Dot Matrix Analysis of Plant Operation Data for Identifying Sequential Alarms Triggered by Single Root Cause

Abstract

Sequential alarms are alarms triggered in succession by a single root cause in a chemical plant. In general, they occur sequentially with specific time lags within a short period of time and, if they are numerous, they reduce the ability of operators to cope with plant abnormalities because critical alarms can become buried under numerous unimportant alarms. In this paper, we propose a method for identifying sequential alarms hidden in plant operation data by using dot matrix analysis. Dot matrix analysis is one of the sequence alignment methods for identifying similar regions in a pair of DNA or RNA sequences, which may be a consequence of functional, structural, or evolutionary relationships. The proposed method first converts plant operation data recorded in a Distributed Control System (DCS) into a single alarm sequence by putting them in order by alarm occurrence time. Then, similar regions in the alarm sequence are identified by comparing the alarm alignment with itself. Finally, the identified regions, which are assumed to be sequential alarms, are classified into sets of similar sequential alarms in accordance with the similarities between them. The method was applied to simulated plant operation data of an azeotropic distillation column. The results showed that the method is able to correctly identify sequential alarms in plant operation data. Classifying sequential alarms into small numbers of groups with this method effectively reduces unimportant sequential alarms at industrial chemical plants.