Risk analysis for consumer-level utility gas and liquefied petroleum gas incidents using probabilistic network modeling: A case study of gas incidents in Japan

Utility gas and liquefied petroleum gas (LPG) are commonly used in Japan for their convenience as fuels for household and commercial needs. Although several efforts have been made to promote the safe use of gas, more gas accidents occur at the consumer level than in gas production facilities or in the supply chain. Incident investigations can acquire facts about the causes and effects of these accidents. In this paper, we propose a probabilistic network modeling approach in which the inherent characteristics of risk factors for consumer-level gas incidents are considered. In the approach, cause-effect chains are formulated for gas incidents, and network diagrams with probabilistic estimations are constructed to indicate the structure behind the occurrence of such incidents. The investigation shows that most gas incidents are caused by more than one risk factor, and one risk factor tends to cascade into others. These risk factors can be clustered according to their nature and can also be classified as originating causes or intermediate risk factors by analyzing their interdependencies in network diagrams. By identifying significant intermediate effects together with their causes, these risk factors can be reduced, which may reduce the occurrence of serious gas incidents at the consumer level.