Risk assessment model using conditional probability and simulation: case study in a piped gas supply chain in Brazil

• Main Author: Silva, Liane Marcia Freitas [UNESP]
• Publication Date: 2020
• Other Authors: de Oliveira, Ana Camila Rodrigues, Leite, Maria Silene Alexandre, Marins, Fernando A. S. [UNESP]
• Format: Article
• Language: eng
• Source: Repositório Institucional da UNESP
• Download full: http://dx.doi.org/10.1080/00207543.2020.1744764; http://hdl.handle.net/11449/200379
• topic: analytical network process (ANP); Monte Carlo simulation; risk assessment; risk management; supply chain risk management (SCRM)

The objective of this article is to present a proposed application for systematic risk assessment considering the dependence between risks. The proposal relies on a systematic literature review (SLR) as the initial phase, in which the risk classes, management phases and the tools that can be applied to the risk assessment are identified, considering the dependence between them. For this, the system adopted includes the identification and later evaluation of the risks. The evaluation involves the analytic network process (ANP), Monte Carlo Simulation and conditional probability by means of Bayes’ theorem. The identification and evaluation of the risks were applied to two links of a piped gas supply chain in Brazil, identified as company X and Y, where six specialists were interviewed in each company in the managerial areas. The ANP indicted that the most critical risk in the links is the demand risk. From this, it was possible through Monte Carlo Simulation to identify the probability of occurrence of events with connection to demand risk: demand (X) / demand risk (Y), with probability of 10%; price risk (X) / demand risk (Y), with probability of 0.64%; and risk of supply (Y) / demand risk (X), with a probability of 0%. This indicates that the highest risk is the risk of demand of firm Y, and therefore mitigation strategies should focus on this risk, as it represents the true cause of supply chain vulnerability, generating risk with the highest probability.