Between MOX and a hard place

It costs more, it’s as dangerous to make as a bomb, and burning MOX creates almost as much plutonium as it gets rid of. Other than that, it’s a great idea.

In September 2000, the United States and Russia signed an agreement to dispose of 68 metric tons of weapons-grade plutonium (34 tons each). The agreement called for 25.5 tons of U.S. plutonium and all of the Russian quantity to be converted into a more diluted form suitable for use in mixed oxide (MOX) fuel in commercial nuclear power reactors. The remaining 8.5 tons from the United States was to be immobilized—until the government decided in 2001 to burn that as fuel as well.

The MOX program elicited controversy well before Vice President Al Gore and Russian Prime Minister Mikhail Kasyanov made it official in 2000. Critics complained that MOX jeopardizes reactor safety and does little to actually reduce plutonium stockpiles, since spent MOX fuel can be separated and the plutonium reused in bombs. The plan, they said, affords terrorists or rogue states greater opportunity to obtain plutonium. Plus, the program is expensive—more than $6 billion to build MOX fabrication facilities in Russia and the United States and convert reactors to handle the new fuel. Experts predict it could take 10 years to retrofit those reactors, extending the disposal process for another decade.

Three years of indecision and foot-dragging by both sides has stalled the MOX agreement and done nothing to quell the debate. And now, at least one alternate plan has emerged: a thorium-plutonium fuel design being developed in Russia with the support of private and government funding from the United States. This new fuel promises significant advantages over MOX, including lower costs, a faster rate of plutonium usage, and greater proliferation resistance (see sidebar).

With a potential alternative on the horizon, the high cost of the MOX fuel plan, the dangers associated with its production, and the slow progress it will make toward reducing global stockpiles of plutonium make it timely to ask: Isn’t there a better way? A closer examination of MOX fuel economics and MOX experiences in other nations clearly demonstrates its inherent limitations.

A brief history of MOX

MOX fuel is an integral part of the closed nuclear fuel cycle, where plutonium is irradiated, reprocessed, and reused. It is regarded as a stepping-stone toward commercial development of the more efficient fast-breeder reactor, which could create more plutonium fuel than it consumes.

To make MOX fuel, plutonium, in the form of plutonium oxide, is mixed with uranium oxides and pressed into ceramic fuel pellets that are loaded into seamless fuel rods. The technology was pioneered by the United States during World War II. The fabrication of MOX fuel is considerably more complicated and more hazardous than that of uranium fuel. Due to the extremely high radiotoxicity of plutonium, it must be processed in isolation: Protective barriers, including remotely operated equipment, need to be built and maintained. After the ceramic MOX pellets are encapsulated into fuel rods, the handling of MOX fuel is essentially the same as that of conventional uranium fuel.

MOX fuel demonstration programs began in the United States and Europe in the 1960s. By the mid-1970s, the U.S. nuclear industry was ready to implement a large-scale MOX utilization program. However, in 1977, concerns over plutonium being diverted to terrorist groups or rogue states caused President Jimmy Carter to issue an executive order that indefinitely deferred the reprocessing of spent nuclear fuel. The administration hoped that other nations would follow its lead and also ban reprocessing. But the effect of the executive order was to halt the development of MOX and fast-breeder reactor technologies in the United States, while other countries proceeded to implement fuel reprocessing in their commercial nuclear programs. As a result, the United States today faces a serious shortage of personnel experienced in evaluating, designing, fabricating, and irradiating MOX fuel.

The experience of U.S. fuel designers, fuel manufacturers, and energy-producing utilities is based on uranium oxide fuel—the type used in commercial U.S. reactors. Working with MOX is entirely different than working with uranium oxide. Scientists must consider different safety, manufacturing, and performance requirements with MOX, starting in the reactor design process. Differences in nuclear characteristics mandate new design strategies to deal with problems like control rod considerations and local power peaking. For example, important parameters such as the moderator temperature coefficient or the Doppler coefficient (both of which affect the stability of the core during accidents) are more negative in a MOX core than in a conventional core, and a MOX core operates with hotter fuel rods.

In addition, the thermal mechanical behavior of MOX fuel is not as well understood as that of uranium oxide fuel. The computer models used to predict the evolution of uranium oxide fuel during irradiation have been extensively benchmarked against data obtained from examining irradiated commercial fuel elements. The data from MOX fuel are much more sparse and were obtained largely in experimental reactors.

One of the parameters calculated with fuel performance codes is the amount of fission gas produced in the fuel and the fraction of this gas that migrates outside the ceramic fuel matrix. This parameter plays a very important role in establishing the safety of the fuel: The fission gas released by the ceramic pellets can exert pressure on the cladding wall, and could potentially cause the fuel rod to increase its diameter and block coolant flow if the gas pressure exceeds the coolant pressure. Because of uncertainty with MOX fission gas release behavior at higher burnup, and the potential impact of high fission gas release on plant safety, the discharge exposure of MOX fuel is not allowed to be as high as that of conventional uranium oxide fuel. Since the higher the discharge burnup, the more economical the fuel, MOX is less economical than uranium oxide.

During irradiation, the cladding barrier on some fuel rods may breach, contaminating the coolant. Aggressive programs by utilities and fuel manufacturers have minimized the number of rods with breached cladding, but the potential is a continual concern. To decrease dose rates, utilities stop operations in mid-cycle to search and remove failed rods at a significant cost. Because the reliability and post-failure behavior of MOX fuel is not as well understood as that of uranium oxide fuels, it cannot be asserted that fuel failures will not have more severe consequences than those that have already been experienced. And that could mean higher operating costs.

The MOX manufacturing process is also more complicated than that of uranium oxide fuels. A major concern is to keep people separated from the plutonium until it is encapsulated in sealed fuel rods. Effective barriers such as glove boxes (closed containers with double-walled gloves that allow operators to manipulate dangerous substances from the outside) must confine the plutonium. Fabricating ceramic pellets involves using oxidized plutonium powder, and grinding operations create fine dust particles. Both necessitate directionally controlled airflow to limit the spread of airborne contaminants. Automatic equipment must be used to manipulate fuel samples for quality control verifications. The safe storage of plutonium mixtures during fabrication requires different containers to prevent criticality accidents. The shipping containers need to be relicensed to account for the presence of plutonium in the fuel rods. All of these complications increase the cost of fuel manufacturing.

Commercial feasibility of MOX

The current price of uranium, in constant dollars, is less than half the price it was 25 years ago, when the economy of the MOX fuel cycle was being evaluated. Projections for uranium demand have not materialized, and discoveries of new sources of uranium ore, combined with a worldwide overcapacity of enrichment facilities to produce low-enriched uranium for standard light-water reactors, have contributed to the collapse of uranium prices. Many uranium mines have closed because prices are too low.

Even if natural uranium cost $700 per kilogram more than it does now, accounting for reprocessing and enrichment, it would still be as cheap or cheaper than MOX fuel. The production of commercial MOX fuel would need to increase considerably for economies of scale to lower the production costs enough to make its unit price competitive with uranium. But the current consumption of MOX fuel represents only about 2 percent of the fuel used by nuclear utilities worldwide, and commercial prospects for increased demand are dim based on recent developments in Britain and Japan.

Japan has been the largest customer of MOX fuel. Until recently, its interest in MOX was perceived by many as a sign of a renaissance for the industry and the only realistic new MOX market. The country’s energy strategy aimed at using plutonium stockpiles to power all its nuclear power plants, and Japan was even contemplating building its own MOX fuel fabrication facilities. However, numerous scandals and accidents over the last four years have sparked an overwhelming public anti-MOX movement.

In 1999, it was discovered that British Nuclear Fuels Ltd. (BNFL), the British fuel manufacturer, had falsified quality control data used to demonstrate the acceptability of a shipment of MOX fuel to Japan. The fuel was returned to Britain, and BNFL lost credibility as a reliable supplier for the Japanese market. In September 2002, Tokyo Electric Power Co. (Tepco), the largest nuclear utility in Japan and the world’s largest privately owned electric utility, admitted that General Electric had found cracks in reactor core components during routine maintenance work. Tepco management kept this information secret for two years, but General Electric reported its findings to Japanese regulators.

As a result, Tepco had to close 13 of its 17 nuclear reactors, which had been supplying 44 percent of Tokyo’s electricity. The company may still shut down the rest of its nuclear fleet in order to perform additional safety checks. Tepco has already announced that it is delaying the irradiation of MOX assemblies indefinitely.

In a different type of accident, in September 1999, workers in the Tokaimura fuel-processing facility operated by the Japanese company JCO were exposed to lethal doses of radiation due to a criticality accident. The accident occurred because the company apparently violated safety procedures in order to meet increased production requirements, and workers were allowed to mix batches of uranium, enriched to 18 percent uranium 235, into an unsafe geometry. Since the higher level of uranium 235 enrichment is required by fast-breeder reactors, the accident in Tokaimura may cause Japan to abandon its breeder program.

Problems in the MOX business are not limited to Japan. Belgian MOX fuel manufacturer Belgonucleaire has been having troubles since 1994, when the Belgian government decided to stop reprocessing for MOX. In 1998, a national court denied Belgonucleaire a license to build another MOX fuel fabrication plant. In 1999, after BNFL falsified data on MOX fuel shipped to Japan, Belgonucleaire was subjected to numerous reviews and technical audits. The recent Tepco incidents have eliminated a large potential market and may challenge the company to keep its MOX fuel fabrication plant and other facilities operational.

France has a well-established MOX program, yet it faces problems associated with MOX fuel’s limited burnup capability, the high cost of fabrication, and the MOX troubles in Japan, which France was hoping would be a large new market for French global energy concern Cogema. The French nuclear regulator limits the maximum number of MOX rods in each assembly, the maximum plutonium content, and the discharge burnup. These safety limitations make MOX fuel less economically attractive than uranium oxide fuel. In June 2001, Electricité de France, the French national utility, submitted a request to the licensing authorities to relax the MOX limits, but at the time of this writing, the limitations had not been lifted.

In Germany, after spending approximately $700 million building a MOX fuel fabrication plant in Hanau, nuclear utilities and Siemens, the German fuel fabricator, decided to dismantle it completely because of concerns about its future, given the anti-nuclear stance of the German government. The dismantling of the plant began in October 2001.

In Britain, BNFL recently completed and opened a new MOX facility at Sellafield, spending upward of $400 million. No fuel orders are being placed at the new plant, and the latest shipment of BNFL’s MOX fuel destined for Japan was returned due to the Tepco scandal. Additionally, negotiations to sell six BNFL reactors and a MOX reprocessing plant to British Energy failed due to that company’s insolvency.

Money matters

The economics of getting the U.S.-Russian MOX fuel agreement off the ground are proving just as questionable as existing MOX operations. The cost of the U.S. disposition program is estimated at $4 billion, while the projected cost of the Russian program is between $1.8 billion and $2.5 billion. Many experts believe that the combined costs will escalate from the projected $5.8–6.5 billion to more than $10 billion. The Russian government agreed to the plan only if the U.S. government, with assistance from other countries in the Group of Eight (G8) most industrialized nations, funded the project. The United States has pledged $400 million toward Russian costs, and another $400 million is expected to be donated from other nations, although the June 2003 G8 summit in Evian, France, ended without any mention of the MOX project, much less a pledge to fund it.

In June 1999, the Energy Department awarded a $130 million base contract to Duke Cogema Stone & Webster (DCS), a consortium of companies equipped to support Energy’s mission to dispose of surplus plutonium. Under the base contract, DCS was to provide full-scope services including design, construction management, operation, and deactivation of a MOX fuel fabrication facility. Construction was scheduled to begin this year at the Savannah River Site near Aiken, South Carolina, but has been pushed back to 2004. And at the time of this writing, the construction permit had not been issued. Given the antagonistic attitude between the U.S. and French governments during the war in Iraq, it is unclear whether Energy will want to continue its association with Cogema in this project, in particular when BNFL was also one of the bidders for the U.S. and Russian MOX contracts. Changing contractors would cause further delays and raise costs.

Other unresolved issues between the United States and Russia could also delay the program’s implementation. Because of Russia’s sensitivity to maintaining the secrecy of the isotopic composition of its weapons-grade plutonium, the country plans to mix it with reactor-grade plutonium before making it available for the program. Such a plan creates verification problems. According to Michael Guhin, U.S. fissile material negotiator for the State Department, the Russians and Americans have not agreed on a process to verify that the plutonium for the MOX program actually comes from weapons stockpiles. This means that billions of dollars could be spent to dispose of the wrong plutonium.

In addition, Russia has not agreed to join the Vienna Convention on nuclear liability, a key U.S. issue. The United States expects Russia to commit to pay for any liability associated with the plutonium disposition agreement and to contribute cash toward the MOX program, but Russia has not yet agreed to these terms.

MOX fuel, although a technically viable solution for the disposition of excess weapons-grade plutonium, is not without problems. The lack of significant experience in the United States with fabricating and irradiating MOX fuel, its high cost, the recent negative experience with MOX fuel in Japan and elsewhere, and the delays in the implementation of the program indicate that parallel alternative advanced fuel designs should also be evaluated for the disposition of U.S. and Russian plutonium inventories.

Adolfo Repáraz

zpět na úvodní stránku