A Pacific Gas and Electric Co. crew installed an inferior piece of pipe without testing its strength. Years later, when the line failed, the utility first dispatched an employee who wasn't properly trained to respond. It later admitted its records about the pipe were flawed.
Those facts fit the events surrounding the Sept. 9 transmission pipeline explosion in San Bruno that killed eight people and destroyed 38 homes. But they also describe an earlier, lesser-known disaster - the explosion of a PG&E gas distribution line Dec. 24, 2008, that killed a 72-year-old man at his home in the Sacramento suburb of Rancho Cordova.
Now, 2 1/2 years later, California regulators have lodged safety violation allegations against PG&E for that explosion that could carry fines of as much as $16 million.
The case, to be decided by the state Public Utilities Commission, is shedding new light on the company's safety practices leading up to the San Bruno disaster. Experts who reviewed the circumstances of both cases say they are disturbingly similar.
"They are mirror images of each other - just fewer people are dead in one of them," said Jim Hall, former chairman of the National Transportation Safety Board and now a pipeline safety advocate in Washington.
"With the information that has come out of these two tragedies, clearly PG&E's safety culture, safety policies, safety procedures should be of concern to all of the regulatory authorities," Hall said. "They need to ascertain if this is pattern failure."
Royce Don Deaver, a pipeline safety consultant in Texas, agreed that the explosions revealed a pattern.
"These are systematic problems - you see it in one case, it happens in another," Deaver said.
Brian Swanson, a spokesman for PG&E, disagreed, saying the root cause of the Rancho Cordova explosion was not the company's way of doing things, but individual employees' failure to follow policies.
He pointed out that federal investigators have yet to say what caused the San Bruno blast. But he says the company has learned lessons from both tragedies.
"We continue to evaluate all new information and improve our operations," Swanson said.
Timeline of disaster
The Christmas Eve 2008 blast that killed Wilbert "Bill" Paana and critically injured his daughter and granddaughter was the culmination of PG&E failures not only that day but dating back two years, state and federal documents indicate.
The first inkling PG&E had that something was wrong came at 9:16 a.m. the day of the explosion, when a resident of Paiute Way in Rancho Cordova called the utility to say she smelled gas outside her home.
Bonnie Stevenson, a 16-year-veteran gas technician, arrived about an hour after that call - without proper equipment or training to "identify, classify and assess outdoor leaks," a state investigation later found.
Four months before, the Public Utilities Commission's safety arm had cited PG&E for not having protocols for responding to outdoor leaks and not properly training or equipping its workers to handle them. Earlier in December 2008, the company told regulators it was still meeting with its unions and had yet to retrain workers.
Even though she did not have the right outdoor gear, Stevenson set about to find the leak. She began at the home of the caller just up the street from the Paana home, where she found some signs of gas.
Ten minutes after she arrived, Stevenson called the maintenance yard for a properly equipped investigator to come and pinpoint the leak. A supervisor contacted a technician with the right gear at 10:42 a.m.
But that technician, Ernest Grider, was busy at a job site in Sacramento. He said he would get to Paiute Way as soon as he could.
Telltale dead grass
Stevenson, meanwhile, checked another home, where a man directed her to a sunken patch of dead grass in front of the Paana house. He said he had smelled gas off and on there ever since a leak at that spot was found two years earlier and repaired.
"I found my leak, I need a crew," Stevenson remembered telling headquarters that morning. She also knocked on Paana's front door. No one answered. Her equipment picked up moderately high, but not explosive, levels of gas out front.
She also spoke briefly by phone to Grider, who promised he would be there by noon.
While Stevenson was in her car and on the phone, she missed seeing Paana and his daughter and granddaughter come into their house.
Separate reports by federal and state investigators subsequently faulted the company for its emergency practices, specifically for not issuing warning notices or yellow caution tape to workers for posting outside homes during leak investigations.
The reports also singled out Stevenson for not calling the Fire Department to the scene to make sure the Paana home and other houses were evacuated.
Stevenson later explained that she had not thought she needed to because she did not smell "massive" amounts of gas near the home, and her instrument did not detect its presence near the home's foundation.
Spotted at the yard
Back at the yard, PG&E's gas compliance supervisor spotted Grider, the leak investigator, between noon and 12:30 p.m. He urged him to go to Paiute Way and sent two other technicians separately to the scene.
Grider finally showed up and met his colleagues at 1:19 p.m. PG&E's own probe faulted two unidentified supervisors in the Sacramento office for not making sure a crew arrived sooner.
Federal officials cited the lengthy delay as a contributing factor in the disaster that followed. PG&E disputed that conclusion, saying the danger was far from clear even after the properly outfitted crew arrived.
Grider explained that he had been delayed by the earlier job, rain, bad traffic and mechanical trouble with a company truck. He and the other PG&E workers did not know that, by the time he finally got to Paiute Way, they had only 16 minutes to act.
Grider checked the home and soon found Paana, who told him a PG&E crew had repaired a gas pipe at the same spot on his lawn about two years before.
Grider's gear detected high levels of gas from the dead area of the lawn. He was doing more checks at 1:35 p.m. when Paana's 17-year-old granddaughter, Sunny Dickson, in a bathroom with her mother, tried to light a cigarette.
"As soon as I flicked my lighter, it felt like my hand blew up," the girl told investigators. "I heard two explosions."
The first, she said, blew out the bathroom walls. The second "took out the rest of the house. ... The walls were gone."
She and her mother, Kimberly Jo Dickson, were seriously burned. Paana was trapped under debris and died of burns and traumatic injuries.
Days later, PG&E dug up the still-intact front lawn and found that a 1 1/4-inch plastic pipe used in that earlier repair had come loose from a connector. The escaping gas had traveled under the home and exploded.
State and federal investigators discovered that PG&E crews went to Paana's home several times in September 2006 to deal with a leaky metal distribution pipe. Rather than digging up the lawn and replacing the pipe, the crew had inserted a smaller plastic pipe into the metal one.
That plastic insert was linked to the metal line with a 6-inch piece ofvplastic pipe whose walls were too thin to qualify as standard ground pipe.
It was this piece that came loose from the connector.
The investigators wondered why the inferior line had not immediately failed the pressure test that the PG&E repair crew was supposed to conduct, which would have alerted workers that they had installed the wrong pipe.
Mark Creighton, the 25-year-veteran foreman who oversaw the repair job, assured federal investigators that he had tested the pipe and that there had been no problem. "I don't leave anything I'm not sure of," Creighton said. "I feel comfortable it was done right."
But investigators soon found that one month after the repair, in October 2006, another PG&E crew installed the same type of substandard pipe on a repair job in nearby Elk Grove. In that case, the repair immediately failed a field pressure test.
PG&E realized after an investigation that the Elk Grove crew had somehow grabbed the wrong type of pipe by mistake, but wrote it off as an isolated incident.
After the Rancho Cordova explosion, a state investigator went to the Sacramento yard, from which the crews in both 2006 repair jobs had been dispatched. The investigator found more substandard pipe - which was to be used only for designating the course of gas lines - stored in a bin near where standard pipe was kept.
The potential for confusing the two types of pipe was obvious, the investigator later concluded, because they were nearly identical.
In late 2010, a staff report filed with the state Public Utilities Commission concluded that PG&E had committed a string of violations and suggested it be fined. Among the company's shortcomings, they said, was failing to check other recent repairs for substandard pipe after the incident in Elk Grove.
Had PG&E done so, they concluded, the Rancho Cordova disaster would have been averted. 'Full responsibility'
In its defense, PG&E filed a document in February in which it took "full responsibility" for the Rancho Cordova explosion and, at the same time, blamed employees for not following company policies to guard against such a disaster.
First, it blamed Creighton, the foreman on the 2006 Rancho Cordova repair job, not only for installing substandard pipe but for failing to pressure-test it.
PG&E revealed it had fired Creighton, as well as the two supervisors running the Sacramento yard the day of the explosion, who the company said had altered key documents on repair jobs. In the case of the Rancho Cordova repair, PG&E said, one of the supervisors changed an entry to make it appear the plastic pipe had been pressure tested when, in fact, it had not.
PG&E also said the two supervisors repeatedly signed off on improperly completed documents or added false information to them. It is not clear how many records were falsified.
PG&E subsequently held training sessions with supervisors and technicians about the need to fill out the forms properly "and the potentially disastrous consequences of failure to do so."
Swanson, the PG&E spokesman, also said that after the Rancho Cordova explosion, the company excavated repair jobs at 72 other sites in the Sacramento area and performed leak surveys throughout its system.
In another February filing, PG&E told the Public Utilities Commission it was prepared to agree to the facts of the Rancho Cordova disaster, so there was no need to have a hearing on them. It sought to move directly to arguments over how much it should be fined.
PG&E still insists, however, that its procedures were not to blame for what happened in Rancho Cordova. It said its handling of the leak was "reasonable, met industry standards and complied with applicable law."
The state commission's staff disagreed. In a recent filing, Robert Cagen, an attorney with the agency's safety arm, argued that the Rancho Cordova blast was "largely attributable to PG&E's procedural failures ... rather than from rogue actions of PG&E employees."
Last month, an administrative law judge with the commission ruled that any hearing should be able to explore whether PG&E practices were at fault.
State regulators, meanwhile, want more time to investigate the firings of Creighton and the two supervisors. They also have questions about a recently disclosed July 2009 internal memo documenting how PG&E had reassigned another employee for "providing inaccurate information" to senior management about the Rancho Cordova blast.
"Clearly, this is critical information" that could call into question whether PG&E told the full story to federal and state investigators, Cagen told the judge.
San Bruno lessons?
Some experts say any flaws in PG&E policies that come to light should help better understand what happened in San Bruno - where the company laid down pipe in the 1950s with a poorly constructed weld, never pressure-tested it, and when it exploded, dispatched an untrained worker to try to turn off the pipe's manual valves. The fire burned for nearly 90 minutes before properly trained and equipped workers shut off the gas.
"Blaming the employees does not get the operator off the hook," said Richard Kuprewicz, a pipeline safety consultant in Redmond, Wash. "There's more than one breakdown here."
He added, "When you see this many breakdowns, in more than one area, it doesn't take long for the public to start to ask the question, do they have confidence in their ability to control this system?"
PG&E spokesman Swanson disagreed that there were obvious parallels between the Rancho Cordova and San Bruno explosions. The company's procedures in the case of Rancho Cordova were sound, he said - employees' shortcomings were the main problem.
Hall, the former National Transportation Safety Board chairman, said both the Rancho Cordova and San Bruno explosions raised so many red flags that regulators should be held to account along with the company.
"This is a public utility," Hall said. "Where is the state oversight? Where is the federal oversight? Where are the individuals looking out for the safety of homeowners?" E-mail Jaxon Van Derbeken at firstname.lastname@example.org.
Copyright 2011 SF Chronicle