Designing a smart risk analysis method for gas chlorination units of water treatment plants with combination of Failure Mode Effects Analysis, Shannon Entropy, and Petri Net Modeling

Abstract

Today, all modern industrial units acknowledge the necessity of efficient and effective safety, health, and environment (HSE) systems. To become practical, these systems must be localized and customized to serve the exact needs of the industry. Nevertheless, most HSE plans are developed upon a set of common presumptions. In the water industry, gas chlorination units require strong HSE plans to mitigate the possibility of chlorine explosion and leak. This study aimed to provide an efficient HSE system for gas chlorination process within water treatment plants. This goal was achieved through a case study performed on a water treatment plant in Razavi-Khorasan province, Iran. In the first stage of this study, the researchers made combined use brainstorming sessions and modified Delphi technique to identify the risk factors of gas chlorination units and classify them into six groups in terms of association with chlorination unit building, gas cylinder storage, technical details of gas cylinders, gas cylinder transport, chlorinator connections, and chlorination unit management. In the second stage, the extracted factors were analyzed by Failure Mode Effects Analysis (FMEA) and Shannon Entropy approaches using two different panels of experts, and the results were compared for validation. Finally, the analysis results were structured by Petri Net modeling. The results showed that, according to FMEA, the risk factors with risk priority number (RPN) of over 46 are of highest importance for the studied unit. Once observed, these factors necessitate shutting down the operation until a risk mitigation solution is reached. Among the analyzed factors, (i) the presence of compounds such as NH3, O2, gas and liquid hydrocarbons and oil in gas chlorine cylinders and (ii) non-vertical and non-mechanized handling of full and empty cylinders during loading and unloading, with RPNs of respectively 160 and 120, were found to be significantly more important than others. In the SE analysis, in addition to the above factors, poor implementation of airflow control mechanism inside the chlorination chamber (W = 0.359), storage of chlorine cylinders near electrical and mechanical installations such as elevators or power panels (W = 0.327), poor pipe placement for connecting the injector to the water inlet and the possibility of air suction (W = 0.433), and failure to provide scientific and practical training to the chlorination staff (W = 0.342) were found to be of highest importance.