Assessment of tanks vulnerability and domino effect analysis in chemical storage plants. (July 2019)

• Record Type: Journal Article
• Authors: Jiang, Dai; Pan, Xu-Hai; Hua, Min; Mébarki, Ahmed; Jiang, Jun-Cheng
• Is Part Of: Journal of loss prevention in the process industries. Volume 60(2019)
• Page Start: 174
• Page End: 182
• Publication Date: 2019-07
• Subjects: Accident -- Domino effect -- Vulnerability -- Assessment method -- Bayesian network; Chemical industries -- Safety measures -- Periodicals; 660.2804
• DOI: 10.1016/j.jlp.2019.04.016
• Languages: English
• ISSNs: 0950-4230

Abstract

Abstract: s afety and security in chemical industrial plants require special attention due to the related concentrated hazards. Actually, primary accidents such as fire and explosions, can escalate into domino effects in the storage tanks depending on the relative distance between tanks, the flammability and explosiveness of the stored material, the fragility to structural defects of the surrounding tank etc. Therefore, reducing the probability of domino effect within an industrial area has become a crucial concern in industrial risk analysis and assessment. The present paper redefines the vulnerability of tanks in the case of domino effect and creates the formation mechanism and evaluation model of multiple accident coupling scenarios. Based on the occurrence, propagation and influence of the primary accident (single accident or multi-accident scenario) and the fusion of matter element extension theory, entropy weight method, complex network model, TOPSIS (technique for order preference by similarity to ideal solution) model and risk matrix, the weight I, weight II of evaluation index and the evaluation system are established which can determine easily the vulnerability level of tanks. For illustrative purposes, the application of Bayesian method indicated that the global site probability of failure (i.e. the failure of the surrounding tanks, after a first sequence accident) can be reduced by removing the most vulnerable tanks, which confirms the effectiveness of the assessment method. Highlights: The probability of accident expansion and the criterion of coupling situation in multi-disaster scene are presented. Complex network maps of accidents in a single accident scenario and multiple accident scenarios are established. The most vulnerable tank is selected effectively from considering both "availability" and "effectiveness".