Incidence investigation of accidents in chemical industries: A comprehensive study based on factor analysis

Abstract This study aimed to identify factors affecting the incidence of chemical accidents in chemical process industries. The present study investigated 840 accidents in 42 chemical process industries over 11 years (2008–2018). Data analysis was conducted using exploratory and confirmatory factor analysis (EFA and CFA) and structural equation modeling (SEM). Data analysis was done using IBM SPSS AMOS v. 22. Moreover, χ 2 / df , RMSEA, CFI, NFI, and NNFA (TLI) indices were used as model fit indices. The descriptive results showed that 45.3% and 32.14% of the accidents were related to transportation and the release of chemicals, respectively. Factor analysis showed that 6 latent factors and 37 indicator variables affected the accidents. SEM findings showed that latent factors, including individual and occupational training, risk management, and their indicator variables, had indirect effects on the chemical accidents ( P < 0.05). In contrast, unsafe conditions and unsafe acts latent factors with their indicator variables had a direct impact on the incidence of chemical accidents ( P < 0.05). The findings confirmed that chemical accidents are affected by different causal layers. Using different methods of accident analysis and combining them with scientific and updated techniques of other sciences, that is, mathematics, as well as statistics, can improve accident analysis and controlling methodologies.