Comparison of shielding calculation codes for used fuels transport/storage casks : Case study with TK-26

Year
2019
Author(s)
Vincent T. Tran - Transnuclear, Ltd
Akihiko Terada - Transnuclear, Ltd
Dai Yokoe - Transnuclear, Ltd.
Ai Saito - Transnuclear, Ltd
Hidenori Sawamura - Transnuclear, Ltd
Jeremy Alt - Transnuclear, Ltd
Boram Lee - Transnuclear, Ltd
Hiroaki Taniuchi - Transnuclear, Ltd.
File Attachment
a1154_2.pdf370.58 KB
Abstract
For the design and application for the transport and storage casks, 2D calculation codes such as DORT, are commonly used in Japan. However, in order to model much more accurately the casks and its contents, and to reduce conservatism, more advanced computational techniques, such as 3D Monte Carlo code, are considered. MCNP, becoming increasingly popular and accepted in the Japanese nuclear industry and the Japanese Nuclear Regulation Authority (NRA), is the 3D code chosen by TN Tokyo for the new shielding calculations.Our last paper presented at the 2016 PATRAM studied the influence of these codes on a simplified model. This paper tries to expand on this research by using the actual design of transport/storage cask to confirm the advantages of using a very accurate model, made possible by using MCNP. The TK-26 transport/storage cask, which can load 26 used PWR fuel assemblies, is selected for this purpose.The codes used in this study are DORT 3.2 (multigroup library: 200N-47G based on ENDF/B-VII.1 from the SCALE6.1 system), both released by the ORNL (Oak Ridge National Laboratory), and MCNP5 v1.60 (continuous energy library: ENDF/B-VII.0) released by the LANL (Los Alamos National Laboratory). For the MCNP5 calculation, ADVANTG which is released by ORNL is applied to generate the automated variance reduction parameter, and reduce the CPU time.This paper discusses for the following topics with respect to the shielding analysis of casks:•theoretical differences between the DORT and MCNP codes,•advantages, limitations and main assumptions of each code,•geometry creation and input verification and CPU time•comparison of the calculated dose rates around the TK-26 in accordance with the Japanese regulations for each codeEven if the capabilities of the two calculation codes are vastly different, the assumptions and associated margins made them both suitable for shielding calculations. However, as the dose rates calculated by MCNP are more accurate, a shift towards MCNP is currently underway at TN Tokyo. Future prospect of research would be the study of the individual impact of geometry modeling (accurate modelling vs homogenizing) and the cross-sections libraries (multigroup vs continuous energy) in more details.