Multiple Gross Error Identification Strategies for Steady State Processes

In the preceding chapter, the different statistical tests for detecting the presence of gross errors and the methods for identifying a single gross error in the data were described. For a well maintained plant, we should generally not expect more than one gross error to be present in the data. Therefore, a fundamental pre-requisite of any gross error detection strategy is that it should have good ability to detect and identify correctly a single gross error. However, if data reconciliation is applied to a large subsystem consisting of many measurements or if the sensors are operating in a hostile environment, and/or plant maintenance procedures are inadequate, then it is possible for several gross errors to be simultaneously present in the data. Thus, there is a need to add a third component to the gross error detection strategy which provides the capability to detect and identify multiple gross errors.