Optimizing Fuel Consumption of a Dual-Fuel Full-Electric Vessel Using Model Predictive Control

Full-electric vessels are a key component in the energy transition of the maritime sector. The design of optimal energy management (EM) is a significant challenge in making this technology more widely applicable. Taking into account developments for sustainable fuels and energy generation components often results in multi-component systems and dual-fuel vessels. The increased flexibility of those heterogeneous energy systems can result in suboptimal solutions for hybrid power-sharing using rule-based control. A solution to this problem is using predictive control strategies for optimal energy management. A promising approach is Model Predictive Control (MPC), which employs an internal model of the system to predict the system behavior over time and, therefore, optimize the operation in the longterm. This is especially helpful for managing the operation of several components simultaneously and can lead to increased energy efficiency and cost reductions overall. This paper presents a novel MPC controller for a dual-fuel DC full-electric vessel, incorporating diesel internal combustion engines, hydrogen fuel cells, and batteries with flexible changing operation conditions. The controller focuses on reducing operational costs by considering the economic weight of the fuels used and the future load demand forecast. The focus lies on the integration of mixed-integer aspects into the non-linear controller. An estimation of the saving potential is assessed in a simulation study comparing to a state-of-the-art rule-based controller and an instantaneous optimization controller. The results demonstrate the control's effectiveness in optimizing the operation. Further, we show the increase in fuel-saving potential by employing a future prediction.