Coupling deterministic and stochastic modeling to unwanted fire

It is difficult for system safety analysts to incorporate fire into their general safety analyses, even though widespread a t tent ion has been given to the evaluation of "fire hazards" in the past few years. For example, in the Rasmussen report , an extensive analysis of nuclear reactor safety [1 ] , there was no thorough consideration of the problem of fire. The subsequent Brown's Ferry Fire Accident [2] pointed up this deficiency, and there has since been an exhaustive study of the subject by the Nuclear Regulatory Commission, and the industry in general. There are a number of cultural and technical reasons for this gap in system safety analysis, and it appears to this author that one of the central reasons is the lack of a simple model, unders tood sufficiently by generalists to help them structure their fire safety problems in a manner enabling fire pro tec t ion specialists to supply the detailed answers. One of the objectives of this paper is to provide a model and a method for coupling both deterministic and stochastic approaches for solving the fire protect ion problem.