Year
2017
Abstract
Tracking the composition of salt as it moves through an electrochemical reprocessing facility is a challenging task because electrochemical material processing is done s e m i - continuously at elevated temperatures, while transfers involve discrete batches of used fuel or solids with adhered salts. Since salt is recycled as part of operations, it is desirable to obtain both accurate and timely measures of the chemical and isotopic compositions of the salt within the electrorefiner and through t h e other operations in the facility to fully close the material mass balance. Several technologies are under development at Argonne National Laboratory to measure the in-process salt composition. One technique is an automated micro-sample generator in combination with rapid chemical and isotopic analysis. Automated processing and analysis of large numbers of samples is made possible by the generation of micro-samples to allow analysis by on-line or at-line room temperature analytical equipment. Micro-sampling also enables quantitative measurements by radiation detectors that would otherwise be saturated by high activity. An advantage of micro- samples is that hundreds or thousands of discrete samples can be collected without affecting the process. Analyzing large numbers of samples permits averaging out of the random errors associated with sampling and measurement. To quantify the statistical benefits of analyzing large numbers of samples, three high-throughput studies were completed using a low-cost XRF as the analysis tool. In the first study, a series of experiments was carried out to study the effects of droplet analysis time. The second study examined the effect of adjusting the total number of droplets analyzed. In the third study, salts with multiple fission-product surrogates were studied to evaluate the benefits gained in the deconvolution of multiple peaks and the associated calibration errors. Based on the success of these studies a high-throughput droplet generator that can be operated remotely and in tandem with a salt processing vessel has been designed and fabricated. Reducing error in this way improves quantification, improves confidence intervals, and can lower the limit of detection for a given analysis technique. The final product combines the best aspects of destructive analysis with the best aspects of non-destructive analysis.