Loss accounting and estimation of leaks and instrument biases using time-series data

In this work we consider the subject of loss accounting, particularly as it applies to the refining industry. The goal of loss accounting is to evaluate a plant's efficiency through the utilization of all process measurements. Central to this activity is the extraction of emissions information from indirect measurement sources (i.e., process flow and hold up measurements). In this work we present a time-series based approach to the leak estimation problem. It is shown that, in a refinery setting, the proposed time-series approach can achieve the precision required to meaningfully estimate the low flow rates associated with process emissions. The subject of integrating the time-series leak estimates with the more traditional direct emission measurements is also discussed.