The U.S. nuclear fuel Gordian knot: From global supplier to vulnerable customer

The federal government has always had an uncomfortable relationship with providing goods and services to the private sector. It has been dithering over privatizing air traffic control for decades. It set up the Post Office as the U.S. Postal Service, a quasi-governmental entity, and almost nobody is happy with the result. It set up Amtrak in a similar way, but the passenger rail service continues to disappoint free market advocates and is effectively on the government dole. And it cut its ties to the enrichment enterprise—arguably not so much freeing it as setting it adrift. And that hasn’t worked as intended, either.

How did it come to this?

We find ourselves in this position as a result of the general reluctance of Washington to get into industrial activities or even industrial policy, even if that means leaving critical industries adrift in an economic environment too rough for the private sector.

Enrichment for the civilian industry began as an add-on to the work of a complex of three giant factories using gaseous diffusion, a technology invented by the Manhattan Project, the wartime sprint to build an atomic bomb. After the war, the United States, pursuing President Dwight Eisenhower’s “atoms for peace” vision, offered enrichment services for civil use to allies around the world.

The government was at first so protective of the enterprise that private companies were not allowed to own enriched uranium—they could only “use” it. Even when they took delivery and put the fuel in a reactor, companies technically were only leasing it from the Atomic Energy Commission (the agency that absorbed the Manhattan Project and a predecessor agency to today’s Department of Energy). It was not until August 1964 that President Lyndon Johnson ended the federal monopoly on owning enriched uranium, allowing private entities to buy it from the government. In 1973, private ownership of fuel became mandatory, according to a July 1983 history on the AEC compiled by the DOE.

Uranium fuel is hardly the only critical energy ingredient that the United States doesn’t supply for itself; rare earth elements, which are essential for assorted energy technologies, are nearly a Chinese monopoly. The nuclear fuel business, however, is the only manufacturing process that the U.S. government deliberately chose to exit. It wasn’t controversial at the time; among members of Congress and the top officials of the Nixon and Ford administrations, there was a consensus.

No good at running factories

One point most agree on is that the government wasn’t good at matching supply to the projected demand. By 1969, as the number of civilian nuclear reactors was on the rise, the General Accounting Office, as the congressional audit organization was then called, was expressing skepticism that Congress would fund expansion of gaseous diffusion plants on a schedule that provided enough production to meet domestic and international market demands. (But government officials weren’t sure that private companies could raise enough money to do that, either.)

Frank G. Zarb, who served in President Richard Nixon’s Office of Management and Budget and was later the “energy czar,” took the position that enrichment did not need to be a government enterprise and that “the responsibility for providing additional capacity for the 1980s and beyond can and should be undertaken by private industry.” In June 1971, Nixon called for private enterprise to “eventually assume responsibility for uranium enrichment.” In 1974, Zarb told President Gerald Ford that the AEC couldn’t keep up with projected demand.

There was one prominent objection based on national security—but not the kind of national security problem we are facing now. Henry Kissinger, when he was the White House national security advisor, warned that getting the private sector involved in uranium enrichment could create proliferation problems. He recognized, however, “the budgetary implications of any new governmental construction and the desire to minimize direct government involvement in commerce.”

Craig Hosmer, who in 1974 was a Republican congressman from California and former AEC lawyer, was advocating for a government corporation, but Zarb thought that would cost the government too much money.

James R. Schlesinger, then AEC chairman, said that having industry take on the enterprise was “related to the federal government’s fiscal capacity and the many demands on that fiscal capacity coming from other sources.”

“This is an area that private enterprise can handle,” he said. The main points of contention were how much the federal government should charge for the enrichment enterprise and how to split the cost of decommissioning the gaseous diffusion plants, located in Paducah, Ky., Piketon, Ohio, and Oak Ridge, Tenn. (The money for completing that task still hasn’t been found.)

The government’s indecision created problems. On the promise of access to American enrichment, companies in Europe began building nuclear reactors, but in the late 1980s, the DOE became convinced that it could not supply all of those customers and closed its order books. The result was two new European competitors: Eurodif, a French firm, and Urenco, a British-Dutch-German consortium. Those companies began with a more modern technology—centrifuges—that had an energy requirement about 90 percent lower than that of gaseous diffusion. (Electricity was 80 percent of the production cost at the gaseous diffusion plants, but since they were already running, simply paying the electric bill seemed better than building a newer-technology system.)

The market went from famine to feast, at least from the customers’ point of view. By 1991, the GAO estimated that enrichment capacity outside of the Soviet Union exceeded demand by 25 percent. Counting Soviet capacity (which nobody could do precisely), the supply might exceed demand by 60 percent, the GAO said. That was a tough environment in which to launch a competing technology. The fact that the Soviets, eager for hard currency, offered materials for export at prices below what a Western competitor could offer made the problem even tougher. In their time, the Soviets—and even today, the Chinese—subsidize production of various materials. Sometimes trade organizations decide that exporters are “dumping” the products (that is, selling in the foreign market for less than the price of production). As punishment, they impose tariffs or duties or quotas. Since 1992, Russian uranium exports to the United States have been limited by a suspension agreement, following a conclusion that uranium was being dumped here.

Competing with giant players backed by government treasuries—and thus the ability to sell products below cost—is a challenge for any private company. Oddly, in GAO reports and congressional testimony from the 1990s, there is not much mention of how a private company with far less access to capital would weather the huge shifts in price resulting from changes in supply and demand. Nevertheless, the United States somehow arrived at a conclusion different from that of every other country with a civil nuclear sector: Enrichment should be private.

The case for going private

One of the factors that moved Congress toward privatization was a report from the Wall Street firm then known as Smith Barney, Harris Upham & Co. Its 1989 Uranium Enrichment report said that the enterprise should be privatized to improve its efficiency and make it more competitive.

A July 1990 hearing of the House Committee on Science, Space, and Technology made clear the concerns at the time: The U.S. enrichment enterprise had gone from a monopoly to having only 45 percent of the world market, and political interference with its operations was preventing it from making businesslike decisions.

“If we don’t act now to prepare for the changing climate, the uranium enrichment enterprise will not accompany us into the 21st century,” said subcommittee chair Rep. Marilyn Lloyd (D., Tenn.). Congress did vote to privatize, but that didn’t solve the problem. The enrichment enterprise was spun off into corporate hands, but in May 2013, the complex’s last enrichment plant, a 1950s gaseous diffusion plant in Paducah, Ky., closed.

In the early 1990s, the issue wasn’t partisan. The ranking minority member, Rep. Sid Morrison (R., Wash.), said, “As a result of lack of market discipline our industry has been saddled with many bureaucratic and non-cost effective requirements.”

“By moving to a more businesslike basis, we will be able to deploy new technologies, which in turn will enhance and strengthen the commercial position of this important energy sector,” he continued.

Meanwhile, the watchdog group National Taxpayers Union complained, not that privatization was a bad idea, but that it left the federal government with legacy costs.

Some people favored establishing a quasi-independent status for the enterprise called USEC, for United States Enrichment Corporation, and giving it autonomy and flexibility while keeping it within the government. William Magwood IV, who was at the time a midlevel staff person at the DOE, recalled that utilities were not happy with the government enterprise. It lacked flexibility and did not understand the market, he said. (Magwood later became a commissioner of the Nuclear Regulatory Commission and is now director-general of the OECD Nuclear Energy Agency.)

“Those of us in the middle thought a government corporation was the answer,’’ he said in an interview. “There was a privatization clause in the law, but no one expected it to be used.” One reason, according to Magwood, was that a private enterprise was unlikely to function well in the nuclear fuel market. Government had always had an outsized influence on the market via decisions about how much uranium it should allow to be imported, or how much it should buy or sell, he said.

Leaving the enrichment business in government hands would have allowed easier access to Treasury financing, although there is no guarantee that Congress would have authorized the necessary investments.

But USEC went from quasi-governmental agency to private company. It happened, Magwood said, because “the Wall Street guys came in and saw a chance to get rich.” In fact, once privatized, it paid executives markedly more. And the private companies that managed its initial public offering in August 1998, which raised $1.4 billion, also took a cut.

In fact, privatization was the brainchild of the private sector, based in large part on the Smith Barney report. Nicholas Timbers, then managing director of the investment house, was the project leader for the financial assessment that led to the report. When the dust settled, Timbers became the CEO of the privatized entity. Today, in retirement, he makes it clear that uranium enrichment was not working well as a government business, but that the current setup isn’t successful either.

Government, according to Timbers, wasn’t set up for running a manufacturing industry. The gaseous diffusion plants were run by materials company Martin Marietta, supposedly with a few government bureaucrats watching. In reality, Timbers said in a telephone interview, it was being run “with 435 people looking over your shoulder,” referring to the members of Congress. There had been no reorganization during 50 years of operation, he said. “They hadn’t lost a single employee in 50 years. You start doing manpower reductions, and you’ve got governors, senators, and congressmen asking what the hell you’re doing.”

One problem was that like all AEC and DOE installations, it was run on a cost-plus basis, he said. “I’d never heard of that before. But whatever your costs are, it was that plus a profit for the contractor.”

“I said, ‘So what’s the incentive to save money?’ And they looked at me like I was from the moon. There is none. The job isn’t to save money and be profitable; it’s to provide contract work for the workforce.”

The accounting system couldn’t track costs, Timber continued. “They didn’t care about accounts receivable and accounts payable. Whatever Martin Marietta wanted, we gave it to them—plus a profit. That’s not the way to run a business.’’

And the government did not want to come up with the money to develop centrifuges or some other more modern enrichment technology, he pointed out. “I thought it was the right decision to privatize at that time, based on what our prospects were,” he said.

Timbers’s view of the operation of a government-owned, contractor-operated enterprise is hardly unusual. The nuclear weapons production complex and other so-called “GoCos” are routinely criticized for inefficient operations.

Improving operations at the gaseous diffusion plants was never going to be more than a short-term help, because those plants were antiquated. The solution was to replace them—something that the privatized entity knew it had to do—but no one was able to accomplish that before the old plants were driven out of the market.

Searching for new technology, and failing to find it

It wasn’t for lack of trying. Work on a replacement began long before privatization. The GAO reported that in 1985, when the DOE shut down the Oak Ridge gaseous diffusion plant, it also halted construction on a centrifuge plant on which it had spent $3.5 billion, that was meant to be the replacement.

The consensus then was that another technology, called Atomic Vapor Laser Isotope Separation (AVLIS), was a better bet. The technology was supposed to require even less energy than centrifuges as well as less chemical processing of the uranium before the enrichment step.

The Smith Barney report agreed that AVLIS would be highly competitive, although it disagreed about when it would be ready.

Today, laser enrichment still is not commercial. Invented at Lawrence Livermore National Laboratory, it worked at bench scale. “We were quite optimistic about that,” said Timbers recently, but “we never got it to a commercial level where we could scale up.”

In 1996, USEC bought exclusive rights to the SILEX process (Separation of Isotopes by Laser Excitation) but dropped the idea in 2003. Later, GE-Hitachi created the company Global Laser Enrichment to commercialize SILEX. Today, the company is co-owned by Canadian uranium mining company Cameco and an Australian firm. (Details of the process have not been publicly revealed, but it would potentially require less capital to build and less energy to run than AVLIS.)

USEC tried again on centrifuges, starting work on a plant at Piketon, Ohio (the site of the former gaseous diffusion plant), with a goal of initial commercial operation by 2009 and full operation by 2012.

It sought $2 billion in loan guarantees from the DOE; instead, the DOE made a loan to AREVA (now called Orano), a French-controlled company and a participant in the Eurodif consortium. The loan provided funding to build an enrichment plant called Eagle Rock, in Bonneville County, Idaho. The NRC gave that project a license in October 2011—just over six months after the Tōhoku earthquake the previous March. That event destroyed the Fukushima Daiichi plant and shut down Japan’s nuclear fleet, delayed work on reactors under construction around the world, and made future reactors look less likely. So, Eagle Rock was never built. (Had the loan gone to USEC, that project would have faced the same headwinds.)

USEC’s primary business, uranium enrichment, was clearly winding down, but the government was not willing to let it die.

In 2012, the DOE arranged a complicated four-way swap of uranium that gave USEC the feedstock it needed to keep its last gaseous diffusion plant—the Paducah facility—running for an extra year. And in 2013, the DOE transferred about 48 metric tons of low-enriched uranium to the company in exchange for 409 metric tons of uranium hexafluoride. USEC gained 399,000 separative work units in the transaction, a DOE contribution to the development of USEC’s proprietary centrifuge.

However, with the failure to field a technology to replace gaseous diffusion, USEC stopped enriching uranium in May 2013. The Paducah plant got a brief stay of execution, and spent its final year making fuel for the Tennessee Valley Authority’s Watts Bar reactors, which makes tritium for nuclear weapons. (The TVA irradiates lithium targets and sends them to the DOE at Savannah River, where the tritium—the “H” in an “H-bomb”—is harvested. Under international treaties, the DOE cannot use uranium enriched by commercial plants to make weapons fuel.) In early 2014, USEC entered bankruptcy. When it reemerged, it had been reorganized and renamed Centrus Energy.

“Maybe the better thing was to keep it as a government corporation, let it run like a business, show a profit, and hire and fire people,” said Timbers, who had also worked on the establishment of Conrail, which was pieced together from the assets of several bankrupt railroads. “Let’s not be second-guessers for 30 years ago,’’ he said. “Hindsight’s always great, but at that time, it seemed like the right thing to do.”

If at first you don’t succeed, fail again

In 2019, Centrus tried again, with a contract from DOE to build a cascade of 16 centrifuges at Piketon to demonstrate production of HALEU, that is, high-assay low-enriched uranium containing up to 20 percent U-235. Centrus announced in February 2023 that it had completed construction and initial testing of a demonstration cascade and was expecting to begin production by the end of this year. The announcement noted that this would be the first time in 70 years that an enrichment project using U.S. technology would be starting up. However, the endeavor is far short of commercial scale. (The sole enrichment operation in the United States currently is a Urenco plant in Eunice, N.M., on the Texas border, which started delivering enriched uranium hexafluoride in 2012.)

One irony of the decision to privatize enrichment is that it did not in fact get the government out of the business. Washington has had to intermittently transfuse it with resources, including uranium and cash, to keep it going. In addition to pre–financial restructuring transactions with USEC, discussed above, the DOE recently announced a $150 million cost-share program with Centrus to demonstrate production of HALEU at Piketon.

The attack of the black swans

So where did private enterprise go wrong in the quest to modernize? According to Magwood, USEC was just too slow in developing its centrifuges—and developing a new technology is already an inherently slow process, he said. “You have to show it works reliably for a long period,” he said. “The competition ate up the market while they were doing what they were doing.’’

There were other factors, too, that caused drastic changes on the supply and demand sides. One, in particular, pushed the enrichment problem further into the future. In February 1993, the United States, throwing a lifeline to the fledgling Russian democracy headed by Boris Yeltsin, signed an agreement to buy high-enriched uranium that had been intended for Soviet bombs. The uranium was downblended to below 20 percent, the legal dividing line between HEU and LEU, and shipped to USEC, which downblended it further to the level commonly used in reactors, 5 percent. Over the next two decades, this enrichment met half the needs of the U.S. nuclear energy industry.

USEC gave away clear plastic paperweights commemorating the downblending, with a colorful logo that said “Megatons to Megawatts” in English and Russian and the two countries’ flags—both red, white and blue—rippling in the breeze. In the waiting room of USEC’s Bethesda, Md., headquarters, glass display cases are filled with samovars, decorated Easter-style eggs, and other Russian trinkets that had been exchanged at various meetings.

USEC earned money as the “executive agent” for the deal, keeping it solvent while it tried to develop a new technology, said Timbers.

Backers referred to that project as an embodiment of the prophesy of Isaiah, who wrote about 2,800 years ago that men would beat their swords into plowshares. But Isaiah was a better prophet than he was a commodities economist; that project was one of the factors that has made investment in the fuel cycle a nightmare from the corporate point of view. The problem is that the reactor business is a roller coaster.

For a while, the roller coaster seemed to be climbing another hill. The price of natural gas got so high in the early 2000s that several utilities dusted off plans for new reactors, and it looked like a “nuclear renaissance” was around the corner. Then came fracking and a price collapse, and what might have been a dozen reactors turned into just two, Vogtle-3 and -4, just now coming on line. Then the Tōhoku earthquake, at the bottom of the sea 45 miles east of Japan, not only wrecked the Fukushima Daichi reactors but essentially shut down Japan’s nuclear fleet. (Ironically, unused fuel from those reactors could help rescue American plants if Russian supplies are lost to the market.)

Now nuclear energy is hot again, as utilities serving more than half of the United States have pledged to cut their carbon emissions to zero or near zero by midcentury but don’t know how to hit that goal. In addition, most of the advanced reactors now in the pipeline are designed for fuel enriched closer to 20 percent instead of the usual 5 percent. That implies more demand for uranium ore and for conversion and enrichment—if that technological revolution actually happens. This situation calls for the government to step in and stabilize commodity prices for uranium, conversion, and enrichment, according to Craig Piercy, executive director and chief executive officer of the American Nuclear Society. The government could, for example, promise to buy uranium or separative work units at a floor price to ensure that the bottom does not fall out of the market. This is an investment in the literal, financial sense; the DOE could earn its money back later by selling at times of need. (The DOE has an LEU reserve of about 230 metric tons, enriched to 4.95 percent. That’s enough for about six core reloads for a 1,000 megawatt reactor.)

“The federal government runs umpteen other programs, like flood insurance or agricultural price supports,” Piercy said. “It backstops private industry in a bunch of different areas.” The government also stockpiles oil to guard against supply interruption, he pointed out. The Strategic Petroleum Reserve has a capacity of 714 million barrels and is about half full. (The country burns about 20 million barrels a day.)

Establishing a floor price and building up a reserve would be “bounding the risk,” Piercy continued. “Sort of like what Price-Anderson [Nuclear Industries Indemnity Act] did for plants, we need a framework for fuel.”

Companies in the mining, conversion, and enrichment business resolutely refuse to do so without contracts in hand. It will take deeper pockets than theirs—namely, government involvement (keep an eye out for part three in this series on nuclear fuel where we look at how Congress dropped the ball last year and what it can do to fix the issue)—to expand the industry and weather the threat of future price fluctuations, which could come from a natural disaster or a technological advance in a competing energy source—like the two in the recent past—or from a changed international situation that brings Russia back into the West’s good graces.

For private companies, these three market-shattering episodes—fracking, Fukushima, and Ukraine—appear like “black swan” events, rare and unpredictable. Others could argue that an earthquake that damages a reactor, a leapfrog in a competing technology, or a war are hardly unpredictable. In any case, at this point, these events are creating a new consensus: that institutions bigger than corporations—namely, governments—should be involved.

Matt Wald is an independent energy writer and consultant. He is a former policy analyst at the Nuclear Energy Institute and for decades was the energy reporter at the New York Times.