Actinide Minimization Using Pressurized Water Reactors

Abstract

Transuranic actinides dominate the long-term radiotoxicity in spent LWR fuel. In an open fuel cycle, they impose a long-term burden on geologic repositories. Transmuting these materials in reactor systems is one way to ease the long-term burden on the repository. Examining the maximum possible burning of trans-uranic elements in Combined Non-Fertile and UO[subscript 2] (CONFU) PWR assemblies is evaluated. These assemblies are composed of a mix of standard UO[subscript 2] fuel pins and pins made of recycled trans-uranics (TRU) in an inert matrix, and are designed to fit in current or future PWRs. Applying appropriate limits on the neutronic and thermal safety parameters, a CONFU-Burndown (CONFU-B) assembly design is shown to attain net TRU destruction in each fuel batch through at least 9 recycles. This represents a time span of nearly 100 years of in-core residence and out-of-core storage time. In this way, when the TRU is multirecycled, only fission products and separation/reprocessing losses are sent to the repository, and the initial inventory of TRU is reduced over time. Thus, LWRs are able to eventually operate in a fuel cycle system with an inventory of transuranic actinides much lower than that accumulated to date. Three recycling strategies are considered, all using a 4.5-year in core irradiation, followed by cooling and reprocessing. The three strategies involve a short-term cooling (6-year) after discharge, a longer-term cooling (16.5-year) after discharge, or a strategy called Remix. The Remix strategy involves partitioning the Pu/Np after 6-year cooling for immediate recycle, and partitioning the Am/Cm for an additional 10.5-year cooling before remixing it into the next CONFU-B batch. At equilibrium, the CONFU-B can burn approximately 1.5 kg to 10.0 kg of TRU per TWhe depending on the recycle strategy used. This represents a net burning rate of 2- 8% of the TRU loaded per assembly, in addition to burning an amount equivalent to the TRU produced in the UO[subscript 2] pins. However, the highly heterogeneous nature of these assemblies can result in fairly high intraassembly pin power peaking. By design, an IMF pin in the assembly carries the highest power to maximize the TRU destruction. For the initial TRU loading, the highest power peaking in an IMF pin is 1.183. This is compensated by having cooler pins in the immediate vicinity. Even so, the pin peaking distribution in the assembly can result in reduced thermal margins. The assembly mentioned above has an MDNBR of 1.43, instead of 1.62 for the all-UO[subscript 2] assembly, based on a core-wide radial peak-to-average assembly power peaking of 1.50. Use of neutron poisons and tailored enrichment schemes reduces the neutronic reactivity of fresh assemblies, while improving MDNBR to 1.51. In addition, RELAP was used to evaluate the fuel behavior under large break LOCA conditions. CONFU-B performance under these conditions was comparable to the standard all-UO2 assembly. Several options for spent fuel recycling in LWRs are compared economically, and all are found to be more costly than making fresh UO2 fuel from mined ore. However, the CONFU-B strategy is less costly on a mills/kWhe basis than other thermal recycling strategies that recycle the full TRU vector. Given OECD estimates for the unit costs of each fuel type, and assuming 10% carrying charge factor, this cost is 10.0 mills/kWhe for the CONFU-B recycle, compared to 22.2 mills/kWhe for MOX-UE and 5.4 mills/kWhe for all UO[subscript 2]. Note that these FCCs assume the 2 disposal fee collected during power generation of a previous cycle can be invested while the fuel is cooling and provide a credit to the cycle that uses the fuel after reprocessing. The fuel handling challenges of multirecycling TRU in CONFU-B assemblies are compared to other multi-recycling strategies. If we assume that the spent fuel from the seventh recycle in each strategy is no longer recyclable and must be sent to the repository in its entirety, the CONFU-B strategy still places much less total burden on the repository than the once-through cycle, and even less burden than the current MOX cycle. Finally, a methodology for calculating the time integrated proliferation risk of a fuel cycle is introduced. An innovation of this methodology is the discounting of future risks to calculate an overall present value risk of a given cycle. Under this methodology, the CONFU-B presents lower risks than other multi-recycling strategies in the first 100 years. For a 10% rate of discount of risk, the CONFU-B risks are comparable to the once-through cycle. The longer term risk favors recycling due to the limited accumulation of repository risk.