PG&E’S 1963 BLAST AN EARLY WARNING
ON LINES’ SAFETY
This story by Jaxon Van Derbeken was publish June 26, 2011 by the San Francisco Chronicle.
A Pacific Gas and Electric Co. gas pipeline running up the Peninsula into San Francisco has a long history of cracked and poorly constructed welds and even exploded once – but it’s not the one that blew up in San Bruno last year.
The pipeline is known as Line 109, and it failed disastrously in 1963 in the Bernal Heights neighborhood in San Francisco. The blast injured nine firefighters and led to the heart-attack death of a battalion chief.
Documents contained in a mountain of just-released PG&E paperwork show that over the years, the company has found flawed welds up and down the 50-mile line. But there is no evidence in the records that in all that time, the company ever revealed the problems to state regulators.
PG&E also says it has lost the key document detailing what caused the 1963 explosion. Although it theorizes what happened, based on records that do exist, it doesn’t know for sure what led to the line’s failure – or whether that problem could be repeating itself on the pipeline’s snaking path from Milpitas to San Francisco’s Dogpatch neighborhood.
Experts who reviewed the reports and photos of the cracked and incomplete welds on Line 109 said the evidence raises doubts about PG&E’s system of aging lines.
“Some of the cracks I looked at were awfully familiar to what was exposed in the San Bruno” explosion, said Bob Bea, a UC Berkeley engineering professor and risk management expert. “This shows the potential for the same kinds of flaws and defects.”
Line never fully tested
Like Line 132, the pipeline that exploded in San Bruno on Sept. 9, Line 109 was never fully tested using pressurized water to check its integrity. Instead, PG&E relied mostly on inspection methods that detect leaks or corrosion, but not bad welds. In one especially troublesome stretch, it installed a plastic liner in 1995 and used water to test the line beforehand.
PG&E says more than half of Line 109 has been replaced in recent years, and the pipe has had no leaks from bad welds in the past decade. The company has also cut gas pressure on the line by 20 percent at the order of the California Public Utilities Commission, which was concerned that a problem such as the one that destroyed Line 132 in San Bruno might be lurking on Line 109.
“We have openly acknowledged the need to significantly improve many of our historic operations practices,” company spokesman Brian Swanson said. Since the San Bruno disaster, he said, “we’ve been working intensively to do exactly that. Today, and going forward, we are taking whatever steps are necessary to bring our operations up to industry-leading levels and to assure our customers that our system is operating safely.”
Line 109’s problems first came to everyone’s attention almost 50 years ago.
On Jan. 2, 1963, the transmission pipe sprang a leak under Alemany Boulevard in San Francisco. About 1,000 homes were evacuated as firefighters rushed in to help.
Before PG&E crews turned off the line, gas spread to a nearby home, which exploded. Two of the nine injured firefighters were critically hurt, and Battalion Chief Frank Lamey, 63, died of a heart attack.
Firefighter recalls blast
One of those critically injured was Anthony Marelich Jr. In an interview last week, he said PG&E had left the line active during the evacuation to avoid cutting off thousands of other customers and believed the gas was safely venting into the atmosphere.
Instead, it was filling a house on Nevada Street. Marelich said he had been standing with several firefighters when the home blew up and a wall “landed on top of me.”
“It was instantaneous,” said Marelich, now 73. His face was crushed, and doctors gave him almost no chance to survive.
He was forced to retire the next year, having lost several teeth and his sense of smell. Surgeons had to wire his jaw back on.
“Safety, right now, is in the limelight because of San Bruno,” Marelich said, adding that he thinks PG&E should have paid a steep price for the 1963 blast, “but they never showed any blame for it.”
“What happened to me and what happened to those people down in San Bruno, it should never have happened,” Marelich said.
Weld failure detected
An investigation showed that the line had fractured at a welded joint – known as a girth weld – that encircles the pipeline and links two segments. Swanson, the PG&E spokesman, said some documents from 1963 indicate that the weld failed because of pressure from dirt that had been piled on the line.
But in releasing almost 250,000 documents about its gas transmission system to the California Public Utilities Commission last week, one record PG&E was unable to find was the metallurgical report that went into detail on what caused the failure.
PG&E has been vexed by such record-keeping problems. Its documents incorrectly showed that the line that exploded in San Bruno, killing eight people and destroying 38 homes, had no longitudinal seams. In fact, the line ruptured at an incomplete weld on such a seam, federal metallurgists have concluded.
The company also is facing potentially millions of dollars in state fines for its inability to locate records to vouch for the safety of about 500 miles of gas-transmission lines in and around urban areas.
“It is always reprehensible and unsafe when PG&E cannot locate a metallurgical report on a weld flaw or failure,” Bob Cagen, an attorney for Public Utilities Commission, said of the missing 1963 document. Cagen is overseeing a state investigation into the company’s recordkeeping.
“It means reduced safety,” Cagen said, “because in order to determine what to do with the system and what portions are vulnerable, a utility must understand and study the causes of past failures.”
More problems with weld
Three years after the 1963 blast, there was another failure involving a Line 109 girth weld, according to one of the documents that PG&E turned over to the state last week. That document does not indicate where the failure occurred or give other details, only that it was caused by a bad-quality weld.
In 1981, PG&E conducted metallurgical strength and durability tests on lines across the Bay Area and found that at least four girth welds under Alemany Boulevard failed at least one of the tests, according to the documents.
Fifteen years later, PG&E set about assessing the state of Line 109’s girth welds. At the time, the company was in the midst of a program to replace its aging pipes across the Bay Area, and Line 109 certainly qualified – its southern portion was installed in 1936, and the pipeline under San Francisco was put in place in 1932.
PG&E had known since 1981 that many of the welds that held pipeline segments together were made using oxyacetylene technology dating from the early 20th century. The technique generated gas bubbles in the welding bond, making pipes brittle and more susceptible to failure in earthquakes.
Records show that PG&E replaced several hundred feet of Line 109 in the city in 1995. At the same time, it told the state Public Utilities Commission that it had decided to reinforce about 2 miles of Line 109 along Alemany Boulevard with an internal plastic liner, rather than replace it.
But what the company did not tell the state was that it was finding problems such as weld cracking all over Line 109.
In an internal report in 1994, PG&E technicians found major weld defects in each of the nine pieces of pipe that workers had recently dug up for analysis.
The test sections were taken from Milpitas, Burlingame, and the Ingleside Heights and Dogpatch neighborhoods of San Francisco, the report said. The most serious problem was from pipe dug up on Chester Avenue near Alemany in San Francisco, which had a 2-inch-long crack in a girth weld that had penetrated two-thirds of the way through the wall.
Regulators not informed
PG&E didn’t mention the girth weld problems to the state in 1995 when it obtained regulators’ permission to install the protective plastic liner inside the pipe along Alemany in Bernal Heights. The company told the state only that it was trying to strengthen the line against earthquakes, according to a Public Utilities Commission document from 1995.
The PG&E records released last week show that the company has replaced some of the segments in San Francisco that had flawed girth welds, and that the remaining 1.83 miles of original 1932 pipe are lined with the pressurized plastic.
When San Francisco officials asked PG&E this year for documents related to Line 109, however, the company did not turn over its records on the problematic welds. The city attorney’s office received the records only last week when PG&E made its filing with the state.
City Attorney Dennis Herrera said the disclosure leaves more questions than answers, and he called the documents incomplete and inadequate.
“PG&E needs to make it clear what steps it has taken to ensure the safety of all” the aging lines in the city and elsewhere, Herrera said. “It’s very, very difficult to reconcile the information that has been provided, to know how it all fits together.”
New failures on pipeline
A year after PG&E installed its pipeline liner in San Francisco, the company found more problems with two of the nine pieces of pipe it first examined in 1994.
The segments, removed either from Burlingame or Milpitas, had not just bad girth welds but flawed longitudinal seam welds as well, the newly released PG&E records show. Seam welds are made at a pipe factory, not in the field. It was a seam weld that failed in the San Bruno disaster.
One of the seam welds double-checked in 1996 had a crack that penetrated three-fourths of the way through the wall of the pipe, and the other had a crack that traveled 44 percent through. The worst crack was covered over by the girth weld, which was all that was holding the pipe together.
The lab report said some of the weld flaws appeared to have been repaired at some point, but only partially, and that the cracks probably originated from when the pipe was put in the ground in 1936.
Those sections of pipe were removed, PG&E’s documents show. But pipeline from the same era remains along 17 miles of the line on the Peninsula and is not due to be replaced until 2014.
‘Many early warning signs’
Bea, the UC Berkeley engineering professor, examined a photograph of the pipe with the worst crack and said, “This looks like something that would have been done in the first day of my welding class.
“If it’s an indication of the quality associated with the work on this line,” Bea said, “it is not something you would want to build your home or hospital over.”
Bea said the overall picture of Line 109’s condition painted by the newly released documents was troubling.
“You have bad welds, bad steel, and it appears to be scattered randomly throughout the system in this case,” Bea said. “In hindsight, we see many early warning signs to indicate what were causative elements in the San Bruno experience. It’s a matter of time and exposure before the weakest links are exposed.”