About this series
This article is the second part in “Lethal Legacy,” The Post and Courier’s investigation into the nation’s plans for disposing of plutonium, the dangerous metal that triggers nuclear weapons. This installment probes the Department of Energy’s failed MOX project, an ambitious but doomed effort to clean up the legacy of the Cold War.
Part I: Why South Carolina is likely stuck with a stockpile of the nation’s most dangerous nuclear materials
Dogged by faulty assumptions and lacking political will, the federal government squandered billions of dollars and an opportunity to dispose of the nation’s most dangerous nuclear material by chasing a massive construction project in South Carolina that was doomed from the start.
The MOX saga reveals an unsettling reality of the nuclear era after the Cold War. The U.S. and the world’s other nuclear powers have proven they are capable of pulling the explosive potential out of atoms, but they have proven unable to dispose of a creation that will retain immense power and be a danger for eternity.
What is MOX? MOX, short for mixed-oxide, is a type of fuel for commercial nuclear reactors. It gets its name from the combination of two oxidized nuclear metals: plutonium and uranium.The U.S. government and Russia agreed to make MOX fuel with highly enriched plutonium, which they made for nuclear weapons during the Cold War. The idea was to make the plutonium less potent and generate electricity by reacting it in power plants; the project’s supporters described it as a way for the countries to turn their “swords into plowshares.”
Instead, the U.S. Department of Energy stranded a huge stockpile of plutonium — the lethal metal at the core of nuclear weapons — at a federal installation on the state’s wooded western edge, with plans to leave it there for decades.
The roadblocks weren’t science or technology. They were politics and poor planning.
It did so over the objection of South Carolina’s political representatives, who prolonged the demise of the plutonium processing project for years and called for Congress to keep pouring money into a program that was careening toward disaster.
This account of the project’s demise is based on more than 10,000 pages of audits, congressional testimony, federal budget documents, internal project records and lawsuits. The Post and Courier also interviewed key Energy Department officials, politicians, watchdogs and people who worked on the massive construction effort.
The newspaper’s reporting found that evidence mounted for years that the project, known as MOX, would cost billions of dollars more than initially thought and take far longer to finish. And it found that Congress and federal officials failed again and again to make a decision about its future, which was caught between spiraling cost estimates and the unified force of South Carolina politicians who saw it as a critical jobs engine.
Each year of indecision cost hundreds of millions of dollars, left in limbo enough nuclear material to build thousands of bombs and delayed the disposal of the most dangerous metal mankind has ever created, one that will take a quarter-million years to decay. Even today, almost a year after construction workers were sent home, the stockpile’s future is uncertain, sitting in the hands of the U.S. Supreme Court.
Chapter 1: A rushed beginning
The MOX project traces its roots to the turn of the century, when the U.S. and Russia signed an ambitious agreement to clean up the legacy of the Cold War.
Both countries had produced massive amounts of plutonium as tensions escalated between them, and now they had agreed to part with some of their stockpiles for good. They planned to get rid of them largely by building mixed-oxide, or MOX, fuel factories that would mix plutonium with other nuclear material to make the pellets that power nuclear reactors.
The federal government decided it would build its MOX plant in South Carolina at the Savannah River Site, a swath of federal land the size of a small county between Aiken and Allendale. The state’s leaders had lobbied for the role because the Savannah River Site needed a new mission to pursue.
The site was created to make material for nuclear weapons in the Cold War, and with its decades-long showdown with the USSR behind it, the U.S. didn’t want to keep hoarding bombs. The site might slowly shut down, and if it did, the local economy would bleed out.
Getting the MOX project came with a trade-off: The site would have to play host to tons of plutonium that was left over from making weapons and conducting nuclear experiments. It came in from California and Colorado, Washington state and New Mexico. South Carolina insisted on a promise that it wouldn’t stay for long.
But after the plutonium arrived, the U.S.’s deal with Russia got off to a rocky start. Russia dragged its feet for years, and MOX’s supporters got restless. South Carolina had received the plutonium but few jobs, and they wanted the Department of Energy to get started.
So when the logjam with Russia finally broke, the Energy Department was eager to begin. It had plenty of reason to press forward: to keep Russia on track, calm MOX’s supporters in Congress and help secure a spot for the project in the federal budget.
The agency and its contractor, a company called MOX Services, rushed to put out their first official estimates for the project. They told Congress that MOX would cost a little under $5 billion, and it would be ready by 2016.
Underneath the surface, those estimates were deeply flawed because the facility’s complicated design was nowhere near ready — somewhere between about one-quarter and two-fifths complete.
A half-baked design meant that the estimates were little more than a ballpark guess. MOX Services would later accuse the government of ordering it to “estimate first and design later.” The government countered that the contractor was still responsible for the promises it made, regardless of the circumstances. Their fight is only now nearing a settlement.
It would take years for the full extent of the problem to set in, for budget writers and lawmakers to realize that the U.S. government’s policy had been no better than “throwing darts in the night and hoping you hit a bull’s-eye,” as U.S. Rep. Donald Norcross, D-New Jersey, put it.
But on the MOX project, it was soon clear that the problem was enormous.
The flaw that would doom MOX revealed itself to Terry Mullins in a stack of paperwork that wouldn’t stop growing.
Mullins was far away from South Carolina, in a factory outside Knoxville, Tenn., a mile uphill from a spindly lake in an Appalachian river valley. His family business occupied only a small corner of the project. But what he was seeing foretold trouble for one of the biggest construction projects the Department of Energy had ever undertaken.
There were design changes, dozens of them.
MOX was a behemoth project with a mission that was at once dangerous and technically complicated. It was supposed to handle enough plutonium to build thousands of bombs the size of the one dropped over Nagasaki, Japan. Even the smallest fleck of plutonium is thought to cause cancer if it’s inhaled — an amount so small it’s measured in millionths of a gram.
The plan was to combine the plutonium with uranium, another dangerous metal, and press the mixture into pellets of fuel for commercial nuclear reactors, tens of thousands of them every day. The plutonium and uranium would come in as a fine powder. To keep it from getting out, the MOX plant would have to do its work in airtight chambers that look like something from a doomsday movie. Workers can only reach in through gloves attached to the sides, which is how the chambers get their name: gloveboxes.
The plant would need hundreds of them. Mullins’ company was hired in 2010 to make three gloveboxes, plus another sealed chamber.
It was supposed to be a straightforward job for a company that was used to designing and building complicated equipment for the government: His company wouldn’t even have to design the chambers. It would just get the finished drawings and build them.
But the designs were not ready. Within two weeks of getting the job, more than 140 design changes came in, Mullins’ company alleged in court. By the end of 2011, it had gotten more than 1,000 in all. Just paying his staff to sort through them cost almost $2 million in wages, Mullins said.
At the same time, another company across the country said it was getting drawings that didn’t show where pieces should be welded together or even what size pieces to use. Some drawings came with instructions that conflicted with each other. Gloveboxes throughout the project ended up taking months longer to make than they should have, and they cost more than expected, adding millions of dollars apiece, an impact multiplied hundreds of times over.
A top official in the company hired to build the MOX plant would later admit to Mullins’ lawyers that the gloveboxes were “inadequately designed.” Everyone in the company’s top ranks knew about the problem, he said under oath.
The issues emerging in places like Tennessee by 2010 were an alarming signal for a project that would ultimately spiral billions of dollars over budget and more than a decade behind schedule. While the MOX plant’s massive, 10-foot concrete walls rose, the equipment inside threatened the entire endeavor. Government auditors would later find that pieces like gloveboxes were one of the main reasons MOX was teetering on the edge of disaster.
The problems in Mullins’ factory were a harbinger of the problems that would doom the entire project: Gloveboxes were the protective layer that would make the MOX undertaking safe. They were designed around the machines that would handle America’s most sensitive nuclear material. The building was designed around the gloveboxes.
So if the gloveboxes’ designs weren’t ready, nothing was ready.
If they were in trouble, the entire project was.
MOX Services had seen Mullins’ problems before.
It needed to order lots of custom-made equipment, but it wanted to do a few test runs first. Before it bought hundreds of gloveboxes, it decided to order two of the simplest ones and see how they turned out.
The work went out shortly after the project broke ground at the end of 2007. Design changes poured in. The gloveboxes went far over budget, and they took nearly an extra year to finish.
The test run predicted the problems Mullins encountered before his company was hired for the job, and MOX Services would later describe the results as “profound and bracing.”
The tests quickly revealed that the project’s entire budget was unrealistic, but MOX Services had these revelations only after it had given the government its first estimates earlier in 2007.
The government and MOX Services blamed each other for the problems they uncovered: The contractor said it hadn’t been allowed to run the tests earlier because of the U.S.’s standoff with Russia. The government contends the company indicated it was “ready, willing and able to commit” to the original estimates all the same.
It would matter little who was at fault for the bad estimates. Both sides were locked in, and billions of taxpayer dollars were pouring into a project built on flawed assumptions. MOX was on a collision course with reality.
It set in again and again, with problems that beset almost every component of the facility.
It was apparent in the air vents: A company making ductwork for the building’s HVAC system realized in 2011 its workers were spending three hours dealing with design changes for every hour they spent making ducts, according to a report obtained by the watchdog group SRS Watch. They handled nearly 8,000 changes between 2009 and 2016, the cost of their work more than doubled, and they took an extra 3½ years to finish, even though the final design ended up calling for fewer air ducts.
It was apparent in the wiring: MOX Services realized in 2012 that its first estimates called for half as much electrical cable as the project would actually need. The extra cable measured 660 miles; unspooled from the MOX plant in South Carolina, it would reach within eyeshot of the Statue of Liberty.
The revelations throughout the building led the head of the Energy Department’s nuclear weapons arm to tell Congress in 2014 that his agency had made a “serious mistake” by pushing forward without a finished design. Lawmakers would order the department to forge ahead for years, led by staunch defenders from South Carolina who were mindful of the plutonium stockpile back home and the economic impact of losing a big project and more than 1,000 construction jobs.
Problems with MOX and other major projects were so profound that the weapons agency, the National Nuclear Security Administration, made an official policy blocking itself from starting another big nuclear project until it has a design that was all but finished.
It set that rule years before the MOX project met its demise, but it was already too late for MOX. The NNSA told The Post and Courier that when it started construction on MOX, the facility’s design “was unquestionably less than our current standard.”
It was only the first time a bad assumption would weigh on the project.
Chapter 2: Faulty assumptions
Before the U.S. government decided to build a MOX plant in the 1990s, engineers and policy wonks considered all kinds of moonshot ideas for getting rid of plutonium, like blasting it into the center of the sun or drilling it into the Earth’s fiery center.
MOX looked straightforward by comparison. It was a technology that had been used before, and officials thought it would be relatively inexpensive. In the program’s early days, the NNSA assured Congress that it had a “high degree of confidence” that the project wouldn’t cost more than a few billion dollars because other countries had built MOX plants decades earlier.
What it didn’t account for was all the differences between what those countries had done and what the U.S. wanted.
Other countries like France and the U.K. built MOX plants to deal with plutonium from their nuclear power plants; the U.S. wanted one to handle the highly enriched metal used in weapons, a much more sensitive job.
Other countries had built their plants under more lenient rules; the U.S. now had strict regulations for building nuclear facilities. They were intended to prevent disaster, and difficult to meet.
The American MOX plan called for a process that was almost entirely automated to minimize workers’ exposure to plutonium. Machines would take over from when plutonium was loaded into the plant to when pellets of fuel came out the other end.
Even before construction began on the plant, outside consultants and government auditors criticized the Energy Department and MOX Services for failing to appreciate the challenges they faced. An internal department review later found that the Energy Department had neglected to study other MOX plants and have outsiders challenge its assumptions.
By the time its assumptions began to crumble, it was too late.
The government had already signed an agreement with Russia promising to use the technology and shipped tons of plutonium to South Carolina with a promise not to leave it there. It had even decided not to pay for a second processing plant built around a different technology.
It was stuck with MOX alone. At the time, it had assumed it wouldn’t need a backup plan.
There was one big question the U.S. failed to ask before it embarked on the MOX project: Was America still capable of building a big nuclear facility?
In the decades since other countries had built their plants, the nuclear industry in the U.S. had been crumbling, a revelation that came to Anne Harrington at so many meetings with MOX Services.
Throughout most of the Obama administration, Harrington was in charge of the Energy Department’s efforts to stop the spread of nuclear weapons, a job that put her in charge of the nation’s most ambitious nonproliferation effort: MOX.
She started the job in 2010, a few years after construction began, and she met with the project’s contractors every few months, each time hearing again how the American nuclear industry was in disarray.
She’d hear about how Americans hadn’t even tried to build a commercial nuclear project in the better part of three decades — not since the near-disaster at the Three Mile Island nuclear plant in Pennsylvania.
She’d hear about how few workers and companies could meet the industry’s exacting standards and the U.S.’s extremely tight regulations — the way concrete had to be poured without air bubbles, the way parts had to be welded just so — with an inspector looking behind.
She’d hear about how the project needed to hire engineers to work on suppliers’ shop floors to make sure they were meeting standards, a conversation she said happened “way more than once.”
MOX Services and the government both had assumed that there would be plenty of companies to do the work they needed. Instead, MOX Services found that the nuclear industry had “atrophied,” and what was left of it was now in high demand.
Soon after the Energy Department started work on the MOX project, two groups of utilities in South Carolina and Georgia set out to build the nation’s first nuclear reactors in a generation, expanding the V.C. Summer power plant near Columbia and the Vogtle plant in Georgia. Both endeavors broke ground within 75 miles.
They all had to fight for workers with skills the government came to regard as “nearly nonexistent,” requiring lots of on-the-job training. MOX got a reputation as a nuclear training ground, a stepping stone for the industry’s newcomers. Some years, one worker in five would leave MOX for another job, forcing the project to pay for a new employee’s training.
“The decision-makers at the time naively thought that U.S. capacity was sufficient to handle this kind of large nuclear project,” Harrington said. “This program helped bring back lost skills, but it brought them back at a huge cost.”
With their newfound leverage, suppliers insisted that MOX Services and the government bear the risk of their work going over budget. MOX Services once approached 72 companies to see if they’d give a fixed price to pour concrete for the project. Only one made an offer, padding it with extra money to blunt the risk.
With lacking expertise and limited options, the project used low-quality parts that didn’t meet safety standards at least three times in the first year of construction, the department’s inspector general found. In one case, they allowed a company to pass off regular rebar as nuclear-grade, a hardened type meant to protect the building in a disaster.
The low-quality rebar was discovered in 2008, only after a worker swinging nothing more than a hammer managed to snap a bar that was supposed to withstand earthquakes and explosions. By then, MOX Services had already bought more than 3,000 tons of the bad rebar and entombed nearly 150 tons of it in concrete.
Auditors criticized the NNSA for years for its loose oversight of the MOX project, starting even before the facility broke ground.
They said the staff in charge of the project was “generally inexperienced” with managing complicated construction projects, that the agency waited years to set up a dedicated project management office for MOX, and that the department had understaffed its oversight efforts.
The hands-off approach was especially problematic because the government’s contract with MOX Services left it on the hook for the project’s mounting problems. It was written so that the government essentially covered any cost the project racked up, plus a fee for the contractor. Auditors found that the department had been warned about the risk when it signed the contract in 1999.
“We would not enter into a contract like this today,” the NNSA’s top project management official, Robert Raines, said in 2017.
It was an old story for the Department of Energy, which has a decades-long reputation for letting projects spin out of control, ending in their cancellation or massive overruns. It had encountered problems with efforts to restore nuclear weapons, clean up environmental messes and build factories and research facilities, among others.
Gregory Friedman, a retired former inspector general for the department, said MOX was a “prime example of grossly inefficient and ineffective use of scarce taxpayer-provided resources” brought on by a “perfect project management storm.”
In an email to The Post and Courier, he said the department was influenced by international pressures and local economic concerns, it failed to estimate costs well or control them, it didn’t keep in check changes to the project, and it didn’t keep close enough tabs on its contractor.
“These factors, to one degree or another, were part of the pattern of failed projects at a number of departmental sites that we examined when I was the inspector general,” Friedman said.
But this time, the NNSA was trying to improve the way it manages projects. Auditors would later observe that the NNSA’s decision to invest in project management would lead it to a conclusion that upset decades of work on plutonium disposal and nuclear nonproliferation.
The agency began to think that the MOX project’s problems ran too deep, that its challenges would be too expensive to overcome.
The soaring cost of MOX then ran headlong into a federal budget crisis.
Revelations about the project’s challenges mounted at a moment when the Obama administration was deciding it wanted to modernize the nation’s nuclear complex and work on making its aging weapons last longer — a costly, long-term endeavor.
And the problems piled up as money was getting tight across the government. Congress had been gridlocked over spending, and lawmakers had let a series of deep budget cuts known as sequestration take effect. They squeezed every department’s budget like a vise.
The budget cuts forced the administration to pick its priorities, to weigh its plan to get rid of plutonium in South Carolina against its hope of investing more in nuclear weapons. Former department officials said in interviews that as they focused on cutting costs, their attention turned to MOX, one of the biggest costs they had.
“If money were not an issue, MOX would be a straightforward way to go. But, unfortunately, as you have indicated, it is an issue,” John MacWilliams, the department’s chief risk officer, told Congress in 2015.
The Department of Energy decided that the money shoveled into MOX would be better used elsewhere. Its leaders decided that they needed to kill MOX.
The question was, would they?