A comprehensive framework for risk probability assessment of landfill fire incidents using fuzzy fault tree analysis

Abstract

Landfill fire is the most frequent type of incident in the waste management complexes. This paper presents a new framework for risk probability evaluation of major fires in landfills using the fuzzy fault tree analysis. The framework starts with construction of the fault tree of landfill fire comprised of 38 basic and 22 intermediate events with the corresponding type of faults under managerial, executive, human, and environmental conditions. Fault tree quantitative analysis is carried out through a combination of fuzzy set theory and experts' judgements to overcome the lack of data. Two new sensitivity analysis approaches are used to identify the critical fault type and critical paths in the fault tree. The proposed framework is demonstrated by its application to a real-world case of a landfill in Iran. The results show the probability of a major "fire incident" is 5.5 %, whereas "fire occurrence" stands for 25 % probability, higher than "lack of preparation for controlling fire" with 21.60 % probability. In addition, "Waste’s uncontrolled dumping" is recognised as the highest critical event with a failure probability of 6 % and importance degree of 24 %. "Executive fault" is also found as the most fault’s critical type through sensitivity analysis. The results also reveal the major impact of the experts’ weights, especially for events related to human or management faults. These results can give decision-makers a profound insight into providing effective intervention strategies for minimising the risk of major landfill fire incidents.