ABSTRACT
The atmospheric storage tank area stores huge amounts of materials. Once a fire accident occurs, it is likely to trigger the domino effect. Although the existing norms and standards have certain universality, there is still a lack of layout optimization to reduce the possibility of the domino effect. In the present work, the dynamic Bayesian network is used to construct an accident dynamic evolution network considering the domino accident in atmospheric storage tank areas under pool fire accidents. The overall escalated probability of domino accidents in the atmospheric storage tank area is calculated, combined with accident consequences, the typical layout of the atmospheric storage tank is analyzed, and the optimal tank area layout is identified to provide a strategy for atmospheric storage tank area layout optimization. According to the calculation, when there are two sizes of storage tanks, the small-size storage tanks are arranged together and the large-size storage tanks are arranged separately. The overall escalation probability and escalation risk of domino accidents in the atmospheric storage tank area are the lowest.
