Automatic generation of accident scenarios in domain specific chemical plants

This paper proposes an automatic generation method of accident scenarios in chemical process industries (CPI) and describes a computer-automated tool. One of the most powerful and widely used concepts in several risk assessment methodologies is the concept of accident scenarios. The generation of accident scenarios is on a strictly experience-based approach, and is a time and labor-consuming procedure. To reduce the effort and resources devoted to this, we developed a systematic method of generating accident scenarios and implemented software called yAGAS (Yonsei Automatic Generator of Accident Scenarios). This system automatically formulates a list of accident events and generic hazardous situations by considering the following process characteristics: process, equipment, materials, operating conditions, status, and accident progress factors. This system yields a qualitative description of accident events related to process hazards. This also provides a qualitative and quantitative ranking of hazardous situations that can be used as the top events of safety analysis.