LWR severe accident source term research in the USA

Fission product releases to the environment, or source terms, arise as a result of a highly diverse group of phenomena involved in any particular severe accident sequence. Because of the multiplicity of accident sequences that can occur for a given plant as well as the diversity of the, as yet, imperfectly understood severe accident phenomena, it is not surprising that reactor accidents such as, for example, those documented in NUREG-1150 have indicated large uncertainties in source terms which represent a significant contribution to the uncertainty in the absolute value of risk. Because of the difficulty and expense involved in performing prototypic experiments, substantial reliance has been placed on the development and validation of detailed mechanistic computer codes for analyzing severe accident phenomena and the source terms associated with them. This paper discusses the extensive research and other efforts that have taken place over the last decade to address the technical issues which have a bearing on being able to describe quantitatively the source term(s) and its characteristics. It also summarizes our present state of knowledge and points out areas where additional research will add further to our understanding. In this context the paper discusses the information that could be provided by the PHEBUS-FP program and its use to assess severe accident integral evaluation codes such as VICTORIA and CONTAIN. Finally, this paper discusses the United States Nuclear Regulatory Commission 's efforts to revise the licensing source term (TID-14844) and the implications of this revision, especially for siting and design of future power plants. (author)