Risk assessment method of gas explosion based on quantification of margins and uncertainties (QMU):a case study on beam structures in buildings

Abstract

Combustible gas explosions are common accidents in residential areas. Key building structural components, such as beams, slabs and columns, can be quickly damaged and destroyed by explosive overpressure loads, which may also result in continual collapse. Risk of interior gas explosions in buildings should be assessed to analyze the ability of structures to withstand disasters, which can also offer a theoretical basis for improving the structural design and construction of buildings. In this study, reinforced concrete beams in buildings were selected as the research object, and the risks to beam structures under mixed methane and air explosions were assessed using the quantification of margins and uncertainties (QMU) method. First, the uncertainty of the existing methods to calculate gas explosion and beam bending loads was analyzed. Through an influence regularity analysis, the expansion factor of combustible gas combustion products in enclosed explosion chamber, the initial density of the gas mixture in the explosion chamber before deflagration, the compressive strength of concrete, and the tensile strength of reinforcement were selected as the key uncertainty parameters. Furthermore, the Monte Carlo method was used to create 5000 samples of the selected uncertainty parameters, and statistical analysis was conducted to obtain the probability distribution of the gas explosion overpressure load and reinforced concrete beam bending load. Based on the sampling results, the beam structure’s safety margin and uncertainty analysis were analyzed to determine the probability of structural failure at the 90% confidence level. Finally, this paper provides guidelines for determining structural failure under different boundary value conditions of combustion explosion overpressure loads and reveals the sensitivity of different uncertainty parameters to structural safety margins using stepwise regression analysis. Compared to previous methods, this work includes the uncertainty features of disasters and disaster-bearing items when analyzing combustible gas explosion accidents; we anticipate it offers a fresh perspective on the quantitative assessment of gas explosion risks in buildings.