Investigation of Optimal Class-8 Fuel Cell Hybrid Electric Truck Design for Improved Fuel Economy

Abstract

In 2022 the transportation sector was the largest source of direct greenhouse gas (GHG) emissions in the US, with medium and heavy-duty trucks representing the second highest GHG emitting vehicle category [1]. Fuel cell hybrid electric trucks (FCHETs) are emerging as a promising solution for the electrification of Class 8 long haul transportation. However, further investigation is required into how to optimize the electrical powertrain of FCHETs. In this thesis, a detailed Class 8 FCHET model is created in MATLAB/Simulink. The electrical powertrain components are modeled using equations that are presented. The model is used to investigate partial power processing topologies for the fuel cell DC/DC converter, the optimal DC-link voltage, and the optimal battery connection topology for Class 8 FCHETs. All the investigations are conducted for SiC and IGBT semiconductor devices, aside from the partial power processing investigation which is only conducted for SiC semiconductor devices. Partial power processing (PPP) converter topologies have been proposed as more efficient and compact alternatives to conventional converters for certain applications, however no investigation has been conducted for the fuel cell DC/DC converter in a Class 8 FCHET. In this thesis, detailed simulations are created to compare the performance of various PPP DC/DC converter topologies to the conventional boost converter. Efficiency maps and power loss breakdowns for multiple PPP DC/DC converter topologies are created using PLECS. Then, the performance of the fuel cell DC/DC converter topologies are assessed with an FCHET vehicle model created in MATLAB/Simulink. The results show that for a 400 V DC-link, the PPP FB converter is slightly more efficient than the boost converter, and for an 800 V DC-link, the PPP FB and PPP DAB converters are slightly more efficient than the boost converter. After considering the results of the PPP investigation, the subsequent DC-link voltage and battery connection topology optimization was conducted with a conventional boost converter for the fuel cell (FC) DC/DC converter, due to the simplicity of the topology and since the PPP DC/DC converters did not show significantly improved efficiency. The FC DC/DC converter is implemented step-up the fuel cell voltage, which allows the other electrical components (inverter, motor) to operate at higher efficiencies. However, a higher voltage gain generally causes a boost converter to operate at a lower efficiency due to higher conduction losses from a higher duty cycle, and higher switching losses which are a function of the output voltage. Thus, this presents an optimization problem: how high should the boost converter boost the fuel cell voltage to maximize the efficiency benefit for other electrical components and minimize the efficiency disadvantage for the boost converter? The DC-link voltage optimization results show that for the topology without DC-link voltage regulation, 800 V is the optimal nominal battery voltage for the SiC implementation, and 600 V is the optimal nominal battery voltage for the IGBT implementation. Additionally, the SiC implementation of this model is 4.5% more efficient than the IGBT implementation. An efficiency comparison is performed between an FCHET model without DC-link voltage regulation (battery connected directly to the DC-link), an FCHET model with constant DC-link voltage regulation, and an FCHET model with a variable DC-link voltage regulation, for various switching frequencies. The investigation results showed that the FCHET topology without DC-link voltage regulation was more efficient for all switching frequencies investigated and for both SiC and IGBT semiconductor devices, although only marginally in some cases. Thus, this thesis has made substantial contributions to the optimal design of Class-8 fuel cell hybrid trucks through the investigation of PPP topologies, electrical architecture, and optimal DC-link voltage. The conclusions can be used by FCHET manufacturers to streamline their design process by quickly narrowing in on proven optimal designs.