Determining the most cost-effective electric infrastructure composition and operations to facilitate the electrification of heavy truck fleets for distribution centres in grid congested areas

Abstract

Emission-free zones are being instituted in cities in the Netherlands to combat the emissions caused by transportation activity causing operators of large commercial vehicle fleets to electrify their vehicle fleet. E-truck fleets need sufficient charging infrastructure and power capacity at the distribution centres they operate from. However, due to the prevalence of grid congestion causes, the capacity of actors’ connection to the power grid is be constrained and solutions behind the electricity meter are necessary to either increase local supply or reduce momentary demand. The main research question is: “What is the most cost-effective electric infrastructural composition and operation to facilitate the electrification of heavy truck fleets of large-scale distribution centres in grid congested areas in various scenarios to ensure that the route schedules can be driven?” This thesis proposes a mathematical mixed-integer linear programming (MILP) model which simultaneously optimizes the composition and operation of the electric infrastructure of a distribution centre, while considering the periodic energy balance, for various scenarios in the electrification process. This has not been done earlier in the literature for distribution centre-like systems, while no similar use case has been studied in the Netherlands or its general region. Before the model was applied to the real-world case, it was verified using a recalculable numerical test and validated using linear regression analysis. In the model validation the K-means method was used to determine a typical day per month, to comprise the simulation period to 288 time periods of an hour. The proposed mathematical model was tested by means of a case study and various scenarios for the distribution centre of a supermarket company to analyse the models’ behaviour in a practical setting. The results show that from a greedy charging subscenario to a variable electricity prices subscenario, the model is able to decrease the yearly total costs and the number of infrastructure components required. The model is able to reduce the yearly total cost by up to 19.4% in the final scenario. The infrastructure investment steps in the electrification process to take, would be to install consecutively or collectively 1, 3 and 7 charging stations and 994, 1623 and 4340 PV panels in 2022, 2025 and 2030, leading to a PV installation with a total rated power of 1.562 MW, which proved to be the more cost-effective than wind turbines and battery storage. The results of the scenarios, the graphs and the sensitivity analysis show that the model is effectively able to optimize the infrastructural composition and operation to minimize the cost in all cases, giving managerial insights into the decisions that need to be made in all steps of a fleet electrification process.