Understanding the effect of buildings and environmental factors in hazardous material transportation using multi-objective Conditional Value at Risk (CVaR)

Abstract

Road transportation is the primary method for transporting Hazardous Materials (HazMat). Its dynamic characteristics can randomly pose significant potential exposure risks to surrounding areas, including populations and buildings. Catastrophic events may occur under low-probability conditions, their consequences often exhibit asymmetric distributions, which traditional risk-averse models fail to quantify, especially in the tail of the risk distribution. Therefore, this study incorporates Conditional Value at Risk (CVaR) into the model construction develops application scenarios considering uncertainties in the transportation environment, and introduces the concept of Building Index (BI) to assign differentiated weights to various building types, thereby quantitatively analyzing the population exposure risk in surrounding areas. The goal is to minimize exposure risks to buildings and area populations, as measured by CVaR, under specific BI and transportation budget constraints. A case study using data extracted from Geographic Information Systems (GIS) is conducted to evaluate and plan optimal routes for typical Liquefied Petroleum Gas (LPG) distribution and retail scenarios in two cities of Jiangsu Province, China. The research findings indicate that the trade-off between transportation cost constraints and factors related to BI determines the optimal area population exposure risk measured by CVaR. Furthermore, the results demonstrate that CVaR-based optimization is superior to other risk-averse methods. This study provides valuable insights for transportation planners and regulatory authorities and supports the generation of optimal routing strategies based on decision-makers’ risk preferences.