The multi-compartment truck and trailer petrol station replenishment problem with domino hazard risks

Petroleum is the foundation of the oil industry and many transportation systems. As a typical hazardous material, transportation of petroleum products by road trucks will create an explosion risk and pose a potential threat to safety. The petrol distribution system with economic viability and safety calls for a well-designed distribution network and effective transportation management. In this paper, we introduce a new multi-compartment truck and trailer petrol station replenishment problem with domino hazard risks, a topic of great practical significance but with limited research attention. The problem jointly optimizes truck routing-scheduling, trailer routing-scheduling, and inventory decisions for oil depots and petrol stations. The evaporation effects of petrol during the transportation, transfer, and inventory process are explicitly considered. Multi-source data are acquired to construct the weighted social factor, including population density, GDP output intensity, POI density, and travel intensity. The comprehensive risk assessment model tailored with the domino effect and social factor is applied to quantitatively measure the risks associated with storage and transportation processes. To tackle this complex non-convex multi-objective problem, we propose a problem decomposition strategy that decomposes the original master problem into two subproblems. A customized multi-objective adaptive large neighborhood search algorithm, combining the benefits of multi-objective random variable neighborhood descent search procedure, archived multi-objective simulated annealing, and hypercube-based selection mechanism, is developed to solve the trailer-related subproblem. For the truck-related subproblem, a first-in-first-out heuristic and a linear relaxation method combined with valid inequalities are employed to rapidly compute the upper and lower bounds. This integrated matheuristic framework effectively coordinates the solutions of the two subproblems. Our model is applied to a metropolitan-wide real-world case study in Guangzhou, China. The results highlight a trade-off between distribution cost minimization and social risk minimization. The detachable mode for trucks and trailers reduces transportation costs, particularly as the distribution network expands. The compartment structure of the trailer greatly affects the efficiency of the distribution network.