transfer hose
135 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 4.00E‐02 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
136 Ship Transfer Hose ‐ Cryogenic
137 Rupture of transfer hose
138 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
139 Ship Transfer Hose ‐ Non‐cryogenic
140 Rupture of transfer hose
141 Release from hole with eff. diam. of 10% diam. with max.of 50mm (2 in)
142 Valve: All diameters
143 9.00E‐06 1.16E‐03 7.55E‐04 5.06E‐04 8.70E‐06 Catastrophic Rupture
327 2.01E‐04 2.01E‐04 2.01E‐04 2.00E‐04 2.00E‐04 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
144 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
328 Release from hole with eff. diam. of 25 mm (1 in) diam.
329 Release from hole with eff. diam. of 50 mm (2 in) diam.
330 Release from hole with eff. diam. of 100 mm (4 in) diam.
145 Manual Valves: All diameters
146 2.07E‐04 1.30E‐04 3.79E‐05 Catastrophic Rupture
331 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
147 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
332 Release from hole with eff. diam. of 25 mm (1 in) diam.
333 Release from hole with eff. diam. of 50 mm (2 in) diam.
334 Release from hole with eff. diam. of 100 mm (4 in) diam.
148 Manual Valves ‐ Cryogenic: All diameters
149 Catastrophic Rupture
335 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
150 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
336 Release from hole with eff. diam. of 25 mm (1 in) diam.
337 Release from hole with eff. diam. of 50 mm (2 in) diam.
338 Release from hole with eff. diam. of 100 mm (4 in) diam.
151 Manual Valves ‐ Non‐cryogenic: All diameters
152 5.82E‐05 3.33E‐05 2.31E‐05 Catastrophic Rupture
339 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
153 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
340 Release from hole with eff. diam. of 25 mm (1 in) diam.
341 Release from hole with eff. diam. of 50 mm (2 in) diam.
342 Release from hole with eff. diam. of 100 mm (4 in) diam.
154 Actuated Valves: All diameters
155 5.55E‐05 4.25E‐05 3.51E‐05 Catastrophic Rupture
343 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
156 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
344 Release from hole with eff. diam. of 25 mm (1 in) diam.
345 Release from hole with eff. diam. of 50 mm (2 in) diam.
346 Release from hole with eff. diam. of 100 mm (4 in) diam.
157 Actuated Valves ‐ Cryogenic: All diameters
158 Catastrophic Rupture
347 Release from hole with eff. diam. of 2 mm (0.08 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 160
Appendix F: "Wisdom of the Crowd" Results with Table 14 Weighting Factors Applied
and Using All Refernces "Specifically Included in Analysis" on Table 11

Example 5 Wgtg:  100% "HSE 

Example 7 Wgtg:  100% "LNE 

Super Category Comparison
Example 1 Wgtg:  Uniform

Example 6 Wgtg:  100% 

Example 8 Wgtg:  100% 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
"IOGP 434‐1 & 434‐3"

Current FRT Category
Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"RIVM BEVI '09"
FRED JUN '12"
CURRENT 
PHMSA 

'09"
Index FRT SPEC
159 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
348 Release from hole with eff. diam. of 25 mm (1 in) diam.
349 Release from hole with eff. diam. of 50 mm (2 in) diam.
350 Release from hole with eff. diam. of 100 mm (4 in) diam.
160 Actuated Valves ‐ Non‐cryogenic: All diameters
161 5.55E‐05 4.25E‐05 3.51E‐05 Catastrophic Rupture
351 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
162 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
352 Release from hole with eff. diam. of 25 mm (1 in) diam.
353 Release from hole with eff. diam. of 50 mm (2 in) diam.
354 Release from hole with eff. diam. of 100 mm (4 in) diam.
438 Valve: 2"≤D<6"
439 Catastrophic Rupture
440 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
441 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
442 Release from hole with eff. diam. of 25 mm (1 in) diam.
443 Release from hole with eff. diam. of 50 mm (2 in) diam.
444 Release from hole with eff. diam. of 100 mm (4 in) diam.
445 Manual Valves: 2"≤D<6"
446 Catastrophic Rupture
447 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
448 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
449 Release from hole with eff. diam. of 25 mm (1 in) diam.
450 Release from hole with eff. diam. of 50 mm (2 in) diam.
451 Release from hole with eff. diam. of 100 mm (4 in) diam.
452 Manual Valves ‐ Cryogenic: 2"≤D<6"
453 Catastrophic Rupture
454 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
455 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
456 Release from hole with eff. diam. of 25 mm (1 in) diam.
457 Release from hole with eff. diam. of 50 mm (2 in) diam.
458 Release from hole with eff. diam. of 100 mm (4 in) diam.
459 Manual Valves ‐ Non‐cryogenic: 2"≤D<6"
460 Catastrophic Rupture
461 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
462 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
463 Release from hole with eff. diam. of 25 mm (1 in) diam.
464 Release from hole with eff. diam. of 50 mm (2 in) diam.
465 Release from hole with eff. diam. of 100 mm (4 in) diam.
466 Actuated Valves: 2"≤D<6"
467 Catastrophic Rupture
468 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
469 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
470 Release from hole with eff. diam. of 25 mm (1 in) diam.
471 Release from hole with eff. diam. of 50 mm (2 in) diam.
472 Release from hole with eff. diam. of 100 mm (4 in) diam.
473 Actuated Valves ‐ Cryogenic: 2"≤D<6"
474 Catastrophic Rupture
475 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
476 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
477 Release from hole with eff. diam. of 25 mm (1 in) diam.
478 Release from hole with eff. diam. of 50 mm (2 in) diam.
479 Release from hole with eff. diam. of 100 mm (4 in) diam.
480 Actuated Valves ‐ Non‐cryogenic: 2"≤D<6"
481 Catastrophic Rupture
482 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
483 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
484 Release from hole with eff. diam. of 25 mm (1 in) diam.
485 Release from hole with eff. diam. of 50 mm (2 in) diam.
486 Release from hole with eff. diam. of 100 mm (4 in) diam.
487 Valve: 6"≤D<12"
488 Catastrophic Rupture
489 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
490 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
491 Release from hole with eff. diam. of 25 mm (1 in) diam.
492 Release from hole with eff. diam. of 50 mm (2 in) diam.
493 Release from hole with eff. diam. of 100 mm (4 in) diam.
494 Manual Valves: 6"≤D<12"
495 Catastrophic Rupture
496 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
497 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
498 Release from hole with eff. diam. of 25 mm (1 in) diam.
499 Release from hole with eff. diam. of 50 mm (2 in) diam.
500 Release from hole with eff. diam. of 100 mm (4 in) diam.
501 Manual Valves ‐ Cryogenic: 6"≤D<12"
502 Catastrophic Rupture
503 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
504 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
505 Release from hole with eff. diam. of 25 mm (1 in) diam.
506 Release from hole with eff. diam. of 50 mm (2 in) diam.
507 Release from hole with eff. diam. of 100 mm (4 in) diam.
508 Manual Valves ‐ Non‐cryogenic: 6"≤D<12"
509 Catastrophic Rupture
510 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
511 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
512 Release from hole with eff. diam. of 25 mm (1 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 161
Appendix F: "Wisdom of the Crowd" Results with Table 14 Weighting Factors Applied
and Using All Refernces "Specifically Included in Analysis" on Table 11

Example 5 Wgtg:  100% "HSE 

Example 7 Wgtg:  100% "LNE 

Super Category Comparison
Example 1 Wgtg:  Uniform

Example 6 Wgtg:  100% 

Example 8 Wgtg:  100% 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
"IOGP 434‐1 & 434‐3"

Current FRT Category
Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"RIVM BEVI '09"
FRED JUN '12"
CURRENT 
PHMSA 

'09"
Index FRT SPEC
513 Release from hole with eff. diam. of 50 mm (2 in) diam.
514 Release from hole with eff. diam. of 100 mm (4 in) diam.
515 Actuated Valves: 6"≤D<12"
516 Catastrophic Rupture
517 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
518 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
519 Release from hole with eff. diam. of 25 mm (1 in) diam.
520 Release from hole with eff. diam. of 50 mm (2 in) diam.
521 Release from hole with eff. diam. of 100 mm (4 in) diam.
522 Actuated Valves ‐ Cryogenic: 6"≤D<12"
523 Catastrophic Rupture
524 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
525 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
526 Release from hole with eff. diam. of 25 mm (1 in) diam.
527 Release from hole with eff. diam. of 50 mm (2 in) diam.
528 Release from hole with eff. diam. of 100 mm (4 in) diam.
529 Actuated Valves ‐ Non‐cryogenic: 6"≤D<12"
530 Catastrophic Rupture
531 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
532 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
533 Release from hole with eff. diam. of 25 mm (1 in) diam.
534 Release from hole with eff. diam. of 50 mm (2 in) diam.
535 Release from hole with eff. diam. of 100 mm (4 in) diam.
536 Valve: 12"≤D<20"
537 Catastrophic Rupture
538 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
539 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
540 Release from hole with eff. diam. of 25 mm (1 in) diam.
541 Release from hole with eff. diam. of 50 mm (2 in) diam.
542 Release from hole with eff. diam. of 100 mm (4 in) diam.
543 Manual Valves: 12"≤D<20"
544 Catastrophic Rupture
545 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
546 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
547 Release from hole with eff. diam. of 25 mm (1 in) diam.
548 Release from hole with eff. diam. of 50 mm (2 in) diam.
549 Release from hole with eff. diam. of 100 mm (4 in) diam.
550 Manual Valves ‐ Cryogenic: 12"≤D<20"
551 Catastrophic Rupture
552 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
553 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
554 Release from hole with eff. diam. of 25 mm (1 in) diam.
555 Release from hole with eff. diam. of 50 mm (2 in) diam.
556 Release from hole with eff. diam. of 100 mm (4 in) diam.
557 Manual Valves ‐ Non‐cryogenic: 12"≤D<20"
558 Catastrophic Rupture
559 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
560 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
561 Release from hole with eff. diam. of 25 mm (1 in) diam.
562 Release from hole with eff. diam. of 50 mm (2 in) diam.
563 Release from hole with eff. diam. of 100 mm (4 in) diam.
564 Actuated Valves: 12"≤D<20"
565 Catastrophic Rupture
566 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
567 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
568 Release from hole with eff. diam. of 25 mm (1 in) diam.
569 Release from hole with eff. diam. of 50 mm (2 in) diam.
570 Release from hole with eff. diam. of 100 mm (4 in) diam.
571 Actuated Valves ‐ Cryogenic: 12"≤D<20"
572 Catastrophic Rupture
573 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
574 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
575 Release from hole with eff. diam. of 25 mm (1 in) diam.
576 Release from hole with eff. diam. of 50 mm (2 in) diam.
577 Release from hole with eff. diam. of 100 mm (4 in) diam.
578 Actuated Valves ‐ Non‐cryogenic: 12"≤D<20"
579 Catastrophic Rupture
580 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
581 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
582 Release from hole with eff. diam. of 25 mm (1 in) diam.
583 Release from hole with eff. diam. of 50 mm (2 in) diam.
584 Release from hole with eff. diam. of 100 mm (4 in) diam.
585 Valve: 20"≤D<40"
586 Catastrophic Rupture
587 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
588 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
589 Release from hole with eff. diam. of 25 mm (1 in) diam.
590 Release from hole with eff. diam. of 50 mm (2 in) diam.
591 Release from hole with eff. diam. of 100 mm (4 in) diam.
592 Manual Valves: 20"≤D<40"
593 Catastrophic Rupture
594 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
595 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
596 Release from hole with eff. diam. of 25 mm (1 in) diam.
597 Release from hole with eff. diam. of 50 mm (2 in) diam.
598 Release from hole with eff. diam. of 100 mm (4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 162
Appendix F: "Wisdom of the Crowd" Results with Table 14 Weighting Factors Applied
and Using All Refernces "Specifically Included in Analysis" on Table 11

Example 5 Wgtg:  100% "HSE 

Example 7 Wgtg:  100% "LNE 

Super Category Comparison
Example 1 Wgtg:  Uniform

Example 6 Wgtg:  100% 

Example 8 Wgtg:  100% 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
"IOGP 434‐1 & 434‐3"

Current FRT Category
Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"RIVM BEVI '09"
FRED JUN '12"
CURRENT 
PHMSA 

'09"
Index FRT SPEC
599 Manual Valves ‐ Cryogenic: 20"≤D<40"
600 Catastrophic Rupture
601 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
602 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
603 Release from hole with eff. diam. of 25 mm (1 in) diam.
604 Release from hole with eff. diam. of 50 mm (2 in) diam.
605 Release from hole with eff. diam. of 100 mm (4 in) diam.
606 Manual Valves ‐ Non‐cryogenic: 20"≤D<40"
607 Catastrophic Rupture
608 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
609 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
610 Release from hole with eff. diam. of 25 mm (1 in) diam.
611 Release from hole with eff. diam. of 50 mm (2 in) diam.
612 Release from hole with eff. diam. of 100 mm (4 in) diam.
613 Actuated Valves: 20"≤D<40"
614 Catastrophic Rupture
615 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
616 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
617 Release from hole with eff. diam. of 25 mm (1 in) diam.
618 Release from hole with eff. diam. of 50 mm (2 in) diam.
619 Release from hole with eff. diam. of 100 mm (4 in) diam.
620 Actuated Valves ‐ Cryogenic: 20"≤D<40"
621 Catastrophic Rupture
622 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
623 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
624 Release from hole with eff. diam. of 25 mm (1 in) diam.
625 Release from hole with eff. diam. of 50 mm (2 in) diam.
626 Release from hole with eff. diam. of 100 mm (4 in) diam.
627 Actuated Valves ‐ Non‐cryogenic: 20"≤D<40"
628 Catastrophic Rupture
629 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
630 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
631 Release from hole with eff. diam. of 25 mm (1 in) diam.
632 Release from hole with eff. diam. of 50 mm (2 in) diam.
633 Release from hole with eff. diam. of 100 mm (4 in) diam.
163 Expansion Joint‐ All diameters
164 4.00E‐03 1.71E‐03 1.50E‐03 1.29E‐03 8.76E‐04 Catastrophic Rupture
355 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
165 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
356 Release from hole with eff. diam. of 25 mm (1 in) diam.
357 Release from hole with eff. diam. of 50 mm (2 in) diam.
358 Release from hole with eff. diam. of 100 mm (4 in) diam.
166 Expansion Joint‐ Cryogenic: All diameters
167 Catastrophic Rupture
359 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
168 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
360 Release from hole with eff. diam. of 25 mm (1 in) diam.
361 Release from hole with eff. diam. of 50 mm (2 in) diam.
362 Release from hole with eff. diam. of 100 mm (4 in) diam.
169 Expansion Joint‐ Non Cryogenic: All diameters
170 Catastrophic Rupture
363 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
171 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
364 Release from hole with eff. diam. of 25 mm (1 in) diam.
365 Release from hole with eff. diam. of 50 mm (2 in) diam.
366 Release from hole with eff. diam. of 100 mm (4 in) diam.
634 Expansion Joint: 2"≤D<6"
635 Catastrophic Rupture
636 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
637 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
638 Release from hole with eff. diam. of 25 mm (1 in) diam.
639 Release from hole with eff. diam. of 50 mm (2 in) diam.
640 Release from hole with eff. diam. of 100 mm (4 in) diam.
641 Expansion Joint‐ Cryogenic: 2"≤D<6"
642 Catastrophic Rupture
643 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
644 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
645 Release from hole with eff. diam. of 25 mm (1 in) diam.
646 Release from hole with eff. diam. of 50 mm (2 in) diam.
647 Release from hole with eff. diam. of 100 mm (4 in) diam.
648 Expansion Joint‐ Non Cryogenic: 2"≤D<6"
649 Catastrophic Rupture
650 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
651 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
652 Release from hole with eff. diam. of 25 mm (1 in) diam.
653 Release from hole with eff. diam. of 50 mm (2 in) diam.
654 Release from hole with eff. diam. of 100 mm (4 in) diam.
655 Expansion Joint: 6"≤D<12"
656 Catastrophic Rupture
657 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
658 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
659 Release from hole with eff. diam. of 25 mm (1 in) diam.
660 Release from hole with eff. diam. of 50 mm (2 in) diam.
661 Release from hole with eff. diam. of 100 mm (4 in) diam.
662 Expansion Joint‐ Cryogenic: 6"≤D<12"
663 Catastrophic Rupture

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 163
Appendix F: "Wisdom of the Crowd" Results with Table 14 Weighting Factors Applied
and Using All Refernces "Specifically Included in Analysis" on Table 11

Example 5 Wgtg:  100% "HSE 

Example 7 Wgtg:  100% "LNE 

Super Category Comparison
Example 1 Wgtg:  Uniform

Example 6 Wgtg:  100% 

Example 8 Wgtg:  100% 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
"IOGP 434‐1 & 434‐3"

Current FRT Category
Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"RIVM BEVI '09"
FRED JUN '12"
CURRENT 
PHMSA 

'09"
Index FRT SPEC
664 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
665 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
666 Release from hole with eff. diam. of 25 mm (1 in) diam.
667 Release from hole with eff. diam. of 50 mm (2 in) diam.
668 Release from hole with eff. diam. of 100 mm (4 in) diam.
669 Expansion Joint‐ Non Cryogenic: 6"≤D<12"
670 Catastrophic Rupture
671 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
672 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
673 Release from hole with eff. diam. of 25 mm (1 in) diam.
674 Release from hole with eff. diam. of 50 mm (2 in) diam.
675 Release from hole with eff. diam. of 100 mm (4 in) diam.
676 Expansion Joint: 12"≤D<20"
677 Catastrophic Rupture
678 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
679 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
680 Release from hole with eff. diam. of 25 mm (1 in) diam.
681 Release from hole with eff. diam. of 50 mm (2 in) diam.
682 Release from hole with eff. diam. of 100 mm (4 in) diam.
683 Expansion Joint‐ Cryogenic: 12"≤D<20"
684 Catastrophic Rupture
685 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
686 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
687 Release from hole with eff. diam. of 25 mm (1 in) diam.
688 Release from hole with eff. diam. of 50 mm (2 in) diam.
689 Release from hole with eff. diam. of 100 mm (4 in) diam.
690 Expansion Joint‐ Non Cryogenic: 12"≤D<20"
691 Catastrophic Rupture
692 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
693 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
694 Release from hole with eff. diam. of 25 mm (1 in) diam.
695 Release from hole with eff. diam. of 50 mm (2 in) diam.
696 Release from hole with eff. diam. of 100 mm (4 in) diam.
697 Expansion Joint: 20"≤D<40"
698 Catastrophic Rupture
699 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
700 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
701 Release from hole with eff. diam. of 25 mm (1 in) diam.
702 Release from hole with eff. diam. of 50 mm (2 in) diam.
703 Release from hole with eff. diam. of 100 mm (4 in) diam.
704 Expansion Joint‐ Cryogenic: 20"≤D<40"
705 Catastrophic Rupture
706 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
707 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
708 Release from hole with eff. diam. of 25 mm (1 in) diam.
709 Release from hole with eff. diam. of 50 mm (2 in) diam.
710 Release from hole with eff. diam. of 100 mm (4 in) diam.
711 Expansion Joint‐ Non Cryogenic: 20"≤D<40"
712 Catastrophic Rupture
713 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
714 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
715 Release from hole with eff. diam. of 25 mm (1 in) diam.
716 Release from hole with eff. diam. of 50 mm (2 in) diam.
717 Release from hole with eff. diam. of 100 mm (4 in) diam.
172 Gasket Flange‐ All diameters
173 3.00E‐02 4.73E‐05 2.91E‐05 2.32E‐05 8.75E‐05 Catastrophic Rupture
367 9.47E‐05 9.73E‐05 9.70E‐05 5.09E‐05 5.09E‐05 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
174 5.15E‐05 5.15E‐05 5.15E‐05 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
368 7.43E‐06 7.31E‐06 7.05E‐06 5.03E‐06 5.03E‐06 Release from hole with eff. diam. of 25 mm (1 in) diam.
369 6.09E‐06 5.77E‐06 5.12E‐06 9.46E‐08 9.47E‐08 Release from hole with eff. diam. of 50 mm (2 in) diam.
370 2.66E‐06 2.66E‐06 2.66E‐06 Release from hole with eff. diam. of 100 mm (4 in) diam.
175 Gasket Flange‐ Cryogenic: All Diameters
176 Catastrophic Rupture
371 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
177 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
372 Release from hole with eff. diam. of 25 mm (1 in) diam.
373 Release from hole with eff. diam. of 50 mm (2 in) diam.
374 Release from hole with eff. diam. of 100 mm (4 in) diam.
178 Gasket Flange‐ Non Cryogenic: All Diameters
179 6.68E‐06 6.44E‐06 6.35E‐06 Catastrophic Rupture
375 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
180 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
376 Release from hole with eff. diam. of 25 mm (1 in) diam.
377 Release from hole with eff. diam. of 50 mm (2 in) diam.
378 Release from hole with eff. diam. of 100 mm (4 in) diam.
718 Gasket Flange: 2"≤D<6"
719 Catastrophic Rupture
720 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
721 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
722 Release from hole with eff. diam. of 25 mm (1 in) diam.
723 Release from hole with eff. diam. of 50 mm (2 in) diam.
724 Release from hole with eff. diam. of 100 mm (4 in) diam.
725 Gasket Flange‐ Cryogenic: 2"≤D<6"
726 Catastrophic Rupture
727 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
728 Release from hole with eff. diam. of 10 mm (0.4 in) diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 164
Appendix F: "Wisdom of the Crowd" Results with Table 14 Weighting Factors Applied
and Using All Refernces "Specifically Included in Analysis" on Table 11

Example 5 Wgtg:  100% "HSE 

Example 7 Wgtg:  100% "LNE 

Super Category Comparison
Example 1 Wgtg:  Uniform

Example 6 Wgtg:  100% 

Example 8 Wgtg:  100% 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
"IOGP 434‐1 & 434‐3"

Current FRT Category
Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"RIVM BEVI '09"
FRED JUN '12"
CURRENT 
PHMSA 

'09"
Index FRT SPEC
729 Release from hole with eff. diam. of 25 mm (1 in) diam.
730 Release from hole with eff. diam. of 50 mm (2 in) diam.
731 Release from hole with eff. diam. of 100 mm (4 in) diam.
732 Gasket Flange‐ Non Cryogenic: 2"≤D<6"
733 Catastrophic Rupture
734 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
735 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
736 Release from hole with eff. diam. of 25 mm (1 in) diam.
737 Release from hole with eff. diam. of 50 mm (2 in) diam.
738 Release from hole with eff. diam. of 100 mm (4 in) diam.
739 Gasket Flange: 6"≤D<12"
740 Catastrophic Rupture
741 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
742 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
743 Release from hole with eff. diam. of 25 mm (1 in) diam.
744 Release from hole with eff. diam. of 50 mm (2 in) diam.
745 Release from hole with eff. diam. of 100 mm (4 in) diam.
746 Gasket Flange‐ Cryogenic: 6"≤D<12"
747 Catastrophic Rupture
748 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
749 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
750 Release from hole with eff. diam. of 25 mm (1 in) diam.
751 Release from hole with eff. diam. of 50 mm (2 in) diam.
752 Release from hole with eff. diam. of 100 mm (4 in) diam.
753 Gasket Flange‐ Non Cryogenic: 6"≤D<12"
754 Catastrophic Rupture
755 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
756 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
757 Release from hole with eff. diam. of 25 mm (1 in) diam.
758 Release from hole with eff. diam. of 50 mm (2 in) diam.
759 Release from hole with eff. diam. of 100 mm (4 in) diam.
760 Gasket Flange: 12"≤D<20"
761 Catastrophic Rupture
762 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
763 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
764 Release from hole with eff. diam. of 25 mm (1 in) diam.
765 Release from hole with eff. diam. of 50 mm (2 in) diam.
766 Release from hole with eff. diam. of 100 mm (4 in) diam.
767 Gasket Flange‐ Cryogenic: 12"≤D<20"
768 Catastrophic Rupture
769 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
770 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
771 Release from hole with eff. diam. of 25 mm (1 in) diam.
772 Release from hole with eff. diam. of 50 mm (2 in) diam.
773 Release from hole with eff. diam. of 100 mm (4 in) diam.
774 Gasket Flange‐ Non Cryogenic: 12"≤D<20"
775 Catastrophic Rupture
776 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
777 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
778 Release from hole with eff. diam. of 25 mm (1 in) diam.
779 Release from hole with eff. diam. of 50 mm (2 in) diam.
780 Release from hole with eff. diam. of 100 mm (4 in) diam.
781 Gasket Flange: 20"≤D<40"
782 Catastrophic Rupture
783 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
784 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
785 Release from hole with eff. diam. of 25 mm (1 in) diam.
786 Release from hole with eff. diam. of 50 mm (2 in) diam.
787 Release from hole with eff. diam. of 100 mm (4 in) diam.
788 Gasket Flange‐ Cryogenic: 20"≤D<40"
789 Catastrophic Rupture
790 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
791 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
792 Release from hole with eff. diam. of 25 mm (1 in) diam.
793 Release from hole with eff. diam. of 50 mm (2 in) diam.
794 Release from hole with eff. diam. of 100 mm (4 in) diam.
795 Gasket Flange‐ Non Cryogenic: 20"≤D<40"
796 Catastrophic Rupture
797 Release from hole with eff. diam. of 2 mm (0.08 in) diam.
798 Release from hole with eff. diam. of 10 mm (0.4 in) diam.
799 Release from hole with eff. diam. of 25 mm (1 in) diam.
800 Release from hole with eff. diam. of 50 mm (2 in) diam.
801 Release from hole with eff. diam. of 100 mm (4 in) diam.
181 Piping:  All diameters
182 3.06E‐06 2.50E‐06 2.12E‐06 5.30E‐07 3.58E‐07 4.66E‐07 2.56E‐08 1.72E‐05 Catastrophic Rupture
183 1.65E‐06 1.35E‐06 1.11E‐06 4.74E‐07 2.29E‐07 7.42E‐06 Release from hole with eff. diam. of 1/3 diam.
184 8.07E‐06 7.71E‐06 7.75E‐06 2.59E‐06 2.59E‐06 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
185 8.42E‐06 9.18E‐06 8.79E‐06 6.86E‐06 1.60E‐06 Release from hole with eff. diam. of 25mm (1 in)
186 Piping:  d < 50mm (2‐inch)
187 1.00E‐06 9.25E‐06 7.87E‐06 6.83E‐06 1.99E‐06 1.01E‐06 9.93E‐07 2.54E‐08 2.69E‐05 Catastrophic Rupture
188 Release from hole with eff. diam. of 1/3 diam.
189 5.09E‐06 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
190 5.00E‐06 5.01E‐06 Release from hole with eff. diam. of 25mm (1 in)
191 Piping:  d < 50mm (2‐inch) ‐ Non‐Cryogenic
192 1.05E‐05 8.22E‐06 7.10E‐06 5.07E‐07 9.93E‐07 2.54E‐08 Catastrophic Rupture
193 Release from hole with eff. diam. of 1/3 diam.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 165
Appendix F: "Wisdom of the Crowd" Results with Table 14 Weighting Factors Applied
and Using All Refernces "Specifically Included in Analysis" on Table 11

Example 5 Wgtg:  100% "HSE 

Example 7 Wgtg:  100% "LNE 

Super Category Comparison
Example 1 Wgtg:  Uniform

Example 6 Wgtg:  100% 

Example 8 Wgtg:  100% 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
"IOGP 434‐1 & 434‐3"

Current FRT Category
Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"RIVM BEVI '09"
FRED JUN '12"
CURRENT 
PHMSA 

'09"
Index FRT SPEC
194 5.09E‐06 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
195 Release from hole with eff. diam. of 25mm (1 in)
196 Piping:  d < 50mm (2‐inch) ‐ Cryogenic
197 Catastrophic Rupture
198 Release from hole with eff. diam. of 1/3 diam.
199 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
200 2.29E‐05 2.29E‐05 2.29E‐05 2.29E‐05 Release from hole with eff. diam. of 25mm (1 in)
201 Piping:  d < 50mm (2‐inch) ‐ Cryogenic ‐ VJ PIPING
202 Catastrophic Rupture
203 Release from hole with eff. diam. of 1/3 diam.
204 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
205 Release from hole with eff. diam. of 25mm (1 in)
206 Piping:  d < 50mm (2‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
207 Catastrophic Rupture
208 Release from hole with eff. diam. of 1/3 diam.
209 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
210 Release from hole with eff. diam. of 25mm (1 in)
211 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch)
212 5.00E‐07 2.49E‐06 2.13E‐06 1.81E‐06 5.44E‐07 4.88E‐07 3.08E‐07 2.36E‐08 7.70E‐06 Catastrophic Rupture
213 7.42E‐06 Release from hole with eff. diam. of 1/3 diam.
214 2.02E‐06 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
215 2.00E‐06 1.04E‐06 Release from hole with eff. diam. of 25mm (1 in)
216 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Non‐Cryogenic
217 2.10E‐06 2.04E‐06 1.90E‐06 2.91E‐08 2.35E‐08 Catastrophic Rupture
218 Release from hole with eff. diam. of 1/3 diam.
219 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
220 Release from hole with eff. diam. of 25mm (1 in)
221 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic
222 Catastrophic Rupture
223 Release from hole with eff. diam. of 1/3 diam.
224 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
225 2.31E‐05 2.31E‐05 2.31E‐05 2.31E‐05 Release from hole with eff. diam. of 25mm (1 in)
226 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic ‐ VJ PIPING
227 Catastrophic Rupture
228 Release from hole with eff. diam. of 1/3 diam.
229 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
230 Release from hole with eff. diam. of 25mm (1 in)
231 Piping:  50mm (2‐inch) ≤ d < 149mm (6‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
232 Catastrophic Rupture
233 Release from hole with eff. diam. of 1/3 diam.
234 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
235 Release from hole with eff. diam. of 25mm (1 in)
236 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch)
237 2.00E‐07 1.73E‐06 1.30E‐06 1.01E‐06 1.40E‐07 1.96E‐07 3.36E‐08 Catastrophic Rupture
238 4.00E‐07 4.28E‐07 3.85E‐07 3.90E‐07 4.32E‐07 3.91E‐07 Release from hole with eff. diam. of 1/3 diam.
239 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
240 7.00E‐07 7.21E‐07 Release from hole with eff. diam. of 25mm (1 in)
241 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Non‐Cryogenic
242 9.18E‐07 9.78E‐07 9.46E‐07 2.89E‐08 3.36E‐08 Catastrophic Rupture
243 Release from hole with eff. diam. of 1/3 diam.
244 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
245 Release from hole with eff. diam. of 25mm (1 in)
246 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic
247 Catastrophic Rupture
248 Release from hole with eff. diam. of 1/3 diam.
249 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
250 2.30E‐05 2.30E‐05 2.30E‐05 2.30E‐05 Release from hole with eff. diam. of 25mm (1 in)
251 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic ‐ VJ PIPING
252 Catastrophic Rupture
253 Release from hole with eff. diam. of 1/3 diam.
254 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
255 Release from hole with eff. diam. of 25mm (1 in)
256 Piping:  150mm (6‐inch) ≤ d < 299mm (12‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
257 Catastrophic Rupture
258 Release from hole with eff. diam. of 1/3 diam.
259 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
260 Release from hole with eff. diam. of 25mm (1 in)
261 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch)
262 7.00E‐08 1.99E‐06 1.39E‐06 1.04E‐06 8.58E‐08 7.20E‐08 2.35E‐08 Catastrophic Rupture
263 2.00E‐07 9.94E‐08 1.16E‐07 1.39E‐07 1.94E‐07 1.94E‐07 Release from hole with eff. diam. of 1/3 diam.
264 4.00E‐07 1.12E‐07 1.10E‐07 1.04E‐07 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
265 5.00E‐07 1.79E‐06 1.57E‐06 1.24E‐06 4.95E‐07 4.95E‐07 Release from hole with eff. diam. of 25mm (1 in)
266 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Non‐Cryogenic
267 8.55E‐07 9.29E‐07 9.38E‐07 2.64E‐08 2.35E‐08 Catastrophic Rupture
268 Release from hole with eff. diam. of 1/3 diam.
269 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
270 Release from hole with eff. diam. of 25mm (1 in)
271 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic
272 Catastrophic Rupture
273 Release from hole with eff. diam. of 1/3 diam.
274 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
275 Release from hole with eff. diam. of 25mm (1 in)
276 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic ‐ VJ PIPING
277 Catastrophic Rupture
278 Release from hole with eff. diam. of 1/3 diam.
279 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 166
Appendix F: "Wisdom of the Crowd" Results with Table 14 Weighting Factors Applied
and Using All Refernces "Specifically Included in Analysis" on Table 11

Example 5 Wgtg:  100% "HSE 

Example 7 Wgtg:  100% "LNE 

Super Category Comparison
Example 1 Wgtg:  Uniform

Example 6 Wgtg:  100% 

Example 8 Wgtg:  100% 

Potential SubCategory1
Potential SubCategory2

Potential SubCategory4
Potential Subcategory3
"IOGP 434‐1 & 434‐3"

Current FRT Category
Example 2 Wgtg.

Example 3 Wgtg.

Example 4 Wgtg.

"RIVM BEVI '09"
FRED JUN '12"
CURRENT 
PHMSA 

'09"
Index FRT SPEC
280 Release from hole with eff. diam. of 25mm (1 in)
281 Piping:  300mm (12‐inch) ≤ d < 499mm (20‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
282 Catastrophic Rupture
283 Release from hole with eff. diam. of 1/3 diam.
284 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
285 Release from hole with eff. diam. of 25mm (1 in)
286 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch)
287 2.00E‐08 1.89E‐06 1.17E‐06 8.38E‐07 4.69E‐08 4.05E‐08 9.39E‐08 2.14E‐08 Catastrophic Rupture
288 1.00E‐07 4.95E‐08 5.74E‐08 6.87E‐08 9.51E‐08 9.51E‐08 Release from hole with eff. diam. of 1/3 diam.
289 2.00E‐07 2.62E‐07 3.02E‐07 3.24E‐07 4.89E‐07 4.89E‐07 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
290 4.00E‐07 9.07E‐07 9.80E‐07 1.05E‐06 4.32E‐07 4.32E‐07 Release from hole with eff. diam. of 25mm (1 in)
291 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Non‐Cryogenic
292 8.01E‐07 8.21E‐07 8.38E‐07 2.69E‐08 2.14E‐08 Catastrophic Rupture
293 Release from hole with eff. diam. of 1/3 diam.
294 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
295 Release from hole with eff. diam. of 25mm (1 in)
296 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic
297 Catastrophic Rupture
298 Release from hole with eff. diam. of 1/3 diam.
299 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
300 Release from hole with eff. diam. of 25mm (1 in)
301 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic ‐ VJ PIPING
302 Catastrophic Rupture
303 Release from hole with eff. diam. of 1/3 diam.
304 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
305 Release from hole with eff. diam. of 25mm (1 in)
306 Piping:  500mm (20‐inch) ≤ d < 1000 mm (40‐inch) ‐ Cryogenic ‐ NON‐VJ PIPING
307 Catastrophic Rupture
308 Release from hole with eff. diam. of 1/3 diam.
309 Release from hole with effective diameter of 10% diameter, up to 50mm (2‐inches)
310 Release from hole with eff. diam. of 25mm (1 in)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 167
Appendix G: Excerpts from Analysis of LNG Peakshaving Facility Release
Prevention Systems (PNL-4153)

The following excerpts from Analysis of LNG Peakshaving Facility Release Prevention Systems
by Pelto, P.J., Baker, E.G., Powers, T.B., Schreiber, A.M., Hoggs, J.M., and Daling, P.M. of
Pacific Northwest Laboratory, PNL-4153, 1982 (“PNL PSRP ‘82”) are provided directly in this
Final Report for the reader’s convenience.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 168
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 169
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 170
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 171
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 172
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 173
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 174
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 175
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 176
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 177
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 178
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 179
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 180
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 181
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 182
Appendix H: Excerpts from Analysis of LNG Import Terminal Release Prevention
Systems (PNL-4152)

The following excerpts from Analysis of LNG Import Terminal Release Prevention Systems by
Baker, E.G. of Pacific Northwest Laboratory, PNL-4152, 1982 (“PNL ITRP ‘82”) are provided
directly in this Final Report for the reader’s convenience.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 183
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 184
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 185
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 186
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 187
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 188
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 189
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 190
Appendix I: Excerpts from LNG Terminal Risk Assessment Study for Oxnard,
California (SAI-75-615-LJ)

The following excerpts from the LNG Terminal Risk Assessment Study for Oxnard, California,
prepared for Western LNG Terminal Co. by Science Applications Inc., SAI-75-615-LJ, Dec. 22,
1975 (“SAI ‘75”) are provided directly in this Final Report for the reader’s convenience. Table
3.1 summarizes the failure rate database used, and Figure 3.16 shows the FTA that is described
in the following excerpt from Section 3.5 “Probabilistic Estimates of Hazards” on p. 3-35:

The results of the study show that the largest probability, 4.0 x 10-6/year, for a
leak of LNG under normal operations occurs for a major rupture of one of the storage
tanks. This probability reflects the compounding of a mean failure rate of 1.0 x 10-6/tank-
year from the data base for a single tank failure. Such a high probability is strictly
applicable only to pressure vessels exposed to a high-pressure environment but was used
because no other information was available.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 191
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 192
Appendix J: Analysis of PHMSA’s Onshore Natural Gas Transmission Pipeline
Incident Data

GTI analyzed the incidents and population data sets between 2002-2015 recorded in PHMSA’s
Natural Gas Transmission and Gathering incident and annual reports. GTI’s analysis built upon
the methodology contained in the 2015 analysis of PHMSA’s Onshore Natural Gas Transmission
data by Chio Lam, Statistical Analyses of Historical Pipeline Incident Data with Application to
the Risk Assessment of Onshore Natural Gas Transmission Pipelines, Master of Science Thesis,
The University of Western Ontario, July 2015 (“PHMSA NGT LAM ’15”). In this report,
PHMSA’s Natural Gas Transmission report data is abbreviated as PHMSA NGT.

As in Lam’s analysis, GTI’s analysis of rupture rates focused on on-shore Natural Gas
Transmission incident and population data, and excluded consideration of: gathering line
incidents and population; and off-shore natural gas transmission incidents and population data.
As in Lam’s analysis, an analysis of leak (vs. rupture) incidents was not performed since only
incidents that meet the reporting threshold of US 49 CFR Section 191.3 are represented in the
data, whereas it is expected that all ruptures are reportable incidents. Gathering line rupture data
was not analyzed since a substantial amount of gathering pipeline materials are non-steel,
whereas virtually all transmission piping is steel; for example, in the 2015 population data set,
9.11% of gathering pipeline materials are non-steel and 99.46% of transmission pipelines are
steel.

In total, we analyzed three cases and included additional subsets of incident data which are
defined below as Cases 1, 2 and 3, in order to build upon the methodology used by Lam:

PHMSA NGT GTI PHMSA NGT PHMSA NGT GTI

Data Entry CASE 1 GTI CASE 2 CASE 3
ONSHORE INCLUDE INCLUDE INCLUDE
PIPE(PIPE BODY or PIPE
SEAM) INCLUDE INCLUDE INCLUDE
TRANSMISSION INLCUDE
SYSTEM INCLUDE INCLUDE
PIPING on
COMPRESSOR STATION INCLUDE
& STORAGE INCLUDE INCLUDE
TRANSMISSION LINE
OF DISTRIBUTION INCLUDE
SYSTEM EXCLUDE INCLUDE
WELD, INCLUDING
HEAT-AFFECTED ZONE EXCLUDE EXCLUDE INCLUDE*
Believed by GTI Final
to Provide a Direct Results
Purpose of Analysis
Comparison with Intermediate “PHMSA NGT GTI
LAM’s Thesis Results ‘16”
Table J.1.1: Data extraction for different cases
* Includes 5 rupture incidents in the 2010-2015 data set, all of which are categorized in “Weld
Subtype” as “Pipe Girth Weld”

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 193
First, we compared the total mileage with PHMSA NGT and the PHMSA NGT online data
(http://www.phmsa.dot.gov/pipeline/library/data-stats/annual-report-mileage-for-natural-gas-
transmission-and-gathering-systems) and showed good agreement:

PHMSA NGT PHMSA NGT PHMSA NGT LAM '15

YEAR Annual Report GTI Analysis (Figure 2.1 in LAM’s
Online[1] Thesis)
2002 477621.7 476420.5 476490.6
2003 475382.8 475374.4 475348.4
2004 477864.2 477869.3 477963.8
2005 474413.0 470079.8 470093.9
2006 472651.5 472697.1 472745.1
2007 474636.3 474640.6 474716.3
2008 478383.4 478389.0 478476.7
2009 481135.2 481125.2 481163.5
2010 481747.7 481760.5 479949.8
2011 482344.3 482350.6 480525.4
2012 480481.7 480533.3 479633.7
2013 480103.6 480110.1 479064.1
Table J.2.1: Total transmission mileage (in km)

The total mileage for each year was computed using the summation of mileage for each diameter
in the population data. Mileage data reported to NPS൑4” was applied in this analysis as
2”൑NPS൑4”. This was considered a reasonable approximation since only 3 of the 37 incidents
(i.e. 8.1%) reported during 2010-2015 for pipe diameter ≤ 4” were for pipe diameters < 2”,
which provides an indication of the underlying distribution of all piping ≤ 4”; in addition, all
piping ruptures reported during 2010-2015 were ≥ 2” in diameter.

Second, we replicated the results of Master’s Thesis by Chio Lam and showed good agreement
with our Case 1 results.

Lam, Chio, Statistical Analyses of Historical Pipeline Incident Data with Application to the 
Risk Assessment of Onshore Natural Gas Transmission Pipelines, Master of Science 
Thesis, The University of Western Ontario, July 2015 [“PHMSA NGT LAM ‘15] 
PHMSA PHMSA PHMSA
Diameter PHMSA NGT NGT GTI NGT GTI NGT GTI
(inch) LAM'15 CASE 1 CASE 2 CASE 3
Unknown 0.40% 0.43% 0.40% 0.39%
0<d൑4 8.60% 8.86% 8.80% 8.93%
4<d൑10 22.20% 22.25% 22.20% 21.94%
10<d൑20 38.60% 38.23% 38.60% 37.67%
20<d൑8 15.30% 15.33% 15.80% 16.70%
28<d 14.90% 14.90% 14.20% 14.37%
Table J.2.2: Distribution of all incidents by diameter (2002-2013)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 194
 45.00%
40.00%
PHMSA NGT LAM '15
Distribution of all incidents 35.00%
PHMSA NGT GTI CASE 1
30.00%
PHMSA NGT GTI CASE 2
25.00% PHMSA NGT GTI CASE 3
20.00%
15.00%
10.00%
5.00%
0.00%
Unknown 0<d<=4 4<d<=10 10<d<=20 20<d<=28 28<d
Diameter (inch)
Figure J.2.2: Distribution of all incidents by diameter (2002-2013)

PHMSA PHMSA PHMSA

Diameter PHMSA NGT NGT GTI NGT GTI NGT GTI
(inch) LAM'15 CASE 1 CASE 2 CASE 3
0<d൑4 2.20E-05 2.22e-5 2.46e-5 2.67e-5
4<d൑10 3.00E-05 2.96e-5 3.12e-5 3.12e-5
10<d൑20 4.60E-05 4.65e-5 5.00e-5 5.06e-5
20<d൑8 2.50E-05 2.49e-5 2.81e-5 3.02e-5
28<d 2.00E-05 2.00e-5 2.00e-5 2.06e-5
Table J.2.3: Onshore rupture rates/km-year by diameter (2002-2013)

6.00E‐05
PHMSA NGT LAM '15
5.00E‐05
PHMSA NGT GTI CASE 1
Rupture rate (/km‐year)

PHMSA NGT GTI CASE 2
4.00E‐05
PHMSA NGT GTI CASE 3
3.00E‐05

2.00E‐05

1.00E‐05

0.00E+00
0<d<=4 4<d<=10 10<d<=20 20<d<=28 28<d
Diameter(inch)
Figure J.2.3: Onshore rupture rates/km-year by diameter (2002-2013)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 195
Third, we confirmed that the rupture rates from 2002 – 2015 are quite similar to the rupture rates
from 2010 – 2015 for three analysis cases. Thus, the 2010-15 incident data is quite
representative of the 2002-2015 incident data.

PHMSA NGT GTI PHMSA NGT GTI PHMSA NGT GTI

Diameter
CASE 1 CASE 2 CASE 3
(inch)
2002 – 2015 2010-2015 2002 – 2015 2010-2015 2002 – 2015 2010-2015
0<d൑4 2.31E-05 1.86E-05 2.51e-5 2.33e-5 2.69e-5 2.33e-5
4<d൑10 2.95E-05 2.74E-05 3.09e-5 3.06e-5 3.09e-5 3.06e-5
10<d൑0 4.56E-05 4.52E-05 4.92e-5 4.77e-5 4.97e-5 4.89e-5
20<d൑28 2.68E-05 2.13E-05 2.96e-5 2.13e-5 3.14e-5 2.55e-5
28<d 1.82E-05 1.71E-05 1.82e-5 1.71e-5 1.99e-5 2.10e-5
Table J.3.1: Onshore rupture rates/km-year by diameter
 

5.00E‐05 PHMSA NGT GTI CASE 1
Rupture rate (/km‐year)

4.00E‐05 2002‐2015
2010‐2015
3.00E‐05

2.00E‐05

1.00E‐05

0.00E+00
0<d<=4 4<d<=10 10<d<20 20<d<=28 28<d
Diameter (inch)

6.00E‐05
PHMSA NGT GTI CASE 2
5.00E‐05
Rupture rate (/km‐year)

2002‐2015
2010‐2015
4.00E‐05

3.00E‐05

2.00E‐05

1.00E‐05

0.00E+00
0<d<=4 4<d<=10 10<d<20 20<d<=28 28<d
Diameter (inch)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 196
 6.00E‐05
PHMSA NGT GTI CASE 3
5.00E‐05 2002‐2015

Rupture rate (/km‐year)
2010‐2015
4.00E‐05

3.00E‐05

2.00E‐05

1.00E‐05

0.00E+00
0<d<=4 4<d<=10 10<d<20 20<d<=28 28<d
Diameter (inch)

Figure J.3.1: Onshore rupture rates/km-year by diameter

 
Fourth, we considered only the pipeline rupture data from 2010 to 2015 in order calculate
the rupture rates for the diameter ranges in Table 4.1 that match the diameter ranges in the
FRT. The data from 2002 – 2009 could not be included in this manner because it does not
contain actual pipeline diameters (and instead only provides a range of diameters such as
0<d<=4, 4<d<=10, 10<d<=20, 20<d<=28 and 28<d).

Diameter PHMSA NGT PHMSA NGT PHMSA NGT

(inch) GTI CASE 1 GTI CASE 2 GTI CASE 3
0൑d<2
2൑d<6 2.31E-05 2.78E-05 2.78E-05
6൑d<12 2.91E-05 3.23E-05 3.23E-05
12൑d<20 3.68E-05 4.03E-05 4.20E-05
20൑d<40 2.68E-05 2.68E-05 2.96E-05
Table J.4.1: Onshore rupture rates/km-year by FRT category diameter (2010-15)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 197
 4.50E‐05
PHMSA NGT GTI CASE 1
4.00E‐05
PHMSA NGT GTI CASE 2
3.50E‐05
Rupture rate (/km‐year)
PHMSA NGT GTI CASE 3
3.00E‐05

2.50E‐05

2.00E‐05

1.50E‐05

1.00E‐05

5.00E‐06

0.00E+00
0<=d<2 2<=d<6 6<=d<12 12<=d<20 20<=d<40
Diameter (inch)
Figure J.4.1: Rupture rates/km-year by FRT category diameter (2010-2015)

The Case 3 results shown in Table J.4.1 above were entered into the database for this project as
“PHMSA NGT GTI ‘16”.
As an additional clarification, no ruptures of valves are identified in this 2010-2015 database.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 198
Appendix K: Analysis of PHMSA’s Onshore Hazardous Liquids Pipeline Accident
Data

GTI analyzed the accident (referred to as incident hereafter in this appendix) and population data
sets between 2010-2015 recorded in PHMSA’s Hazardous Liquids accident and annual reports,
to provide directly comparable data to our analysis of on-shore natural gas transmission pipeline
ruptures during 2010-15 as summarized in Appendix J.

In total, we analyzed two cases with different subsets of onshore rupture incidents data as shown
in Table K.1.1. One case included those incidents occurring onshore and pipe related (pipe body
or pipe seam) or weld, including the heat affected zone but excluded those occurring on piping at
pump/meter station and terminal/tank farm. Another case included those incidents occurring
onshore and pipe related (pipe body or pipe seam) or weld, including the heat affected zone and
also included piping at pump/meter station and terminal/tank farm. In this report, PHMSA
Hazardous Liquid is abbreviated as PHMSA HL.

PHMSA HL GTI PHMSA HL GTI

Data Entry CASE 1 CASE 2
ONSHORE INCLUDE INCLUDE
PIPE(PIPE BODY or PIPE
SEAM) INCLUDE INCLUDE
WELD, INCLUDING HEAT-
AFFECTED ZONE INCLUDE INCLUDE
PUMP/METER STATION EXCLUDE INCLUDE
TERMINAL/TANK FARM EXCLUDE INCLUDE
Table K.1.1: Data extraction for Case 1 and Case 2

First, we compared the total mileage with PHMSA NGT and the PHMSA HL online data
(http://www.phmsa.dot.gov/pipeline/library/data-stats/annual-report-mileage-for-hazardous-
liquid-or-carbon-dioxide-systems) and found good agreement as shown below in the first two
columns of Table K.2.1. The total mileage for each year was computed using the summation of
mileage for each dimeter in the population data. Column 3 of Table K.2.1 also shows the
mileage contained in PHMSA’s Natural Gas Transmission reports, to provide an understanding
of the comparative sizes of these two databases.

PHMSA HL GTI PHMSA HL PHMSA NGT GTI

Year
Online[1]
2010 284514.1 292864.5 481760.5
2011 287247.5 295429.7 482350.6
2012 291921.9 299679.8 480533.3
2013 301735.9 309642.7 480110.1
2014 313150.2 321305.1 479457.4
2015 327165.0 335523.0 478389.9
Table K.2.1: Total transmission mileage (in km)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 199
Second, we computed the breakdown of the number of onshore rupture incidents by predominant
commodity carried in the pipeline.

Number of Rupture Incidents

Predominant
Commodity PHMSA HL GTI PHMSA HL GTI
CASE 1 CASE 2
Crude Oil 12 19
HVL 15 15
Non-HVL 12 13
CO2 1 1
Table K.3.1: Breakdown of the number of onshore rupture incidents
by predominant commodity (2010-2015)

20
18
PHMSA HL GTI CASE 1
16
# of rupture incidents

PHMSA HL GTI CASE 2
14
12
10
8
6
4
2
0
Crude Oil HVL Non‐HVL CO2
Commodity
Figure K.3.1: Breakdown of the number of onshore rupture incidents
by predominant commodity (2010-2015)

Third, we computed the rupture rates by predominant commodity carried, and confirmed that
there were not substantial (e.g. >5x) differences in overall rupture rates for the different
predominant commodities carried.
Rupture Rate (/km-year)
Predominant
Commodity PHMSA HL GTI PHMSA HL GTI
CASE 1 CASE 2
Crude Oil 2.21E-05 3.58E-05
HVL 2.53E-05 2.53E-05
Non-HVL 1.95E-05 2.11E-05
CO2 2.00E-05 2.00E-05
Table K.4.1: Onshore rupture rates/km-year by commodity (2010-2015)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 200
 4.00E‐05
3.50E‐05

Rupture rate ( /km‐year)
PHMSA HL GTI CASE 1
3.00E‐05
PHMSA HL GTI CASE 2
2.50E‐05
2.00E‐05
1.50E‐05
1.00E‐05
5.00E‐06
0.00E+00
Crude Oil HVL Non‐HVL CO2
Commodity

Figure K.4.1: Onshore rupture rates/km-year by commodity (2010-2015)

Fourth, we computed the rupture rates by diameter ranges that match those in the FRT for all of
the data in the dataset (i.e. including all predominant commodity carried). The Case 2 results
shown in Table K.5.1 above were entered into the database for this project as “PHMSA HL GTI
‘16”. As an additional clarification, one valve rupture is identified in this 2010-2015 database
(in crude oil service) but the valve size is unidentified and the “Rupture Details” column
identifies that “DIAPHRAGM WAS CRACKED. REPLACED DIAPHRAGM”; this sole valve
rupture was not included in the analysis but if it had it would increase the rate by 2% (i.e. 49/48).
Rupture Rate (/km-year)
Diameter
(inch) PHMSA HL GTI PHMSA HL GTI
CASE 1 CASE 2
0൑d<2
2൑d<6 3.95e-5 3.95e-5
6൑d<12 2.22e-5 2.45e-5
12൑d<20 1.37e-5 2.01e-5
20൑d<40 3.51e-5 3.90e-5
Table K.5.1: Onshore rupture rates/km-year by FRT diameter category (2010-2015)
4.50E‐05
4.00E‐05 PHMSA HL GTI CASE 1
PHMSA HL GTI CASE 2
Rupture rate (/km‐year)

3.50E‐05
3.00E‐05
2.50E‐05
2.00E‐05
1.50E‐05
1.00E‐05
5.00E‐06
0.00E+00
2<=d<6 6<=d<12 12<=d<20 20<=d<40
Diameter (inch)

Figure K.5.1: Onshore rupture rates/km-year by FRT diameter ranges (2010-2015)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 201
Fifth, for PHMSA’s and FERC’s additional information we computed the rupture rates by any
diameter broken down by commodity for case 1.
Diameter  Rupture Rated   Breakdown by  Rupture 
(inch)  (All  Commodity  Rates 
Commodities)  (Breakdown 
by 
Commodity) 
0൑d<2       
2൑ d< 6    Crude Oil  1.30e‐5 
  HVL  1.34e‐5 
3.95E‐05  Non‐HVL  0 
CO2  1.31e‐5 
    Crude Oil  2.23e‐6 
    HVL  1.12e‐5 
6൑d<12  2.22E‐05  Non‐HVL  8.80e‐6 
CO2  0 
    Crude Oil  1.78e‐6 
    HVL  5.63e‐6 
12൑d<20  1.37E‐05  Non‐HVL  6.32e‐6 
CO2  0 
    Crude Oil  2.85e‐5 
    HVL  2.94e‐6 
20൑d<40  3.51E‐05  Non‐HVL  3.70e‐6 
CO2  0 
Table K.6.1: Onshore rupture rates/km-year by FRT diameter category; breakdown by commodity
(2010-2015)

4.50E‐05
CO2
4.00E‐05 Non‐HVL
3.50E‐05 HVL
Rupture rate (/km‐year)

Crude Oil
3.00E‐05

2.50E‐05

2.00E‐05

1.50E‐05

1.00E‐05

5.00E‐06

0.00E+00
2<=d<6 6<=d<12 12<=d<20 20<=d<40
Diameter (inch)
Figure K.6.1: Onshore rupture rates/km-year by FRT diameter category; breakdown by
commodity (2010-2015)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 202
Sixth, for PHMSA’s and FERC’s additional information we computed the rupture rates by FRT
diameter category broken down by commodity for case 2.
Diameter  Rupture Rated   Breakdown by  Rupture 
(inch)  (All  Commodity  Rates 
Commodities)  (Breakdown 
by 
Commodity) 
0൑d<2       
    Crude Oil  1.30e‐5 
    HVL  1.34e‐5 
2൑ d< 6  3.95E‐05  Non‐HVL  0 
CO2  1.31e‐5 
    Crude Oil  4.49e‐6 
    HVL  1.12e‐5 
6൑d<12  2.45E‐05  Non‐HVL  8.80e‐6 
CO2  0 
    Crude Oil  5.98e‐6 
    HVL  5.63e‐6 
12൑d<20  2.01E‐05  Non‐HVL  8.46e‐6 
CO2  0 
    Crude Oil  3.23e‐5 
    HVL  2.94e‐6 
20൑d<40  3.90E‐05  Non‐HVL  3.70e‐6 
CO2  0 
Table K.7.1: Onshore rupture rates/km-year by FRT diameter category; breakdown by commodity
(2010-2015)

4.50E‐05

4.00E‐05

3.50E‐05
Rupture rate (/km‐year)

3.00E‐05

2.50E‐05
CO2
2.00E‐05
Non‐HVL
1.50E‐05 HVL

1.00E‐05

5.00E‐06

0.00E+00
2<=d<6 6<=d<12 12<=d<20 20<=d<40
Diameter (inch)
Figure K.7.1: Onshore rupture rates/km-year by FRT diameter category; breakdown by
commodity (2010-2015)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 203
Appendix L: Visualization of UK HSE Hydrocarbon Releases Database (HCRD)

The UK HSE Hydrocarbon Releases Database (HCRD) System is one often-cited equipment
failure rate database for oil and gas, CPI and related applications. The HCRD System has
continuously collected reliability data from UK off-shore oil and gas operations since 1992. The
HCRD database is the basis for DNV’s commercial LEAK software that estimates piping and
equipment leak frequencies, as described in “DNV FFG HCRD ‘13”.

GTI generated the figures in this Appendix directly from the data in the HCRD from fiscal years
1992/1993 to 2015/2016, in order to help PHMSA and FERC better understand the distribution
of some of the components and leak incidents from this large public database. GTI used the data
that is publicly available in Microsoft Excel® files entitled “Offshore Hydrocarbon Population
Data 1992 - 2015” and “Offshore Hydrocarbon Releases 1992 – 2015” which are available as of
the date of this report at http://www.hse.gov.uk/offshore/statistics.htm.

The HCRD uses the term “Pressure Vessel” for equipment that is in service such as a: knock-out
drum, reboiler, separator, or other. Because these uses are primarily process-related we use the
terminology “Pressure (Process) Vessel” in this Appendix for this class of equipment in the
HCRD.

1. Piping

800

700

600

500

400

300

200

100

0
0D3 3D11 11D+
Steel piping size class

Figure L.1: Number of steel piping leak incidents reported by size category (D≤3”, 3”<D≤11”,
D>11”)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 204
 0
50
100
150
200
250
300
0
50
100
150
200
250

0.011
0.0196850393700787
0.081
0.181 0.196850393700787
0.247 0.25
0.348
0.37
0.4416376
0.579 0.375196850393701
0.686 0.393700787401575
0.81
0.5
1.041542
1.184 0.625
1.3 0.78740157480315
1.496181
1.639 0.874015748031496
1.876991 1.00393700787402
2.1
1.49606299212598
2.327
2.7 1.96850393700787
3 2.00787401574803
3.4
2.99212598425197
3.755
4.019 3.50393700787402
Steel piping size (inches)

4.514244 4.01574803149606
5.178
5.90551181102362

Steel piping equivalent hole diameter (mm)
5.828198

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table
6.638 6.00393700787402
7.5 7
9
10.154 8
12 11
15.772
13.7795275590551
Figure L.2: Number of steel piping leak incidents reported by size (inches)

19.1

Figure L.3: Number of steel piping holes by equivalent hole diameter (mm)

Page 205
25 16
36.672 20
65
28
264
36
 2. Flanges

250

200

150

100

50

0
HAMMER UNION SPIRAL WOUND OTHER CLAMP (e.g., RING TYPE (RTJ) COMPRESSED
(CHICSAN) GRAYLOC)

Flange type

Figure L.4: Number of flange leak incidents reported by type

250

200

150

100

50

0
0D3 3D11 11D+

Flange size class

Figure L.5: Number of flange leak incidents reported by size category (D≤3”, 3”<D≤11”, D>11”)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 206
 80
70
60
50
40
30
20
10
0

Flange size (inches)

Figure L.6: Number of flange leak incidents reported by size (inches)

140

120

100

80

60

40

20

0
0.01
0.12
0.2
0.332
0.4
0.51
0.598343
0.751
0.942
1.13
1.25
1.479328
1.6
1.90877
2.1
2.3
2.747
3.242
3.8
4.7
4.98
5.3
5.8
6.753
7.3
8.6
10.1
12.787
17
24
38.1
76.2
138.1
401

Flange equivalent hole diameter (mm)

Figure L.7: Number of flange holes by equivalent hole diameter (mm)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 207
 4.822

WOUN
SPIRAL

D
0.621
100
21.5
10.1
6
RING TYPE (RTJ) 3.242
2.125
1.7
1.4
1
0.563
0.4
0.247
0.113
CSA OTHE

1
R

0.2
HAM
MER

(CHI
UNI
ON

N)

1.188792
139.7
50
24
15
9.2
8
7
5.89
COMPRESSED

5.418
4.98
4.3
3.5
2.5
2
1.6
1.233
0.8
0.351
0.15
254
CLAMP (e.g.,

5.08
GRAYLOC)

3
1.9
0.77
0.01
0 10 20 30 40 50 60 70

Figure L.8: Number of flange holes by equivalent hole diameter (mm) and type

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 208
 0
5
10
15
20
25
30
35
40
45
14.5038
150
181.2975
250.065527244
274.992048
275.5722
300.22866
507.633
700.183418268
720.403746
899.2356
960.15156
1440.22734
1480.3883622
1653.4332
2102.8624506

Flange rating (psig) 
2160.631086
2485.95132
2741.2182
3016.7904
3190.836
3600.858426

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table
3698.469
4641.216
5003.811
6001.57236378
Figure L.9: Number of flange leak incidents reported by rating (psig)

6526.71

Page 209
8702.28
10002.62108976
21755.7
 600.157236378

SPIRA

WOU
ND
L
3601.003464
180.047127402
957.2508
1450.38
145.038
RING TYPE (RTJ)

290.076
1319.8458
3600.858426
15003.9345354
2175.57
900.235927086
1440.22734
4351.14
N) OTHER

5001.31054488
1440.22734
HAM
MER

(CHI
CSA
UNI
ON

2175.57
700.183418268
7501.96581732
3016.7904
2088.5472
6001.57236378
2132.0586
COMPRESSED

14.5038
3161.8284
3596.9424
1595.418
275.282124
2500.65527244
4989.3072
2160.631086
6004.5732
580.152
CLAMP (e.g.,

275.072108976
GRAYLOC)

3176.3322
2146.5624
1500.39345354
275.137086
0 5 10 15 20 25

Figure L.10: Number of flange ratings by rating (psig) and type

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 210
 3. Actuated Valve

180
160
140
120
100
80
60
40
20
0
BLOCK

BLOWDOWN

CHOKE

CONTROL

ESDV

PLESDV

PLSSIV ASSEMBLY

RELIEF
Actuated valve type

Figure L.11: Number of actuated valve leak incidents reported by type

250

200

150

100

50

0
0D3 3D11 11D+

Actutated valve size class

Figure L.12: Number of actuated valve leak incidents reported by size category (D≤3”, 3”<D≤11”,
D>11”)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 211
90
80
70
60
50
40
30
20
10
0

Actuated valve size (inches)

Figure L.13: Number of actuated valve leak incidents reported by size (inches)

120

100

80

60

40

20

0
0.03
0.11
0.151
0.1986845
0.26
0.3
0.379
0.476
0.585
0.7
0.804
0.89
1.21
1.332018
1.5875
1.707
1.863769
2.2
2.9
3.17
3.31
3.6
4.489769

5.451
6.245
7.1
8.4
11.817
19.05
26.7
41.91
97.324
5

999

Actuated valve equivalent hole diameter (mm)                    ( note: 999 represents >100 mm)

Figure L.14: Number of actuated valve holes by equivalent hole diameter (mm)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 212
 4. Manual Valve

100
90
80
70
60
50
40
30
20
10
0
NOT KNOWN

BALL

GATE

NOT KNOWN

BALL

GATE

NOT KNOWN

BALL
GATE

NOT KNOWN
BUTTERFLY

BUTTERFLY
OTHER
PLUG

OTHER

PLUG
NEEDLE

GLOBE

BLANK

NEEDLE

GLOBE

OTHER
PLUG
NEEDLE
CHOKE CHECK BLEED BLOCK

Manual valve type

Figure L.15: Number of manual valve leak incidents reported by type

160

140

120

100

80

60

40

20

0
0D3 3D11 11D+
Manual valve size class

Figure L.16: Number of manual valve leak incidents reported by size category (D≤3”, 3”<D≤11”,
D>11”)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 213
 60

50

40

30

20

10

Manual valve size (inches)

Figure L.17: Number of manual valve leak incidents reported by size (inches)

80

70

60

50

40

30

20

10

0
0.1
0.149
0.25
0.32
0.38
0.51
0.555
0.663
0.73
0.86
1
1.1938
1.35
1.681
1.954
2.31
2.7
3
3.2
3.6
3.9
4.2
4.98
6.2
6.6
7.455
8.333
10
10.574
12.7
19
21.333
26.5
30
62
203.2

Manual valve equivalent hole diameter (mm)

Figure L.18: Number of manual valve holes by equivalent hole diameter (mm)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 214
 5. Heat Exchanger

60

50

40

30

20

10

0
HC in Shell HC in Tube Plate
Heat exchanger type

Figure L.19: Number of heat exchanger leak incidents by type

16

14

12

10

0
0.5
1.329
0.1
1.3
1
0.2
0.15
50
2.81
2
1.1
0.64
605
10.6
0.04
1.597
0.286
15.194
0.3881452
0.1690388
0.4631117
4.356201
2.349697
1.81552
0.7517739
0.1330848
4.7752
999
25
6.35
5
10
4
9.5
3
6.7

Heat exchanger equivalent hole diameter (mm)

Figure L.20: Number of heat exchanger leak incidents by hole diameter (mm)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 215
 6. Fin Fan Coolers

0
Fin Fan Cooler

Figure L.21: Number of fin fan cooler leak incidents

2.5

1.5

0.5

0
0.1 0.5 0.645 1 2.6
Fin Fan cooler equivalent hole diameter (mm)

Figure L.22: Number of fin fan cooler leak incidents by hole diameter (mm)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 216
 7. Pressure (Process) Vessels

60

50

40

30

20

10

0
K.O.DRUM OTHER REBOILER SCRUBBER SEPARATOR
Pressure (process) vessel type

Figure L.23: Number of pressure (process) vessel leak incidents by type

12

10

0
0.267
0.453
0.564
0.75
1.2
1.37
1.91
2.108
2.5
2.9
4.92005
5.15
6.7
8.08177
10.4
11.2
11.7647
12.6
13
24.3
25.4
29.7
50.8
76.2

Pressure (process) vessel equivalent hole diameter (mm)

Figure L.24: Number of pressure (process) vessel leak incidents by hole diameter (mm)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 217
Comment re: HCRD Equipment Counts

GTI concurs with the following statement made by DNV on p. 5 of “DNV FFG HCRD ‘13” that
there are uncertainties associated with the equipment count in the HCRD (which of course
directly impacts the calculation of failure frequencies from the HCRD database):

Determining the number of leaks that have occurred off-shore provides only one part of
the data that is required to calculate leak frequency. The number of different types of
equipment offshore has also been recorded and quantified since 1992, although HSE has
recorded no change in the equipment count since 2003 (Regarding system and equip-
ment population data, HSE notes that the responsibility for maintaining the currency of
this data rests with duty holders. The population data in HCRD is provided by duty
holders on a voluntary database and it is not HSE’s role to update, or verify this particular
data. Use of this population data would need to be made with caution). It is questionable
that the amount of equipment has remained the same offshore since 2003. Maintaining an
accurate equipment count is not straightforward, for example the count of equipment on
mobile rigs would require the database operators to keep track of the position of MOUs
and their movements. The equipment count on the UKCS is provided by the operators on
a voluntary database and it is not part of HSE’s role to monitor or verify the equipment
count. Therefore there are uncertainties associated with the equipment count.

As one example, the population of steel piping in the HCRD database is illustrated below:

Population of Steel Piping (m) in HCRD Database
600,000.0

500,000.0

400,000.0

300,000.0

200,000.0

100,000.0

0.0

PIPING_STEEL_3D11 PIPING_STEEL_D11 PIPING_STEEL_D3

Figure L.25 Population of Steel Piping in HCRD Database

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 218
In comparson, total UK petroleum and other liquids production has declined approximately 60%
since 2000, according to a 2016 analysis by the U.S. Energy Information Administration:69

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 219
Appendix M: Cumulative Leak Frequency Curves Developed from UK HSE
Hydrocarbon Releases Database (HCRD)

GTI developed cumulative failure frequency curves for flanges, heat exchangers, piping-steel,
pressure (process) vessels, fin fan coolers, actuated valves, and manual valves in the UK HCRD
database in order to help enhance PHMSA’s and FERC’s understanding of this database. GTI
used the data that is publicly available in Microsoft Excel® files entitled “Offshore Hydrocarbon
Population Data 1992 - 2015” and “Offshore Hydrocarbon Releases 1992 – 2015” which are
available as of the date of this report at http://www.hse.gov.uk/offshore/statistics.htm.

GTI analyzed the raw data in this public database and calculated cumulative failure rate
distribution curves by considering the total leak incidents and the total equipment populations
over the each of the years from 1992/1993 to 2015/2016. The results are illustrated in the graphs
below where x-axis represents a hole size diameter and y-axis represents the cumulative failure
rate of a leak hole that is equal to or greater than that diameter. The cumulative leak frequencies
provided in Appendix M were calculated directly from all of the raw data, and do not incorporate
any differentiation of type of leak (e.g. “Full Pressure” versus “Zero Pressure” leaks).

The HCRD uses the term “Pressure Vessel” for equipment that is in service such as a: knock-out
drum, reboiler, separator, or other. Because these uses are primarily process-related we use the
terminology “Pressure (Process) Vessel” in this Appendix for this class of equipment in the
HCRD.

There are 2401 leak incidents involving these equipment categories from 1992/1993 to
2015/2016, out of which 52 incidents (i.e., 2.2%) do not provide hole size diameter data. In
order to determine total failure frequencies more accurately, we included these 52 incidents and
applied them uniformly across the failure rate distribution generated from the incidents in their
respective equipment categories that did have hole size data. As shown below, the Pressure
(Process) Vessels category was by far the equipment category most impacted, with with 22 of 99
entries (22.2%) lacking hole size data.

Equipment Number of incidents with no hole size

Flanges 0 out of 430
Heat exchangers 2 out of 107
Piping - steel 22 out of 1012
Pressure (Process) vessels 22 out of 99
Fin Fan coolers 0 out of 6
Valve actuated 4 out of 438
Valve manual 2 out of 309

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 220
 Figure M.1: Cumulative Failure Rate of Flanges by size category (D≤3”, 3”<D≤11”, D>11”)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 221
 Figures M.2: Cumulative Failure Rate of Piping – Steel by size category (D≤3”, 3”<D≤11”, D>11”)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 222
 Figures M.3: Cumulative Failure Rate of Actuated Valves by size category (D≤3”, 3”<D≤11”,
D>11”)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 223
Figures M.4: Cumulative Failure Rate of Manual Valves by size category (D≤3”, 3”<D≤11”, D>11”)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 224
Figures M.5: Cumulative Failure Rate of Heat Exchangers by Heat Exchanger Type (excludes Fin-
Fan)

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 225
 Figure M.6: Cumulative Failure Rate of Pressure (Process) Vessels

Figure M.7: Cumulative Failure Rate of Fin Fan Coolers

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 226
Appendix N: Subject Matter Experts Contacted in Addition to Technical Advisory
Panel

The purpose of Task 4 of the project was to contact Subject Matter Experts (SMEs) and other
stakeholders beyond the project’s Technical Advisory Panel (TAP), in order to ensure that no
significant existing failure rate data sources were overlooked and to gather additional key input
into this PHMSA-funded project. The project targeted to contact a representative sampling of at
least 15 entities beyond those in the TAP, such as additional pipeline companies and operators of
LNG facilities, operators of related process facilities (e.g. industrial gas, chemical processing,
etc.), standard-developing organizations, and perhaps a representative major oil and gas
company. The goal was to seek their voluntary input and insights into this project, our analysis,
and failure rate data sources (and inquire if they have proprietary data that they may publically
share under this or subsequent research).

For this task the project team leveraged its contacts in the LNG industry through its long history
in LNG. In addition, the project team leveraged major events such as: the LNG18 conference in
Perth, Australia during April 2016 (this tri-annual conference is one of the largest LNG
conferences held in the world, with approximately 6,000 attendees); meetings of the American
Gas Association’s Supplemental Gas Technical Committee during 2016 (this committee includes
LNG plant peak shaving plant management personnel, LNG export facility personnel, LNG-
focused engineering firms, LNG equipment manufacturers, and others); and Natural Gas for
High Horsepower (HHP) Summits and Expositions. By leveraging these LNG-related events in
an opportunistic manner, the project team was able to personally discuss LNG equipment failure
rate data issues and PHMSA’s project with a diverse number of individuals at key companies.

More than 50 SMEs were contacted. They represented global perspective and expertise as
operators of LNG plants and related process facilities, pipeline companies, LNG equipment
manufacturers, LNG facility and storage tank designers and builders, and others.
All of the following SMEs were contacted by email, and most also in person:

1. AGL Resources, Managing Director Storage & Peaking Services

2. Air Liquide Global E&C Solutions, LNG Product Manager and also LNG Technology
Lead
3. Air Products and Chemicals Inc. (APCI), General Manager Global LNG &
CyroMachinery Division, and also Manager LNG Mechanical Design Engineering
4. Applied Cryo Technologies, Chief Operating Officer
5. Baltimore Gas & Electric Co., Senior Engineering Technical Specialist
6. Bechtel, Engineering Manager, Tank Business Line
7. Bestobell Valves, div. of Parker Hannifin, Product Manager and also Business
Development Manager
8. Black & Veatch Corp., Vice President - LNG Technology, B&V Energy
9. Braemar Engineering, Vice President LNG Technical Services
10. British Compressed Gas Association, Technical Manager
11. CB&I, Senior Vice President of LNG
12. Chart Industries, Vice President of Sales, LNG and also Vice President of Engineering
13. Cheniere LNG O&M Services, LLC, Vice President, Government and Regulatory

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 227
 Affairs, and also Principal Process Engineer
14. Chevron Energy Technology Pty Ltd., Perth Global Technology Centre Manager
15. Compressed Gas Association, Technical Manager
16. ConocoPhillips, Licensing Director, LNG Technology and Licensing
17. Cryosys, President
18. Dominion Cove Point, Vice President - LNG Operations, and also Director -
Liquefaction
19. EcoElectrica L.P., Plant Manager
20. EMCO Wheaton by Gardner Denver, Engineering Manager
21. Elengy, a company of GDF Suez, Head of LNG Technical Department
22. Evergas, Senior LNG Operations Manager
23. European Industrial Gas Association, Deputy General Secretary
24. Excelerate Energy LP, Senior Vice President, Operations
25. ExxonMobil Development Company, Risk and Loss Prevention Engineer
26. Fluor, Technical Director, Process Engineering and Senior Fellow, LNG & Gas
Processing
27. FMC Technologies, Business Development Manager, Loading Systems
28. Freeport LNG Development, LP, FERC Compliance Manager - Liquefaction
29. GTT, Structure Engineer and also Business Development Manager
30. Hitachi High-Tech AW Cryo, Inc., President
31. INOXCVA, President and CEO
32. International Group of Liquefied Natural Gas Importers (GIIGNL), Technical Study
Group Lead
33. Jereh Group, Deputy Director
34. Kawasaki Heavy Industries, Manager Overseas Sales Section, Cryogenic Storage
System Sales Department
35. Kellogg Brown & Root (KBR), Director LNG and FLNG Engineering and Construction
36. Korea Gas Research Institute, Head
37. Linde AG Engineering Division, Executive Vice President, LNG and Natural Gas
Plants
38. Matrix Service, Senior Director, Business Development
39. Perma-Pipe Inc., Division Vice President, Oil & Gas
40. Petronas, Head - Performance, Portfolio & Governance, Technology, Technical Global
41. Petronet LNG Limited, General Manager – Maintenance
42. PHPK Technologies, Director of Sales and Marketing
43. RegO Products, Director of Sales
44. Shell Global Solutions International BV, Global Manager - LNG Market Access and
also Distribution Engineering Manager - LNG Market Access
45. Society of International Gas Tanker and Terminal Operators Ltd (SIGGTO),
General Manager
46. Stabilis Energy, Vice President, Plant Construction/Business Development
47. Technodyne International Ltd., Founder and past Technical Director
48. Tokyo Boeki Machinery Ltd. (Niigata Loading Systems, Ltd.), Subsection Chief,
Overseas Plant Equipment Group
49. Toyo Kanetsu K.K. (TKK), Sales Department
50. U.S. Coast Guard, Liquefied Gas Carrier National Center of Excellence, Detachment
Chief
51. Valco Group, Director Australia
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 228
 52. Wartsila North America, Gas Initiatives Vice President
53. Woodside Energy Ltd., VP Technology

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 229
Appendix O: Acronyms

Acronyms often used in this report and in the referenced literature are listed below. See also Section 3 for
acronyms of references as used in this project.

Acronym Description (using English translations)

AGA American Gas Association
AGS Dutch Hazardous Substance Council
AHJ Authority Having Jurisdiction
AIChE American Institute of Chemical Engineers
API American Petroleum Institute
ASME American Society of Mechanical Engineers
BEVI Dutch External Safety Decree for Establishments
BCF Billion Cubic Feet of Natural Gas
BCGA British Compressed Gas Association
BAC Break-Away Coupling
BRZO Dutch Hazards of Major Accidents (risks) Decree
CCPS AIChE Center for Chemical Process Safety
CEN European Norm (i.e., European Community standard)
CFR U.S. Code of Federal Regulations
CGA Compressed Gas Association
CH-IV CH-IV International Inc.
COMAH Control of Major Accident Hazards
COTP USCG Captain of the Port
CPI Chemical Process Industry
CSA CSA Group (formerly Canadian Standards Association)
DOE U.S. Department of Energy
DOT U.S. Department of Transportation
DNV Det Norske Veritas
EGIG European Gas pipeline Incident data Group
EIGA European Industrial Gas Association
EPA U.S. Environmental Protection Agency
EPRI Electric Power Research Institute
ERC Emergency Release Coupling
ERP Emergency Response Plan
ERS Emergency Release System
ESD Emergency Shutdown system
FERC U.S. Federal Energy Regulatory Commission
F-N Frequency-Number of Fatalities
FRED Failure Rate and Event Data
FRT Failure Rate Table
FSRU Floating Storage and Regasification Unit

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 230
 FSU Floating Storage Unit
FTA Fault Tree Analysis
GIIGNL International Group of LNG Importers Ltd.
GRI Gas Research Institute
GTI Gas Technology Institute (formerly GRI and Institute of Gas Technology)
HAZID Hazard Identification analysis
HAZOP Hazard and Operability analysis
HCRD UK Hydrocarbon Releases Database System
HSE UK Health and Safety Executive
HSL UK Health and Safety Laboratory (of HSE Science Division)
IMDG Code International Maritime Dangerous Goods Code
IMO International Maritime Organization
INL Idaho National Laboratory
IOGP International Association of Oil & Gas Producers (formerly OGP)
IR Individual Risk
ISO International Organization for Standardization
LNE Flemish Environment, Nature and Energy Department
LNG Liquefied Natural Gas
LNGC LNG carrier
LPG Liquefied Petroleum Gas
MARAD U.S. DOT Maritime Administration
NB Nominal Bore
NFPA National Fire Protection Association
NG Natural Gas
NVIC USCG Navigation and Vessel Inspection Circular
OCIMF Oil Company International Marine Forum
OREDA Offshore and Onshore Reliability Data project
PERC Powered Emergency Release Coupling
PGS Dutch Publication Series on Dangerous Substances
PHMSA U.S. DOT Pipeline and Hazardous Materials Safety Administration
PNL Pacific Northwest Laboratory
PS Peak Shaving LNG Plant
QRA Quantitative Risk Assessment
RAM Reliability Availability Maintenance analysis
RIVM Dutch National Institute of Public Health and the Environment
SALS Single Accidental Leakage Source
SIGTTO Society of International Gas Tanker and Terminal Operators
SINTEF Norwegian Foundation for Scientific and Industrial Research
SME Subject Matter Expert
SR Societal Risk
TNO Dutch Organization for Applied Scientific Research
USCG U.S. Coast Guard
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 231
Appendix P: Bibliography

The following references were evaluated for relevancy to this research during project DTPH56-
15-T-00008. The abbreviated name that is [included in brackets] is used elsewhere in this project,
including in the project’s database.

Additional Failure Rate Quantitative Data References Evaluated for Relevancy to

Components in PHMSA’s LNG Failure Rate Table:

1. API, Risked-Based Inspection, Recommended Practice 580, Third Edition, American

Petroleum Institute, 2016.
2. Blanchard, A. and Roy, B.N., Washington Savannah River Company, Savannah River
Site Generic Data Base Development, DOE Contract No. DE-AC09-89SR18035, WSRC-
TR-93-262, May 1998 [“SAV ‘98”]
3. Cadwallader, Lee Charles. Selected component failure rate values from fusion safety
assessment tasks. Idaho National Engineering and Environmental Laboratory, USA,
67p (1998) [“INO FUSION ‘98”]
4. Fecht, B.A. Gates, T.E., Nelson, K.O., Marr, G.D, Comparative Safety Analysis of LNG
Storage Tanks, Pacific Northwest Laboratory, PNL-4399, 1982. [“PNL CSA ‘82”]
5. EIReDA, Procaccia, H., Arsenis, S.P., Aufort, P., European Industry Reliability Data
Bank EIReDA 1998. (1998). [“EIReDA ‘98”]
6. GIIGNL’s representatives Anthony Acton, Deborah Brown and Pieere Langry, “LNG
Incident Identification – Updated Compilation and Analysis by the International Group of
LNG Importers (GIIGNL)”, LNG17 conference paper and presentation, April 2013.
7. Goodfellow, Graham, Turner, S., Haswell, J., Espiner, R., “An Update to the UKOPA
Pipeline Damage Distributions 2012”, 9th International Pipeline Conference. American
Society of Mechanical Engineers, 2012. [“UKOPA ‘12”]
8. HSE Failure Rate and Event Data for use within Land Use Planning Risk Assessment,
undated but estimated as issued April 2012. [“HSE FRED APR‘12”]
9. HSE Failure Rates for Underground Gas Storage, 2008 [“HSE UG ‘08”]
10. HSE Fire, Explosion and Risk Assessment Topic Guidance, Issue 1 2003 [“HSE
GUIDANCE ‘03”]
11. HSE Hill, RT, Catmur, J.R., Risks from Hazardous Pipelines in the UK, Contract
Research Report No. 82, 1994 [“HSE PIPELINES ‘94”]
12. HSE Societal Risk: Initial Briefing to Societal Risk Technical Advisory Group, Research
Report RR703, 2009
13. HSE The Major Accident Failure Rates Project, Report RR915l, 2012 [“HSE MAJ
ACCID ‘12”]
14. HSE Update of pipeline failure rates for land use planning assessments, Research Report
RR9035, 2015

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 232
 15. IOGP Risk Assessment Data Directory, Riser and Pipeline Release Frequencies,
International Association of Oil and Gas Producers Report No. 434 – 4, March 2010.
[“IOGP 434-4”]
16. IOGP Risk Assessment Data Directory, Guide to Finding and Using Reliability Data for
QRA, International Association of Oil and Gas Producers Report No. 434 – 20.1, March
2010. [“IOGP 434-20.1”]
17. IOGP Risk Assessment Data Directory, Appendix 1, Risk Assessment Data Directory,
International Association of Oil and Gas Producers Report No. 434 – A1, March 2010.
[“IOGP 434-A1”]
18. ISO Guidance on performing risk assessment in the design of onshore LNG installations
including the ship/shore interface, ISO/TS 16901:2015
19. Ketchell, N., Robinson, R., AEA Technology and Genoud, P., S.N. Technigaz,
“Quantification and Comparison of the Risks of LNG Storage Concepts – Membrane and
Full Containment”, LNG12 Conference proceedings, Perth, 1996. [“AEA TG ‘06”]
20. Lee, Seung Rim, and Han Sang Kim. “The comparative risk assessment of LNG tank
designs using FTA”, Journal of the Korean Institute of Gas 16.6 (2012): 48-54 [“KIGAS
‘12”].
21. NFPA Standard for the Production, Storage, and Handling of Liquefied Natural Gas
(LNG), NFPA 59A, 2001 Edition. [“59A ‘01”]
22. NFPA Standard for the Production, Storage, and Handling of Liquefied Natural Gas
(LNG), NFPA 59A, 2009 Edition. [“59A ‘09”]
23. NFPA Vehicular Gaseous Fuel Systems Code, NFPA 52, 2016 Edition.
24. Pelto, P.J., Baker, E.G., Holter, G.M., and Powers, T.B., An Overview Study of LNG
Release Prevention Systems, Pacific Northwest Laboratory, PNL-4104, March 1982.
[“PNL RP ‘82”]
25. Rath, S and Krol, M, “Comparative Risk Assessment for Different LNG-Storage Tank
Concepts”, Chemical Engineering Transactions, published by Italian Association of
Chemical Engineering Vol 31, 2013.
26. SAI, LNG Terminal Risk Assessment Study for Los Angeles, California, Science
Applications, Inc., SAI-75-614-LJ, 1975.
27. SAI, LNG Terminal Risk Assessment Study for Point Conception, California, Science
Applications, Inc., SAI-75-616-LJ, 1976.

Additional References:

28. ABS Consulting, Bunkering of Liquefied Natural Gas-fueled Marine Vessels in North
America, Houston, 2nd Edition, March 2015.
29. AGS Risk calculations by prescription: rituals for granting permits and land-use planning,
Hazardous Substances Council, The Netherlands, The Hague, 2010.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 233
 30. Al-Kinari, A, Cakir, Nihal, Baumgartner, T., Stundner, M., “Adaptive Advisory Systems
for Oil and Gas Operations” in Integrated Operations in the Oil and Gas Industry:
Sustainability and Capability Development, 262-284 (2013).
31. Akhmedjanov, Farit. Reliability databases: State-of-the-art and perspectives. (Denmark.
Forskningscenter Risoe. Risoe-R; No. 1235(EN) 2001.
32. Atallah, S., J. N. Shah, and M. Betti. Reduction of LNG operator error and equipment
failure rates. Topical report, 20 April 1990. No. PB-90-228131/XAB. Risk and Industrial
Safety Consultants, Inc., Des Plaines, IL (USA), 1990.
33. Ayello, Francois, Alfano, T., Hill, D., Sridhar, N., DNV, “A Bayesian network based
pipeline risk management” CORROSION 2012. NACE International, 2012.
34. Barle, Jani, and Dario Ban. “Maritime component reliability assessment and maintenance
using Bayesian framework and generic data” Advanced Ship Design for Pollution
Prevention. Taylor & Francis Group, 2010.
35. Barua, S., Gao, X., Mannan, S., “Dynamic Operational Risk Assessment with Bayesian
Networks”, Proceedings of 8th Global Congress on Process Safety, Houston, Texas,
April 1-4, 2012.
36. Barua, S., Dynamic Operational Risk Assessment with Bayesian Network, MS Thesis,
Texas A&M University, August 2012.
37. Barua, Shubharthi, Gao, Xiaodan, Pasman, Hans, and Mannan, M. Sam, “Bayesian
network based dynamic operational risk assessment”, Journal of Loss Prevention in the
Process Industries, Available online 30 November 2015.
38. Bearfield, G.J., Using Bayesian networks to represent parameterised risk models for the
UK railways. 2009, Queen Mary, University of London.
39. Berg, H. P., R. Gersinska, and J. Sievers, Improved approach for estimating leak and
break frequencies of piping systems in probabilistic safety assessment. Electronic Journal
Reliability & Risk Analysis: Theory & Applications 4 (2009): 15.
40. Bolt, R., Hilgenstock, A., Kolovich, C., Velez Vega., D., Cappaner, A., Rasmussen, O. A
Guideline: Using or Creating Incident Databases for Natural Gas Transmission
Pipelines 2006 International Pipeline Conference. American Society of Mechanical
Engineers, 2006.
41. Bolsover, Andy, Falck, A., Pitblado, R., DNV. “A Public Leak Frequency Dataset for
Upstream and Downstream Quantitative Risk Assessment”, AIChE Spring National
Meeting, (2013).
42. Bragatto, P. Ansaldi, S., Delle Site, C., “A Pooled Knowledge Basis on Pressure
Equipment Failures to Improve Risk Managementin Italy” Chemical Engineering
Transactions 33, 433-438 (2013).
43. British Standard, Installation and equipment for liquefied natural gas. Design of onshore
installations, BS EN 1473:2016.
44. Bureau Veritas, Guidelines on LNG Bunkering, Guidance Note NI 618 DT R00E, July
2014.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 234
 45. Caldwallader, L.C., Selected Component Failure Rate Values from Fusion Safety
Assessment Tasks, Idaho National Engineering and Environmental Laboratory,
INEEL/EXT-98-00892, Sept. 1998.
46. Center for Energy Economics, The University of Texas at Austin, LNG Safety and
Security, June 2012.
47. Chang, James I., and Cheng-Chung Lin. “A study of storage tank accidents” Journal of
loss prevention in the process industries 19.1 (2006): 51-59.
48. ClassNK, Guidelines for Gas Fueled Ships, Version 4, April 2016.
49. Cowell, R.G., Probabilistic networks and expert systems: exact computational methods
for Bayesian networks. Information Science & Statistics. 2007, New York, NY: Springer.
xii 321.
50. Cox, Andrew William, Studies in Support of a Quantitative Approach to Hazardous Area
Classification, Doctoral Thesis, Loughborough University of Technology, 1989.
51. CSA Liquefied natural gas (LNG) – Production, storage, and handling, CAN/CSA-
Z276-15, approved January 2016.
52. de Andreade Melani, Arthur Henrique, et al, Univ. of Sao Paulo, Use of Bayesian
Network to Support Risk-Based Analysis of LNG Carrier Loading Operation.
Probabilistic Safety Assessment and Management Conf. 12, June 2014.
53. de Andrade Melani, Arthur Henrique, Dennis Wilfredo Roldán Silva, and Gilberto
Francisco Martha Souza. “Use of Bayesian Network to Support Risk-Based Analysis of
LNG Carrier Loading Operation”, 12th Probabilistic Safety Assessment and Management
Conference. June 22- 27, 2014.
54. Dawotola, A. W, Risk Based Maintenance of Petroleum Pipelines. Diss. TU Delft, Delft
University of Technology, 2012.
55. DNV GL, Recommended Practice: Development and Operation of Liquefied Natural Gas
Bunkering Facilities, DNVGL-RP-G105, Oct. 2015.
56. DNV GL, Follow-up Recommendations HAZID LNG Safety Program, Report PP132344-
1, April 2016.
57. EIGA Determination of Safety Distances, AISPL, European Industrial Gases Assocation
IGC Doc 75.07.E, 2007.
58. EIGA Safe Practices Guide for Cryogenic Air Separation Plants, European Industrial
Gases Assocation IGC Doc 147/13/E, 2007.
59. Falck, Andreas, DNV, “Leak Frequency Modeling for Offshore QRA based on the
Hydrocarbon Released Database”, Institution of Chemical Engineers, Hazards XXI,
2009.
60. Fenton, N. and M. Neil, Risk assessment and decision analysis with Bayesian networks.
2012: CRC Press.
61. Fitzgerald, G.A., “The Maximum Design Leak (MDL) Approach to Leak Size Selection”,
2011 Mary Kay O’Connor Process Safety Center International Symposium, October 25,
2011.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 235
 62. Fitzgerald, Gary Allen. “Calculating Facility Siting Study Leak Sizes - Applications of the
Maximum Design Leak (MDL) Approach”, 11th Global Congress on Process Safety,
Texas, Austin, April 26-20, 2015.
63. GeunWoong Yun, William J. Rogers, M. Sam Mannan, “Risk Assessment of LNG
Importation Terminals using the Bayesian–LOPA Methodology”, Journal of Loss
Prevention in the Process Industries, vol. 22, no.1, pp. 91–96, January 2009.
64. Guidelines for Developing Quantitative Safety Risk Criteria, American Institute of
Chemical Engineers, Center for Chemical Process Safety, New York, NY, August 2009.
65. Guidelines for Improving Plant Reliability through Data Collection and Analysis,
American Institute of Chemical Engineers, New York, NY, 1998.
66. Hölzl, Reinhold. “Lifetime estimation of aluminum plate fin heat exchangers”, ASME
2012 Pressure Vessels and Piping Conference. American Society of Mechanical
Engineers, 2012.
67. IIASA, Mandl, C. and Lathrop, J.W., Liquefied Energy Gas Terminal Risk: A
Comparison and Evaluation, International Institute for Applied Systems Analysis, IIASA
Research Report, Dec. 1983.
68. IOGP Guidance on Performing Risk Assessment in the Design of Onshore LNG
Installations Including the Ship/Shore Interface. International Association of Oil and Gas
Producers Draft 116901, 2013.
69. IOGP Guidance for Systems and Installations for Supply of LNG as Fuel to Ships.
International Association of Oil and Gas Producers Draft 118683, 2013.
70. IOGP Risk Assessment Data Directory, Human factors in QRA, International Association
of Oil and Gas Producers Report No. 434 – 5, March 2010.
71. IOGP Risk Assessment Data Directory, Ignition Probabilities, International Association
of Oil and Gas Producers Report No. 434 – 6.1, March 2010.
72. IOGP Risk Assessment Data Directory, Land Transport Accident Statistics, International
Association of Oil and Gas Producers Report No. 434 – 9, March 2010.
73. IOGP Risk Assessment Data Directory, Mechanical Lifting Failures, International
Association of Oil and Gas Producers Report No. 434 – 8, March 2010.
74. IOGP Risk Assessment Data Directory, Water Transport Accident Statistics, International
Association of Oil and Gas Producers Report No. 434 – 10, March 2010.
75. Jensen, F.V. and T.D. Nielsen, Bayesian networks and decision graphs. 2nd ed.
Information science and statistics. 2007, New York: Springer. Xvi.
76. Jain, S., Ayello, F., Beavers, J.A., Sridhar, N. DNV, “Development of a probabilistic
model for stress corrosion cracking of underground pipelines using Bayesian networks: a
concept”, 2012 9th International Pipeline Conference. American Society of Mechanical
Engineers, 2012.
77. Jain, S., Ayello, F., Beavers, J.A., Sridhar, N., DNV, “Probabilistic model for stress
corrosion cracking of underground pipelines using Bayesian networks”, Proceedings of
the NACE International Corrosion. Conference Series, Corrosion, 2013.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 236
 78. Kleinhammer, R. K., and J. C. Kahn. “Constructing the Best Reliability Data for the Job”.
8th Annual Mars Event - Maintenance and Reliability Symposium, Galveston, TX, 13-15
Aug. 2014.
79. Khan, Faisal, Development of a Tool for Risk Based Integrity (RBI) Assessment of
Process Component, Final Report C9-07, University of Newfoundland.
80. Kjærulff, U.B. and A.L. Madsen, Bayesian networks and influence diagrams: a guide to
construction and analysis. 2nd ed. Information science and statistics, 2013, New York,
NY: Springer. xvii, 382
81. Kohout, Andrew Joseph. Evaluation of Fire Dynamics Simulator for Liquefied Natural
Gas Vapor Dispersion Hazards. Master of Science Thesis, University of Maryland at
College Park, 2011
82. Lee, Seung Rim, Jang, SP and Jeong, RK. “Development of technologies and standards
for the life time management of LNG storage tanks” 2010 AIChE Spring Meeting & 6th
Global Congress on Process Safety.
83. Lloyd’s Register, Guidance Notes for Risk Based Analysis: Cryogenic Spill, August
2015.
84. Martins, Marcelo, Schleder, A.M., Droguett, E.L., “A Methodology for Risk Analysis
Based on Hybrid Bayesian Networks: Application to the Regasification System of
Liquefied Natural Gas Onboard a Floating Storage and Regasification Unit”, Risk
Analysis, Vol. 34, No. 12, 2014.
85. Miranda, Marcelo Accorsi, and Orlando Guerreiro Meira. Lifecycle Assessment of
Turbomachinery for Offshore Applications – Updated with Field Data. Proceeding of the
37th Turbomachinery Symposium. 2008.
86. PGS Natural gas: liquefied natural gas (LNG) delivery installations for vehicles,
Publicatiereeks Gevaarlijke Stoffen 33-1, 2013.
87. PGS Natural gas: liquefied natural gas (LNG) delivery installations for ships,
Publicatiereeks Gevaarlijke Stoffen 33-2, 2014.
88. Richardson, S. A. Interpretation of the HCR for QRA-and its application beyond the
North Sea Institution of Chemical Engineers Symposium Series. Vol. 154. Institution of
Chemical Engineers; 1999, 2008.
89. Rosendahl, Tom Hepsø, Vidar. Integrated Operations in the Oil and Gas Industry -
Sustainability and Capability Development: Chapter 16 Adaptive Advisory Systems for
Oil and Gas Operations. IGI Global, 2013.
90. Shafaghi, Ahmad, ABS Consulting, “Equipment failure rate updating – Bayesian
estimation”, Journal of Hazardous Materials, 159 (2008) 87-91.
91. Skangass, AS, by DNV, Appendix D – Leak Frequencies Analysis, Report no./DNV Reg
No.: 2013-4091 / 17TLT29-3 REV 1, 11.06.2013
92. Spouge, John, DNV, “Leak frequencies from the hydrocarbon release database,”
Institution of Chemical Engineers, Symposium Series. Vol. 151, 2006.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 237
 93. Stahl, M.W., Kenady, J.K., ABS, “A Frequency Based Approach to Hole Size Selection
for Consequence Analysis”, American Institute of Chemical Engineers, 7th Global
Congress on Process Safety, Illinois, 2011.
94. Techniques for Assessing Industrial Hazards: A Manual, Technica, Ltd. World Bank
Technical Paper Number 55 ©1988 The World Bank.
95. Taveau, Jerome, “Risk assessment and land-use planning regulations in France following
the AZF disaster”, Institute for Radioprotection and Nuclear Safety, Journal of Loss
Prevention in the Process Industries, Elsevier Ltd., 2010, 813-823.
96. TNO, Methods for the Calculation of Physical Effects (TNO Yellow Book), Committee
for the Prevention of Disasters (CPR), National Institute of Public Health and the
Environment (R1VM), The Netherlands Organization for Applied Scientific Research
(TNO). Third edition. Second revised print. 2005. [“TNO YELLOW ‘05”]
97. U.S. Coast Guard, Guidance Related to Vessels and Waterfront Facilities Conducting
Liquefied Natural Gas (LNG) Marine Fuel Transfer (Bunkering) Operations, CG-OES
Policy Letter No. 02-15, Feb. 2015.
98. U.S. Coast Guard, Guidance for Liquefied Natural Gas Fuel Transfer Operations and
Training of Personnel on Vessels Using Natural Gas as Fuel, CG-OES Policy Letter No.
01-15, Feb. 2015.
99. U.S. Navy, Naval Surface Warfare Center, Carderock Division, Handbook of reliability
prediction procedures for mechanical equipment, 1998.
100. Wierman, T, Rasmuson, F.M., Mosleh, A., Common-cause failure database and
analysis system: event data collection, classification, and coding. Division of Risk
Assessment and Special Projects, Office of Nuclear Regulatory Research, US Nuclear
Regulatory Commission, 2007.
101. Wilhelm, S. Mark. “Risk analysis for operation of aluminum heat exchangers
contaminated by mercury,” Process Safety Progress, 28.3, 259-266, 2009.
102. Yue, Meng, and Tsong-Lun Chu. “Estimation of failure rates of digital
components using a hierarchical Bayesian method”. International Conference on
Probabilistic Safety Assessment and Management. New Orleans. 2006.
103. Yun, G.W., Bayesian-LOPA Methodology for Risk Assessment of An LNG
Importation Terminal, Master of Science Thesis, Texas A&M University, August 2007.
104. Yun, G.W., “The Risk Assessment of Bayesian-LOPA Methodology for an LNG
Importation Terminal”, AIChE Spring Meeting, April 2008 8th Topical Conference on
Natural Gas Utilization New Orleans, Louisiana, April 6-10, 2008.
105. Yu, Wei, Popva, E., Kee, E., Sun, A., Equipment data development case study -
Bayesian Weibull analysis., Univ. of Texas at Austin joint research with South Texas
Project Nuclearn Operating Co., July, 2004.
106. Zavadskas, Edmundas Kazimieras, and Egidijus Rytas Vaidogas, “Bayesian
reasoning in managerial decisions on the choice of equipment for the prevention of
industrial accidents”, Engineering Economics 60.5 (2015).

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 238
Appendix Q: References

1
https://primis.phmsa.dot.gov/lng/docs/Failure%20Rate%20Table%201.pdf
2
http://primis.phmsa.dot.gov/lng/faqs.htm
3
PHMSA, Kowalewski, Rick, A Data Quality Assessment Evaluating the Major Safety Data
Programs for Pipeline and Hazardous Materials Safety. November 10, 2009.
4
NTSB, Integrity Management of Gas Transmission Pipelines in High Consequence Areas, US
National Transportation Safety Board Safety Study, NTSB/SS-15/01, Jan. 27, 2015. See e.g.
Safety Recommendation #P-15-13, p. 69.
5
FERC, Kohout, Andrew, “U.S. Regulatory Framework and Guidance for Siting Liquefied
Natural Gas Facilities – A Lifecycle Approach”, Proceedings of Mary Kay O'Connor Process
Safety Center, 15th International Symposium, College Station, Texas, 2012.
6
ABS, Stahl, Michael and Kenady, Kollin, “A Frequency Based Approch to Hole Size Selection
for Consequence Analysis”, AIChE 7th Global Congress on Process Safety, Chicago, IL. March
13-16, 2011.
7
ABS, Fitzgerald, Gary, Stahl, Michael, Campbell, David, Nouri Farzin, and Montgomery,
Randall, “The Maximum Design Leak (MDL) Approach to Leak Size Selection”, Proceedings
of Mary Kay O'Connor Process Safety Center, 14th International Symposium, College Station,
Texas, 2011.
8
FERC, Kohout, Andrew, “U.S. Regulatory Framework and Guidance for Siting Liquefied
Natural Gas Facilities – A Lifecycle Approach”, Proceedings of Mary Kay O'Connor Process
Safety Center, 15th International Symposium, College Station, Texas, 2012
9
Lever, Ernest and Ersoy, Dan, Gas Techology Institute, “Approaches for Preventing
Catastrophic Events”, Final Report, June 15, 2016, US DOT PHMSA contract
DTPH5615T00001L.
10
Flamberg, Stephanie, Rose, Susan, Kiefner/Applus-RTD and Kurth, Bob and Sallaberry,
Cedric, EMC2, “Paper Study on Risk Tolerance”, Final Report No. 16-092, Final Report, June
30, 2016, US DOT PHMSA contract DTPH5615T00002L.
11
FERC, Risk-Informed Decision Making Guideline, Chapter 3 Risk Assessment, Version 4.1,
March 2016, pp. 3-11 and 3-14.
12
DoD, Common Risk Criteria Standards for National Test Ranges, Standard 321-16, August
2016.
13
US Army Corps of Engineers and U.S. Department of the Interior Bureau of Reclamation, Best
Practices in Dam and Levee Safety Risk Analysis, Part IX – Risk Assessment/Management,
Version 4.0, July 2015.
14
Suter, Phil, CH-IV International Inc., “Comparison of Exclusion Zone Calculations and Vapor
Dispersion Modeling Tools”, Final Report, US DOT PHMSA contract DTPH5615T00001L.
15
http://www.phmsa.dot.gov/pipeline/library/data-stats/distribution-transmission-and-gathering-
lng-and-liquid-accident-and-incident-data
16
ibid
17
Singapore Government National Environment Agency, QRA Technical Guidance, Revision
No. 2, 31st August 2016, http://www.nea.gov.sg/docs/default-source/anti-pollution-radiation-
protection/central-building-planning/qra-technical-guidance_final_31aug16.pdf
18
Personal communication of Rich Kooy with Ik Keun Yoon, Functional Safety Engineer,
Process Hazard Analysis Leader, KOGAS, May 20, 2016.

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 239
19
Blanchard, A., and Roy, B.N., Savannah River Site Generic Data Base Development, SWRC-
TR-93-262, Rev.1, pp.3-4.
20
Eide, S.A, Khericha, S.T., Calley, M.B, and Johnson, D.A., Component External Leakage and
Rupture Frequency Estimates, Idaho National Engineering Laboratory, EGG-M-92256, 1992.
21
Helle, Henk, Five fatal flaws in API RP 581, 14th Middle East Corrosion Conference &
Exhibition, Feb. 12-15, 2012.
22
RIVM, de Haag, Paul, “QRAs for Dutch Installation. I. New Dutch Risk Regulations and Risk
Methodology by the New Purple Book” IChemE Symposium Series No. 153, 2007
23
DNV GL. Liqueified Natural Gas (LNG) Bunkering Study for US DOT Maritime
Administration, Washington DC, Report No. PP087423-4, Sept. 2014.
24
DNV, Navigational and Safety Risk Assessment for Washington State Ferries, Appendix 1,
Safety Risk Assessment Study Basis, 09 October 2013, p. 16.
http://www.wsdot.wa.gov/nr/rdonlyres/0d01607b-7b65-4ac5-a86d-
c8c688964536/0/reducedpp061307wsfwsareportappendix1rev01.pdf
25
DNV, Report for Shell E&P Ireland Ltd., Corrib Onshore Pipeline EIS, DNV Reg. No.
12LKQW5-2, Rev. 0, 2010-05-18, p. 47,
http://www.epa.ie/licences/lic_eDMS/090151b2804d1b02.pdf
26
DNV Pembina Propane Export Terminal Facility QRA Report, DNV Report PP124992
DRAFT, Rev. 1, March 16, 2015,
http://media.oregonlive.com/portland_impact/other/Pembina%20QRA%20Draft%20Report.pd
f
27
“BCGA Policy Regarding Internal Examination and Proof Pressure Testing of Static
Cryogenic Liquid Storage Tanks”, British Compressed Gases Association (BCGA) Technical
Information Sheet 23, published 20Sep12.
28
“Periodic Inspection of Static Cryogenic Vessels” European Industrial Gas Association
(EIGA) IGC Doc 119/04/E, ©2004.
29
http://www.safelng.nl/1118-2/ content excerpt as of Oct. 8, 2016. See also
http://www.safelng.nl/wp-content/uploads/2016/04/Test-programme-Heat-load-resistance-of-
cryogenic-storage-tanks.pdf
30
Personal telephone conversation, Rich Kooy of GTI to Tom Drube, Vice President of
Engineering for Chart Industries, Inc. May 12, 2016.
31
Kobayashi, Hideo, “Explosion of Liquefied Nitrogen Storage Tank by Closing Shutoff Valve
for Safety Valve”, Tokyo Institute of Technology, 1992.
32
http://www.engie.com/en/news/everett-lng-terminal-united-states/ “GDF SUEZ’s subsidiary
Distrigas is the largest LNG truck-loading company in the United States. "We deliver between
6,000 and 12,000 trucks of LNG per year to different customers," says Anthony Scaraggi,
Vice-President of Operations at GDF SUEZ Gas North America.” excerpted Oct. 20, 2016.
33
HSL Hose and Coupling Failure Rates and Human Error, 2000 [“HSE HOSE ‘00”].
34
IGU 2016 World LNG Report
35
GIIGNL, Annual Report, 2016 edition
36
FERC Issuance Order Granting Section 3 and 7 Authorizations to Dominion Cove Point LNG,
LP, issued Sept. 29, 2014, p. 7.
37
Baker, E.G, Analysis of LNG Import Terminal Release Prevention Systems, Pacific Northwest
Laboratory, PNL-4152, 1982. [“PNL ITRP ‘82]. Page 56 identifies two plies.
38
Hydrocarbon Releases System Internet Help File accessed via https://www.hse.gov.uk/hcr3/ ;
reference item #33 of “Explanation of Terms used in Data Download” under Part Three:
Equipment Data
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 240
39
Park, Jai Hak, Cho, Young Ki, Kim, Sun Hye and Jin, Ho Le, “Estimation of Leak Rate
Through Circumferential Cracks in Pipes in Nuclear Power Plants”, Nuclear Engineering and
Technology, 47 (2015) 332-339.
40
Lam, Chio, Statistical Analyses of Historical Pipeline Incident Data with Application to the
Risk Assessment of Onshore Natural Gas Transmission Pipelines, Master of Science Thesis,
The University of Western Ontario, July 2015 [“PHMSA NGT LAM ‘15”]
41
As per pages 25, 26 and 31 of Concawe’s Report no 7/16 “Performance of European cross-
country oil pipelines” (“CONCAWE ‘16”).
42
https://www.neb-one.gc.ca/sftnvrnmnt/sft/dshbrd/mp/dt-eng.html
43
https://www.neb-one.gc.ca/sftnvrnmnt/sft/pplnrptr/index-eng.html
44
http://www.epri.com/abstracts/Pages/ProductAbstract.aspx?productId=000000003002000079
45
http://www.agenarisk.com/
46
HSE Nussey, “Failure Frequencies for major failures of high pressure storage vessels at
COMAH sites: A comparison of data used by HSE and the Netherlands,” Dec. 2006.
47
Pitiglio, Paolo, Bragatto, Paolo, and Delle Site Corrado, “Updated failure rates and risk
management in process industries”, 68th Conference of the Italian Thermal Machines
Engineering Association, p. 130.
48
http://www.safelng.nl/failure-frequencies-composite-hoses/ content excerpt as of Oct. 8, 2016.
49
http://www.safelng.nl/
50
TNO, van der Weijde, G.D., and Klootwijk, C.W., Proving leak before burst for small scale
LNG transfer hoses, Phase 1, TNO 2015 R10689, August 2015.
http://www.safelng.nl/proving-eak-befor-burst-small-scale-lng-transfer-hoses/
51
TNO, van der Weijde, G.D., Proving leak before burst for small scale LNG transfer hoses
Phase 2, TNO 2016 R10126, January 2016. http://www.safelng.nl/1114-2/
52
FERC Guidance Manual for Environmental Report Preparation, Vol. II, LNG Facility
Resource Reports 11 & 13 Supplemental Guidance, DRAFT, December 2015.
53
Personal conversation of Rich Kooy with Secil Toran, Small-Scale LNG Key Program,
ENGIE, April 26, 2016.
54
http://www.gastechnology.org/Training/Documents/LNG17-proceedings/5-2-
Anthony_Acton.pdf
55
Lagarrigue, Vincent, “Transforming the LNG transfer zone”, LNG World Shipping, April 27,
2016, http://www.lngworldshipping.com/news/view,transforming-the-lng-transfer-
zone_42757.htm
56
TNO, van der Weijde, Gerard, and Mallon, Niels, “Qualification of Multi-Composite Hoses
for STS LNG Transfer”, Proceedings of the ASME 2011 30th Internationl Conference on
Ocean, Offshore and Arctic Engineering (OMAE2011), June 19-24, 2011, pp. 209-215
57
DNV, Report to Skangass AS, “Appendix A – Assumptions Register”, June 11, 2013, page
19, http://dsbinfo.no/Global/7%20Vedlegg%20A.pdf and DNV, Report to Skangass AS,
“Appendix D – Leak Frequencies Analysis”, June 11, 2013, page 12,
http://dsbinfo.no/Global/10%20Vedlegg%20D.pdf
58
Eide, S.A, Khericha, S.T., Calley, M.B, and Johnson, D.A., Component External Leakage and
Rupture Frequency Estimates, Idaho National Engineering Laboratory, EGG-M-92256, 1992,
p. 1.
59
CH-IV International Inc., Safety History of International LNG Operations, Rev. 12, Feb. 2012,
p. 4.
60
FERC Issuance 20050615-0176
61
FERC Issuance 20050615-0177
Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 241
62
Cadwallader, L, and Pinna, T., “Reliability Estimation for Double Containment Piping”, 20th
ANS Topical Meeting on the Technology of Fusion Energy, NIL/CON-12-24848, August
2012.
63
RIVM, Beerens, H.I., Post, J.G., and Uijt de Haag, P.A.M., “The use of generic failure
frequencies in QRA: The quality and use of failure frequencies and how to bring them up-to-
date.”, Journal of Hazardous Materials, Vol. 130, Issue 3, 2006, pp. 265-270.
64
Falck, Andreas, Bain, Brian, and Rodsaetre, Laila, “Leak Frequency Modelling for Offshore
QRA based on the Hydrocarbon Release Database”, Hazards XXI Symposium Series No. 155,
© 2009 IChemE, p. 644.
65
ISO Guidance on performing risk assessment in the design of onshore LNG installations
including the ship/shore interface ISO/TS16901, First Edition, 2015.
66
EIGA Seveso Documents – Guidance on Applicability, Assessment and Legal Documents for
Demonstrating Compliance of Industrial Gases Facilities with SEVESO Directive(s), European
Industrial Gases Association DOC 60/15, 2015.
67
Williams, Ted, and Assing, Nneka, “Risk Based LNG Facility Siting and Safety Analysis in
the U.S.: Recent Developments”, Proceedings of LNG17, Houston, TX, 2013.
68
McInerney, Ellen H., Hart, Ryan J., Morrison, Delmar “Trey” III, and Kytomaa, Harri K.,
“New Quantitative Risk Criteria for US LNG Facilities”, Process Safety Progress, March
2014.
69
US EIA Today in Energy, “United Kingdom increases oil production in 2015, but new field
development declines”, March 28, 2016.
http://www.eia.gov/todayinenergy/detail.php?id=25552#

Public Final Report DTPH5615T00008 Statistical Review and Gap Analysis of LNG Failure Rate Table Page 242