http://www.seattletimes.com/news/local/html98/pipe_19991027.html Link

NTSB ties rupture to water-line project

by James V. Grimaldi Seattle Times Washington bureau

WASHINGTON - Federal investigators have preliminarily determined that a pipeline in Bellingham burst at a gouge on the pipe near where construction workers installed city water lines five years ago.

National Transportation Safety Board (NTSB) Chairman Jim Hall also said in testimony at a hearing here today that the gasoline pipeline ruptured "in the immediate vicinity" of where inspectors for Olympic Pipe Line found an anomaly in 1996, two years after the construction had been completed.

Olympic did not excavate to examine the "possible wrinkle bend" in the pipeline because it determined no action was needed, Hall said, and pipeline officials who made the decision have refused to answer questions, citing their Fifth Amendment right against self-incrimination.

The pipeline burst June 10, sending 277,000 gallons of gasoline into Whatcom Creek. Liam Wood, 18, was overcome by fumes and drowned. Wade King and Stephen Tsiorvas, both 10, died after inadvertently igniting the fumes into a massive fireball.

Meanwhile, the U.S. Department of Transportation announced today it would inspect every underground pipeline in Washington state in the aftermath of the explosion.

"These measures . . . will lead to lasting improvements in pipeline safety," according to U.S. Transportation Secretary Rodney Slater.

The plan calls for the federal department to work with the Washington Utilities and Transportation Commission to conduct the inspections, expected to be completed in six months, said Kelley Coyner, the federal agency's research and special programs administrator.

The federal department's testimony at the hearing before the House Transportation Committee today provides the most complete account to date of the NTSB's continuing investigation into the blast.

The latest findings further confirm a theory investigators developed soon after the blast - that construction of water lines to a nearby treatment plant might have damaged the pipe. While the NTSB has said the rupture occurred near gouges, Hall's latest comments mark the first time he has said the rupture "originated at a gouge on the surface of the pipe."

Testimony today also raises new doubts about Olympic's statements that a computer "slowdown" contributed to the rupture and fuel leak. And Hall raised fresh questions about operation of a key Olympic fuel-storage facility near Mount Vernon.

At the hearing, Frank King recounted the wrenching events of June 10 and his mission to find "justice for my son's death."

King says justice will come only "if I could go up to my son's grave and tell him we succeeded in making Olympic Pipe Line Co. accountable."

As observers wiped tears from their eyes, King and his wife, Mary, asked the committee to put in place regulations that would prevent such an accident from happening again.

"I hope I've made my son come alive to you today," King said as he held his wife's hand, `because I don't want pipeline-safety issues to be buried along with him."

While Fred Crognale, president of the pipeline company, has outlined how Olympic is meeting Department of Transportation requirements to reopen the line, Hall of the NTSB cast doubt over Olympic's public explanation that a computer "slowdown" occurred at the time of the accident and contributed to the pipeline being accidentally reopened for 17 minutes after the rupture.

"Restarting a pipeline after a rupture suggests a significance performance failure," Hall said. "We don't yet know whether this can be traced to training, qualifications, equipment malfunctions, poor design in the computer-based control system or some other undetermined culprit."

Committee members questioned Crognale about the anomalies found in the pipeline section that ruptured and asked why Olympic did not excavate to check them out. Crognale said: "I can't tell you. I was not personally there and I cannot tell you what the judgment was."

Rep. Bob Franks, R-N.J., the subcommittee's chairman, expressed frustration at Crognale's response and said he would be sending written questions to the company to get more complete answers.

Crognale testified that "during the computer slowdown, the controllers were unable to obtain current pipeline information on the computer screens and to process commands to equipment, such as pumps along the pipeline."

But Hall says a report prepared by Olympic for federal regulators "acknowledges that the alleged (computer) system slowdown could not be verified or reproduced."

"Our preliminary review has not identified that a slowdown actually occurred on the computer system the day of the accident," Hall said.

Crognale says Olympic has beefed up its faltering computer system, but Hall doused the significance of those claims. "Until we fully understand what happened during the accident sequence," Hall said, "the impact of these (computer) changes on future system operations cannot be fully evaluated."

Hall also raises questions about a new storage station and a series of valves there. The investigation is focusing on a relief valve installed when the station was built near Mount Vernon last year.

NTSB investigators found that Olympic made key design changes on the relief valve but might have failed to replace a piston and cap recommended by the valve's manufacturer.

"Since valves of this type or those with a similar design are commonly used throughout the liquid pipeline industry, it is extremely important to fully understand what occurred," Hall said.

Tests on the relief valve appear to be on hold, pending a criminal investigation into the accident, Hall said.

Investigators also have put off, for now, microscopic fatigue tests on the ruptured pipe. Federal prosecutors are worried the tests, which require destruction of the pipe, could weaken their case and provide fodder for defense attorneys.

Hall said the U.S. attorney's criminal investigation into the pipeline accident was making it difficult for investigators to determine what caused the accident. Hall displayed a chart detailing the various ways the criminal probe had stymied the accident investigation. Members of the committee took note.

"It looks to me like you're pretty much out of business in some of the most significant aspects of this investigation," said Bob Wise, D-W.Va.

However, some members of Congress and Hall criticized the Office of Pipeline Safety for its slow response to calls for tighter regulation.

"It is a sad state of affairs that regulatory oversight is basically coming out of the Department of Justice and not the Department of Transportation," Hall said.

The NTSB investigation also is stymied in part because of Olympic employees who refuse to cooperate. Hall is especially concerned about "four key individuals who may have direct knowledge of the events that occurred in the control room during the accident sequence."

The workers include two controllers, their supervisor and a former controller who now is responsible for maintaining the pipeline computer system. The former controller "was reportedly performing modifications to the computer programming."

Copyright ) 1999 The Seattle Times Company

-- Homer Beanfang (Bats@inbellfry.com), October 28, 1999

Answers

" ... performing modifications to the computer programming."

coming soon to pipelines near you, globally, locally, everywhere

-- rupture (not@rapture.burstbubble), October 28, 1999.

Within hours of the explosion, the Bellingham paper (I believe) reported that computer upgrades were underway at the pipeline company, that the computer had just crashed, and was being rebooted when the leak was discovered. There has been zero investigative followup on that issue.

Because the leak timing is uncertain, we don't know if some computer error caused a transient spike that blew the line, or what.

See one thread and another detailed one.

The pipeline company was very quick to talk about the city maintenance, and slow to talk about their computer. That I understand - any company wants to shed blame. But let's consider the meaning, if this was in fact a date/time-related computer error, indicative of Y2k problems. What should be the level of blame, if a hundred similar failures take down the US pipeline system as a whole? What should be the penalty, if this company gave no warning when it could have?

Interesting problem, yes?

-- bw (home@puget.sound), October 28, 1999.

The pipeline company was very quick to talk about the city maintenance, and slow to talk about their computer. That I understand - any company wants to shed blame. But let's consider the meaning, if this was in fact a date/time-related computer error, indicative of Y2k problems.

That has been gone through before. It was not a Y2K problem. It was a problem with the upkeep, maintenance and training of the personel. One idiot decided not to believe the warnings generated and restarted the pipeline. The "damage" to the pipe from the equipment putting in the water line was not investigated or considered serious. When they tested a part of the line by filling it with water reciently, the pipe burst.

Now every INCH of all pipelines in Washington state will be checked.

By assuming this is a Y2K problem, you discount where the problem really lies. In the bad maintence practices of the company and the complacent attitude of the workers.

-- Cherri (sams@brigadoon.com), October 29, 1999.

It hasn't been gone through before, it's merely been reported. We have no idea if the reports are accurate, or what really happened. Reporting on Y2k-related issues has been pretty poor; reporters are frequently illiterate regarding computer problems, and they didn't know what questions to ask on this one. The right questions, early on, might have been useful.

Note that the computer operators are pleading the fifth and refusing to say what happened in the control room. If it was just poor training, they could say that and deflect blame from themselves, right? They were doing something to the software at the time.

It didn't HAVE to be a y2k problem. (And in that category, I include new software installed in a rush, to prevent Y2k problems later.) But we certainly have evidence that should make us look closer, and there has been virtually no discussion of that evidence. So it's quite possible that it WAS a Y2k problem.

It's been gone through? Rubbish.

-- bw (home@puget.sound), October 29, 1999.