• GIS- and expert-risk-aware MILP for 4-level EV battery reverse logistics in Idlib. • Operational (GIS) and environmental (expert) risks are optimized as separate goals. • Fuzzy-FUCOM weights and goal programming balance cost, CO₂, and risk trade-offs. • Carbon policies (tax, subsidy, cap-and-trade) evaluated via scenario optimization. • Facility selections remain stable for the 39–14–5–3 network under sensitivity tests. This paper develops a policy- and risk-aware MILP for designing a four-level reverse logistics network for end-of-life EV/HV batteries in Idlib city, northwest Syria, a conflict-affected setting where formal datasets are limited and must be supported by field collection and expert elicitation. The contribution is to embed two distinct risk dimensions directly in the optimization—a GIS-derived operational risk and an expert-derived environmental vulnerability risk—kept separate to reflect different mechanisms of disruption and harm. Goal importance is derived using fuzzy-FUCOM, and the network is solved via non-preemptive weighted goal programming over four goals: total cost, CO₂, operational risk, and environmental risk, under carbon-policy scenarios including tax, subsidy, and cap-and-trade. Using a 39–14–5–3 candidate structure, the compromise solution attains 101% of the environmental-risk target and 111% of the operational-risk target, with cost at 128% and CO₂ at 147% of their aspiration levels, while the echelon structure remains stable across weight-sensitivity profiles, although exact site selection varies under some preference structures.
Jamil Hallak (2026) studied this question.