Fleet-mix electric vehicle routing problem for the e-commerce delivery with limited off-hour delivery implementation

Operations with battery electric trucks can reduce total energy and transportation costs on daily delivery tours although an electric truck has operational limitations with lower driving ranges and higher purchasing cost than a conventional truck. In this paper, we propose a fleet-mix and multi-shift electric vehicle routing problem for joint implementation of fleet electrification and off-hour delivery in urban e-commerce delivery systems. Every electrified truck is assumed to have two shifts for both daytime and nighttime delivery operations while conventional trucks can operate during daytime only because of municipal restrictions on nighttime deliveries, which are related to engine noise. Also, every electrified truck must recharge between shifts at its depot. A fleet owner decides the best electrification ratio of the fleet and the proper number of chargers which gives the minimum total cost. The optimization problem is described as mixed-integer linear programming model including common constraints for vehicle routing problem, recharging constraints, and two-shift operation of electrified trucks. A bi-level meta-heuristic algorithm is also suggested for efficient solution search. The upper-level problem assigns trucks with engine type and brief tour information using variable neighborhood search heuristic, and the lower-level problem finds the best tour of each assigned truck using a tabu search heuristic. Scenarios with different driving ranges and nighttime operation availabilities are developed and evaluated with the POLARIS transportation simulation framework, and results are reported.