California has a long history of reducing greenhouse gas (GHG) emissions, and has been working to accelerate the adoption of battery electric heavy-duty trucks (BEHDTs). Unlike diesel heavy-duty trucks (DHDTs), which have hundreds of miles of range per refill, BEHDTs have a restricted, load-dependent driving range, which makes charging planning an important role in the use of BEHDTs as an alternative to DHDTs. This research study investigates a mixed fleet drayage routing problem (MFDRP) with non-linear charging times. The study extends existing mixed fleet drayage routing models by considering multiple charging locations and allowing for more flexible routes for freight pickup and delivery. We formulate the MFDRP as a mixed integer programming model. After linearization and variable elimination, the model can be solved by commercial optimization solvers. However, the model becomes inefficient to solve when the problem size increases. Therefore, we develop a modified adaptive large neighborhood search algorithm, which can solve the problem with hundreds of units of demand in a few CPU minutes. Finally, we simulate one-day drayage operations with different BEHDT shares in the fleet for the years 2022, 2025, and 2030 to assess the potential for substituting DHDTs with BEHDTs. The numerical experiments indicate that employing BEHDTs as substitutes for DHDTs will increase the fleet size under the same level of demand. To reach the maximum share of BEHDTs in the truck fleet, the fleet size increases by 47.2%, 3.4%, and 3.4% in 2022, 2025, and 2030, respectively. Over 50% (90%) CO2 (NOx) emission reductions can be achieved by employing BEHDTs to the maximum share in the fleet.
