Safety alarms raised at nuclear weapons plant

Federal investigators are looking into deteriorating conditions at the Pantex plant in Texas. Energy Department officials say there's no danger.

By Ralph Vartabedian

Times Staff Writer

February 21, 2007

AMARILLO, TEXAS — Electrical failures have shut down the plant. The roof has leaked. Decrepit machinery dates back more than 40 years. Safety lapses led inspectors to levy fines twice within two years. And employees, under deadline pressure, complain they are often worked past the point of exhaustion.

If this factory were producing medical devices or refining gasoline, the conditions would be serious enough. But this is where they work on nuclear bombs.

Pantex is the Energy Department's main nuclear weapons factory, a linchpin of the nation's defense for half a century. The nation no longer makes nuclear weapons, so the plant's chief roles are servicing them or dismantling them to meet the terms of disarmament pacts.

On a 25-square-mile swath of the Texas Panhandle, a series of massive white concrete domes mark the places where live nuclear weapons are opened up. The rituals and procedures inside those cells are supposed to be as strict as in any operating room, part of a safety culture that reduces any chance of an accidental nuclear explosion to one in 100 million.

But lately, outside experts are questioning whether safety margins are eroding. Federal investigators are trying to assess the overall safety of the plant, which employs 3,300 people, amid troubling safety snafus and what employees call an atmosphere of intimidation.

Energy Department officials acknowledge that the plant has fallen behind schedule on reliability testing of weapons. Long delays have occurred in decommissioning thousands of surplus warheads. They also concede the plant has maintenance problems and has violated safety procedures. But they insist there is virtually no danger of a conventional or nuclear explosion.

"Pantex is safe, no doubt," said Marty Schoenbauer, the acting chief of the Energy Department's nuclear weapons program.

Safety has improved in recent years, he said, thanks to better procedures. But outside experts, union officials and watchdog groups say the opposite is true — that safety has regressed since 2000 as the most knowledgeable senior safety experts of the Cold War era retire and the plant's condition deteriorates. Energy Department Inspector General Gregory H. Friedman is investigating safety conditions at Pantex.

"You can't run a plant on glittering platitudes and generalities and call that a safety program," said Bob Alvarez, a former deputy assistant secretary of Energy and now a senior scholar at the Institute for Policy Studies, a Washington think tank. "A nuclear detonation accident is a low probability, but it is not incredible."

The backdrop to problems at Pantex is a growing concern that the Energy Department has mismanaged the nuclear weapons program. Last year, the Defense Department bluntly said that it had lost confidence in the Energy Department, Energy Secretary Samuel Bodman has acknowledged.

"We have constraints," Bodman said in an interview, conceding the department hasn't met all of its commitments to the Pentagon. Last month, he fired the head of the nuclear weapons administration.

Conditions at Pantex began deteriorating at the end of the Cold War in 1989, when federal managers started starving the plant of funds. Billions of dollars were instead funneled into nuclear weapons laboratories, giving scientists new supercomputer centers, powerful lasers and physics instruments.

By about 2000, the leaks in Pantex's roof were so bad that workers had to cover bombs with plastic when it rained. In summer 2004, a power overload tripped transformers, causing a plant-wide blackout. Last July, another electrical failure occurred when rats gnawed through wiring, according to weekly safety reports. And in August, a storm swept over the plant that left standing puddles in nuclear production areas.

Although such conditions don't necessarily lead to accidents, the Energy Department has levied fines totaling nearly $234,000 against the contractor that operates Pantex, BWX Technologies Inc., for safety violations.

In one case, involving the disassembly of a missile warhead, technicians improperly used red vinyl tape to secure a crack in the high explosives surrounding the plutonium sphere of the hydrogen bomb. The use of the tape itself was not faulted, but technicians misread engineering instructions and caused an even bigger crack.

Federal safety inspectors found that the flawed operation "increased the opportunities for dropping all or part of the explosive during handling and hence increased the potential for a violent reaction," a finding that ran against assurances such a detonation was virtually impossible.

In the second case, technicians were extracting an assembly of high explosives and plutonium from the casing of a different missile warhead for servicing. Using a jackscrew to apply several thousand pounds of force to the explosives, technicians exceeded the allowable loads and over a three-day period violated strict safety protocols.

Dan Swaim, BWX Technologies president at Pantex, acknowledged that both incidents broke safety rules and were unacceptable, but he said there was no risk of a disaster.

"No nuclear yield was possible," Swaim said, asserting the company maintains huge safety margins that preclude any potential for a conventional or nuclear explosion. "As a guy who works here every day, I want it to be darn safe."

The problems at Pantex came to light last fall after Danielle Brian, executive director of the Project on Government Oversight, a Washington watchdog group, wrote two letters to Bodman complaining about safety. The group has cited a lengthy report in 2000 by former Energy Department safety expert Frank Rowsome, who said that a detonation caused by lightning strikes, solvent fires or other incidents at Pantex was more probable than the Energy Department was admitting.

In an interview, Rowsome, who retired in 2004, said he did not want to alarm the public, but he believed Energy Department officials were so "overly confident" and "complacent" about safety that they were not alert to deteriorating safety conditions.

Where some see problems, others see progress. Last month, Pantex finally began overhauling the B83, a nuclear bomb designed to drop from a plane, after an 18-month delay triggered when scientists discovered potentially dangerous static electricity in work cells. Schoenbauer said the delay showed how far the Energy Department would go to protect safety. Critics say it shows that hidden safety problems can still exist.

Meanwhile, Pantex also has fallen behind schedule in performing crucial surveillance tests required by laboratory scientists to certify the reliability of the bombs, Schoenbauer acknowledged. "That backlog has not affected the lab's ability to certify weapons," he said.

But Ralph Levine, who once ran the Energy Department's nuclear weapons surveillance testing, wrote a letter in 2005 asserting the backlog would allow defects in nuclear weapons to go undetected for years. As a result, he said, Energy officials removed him as manager of the program, and he retired last year.

John Duncan, who until four years ago headed surveillance testing at Pantex for Sandia National Laboratory, agreed that testing problems at Pantex are undermining confidence in the stockpile. Even today, the certifications of nuclear weapons are being made with less certainty than scientists should have, Duncan and Levine said.

"I knew we were in trouble when I started attending meetings in Washington and was told to work better, faster, cheaper," said Duncan. "They started sending people to the plant with little weapons experience."

The Energy Department confirmed Friday it planned to reduce the number of annual surveillance tests, saying the plan was better suited to an aging bomb stockpile. But Duncan said this would further erode the reliability of the weapons.

A senior federal manager at Energy Department headquarters, speaking without attribution because the interview was not authorized, endorses those concerns.

"The delays in testing and lack of resources are so significant," he acknowledged. "The real question is when the laboratories can no longer certify the reliability of the stockpile."

The workload at Pantex is likely to grow even more onerous in coming years.

Under the Moscow Treaty, the U.S. needs to get rid of about 4,000 surplus nuclear warheads. At current rates, that could take Pantex until 2050, said Harvard University nuclear weapons expert Matthew Bunn. Though there is no firm deadline, that pace is unacceptable to arms control experts, Bunn said.

The Bush administration has ordered the plant to increase dismantlements by 50% this year.

Another task looming for Pantex is modernizing the W76 missile warhead used on the Trident submarine. Hundreds of W76 warheads will have to be disassembled and rebuilt with new parts. Swaim said the W76 program would begin on schedule this year.

Pantex — named for the panhandle of Texas — has a largely blue-collar culture. The people who actually touch the weapons are known as "production technicians." They earn about $24 per hour. Engineers average $84,000 per year.

Though the jobs are sought after in Amarillo, an anonymous letter surfaced in November alleging that the plant was in serious disrepair, BWX Technologies management was letting safety slip and employees were forced to work more than 80 hours a week in some cases.

The stress of working with nuclear weapons has been exacerbated by an abusive management, said Henry Bagwell, the former chief of the Metals Trade Council, the principal union at the plant. "They treat people badly," said Bagwell, who left last year after 24 years at the plant.

Bagwell said that when he attempted to raise a health and safety problem involving toxic beryllium dust in 2003, he was demoted from X-ray technician to janitor.

"It sent a message," Bagwell said. "It was a public humiliation. Safety seriously took a back seat."

Swaim takes sharp exception to allegations that the roof leaks, labor relations are strained or that severe amounts of overtime are being forced on workers. After the letter surfaced, Swaim said an audit showed overtime averaged just 15% over the normal 40-hour work week, though some individuals racked up more.

"I told my guys no more than 72 hours a week," he said.

ralph.vartabedian@latimes.com