Explainable BN-MC Modeling of Safety-Ergonomic Risk: An Iran-Nigeria Comparison

Occupational risk in the oil and gas industry is driven by a complex interplay between technical failures and human vulnerabilities, yet current assessments typically evaluate these domains in isolation. This study aims to quantify and compare integrated safety and ergonomic risks in the South Pars region of Iran and the Niger Delta of Nigeria, and to dynamically measure how safety culture influences systemic risk using an explainable Bayesian Network-Monte Carlo framework integrated with Fuzzy Cognitive Maps. An applied, model-based approach utilizing secondary data was employed. The framework integrated process safety and ergonomic variables into a unified probabilistic topology, using Fuzzy Cognitive Maps to translate ordinal safety culture data into continuous dynamic covariates, followed by ۱۰,۰۰۰-iteration Monte Carlo simulations. The simulations revealed that the Niger Delta operates with a higher baseline risk probability (۰.۹۴۷۹) but significantly lower systemic uncertainty compared to South Pars (۰.۷۷۳۹), which exhibits high volatility. Sensitivity analysis identified physical and ergonomic factors as the primary risk drivers in Iran, whereas behavioral factors dominate in Nigeria. While the dynamic cultural covariate successfully reduced integrated risk by ۳.۴۴% in Nigeria, it yielded a zero-impact anomaly in Iran, exposing a structural gap in the relational quality of the extracted behavioral data. These findings demonstrate that integrating human factors into probabilistic models is mathematically feasible but strictly dependent on the structural utility of the source data. The proposed framework provides a transferable *in silico* protocol for identifying true systemic leverage points, suggesting that safety investments must prioritize mathematically relational data over descriptive surveys to effectively alter complex risk topologies in high-hazard industries.

کلیدواژه ها:

Bayesian Network ، Monte Carlo Simulation ، Fuzzy Cognitive Map ، Occupational Risk Assessment ، Safety-Ergonomic Integration ، Oil and Gas Industry