Jim Hall, Chairman
National Transportation Safety Board
Committee on Transportation and Infrastructure
Subcommittee on Economic Development, Public Buildings, Hazardous Materials, and Pipeline Transportation
House of Representatives
July 27, 1999
Mr. Chairman, I want to thank you and the Committee for inviting the National Transportation Safety Board to comment on pipeline safety issues and the performance of the Research and Special Programs Administration's Office of Pipeline Safety.
RSPA should play a crucial role in pipeline safety as the federal regulators in this area. It is their mission to protect the American people and the environment. They have failed in that mission - primarily because they have failed to respond aggressively to Safety Board pipeline recommendations. At 68.9 percent, RSPA has the lowest acceptance rate of any modal administration in the Department of Transportation. That statistic represents RSPA's failure to take measures which I believe would prevent future accidents, save lives, and fulfill their stated mission.
There are too many examples of this regulatory inaction. I believe that the accidents that I'll discuss today could have been prevented or their consequences reduced - if our recommendations had been accepted. The most recent example is the pipeline rupture in Bellingham, Washington, on June 10, 1999, that released approximately 250,000 gallons of gasoline. The gasoline flowed into Whatcom Creek which runs through a Whatcom Falls Park, a recreational area. The gas then ignited, resulting in three deaths, and substantial property and environmental damage.
Although our investigation is in its early stages, the long-standing safety issues which I'll also discuss today appear to be relevant to the Bellingham accident, based on our early investigation, and have been previously addressed in other NTSB accident investigations.
The first issue I want to address is pipeline integrity. In 1987, as a result of investigations into three pipeline accidents, the Safety Board recommended that RSPA require pipeline operators to periodically determine the adequacy of their pipelines to operate by performing inspections or tests capable of identifying, corrosion, mechanical damage, or other time-dependent defects that could be detrimental to the safe operation of pipelines.
Nine years later, in 1996, nearly a million gallons of fuel oil were released into the Reedy River near Fork Shoals, South Carolina when a corroded section of pipe ruptured. That same year, almost 500,000 gallons of gasoline were released into marshland and the Blind River near Gramercy, Louisiana when a damaged section of pipeline ruptured.
In the Bellingham accident, our investigators found indications of previous external damage that may have weakened the pipeline near the rupture. Yet, 12 years after our initial recommendation, no regulations require pipeline operators to perform periodic inspections or tests to locate and assess whether the type of damage that was found in the accidents I just mentioned, exists on other pipelines.
The second area I want to focus on is the lack of training for pipeline personnel. In 1987, after several accidents in which inadequate training of pipeline personnel was an issue, the Safety Board recommended that RSPA require operators to develop training procedures for their employees. Because of RSPA's inaction over the intervening years, inadequate training continues to be a factor in pipeline accidents. Three 1996 accidents provide ample evidence of the problem.
Following the San Juan accident, in January 1998, the Safety Board recommended that RSPA complete a final rule on employee qualification, training, and testing within one year. That October, RSPA published a Notice of Proposed Rulemaking (NPRM) to require pipeline operators to develop a written qualification program for individuals operating pipelines. However, the NPRM fails to establish any training requirements; rather, it allows companies to evaluate performance using methods such as oral examinations and on-the-job observation. We are currently awaiting RSPA's final rule.
In the Bellingham accident, preliminary indications are that the pipeline controllers continued to operate the pipeline after it ruptured. We are in the process of determining what the circumstances were and how the controllers were trained to recognize and handle emergency conditions. However, we do know that there are still no mandatory training requirements. This is unacceptable. No segment of the transportation system should have untrained personnel operating safety-sensitive equipment - placing themselves and others at risk.
The third issue I want to address is one that the NTSB first identified some 29 years ago - the need for rapid shutdown of failed pipelines in order to limit the release of product following a pipeline rupture. The increased use of automatic and remotely controlled valves to reduce the consequences of pipeline failures was discussed in our 1970 study, Effects of Delay in Shutting Down Failed Pipeline Systems and Methods of Providing Rapid Shutdown. Since then, there have been a number of other accidents which highlight the need to reduce the release of hazardous gases or liquids.
In 1986, in Mounds View, Minnesota, gasoline spewed from a pipeline and flowed down a city street before igniting and seriously burning three people, two of whom later died. The Safety Board found that the pipeline operator could not promptly stop the release of gasoline, which greatly contributed to the magnitude of the loss suffered by the community. In 1994, in Edison, New Jersey, a high-pressure natural gas pipeline exploded; a massive fire ensued which then ignited several building roofs in an apartment complex. The Safety Board again found that the inability of the pipeline operator to promptly stop the flow of natural gas contributed to the devastation. The following year, the Safety Board recommended that RSPA expedite requirements for rapid shutdown of failed pipeline segments on high-pressure pipelines in urban and environmentally sensitive areas. RSPA continues to study potential technological applications to rapidly shut down pipelines; however, it still does not require these systems. Despite RSPA's inaction, several of the more progressive pipeline companies have voluntarily installed shut-off valves.
Following the Bellingham accident, RSPA ordered the pipeline company to install an automatic check valve just downstream of the rupture location so that the volume of product released would be limited in the event of a future pipeline rupture in that area. A case of too little - too late.
Finally, there is one area in which RSPA has responded to Congressional action - excavation damage prevention. As you may know, Mr. Chairman, excavation damage is the leading cause of pipeline accidents. And, excavation damage prevention is on the Safety Board's Most Wanted list. Just last month, at the request of Congress, RSPA held a joint symposium on excavation damage with the Safety Board and sent a report to Congress on best practices for preventing damage to underground facilities.
However, RSPA still has more work to do in order to reduce the number of excavation damage accidents. At Bellingham, our investigators found several indicators of external damage in the vicinity of the rupture. That segment of the pipeline is now in our laboratory and we are closely examining it to determine what part excavation damage may have played in the rupture.
Mr. Chairman, let me close by saying that the RSPA Administrator has promised to be more proactive and to improve the Office of Pipeline Safety's responsiveness to our safety recommendations. However, as I've outlined, RSPA still has much to do. They must take more aggressive action to protect public safety and our environment.
Mr. Chairman, this concludes my testimony. I would be happy to answer the Committee's questions.
for the record:
Good morning Mr. Chairman and Members of the Committee. I am pleased to represent the National Transportation Safety Board before you today to discuss pipeline safety issues. Because of Board activity over the past ten days, I have not had an opportunity to share this statement with the Board's other members and, therefore, the comments below do not have Board concurrence.
Before I begin, I would like to update the Committee on the status of the Safety Board's investigation of the pipeline rupture that occurred June 10, 1999, in Bellingham, Washington, that resulted in the release of approximately 250 thousand gallons of gasoline. The released gasoline flowed down a creek and ignited, resulting in three deaths as well as property and environmental damage. Safety Board investigators were on-site for over a month because of several difficulties. Exposing and conducting an initial examination of the failed segment of pipe was complicated, because gasoline continued to seep from the failed pipeline and fueled lingering fires at the rupture site. In addition, the failed segment of pipe was within a water treatment plant yard.
The gasoline pipeline crossed directly underneath major water pipelines that supplied water to portions of the city and its surrounding areas. Because of the need by pipeline company and city officials to construct a temporary pumping station to bypass the damaged pumping station and allow the water pipelines to be excavated, excavation of the ruptured pipeline was delayed. Further, the ability to gather information from key pipeline company personnel was complicated because they declined be interviewed by Board investigators. We have, however, collected a large amount of information, including the electronic data from the pipeline company's operating system, that will help us assess the conditions leading up to the accident.
The investigation is still in its early stages; however, we will closely examine the failed segment of the pipeline, the design and operation of the pipeline, the adequacy of pipeline company and federal inspection procedures, as well as the actions and training of the pipeline controllers.
I would now like to discuss general pipeline safety issues. As the Federal regulatory oversight agency for pipeline safety, the Research and Special Programs Administration (RSPA) plays a crucial role. It is the Board's view, however, that RSPA has not responded as aggressively as we and the American people would expect. RSPA's implementation rate of pipeline safety recommendations is 68.9 percent, the lowest acceptance rate of any modal administration in the Department of Transportation. We do not think this low percentage is a result of ill-conceived recommendations. In fact, the acceptance rate of our pipeline safety recommendations issued to the pipeline community as a whole is 86.9 percent.
RSPA's acceptance rate of Safety Board recommendations also reflects the tenuous relationship between our two agencies over the years. In an April 14, 1998, letter to Secretary Rodney E. Slater I stated: "... I am ... troubled by OPS' lack of concern and responsiveness to open pipeline safety recommendations issued by the Safety Board. OPS had not provided any written update on actions taken on some of these recommendations since 1992. In October 1997, the NTSB requested such an update on 28 recommendations in preparation for upcoming investigations reports. As of this date, we have received updates on only 7 of these recommendations."
We believe that RSPA's lack of action continues to place the American people at risk. Ms. Kelley Coyner, the new RSPA Administrator, has met individually with our Board members and has made a commitment to improve RSPA's response rate to Safety Board safety recommendations. As a result, we have seen improvement in some areas. However, we are still concerned about the lack of timely action on some much needed safety improvements, and we feel the areas listed below need improvement:
The continued operation of pipelines with integrity problems is a recurring issue in accidents investigated by the Safety Board. There are over 1.7 million miles of natural gas pipelines, and over 165,000 miles of liquid pipelines crisscrossing this country. A mechanism needs to be in place to find problems with these pipelines before defects can grow to a critical size and result in catastrophic failure.
In 1987, as a result of investigations into three pipeline accidents (in Beaumont, Kentucky; Lancaster, Kentucky; and Mounds View, Minnesota), the Safety Board recommended that RSPA require that pipeline operators periodically determine the capability of their pipelines to safely operate by performing inspections or tests capable of identifying corrosion, mechanical damage, or other time-dependent defects that could be detrimental to the safe operation of pipelines. Since the Safety Board recommended this action, RSPA has been studying the issue, but has yet to reach any conclusions. Due to the length of time that has passed without final RSPA action, the Safety Board in June of this year classified its recommendation as "Open-Unacceptable Response."
In 1996, nearly a million gallons of fuel oil were released into the Reedy River near Fork Shoals, South Carolina, when a corroded section of pipe ruptured. Also in 1996, almost 500 thousand gallons of gasoline were released into marsh land and the Blind River near Gramercy, Louisiana, when a damaged section of pipeline ruptured. Both of these failures occurred at time-dependent damage locations.
In addition, the Safety Board is currently investigating two other pipeline accidents with potential pipeline integrity problems that occurred this year. One is the Bellingham, Washington, pipeline accident where we found indications of previous external mechanical damage in the vicinity of the rupture. The other occurred in February in Knoxville, Tennessee. Approximately 45,000 gallons of diesel fuel were released into the Tennessee River. In the Knoxville accident, we are also studying the effects of corrosion and metal fatigue on older pipe.
Training Of Pipeline Personnel
The Safety Board has long been concerned about the need to adequately train personnel in all transportation modes, including pipeline. In 1987, after several pipeline accidents in which inadequate training was an issue, the Safety Board recommended that RSPA require operators to develop training programs for pipeline personnel. After 11 years had passed since the recommendation was issued without final action, the Safety Board classified the recommendation as "Closed-Unacceptable Action."
However, inadequate training continues to be a factor in pipeline accidents. In the 1996 Fork Shoals, South Carolina, pipeline accident, the Safety Board found that pipeline controllers had been inadequately trained to recognize and handle emergency conditions. In that accident, the controller mistakenly shut down a pump station, failed to recognize his mistake, and continued to operate the pipeline after it ruptured. As mentioned earlier, this action resulted in the release of nearly one million gallons of fuel oil into the Reedy River.
On November 21, 1996, a pipeline accident in San Juan, Puerto Rico, resulted in 33 fatalities and 69 injuries. Our investigation determined that the gas company's employees were not properly trained to survey, pinpoint, or test for pipeline leaks, and failed to locate a reported leak before the explosion occurred. In January 1998, the Safety Board recommended that RSPA complete a final rule on employee qualification, training, and testing within one year.
In October 1998, RSPA published a Notice of Proposed Rulemaking (NPRM) to require pipeline operators to develop a written qualification program for individuals operating pipelines. Although the Safety Board was told that the rule would meet the intent of our recommendation, it does not. The NPRM does not establish training requirements for personnel. Rather, it allows a company to evaluate an individual's ability to perform tasks using such methods such as oral examinations, or observations of on-the-job performance.
As you are aware, Mr. Chairman, observation of on-the-job performance is a routine supervisory function. The Safety Board believes that strong training and testing requirements are needed to ensure that employees can properly perform their tasks. Tests must be administered in conjunction with training so that an objective assessment can be made of the training's success. In January 1999, the Safety Board provided comments to RSPA on this rulemaking and again urged RSPA to amend its final rule to require that individuals be trained, that they be tested to assess the success of the training, and that they be periodically retrained and retested. In February 1999, the Safety Board classified its recommendation as "Open-Unacceptable Action," because the NPRM does not require adequate training or testing. At this point, we are still awaiting RSPA's final rule.
The third area of concern I would like to discuss is the lack of adequate requirements for corrosion protection on pipelines.
The Safety Board investigated a pipeline accident that occurred in Lively, Texas, on August 24, 1996, that sent a butane vapor cloud into a residential area. The resulting fire killed two residents. The Safety Board concluded that the pipeline was inadequately protected from corrosion. In addition, the Safety Board identified weaknesses in federal regulations concerning corrosion protection, and in November 1998, we recommended that RSPA strengthen these requirements. For example, the Board recommended that RSPA provide performance measures so that one would know when an acceptable level of corrosion protection exists. Based on information provided in meetings by RSPA staff, we are encouraged that RSPA will soon take action on this issue.
The fourth area I would like to discuss is the need to limit the release of product into the environment following a pipeline rupture. The increased use of valve automation to protect public safety and the environment by reducing the consequences of pipeline failures has been a long-standing concern of the Safety Board. We first addressed this issue 29 years ago in a study entitled Effects of Delay in Shutting Down Failed Pipeline Systems and Methods of Providing Rapid Shutdown.
Since then, there have been a number of additional accidents which have highlighted the need to reduce the release of product. On July 8, 1986, in Mounds View, Minnesota, gasoline spewed from a pipeline and flowed down a city street before igniting and seriously burning three people, two of whom later died. The Safety Board found that the pipeline operator could not promptly stop the release of gasoline.
On March 23, 1994, in Edison, New Jersey, a high-pressure natural gas pipeline exploded and a fire ensued. Heat from a fire then ignited several building roofs in an apartment complex. The Safety Board again found that the inability of the pipeline operator to promptly stop the flow of natural gas contributed to the severity of the accident. In February 1995, the Safety Board recommended that RSPA expedite requirements for rapid shutdown of failed pipeline segments on high-pressure pipelines in urban and environmentally sensitive areas. RSPA held a public workshop on this subject later in 1995, and they continue to study this issue. Although RSPA still does not require these systems, we are pleased that several pipeline companies have voluntarily put in valves that allow them to rapidly shut-down failed pipelines.
In an accident that occurred in May 1996 near Gramercy, Louisiana, it took the pipeline company approximately 4½ hours to manually close the valves on either side of a ruptured pipeline. Almost 500 thousand gallons of gasoline were ultimately released into the environment. In September 1998, the Board recommended that the pipeline operator evaluate and install a higher degree of valve automation into its pipeline system. The pipeline operator has advised the Safety Board that it is using risk management principles to evaluate existing valves to automate. The company also plans to install this technology into a new pipeline that may run from Kenova, West Virginia, to Columbus, Ohio.
Mr. Chairman, this technology is available and is obviously being used. But we should not have to rely on the industry's altruism. RSPA needs to finally act to require the installation of these systems to limit the release of product from major pipeline ruptures before the next accident, the next environmental release, and the next death occurs.
Excavation Damage Prevention
As you may know, excavation damage is the leading cause of pipeline accidents. This issue was added to the Safety Board's "Most Wanted" list of transportation issues in 1997, and in December 1997, we published a study entitled Protecting Public Safety Through Excavation Damage Prevention. As a result, the Board issued 26 recommendations aimed at improving excavation damage prevention covering such areas as:
RSPA has taken some steps in excavation damage prevention. At Congress' direction, in June RSPA held a joint symposium on excavation damage with the Safety Board. In addition, it has forwarded to Congress a report on best practices for preventing damage to underground facilities. It is our understanding that RSPA will use the best practices to evaluate State damage prevention programs.
Let me close by saying that we are encouraged by commitments made by the new RSPA Administrator. She has advised she will be more proactive and will improve the Office of Pipeline Safety's responsiveness to our safety recommendations, and we look forward to better communications with RSPA regarding the Board's recommendations.
Mr. Chairman, this concludes my testimony. I would be happy to answer the Committee's questions.