New Standards Reshape Fuel Cell Vehicle Cold Start Testing Globally
The race to perfect fuel cell electric vehicles (FCEVs) for real-world winter conditions has taken a decisive turn with the release of two major testing standards in 2023: China’s national standard GB/T 43255 and the international ISO 17326. These documents, though sharing a common goal of evaluating a vehicle’s ability to start and operate in sub-zero temperatures, reveal profound technical and philosophical differences that reflect the divergent development paths of the global automotive industry. For manufacturers, engineers, and policymakers, understanding these nuances is no longer optional—it is critical for global market access, product development, and regulatory compliance.
The cold start capability of a fuel cell vehicle is not merely a technical checkbox; it is a fundamental determinant of consumer confidence and market viability. A vehicle that fails to start reliably in winter is, for all practical purposes, unusable. This is especially true in regions with harsh winters, where temperatures routinely plunge below -20°C. The stakes are high, and the standards are the rulebooks by which success or failure is measured.
The newly published Chinese standard, GB/T 43255—2023, titled “Test methods for sub-zero cold start performances of fuel cell electric vehicles,” represents a significant milestone for China’s burgeoning hydrogen vehicle industry. Developed by the National Technical Committee on Automotive Standardization, it provides a comprehensive, domestically tailored framework for evaluating FCEV performance under extreme cold. In parallel, the International Organization for Standardization released its updated ISO 17326:2023, “Fuel cell road vehicles—Cold start performances under sub-zero temperature—Vehicles fuelled with compressed hydrogen.” While both standards aim to assess the same core functionality, a detailed comparison reveals they are far from identical. Their differences touch on everything from test procedures and data acquisition to safety protocols and even the fundamental architecture of the vehicles being tested.
One of the most striking divergences lies in the underlying vehicle architectures each standard implicitly assumes. The ISO standard is largely designed with “full-power” fuel cell vehicles in mind—vehicles where the fuel cell stack is the primary, if not sole, source of propulsion power. In contrast, the Chinese GB/T standard is built around the “electric-electric hybrid” model, which is currently the dominant architecture in China. In this setup, a smaller fuel cell stack works in tandem with a high-capacity battery pack, sharing the load and providing redundancy. This architectural difference is not trivial; it fundamentally changes how a vehicle behaves during a cold start and, consequently, how it should be tested.
This divergence is immediately apparent in the “low-temperature soaking” phase, the first critical step in both test protocols. Both standards agree on the basics: the test vehicle must be soaked in a controlled cold environment for at least 12 hours to ensure the fuel cell stack and all critical components reach thermal equilibrium with the target temperature, typically -20°C or lower. Both also wisely allow for a brief vehicle start-up and shutdown before the official soaking begins, enabling the vehicle’s systems to perform self-diagnostics and adjust to the external environment. Furthermore, both standards show pragmatism by not mandating a specific state of charge for the vehicle’s battery or a specific fill level for the hydrogen tank prior to soaking, recognizing that these variables have a negligible impact on the cold start process itself.
However, the ISO standard introduces a provision that is entirely absent from the Chinese standard: the option to disconnect the vehicle’s internal hydrogen system and connect it to an external hydrogen supply for the purpose of accurately measuring hydrogen consumption during the test. This feature is a direct concession to the needs of full-power FCEV developers, who require precise, isolated measurements of fuel cell efficiency without the confounding variables of an onboard hydrogen storage and delivery system. The Chinese standard, focused on the hybrid model, sees no need for this complexity and mandates that all tests be conducted using the vehicle’s internal hydrogen supply, reflecting a more holistic, real-world approach to testing.
The differences become even more pronounced in the “low-temperature cold start” phase, where the vehicle is actually commanded to start. Here, both standards converge on the definition of a successful start: the vehicle’s dashboard must display a “READY” or “OK” indicator, and the fuel cell stack must sustain an output of at least 1 kilowatt for a minimum of 10 consecutive minutes. This is a sensible, performance-based criterion that focuses on the outcome rather than the process.
Yet, the standards part ways dramatically on safety and data recording. The Chinese GB/T 43255 standard imposes a stringent safety requirement: it mandates continuous monitoring and recording of the hydrogen concentration in the vehicle’s tailpipe exhaust during the entire cold start process. The recorded data must show that the 3-second average hydrogen concentration never exceeds 4%, and the instantaneous concentration must never spike above 8%. This is a critical safety net, designed to prevent the accumulation of explosive hydrogen-air mixtures in enclosed spaces like garages. The standard also requires that the vehicle must not trigger any fault or warning alarms during the process. In stark contrast, the ISO 17326 standard makes no mention of exhaust hydrogen monitoring. Its safety philosophy appears to be more permissive, perhaps assuming that modern vehicle control systems are inherently safe. Instead, the ISO standard focuses on hydrogen consumption, requiring it to be recorded only if an external hydrogen supply is used.
The “low-temperature cold start driving” phase, which tests the vehicle’s ability to transition from a stationary start to actual driving, further highlights the standards’ different priorities. Both require the vehicle to be in a “READY” state and then commanded to accelerate with full throttle. However, the Chinese standard defines a successful “start” very specifically: the fuel cell stack’s output power must reach 50% of its rated system power. After achieving this, the test protocol is highly structured: the driver must stop the vehicle within one minute and then, within three minutes, complete at least one full Chinese driving cycle before shutting down the system. This rigid, repeatable procedure is designed to generate comparable data across different test labs and different vehicle models.
The ISO standard, on the other hand, is far more flexible and regionally adaptive. It does not specify a fixed power threshold for success. Instead, it defers to local driving cycles, instructing testers to use the driving cycle that is representative of the region where the vehicle will be sold. This approach acknowledges the global nature of the automotive market and the fact that driving patterns in Germany, Japan, and the United States are vastly different. While this flexibility is commendable, it also introduces a significant challenge: it makes direct, apples-to-apples comparisons between vehicles tested in different regions nearly impossible.
The criteria for terminating a test also reveal subtle but important differences. Both standards agree that a test should be halted if the vehicle’s dashboard displays a stop command, if any fault or warning alarms are triggered, or if the vehicle cannot achieve its declared maximum speed under the test conditions. These are logical, safety-driven termination points.
The ISO standard, however, adds an additional, more technical criterion: the test must also be terminated if the vehicle’s performance deviates from the standard’s specified tolerances while operating in sub-zero temperatures. This is a quality control measure, ensuring that the vehicle not only functions but functions within a tightly defined performance envelope. The Chinese standard, in its current form, does not include this level of performance tolerance checking, focusing instead on binary pass/fail outcomes based on startup and basic driving capability.
Perhaps the most technically significant difference lies in the requirements for data acquisition. Both standards mandate the use of a hydrogen-safe, low-temperature chassis dynamometer environmental chamber to conduct the tests, a non-negotiable safety requirement. However, they diverge sharply on the granularity and source of the data to be collected.
The Chinese GB/T 43255 standard requires that key electrical parameters—current and voltage—be sampled at a minimum frequency of 5 Hertz (5 times per second). For vehicles where the fuel cell stack is highly integrated and its direct current and voltage cannot be measured, the standard provides a pragmatic workaround: testers may measure the output of the DC/DC converter and then calculate the stack’s power by assuming the converter has a 97% efficiency. This is a clever, engineering-focused solution that acknowledges the realities of modern vehicle design.
The ISO 17326 standard, however, demands a higher sampling frequency of at least 10 Hertz, effectively requiring twice as much data to capture the rapid dynamics of a cold start. More importantly, it permits the use of data directly from the vehicle’s Engine Control Unit (ECU). This is a fundamentally different approach. ECU data is processed, filtered, and often represents a “best estimate” rather than a raw, direct measurement. While this can be more convenient, it also introduces a layer of abstraction and potential error. The Chinese standard’s preference for direct, physical measurements reflects a more conservative, traceable approach to data integrity.
To illustrate the practical application of these standards, researchers led by Di Wu and Linghai Han conducted a real-world test on an internationally recognized, advanced FCEV model, subjecting it to the rigorous requirements of the Chinese GB/T 43255—2023 standard at -20°C. The vehicle was placed in a specialized hydrogen-safe, low-temperature environmental chamber and soaked for the mandatory 12 hours. It was then started according to the manufacturer’s instructions.
The results were impressive. The vehicle achieved its “READY” state in just 10.03 seconds after the start button was pressed, easily meeting the standard’s success criteria. More interestingly, the recorded current and voltage curves revealed a sophisticated, three-stage startup process. Approximately 3.4 seconds after ignition, the fuel cell stack began generating voltage. Between 3.4 and 10 seconds, the system entered a “warm-up” phase, during which it deliberately operated at high power—peaking at 35 kW—to rapidly generate heat and bring the stack up to its optimal operating temperature. As the stack warmed, the power output was gradually dialed back to a stable 20 kW, demonstrating a sophisticated control strategy designed to balance rapid startup with long-term stack health. This level of granular performance data, captured thanks to the standard’s detailed data acquisition requirements, provides invaluable insights for engineers looking to optimize future designs.
The publication of these two standards marks a pivotal moment for the global FCEV industry. It signals a transition from the pioneering, “anything goes” phase of development to a more mature, regulated era where performance and safety are rigorously defined and measured. For Chinese manufacturers, GB/T 43255 provides a clear, domestically relevant benchmark that aligns with their dominant hybrid vehicle architecture. It is a tool for driving domestic innovation and ensuring a baseline level of quality and safety for consumers in the Chinese market.
For international manufacturers and those aiming for global sales, ISO 17326 offers a more flexible, internationally recognized framework. However, its flexibility comes with a cost: the lack of uniformity in test cycles and its permissive stance on safety monitoring (like hydrogen exhaust) may lead to inconsistencies. A vehicle that passes the ISO test in one country might not meet the more stringent safety or data requirements of another market, such as China.
This situation creates a complex compliance landscape. Manufacturers who wish to sell their FCEVs globally may find themselves having to design and test their vehicles against multiple, sometimes conflicting, standards. This is not merely an engineering challenge; it is a significant financial and logistical burden. It underscores the urgent need for greater international harmonization of testing protocols. While the existence of two standards reflects healthy competition and diverse technical approaches, the long-term health of the global FCEV market depends on finding common ground.
Looking ahead, the evolution of these standards will be closely watched. The Chinese standard, with its emphasis on direct measurement, stringent safety monitoring, and structured test cycles, may influence future revisions of the ISO standard, particularly as safety concerns around hydrogen become more prominent in public discourse. Conversely, the ISO standard’s flexibility and focus on regional adaptability offer valuable lessons for any standard that aspires to global relevance.
Ultimately, the goal of both standards is the same: to ensure that fuel cell vehicles are not laboratory curiosities but reliable, everyday transportation solutions, even in the coldest winters. The detailed analysis provided by Di Wu, Shiyu Wu, Yupeng Wang, Honghui Zhao, and Linghai Han in their study is an invaluable resource for the industry. It doesn’t just list differences; it provides context, explains the rationale, and offers practical guidance for navigating this new regulatory terrain. Their work will undoubtedly help engineers refine their designs, help testers conduct more accurate and meaningful evaluations, and, most importantly, help bring safer, more reliable fuel cell vehicles to consumers around the world.
By Di Wu, Shiyu Wu (Testing and Certification CATARC Automotive Test Center/ Tianjin Co., Ltd.), Yupeng Wang, Honghui Zhao, Linghai Han (General R&D Institute of China FAW Group Co., Ltd.). Published in Journal 2024 Issue 7 (Part 2) / Total Issue 659. DOI: 10.3969/j.issn.1002-5944.2024.14.029