Simultaneous Optimization of Electric Ferry Operations and Charging Infrastructure

Abstract

Electrification of marine transport is a promising solution to reduce sector greenhouse gas emissions and operational costs. However, the large upfront cost of electric vessels and the required charging infrastructure can be a barrier to the development of this technology. Optimization algorithms that jointly design the charging infrastructure and the operation of electric vessels can help to reduce these costs and make these projects viable. In this paper, we present a mixed-integer linear programming optimization framework that jointly schedules ferry operations, charging infrastructure and ship battery size. We analyze our algorithms with the case of the China Zorrilla, the largest electric ferry in the world, which will operate between Buenos Aires and Colonia del Sacramento in 2025. We find that the joint system and operations design can reduce the total costs by 7.8\% compared to a scenario with fixed power limits and no port energy management system.

Figures (4)

• Figure 1: Problem sketch showing the main optimization design variables in green. Maximum photovoltaic production $P_{\mathrm{pv},i}$, peak grid connection $P_{\mathrm{g},i}$ and battery storage capacity $E_{\mathrm{b},i}$ and power $P_{\mathrm{b},i}$. On the vessel side, the maximum battery $E^{\mathrm{max}}_{v}$. Operational variables include the power dispatch variables from each source $P_{\mathrm{b},i}$,$P_{\mathrm{g},i}$,$P_{\mathrm{pv},i}$, and the operation speed of each vessel $s_{v}$.
• Figure 2: The route to be covered by the China Zorrilla ferry between the port of Buenos Aires (BA) and port of Colonia del Sacramento (CO).
• Figure 3: Energy and power profiles for vessels $V = 1$ and $V = 2$ over a four-day horizon for experiment 4. The figure shows the battery energy $E_v$ and the power flows $P_{\mathrm{g2v}}, P_{\mathrm{b2v}}, \text{ and } P_{\mathrm{pv2v}}$, along with mooring periods.
• Figure 4: Power flows and battery energy $E_b$ at ports BA and CO over a four-day period for experiment 4. The power profiles include photovoltaic generation $P_{\mathrm{pv}}$, grid-to-battery $P_{\mathrm{g2b}}$, battery-to-grid $P_{\mathrm{b2g}}$, grid-to-vessel $P_{\mathrm{g2b}}$, and battery-to-vessel $P_{\mathrm{b2v}}$.