By Susan A. Olenchuk, James B. Curry, and Katie E. Leesman
On Sept. 26, 2011, the National Transportation Safety Board (Board or NTSB) released its final report on the Sept. 9, 2010 San Bruno, CA pipeline explosion. The incident occurred on a 30-inch diameter intrastate gas transmission line (Line 132) operated by Pacific Gas & Electric Co. (PG&E). The explosion and fire destroyed or damaged 70 homes and killed 8 people.
The NTSB found that the San Bruno accident was caused by numerous deficiencies, including three key lapses. First, the Board found that PG&E’s oversight of the Line 132 relocation in 1956 was inadequate. Consequently, PG&E installed pipe that did not conform to any known standards and which contained a long seam weld flaw that would have been visible during construction. This flaw grew over time, causing the rupture when electrical work at an upstream terminal caused a modest pressure increase in the pipeline.
Second, the Board found that PG&E’s integrity management program was inadequate and failed to detect or repair the defective pipe. Third, the NTSB found that the regulations and oversight of the California Public Utilities Commission (CPUC) and the Pipeline and Hazardous Materials Safety Administration (PHMSA) were contributing factors because regulations exempt older pipelines from hydrostatic testing requirements and because neither agency detected the inadequacies in PG&E’s integrity management program.
The NTSB issued several safety recommendations to PG&E, the CPUC, PHMSA, the U.S. Department of Transportation (DOT), and pipeline industry trade groups. The NTSB’s recommendations are not mandatory, but they carry substantial weight with regulators, Congress, and industry, and could influence pipeline safety regulatory policy, particularly with respect to regulatory exemptions for pipe constructed before 1970, integrity management implementation and oversight, and the PHMSA-state relationship.
Construction Defects and the Grandfather Clause
The NTSB found that a latent construction defect on Line 132 was a primary cause of the San Bruno incident, and that multiple opportunities to discover and remedy this defect were missed in the decades between construction and the incident. For example, because pipelines constructed before adoption of federal safety rules in 1970 and California’s state safety rules in 1961 are exempt (grandfathered) from post-construction hydrostatic pressure testing requirements to establish maximum allowable operating pressure (MAOP), Line 132 did not undergo a hydrotest that would likely have exposed a defective long seam weld. The Board further noted that the grandfather clause results in reduced safety margins because pipeline MAOP is the basis for important decision making driven by other pipeline regulations.
Finding no safety justification for the exemption, the NTSB recommended that PHMSA eliminate the grandfather clause from its regulations and require that all gas transmission pipelines constructed before 1970 be subjected to a hydrostatic pressure test that incorporates a spike test.(1)
Legislation proposed in the 112th Congress(2) and a PHMSA Advanced Notice of Proposed Rulemaking (ANPRM)(3) initiated this year both anticipated this NTSB recommendation. Proposed legislation would effectively eliminate the grandfather clause by requiring pipeline operators to verify MAOP on many gas transmission lines, and PHMSA’s ANPRM solicits public input on whether the grandfather clause should be explicitly removed.
Eliminating or limiting the grandfather clause could have a significant impact on industry because, according to PHMSA, 61% of onshore natural gas transmission lines were constructed before 1970. If the provision is eliminated, the pipeline industry could be required to hydrotest thousands of miles of pipe, potentially resulting in pressure reductions, service outages, environmental issues associated with hydrotest water disposal and significant costs.
The NTSB found significant flaws in all elements of PG&E’s integrity management (IM) program and in the oversight and enforcement practices of the CPUC and PHMSA. IM regulations require that operators apply additional protective measures to segments of pipeline located in high consequence areas where people live or congregate. An effective IM program requires accurate data, an assessment of the threats to the pipeline and the risks posed by those threats, a process for evaluating threats and remediating and repairing defects, and a procedure for measuring the effectiveness of the IM program itself. The section of Line 132 that ruptured is subject to the IM regulations.
The Board found that data used to support PG&E’s IM program was inaccurate, and no steps were taken to improve data quality; the magnitude of various pipeline threats did not correspond to actual data; the IM program did not consider known long seam cracks and at least one past seam leak; assessment tools that could actually detect seam defects were not selected; known seam defects were improperly considered stable; and IM program self-measurements were not meaningful.
For example, the Board found that PG&E’s records contained obvious errors in key pipeline parameters, including long seam type, specified minimum yield strength (SMYS), and depth of cover. The NTSB also found that PG&E used insufficiently conservative assumptions where data was unknown. On one pipeline segment, PG&E assumed a SMYS of 52,000 psi, contrary to regulations requiring operators to use a value of 24,000 psi when SMYS is unknown.
The Board also found that PG&E missed opportunities to correct pipeline data in the course of normal information collection activities. Records for many pipeline segments had not been updated with data collected when lines were exposed during external corrosion direct assessment (ECDA) excavations.
The NTSB discovered that several known defects on Line 132, such as longitudinal seam weld cracks found during radiography of the girth welds during the 1948 construction, were documented in PG&E’s records but not considered in its IM program. Also, PG&E used only one method, ECDA, to assess all 322 segments of Line 132 even though ECDA is incapable of detecting construction defects, including seam flaws.
The Board found that, although the relative weights used by PG&E to assess total risk were generally consistent with industry averages, the weights were not consistent with PG&E’s leak, failure and incident experience. In particular, the NTSB found that PG&E’s IM program significantly understated threats due to external corrosion and design and materials defects, and overstated the threats due to third-party damage and ground movement.
As a result of these and other findings, the NTSB recommended that PG&E perform a comprehensive assessment and revision of its IM program. The Board also recommended broad changes to PHMSA's oversight of compliance with IM regulations, in particular, that PHMSA verify the completeness and accuracy of operator information on pipeline attributes and require operators to have procedures to ensure the accuracy and completeness of such information; develop clear metrics for IM program performance and require operators to regularly measure their programs against those metrics; audit operator performance measures against the pipeline’s risk model, presumably to ensure the model is appropriate; and set performance goals for operators at each inspection and follow up in subsequent inspections.
If adopted, the Board’s recommendations could increase the scrutiny of IM programs and result in a shift in PHMSA’s role. Involving the agency more deeply in integrity decision making could reduce operator discretion. In particular, reviewing performance measures against the pipeline risk model places PHMSA in the role of judging the inner workings of operator risk models, algorithms and engineering justifications. It is unclear whether PHMSA has the framework in place to provide inspection personnel with information on operator systems at a level of detail sufficient to engage in such a review.
The NTSB recommendation that PHMSA require that pipelines verify that they are relying on accurate and complete information also does not appear consistent with a tenant of IM that allows operators to use conservative assumptions when data is missing or inaccurate. This regulatory concept reflects the reality that many pipeline systems are decades old and have changed owners multiple times, and that operators may not have complete original design and construction documentation.
Many of the NTSB’s recommendations have parallels in PHMSA’s recent ANPRM, which sets forth numerous questions concerning whether more stringent requirements for data collection and validation, risk modeling, assessment tool selection and IM program self-assessment are appropriate.
Regulatory Oversight and the Federal-State Relationship
PHMSA has the authority to regulate both interstate and intrastate pipeline facilities, but states may opt to regulate intrastate facilities pursuant to an annual certification to PHMSA, certifying among other things, that the state has jurisdiction, has adopted PHMSA’s pipeline safety standards, and is enforcing each standard. If a state gas pipeline safety program is operating under an annual certification, PHMSA is prohibited from prescribing or enforcing standards for intrastate facilities in that state. In California, the CPUC regulates intrastate pipelines, including PG&E’s Line 132, pursuant to such a certification.
The NTSB found that the CPUC did not 1) uncover PG&E’s inadequate IM program and other safety weaknesses within the company and 2) historically failed to take meaningful enforcement action. Finding that PHMSA did not recognize the CPUC’s weak oversight, the NTSB questioned the process used to certify state pipeline safety programs. For example, the NTSB found that a joint audit conducted by PHMSA and the CPUC in 2005 did not uncover “systemic problems” in PG&E’s IM program.
Therefore, the NTSB recommended, among other things, that DOT audit PHMSA’s enforcement policies and procedures, including its oversight of performance based regulations like IM, as well as PHMSA’s state pipeline safety program certification program. The Board further recommended that DOT ensure that the certification program is amended, as necessary. These tough recommendations suggest the potential for significant change in the federal-state pipeline safety enforcement relationship.
PHMSA relies on its state partners to enforce and oversee intrastate pipeline safety compliance. In turn, the states depend on PHMSA to fund substantial portions of their pipeline safety programs and - under the current framework - funding levels depend in large part on state program performance. Raising the bar for states could result in reduced funding for some states and could threaten effective pipeline safety oversight, at least in the short term, as PHMSA itself lacks the resources to backstop state enforcement.
Taken together, the NTSB recommendations and the potential legislative and regulatory changes under consideration by Congress and PHMSA could result in a significant increase in the federal government’s pipeline safety oversight. It bears watching how PHMSA and Congress consider and incorporate the NTSB recommendations into the ongoing pipeline safety debate.
1. The NTSB noted that research demonstrated that a spike test can identify certain defects that may not be revealed by a hydrotest alone, but which may fail when the pipeline is returned to service after the hydrotest through a phenomenon known as pressure reversal.
2. At this writing, three main pipeline bills are pending in the Senate and House of Representatives. On July 7, 2011, the Senate Commerce, Science & Transportation Committee reported S.275; on Sept. 8, 2011, the House Transportation and Infrastructure Committee reported H.R. 2845; and on Sept. 21, 2011, the House Energy & Commerce Committee reported H.R. 2937.
3. Pipeline Safety: Safety of Gas Transmission Pipelines, 76 Fed. Reg. 53,086 (Aug. 25, 2011).