MODELING NUCLEAR MATERIALS PROCESSES

At Los Alamos National Laboratory the Technology Modeling and Analysis group is currently modeling nuclear material (NM) processing at the DOE's Rocky Flats Plant. Modeling methodology employs artificial intelligence and discrete-event simulation techniques. The goals of this project are to predict material flows, manpower requirements, equipment utilization, radiation exposure, and the estimation of waste production, all under varying scenarios. Given the flexibility in the design of the model, these goals can be readily extended. In the model, one follows the movement of NM through a processing plant, with specific amounts of NM predicted at each stage. As the model is developed and analysis performed, optimum sampling points are realized, as are better methodologies for measurement of NM holdup. The model could also be used to flag possible diversion points throughout the plant. Prediction of inventory differences, holdup, and analysis of measurement needs are further applications. The level of detail that such a model provides also allows for the prediction of future NM feed and storage needs.