Application of Data Reconciliation and Gross Error Detection to Nuclear Material Accounting for Spent Fuel Reprocessing

Data reconciliation and gross error detection are applied to nuclear material accounting of spent fuel reprocessing in order to demonstrate their usefulness for verification activities in a safeguards system. In a simple flowsheet of the Purex process including dissolution, co-decontamination, and uranium-plutonium partition processes, the measurements of the steady-state mass flow rate, which include artificial random errors imitating measurement errors and/or process disturbances, are reconciled by a least-squares method with linear constraints of mass balance equations. The reconciled measurements satisfy the mass balance equations and their random errors are reduced. By use of another set of measurements, which includes both random and systematic errors, the gross error detection with � 2 test statistic is tested in order to find the systematic errors, which can correspond to the failure of measurement devices and the loss of nuclear materials. The systematic error is successfully detected with a given confidence limit.