Comparison and Evaluation of Calculation Codes Used for Shielding Analysis of Spent-Fuel Packages

Transport package shielding calculations are performed to predict external dose rates from gamma rays and neutrons. In Japan, for gamma ray shielding calculations, the QADCGGP2 code (Cain 1977) has generally been used, and for neutron shielding calculations ANISN (Engle I 967) or DOT (Rhoades 1973) codes have been used in many cases. However, when those codes are used there are many limiting calculation conditions. QAD is a three-dimensional code which uses the point-kernel method. It is easy to treat the complex geometry problems, and the computing time is short, but it is only possible to use one build-up factor for one shielding model. ANISN is a one-dimensional code following transport theory using the discrete ordinate (Sn) method. The calculations, which include the effect of the direct ray and scattering are performed exactly but the computing time is longer and, of course, it is only available to treat one-dimensional problems. The oor code is the same as the ANISN code except that it is a two-dimensional code. In this study, the MARMER (Kloosterman 1990)and MCNP (Briesmeister 1991) codes, which were recently developed in The Netherlands and the United States were investigated and examined to see whether they could be used instead of the QAD and ANISN codes. The MARMER code is three-dimensional code which employs the point-kernel method. The MCNP code is a three-dimensional code using Monte Carlo methods, and it can treat any complex geometry, although the computing time is the longest of all the four previously mentioned codes. In this paper, the calculation results obtained by using those four codes on a benchmark problem are compared with measured values from experiments. Those results are then compared and evaluated on a simulated actual multilayer packaging body wall problem.