With the end of the Cold War, the US and Russia dismantled a number of their nuclear warheads, resulting in the creation of a stockpile of weapons-grade material. The US Department of Energy (DOE) was given the task of disposing of this plutonium. It is adopting a strategy of immobilising some of the plutonium, and converting the rest to MOX fuel. As a result, while much of Europe is suspending involvement in MOX, the US is showing an interest it had not displayed before.

The DOE chose a consortium of Duke Engineering, Cogema, and Stone & Webster (DCS) to design a MOX fuel fabrication facility (MFFF) based on those of Cogema and Belgonucleaire.

The DOE originally planned to dispose of 33t of plutonium at six plants but later reduced that amount to 25t. This will be burned at four Duke Power plants, McGuire 1 and 2 (2x1220MWe PWRs) and Catawba 1 and 2 (2x1205MWe PWRs). All the mission reactors use 17×17 fuel in operating cycles of approximately 18 months.

DCS has developed a three-step mission reactors irradiation plan (MRIP) to examine the feasibility of disposing of 25t of plutonium between 2007 and 2022.

In the first step DCS, identified the key fuel management objectives. There were several operational assumptions and constraints that were considered. The MFFF will be designed, constructed and started up in time to produce batch quantities of MOX fuel to support batch implementation in September 2007. This is planned to coincide with the issuing of regulatory approvals.

In addition, the mission reactors must be modified to allow them to receive and use MOX fuel by September 2007. It was assumed that, as the mission reactors have not received licence extensions, they will cease operating at or near the end of their 40 year licenced operating period. It was also assumed that the reactors will operate in accordance with current plans for cycle length and that they will not experience significant unplanned outages. It was also assumed that MOX fuel would continue to be available.

The technical and licensing risks are being minimised by using MOX fuel designs and core designs based on European experience.

In the second step Duke Power established a fuel management strategy to accomplish the stage 1 objectives. It took a number of decisions on the constraints associated with the MOX fuel core designs. These constraints are as follows:

•MOX fuel assembly designs must be compatible with and mechanically similar to current low enriched uranium fuel designs.

•MOX fuel burnup will be limited to 45MWd/kgHM (assembly average) and 50MWd/kgHM (peak rod).

•Uranium fuel burnup will be limited to 60GWd/kgHM (lead rod).

•MOX fuel will be discharged after

two cycles.

•MOX fuel peaking limits will be identical to uranium fuel.

•Fuel cycles will be consistent with uranium fuel management plans (currently

18 month cycles).

•MOX fuel core fractions will be less

than 40%.

•Low leakage core designs will be used.

•MOX fuel will receive a minimum of 20MWd/kgHM burnup.

•The MOX fuel will contain no integral burnable absorbers.

•The plutonium isotopic concentrations are characteristic of weapons grade material (<7% Pu-240).

•The uranium portion of the MOX fuel is composed of depleted uranium with a

nominal enrichment of 0.25% U-235.

In step three, Duke Power performed fuel cycle studies to verify that the core designs could accomplish the programme objectives and meet all the assumptions and constraints. Typical PWR cores are a mixture of fuel assemblies that are in their first, second or third cycle of irradiation. Duke Power performed its nuclear analyses using the Casmo-4 and Simulate-3 MOX computer codes.


The MFFF is based on the successful Melox plant at Marcoule in southern France. The design also includes an aqueous polishing process, as a result of Cogema’s experience at its La Hague facility in northern France.

The process design is being carried out by a joint team from Cogema and Belgonucleaire, while the facility design is being carried out by Duke Engineering and Stone & Webster.

The project design started in March 1999, and is due for completion in March 2002. The facility will be licensed by the NRC, and owned by the DOE. The NRC will apply for construction authorisation in late 2000, providing supplemental information as the design evolves. Final licence application is planned for 2002, construction is scheduled to start in 2003, and plant startup is scheduled for 2006. The first fuel from the plant will be delivered to the mission reactors in 2007.

The MFFF has two major process operations: the aqueous polishing process to remove gallium from the plutonium, and the MOX fuel fabrication process that processes the oxides into pellets and manufactures the fuel assemblies.

The three main steps of the aqueous polishing process are dissolution, purification and conversion:

•The dissolution step involves silver catalysed dissolution and filtration. The process was selected because it is very efficient, and independent of PuO2 powder characteristics. It results in the complete dissolution of the PuO2 powder according to kinetics governed solely by the rate of Ag(II) generation. PuO2 powder is dissolved by electro-

generated Ag(II) in a nitric acid medium.

•The purification step involves plutonium extraction, solvent regeneration, acid recovery and silver recovery. The process was selected because it yields very little plutonium leakage and has a very high gallium decontamination factor. The main functions of the plutonium extraction step are plutonium extraction and impurities scrubbing and plutonium stripping. These operations are performed in pulsed columns.

•The conversion is a continuous oxalate conversion process. This was selected because it yields a PuO2 powder routinely used for MOX fabrication at Cogema

facilities. There are several main operations to convert to plutonium oxide:

precipitation, evaporation of the mother liquor, filtration, drying and calcination and plutonium conditioning.


Belgonucleaire and Cogema produce pellet fuel characterised by an intimate mix of the PuO2 and UO2 powders, using the

A-MIMAS process.

The fuel fabrication process includes four major steps: the powder master blend and final blend production, pellets production, rods production and fuel rod assembly.

Powder master blend and final blend

The first blending operation in the A-MIMAS process is the production of a master blend with 20% plutonium content – higher than the content in the final blend. Three different powders are blended: PuO2, UO2 and recyclable scraps of MOX fuel pellets. Primary milling is one of the most important processes and the ball mill is the best device to meet all the necessary requirements to obtain a good master blend before sieving and final blending.

The final blending is the last operation

of the A-MIMAS process to adjust the final Pu content. After the final blending, the master blend and the dilution UO2 are homogenised to satisfy the requirements

of MOX fuel pellets.

Pellet production

Most of the final characteristics of green pellet production are defined during this step. The parameters of the pelletising operation must be adjusted and controlled to avoid pellet defects.

By-products – such as recovered powder and discarded pellets – are recycled. This includes

crushing and ball milling of the discarded pellets, and pelletising and sintering recovered powder before following the same process as the discarded pellets. The

sintering step removes organic products that are dispersed into the pellets and removes the

poreformer necessary to reach the required pellet specific gravity. The dry centreless grinding machines grind the sintered

pellets to the final diameter. The system also handles the discarded pellets and the dust from the grinding process.

Rod production

Rods are loaded to an adjusted pellet length column, TIG welded, helium pressurised and then decontaminated. Rod inspection then verifies the helium tightness, the welding quality, and the correct Pu content in the pellet column.

Fuel rod assembly

Rods of different Pu content are assembled and the final bundles are stored before packaging and shipping.

Design challenges

There are a number of challenges associated with the design of the MFFF. It is the first of its kind in the US that will use weapons-grade plutonium as the feedstock; it is owned by the US DOE but is licensed by the Nuclear Regulatory Commission. It involves technology transfer from Europe to the US.

NRC licensing

The MFFF will be a DOE-owned, contractor-operated facility, located alongside other DOE-owned facilities at Savannah River. The MFFF will be subject to NRC regulation, and the consortium will ensure that the site will meet regulations imposed by

the NRC, the Environmental Protection Agency, the Occupational Safety and Health Administration, and the state and local authorities.

Licensing for the MFFF will be in accordance with 10 CFR 70, which places a heavier burden on applicants for plutonium facilities than for other fuel cycle facilities. Authorisation to start construction and issuance of a possession-and-use licence are separate licensing actions.

Process hazard analysis

An integrated safety analysis (ISA) is

being prepared to establish the safety design basis.

As a starting point for identifying hazards, an initial hazard list based on Cogema facilities was adapted for US regulations and standards. The first step in the process is the preparation of a preliminary process hazard analysis. This provides a qualitiative assessment of postulated accident scenarios in the MFFF and identifies an initial set of items relied on for safety.

In later phases, other methods such as hazard and operability analysis (HAZOP) and fault tree analysis may also be used to analyse areas in more detail.

Criticality control

Criticality safety evaluations are performed in accordance with standard US procedures.

The facility is divided into a number of criticality control units. For each unit, the reference fissile material is defined, along with the criticality control mode. Geometry control is used whenever this is possible. In addition to this, fissile material mass with moderation control will be used for process and operability reasons. This double contingency principle requires that there be at least two unlikely, independent, and concurrent changes in the process conditions before it is possible for a criticality accident to occur.

The MFFF uses appropriate criticality benchmark experiments for each type of physical situation. This is important given the diversity in applications including high, medium and low moderated PuO2 and mixed oxide powder, nitrate, oxalate solutions, and arrays of pellets and rods. Benchmark experiments will come primarily from the international criticality benchmark handbook, although some additional recognised experiments will be used. Validation of benchmark experiments will be performed using traditional statistical methods along with newly developed methods if they are necessary and available.

Static and dynamic confinement systems

The confinement systems are designed so as to prevent any

permanent contamination in the areas where personnel can be

present and to keep releases to below acceptable limits in the event of an accidental release, spill or system failure.

The facility is designed with multiple confinement systems. Each of these confinement systems consist of both static confinement and dynamic confinement subsystems.

The static confinement systems include building walls, barriers, glove boxes, enclosures, filters, hoods, piping, tanks, valves, exhaust ductwork, plenums and vessels. The dynamic confinement consists of the HVAC exhaust subsystems.

The dynamic confinement systems

supplement the static systems by maintaining the pressure gradients between the

different contamination level zones in order to induce airflow leakage from the zones of lowest contamination potential towards the zones of increasing contamination potential.

The dynamic and the static confinement portions are not redundant or back up to each other, but complement each other’s functions.