Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
4 Review of Pipeline Incident Data T his chapter reviews U.S. and Canadian pipeline incident statistics and investigations for insight into whether transmission pipelines expe- rience more releases when they transport diluted bitumen than when they transport other crude oils. U.S. AND CANADIAN INCIDENT DATA The Pipeline and Hazardous Materials Safety Administration (PHMSA) requires that all regulated pipeline operators report unintended releases that meet certain thresholds of release quantities or impact severity. PHMSA tracks and analyzes these reports to inform its inspec- tion, investigation, and enforcement activities.1 PHMSA inspectors also conduct more in-depth investigations of selected incidents. Incidents involving especially severe consequences, such as deaths, injuries, evac- uations, and environmental damage, may also be investigated by the National Transportation Safety Board (NTSB). Through field and foren- sic investigations, NTSB assesses both causal and contributing factors and recommends preventive and follow-up actions, including regula- tory responses.2 The National Energy Board (NEB) and Transportation Safety Board (TSB) serve similar functions, respectively, for incidents involving pipelines in Canada. PHMSA and NEB incident statistics and investigations, as well as relevant investigations by NTSB and TSB, are reviewed next. âMore discussion of PHMSA safety oversight programs can be found in Appendix B. 1 2 âNTSB recommendations pertaining to PHMSAâs pipeline safety authorities can be found at http://www.phmsa.dot.gov/pipeline/regs/ntsb. 51
52 Effects of Diluted Bitumen on Crude Oil Transmission Pipelines Annual releases 200 180 160 140 Small releases (less than 50 barrels) 120 100 80 60 40 Large release (50 barrels or more) releases (50 barrels or more) 20 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Year FIGURE 4-1 Crude oil pipeline incidents reported to PHMSA, 2002 to 2011. Source: Incident data were provided to the committee by PHMSA during the October 23, 2012, committee meeting. PHMSA Incident Data and Investigations PHMSA regulations require that operators of hazardous liquid pipe- lines, which include crude oil pipelines, report any incident that involves a release of 5 gallons or more or explosion, fire, serious injury, or sig- nificant property damage.3 Incidents that involve any component of the pipeline facility, including line pipe, tanks, valves, manifolds, and pumps, must be reported. A short reporting form is required for notifying the agency of small releases, and a longer form is required for larger releases and any release into water exceeding 5 gallons. Before 2002 the thresh- old for reporting releases was 50 barrels. The reporting changes make comparisons of recent release data with historical performance difficult. A further complication of the reporting system is that while PHMSA reporting covers most crude oil pipelines, there are exceptions to cover- age, such as some intrastate pipelines and gathering systems. The number of incidents reported for regulated crude oil pipelines during 2002 to 2011 is shown in Figure 4-1. During the 10-year period, the number of large incidents fluctuated from about 80 to 120 per year. â49 CFR 195.50. 3
Review of Pipeline Incident Data 53 Mainline pipe with unknown diameter 3% Other equipment (e.g., pumps and gauges) 65% Mainline pipe diameter <16 in. 18% Mainline pipe diameter >16 in. 5% Tanks 7% Valves 2% FIGURE 4-2 System components involved in crude oil pipeline incidents reported to PHMSA, 2002 to 2012. Source: Data were obtained from analysis of PHMSA data from the Environmental Impact Statement of TransCanada XL permit application (U.S. Department of State 2013, Volume IV, Appendix K). Total releases trended downward from about 190 to 150 per year, with small releases accounting for between one-third and one-half of the total. System components involved in the releases are shown in Fig- ure 4-2. Main-line pipe and tanks were involved in about one-third of the incidents, while all other equipment, such as pumps, valves, and fit- tings, accounted for the rest. A generalization that can be made is that the larger releases tend to be associated with main-line pipe, and sometimes with tanks, whereas the other system components tend to experience smaller releases on average. For 2002 to 2012, the pattern of releases by system component and cause is shown in Figure 4-3 and Table 4-1. The causal distribution differed by component. For main-line pipe, inter- nal corrosion was the cause of about one-third of releases, while external corrosion and outside force damage accounted for most of the remain- der. For most other pipeline components, incorrect operation and
54 Effects of Diluted Bitumen on Crude Oil Transmission Pipelines Unspecified cause Pipe External corrosion Tanks Valves Internal corrosion Pumps Unspecified component Unspecified corrosion Manufacture/construction Equipment malfunction Outside force Incorrect operations Weather or natural 0 100 200 300 400 500 600 FIGURE 4-3 Crude oil pipeline incident reports to PHMSA, by cause of release and system component involved, 2002 to 2012. Source: U.S. Department of State 2013, Volume IV, Appendix K. table 4-1 Crude Oil Pipeline Incident Reports to PHMSA, by Cause of Release and System Component Involved, 2002 to 2012 System Component Involved Unspecified Cause of Release Pipe Tanks Valves Pumps Component Total Weather or natural force 10 10 â0 â29 â20 69 Incorrect operations 5 16 â1 â80 â58 160 Outside force 80 â0 â2 â17 â11 110 Equipment malfunction 1 29 17 491 ââ1 539 Manufacture or construction 31 â7 â1 â67 â41 147 Unspecified corrosion 1 â1 â0 ââ0 191 193 Internal corrosion 103 â7 â3 165 ââ3 281 External corrosion 82 â7 â0 â23 ââ0 112 Unspecified cause 8 16 â1 â37 â22 84 Total 321 93 25 909 347 1,695 Source: U.S. Department of State 2013, Volume IV, Appendix K.
Review of Pipeline Incident Data 55 malfunctioning equipment were the main causes of incidents. Most of the corrosion-related incidents reported to PHMSA occurred in pipes and pumps. Main-line pipe was the dominant location for external corrosion. Whereas main-line pipe also accounted for about one-third of incidents involving internal corrosion, more of these incidents occurred in pumps. Each year, PHMSA inspectors select as many as two dozen pipeline incidents for more thorough investigation on the basis of the severity of the consequences, the nature of the suspected failure modes, and the incident and compliance history of the pipeline system involved. The investigations normally consist of site visits, forensic tests, interviews with operating personnel, and reviews of operator records. Since 2005, PHMSA has conducted 63 investigations of natural gas and hazardous liquid pipelines, including 14 incidents involving onshore crude oil trans- mission pipelines.4 The latter incidents are referenced in Table 4-2. In the two cases found to have involved internal corrosion, factors other than the properties of the crude oils transported were cited as causes. In three other cases, investigators reported that internal pressure cycles and associated stress loadings may have contributed to the formation and growth of cracks initiated at sites of external corrosion. Apart from providing some examples of possible failures related to the transported product, the PHMSA investigations do not provide evi- dence that pipelines transporting diluted bitumen are more susceptible to release. In the next chapter, the chemical and physical properties of diluted bitumen are examined to deduce possible susceptibilities to pipeline damage. NEB Incident Statistics NEB regulates interprovincial pipelines in Canada. The regulated net- work consists of 11,000 miles of crude oil pipeline, nearly all of which are in transmission systems. Regulated operators must file an âaccidentâ record if a pipeline facility experiences a fatal or serious injury, fire, or explosion due to a release; any other damage to the pipeline that causes a release; and any form of outside force damage, even if it does not lead to a release. In addition, operators are required to file an âincidentâ report âhttp://phmsa.dot.gov/pipeline/library/failure-reports. 4
TABLE 4-2 PHMSA Crude Oil Pipeline Incident Investigations, 2005 to 2012 Date of Commodity Attributed Failure Operator Location Released System Component Cause Summary 4/12/05 Jayhawk Stevens, Crude oil 7-in. main-line Internal Sand and saltwater collected in a low point in the Pipeline Kansas pipe section corrosion pipeline, resulting in corrosive conditions. 1/1/07 Enbridge Clark Crude oil 24-in. main-line Defect in Weld seams did not fuse during pipe manufacture. Energy County, from Canada pipe section manufacture The defect grew to a critical size by fatigue from Partners Wisconsin operating pressure cycles. 11/13/07 Enbridge Clearbrook, Crude oil 34-in. main-line Defect in Pipe was transported to the construction site on Energy Minnesota from Canada pipe section manufacture rail cars, causing fatigue cracks from cyclical load- Partners ing. Pressure cycling during operations may have caused the cracks to grow to failure. 2/18/09 Mid-Valley Cygnet, Ohio Crude oil 12-in. branch Material The combined loading of the branch connection, Pipeline connection to failure valve, and flanging caused the branch attachment to main line crack at the weld. 6/9/09 Enbridge Gowan, Crude oil 26-in. main-line Material A sleeve installed 20 years earlier to repair a pipe Energy Minnesota from Canada pipe section failure split opened at a deficient weld. Partners 12/23/09 Enterprise Galveston, Crude oil Meter station Material Cap screws on a stainless steel pressure switch Products Texas from offshore component failure in a failed because of hydrogen-assisted cracking fitting promoted by galvanic corrosion. 3/1/10 Mid-Valley Gregg Crude oil Tank farm mani- Internal Internal corrosion occurred in a dead-leg section of Pipeline County, fold piping corrosion pipe with no flow during normal operations. Texas
6/11/10 Chevron Salt Lake Crude oil 10-in. main-line Outside force An electric charge jumped from a metal fence to the Pipe Line County, Utah pipe section damage pipe, creating a 0.5-in. hole in the top of the pipe. 6/14/10 Suncor Laramie, Crude oil Breakout tank Incorrect Operating personnel did not respond to an alarm Energy Wyoming operation indicating tank capacity had been reached. Pipeline 11/16/10 Shell Vinton, Crude oil 22-in. main-line Material The coating disbonded at a bend in the pipe allow- Pipeline Louisiana from offshore pipe section failure ing the onset of corrosion. Cyclical loading due to normal batch operations may have contributed to crack growth. 12/1/10 Chevron Salt Lake Crude oil Valve used for Incorrect Water was not properly drained from the valve. Pipe Line County, Utah (condensate) water injection operation Internal pressure brought on by freezing water in main line caused the valve connection to leak. 1/26/11 Chevron Plaquemines Crude oil 10-in. main-line Excavation The pipeline was being lowered while in service. Pipe Line Parish, from offshore pipe section at damage Stress concentrations from the procedure caused Louisiana river crossing fracturing in an area with preexisting dents. 2/21/11 Enterprise Cushing, Crude oil 8-in. pipe within Incorrect Personnel purging a pipe failed to shut down the Products Oklahoma terminal area operation pump, which resulted in the delivery being pumped against a closed valve, causing a pipe with preexist- ing manufacturing defects to fail. 7/1/11 ExxonMobil Laurel, Crude oil 12-in. main-line Outside force River flooding caused debris to strike and rupture Pipeline Montana pipe section damage the line. Source: PHMSAâs pipeline failure investigation reports can be found at http://phmsa.dot.gov/pipeline/library/failure-reports.
58 Effects of Diluted Bitumen on Crude Oil Transmission Pipelines Other causes Outside 3% interference 8% Cracks Metal loss 30% 16% Material, manufacturing, Equipment or or construction component deficiency failure 22% 21% FIGURE 4-4 Causes of crude oil transmission pipeline incidents reported to NEB, 2004 to 2011. Source: TSB 2012, Table 5. in the event of an uncontrolled release, operations that exceed design limits, an abnormality that reduces structural integrity, or a shutdown for safety reasons. These reported incidents do not necessarily involve releases. From 2004 to 2011,5 NEB received 12 accident reports and 292 inci- dent reports involving crude oil transmission pipelines (TSB 2012, Table 5). Of the 292 incidents involving pipeline integrity issuesâsuch as internal and external degradationâcracks accounted for the larg- est share, almost 30 percent (see Figure 4-4). Metal loss, mainly from corrosion, was reported in 16 percent of incidents. Of the 12 accident reports, one involved combined corrosion and cracking (stress corrosion cracking), as discussed in more detail below. 5 âBefore 2004, the definition of reportable incident used by NEB was different from that used today. The reporting change makes longer-term trend analysis less meaningful.
Review of Pipeline Incident Data 59 NTSB and TSB Investigations The main transportation safety investigative bodies in the United States and Canada are NTSB and TSB, respectively. Although their pipeline investigations are thorough, they are infrequent and selective. For exam- ple, over the past decade NTSB has investigated fewer than a dozen pipe- line incidents, most involving pipelines carrying volatile commodities such as natural gas and refined products.6 The investigations are help- ful in understanding factors that can interact to cause pipeline damage and failures, but they produce limited information useful in assessing the effect of specific crude oil types or crude oil properties on pipeline release probabilities. In 2012, NTSB completed an investigation of a pipeline failure in which diluted bitumen was reported to have been released. The incident involved a 30-inch transmission pipeline that ruptured and released 20,000 barrels of product into a river near Marshall, Michigan (NTSB 2012). The investigators determined that the cause of the rupture was cracks that had formed in a corrosion pit on the outside of the pipe under a disbonded polyethylene tape coating. The cracks coalesced and grew as a result of stresses on the pipe, a process known as environmentally assisted cracking (EAC), which is described in more detail in Chapter 5. The Marshall release attracted considerable attention because of the consequences of the release and the actions of the operator. However, NTSB did not report that specific properties of the products transported through the pipeline at the time of the event or in the past had caused or contributed to the pipeline damage. As noted above, one of the 12 crude oil pipeline accidents reported to NEB since 2004 involved a corroded and cracked pipeline. This release, which occurred in 2007, was investigated by TSB.7 The release was from a 34-inch transmission pipeline originating in Alberta and transporting crude oil to the United States (TSB 2007). A forensic analysis of the rup- tured pipe joint detected a shallow corrosion pit at a weld on the outside 6 âNTSB pipeline investigation reports are available at http://www.ntsb.gov/investigations/reports_ pipeline.html. 7 âNEB may conduct its own investigations of a reported incident to ensure that safety regulations are being followed and to determine the need for remedial actions.
60 Effects of Diluted Bitumen on Crude Oil Transmission Pipelines of the pipe that led to a stress corrosion crack, which eventually spread and fractured the pipe. TSB investigators determined that the polyeth- ylene tape coating had tented over the weld, shielding the pipe from the beneficial effects of the cathodic protection current.8 The corrosion pit that developed because of the tape failure became a stress concentration site where cracks formed and grew. TSB noted that 2 years earlier the operator had converted the pipeline to batch operations and surmised that this operational change may have contributed to crack growth as a result of more cyclic stress loadings from internal pressure fluctuations. Whether specific varieties of crude oil in the stream had properties that contributed to more severe pressure cycling was not reported by TSB. A review of other NTSB and TSB investigations over the past decade did not indicate any cases in which specific crude oil types or shipment properties were associated with causes of pipeline damage or failure. Assessment of Information from Incident Reports The causes of pipeline incidents reported to PHMSA are proximate and broadly categorized. Incidents categorized as corrosion damage, for example, do not distinguish among those occurring as a result of the action of microorganisms, in combination with stress cracking, or at sites of pre- existing mechanical damage. Some types of damage, such as EAC, may be categorized alternatively as caused by corrosion, a manufacturing defect, or a material failure. Whereas NTSB and TSB investigations pro- vide detailed information on factors causing and contributing to pipeline releases, the investigations are too few in number to assess the causal effects of specific crude oil types and their properties. Because of the potentially large number of factors associated with a given release, it is often difficult to isolate the role of any single causative factor, such as the effect of the specific crude oil being transported on time-dependent mechanisms such as corrosion and cracking. Sources of pipeline damage affected by the crude oils transported, either at the time of the release or in earlier shipments, are most pertinent to this 8 âWhen the tape disbonds from the pipe steel, moisture can accumulate beneath the tape surface. Because the tape has fairly high electrical insulation properties, it can prevent cathodic protection current from reaching the exposed steel subject to corrosion.
Review of Pipeline Incident Data 61 study. Neither PHMSA nor NEB incident data contain information on the types of crude oils transported or the properties of past shipments in the affected pipeline. STATE AND PROVINCIAL INCIDENT DATA Some U.S. states and Canadian provinces maintain reporting systems for incidents in intrastate and intraprovincial pipeline systems, includ- ing gathering lines. The Energy Resources Conservation Board (ERCB) holds this responsibility in Alberta. In the United States, several state regulators have authority over intrastate pipelines, including the state fire marshal of California. Pipeline incident data and analyses derived from both of these jurisdictions were considered. Alberta ERCB Incident Data The Alberta ERCB regulates and monitors the safe performance of oil pipelines in the province, with the exception of approximately 700 miles of NEB-regulated transmission pipeline crossing into other provinces and the United States.9 ERCB mandates reporting of all pipeline inci- dents involving a release or damage from an outside force. In 2007, the agency reviewed the causes of 411 crude oil pipeline incidents reported from 1990 to 2005 (EUB 2007). The ERCB analysis showed that the larg- est single cause was internal corrosion, which the agency ascribed to the effects of the large percentage of gathering pipelines in the province. These small-diameter lines were described as susceptible to internal corrosion because of repeated low-flow conditions; frequent stopping and idling of movements; and the mixture of raw crude oil, gases, sedi- ments, and waters carried from production fields (EUB 2007, 30). About 29 percent of the roughly 11,000 miles of ERCB-regulated pipeline mile- age consisted of pipe with a diameter of 4 inches or less, and 73 percent had a diameter of 12 inches or less. Only about 1 percent of the mileage consisted of pipelines having a diameter of more than 22 inches. 9 âThe Energy and Utilities Board regulated pipelines in Alberta until it was replaced in 2008 by ERCB.
62 Effects of Diluted Bitumen on Crude Oil Transmission Pipelines Although ERCB release statistics have at times been cited as evidence of a corrosive effect of diluted bitumen on pipelines (Swift et al. 2011), the regulated systems represented by these incident statistics are not com- parable with transmission pipelines in size, operations, or, most impor- tant, contents. As a result, the committee concluded that the ERCB data were not useful for the purposes of this study. California Pipeline Safety Study Pipeline operators in California have a long history of transporting crude oils with physical properties similar to those of Canadian crude oils and diluted bitumen. Most of the oil from the San Joaquin Valley, for instance, has an American Petroleum Institute (API) gravity of 18 degrees or less, with the Kern River field producing especially dense crude oil with an API gravity of about 13 degrees (Sheridan 2006). Like bitumen producers, California oil producers commonly use thermal recovery techniques, such as injecting steam through the wellbore, to reduce crude oil viscosity and facilitate pumping to the surface. Heavier California crude oils are often transported undiluted through heated pipelines. This is not the case for Canadian bitumen, which is diluted for transportation.10 California has nearly 3,300 miles of transmission pipelines subject to federal safety regulation.11 In addition, the state contains 3,000 to 4,000 miles of state-regulated pipeline, most of it in gathering systems. Responsibility for regulating the safety of hazardous liquid pipelines in California is shared by PHMSA and the California State Fire Marshal (CSFM). In 1993, CSFM issued a report of the incident history of hazardous liquid pipelines in the state from 1981 to 1990 (CSFM 1993). The report examined releases from state and federally regulated lines, includ- ing those transporting refined petroleum products. Operators were required to submit records of releases during the period regardless of 10 âAs discussed in Chapter 2, California oil fields are served by transmission pipelines that connect to refineries elsewhere in the state. The transmission pipelines do not cross state borders. 11 âPipeline mileage by state is available at the following PHMSA website: http://primis.phmsa.dot. gov/comm/reports/safety/CA_detail1.html?nocache=9253#_OuterPanel_tab_5.
Review of Pipeline Incident Data 63 release quantity or consequences, along with information on pipeline diameter, length, age, operating temperature, and external coating type. Although the report is now 20 years old, its results have been cited as indicative of the potential effects of diluted bitumen on pipeline integ- rity (NRDC 2011). The CSFM study documented 502 releases from hazardous liquid pipelines in California during the 10-year period. Analyses of the inci- dent records indicated that external corrosion was the leading cause of releases, accounting for 59 percent, followed by third-party damage (20 percent), equipment malfunctions (5 percent), and weld failures (4 percent). Internal corrosion accounted for 3 percent, while operator error accounted for 2 percent.12 Crude oil pipelines generated 62 percent of total releases, including 70 percent of the releases attributed to exter- nal corrosion. While the CSFM study did not investigate each reported incident in depth, statistical analyses of the 502 records presented some patterns of interest. The age of the pipeline was correlated with a higher release rate. For example, 62 percent of the releases occurred in pipelines con- structed before 1950, even though these lines accounted for only 18 per- cent of pipeline mileage. CSFM noted that many of the pipelines built in California during the first half of the 20th century lacked cathodic protection for most of their service lives, which suggests that the lack of cathodic protection, coupled with the absence of coatings or use of older coating materials, may have led to the high incidence of external corro- sion relative to other failure causes.13 The CSFM analysis revealed that 22 percent of the external corrosion incidents occurred in pipelines that were uncoated, and another 53 percent occurred in pipelines coated or wrapped with certain materials, most often asphalt and tar. One finding that stood out among pipelines experiencing external corrosion was the disproportionate number of small-diameter pipelines that were operating at relatively high temperatures. Operating tem- perature was highly correlated with external corrosionâmore than half the releases from external corrosion occurred in the 21 percent of pipe- line mileage in which the operating temperature regularly reached or âAll other causes accounted for 7 percent of releases. 12 âAs is discussed in Chapter 5, some older coating technologies shield cathodic protection currents. 13
64 Effects of Diluted Bitumen on Crude Oil Transmission Pipelines exceeded 55Â°C (130Â°F). In addition, a large portion of the pipelines expe- riencing external corrosion consisted of small-diameter pipe. Although they accounted for only 13 percent of pipeline mileage, pipelines with diameters of less than 8 inches accounted for 21 percent of external cor- rosion incidents. Larger pipelines, with diameters of 16 inches or more, accounted for 23 percent of mileage but only 6 percent of the external corrosion incidents. The preponderance of external corrosion incidents in smaller- diameter pipe and pipelines with high operating temperatures does not indicate that transmission pipelines contributed to the high rate of pipe- line releases in California during the 1980s. Instead, the results suggest that older lines, many of which lacked modern coatings and cathodic pro- tection for much of their operating history, were the main source of the releases. The high operating temperatures of many of these pipelines can be attributed to the thermal recovery methods used for California crude oil production. While the California experience illustrates the problems that can arise when pipelines are not properly protected against external corrosion, it is not indicative of the protections afforded crude oil trans- mission pipelines today.14 SUMMARY A logical step in addressing the question of whether shipments of diluted bitumen have a greater propensity to cause pipeline releases than ship- ments of other crude oils is to examine historical release records. The incident statistics can be used to identify the general sources of pipe- line failure. However, the information contained in the U.S. and Cana- dian incident records is insufficient to draw definitive conclusions. One reason is that the causal categories in the databases lack the specificity needed to assess the particular ways in which transporting diluted bitu- men can affect the susceptibility of pipelines to failure. Another rea- son is that incident records do not contain information on the types of crude oil transported and the properties of past shipments in the affected pipeline. Because many pipeline releases involve cumulative and time- 14 âAll hazardous liquid transmission pipelines are required by federal regulation to have cathodic protection.
Review of Pipeline Incident Data 65 dependent damage, there is no practical way to trace the transportation history of a damaged pipeline to assess the role played by each type of crude oil and its properties in transport. Incident reporting systems in Canada and the United States do not have uniform reporting criteria and coverage. Given the relatively small number of pipeline incidents, even minor variations in reporting criteria can lead to significant differences in incident frequencies and causal pat- terns. Some reporting systems combine incident reports from oil gather- ing and transmission systems, while others do not. Variation in reporting coverage is problematic because gathering pipelines are fundamentally different from transmission pipelines in design, maintenance, and oper- ations and in the quality and quantity of the liquids they carry. REFERENCES Abbreviations CSFM California State Fire Marshal EUB Energy and Utilities Board NRDC Natural Resources Defense Council NTSB National Transportation Safety Board TSB Transportation Safety Board of Canada CSFM. 1993. Hazardous Liquid Pipeline Risk Assessment. Sacramento, Calif. http:// osfm.fire.ca.gov/pipeline/pdf/publication/pipelineriskassessment.pdf. EUB. 2007. Pipeline Performance in Alberta, 1990â2005. Report 2007-A. Alberta, Canada, April. http://www.ercb.ca/reports/r2007-A.pdf. NRDC. 2011. Say No to Tar Sands Pipeline: Proposed Keystone XL Project Would Deliver Dirty Fuel at a High Cost. Washington, D.C., March. http://www.nrdc. org/land/files/TarSandsPipeline4pgr.pdf. NTSB. 2012. Enbridge Incorporated Hazardous Liquid Pipeline Rupture and Release, Marshall, Michigan, July 25, 2010. Report NTSB/PAR-12/01. Washington, D.C. http://www.ntsb.gov/doclib/reports/2012/PAR1201.pdf. Sheridan, M. 2006. California Crude Oil Production and Imports. Staff Paper CEC- 600-2006-006. Fossil Fuels Office, Fuels and Transportation Division, California Energy Commission, April. http://www.energy.ca.gov/2006publications/CEC- 600-2006-006.pdf. Swift, A., S. Casey-Lefkowitz, and E. Shope. 2011. Tar Sands Pipelines Safety Risks. Natural Resources Defense Council, National Wildlife Federation, Pipeline
66 Effects of Diluted Bitumen on Crude Oil Transmission Pipelines Safety Trust, and Sierra Club, Washington, D.C. http://www.nrdc.org/energy/ files/tarsandssafetyrisks.pdf. TSB. 2007. Pipeline Investigation Report: Crude Oil Pipeline Rupture, Enbridge Pipelines, Inc., Line 3, Mile Post 506.2217, near Glenavon, Saskatchewan, 15Â April 2007. Report P07H0014. http://www.bst-tsb.gc.ca/eng/rapports-reports/ pipeline/2007/p07h0014/p07h0014.pdf. TSB. 2012. Statistical Summary, Pipeline Occurrences 2011. Gatineau, Quebec, Canada. http://www.tsb.gc.ca/eng/stats/pipeline/2011/ss11.pdf. U.S. Department of State. 2013. Draft Supplementary Environmental Impact State- ment for the Keystone XL Project Applicant for Presidential Permit: TransCanada Keystone Pipeline, LP. Bureau of Oceans and International Environmental and Scientific Affairs, Washington, D.C. http://keystonepipeline-xl.state.gov/draftseis/ index.htm.