Detection of equipment faults in automatically controlled processes

Based on the UBET, a new strategy for identifying faulty equipment for dynamic chemical processes under automatic control is presented. The strategy is designed to distinguish between measurement biases, controller biases, and process leaks. For illustration purposes, application is given to a level control process under pseudo‐steady state. This approach was inspired by the work of Rollins and Devanathan to identify biased measurements under dynamic conditions. Advantages of this method are that it is not computationally intensive, can accurately detect and specifically identify the type of fault, and can accurately determine the time when the fault occurs.