PROCESS MONITORING CONSIDERATIONS FOR REPROCESSING

Process monitoring (PM) for reprocessing facilities has been examined as a way to provide additional safeguards data and/or reduce the burden of traditional safeguards measurements. For the purpose of this work, PM is defined as any non-nuclear bulk measurement in a facility, including mass, density, level, current, voltage, temperature, etc. or an on-line measurement that can provide more continuous data of process streams including spectroscopy, pH monitors, flowmeters, off-gas monitors, etc. Historically, PM data has been in the domain of the plant operator to ensure correct operations, but this data could be better used to improve both domestic and international safeguards. This paper summarizes recent work on integrating PM with safeguards in both aqueous and electrochemical reprocessing plants. Bulk PM data can be useful for increasing detection probability for direct material loss, but it is likely more useful in the domain of domestic safeguards since authentication of a large number of process monitoring instruments poses challenges for international safeguards. Also, the addition of a large number of PM material balances can either lead to the potential for more false alarms or a reduction in detection probability in order to keep the overall false alarm rate constant. Coupling of on-line stream monitors using spectroscopy along with bulk material balances can provide a robust PM system for detecting both direct and substitution diversion scenarios. However, this approach is difficult to retrofit and so would be more useful for new builds. The use of PM data only to estimate actinide content can require monitoring a large number of measurements (much larger than relying on sampling only), which could add considerably to the burden of authentication for international safeguards. Also, the exclusive use of bulk process data requires some type of model that has had extensive validation, which requires significant research dollars. PM approaches need to be optimized to provide the right balance of useful additional data without adding considerably to the burden of safeguards systems. Examples of the approaches studied will be presented.