The San Bruno gas pipeline explosion of 2010 brought unprecedented scrutiny to the state’s network of natural gas pipelines.
Now, two breaks in an East Bay pipeline have sparked questions about whether the state office charged with overseeing oil pipeline safety is up to the job.
KQED has learned that the state investigation into the rupture of a 24-inch Shell Oil Co. pipeline on the eastern edge of the Bay Area in September 2015 relied heavily on the energy company’s own analysis of the incident — some critics say too heavily.
The Office of the State Fire Marshal, a division of Cal Fire charged with investigating oil pipeline incidents inside the state, never took possession of the portion of the line that broke and did not conduct its own inspection of the line after it ruptured.
In assessing the rupture’s cause, the fire marshal’s report merely quoted a three-sentence summary provided by a firm Shell hired to look into the incident. The fire marshal’s review failed to document when, where or by whom the pipe was manufactured or how it got to the West Coast decades ago — a trip that could have weakened the line.
The fire marshal’s office says the company did not violate state regulations in connection with the rupture and did not levy any penalties.
Eight months later, in May 2016, the Shell pipeline broke again and spilled crude oil in the same area, just off Interstate 580 near Tracy.
Although state reports say the spills were cleaned up and “no water was affected,” each rupture along the company’s San Pablo Bay Pipeline, which brings crude oil from the Central Valley to the Bay Area, led to tens of thousands of gallons of crude to spill onto ranch land in the Altamont Pass, alarming environmentalists and prompting concern from at least one state lawmaker.
KQED obtained Shell’s analysis of the break and the fire marshal’s report on the incident through a California Public Records Act request, to learn more about the investigation into the Sept. 16, 2015, spill.
Pipeline experts, safety advocates and environmentalists who have reviewed those records say they reveal weaknesses in the way many crude pipelines in California are regulated. They also raise questions about whether the state fire marshal’s office is equipped to oversee this part of the oil industry.
Cal Fire says 752 intrastate pipelines are under its jurisdiction. One of the nation’s leading experts on pipeline safety says that should change.
“Cal Fire does not have the appropriate tools to regulate and inspect something as sophisticated and intricate as a pipeline,” says Robert Bea, professor emeritus at the Center for Catastrophic Risk Management at UC Berkeley’s College of Engineering. “For us to turn to our heroes in Cal Fire and expect them to do this is not fair.”
Bea says the California Public Utilities Commission and the federal Pipeline and Hazardous Materials Safety Administration are better suited for such investigations.
“The understanding of cracking threats and what are the appropriate assessment methods may not be well understood by state regulators,” says Richard Kuprewicz, president of Accufacts Inc., a Redmond, Washington-based firm that specializes in pipeline investigations. He notes that part of the cause of the September break was a small crack in the line that expanded over time.
“The California fire marshal used to be considered one of the better state regulators in the country, and then they went through some funding issues,” says Carl Weimer, executive director of the Pipeline Safety Trust, a Bellingham, Washington-based nonprofit.
“They actually backed away from a lot of their regulatory authority,” Weimer says.
“Cal Fire seems to almost have completely outsourced the investigation into this very serious incident to Shell itself,” adds Patrick Sullivan, the climate media director with the activist Center for Biological Diversity.
“We have a pipeline here that has failed again and again,” Sullivan says. “The state ought to be taking a much harder look and do its own hands-on investigation to figure out what’s really going on here. Cal Fire doesn’t have enough staff, it doesn’t have enough resources and it does not have the will to really keep a close eye on pipelines.”
The state’s director of pipeline safety acknowledged after a major crude oil pipeline rupture last year near Santa Barbara that the agency has been working with a minimum staff for years — but Cal Fire emphasizes that its fire marshal’s office followed state regulations in investigating last September’s Altamont spill and note that Shell commissioned a well-respected third party to investigate the break.
Cal Fire argues that federal law calls for pipeline operators to analyze the cause of accidents involving their lines.
But according to the fire marshal’s website, the office’s Pipeline Safety Division “is also responsible for the investigation of all spills, ruptures, fires, or pipeline incidents for cause and determination of probable violations.”
Cal Fire spokeswoman Lynne Tolmachoff says because the September spill uncovered no violations, there was no need for state regulators to conduct their own inspection of the pipe afterward.
‘Cause for Alarm’
The two spills took place at the border between Alameda and San Joaquin counties on a pipeline that stretches from Coalinga in Fresno County to Martinez.
After learning of the two breaks, San Joaquin County Assemblywoman Susan Eggman met with Shell executives.
“I think it’s really cause for alarm up and down that pipeline,” Eggman said following the meeting.
The fire marshal’s office is still investigating the most recent break, which occurred on May 20. That rupture spilled what Shell estimates to be at least 20,000 gallons of crude.
Initially, Shell said the September break spilled the same amount. Shell spokesman Ray Fisher wrote KQED in an email in late May that 500 barrels had spilled, a volume equivalent to 21,000 gallons. But Shell’s own investigative documents show that 900 barrels, or close to 38,000 gallons, spilled.
It’s common for oil companies to provide lower estimates on the amount of fuel lost in a pipeline break, according to Kuprewicz.
“When the number turns out to be way different than what you originally said without a good justifiable reason, credibility starts to get lost on all sides,” he says. “That’s a problem across the country.”
Fisher said the initial figure was based on engineering and field data in September. He said the company updated the estimate in its report to federal regulators in October.
The emergency response, cleanup and pipeline repairs stemming from the September spill cost $1.2 million, according to Shell.
Pipeline Had Safety Issues
After the first rupture the energy company hired a risk management firm, Det Norske Veritas (DNV), to investigate. According to DNV and Shell’s documents, the September and May spills were not the only dangerous incidents associated with the pipeline.
In 1998, less than a decade after the pipe was installed, an unspecified defect in the pipe caused a failure on the line. Forty feet of the pipe had to be replaced. Shell says it’s unable to find more information about the incident.
“Records were basic and did not have modern level of report detail,” Shell’s analysis said.
The 1998 incident, along with a spill the following year, prompted Cal Fire to place the pipeline on its list of “high-risk” lines, a classification that requires a different level of inspections. For the next several years, crews found possible dents, corrosion and grind marks on the pipeline.
Shell’s Fisher calls those kinds of observations normal and says the company’s willingness to point them out represents “proactive steps to preserve the integrity and reliability of the pipeline.”
In July 2004, the state removed the line from its list of high-risk pipes.
In May 2015, pumps connected to the pipeline were blocked, resulting in a fire. While Shell said that the fire was a significant event leading up to last September’s spill, company officials at the time determined that the fire “did not likely affect the pipeline from a heat or pressure perspective.”
An Industrywide Problem
According to Shell, the September 2015 rupture began at a pre-existing fatigue crack that was 36 inches long and 6 inches wide. A “longitudinal seam failure” grew over time in a corroded section of pipe and eventually ruptured after line pressure fluctuated to accommodate different grades of oil.
“The pipeline has been aggressively cycled in the last year of operation,” Shell’s report said.
Bea says the kind of cracking involved in the rupture is a common cause for pipeline breaks. “It is an industrywide problem,” he says.
“They are moving very heavy crude through this pipeline at very high temperatures and they’re doing it at very high pressures, and they’re alternating those pressures so they’re putting a ton of stress on this pipeline,” adds CBD’s Sullivan.
Shell says the needs of the market are responsible, in part, for the pressure cycling.
“Ideally, a pipeline would operate at a constant pressure,” Fisher wrote in an email. “However, customer demands, especially on common carrier pipelines, dictate loads on the pipeline. These resulting variations in load requirements result in pressure cycling.”
The September spill led Shell to take “steps to reduce the pressure cycling of this pipeline,” Fisher wrote.
Just over a month before the September spill, Shell crews conducted in-line inspections of the pipe and reported that there were no problems.
Shell’s analysis showed that the rupture took place four minutes after crews raised pressure on the pipeline in the middle of the night. More than eight hours later, the company confirmed that the rupture caused a spill with crude oil being released onto pasture land.
Holes in Pipe’s History
DNV’s report speculates that the crack could have developed from in-service operation or when the pipe was transported to the West Coast after it was manufactured, which Shell now says — contrary to its initial report — was in 1982.
In response to KQED inquiries, Shell provided several other details of the pipe’s history omitted from the report provided to — and relied upon by — the Office of the State Fire Marshal.
The original document said the pipe manufacturer was unknown, for instance. Company spokesman Ray Fisher said in a later email that the pipe was made by Houston’s Armco Steel Corp.
From there, Fisher and the company’s report says, the pipe took a circuitous route to California. After Armco made it, the pipe was apparently purchased by and shipped to Columbia Gas, a big East Coast provider of natural gas. For about six years, “the pipe sat in the yard exposed to elements in Delaware, Maryland or Pennsylvania,” the company’s report said.
Then Texaco bought the pipe, which Fisher says was shipped cross-country by rail to Coalinga. The 24-inch pipe, which had a wall thickness of about a quarter-inch, may have suffered a “transportation fatigue crack” when it was moved to Coalinga.
Shell’s lack of consistent paperwork is part of growing problem in the pipeline industry, according to Kuprewicz.
“We’re finding more and more companies that have undergone pipeline rupture failures where all of the sudden records have disappeared or they can’t find them,” said Kuprewicz.
Shoddy documentation involving a ruptured pipeline was an issue that came to light after PG&E’s San Bruno disaster in 2010. The incident involved a poorly welded natural gas pipeline, installed in the 1950s, that ruptured and exploded. The blast and subsequent fire killed eight people destroyed 38 homes.
Inspection Methods Questioned
Experts and safety advocates also expressed concern that Shell’s inspections did not reveal the crack that led to the rupture — all the more dangerous since a break took place eight months afterward.
The crack that eventually led to the September break could have been there in 1990, the year the pipe underwent its last hydrotest, when a section of a line is filled with water at high pressure to detect problems. The flaw was not detected then or in subsequent examinations using other inspection methods.
Before the September 2015 spill, Shell conducted inspections every two years on the line using electronic tools that seek pipeline defects, according to the company. Cal Fire says Shell has “historically” asked the office for permission to conduct inspections using in-line mechanical probes, instead of hydrotests.
State law requires pipelines considered high risk to be hydrotested every year. Pipelines that are at least 10 years old, like the San Pablo Bay Pipeline, are supposed to be hydrotested every three years, unless they get an exemption from the OSFM. In fact, the OSFM prefers the in-line inspections.
“Hydrotests only test the pipeline integrity at that moment in time,” says Cal Fire’s Tolmachoff. “In-line inspection tools are capable of detecting anomalies such as dents, gouges, mental loss, etc., that could lead to a release.”
But some experts say that the way in which the company is assessing risks on the pipe has not worked well.
“Those internal inspection computers they run through pipelines — it’s pretty well-known in the industry that they have a real hard time identifying this type of crack on those types of seam welds,” Weimer says.
“If the in-line inspection tool didn’t spot the crack and two months later or less it ruptured, something’s wrong here,” Kuprewicz says.
After the September spill, Shell conducted more in-line inspections of the pipeline, which found “no additional defects similar to that which caused the September release,” Fisher said.
But, in June, a month after the pipeline’s most recent rupture, Shell conducted a hydro test in the line, the first such exam on the pipe in more than 25 years, he said.
“Shell has had trouble with this pipeline before,” Bea says. “You can expect to see similar trouble with this same pipeline and with others like it again.”