Modeling emergent behavior for socio-technical probabilistic risk assessment

Socio-Technical Probabilistic Risk Assessment (Socio-Tech PRA) extends PRA modeling frameworks to include the effects of human and organizational factors in a more systematic way. The field requires essential improvements with respect to "theoretical modeling" and "techniques" in order to depict emergent organizational safety behaviors. This calls for integrating a more complete theoretical model of how organizations perform with technical system PRA models. It also needs appropriate techniques to make the theoretical framework capable of capturing the dynamic interactions of causal factors within their possible ranges of variability and across different levels of analysis. Due to the equifinality characteristic of organizations, it is not always possible to identify any single factor as being responsible for flaws or deviations. What is important is whether these factors are "fitted" and work properly as a whole. This paper utilizes a framework called Socio-Technical Risk Analysis (SoTeRiA), as a unique theoretical foundation for integration of both social (safety culture and safety climate) and structural (safety practice) aspects with the technical system PRA models. The issues of developing assessment tools for the SoTeRiA framework are discussed. This paper also highlights the importance of hybrid approaches that combine deterministic (e.g., System Dynamics) and probabilistic (e.g. Bayesian Belief Network) modeling techniques. It introduces the configurational method, which relies heavily on analyzing patterns of factors, as a good candidate to bring the collective effect of factors into the hybrid environment. The proposed theories and methodologies are generic and can be applied in diverse, high-risk industries.