Ideas and concepts for diagnosis of performance and evaluation of data reliability based upon ARSA state-of-health (SOH) data

At the current time, the Pacific Northwest National Laboratory (PNNL) prototype for the Automated Radioxenon Sampler/Analyzer (ARSA) automatically transmits, on a daily basis, a subset of all state-of-health (SOH) data in an e-mail data file to a limited number of recipients. These variables represent what were considered the most critical physical parameters for the ARSA's operation at the beginning of the field demonstration in Freiburg, Germany. Operators at PNNL perform a daily review of the information in the data file for anomalous operational conditions as evidenced by sensor readings. The initial review is easily implemented by plotting the various sensor data versus time and looking for gross deviations in the periodicity of the variables compared to previous sample sensor data. After viewing the 24-hr graphical plots, if necessary, a review is conducted of the tabular data of specific sensor anomalies. In most cases, the experience has been that when there is an ARSA operational problem the data file will have multiple sensor readings that reflect some aspect of the problem.