No abstract

The sensitivity of parameters related with reactor physics on the source terms of decommissioning wastes from a CANDU reactor was investigated in order to find a viable, simplified burned core model of a Monte Carlo simulation for decommissioning waste characterization. First, a sensitivity study was performed for the level of nuclide consideration in an irradiated fuel and implicit geometry modeling, the effects of side structural components of the core, and structural supporters for reactive devices. The overall effects for computation memory, calculation time, and accuracy were then investigated with a full-core model. From the results, it was revealed that the level of nuclide consideration and geometry homogenization are not important factors when the ratio of macroscopic neutron absorption cross section (MNAC) relative to a total value exceeded 0.95. The most important factor affecting the neutron flux of the pressure tube was shown to be the structural supporters for reactivity devices, showing an 10% difference. Finally, it was concluded that a bundle-average homogeneous model considering a MNAC of 0.95, which is the simplest model in this study, could be a viable approximate model, with about 25% lower computation memory, 40% faster simulation time, and reasonable engineering accuracy compared with a model with an explicit geometry employing an MNAC of 0.99.

There are now twenty commercial nuclear power reactors operating as of May 2010 in South Korea. As nuclear capacity becomes higher and installations age, the Korean government and industry have launched R&D to estimate appropriate decommissioning costs of power reactors. In this paper, MCNP/ORIGEN2 code system which is being developed as a source term evaluation tool was verified by comparing the estimated nuclide inventory from MCNP/ORIGEN2 simulation with the measured nuclide inventory from chemical assay in an irradiated pressure tube discharged from Wolsong Unit 1 in 1994. Equilibrium core model of Wolsoung unit 1 was used as a neutron source to activate in-core and ex-core structural components. As a result, the estimated values from the analysis system agreed with measured data within 20% difference. Therefore, it can be concluded that MCNP/ORIGEN system could be a reliable tool to estimate source terms of decommissioning wastes from CANDU reactor, although this system assumes constant flux irradiation and snapshot equilibrium core model as a reference core.

The characteristics of a geological disposal system that can accommodate increasingly higher burn-up levels of spent fuel were assessed based on the Korea reference disposal system concept. First, a status investigation that included a projection of spent fuel quantity versus burn-up was carried out to demonstrate the trend toward higher burn-up levels. Next, the main features of the Korea reference disposal system were introduced. Finally, the disposal tunnel length, excavation volume, and raw materials (e.g., a cast insert, copper, bentonite and backfill) necessary for a disposal system were comprehensively analyzed to define the characteristics and overall effects on geological disposal at increasingly higher burn-up levels. Our study determined that it is reasonable to use a canister containing 4 spent fuel assemblies with burn-up levels up to 50 GWD/MTU, while a canister containing 3 spent fuel assemblies can accommodate burn-up levels beyond 50 GWD/MTU. A remarkable increase of 33% in disposal tunnel length and that of 30% in excavation volume were observed as the burn-up increased from 50 to 60 GWD/MTU. However, this was offset by a reduction of 17% in raw materials used in canister fabrication. Therefore, it seems that spent fuel at increasingly higher burn-up levels is not a serious concern for deep geological disposal in Korea.

No abstract