Methodology for estimating regional production capacity loss rate in industrial sectors caused by disasters: A case study of the 2016 Kumamoto earthquakes

Abstract

The estimation of regional Production Capacity Loss Rate (PCLR) in industrial sectors after disasters is critical for disaster mitigation, resource allocation and restoration. Recent studies to estimate the PCLR in industrial sectors were based on hypothetical disaster survey data for resistance analysis and the linear regression model for recovery analysis, which may cause non-negligible bias. Moreover, these models did not incorporate the impact of lifeline disruption and restoration, thus further amplifying the estimation bias. To fill this gap, this study proposes a comprehensive regional PCLR estimation framework that incorporates the impacts of lifeline disruptions into resistance analysis and lifeline restoration into recovery analysis to improve the PCLR estimation accuracy. Specifically, this study utilizes lifeline resilience factors (LRFs) to quantify the impact of lifeline disruption on the production capacity rate immediately after the earthquake, and apply a semi-Markov recovery function to model the recovery rates considering the initial damage and lifeline availability influences. Then, the proposed methodology is applied to a case study of 2016 Kumamoto earthquakes to verify its effectiveness and accuracy. In this case study. Results shows that the non-manufacturing sector is more resistant than the manufacturing sector, yet lower recovery ability, ultimately resulting in greater production losses. The Mean Square Errors between the estimated PCLR and the official statistical of Kumamoto prefecture's Index of Industrial Production in each industry are less than 0.05, proves the validity of the proposed model and indicates the improved accuracy of loss estimation compared to previous studies. In addition, comparison results indicate improved accuracy of loss estimation compared to previous studies.