Errors, Costs Stall Nuclear Waste Project
On a desert plateau seven miles from the Columbia River, a massive federal project to clean up a Cold War-era nuclear weapons plant is deeply troubled.
The effort to avoid a future environmental calamity here, at the most polluted site in North America, is a priority of the Energy Department but has foundered because of engineering mistakes and runaway costs. Fifty-three million gallons of radioactive sludge, most of it the texture of ketchup, is stored in scores of underground tanks, some of which have leaked for years.
The Energy Department and its lead contractor Bechtel Corp. are trying to build a sophisticated waste treatment complex -- a small-scale industrial city -- that would transform the sludge into radioactive glass. After spending $4 billion since 1989 and getting rid of three previous contractors, the program has yet to transform a gallon of sludge.
“We have had some world-class technical issues,” acknowledged John Eschenberg, the federal manager for construction. “I have made mistakes. Bechtel has made mistakes. If I could relive the last three years, there are things I would do differently.”
The project is a long-distance race to empty the leaky tanks and secure the radioactive waste before it becomes a greater menace to the Columbia River. The job is likely to take decades, and the price tag could approach $100 billion.
In January, the Energy Department stopped construction on the two most important parts of the project after it realized it had miscalculated the earthquake risks at the sprawling federal facility, known as the Hanford Site. In recent weeks, it put off any resumption of construction until after October 2007. At best, the plant would be finished in 2019.
What remains uncertain is whether the plant’s remarkably complex technology will work as planned. Shortly after construction was halted, a team of experts delivered a sobering report that warned of a large number of other potential technical issues that could undermine the plant’s operation.
In addition, a long list of major safety problems has been discovered -- though these problems are fixable, construction managers say. They include the potential for explosive hydrogen gas to build up inside the plant’s pipes; concerns that the steel frame had inadequate fireproofing; and the discovery of faulty welds in tanks designed to hold dangerous waste.
The cumulative effect of all the problems and challenges has been staggering.
Energy Department officials disclosed in May that the plant would probably cost $11.6 billion to build, double the estimate of only three years ago. An independent cost estimate due in coming weeks from the Army Corps of Engineers is expected to exceed $13 billion.
“You want to take somebody out and hang them,” said Rep. David L. Hobson (R-Ohio), the chairman of the House Appropriations subcommittee that pays for the project. “It is already outrageous what it is costing.”
Hobson, who has led an investigation into the problems at the Hanford plant, said the Energy Department had ignored problems until they mushroomed out of control and had adopted a “fast-track” construction strategy that backfired.
Meanwhile, officials in Washington state are furious about the continued delays. The state has a legal agreement with the Energy Department that promised the plant would be operating by 1999, meaning it is now 20 years behind schedule.
“We are extremely frustrated,” said Suzanne Dahl, the top official at Hanford from Washington’s Department of Ecology. “It is becoming impossible to accept more delays.”
Construction of the waste treatment complex, consisting of two dozen massive buildings, is only 30% completed, and engineering work is about 70% completed, Eschenberg said. Building and designing the plant at the same time was necessary, he said, to get the cleanup done as quickly as possible. The decision to halt construction was a prudent step that will give engineers time to solve all the problems, he said.
James Rispoli, assistant Energy secretary for the nuclear waste cleanup program, acknowledges the program has had setbacks but says it is not facing any problems that would derail the project.
Although work has stopped on the pretreatment plant and high-level-waste plant, construction is continuing on 20 other facilities in the complex, Rispoli said. “We are keeping the forward momentum,” he said.
Of course, there is nothing new about federal projects hitting technical problems that cause skyrocketing costs. In most cases, the problems are overcome, and often the projects even become examples of U.S. engineering prowess. But in the case of Hanford, there are some deeper concerns.
Bechtel says it underestimated how much U.S. expertise in nuclear engineering has atrophied. Academic experts agree that the U.S. has lost much of its nuclear know-how.
The history of problems at Hanford raises questions about how effectively the radioactive waste dumps left over from the Cold War can be cleaned up -- even with the best technology and with almost unlimited federal spending.
The project’s current problems come on the heels of a whole other set of missteps dating to the 1990s, when three different efforts to build a treatment system failed. British Nuclear Fuels, the contractor that preceded Bechtel, was fired when it submitted a $15.2-billion estimate to build a plant in 2000. Bechtel promised to build the plant for less.
Rough estimates for building and operating the plant -- then decommissioning the facility when the job is done -- range from more than $50 billion to $100 billion.
“This is the environmental equivalent of sending men to the moon,” said Bob Alvarez, a senior scholar at the Institute for Policy Studies in Washington, D.C., and a former Energy Department official. “It costs about the same.”
There is also an urgency to the mission, given the risk that radioactive waste will someday reach the Columbia River, the largest river in the West. About 1 million gallons of the waste has already leaked into the ground at Hanford, though government experts are confident the rate of leakage has slowed or stopped.
Government hydrologists say they have no evidence that any leaked sludge has reached the water table 250 feet below ground, and they cannot calculate when -- or whether -- the radioactivity will reach the Columbia River.
Such assurances are rejected by some outside experts, including geotechnical engineer John Brodeur, who conducted a comprehensive study of the tanks in the late 1990s for the Energy Department.
“Some of the ground under the tanks is screaming hot,” said Brodeur. “The groundwater is already contaminated.”
By 2019, the plant is supposed to be ready to transform the waste into glass, a process called vitrification. New pipelines would carry the waste to a facility consisting of three huge radioactive waste treatment plants, a water treatment plant, a laboratory, a power distribution center and a maintenance shop.
The idea is to separate the highly radioactive materials into two waste streams: a small amount of high-level waste that will be vitrified and shipped to a future dump in Nevada; and a much larger volume of lower-level waste to be vitrified and buried at Hanford. Eventually, there would be 10,000 canisters of high-level vitrified waste and 100,000 canisters of low-activity waste.
The separation process is the biggest challenge, because the waste composition is so complex. Some of the highly lethal waste is a thick red or green sludge, and some is made of hard salt cakes. A teacup of the waste would deliver a lethal dose in minutes, nuclear waste experts say.
The separation occurs in a huge pretreatment plant, which by itself will cover the size of four football fields and reach 12 stories high. For safety, the concrete walls are 4 feet thick. The separate processes use both fine metal filters and two chemical processes, which have never been tested together on a large scale.
“They are taking a real risk the thing won’t work and they will have a $11.5-billion white elephant sitting in the desert,” said Tom Carpenter, nuclear oversight program director at the Government Accountability Project, a Washington, D.C., watchdog group.
Amid growing congressional concerns about Hanford’s technology, Bechtel assembled a team of the top nuclear experts in the nation.
In a March report, they cited a number of defects that would have to be fixed for the plant to work. They said one of the two chemical processes was “undemonstrated” and the other “will not provide acceptable performance.” The whole pretreatment facility “will be difficult to reliably operate.”
The team of outside experts also raised concerns with the vitrification processes. Once the waste streams are separated, they are sent to two different final treatment plants for vitrification. The melters in the low-level plant could wear out or fail prematurely, while the piping in the high-level plant could get plugged up, they said.
The report raised the prospect that the Hanford treatment plant might wear out before all the waste was treated, particularly if it could not operate reliably and avoid shutdowns.
The plant is designed to turn out 6 metric tons of vitrified high-level waste per day and 30 tons of low-level waste. If it can do that, it will complete the job 40 years after starting up, or 2059. If the job lasts longer than that, the plant could wear out first.
Rispoli, the Energy Department environmental chief, believes the outside assessment shows that the plant will work. All the Energy Department has to do is solve the problems identified in the report, he said.
The problems at Hanford reflect a troubling reality about the U.S. nuclear workforce: Its skill has declined over the last several decades as nuclear construction wound down, said William Elkins, Bechtel’s top officer managing the Hanford construction.
“There is no question we underestimated the amount of atrophy, how much the nuclear supply industry had atrophied ... how difficult it would be to find the people with the required skills,” he said.
Academic experts agree the U.S. is losing its expertise in nuclear engineering. There are 35 universities in the U.S. with nuclear engineering departments, half as many as in the 1970s, said Joonhong Ahn, a UC Berkeley expert on nuclear waste. And enrollment has plunged, said Jim Stubbins, chairman of nuclear engineering at the University of Illinois in Urbana-Champaign.
Bechtel is still trying to determine whether the waste treatment plant could stand up to the ground motion of a magnitude 9 earthquake on the Washington coast a couple of hundred miles away.
The issue was raised in 2004 by the Defense Nuclear Facilities Safety Board, an independent federal agency that oversees Energy Department sites. As engineers examined the issue, they realized the ground motion at Hanford could be 38% greater than they had anticipated.
Critics, including board Chairman A.J. Eggenberger, say Bechtel and the Energy Department took two years to act on that information, allowing the problem to mushroom. So far, it appears Bechtel designed safety margins large enough so that no walls will have to be rebuilt. But costly equipment will have to be upgraded, and engineers are still studying the issue.
Rep. Hobson held a blistering oversight hearing on the problems in April and then wrote new constraints on the program in the 2007 appropriations bill. He would require Bechtel to complete 90% of the engineering on the plant before it could resume construction, and he wants the Nuclear Regulatory Commission to oversee the project.
The Energy Department and Bechtel have resisted, saying those ideas would not help the situation. Bechtel, which is already facing a demand from the government to forfeit $48 million in fees, says its professional reputation has been tarnished.
“We don’t like any inference that we are soaking the taxpayer or don’t know what we are doing or are lining our pockets,” Elkins said. “Nothing could be further from the truth.”
But Hobson is impatient with Bechtel’s concerns about its reputation and with its explanations, particularly its claim that nuclear technology has eroded in the U.S.
“Of all the people in the world, they have the most expertise,” Hobson said. “They took their best people and put them on other projects in the world. They took this one for granted.”