Root Cause Analysis of Process Faults Using Alarm Data

Alarm system is a vital component of industrial complexes and process plants. Once a fault occurs in a plant, some variables exceed their predetermined limits, which will trigger an alarm. However, because of immense interconnection between process variables, a fault might cause many alarms. This makes the task of finding the root cause fault difficult. One approach to find the root cause is to analyze time-trend of process data. This approach, however, is time and resource consuming and requires advanced infrastructure. In this paper, first the root cause analysis methods are reviewed. Then a technique is proposed to identify the root cause of a fault using alarm data only. The method is applied on the Tennessee Eastman process as a well-known benchmark.