Reactor Operation with Mixed-oxide (MOX) Fuel
Presented by David A. Lochbaum, Nuclear Safety Engineer for the Union of Concerned Scientists at "People's Forum on Plutonium Fuels," November 9, 2000, at Winthrop University, Rock Hill, SC (event co-sponsored by Carolina Peace Resource Center and Blue Ridge Environmental Defense League).
Consider arriving home from the grocery store after buying a carton of milk. You are likely to place the milk into the refrigerator fairly soon after getting home because you don't want it to spoil. Of course, the milk will eventually spoil, even when stored inside the refrigerator. The refrigerator is therefore no guarantee against spoilage, but it does delay the inevitable. You hope to use the milk before it goes bad. 
This spoiled milk saga is only one of countless examples of aging. Living creatures have lifetimes defined by birth and death. Inanimate objects have lifetimes defined by their usefulness. For many objects, usefulness ends with failure as when milk goes bad or a light bulb burns out. Life expectancy can be represented graphically by what is termed the "bathtub curve" due to its familiar shape. The early part of life (Region A) is dominated by infant mortality. The peak health period (Region B) is marked by random failures or accidents. The "wear-out phase" best describes the latter stage of life (Region C). The simple act of refrigerating a carton of milk postpones, but does not prevent, entry into Region C. Conversely, leaving the carton on the kitchen counter hastens the arrival of Region C and spoilage.
What does the "bathtub curve" and milk spoilage have to do with nuclear power plants operating with mixed-oxide (MOX) fuel? In today's nuclear plants and in those proposed to be fueled by MOX, power is created by the splitting, or fissioning, of uranium and plutonium atoms. More plutonium atoms will be split in MOX-fueled plants than in non-MOX-fueled plants. The splitting of plutonium atoms releases more neutrons at higher energy levels than the splitting of uranium atoms. More neutrons traveling at greater speeds may increase the number that strike the metal reactor vessel containing the reactor core. Neutrons weaken metal by making it more brittle.
A hot, brittle reactor vessel might shatter like hot glass when placed into cold water. The reactor vessel metal can reach nearly 550°F when the plant is running. In case of a small problem, the plant's emergency systems may automatically start sending 40°F water to the reactor vessel. If the reactor vessel has become too brittle and shatters, all the King's men and all the King's horses would be no more successful putting it back together than they were with Humpty Dumpty. A shattered reactor vessel makes it virtually impossible to cool the reactor core, resulting in a reactor accident significantly worse than at Three Mile Island.
Reactor vessel embrittlement is a genuine safety concern. The Yankee Rowe nuclear plant in Massachusetts was closed in 1992 after its owners were unable to convince the NRC that its reactor vessel had not become too brittle. The Yankee Rowe reactor vessel was supposed to last for 40 years, but it wore out many years sooner than expected. In other words, it was expected that Yankee Rowe would operate for 40 years before its reactor vessel reached Region C of the "bathtub curve" but that expectation was not met. Because operation with MOX fuel may cause reactor vessels to get brittle even faster, there may be more nasty surprises.
But can a nuclear plant really have a serious accident? After all, the plants are designed with defense-in-depth. There are multiple barriers between radioactive materials and the environment. There are systems backed up by emergency systems with backups to the backups. Like the hull of the Titanic, it's not the length or breadth of the barrier but the size of the holes in the defense-in-depth concept that really matter. One of those holes is the reactor vessel. Hopefully not in the reactor vessel because there is only one. No nuclear plant has a backup for the reactor vessel. Nuclear plants operate on the assumption that the reactor vessels never fail.
Another essential difference between a conventional reactor core and one with MOX fuel is the delayed neutron fraction. "Delayed neutron fraction" is a fancy nuclear physics term defining when neutrons are emitted after an atom splits. The power from a nuclear plant is produced by the splitting or fissioning of uranium and plutonium atoms. Splitting atoms releases energy and it also releases atomic particles called neutrons that are needed to sustain the nuclear chain reaction. These neutrons are released within fractions of a second after a uranium atom splits. When a plutonium atom splits, neutrons are released even sooner.
Why does this timing matter? It affects the ability to control reactor power changes. Perhaps the easiest way to explain this effect is to use the example of driving a car. When you depress the gas pedal a little, the car will gradually accelerate until it reaches a steady though faster speed. The time it takes to reach the new speed is what allows you to operate the car. Imagine how often you would hit a car in front of you if speed changes were instantaneous.
The same principle applies to the delayed neutron fraction and reactor control. There is less response time to control a reactor with plutonium fuel than with uranium fuel. Reduced response time corresponds to increased chances that the reactor core will get out of control. Therefore, nuclear plants operating with MOX fuel may be more likely to experience an accident than plants that do not use MOX fuel.
Last but certainly not least, MOX fuel represents a threat even before it is placed into the reactor core. MOX fuel is made by taking plutonium from atomic bomb warheads and mixing it with uranium. Until the MOX fuel has been used inside a reactor core, it is a relatively simple procedure to retrieve its plutonium and use it to make an atomic bomb again. Therefore, security must be tight at plants using MOX fuel. But tests conducted by the NRC over the past eight years demonstrated that nuclear plant security leaves a lot to be desired. Roughly half of the tests revealed one or more significant breakdowns.
For example, during a test conducted in May 2000 at the Quad Cities nuclear plant near Chicago, a three-man NRC team simulated destruction of enough equipment inside the plant to trigger a major accident. The nuclear industry has responded to these demonstrated failures by insisting on conducting the tests themselves. The industry whines that the NRC tests are unfair because the NRC's simulated terrorists do not follow rules established by the Occupational Safety & Health Administration (OSHA). For example, OSHA rules require plant employees working more than six (6) feet off the ground to be protected by scaffolding and/or safety belts. But the NRC's simulated terrorists do not use safety belts when they scale walls and crawl along elevated equipment to elude plant security personnel and gain access to vital areas. The industry is merely trying to change these security tests so that there won't be any more failures. No nuclear plant in this country, especially one with MOX fuel onsite, should be allowed to operate if NRC turns security tests into a charade.
Perhaps there's no use crying about spilled milk. But in the case of nuclear power plants operating with MOX fuel, there's real use crying about spoiled milk. The concerns about MOX are real. The chances of an accident may be increased by MOX. The consequences of an accident may also be increased by MOX. No plant in the United States should operate with MOX fuel until each and every concern has been fully resolved in a public forum.
December 29, 2000
