Regional Closed-Loop Supply Chain Optimization for Retired Electric Vehicle Batteries with Performance Grading and Safety-Constrained Pathways

Retired electric vehicle batteries leave service with different performance and safety conditions, so a regional network must decide both where to process them and which pathways remain eligible. This study formulates a cost-oriented mixed-integer linear programming model for nine mainland cities in the Guangdong–Hong Kong–Macao Greater Bay Area. The observation-bounded T2 model is the primary case. Four official 2024 municipal indicators are mapped to prespecified ±30% operating ranges, and a Dongguan second-life indicator is withheld as a consistency check. Under a common 3.288 GWh supply, performance grades raise NetCost by 5.82% without changing topology. The Category D safety rule routes 0.164 GWh to controlled treatment at M06, replaces M03, and raises NetCost and transport emissions by 6.46% and 24.67% relative to the grade-only model. A distance-first rule cuts transport work and emissions by 59.54% but raises NetCost by 25.37%; the nearest-facility rule is infeasible. The shared 13.84% Dongguan error does not test the Category D mechanism. Within the tested ranges, condition grading mainly changes economic accounting, whereas safety eligibility changes siting and logistics.