Optimization and sizing of a fuel cell range extender vehicle for passenger car applications in driving cycle conditions

Highlights

• Passenger FCREx components sizing and evaluation was performed for the first time.

• As a novelty, the optimization comprises both the EMS and the FC system operation.

• Space designs for plug-in FCV (FCREx) were generated.

• Equivalent-in-range FCREx designs were compared to commercial FCV.

• FCREx architecture potentially increases FCV range and efficiency (16.8%–25%).

Abstract

Aiming to reduce global warming and emissions in general, cleaner technologies are the spotlight of research and industry development. Among them, fuel cell vehicles (FCV) are gaining interest to decarbonize the transport sector. Plug-in FCV or FCV in range-extender configuration (FCREx) is an interesting option to reduce the total cost of ownership (TCO) and the energy usage per km. The aim of this study was to generate design spaces of FCREx by varying the FC stack maximum power output, the battery capacity, and the H₂ tank capacity to understand the implications of this architecture in range, consumption, and cost (estimated with a WLTP driving cycle). Unlike other studies, the approach was focused on a novel architecture for passenger vehicles and was focused on the development of the validated FC system model and the energy management strategy (EMS) optimization for each design, based on the Pontryagin Minimum Principle (PMP). Consumption was found to decrease with increasing battery capacity and FC maximum power due to the higher efficiency of the systems. The design spaces showed how with 5 kg of H₂ and $\ge$50 kWh of battery capacity the maximum range of FCREx could be over 700 km. The results of this study showed how FCREx architecture could provide overall energy consumption saving up to 6.8% and H₂ consumption saving ranging from 16.8% to 25%, compared to current commercial FCVs. The optimum FCREx design, not only based on performance, should have $\sim$30 kWh of battery capacity and $\ge$80 kW of FC maximum power to minimize manufacturing costs while maximizing efficiency.

Introduction

With the growing interest in low environmental impact technologies for mobility, hydrogen fuel cell vehicles (FCVs) are getting relevant and have gained market share in the automotive industry [1]. This technology is not only relevant because it is relatively carbon-free, but also since the fuel (H₂) has many advantages as an energy carrier (Section 2.1), relative to electricity for battery electric vehicles (BEVs).

As in any relatively new technology application, there exist several system architecture variations of the same technology that may improve or worsen the capabilities and performance of FCVs. This is the case of plug-in FCVs or FCVs in range-extender configuration (FCREx). FCREx configuration is a combination of BEVs and FCVs and has not yet been extensively explored for light passenger vehicles but has high potential to improve energy usage and may be the solution to extend the range of FCVs until enough H₂ refueling stations are built [2]. At present, the only architecture that was considered for light-duty passenger vehicles is that combining an FC system with a low-capacity battery. As such, the performance of FCREx architecture for passenger vehicles and how it changes with systems sizing remains unexplored, thus neglecting the potential of an FCV architecture that may be key in the context of low availability of hydrogen refueling stations.

The sizing of FCREx is relatively more complex than that of a conventional FCV since the battery capacity also affects significantly the optimum energy management strategy of the vehicle systems, the cost, and the range. As such, for this type of vehicles, it is imperative to provide a detailed and wide analysis on the performance, range, and cost of systems for different combinations of FC system, battery capacity, and H₂ tank storage in order to understand the real potential and limits of such configuration, relative to simple FCVs.

In the literature, most of the studies focus on sizing non-plug-in FCV components [3], and those focused on FCREx [4] are not oriented towards light passenger vehicles or do not consider the same parameters as those in this study. Therefore, there is a clear lack of data regarding the sizing of FCREx systems for light-duty passenger cars.

The state-of-the-art research about FCREx for passenger vehicles is limited. As a consequence, it is difficult to assess the state-of-the-art focusing only on light-duty applications. Mainly, the recent related research lines have been focused on the use of FCREx on bus and heavy-duty applications and on the energy management optimization of different FCV architectures to maximize performance. The studies focused on the use and sizing of FCREx architecture for bus applications use different EMS such as the CDCS (charge depleting and charge sustaining strategy) or two-step algorithms based on dynamic programming to analyze and optimize the vehicle costs and performance. With this, it was concluded that to minimize H₂ in FCREx the priorities are in order: reducing auxiliary power, braking energy recovery, increase FC stack efficiency, and decreasing battery losses [5]. Furthermore, following these methodologies, it was found that the optimum systems sizing design for city buses should be close to 150 Ah for the battery capacity and 40 kW for the FC system maximum power output [6]. Nonetheless, the conclusions extracted from these studies are only applicable to city buses and the performance results are far from those expected for an FCREx light passenger vehicle.

Similar to city bus application, FCREx architecture was also explored for urban logistics vehicles, using tools such as convex programming or fuzzy logic controllers to solve the sizing problem. The combination of FC systems together with moderate-capacity batteries showed that the range of urban logistics vehicles could be extended with respect to BEV and the H₂ consumption decreased by half [7]. Differently from the city bus application, the optimum battery capacity was estimated to be around 29 kWh, while the optimum FC stack maximum power depended on H₂ price [8]. The dependence of FC sizing with H₂ cost to minimize the TCO showed how sensitive the performance of FCREx vehicles is to FC stack sizing since higher FC stack maximum power implies lower H₂ consumption due to the higher system efficiency.

Among the heavy-duty applications, the use of FC for trucks is considered to provide the most advantages with respect to BEV and ICEV trucks due to the high range and carbon-free emissions. FCREx architecture is very compatible with these heavy-duty vehicles since it enables flexible operation and lower consumption. Recent research showed that using FCREx architectures for trucks could reduce the TCO by 1.3% with respect to conventional FCV architectures [9], but the result is still dependent on H₂ costs. Furthermore, different EMS were explored and compared for FCREx trucks considering 8*CHTC-HT and 7*C-WTVC Chinese truck driving cycles, concluding that convex-optimization-based EMS could provide minimal H₂ consumption and be used in on-line driving. FCREx architecture was also used in mining truck applications, where the decrease in emissions is critical to ensure the safety of mining operations, given the small space and the potential gases build-up. For these vehicles, with an optimized FC-battery hybrid powertrain design the battery life was extended, the H₂ consumption reduced, and the mining cost decreased by 8.7% [10].

Complementary to the FCREx-focused research lines, there have also been several studies also focused on EMS optimization using driving cycles simulations or conventional FCV systems sizing to improve fuel economy and system durability, but they used other components such as supercapacitors [11] or low-capacity batteries [12].

In light of the studies presented that represent the state-of-the-art of FCREx, it is worth noting that most of the use low-order models to express the FC system performance such as simple and constant polarization curves [10], [11], simple polynomials [8], [13] or simply straight lines expressing constant FC efficiency [12] that do not capture the physics behind the FC performance variation with operating conditions. In most of these studies, the FC system management was not optimized nor validated, while most of the research was focused on EMS optimization. This implies that the results were only partially-optimized and could be further improved.

The overview of the state-of-the-art research shows that currently, FCREx architecture has mostly been considered for heavy-duty applications and captive fleets such as urban logistics vehicles. Sizing studies for this architecture and applications already exist, but the conclusions and optimal designs do not apply to passenger vehicles. Furthermore, sizing studies focused on the use of FC for passenger applications do not consider FCREx architecture.

In conclusion, the literature regarding the sizing of FCREx is still limited, particularly for light passenger vehicles, and mostly omit the fundamental behavior and optimization of the FC system.

From the analysis of the state-of-the-art, some conclusions can be extracted to provide an idea of the knowledge gaps in the literature:

• FCREx architecture has been explored for heavy-duty vehicles such as city buses or trucks but the literature focused on using this architecture for light-duty passenger cars is limited.

• Most of the studies do not provide the space designs generated from their sizing analyses. The results are usually based on the optimum design based on the criteria of each particular study. Generating and showing the space designs is very important to provide an estimation of the capabilities of a system, given a wide range of design combinations.

• Range is usually not estimated for the different designs produced in the sizing analyses. In the case of FCV and FCREx, there is an actual need to understand and quantify how the sizing of the components affects the range and consumption. By showing the range estimation in design spaces, it is possible to provide passenger car manufacturers an estimation of the preliminary design they should aim for with a chosen range.

• Most of the studies consider the FC system maximum net output power, the battery capacity, or the H₂ mass in the deposit, but very few consider these three parameters simultaneously as sizing parameters and, in the case they do, the target vehicle is a city bus instead of a light-duty passenger car.

• The studies usually used FC system models that are not validated or, in the cases where they were validated or obtained experimentally, have not been optimized previously. Frequently, the optimization of each design was performed by optimizing only the EMS, which has a significant impact on consumption and costs reduction but does not focus on prior-optimizing the FC system behavior. Therefore, the sizing analyses usually omit the fundamental behavior and optimization of the FC system.

• Sizing and EMS optimization are strongly coupled to provide a representative benchmarking of different designs. Some studies use the same EMS for different designs even though the load demand and the system efficiencies also change with load.

• The resulting optimum designs from the sizing studies were not compared against commercial FCV to prove the increase in fuel economy or overall performance.

The motivation and contribution of this paper provide an understanding of the performance and costs of vehicles with FCREx architecture depending on the systems sizing and to identify how the battery capacity, the FC stack maximum power, and the H₂ tank capacity should be dimensioned depending on the target range and/or consumption. To fulfill these objectives, space designs for light-duty passenger FCREx were generated and analyzed considering as sizing variables the FC stack maximum power, the battery capacity, and the H₂ tank capacity (knowledge gaps 1, 2, 3 & 4). Unlike other studies, the FC stack model was validated at different operating conditions, the BoP operation was optimized, and the EMS between the FC stack and the battery was optimized independently for each design with the PMP (knowledge gaps 5 & 6). This means that the optimization of the FCREx was performed comprising the FC system operation and the EMS. The resulting FC system model was fully-scalable. The design spaces showing the range, the estimated systems cost, and the H₂ consumption were generated considering the WLTP driving cycle WLTC class 3b since the power-to-mass ratio of most of the designs was over 34 (this WLTC driving cycle was also chosen so that the final results can be compared against current commercial FCVs). Finally, state-of-the-art commercial FCVs were compared against equivalent-in-range optimum FCREx designs to understand the capabilities of this FCV architecture (knowledge gap 7).

This study comprises the following parts: introduction (Section 1), theoretical foundations (Section 2), methodology (Section 3), BoP operating conditions optimization (Section 4), FCREx systems sizing (Section 5), and conclusions (Section 6). In the Introduction and the theoretical foundations sections, the objectives, background, and motivation of the study are defined and explained. The simulation tools and procedures were described in the methodology section. The results are presented and discussed in BoP operating conditions optimization and FCREx systems sizing sections, where the optimum operating conditions and energy balance of the FC system and the consequences of varying the FCREx design are analyzed respectively. Finally, the main conclusions of this study were summarized in the conclusions section.