{"operation":"document","citation":"PHMSA FIR, TransCanada Northern Border Inc., 2011-07-20","title":"Failure Report- TransCanada Northern Border Inc. 7/20/11","source_type":"incident","agency":"Pipeline and Hazardous Materials Safety Administration","status":"historical","official":true,"published_on":"2013-02-26","effective_on":"2011-07-20","summary":"TransCanada Northern Border Inc.; Gas Transmission; WY; failure 2011-07-20; apparent cause: Construction Damage.","machine_formats":{"json":"https://regulus.evalyn.ai/document/phmsa-fir-failure-report-transcanada-northern-border-inc-72011.json","markdown":"https://regulus.evalyn.ai/document/phmsa-fir-failure-report-transcanada-northern-border-inc-72011.md"},"app_url":"https://regulus.evalyn.ai/document/phmsa-fir-failure-report-transcanada-northern-border-inc-72011","source_url":"https://www.phmsa.dot.gov/safety-reports/failure-report-transcanada-northern-border-inc-72011","body":"<<<PAGE 1>>>\n\nDOT US Department of Transportation\nPHMSA Pipelines and Hazardous Materials Safety Administration\nOPS Office of Pipeline Safety\nWestern Region\nPrincipal Investigator\nRegional Accident\nCoordinator\nPeter J. Katchmar\nPeter J. Katchmar\nRegion Director Chris Hoidal\nDate of Report 11/07/2012\nSubject Failure Investigation Report – TransCanada/Bison Pipeline\nNatural Gas Transmission Release near Gillette, WY\nOperator, Location, & Consequences\nDate of Failure 7/20/2011\nCommodity Released Natural Gas\nCity/County & State Gillette/Campbell/Wyoming\nOpID & Operator Name 32487 – TransCanada Northern Border Inc. (TransCanada)\nUnit # & Unit Name 73819 - Bison Pipeline (MP0 to MP 64.5)\nSMART Activity # 135336\nMilepost / Location MP 16.2\nType of Failure Material Failure of Pipe or Weld/Construction Related\nFatalities 0\nInjuries 0\nDescription of area\nimpacted\nRural, High Desert with rolling hills\nProperty Damage $6,700,000 Total\nPage 1 of 10\n\n<<<PAGE 2>>>\n\nExecutive Summary\nOn July 20th, 2011, Bison Pipeline, operated by TransCanada Northern Border Inc.\n(TransCanada) experienced a rupture at Milepost 16.2 at approximately 7:30 pm Mountain\nDaylight Time (MDT). The release site was in a remote, high desert area of rolling hills located\nin Campbell County, Wyoming approximately 17 miles northwest of Gillette, Wyoming. The\nrelease location was about 1 mile uphill and west of the landowner’s house. The pressure drop\nwas detected by the line break system and isolation valves were automatically activated.\nConcurrently, the pressure drop at the Buffalo meter station sixteen miles upstream was\nidentified by TransCanada’s supervisory control and data acquisition (SCADA) operations\ncontrol center (OCC) in Canada. Near the same time, the landowner called TransCanada\npersonnel to report the explosion.\nTransCanada personnel were dispatched to the area to assess the situation. The first responders\nconfirmed that a pipeline failure had occurred and that automated isolation valves operated as\ndesigned. They secured the site and coordinated with local emergency responders. There was\nno ignition of the gas or fire and there were no reported injuries or third party property damage.\nThe National Response Center was notified on July 21, 2011 at 13 minutes past midnight EDT\n(10:13 pm MDT).\nThe line failed after being in service for approximately six months. The apparent cause of the\nfailure was pipe body rupture that initiated at a defect caused by mechanical damage during\nconstruction. The defect was a dent with a crack. The section of line containing the failure\nfeature was hydrostatically tested on November 5, 2010. The maximum hydrostatic test pressure\nof 1890 psi measured at the test point resulted in a maximum pressure at the failure site of\napproximately 2004 psi. The pressure differential was due to the elevation difference. The post\nconstruction hydrotest caused the pipe dent to re-round and opened a crack in the gouge. A full\nmetallurgical analysis was conducted which determined that the crack was near failure at the end\nof the hydrostatic test and it was just a matter of time until the internal pressure in the pipe from\nnormal operations drove the defect to failure.\nSystem Details\nThe Bison Pipeline is a 302 mile long natural gas transmission pipeline. It originates at the\nAnadarko compressor station northwest of Gillette, Wyoming and travels northeast through\nWyoming into southeastern Montana and southwestern North Dakota where it connects with\nTransCanada’s Northern Border Pipeline system (Northern Border) near Northern Border's\nCompressor Station No. 6 in Morton County, North Dakota. The Bison Pipeline was constructed\nin 2010 and commenced operation on January 14, 2011. Bison is comprised of 30-inch diameter\npipe with a nominal wall thickness of 0.438\" (0.72 design factor) between MP 0.00 and MP 64\nand a nominal wall thickness of 0.386\" (0.80 design factor) between MP 64 and MP 302. The\npipeline between MP 64 and MP 302 was commissioned and approved for a maximum operating\npressure (MAOP) of 1440 psig under the alternative maximum allowable operating pressure for\ncertain steel pipelines rule pursuant to §192.620.\nThe line pipe for the Bison Pipeline was manufactured by two companies; Welspun, in Dahej,\nIndia, and EVRAZ at their mill in Regina, Canada. The EVRAZ 0.386\" wall thickness line pipe\nwas used for the entire pipe downstream of Milepost (MP) 64 to the end of the line at MP 302, as\nwell as a few heavy wall pipe sections installed within the first 64 miles. The failed pipe was\nPage 2 of 10\n\n<<<PAGE 3>>>\n\nrolled by Welspun. This pipe was used only within the first 64.7 miles of the line in Wyoming.\nWelspun obtained the steel plate from Ilyich Iron and Steel Works of Mariupol in the Ukraine.\nThe pipe was rolled using a double submerged arc welding (DSAW) process for the longitudinal\nseam according to the 43rd edition of API 5L PSL2 in 2007. The pipe was shipped to the USA\nin 2008-2009, and upon acceptance, the pipe was double-jointed and coated with fusion bonded\nepoxy (FBE) external protective coating and an internal friction reduction coating, at Bayou\nCoating LLC in New Iberia, Louisiana, in 2009.\nEvents Leading up to the Failure\nPHMSA’s Western Region (WR) and Central Region worked closely with TransCanada during\nthe construction of the Bison Pipeline in 2010. The WR spent 47 days of field inspections and\n17 days of associated in office work during the construction and the CR spent an additional 49\nfield inspections days and 9 days of associated in office work during the construction. Because\nthe last 238 miles of the Bison line was to be an alternate Maximum Allowable Operating\nPressure (AMAOP) pipeline (to be operated at 80% SMYS), TransCanada decided to impose all\nof the more stringent post construction requirements on the entire pipeline including the first 64\nmiles of regular MAOP pipeline (to be operated at 72% SMYS). The post construction\nrequirements were to conduct an MFL and geometry in-line-inspection (ILI) survey within six\nmonths of commissioning the line as well as conducting a cathodic protection (CP) Close-\nInterval-Survey (CIS) and a direct current voltage gradient (DCVG) survey within six months of\nenergizing the CP system.\nThe pipeline was post-construction hydrostatically tested in November 2010 with no leaks or\nfailures. A post construction caliper ILI survey was completed in November 2010 and there\nwere no actionable anomalies identified per 49 CFR Part 192.\nThe installation of four of the eight impressed current cathodic protection ground beds were\ndelayed due to weather and a 120 day extension to complete the installation of the cathodic\nprotection system was requested and granted by the PHMSA Western Region Director. With the\ndelay of installation of the impressed current cathodic protection systems, the CIS and DCVG\nsurveys were delayed. The DCVG survey was completed on July 22, 2011 and the CIS was\ncompleted on September 23, 2011. Results of the CIS and DCVG surveys were provided to\nPHMSA Western Region after the release.\nOn the day of the failure, the pipeline pressure was generally consistent at 1340 psig up to the\napproximate 7:25 pm MDT rupture. TransCanada’s Gas Control in Canada observed the drop in\npressure and initiated telephone conversations with Anadarko, the gas supplier. The rupture was\nconfirmed and the line was shut in by 8:50 pm MDT. The maximum pressure on the day of the\nfailure was 1396.8 psi as logged by the SCADA system at the upstream Buffalo Meter Station.\nAccording to SCADA the failure was estimated to have occurred at a pressure of 1340 psig.\nPrior to the incident, the pipeline was operating normally.\nEmergency Response\nTransCanada was made aware of the release by their SCADA system in their Gas Control\nCenter. The on duty controller initiated telephone calls to their gas supplier in an attempt to\nunderstand the situation. At nearly the same time, the landowner called TransCanada to report\nthe release. There is an automatic valve just downstream of the release point that closed just\nPage 3 of 10\n\n<<<PAGE 4>>>\n\nafter the release. It took some time for the gas supplier to shut down their compressor station at\nthe beginning of the line so TransCanada could close the upstream valve to isolate the release\nlocation. TransCanada personnel notified local emergency responders and mobilized personnel\nto the site. The National Response Center (NRC) was notified at thirteen minutes after midnight\non the 21st of July Eastern Daylight Time (EDT). This corresponds to 10:13 pm Mountain\nDaylight Time (MDT).\nPHMSA was made aware of the release by an e-mail report from the National Response Center\n(NRC) that came in at 10:39 pm MDT on July 20, 2011. By 10:58 pm, PHMSA’s Western\nRegion after hour’s duty officer had called and talked with TransCanada personnel who\nconfirmed the release but had little further information. PHMSA’s WR Accident Coordinator\nwas notified and made contact with the WR Director and by 11:00 pm, it was decided that\nbecause the Bison line had just been commissioned, the WR should go to the site and conduct an\ninvestigation into the cause of the release. The WR Accident Coordinator left early the next\nmorning and drove to the site, arriving at 10:00 am MDT on July 21, 2011. The site had been\nsecured by TransCanada/Bison representatives the night before and nothing had been touched or\nremoved. PHMSA started the investigation by photographing the location, looking over the\nfracture surface, generally measuring the pipe to make sure it was all there and conducting\ninformal interviews of first responders and the landowner. The landowner mentioned that his\nwife had taken some pictures immediately after the release occurred and PHMSA secured a copy\nof those photographs. The interviews gave little information as to what may have occurred\nimmediately prior to the release that gave any indication as to the apparent cause.\nIt was determined that there was no nearby construction work that predicated the release so 3rd\nparty damage was removed as a probable cause. Also, the pipe had no apparent internal or\nexternal corrosion and the operator reported that there were no apparent pressure excursions\nprior to the failure. There were some areas of possible external damage located on the pipe\nwhich could have occurred during the pipe thrashing about during the release or they could have\noccurred during construction.\nA few hours after PHMSA arrived on site, a metallurgist from TransCanada arrived at the site.\nThe metallurgist and the PHMSA representative conducted a combined review of the failed\npipe’s fracture surface. The metallurgist indicated that the failure involved a classic 45 degree\nshear fracture on both ejected pipe fragments and the two arrest ends. He indicated that that\nfracture mode is indicative of high steel toughness. A number of abrasions on fracture surfaces\nwere interpreted to be secondary abrasions that occurred during the failure event as the pipe\ncame through the ground and landed. He also identified a location where he observed an area of\nflat fracture that coincided with mechanical damage that he thought might be the origin of\nfailure. He cleaned the fracture surface with water and a tooth brush and photographs were taken\nof that area as well as the other probable mechanical damage pipe a few feet away. The fracture\nat the probable origin of failure was observed to have a 'woody' morphology and to have a dark\nstain line on the surface with a curved profile resembling a possible crack arrest line. The\nmetallurgical laboratory concluded later that the length of the dark stain line was measured to be\n4.5\" long with a maximum depth of approximately 0.372- inch (89%) of wall thickness. Based\non the positions of the long seam welds on the arrest ends and on the elected pipe, the damage\nwas estimated to be located on the top of the pipe at the 10:15 o'clock position (looking\nPage 4 of 10\n\n<<<PAGE 5>>>\n\ndownstream). The TransCanada metallurgist explained to the PHMSA investigator that he had\nseen this type of mechanical damage grow to failure in the past but that it usually took some 30\nto 40 years. He found it curious that this mechanical damage grew to failure after the post\nconstruction hydrotest in just under 7 months.\nPHMSA’s investigator released the site to the operator on the afternoon of July 21, 2011.\nTransCanada personnel had a third party survey crew come in and survey the site, official\nphotographs were taken of the site and the crater size and the location of all pipe thrown from the\nditch was documented. Several pieces of the pipe containing the probable failure initiation point\nand other pipe containing mechanical damage were cut at the site and transported to a third party\nlaboratory for analysis.\nSummary of initial start-up plan and return-to-service, including preliminary safety\nmeasures\nDue to this release occurring within 6 months of commissioning, and the fact that the pipe\nwithstood a post construction hydrotest, PHMSA was concerned with the quality of pipe used to\nconstruct the Bison Pipeline. Specifically, PHMSA was concerned that the pipe may have lower\nthan required physical properties which may have allowed the mechanical damage to grow to\nfailure in a very short amount of time. PHMSA issued a Corrective Action Order (CAO) to\nTransCanada on July, 21, 2011. The CAO required TransCanada to develop a written restart\nplan prior to repairing and resuming operation of the pipeline; drop the operating pressure to\n80% of the failure pressure (not to exceed 1072 psig) at the rupture location; develop a remedial\nwork plan to identify and address any additional anomalous conditions in the pipe; and provide\nquarterly progress reports to PHMSA’s WR.\nOn July 21st, 2011, TransCanada mobilized an excavation contractor to the site at MP 16.2. The\nwork consisted of repairs to the pipeline and remediation of the terrain. The total length of the\npipeline exposed to make repairs to the pipeline was approximately 140'. The failed section was\nreplaced with approximately 109 feet of new pipe that was hydrostatically tested at the site and\nwitnessed by TransCanada Engineering personnel. Pipe used in the repair consisted of 30\"\ndiameter, 0.514\" wall thickness, Grade X-70. The final tie-in was completed on July 22nd, 2011.\nOn July 26, 2011, TransCanada submitted a letter to PHMSA requesting to be allowed to return\nto service at 1200 psig in order to run ILI surveys in order to identify any dents in the pipe\nsimilar to the one that failed. PHMSA agreed with the stipulation that TransCanada would\nimmediately reduce the pressure to the CAO requirement of 1072 psig immediately after the ILI\nsurveys were complete.\nBecause the preliminary cause of the failure was identified as a top-side dent with metal loss\nPHMSA approved the return to service request contingent upon TransCanada excavating and\nevaluating all known top side (9 o’clock to 3 o’clock) dents on the pipeline as revealed by either\nof the 2010 and 2011 ILI runs. Based on information provided by TransCanada after the release,\nthere were at least four (4) topside dents between milepost 0 and milepost 161. PHMSA also\nrequired that prior to operating the pipeline that TransCanada provide documentation on the\n2010 and 2011 ILI results and any subsequent excavation, assessment, and evaluation of\nanomalies including dents. Each dent was to be evaluated to confirm that it did not pose a safety\nPage 5 of 10\n\n<<<PAGE 6>>>\n\nrisk to operating the line at the modified pressure. PHMSA’s WR observed each anomaly dig\nand confirmed the non-destructive testing of each possible dent.\nInvestigation Findings & Contributing Factors\nImmediately after learning of the release on the Bison line, PHMSA’s WR Director requested\nthat his Accident Coordinator drive to the site and initiate an investigation into the cause of the\npipe failure. PHMSA’s WR staff immediately reviewed all construction documentation in an\neffort to identify any inspector identified anomalous information with respect to the rupture\nlocation. None was identified. Information on the pipe used for constructing the pipeline was\nalso reviewed. It appeared that the pipe met the applicable manufacturing specifications for the\n72% MAOP section of line from MP 0 to MP 64 and in fact met the TransCanada specification\nfor the 80% pipe from MP 65 to MP 302 (TES-PIPE-SAW-US). PHMSA requested, received\nand reviewed all post construction integrity surveys. These included the post construction\nhydrotest, the post hydrotest caliper ILI survey, the DCVG Survey and the July partial\ncombination ILI survey.\nThe post construction caliper tool was a rudimentary tool which identified a reported 1.1%\noutside diameter (OD) dent at the release location. A combination MFL and geometry ILI\nsurvey was conducted in July 2011; 2 weeks prior to the failure. The results were not available\nbefore the release. After full review of the ILI survey, it was determined that the tool collected\ndata for only a portion of the line surveyed and so a rerun had been scheduled. The tool did\ncollect data for the first section of pipeline that contained the release. That ILI survey identified\na 0.5% OD dent at the release location. A DCVG survey was conducted on the Bison line in\nJune 2011. This survey identified a coating holiday DCVG indication of 14.57% voltage drop\n(IR) coincident with the slight dent at the rupture location.\nNone of these anomalies were actionable per the criteria listed in PHMSA’s regulations or in\nindustry standards at the time they were detected. It was not until after the July 20th release that\nthe different survey results were overlaid and only then did it become obvious that a criterion for\naction, a dent with metal loss requiring immediate mitigation, was identified.\nTransCanada contracted with a third party metallurgical laboratory to perform a full\nmetallurgical analysis on the failed pipe in an effort to determine the probable cause of the\nrelease and to make a determination of why the pipe failed within 6 months of commissioning.\nThe first metallurgical report contained a detailed description of the pipe properties and the\nanomaly at the rupture location. The anomaly was confirmed as the origin of the failure but the\nreport was inconclusive as to the reason for the rapid growth of the anomaly to failure within 6\nmonths after the post construction hydrotest.\nBecause the first metallurgical laboratory could not specifically determine the mechanism that\ncaused the anomaly to grow to failure in the apparent rapid manner that it did, PHMSA was\nreluctant to accept the results as final and close the investigation. Discussions with TransCanada\nled to further metallurgical testing by a laboratory that had the ability to clarify if hydrogen\ncracking was the cause of the rapid crack growth of the anomaly involved in the release. Three\ncritical tests were performed to clarify if hydrogen cracking was involved in the rapid crack\nPage 6 of 10\n\n<<<PAGE 7>>>\n\ngrowth: (1) slow strain rate tensile (SSRT) testing in NS4 solution under hydrogen charging\ncondition; (2) SSRT in air, and (3) SEM characterization of the nature of blisters and cleavage.\nThe following is the executive summary from the first laboratory report:\n“The failure originated in a 6.4-inch (12.7 cm) long area of mechanical damage. The total extent\nof mechanical damage at the failure origin was approximately 13 inches (33 cm). The pipeline\nfailed as a rupture that propagated longitudinally approximately 81 feet (25m) prior to\ntermination. Two pipe sections were ejected from the ditch line during the failure, one\napproximately 68 feet long (21 m) and one approximately 13 feet long (4 m). The mechanical\ndamage removed up to 11-percent of the pipe wall thickness. Rerounding cracks to a maximum\ndepth of 14-percent of the pipe wall thickness developed in the mechanical damage that\npropagated to failure in less than 7 months of operation. The combination of the mechanically\ndamaged material removed from the pipe outside surface and the re-rounding cracks reduced the\neffective pipe wall thickness by 25-percent with an initial length of 3.2 inches (8 cm).\nMechanically damaged areas on high strength line-pipe can be susceptible to crack propagation\ndue to a number of possible individual degradation mechanisms or a combination of degradation\nmechanisms. The microstructure of the pipe where the failure occurred consists of bands of fine\ngrained ferrite and pearlite. The morphology of the propagated crack portion of the fracture\nsurface was layers of cleavage bounded by layers of shear. The maximum depth of the cracking\nwas 89-percent of the nominal pipe wall thickness. The final portion of the fracture was ductile\nshear adjacent to the inside pipe surface. Further testing is being performed to determine the\ndegradation mechanisms responsible for the rapid crack growth from the initial mechanical\ndamage.\nThe pipe specimen met the requirements of API Specification 5L, 43nd Edition, March 2004, for\nGrade X70 Welded Line Pipe, the API specification in effect the year the pipe was\nmanufactured.”\nThe following is the executive summary from the second laboratory report:\n“Kiefner Associates, Inc. (KAI) reported in the analysis of the Bison Pipeline incident a rapid\ncrack propagation of 500 mils per year for the six month in-service operation after\ncommissioning. Since the presumed rapid crack growth rate could not be explained with any of\nthe common steel degradation mechanisms, such as external corrosion, SCC, cold creep, fatigue\nand corrosion fatigue, KAI in its first draft report attributed it to hydrogen assisted cracking in\nvery aggressive environments. The observational evidence for this mechanism was the partial\nmaterial cleavage on the fracture surface and microvoids assumed to be hydrogen blisters in the\nvicinity of fracture surfaces. However, from the six months operating history, the environment\nat the rupture location could not be classified as aggressive. Therefore, the presumed “hydrogen\nassisted cracking” mechanism was removed from KAI’s third draft, i.e., the final report but the\nrapid crack growth issue remained unresolved.\nTransCanada requested Blade Energy Partners to perform a supplemental investigation with an\nobjective to provide factual evidence for the rapid crack propagation mechanism. The scope of\nwork consisted of six investigation components, including gap analysis of the previous findings\nby KAI and developing approaches and testing procedures that could identify and verify possible\nfailure mechanism(s).\nPage 7 of 10\n\n<<<PAGE 8>>>\n\nThree critical tests were performed to clarify if hydrogen cracking was involved in the rapid\ncrack growth: (1) slow strain rate tensile (SSRT) testing in NS4 solution under hydrogen\ncharging condition; (2) SSRT in air, and (3) SEM characterization of the nature of blisters and\ncleavage.\nComparisons of the fracture surface morphology produced by SSRT tests and field failure\nindicated that hydrogen was not involved in the rapid crack propagation of Bison failure. SEM\nconfirmed that the microvoids near the fracture surfaces were not hydrogen blisters but formed\nby severe plastic deformation. The cleavage reported by KAI was not because of the presumed\nhydrogen ingress but associated with the inherit nature of pearlite cracking. The investigations\nconcluded that hydrogen played no role in any of the cracking steps.\nBased on the above findings, further study proceeded to address the unresolved mechanism for\nrapid crack growth, including (1) characterization of fracture surfaces to determine the processes\ninvolved in the Bison failure; (2) determination of the possible mechanism for rapid crack\ngrowth; (3) small scale confirmatory testing to demonstrate that large crack growth by ductile\ntearing during hydrostatic test is possible, and (4) additional metallurgical testing for material\nqualification. The findings are summarized as follows:\n1. As evidenced by detailed fractographic analysis, the fracture surface consisted of three zones\nthat correspond to three processes involved in the Bison failure:\n‒ Zone 1: 27% wt initial crack caused by mechanical damage and dent rebounding. The crack\nwas underneath the gouge, about 10% wt deep, and demarcated by black stain on the\nfracture surface. The crack was associated with 17% wt metal loss (wall thinning) by\ngouging.\n‒ Zone 2: subcritical crack growth from 27% wt to 85% wt by two different mechanisms:\ni. Zone 2a: a semi-elliptical-shaped region demarcated by presence of banded fracture\nsurface alternating between microvoids (dimples) and quasi cleavage. Zone 2a is the\none related to the rapid crack growth and was the focus of the study.\nii. Zone 2b: a narrow (1mm wide) region demarcated by black stain, i.e., thick corrosion\nscale. The appearance and chemical composition of the corrosion scale of Zone 2b\nwere distinctly different from Zone 1, suggesting that Zone 1 formed prior to pre-\ncommissioning hydrostatic test while Zone 2b formed after hydrostatic test.\n‒ Zone 3: final rupture occurred when the crack reached its critical size (3.8” x 85%wt) at the\noperating pressure.\n2. The mechanism for rapid crack growth in Zone 2a was identified to be purely mechanical.\nFree of corrosion and corrosion scale on the fracture surface indicate that Zone 2a was not\nformed by progressive propagation of the crack from its initial depth of 27% wt to about 76%\nwt or more during six months operation; instead, it reflects a large crack increment by ductile\ntearing due to pre-commissioning hydrostatic testing followed by a pressure cycle from zero\nto MAOP.\nPage 8 of 10\n\n<<<PAGE 9>>>\n\n‒ Fracture mechanics calculation showed that a large crack increment up to a depth of 61%\nwt - 71% wt is possible during pre-commissioning hydrostatic test at the pressure of 2004\npsi.\n‒ Fracture mechanics model also showed that a second increment up to a depth of more\nthan 80% wt is possible after hydrostatic test followed by a pressure cycle from zero to\nMAOP (1440 psi).\n‒ Due to uncertainties in dent-gouge geometry, residual stresses, and material properties, it\nis difficult, if not impossible; to precisely quantify crack increment by pre-commissioning\nhydrostatic testing.\n‒ Finally, the crack reached its critical size by a small amount of crack increment, 1mm\n(0.04”) or less, during six-month operation either due to corrosion or slow ductile tearing,\nor a combination. A flaw of this size would fail at the operating pressure as evidenced by\nthe failure and fracture mechanics calculation.\n3. The small scale single edge notched tensile tests (SENT) at a constant stress level of 94%\nSMYS demonstrated and provided a direct evidence that large crack growth by ductile tearing\nduring hydrostatic test is feasible. The fracture surface morphology produced by ductile\ntearing test is identical to that produced by Bison failure (Zone 2a).\n4. Metallurgical and mechanical testing confirmed that the material meets API 5L PSL2\nspecification for X70 as reported by KAI.\nIn summary, the root cause for Bison pipeline rupture was severe mechanical damage in the form\nof a gouged dent containing cracks. Hydrogen played no role in the Bison failure incident. The\nmechanism for rapid crack growth in Bison failure was determined to be purely mechanical.\nFree of corrosion and corrosion scale on the fracture surfaces indicates that the Zone 2a was not\nformed by progressive propagation of the initial crack during six months operation; instead it\nreflects a large crack increment by ductile tearing due to pre-commissioning hydrostatic testing\nfollowed by a pressure cycle from zero to MAOP. This conclusion is supported by fracture\nmechanics calculations; single edge notched tensile tests, and is consistent with the well\ndocumented phenomenon that time-dependent crack growth by slow ductile tearing could occur\nduring hydrostatic test.”\nConclusions:\nPHMSA reviewed the final reports and held a meeting with TransCanada and representatives\nfrom the second laboratory in order to fully understand the testing preformed and the conclusions\nmade. PHMSA concurs with the scientific methods utilized in the further testing as well as the\nconclusions drawn from the testing.\nPHMSA finds that an anomaly was introduced to the pipe at the ultimate failure location during\nfinal phases of construction. This anomaly caused a dent with metal loss in the pipe wall. The\npost construction hydrotest caused the dent to reround and the damaged pipe to crack to a point\nPage 9 of 10\n\n<<<PAGE 10>>>\n\njust prior to failure. Subsequent commissioning and operation for 6 months caused the crack to\npropagate to failure with no additional outside forces acting on it.\nAppendices\n1. Maps and Photographs\n2. NRC Report\n3. Operator Accident/Incident Report to PHMSA\n4. Metallurgical Analysis Report #1\n5. Metallurgical Analysis Report #2\nPage 10 of 10\n\n<<<PAGE 11>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nGoogle Map of Topography of release location NE of Gillette, WY.\nPage 1 of 11\n\n<<<PAGE 12>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nPicture from Landowner’s House immediately after rupture.\nPage 2 of 11\n\n<<<PAGE 13>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nPicture from Landowner’s House immediately after rupture.\nPage 3 of 11\n\n<<<PAGE 14>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nLooking South from Rupture with long piece of ejected pipe in the foreground.\nLooking North at Rupture Location with long piece of ejected pipe in the foreground.\nPage 4 of 11\n\n<<<PAGE 15>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nClose up of North end of failed pipe looking SE\nFull rupture crater looking South\nPage 5 of 11\n\n<<<PAGE 16>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nCoating repair and additional external damage (post rupture) on long piece of ejected pipe.\nClose up of Probable Failure Initiation showing external damage and visible cracking.\nPage 6 of 11\n\n<<<PAGE 17>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nClose up of Probable Failure Initiation showing external damage and visible cracking.\nClose up of Probable Failure Initiation showing external damage.\nPage 7 of 11\n\n<<<PAGE 18>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nTransCanada Metallurgist cleaning probable fracture initiation point with toothbrush and water.\nClose-up TransCanada Metallurgist cleaning probable fracture initiation point with toothbrush and\nwater.\nPage 8 of 11\n\n<<<PAGE 19>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nClose-up of probable fracture initiation point after cleaning with measuring tape.\nPage 9 of 11\n\n<<<PAGE 20>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nClose-up of probable fracture initiation point after cleaning with measuring tape.\nPage 10 of 11\n\n<<<PAGE 21>>>\n\nMaps and Pictures – TransCanada Bison Pipeline Rupture 7/20/2011\nExtreme Close-up of probable fracture initiation point after cleaning with measuring tape.\nPage 11 of 11\n\n<<<PAGE 22>>>\n\nPage 1 of 2\nNATIONAL RESPONSE CENTER 1-800-424-8802\n*** For Public Use ***\nInformation released to a third party shall comply with any\napplicable federal and/or state Freedom of Information and Privacy Laws\nIncident Report # 983351\nINCIDENT DESCRIPTION\n*Report taken at 00:13 on 21-JUL-11\nIncident Type: PIPELINE\nIncident Cause: UNKNOWN\nAffected Area:\nThe incident was discovered on 20-JUL-11 at 19:15 local time.\nAffected Medium: AIR ATMOSPHERE\n____________________________________________________________________________\nSUSPECTED RESPONSIBLE PARTY\nOrganization: BISON PIPELINE LLC\nHOUSTON, TX\nType of Organization: PRIVATE ENTERPRISE\n____________________________________________________________________________\nINCIDENT LOCATION\nSEE LAT /LONG County: CAMPBELL\nCity: GILLETTE State: WY\nLatitude: 44° 00' 23\" N\nLongitude: 105° 00' 50\" W\nPIPELINE MP 16.2\n____________________________________________________________________________\nRELEASED MATERIAL(S)\nCHRIS Code: ONG Official Material Name: NATURAL GAS\nAlso Known As:\nQty Released: 0 UNKNOWN AMOUNT\n________________________________________________________________________\nDESCRIPTION OF INCIDENT\nCALLER IS REPORTING A RELEASE OF NATURAL GAS IN A PIPELINE DUE TO TWO JOINTS OF\nPIPELINE COMING OUT OF THE GROUND. THE CAUSE IS UNKNOWN AT THIS TIME.\n____________________________________________________________________________\nINCIDENT DETAILS\nPipeline Type: TRANSMISSION\nDOT Regulated: YES\nPipeline Above/Below Ground: BELOW\nExposed or Under Water: NO\nPipeline Covered: UNKNOWN\n____________________________________________________________________________\nDAMAGES\nFire Involved: NO Fire Extinguished: UNKNOWN\nINJURIES: NO Hospitalized: Empl/Crew: Passenger:\nFATALITIES: NO Empl/Crew: Passenger: Occupant:\nEVACUATIONS: NO Who Evacuated: Radius/Area:\nDamages: NO\nLength of Direction of\nClosure Type Description of Closure Closure Closure\nAir: N\nRoad: N\nWaterway: N\nMajor\nArtery:\nN\nTrack: N\nPassengers Transferred: NO\nEnvironmental Impact: UNKNOWN\nMedia Interest: NONE Community Impact due to Material:\nhttp://www.nrc.uscg.mil/reports/rwservlet?standard\nweb+inc\n_\n_\nseq=983351\n9/7/2012\n\n<<<PAGE 23>>>\n\nPage 2 of 2\n____________________________________________________________________________\nREMEDIAL ACTIONS\nPIPELINE IS ISOLATED.\nRelease Secured: YES\nRelease Rate:\nEstimated Release Duration:\n____________________________________________________________________________\nWEATHER\nWeather: SUNNY, 90ºF\n____________________________________________________________________________\nADDITIONAL AGENCIES NOTIFIED\nFederal: NONE\nState/Local: NONE\nState/Local On Scene: NONE\nState Agency Number: NONE\n____________________________________________________________________________\nNOTIFICATIONS BY NRC\nUSCG ICC (ICC ONI)\n21-JUL-11 00:25\nCGIS RAO ST. LOUIS (COMMAND CENTER)\n21-JUL-11 00:25\nCOLORADO INFO ANALYSIS CENTER (FUSION CENTER)\n21-JUL-11 00:25\nDOT CRISIS MANAGEMENT CENTER (MAIN OFFICE)\n21-JUL-11 00:25\nU.S. EPA VIII (MAIN OFFICE)\n21-JUL-11 00:31\nNE INFORMATION ANALYSIS CENTER (MAIN OFFICE)\n21-JUL-11 00:25\nNATIONAL INFRASTRUCTURE COORD CTR (MAIN OFFICE)\n21-JUL-11 00:25\nNOAA RPTS FOR WY (MAIN OFFICE)\n21-JUL-11 00:25\nNTSB COMM. CTR (MAIN OFFICE)\n21-JUL-11 07:53\nDOI/OEPC DENVER (MAIN OFFICE)\n21-JUL-11 00:25\nWY DEPARTMENT OF ENVIRON QUALITY (MAIN OFFICE)\n21-JUL-11 00:25\nWYOMING CRIMINAL INTEL CENTER (SR INTELLIGENCE OFFICER)\n21-JUL-11 00:25\nWYOMING OFFICE OF HOMELAND SECURITY (OPERATIONS DIVISION)\n21-JUL-11 00:25\n____________________________________________________________________________\nADDITIONAL INFORMATION\nNO ADDITIONAL INFORMATION.\n___________________________________________________________________________\n*** END INCIDENT REPORT # 983351 ***\nhttp://www.nrc.uscg.mil/reports/rwservlet?standard\nweb+inc\n_\n_\nseq=983351\n9/7/2012\n\n<<<PAGE 24>>>\n\nNOTICE: This report is required by 49 CFR Part 191. Failure to report can result in a civil penalty not to\nexceed 100,000 for each violation for each day that such violation persists except that the maximum civil\npenalty shall not exceed $1,000,000 as provided in 49 USC 60122.\nOMB NO: 2137-0522\nEXPIRATION DATE: 01/31/2013\nReport Date: 08/16/2011\nNo. 20110294 - 15608\nU.S Department of Transportation\nPipeline and Hazardous Materials Safety Administration\n--------------------------------------------------\n(DOT Use Only)\nINCIDENT REPORT - GAS TRANSMISSION AND\nGATHERING PIPELINE SYSTEMS\nA federal agency may not conduct or sponsor, and a person is not required to respond to, nor shall a person be subject to a penalty for failure to comply\nwith a collection of information subject to the requirements of the Paperwork Reduction Act unless that collection of information displays a current valid\nOMB Control Number. The OMB Control Number for this information collection is 2137-0522. Public reporting for this collection of information is estimated\nto be approximately 10 hours per response, including the time for reviewing instructions, gathering the data needed, and completing and reviewing the\ncollection of information. All responses to this collection of information are mandatory. Send comments regarding this burden estimate or any other aspect\nof this collection of information, including suggestions for reducing this burden to: Information Collection Clearance Officer, PHMSA, Office of Pipeline\nSafety (PHP-30) 1200 New Jersey Avenue, SE, Washington, D.C. 20590.\nINSTRUCTIONS\nImportant: Please read the separate instructions for completing this form before you begin. They clarify the information requested and provide specific\nexamples. If you do not have a copy of the instructions, you can obtain one from the PHMSA Pipeline Safety Community Web Page at\nhttp://www.phmsa.dot.gov/pipeline.\nPART A - KEY REPORT INFORMATION\nReport Type: (select all that apply) Original: Supplemental: Final:\nYes Yes\nLast Revision Date: 09/10/2012\n1. Operator's OPS-issued Operator Identification Number (OPID): 32487\n2. Name of Operator TRANSCANADA NORTHERN BORDER INC\n3. Address of Operator:\n3a. Street Address 717 Texas Avenue\n3b. City Houston\n3c. State Texas\n3d. Zip Code: 77002\n4. Local time (24-hr clock) and date of the Incident: 07/20/2011 19:30\n5. Location of Incident:\nLatitude: 44.39328\nLongitude: -105.8387\n6. National Response Center Report Number (if applicable): 983351\n7. Local time (24-hr clock) and date of initial telephonic report to the\nNational Response Center (if applicable): 07/21/2011 00:13\n8. Incident resulted from: Unintentional release of gas\n9. Gas released: (select only one, based on predominant volume\nreleased) Natural Gas\n- Other Gas Released Name:\n10. Estimated volume of commodity released unintentionally - Thousand\nCubic Feet (MCF): 50,555.00\n11. Estimated volume of intentional and controlled release/blowdown -\nThousand Cubic Feet (MCF) 41,938.00\n12. Estimated volume of accompanying liquid release (Barrels):\n13. Were there fatalities? No\n- If Yes, specify the number in each category:\n13a. Operator employees\n13b. Contractor employees working for the Operator\n13c. Non-Operator emergency responders\n13d. Workers working on the right-of-way, but NOT\nassociated with this Operator\n13e. General public\n13f. Total fatalities (sum of above)\n14. Were there injuries requiring inpatient hospitalization? No\n- If Yes, specify the number in each category:\n14a. Operator employees\n14b. Contractor employees working for the Operator\n14c. Non-Operator emergency responders\n14d. Workers working on the right-of-way, but NOT\nassociated with this Operator\n14e. General public\n14f. Total injuries (sum of above)\n15. Was the pipeline/facility shut down due to the incident? Yes\n- If No, Explain:\nPage 1 of 13\n\n<<<PAGE 25>>>\n\n- If Yes, complete Questions 15a and 15b: (use local time, 24-hr clock)\n15a. Local time and date of shutdown 07/20/2011 19:40\n15b. Local time pipeline/facility restarted 08/05/2011 19:30\n- Still shut down? (* Supplemental Report Required)\n16. Did the gas ignite? No\n17. Did the gas explode? No\n18. Number of general public evacuated: 0\n19. Time sequence (use local time, 24-hour clock):\n19a. Local time operator identified Incident 07/20/2011 20:15\n19b. Local time operator resources arrived on site 07/20/2011 20:15\nPART B - ADDITIONAL LOCATION INFORMATION\n1. Was the origin of the Incident onshore? Yes\n- Yes (Complete Questions 2-12)\n- No (Complete Questions 13-15)\nIf Onshore:\n2. State: Wyoming\n3. Zip Code: 82716\n4. City Gillette\n5. County or Parish Campbell\n6. Operator designated location Milepost/Valve Station\nSpecify: 16.2\n7. Pipeline/Facility name: Bison Pipeline\n8. Segment name/ID: MLV 0 - MLV 17\n9. Was Incident on Federal land, other than the Outer Continental Shelf\n(OCS)? No\n10. Location of Incident : Pipeline Right-of-way\n11. Area of Incident (as found) : Underground\nSpecify: Under soil\nOther Describe:\nDepth-of-Cover (","truncated":true,"body_characters":67367}