Sustainability strategies in location-routing: A comparison of electric and bi-objective CLRP formulations

The adoption of Electric Vehicles (EVs) has become a central strategy for reducing emissions in the logistics and transport sector. However, practical barriers, including limited vehicle availability and inadequate or unreliable charging infrastructure, can constrain large-scale electrification. This study investigates whether an alternative routing-location strategy can deliver substantial CO 2 reductions without full fleet electrification. To this end, we formulate a comparative decision-support framework based on two green variants of the classical Capacitated Location-Routing Problem (CLRP): (i) the CLRP with EV (CLRPwEV), representing policy-constrained electrification with depot-based charging, and (ii) bi-objective CLRP that jointly minimises economic cost and environmental impact. A Mixed-Integer Linear Programming (MILP) formulation for the classical CLRP is first validated using Gurobi by reproducing benchmark solutions reported in the literature. A Genetic Algorithm (GA) is then benchmarked on the same instances, achieving an average optimality gap below 2.2%, and is subsequently used as a consistent solution method to ensure comparability across model variants. Computational experiments indicate that optimisation-based routing-location decisions can yield meaningful emissions reductions, although the magnitude depends on operational conditions and infrastructure effects. In the Nanning case study, a cost-emissions weighted bi-objective configuration reduced vehicle-related emissions by 16.4% but increased total system cost by nearly 50%, whereas the CLRPwEV achieved approximately 69% lower overall emissions at comparable cost. These results highlight the importance of evaluating both vehicle- and infrastructure-level emissions when selecting sustainability strategies for logistics network design.