[ Article ]

Transactions of the Korean Hydrogen and New Energy Society - Vol. 36, No. 6, pp.730-737

ISSN: 1738-7264 (Print) 2288-7407 (Online)

Print publication date 30 Dec 2025

Received 10 Nov 2025 Revised 24 Nov 2025 Accepted 25 Nov 2025

DOI: https://doi.org/10.7316/JHNE.2025.36.6.730

Experimental Study on Fuel Cell Performance Under Varying Load Conditions in an Electric Tricycle

CESARE APRILEND ABDI KHALIRIZ¹ ; OCKTAECK LIM²^{, †}

1Graduate School of Mechanical Engineering, University of Ulsan, San 29, Mugeo2-dong, Nam-gu, Ulsan 44610, Korea
2School of Mechanical Engineering, University of Ulsan, San 29, Mugeo2-dong, Nam-gu, Ulsan 44610, Korea

전기 삼륜차의 다양한 부하 조건에서 연료 전지 성능에 대한 실험 연구

할리리스 체사레 아프릴렌드 압디¹ ; 임옥택²^{, †}

1울산대학교 기계공학과 대학원
2울산대학교 기계공학과

Correspondence to: ^† otlim@ulsan.ac.kr

Abstract

This study explores how a Proton Exchange Membrane Fuel Cell (PEMFC) performs in an electric tricycle when subjected to various levels of electrical load. The experiment focuses on two key factors: how much hydrogen the fuel cell system consumes and how the voltage output changes as the load increases. A series of tests were conducted using a 300W proton exchange membrane fuel cell (PEMFC), with the load gradually increased to simulate real driving conditions, from light usage (25%) up to full usage (100%). Throughout the tests, hydrogen consumption was measured using a calibrated mass flow meter, while the voltage was recorded in real-time using a digital Data Acquisition System (DAQ). The results show that hydrogen consumption rises with load, but not always in a straight line, and that higher loads can also cause a noticeable drop in voltage. These findings offer useful insights into the behavior and efficiency of fuel cells in small electric vehicles, especially in urban traffic conditions where load fluctuation can happen frequently. These findings offer a practical insight into optimizing energy management strategies and fuel economy in fuel-cell-based electric mobility solutions, especially in developing regions where lightweight, low-speed transportation like electric tricycles are widely used.

Keywords:

Hydrogen, PEMFC, Fuel Cell Electric Tricycle, Load Variation

1. Introduction

As the world moves toward cleaner and more sustainable energy solutions, transportation remains one of the most important and challenging sectors to decarbonize. Electric vehicles (EVs) have seen quite rapid growth in recent years, offering a clear alternative to the traditional combustion engines^1,2). However, while battery-powered EVs dominate the market right now, hydrogen fuel cells have become an increasing an attractive and quite interesting option, especially for vehicles that need quick refueling, extended range, or consistent power under varying loads³⁾.

Proton Exchange Membrane Fuel Cells (PEMFCs) are known for their efficiency, zero emissions, and quite operation. The qualities they offer make them a good prospect not just for large vehicles such as trucks and buses, but also for smaller vehicles like electric tricycles. Electric tricycles are widely used in urban and semi-urban areas, especially in the developing countries like Southeast Asia, where they were used as affordable and practical alternative solutions. They are lightweight, easy to operate, and often used for short to medium range travel. Replacing or supplementing traditional batteries in these vehicles with PEMFC systems could help reduce emissions in urban areas and improve everyday mobility to become more sustainable^4,5,10). At the same time, it is important to acknowledge that the current cost of PEM fuel cell systems remains one of the main challenges for widespread adoption, especially in low-cost mobility markets.

Fuel cell systems also have their own challenges. One of the biggest issues is how they respond to the changing power demands. In other words, when the motor demands more or less power, like during acceleration, climbing a slope, or stopping, the fuel cell’s hydrogen consumption and output voltage can vary significantly⁶⁾. These changes can affect not only the vehicle’s performance but also the long-term durability of the fuel cell stack such as voltage drops, higher hydrogen consumption, and reduced system efficiency or wear on the fuel cell stack itself overtime^7,8). In many real applications, PEM fuel cells are supported by a battery or supercapacitor because these auxiliary devices can respond quickly to sudden load changes and help protect the stack from unnecessary stress during transient operation. While some studies have looked at how PEMFCs behave during such transitions, most of them are simulation-based research, while the experimental studies under real-world or near-real-world conditions are still limited^4,6,9).

Understanding how PEMFCs behave during dynamic operation is important for several reasons. It helps predict fuel usage more accurately, which is critical in areas where hydrogen refueling stations are still rare. It also allows engineers to design a better control system that can maintain performance and protect the fuel cell stack over time. This is especially important when deploying fuel cells in small or lightweight vehicles like electric tricycles, where weight and space limitations make performance optimization and better fuel cell system design even more critical.

This study aims to contribute to that effort by taking a hands-on, experimental look at how a PEMFC system performs when installed on an electric tricycle and subjected to varying load conditions. By observing the changes in hydrogen consumption and voltage output as the load increases, we hope to provide useful experimental data that reflects real-driving conditions. The goal is not only to evaluate system efficiency, but also to highlight the kinds of performance trade-offs that need to be considered when applying fuel cells to lightweight electric vehicles. The results obtained from this study can help support broader application of hydrogen-based transport, especially in regions like developing countries where small vehicles are an important part of their daily life.

2. Method

2.1 Experimental Setup

The full experimental setup shown in Fig. 1, where the left side of the figure shows the actual tricycle that was used, and in the right side displays the test bench consisting of the hydrogen supply system, fuel cell components, and instrumentation.

Fig. 1.

Experimental setup of the fuel cell electric tricycle and testing station

The fuel cell being used in the center of the system is Horizon PEM FC stack. The fuel cell is supplied with compressed hydrogen from a tank, and its flow is controlled through a simple valve line and regulators to ensure its consistency during operation.

The stack requires an external 13V DC power source to activate its controller. This controller handles operations like hydrogen purging, startup safety checks, and general output regulation. Once powered on, the system’s output is directed to the electric tricycle’s drivetrain, allowing the tricycle itself to serve as the primary load during testing.

The connection between components is outlined in the schematic shown in Fig. 2. The layout includes both the fuel flow and electrical connections between the components. It also shows how the hydrogen source, controller, fuel cell and tricycle (as a load) are all integrated into a single experimental loop.

Fig. 2.

System schematic connecting fuel cell, controller, tricycle, and hydrogen tank

The tricycle used in the experiment was originally powered by a 36 V battery. To adapt it for the fuel cell integration, we modified it by installing a 36 V 350 W Brushless DC (BLDC) motor controller and a DC-DC converter as mentioned in Table 1. Since the FC stack can produce voltages exceeding 40 V during operation, the DC-DC converter was necessary to step down the voltage to 36 V, making it safe to be received by the tricycle’s motor and control electronics. For this test, the tricycle remained stationary, with the front wheel lifted to simulate load without actual movement. This made it easier to monitor real-time data. To evaluate performance, we ran the fuel cel system under four load conditions, 25%, 50%, 75%, and 100%. Instead of riding the tricycle, we create an electrical resistance to simulate the different load levels. The other testing condition details are shown in Table 2 below.

Table 1.

Fuel Cell Electric Tricycle Setup Specification

Table 2.

Experimental Conditions

Each test lasted around 1 minute, during which we collected the voltage and current data in real time using NI LabVIEW. We built a simple LabVIEW interface with a DAQ Assistant block and custom calibration for the sensors, as shown in Fig. 3.

Fig. 3.

LabVIEW-based data acquisition for real-time voltage, current and power monitoring

This setup allowed us to:

• Monitor voltage and current live.
• Calculate power output on the fly.
• Store all data for later analysis.

The voltage and current values were acquired through analog sensors connected to the NI DAQ module. Power output then calculated using Eq. (1).

P = V × I,

and these values were displayed on the LabVIEW interface using real-time charts. To improve accuracy, we applied linear calibration on the current signal, shown in the lower section of the LabVIEW block diagram. Using the values of the power we got previously, we can get the efficiency of the system relative to the stack’s 300 W rated power that was calculated using Eq. (2).

η r a t e d = P 300 W × 100 % .

One limitation of the current setup is the manual measurement of hydrogen consumption using stopwatch timing and visual readings from the flow meter. This approach introduces a degree of human error, particularly for short sampling durations (1 minute). In future work, incorporating digital mass flow sensors with automated data logging could significantly improve measurement accuracy and repeatability. Through this process, we were able to identify:

• How voltage dropped under sudden load increases.
• How current and power output changed with demand.
• How much hydrogen was consumed at each load level.

These observations provide a good starting point for evaluating the fuel cell’s efficiency, especially in practical vehicle scenarios.

3. Results & Discussions

3.1 Voltage Characteristics

The voltage trends of the PEM fuel cell under different intake pressures and load conditions are shown in Fig. 4. Across all scenarios, the voltage gradually decreases as the load increases. This behavior follows typical PEMFC operation, where higher current demand leads to greater activation, ohmic, and concentration losses. At lower loads, the voltage remains relatively stable because the electrochemical reactions proceed without significant resistance or mass-transfer limitations.

Fig. 4.

Fuel cell output voltage versus load percentage under different hydrogen supply pressures

Among the tested pressures, 0.5 bar maintains slightly higher voltages, especially in the mid-load range. This suggests that moderate pressurization provides sufficient oxygen supply to the cathode without introducing unnecessary stress on the system. In contrast, 0.4 bar consistently produces lower voltages. While the 0.6 bar condition shows improved performance at low loads, though the benefit becomes less noticeable as the load increases. This diminishing return implies that increasing the intake pressure beyond an optimal point does not continue to enhance voltage performance. Overall, the voltage characteristics point to 0.5 bar as the most balanced pressure setting for stable stack operation.

3.2 Current Draw

Fig. 5. presents the current output of the fuel cell at each pressure and load level. As expected, the current increases proportionally with load, reflecting the direct control of reaction rate by the external electrical demand. The current output remains steady and predictable, indicating stable load-step operation and consistent stack behavior throughout the tests.

Fig. 5.

Current draw from the fuel cell system under varying load conditions and supply pressures

Differences between pressure conditions become more noticeable at higher loads. While pressure has little influence at 25% load, 0.5 bar supports slightly higher current levels in the 50-75% range/ at 0.6 bar, the current output is comparable to – or slightly below – the 0.5 bar case, suggesting that additional pressure does not provide meaningful gains once stack approaches higher operating currents. At full load, the current values across all pressures converge, showing that the stack reaches a point where further improvements in reactant pressure no longer translate into higher current output. These observations reinforce the idea that moderate intake pressure provides the most efficient reaction environment for the tested operating range.

3.3 Power Output

Fig. 6. summarizes the power output calculated using Eq. (1). Power increases consistently with load for all pressure settings and reaches its maximum at 100% load. Among these conditions, 0.5 bar delivers the highest power output, particularly at middle and upper load levels. This result aligns with the earlier voltage and current findings, confirming that moderate pressurization enhances the electrochemical reaction rate and reduces concentration-related losses.

Fig. 6.

Power output of the fuel cell system at different loads and supply pressures

The 0.4-bar condition produces the lowest power output across the board. Its limited oxygen supply restricts both voltage and current, leading to reduced overall performance. Meanwhile, 0.6 bar shows improved performance at lower loads, but the advantage diminishes as the system approaches full load. This indicates that pressure increases above 0.5-bar does not further enhance the stack’s ability to deliver power and may instead introduce unnecessary parasitic effects. In general, the power output trends highlight the importance of identifying an optimal intake pressure to achieve the best balance between performance and system efficiency.

3.4 Hydrogen Consumption

Hydrogen consumption at different loads and intake pressures is illustrated in Fig. 7. As expected, consumption rises with increasing load, since higher reaction requires more fuel. The overall relationship is nearly linear, indicating stable and predictable fuel utilization throughout the tests.

Fig. 7.

Hydrogen consumption rate (L/min) at varying load percentages across different supply pressures

The 0.6-bar condition shows slightly higher hydrogen consumption, which is reasonable given the increased supply pressure and potential for additional parasitic gas flow. Although higher pressure helps the stack at low loads, it also tends to increase fuel usage without offering significant benefits at higher loads. On the other hand, 0.5-bar demonstrates the most efficient balance between consumption and delivered power, aligning with its superior voltage and power performance. The 0.4-bar condition shows lower hydrogen consumption, but this reduced usage comes at the cost of noticeably lower power output. When normalized to performance, operating at 0.4 bar is therefore less favorable.

Taken together, these results suggest that 0.5 bar provides the most efficient fuel utilization relative to electrical output, supporting its selection as the optimal intake pressure for this system.

3.5 Efficiency

The system efficiency relative to the 300 W rated output of the H-300 PEM fuel cell was calculated using the power obtained from Eq. (1), and the efficiency was expressed using Eq. (2).

Fig. 8. presents the resulting efficiency values for all intake pressures and load levels. A clear upward trend is observed: efficiency increases with load because the fuel cell delivers a larger proportion of its rated capacity as electrical demand rises.

Fig. 8.

Efficiency of the PEM fuel cell relative to its rated power under varying load and supply pressures

Among the pressure conditions, 0.4 bar consistently produces the lowest efficiencies, with minimal improvement until the highest load. The 0.5 bar condition achieves the highest and most stable efficiencies, especially at 50-75% load, reaching around 16% at full load. This aligns with the earlier findings that 0.5 bar provides the most balanced operating environment. The 0.6 bar condition shows the highest efficiency at low load. However, as the load increases, the advantage diminishes and the efficiency converges with the 0.5-bar results. This again confirms that additional pressurization does not yield proportionally better performance at higher currents.

4. Conclusions

This study examined how a 300 W PEM fuel cell performs under different electrical load levels when integrated into an electric tricycle system. The results show a clear and predictable trend: as the load increases, the stack voltage steadily declines, driven by the combined effects of activation, ohmic, and concentration losses. Adjusting the hydrogen supply pressure helps at lower loads, but at higher loads the voltage drop remains largely unavoidable, indicating that load demand—rather than intake pressure—is the dominant factor shaping the stack’s electrical behavior.

Hydrogen consumption also increased with load, following a near-linear pattern. Among the tested conditions, 0.5 bar consistently provided the most balanced outcome, offering a stable voltage profile, strong power delivery, and more efficient hydrogen usage. The results also highlight the role of the load itself as the primary driver of current draw from the fuel cell. This insight is important for lightweight electric vehicles, because it underscores the need for coordinated power management between the fuel cell, DC-DC converter, and motor controller. Taken together, the findings indicate that operating the system at a moderate intake pressure, particularly around 0.5 bar, provides the best balance between performance, fuel efficiency, and system stability.

For real-world applications, this suggests that designers and manufacturers should emphasize maintaining optimal pressure levels and consider integrating automatic flow-control mechanisms that respond to the vehicle’s power demand in real time. Such improvements could reduce unnecessary hydrogen consumption and make fuel-cell-powered tricycles a more practical and cost-effective option, especially in regions where affordability and infrastructure limitations remain key challenges.

Acknowledgments

• This work was supported by Regional Innovation Cluster Development (R&D) by the Ministry of Trade, Industry and Energy (MOTIE, Korea) [Project Name: Open Innovation Project for Convergence Industry of Battery/Fuel Cell for Mobility Electrification and Energy Production/Storage (P0025406)].
• This result was supported by the “Regional Innovation System & Education (RISE)” through the Ulsan RISE Center, funded by the Ministry of Education (MOE) and the Ulsan Metropolitan City, Republic of Korea. (2025-RISE-07-001).
• This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. RS-2023-00217778).
• This work was supported by the Technology Innovation Program (or Industrial Strategic Technology Development Program-The Safety Based Technology Development and Substantiation of Small Hydrogen Powered Vessel) (‘RS-2022-00142947, The technology development on fuel cell electric propulsion system using Land Based Test Site) funded By the Ministry of Trade, Industry & Energy (MOTIE, Korea).

References

J. Yan, C. Zhou, Z. Rong, H. Wang, H. Li, and X. Hu, “Simulation of the dynamic characteristics of a PEMFC system in fluctuating operating conditions”, Energies (Basel), Vol. 13, No. 14, 2020, pp. 3596. [https://doi.org/10.3390/en13143596]
A. Loskutov, A. Kurkin, A. Shalukho, I. Lipuzhin, and R. Bedretdinov, “Investigation of PEM fuel cell characteristics in steady and dynamic operation modes”, Energies (Basel), Vol. 15, No. 19, 2022, pp. 6863. [https://doi.org/10.3390/en15196863]
M. Tomas, P. Novotny, F. Gholami, O. Tucek, and F. Marsik, “A comparative study of dynamic load response of high temperature PEM fuel cells”, Environmental and Climate Technologies, Vol. 24, No. 1, 2020. [https://doi.org/10.2478/rtuect-2020-0033]
R. Eini and S. M. Rakhtala, “Nonlinear modeling of a PEM fuel cell system; a practical study with experimental validation”, International Journal of Mechatronics, Electrical and Computer Technology, Vol. 4, No. 10, 2014. [https://doi.org/10.48550/arXiv.2501.08420]
A. Kulikovsky, “Performance of a PEM fuel cell with oscillating air flow velocity: a modeling study based on cell impedance,” eTransportation, Vol. 7, 2021, pp. 100104. [https://doi.org/10.1016/j.etran.2021.100104]
S. Ekici and M. Kabir, “Enhanced performance prediction for proton exchange membrane fuel cells: A comprehensive study with different load profiles”, International Journal of Hydrogen Energy, Vol. 143, 2025, pp. 1031-1042. [https://doi.org/10.1016/j.ijhydene.2024.12.348]
W. Chen, Y. Liu, and B. Chen, “Numerical simulation on pressure dynamic response characteristics of hydrogen systems for fuel cell vehicles”, Energies (Basel), Vol. 15, No. 7, 2022, pp. 2413. [https://doi.org/10.3390/en15072413]
U. R. Singh and S. Bhogilla, “Experimental evaluation of a 1 kW proton exchange membrane fuel cell with resistive, DC and AC loads”, International Journal of Hydrogen Energy, Vol. 52, 2024, pp. 1345-1355. [https://doi.org/10.1016/j.ijhydene.2023.06.162]
Y. Li, X. Chen, Y. Liu, D. Xiong, J. Li, S. Yin, L. Chen, C. Li, and J. Xu, “Analytical modeling framework for performance degradation of PEM fuel cells during startup-shutdown cycles”, RSC Advances, Vol. 10, No. 4, 2020, pp. 2216-2226. [https://doi.org/10.1039/c9ra09572a]
H. Ryu, W. Yoon, S. Lee, J. Hwang, J. Oh, J. Park, and H. S. Kim, “A study on power distribution strategy between fuel cell and battery for 50 kW construction equipment using 1D simulation”, Journal of Hydrogen and New Energy, Vol. 35, No. 6, 2024, pp. 706-714. [https://doi.org/10.7316/JHNE.2024.35.6.706]

Parameters	Value
Fuel Cell Rated Power	300 W
Fuel Cell Rated Voltage	36 V
Fuel Cell Controller	Horizon proprietary controller with start/stop, purge valve, and safety features
DC-DC Converter	Input: 52 V DC, Output: 36 V DC (without load)
BLDC Motor Controller	Rated for 36 V, 350 W BLDC motor
Tricycle	AU Tech Scanic Trike 36 V 10 A Electric Tricycle
Motor Type	36 V 350 W
DAQ System	National Instruments (NI) with LabVIEW
Sensors Used	Voltage Sensor, Current Sensor, configured via DAQ Assistant

Specification	Value
Load levels	25%, 50%, 75%, 100%
Pressure (bar)	0.4, 0.5, 0.6
Sampling time (min)	1