UCS: AREVA Fuel Assembly Test Failure Dooms Plutonium Fuel Test

August, 2008 -- The Department of Energy's (DOE) plan to turn 34 metric tons of surplus nuclear weapons plutonium into "mixed oxide" (MOX) fuel for irradiation in nuclear power reactors has stumbled into yet another serious roadblock. Duke Energy has prematurely terminated a multiyear test of four experimental MOX fuel assemblies in its Catawba 1 reactor in South Carolina because of a potentially serious defect in the fuel design.

As a result, if DOE's MOX test is to produce the necessary data to support large-scale use of MOX fuel, the fuel manufacturer – the state-owned French nuclear company AREVA – will have to change the defective fuel design, manufacture new MOX fuel, and repeat the irradiation test. A complete irradiation test on fuel assemblies identical to those that would be actually used in the full-scale campaign is essential before any MOX program can go forward.

A new test, however, would cause additional delays and cost overruns for a DOE program that is already a decade behind schedule and projected to cost billions of dollars more than when it was first proposed.

A far simpler, quicker and cheaper way to dispose of separated plutonium would be to "immobilize" it by mixing it with high-level waste and turning it into glass blocks. DOE, however, is planning to cancel all immobilization technology development.

Reactor Fuel

For the past three decades, U.S. nuclear power reactors have used uranium fuel. Long metal tubes filled with uranium fuel pellets are placed in metal frames called fuel assemblies, which are shipped to nuclear plant sites and loaded into the reactor cores.

MOX fuel assemblies are similar to conventional uranium assemblies. The difference is that the MOX fuel pellets are made from both uranium and plutonium.

Fuel assemblies typically reside in reactor cores for three operating cycles. An operating cycle is the period between refueling, which is 18 months for the Catawba reactor. During refueling, about 40 percent of the fuel assemblies are removed from the reactor core and replaced with new fuel assemblies.

Duke Energy informed the Nuclear Regulatory Commission (NRC) about the aborted MOX test in a June 10 report about the reactor's recent refueling outage. The report stated: "A re-design of the Catawba Unit 1 Cycle 18 core design was required to remove MOX fuel assemblies from the core due to excessive assembly growth…."[i]

The problem that terminated the MOX experiment apparently was due to AREVA's use of a proprietary experimental alloy, called M5, in the fuel assembly control rod guide tubes. Dozens of AREVA conventional uranium fuel assemblies using M5 guide tubes are currently in the cores of several other U.S. reactors, including Three Mile Island-1 in Pennsylvania, Davis-Besse in Ohio, Oconee in South Carolina, and Crystal River-3 in Florida.

A Flawed Experiment

In June 2005, Duke Energy loaded four experimental lead test assemblies (LTAs) containing weapon-grade plutonium (MOX) fuel into the Catawba 1 reactor near Rock Hill, South Carolina. AREVA manufactured the LTAs using U.S.-origin plutonium from Los Alamos National Laboratory that it shipped to France from Charleston. The Catawba experimental program was intended to demonstrate that the MOX LTAs could be irradiated safely in U.S. reactors as a precondition for obtaining an NRC license to load MOX fuel into as many as 40 percent of the cores of the four Catawba and McGuire reactors.

The MOX LTAs were based on an AREVA design called Advanced Mark-BW. Instead of using conventional Zircaloy-4 metal, this type of fuel utilizes M5 for the tubes containing the fuel pellets and for the guide tubes in which the reactors' control rods are inserted to shut down the reactor. AREVA used M5 instead of Zircaloy because it generally exhibits a slower rate of corrosion during reactor operation.

Duke Energy reloaded all four assemblies for a second 18-month cycle in December 2006. In an April 2008 presentation to the NRC, AREVA stated that it expected that the MOX LTAs would be irradiated for three cycles.[ii] (Duke Energy irradiates non-MOX fuel for up to three cycles.) However, during the refueling outage in June of this year, Duke removed all the MOX LTAs from the Catawba 1 core "due to excessive assembly growth," and instead had to redesign the core to accommodate the change.[iii] The unexpected guide tube growth could deform and damage affected fuel assemblies, potentially degrading fuel cooling and interfering with the ability to insert the control rods needed to shut down the reactor. As a result, Duke and AREVA will not be able to obtain any irradiation data on the MOX LTAs for burnups as high as those currently achieved with non-MOX fuel. Moreover, the data obtained with the defective fuel during the past three-year irradiation will not be usable, because AREVA will have to redesign the MOX fuel and start the qualification process all over again.

The defective MOX fuel assemblies are currently in the Catawba 1 spent fuel pool, and future plans for them are uncertain, according to statements made by Duke Energy at an August 1 meeting with the NRC in Atlanta.

In July 2007, Katie Vogler, a National Nuclear Security Administration (NNSA) official, reported that "the MOX fuel lead assemblies … are performing well."[iv] Vogler was misinformed. NNSA and its contractors, Duke Energy and AREVA, should have been aware at that time of a potential problem with the MOX LTAs. The problem with excessive M5 guide tube growth in AREVA fuels – a growth rate "much higher than expected," according to AREVA – was actually first observed in April 2007 for 20 AREVA Mark B12 conventional uranium assemblies that had been irradiated for two cycles in Three Mile Island-1. At the time this was observed, at least 94 of the defective assemblies were in the Three Mile Island-1's core.

An April 2008 AREVA fuel performance presentation documented that a number of reactors – including Oconee 2 & 3, TMI-1, Davis Besse and Crystal River – had follow-up inspections for "unexpected" Mark B-12 M5 guide tube growth.[v]

Despite the fact that the assembly growth exceeded design limits, TMI's operator, AmerGen, decided to reload all but two of the assemblies for a third cycle after changing the upper end fittings. Operators did the same thing at the Davis-Besse reactor in Ohio. Conversely, Duke Energy decided not to go ahead with reloading any of the defective MOX assemblies in Catawba-1 for a third cycle.

Should the MOX Program Continue?

The AREVA MOX fuel fiasco is yet another demonstration of the MOX program's technical uncertainties, dangers and costs. AREVA has not yet determined the cause of the excessive assembly growth, but it is likely that the Mark B series with M5 guide tubes will have to be redesigned or abandoned. That means that the MOX test has failed and that AREVA will need to go back to the drawing board for MOX fuel. Lacking MOX fabrication capability in the United States, new LTAs would have to be manufactured in France, a politically and technically challenging proposition requiring new environmental documentation, export licenses, transport security plans, and new fuel fabrication plans. Further complications would result from the fact that the facility in Cadarache, France, where AREVA fabricated the MOX fuel pellets used in the defective assemblies, is now closed. The Union of Concerned Scientists (UCS) and Friends of the Earth (FOE) estimate that this problem will cause a delay of at least eight years in the weapon-grade MOX fuel qualification program.

Congress should immediately begin an inquiry into the MOX program and consider terminating it. The Government Accountability Office (GAO), charged by Congress to provide quarterly reports on the MOX program, also must investigate the use of M5 materials in all U.S. reactors, and whether those reactors should be shut down and the defective material removed.

Source: UCS

Bookmark/Search this post with:

 •    •    •    •    •    •    •    •    •    •    •    •