Methodology for worker neutron exposure evaluation in the PDCF facility design

A project headed by Washington Group International is meant to design the Pit Disassembly and Conversion Facility (PDCF) to convert the plutonium pits from excessed nuclear weapons into plutonium oxide for ultimate disposition. Battelle staff are performing the shielding calculations that will determine appropriate shielding so that the facility workers will not exceed target exposure levels. The target exposure levels for workers in the facility are 5 mSv y(-1) for the whole body and 100 mSv y(-1) for the extremity, which presents a significant challenge to the designers of a facility that will process tons of radioactive material. The design effort depended on shielding calculations to determine appropriate thickness and composition for glove box walls, and concrete wall thicknesses for storage vaults. Pacific Northwest National Laboratory (PNNL) staff used ORIGEN-S and SOURCES to generate gamma and neutron source terms, and Monte Carlo (computer code for) neutron photon (transport) (MCNP-4C) to calculate the radiation transport in the facility. The shielding calculations were performed by a team of four scientists, so it was necessary to develop a consistent methodology. There was also a requirement for the study to be cost-effective, so efficient methods of evaluation were required. The calculations were subject to rigorous scrutiny by internal and external reviewers, so acceptability was a major feature of the methodology. Some of the issues addressed in the development of the methodology included selecting appropriate dose factors, developing a method for handling extremity doses, adopting an efficient method for evaluating effective dose equivalent in a non-uniform radiation field, modelling the reinforcing steel in concrete, and modularising the geometry descriptions for efficiency. The relative importance of the neutron dose equivalent compared with the gamma dose equivalent varied substantially depending on the specific shielding conditions and lessons were learned from this effect. This paper addresses these issues and the resulting methodology.