Frank von Hippel (Princeton University) and Edwin Lyman (Union of Concerned Scientists)

In November 2015, High Bridge Associates released a report questioning the feasibility of putting more plutonium into the deep-underground Waste Isolation Pilot Plant (WIPP) in New Mexico, and raised the specter that the plutonium might undergo a chain reaction such as happens in a nuclear reactor or nuclear explosion. The report thereby questioned the feasibility of the plan being considered by the U.S. Department of Energy (DOE) to dispose in WIPP dozens of tons of plutonium that it no longer needs for nuclear weapons. The scenario raised in High Bridge report had been studied by Sandia National Laboratories in 2001, however, and found to be impossible.

The High Bridge report was prepared under contract with MOX Services, the prime contractor for the construction at the DOE's Savannah River Site of the hugely over cost, greatly delayed and very controversial MOX Fuel Fabrication Facility, which was to turn at least 34 tons of excess U.S. weapons plutonium into reactor fuel for use in nuclear power reactors.

Three different DOE-commissioned studies have concluded that disposing of the 34 tons of excess US weapon plutonium in WIPP would be a much less costly disposal approach than completing the MOX project (Report of the Plutonium Disposition Working Group, Plutonium Disposition Study Options, and Final Report of the Plutonium Disposition Red Team). MOX Services has been fighting this option and the new High Bridge report is part of its struggle to keep the MOX plant option alive.

In December, the U.S. Department of Energy (DOE) announced that dilution and disposal in WIPP is its preferred option for disposal of 6 metric tons of plutonium currently in storage at its Savannah River Site. The proposed plutonium waste packages (known as Criticality Control Overpacks, or CCOs) would each contain up to 380 grams of plutonium-239 (Pu-239) equivalent in a mixture that is less than 10 percent plutonium by weight inside a stainless steel pipe emplaced inside a drum with a volume of 208 liters (55 gallons).

The High Bridge report argues, however, that

"A 55-gallon drum is not designed to survive for thousands of years in this environment and the weight of the salt will eventually crush the 55-gallon drums potentially moving the plutonium in the CCOs [Criticality Control Overpacks] into a critical geometry. Since each shipment of CCOs containing surplus weapons plutonium would contain approximately five critical masses, this process needs to be analyzed and evaluated as part of a Supplemental FEIS. It is likely that this issue applies to any quantity of surplus weapons plutonium shipped to WIPP."

In fact, Sandia National Laboratories' Nuclear Waste Management Program published an analysis of exactly this scenario in 2001. It found that, if water invaded the repository, a mixture of Pu-239 and rust in water could not go critical until the average concentration of Pu-239 reached 20 kilograms per cubic meter (kg/m3). It estimated that crushing could reduce the volumes of the drums by a factor of up to six. On that basis, the average concentration of plutonium would be 11 kg/m3, still about a factor of two below the minimum average concentration required for criticality. If a greater margin were desired, some relatively non-crushable and insoluble material (e.g. pebbles) could be added to the containers to reduce the factor by which they could be compressed.

Finally, it is important to put the possibility of a criticality in WIPP into perspective. Even if it occurred, it would almost certainly be confined in the repository. The behavior of the famous natural reactors that went critical in uranium deposits in Gabon 1.8 billion years ago provides a possible model. The criticalities occurred because natural uranium in that era contained four times as much chain-reacting U-235 as it does today. The reactors released their energy at low rates, cycling on and then off again, as the fission heat increased the temperature - and perhaps even drove off the neutron-slowing water that made the chain reactions possible. This self-regulating effect would be increased if the plutonium were mixed with depleted uranium (U-238) due to increased neutron absorption in U-238 resonances as it heats up. Mixing with U-238 also would prevent the plutonium from decaying into highly enriched uranium (Pu-239 decays into U-235 with a 24,000-year half-life).