Calculational Methodology and Associated Uncertainties

Abstract

The reliable prediction of performance, safety and economic parameters requires a knowledge of the neutron distribution in space, energy and time from the level of a single fuel pin to that of the entire reactor. The behavior of the neutrons is formally governed by the Boltzmann equation (1,2). This equation is so complicated that there is no possibility of obtaining exact solutions, but for the most highly idealized situations. As a result, much effort has gone into developing methods and computational algorithms to approximate its solution. New and more efficient approaches are continually being developed to attack problems for which the conventional methods are not adequate or needlessly time consuming. This process is closely tied to the computer capability. If it were possible to analyze the core physics at little cost, further research in code development and design methodology would scarcely be warranted. However, this is not the present state of the art for reactor design. It is interesting to speculate about the future, and although the “1984 date” is highly optimistic, we quote Lathrop (3): “Hence, it is possible that by 1984, say, the ingenuity of methods developers combined with increased computer power may provide the final solution to multidimensional core analysis problems.”