Life Cycle Assessment Based on Whole Industry Chain Assessment of FCEVs

Abstract

Fuel cell electric vehicles (FCEVs) offer a promising solution for energy saving and emission reduction in transportation. However, several challenges must be addressed for their application. This study conducts a full life cycle assessment (LCA) of FCEVs, dividing it into the fuel cycle and vehicle cycle to separately assess energy consumption (EC) and emissions. The fuel cycle examined 18 hydrogen production–storage–transport pathways, while the vehicle cycle evaluates energy use and emissions associated with vehicle component production, assembly, disposal, battery production, and fluid consumption. Based on the GREET database, total energy consumption and emissions over a lifetime were calculated. Five environmental impact indicators were used for evaluation, and a comprehensive environmental assessment (CEA) indicator was established for different scenarios. Results indicate that nuclear thermochemical water splitting is the best hydrogen production method, and pipeline transportation is the most efficient for hydrogen transport. Additionally, water electrolysis for hydrogen production is only practical when paired with renewable energy. The study also identified that the Hydrogen production method, “Body”, “Proton Exchange Membrane Fuel Cells (PEMFCs) System”, “Chassis”, “Hydrogen Storage System” and lifetime significantly impact energy consumption and emissions. These stages or products represent high-impact leverage points for enhancing the lifecycle sustainability evaluation of FCEVs.

1. Introduction

Reducing carbon emissions is a primary global strategy for combating climate change [1]. As global energy demands increase, carbon dioxide emissions are intensifying, underscoring the limitations of fossil fuels and the urgent need for reform in energy systems, particularly in the transportation sector. Solaymani et al. [2] investigated the significance of the transportation sector in global fossil energy consumption and CO₂ emissions; their study revealed that in 2024, global CO₂ emissions totaled 34,116.8 Mt, with the transportation sector accounting for 7941.0 Mt. The results underscore that the transportation sector remains a critical contributor to global CO₂ emissions, representing 23.0% of the total emissions. Therefore, a global transformation of the energy system is imperative, particularly within the transportation sector. Renewable energy sources are being developed and deployed at visible scale and pace to replace conventional fossil fuels. The International Energy Agency (IEA) and the Hydrogen Council emphasize hydrogen’s potential role in reducing emissions [3,4]. FCEVs, which offer high efficiency and zero emissions, are pivotal for emission reduction in the transportation sector. As of February 2025, a total of 18,620 FCEVs have been sold and leased in the United States [5]. Japan’s hydrogen energy strategy indicates that the number of FCEVs in Japan is expected to reach 800,000 by 2030, and there will be 900 hydrogen refueling stations. [6]. Guo et al. [7], in their study, utilized the generalized Bass diffusion model, incorporating historical FCEVs production and sales volumes in China, the expansion of hydrogen refueling stations, and potential price reductions, to project that the cumulative demand for FCEVs in China will reach approximately 49,900 units by 2025.

In road transportation, the focus of vehicle carbon emissions has shifted from usage to raw material acquisition with new energy vehicles. Therefore, evaluating the environmental impact of new energy vehicles is essential. To enable more comprehensive regulation of automotive energy consumption and emissions, the European Union formally enacted the “Proposal for a Regulation on Life Cycle Assessment of Vehicles (COM/2023/321)” in 2023. This regulation mandates that, effective 1 January 2026, all new vehicles must disclose emissions data based on standardized LCA methodologies, covering the entire lifecycle from raw material extraction to production, use, and end-of-life recycling.

FCEVs are characterized by zero-emission operation during use; however, their environmental impact is critically dependent on hydrogen production methods. When renewable energy sources are employed for hydrogen generation, greenhouse gas emissions can be substantially reduced. Conversely, hydrogen derived from fossil fuels without carbon mitigation measures may even result in higher lifecycle emissions compared to internal combustion engine vehicles (ICEVs) [8]. Disappointingly, hydrogen saw global demand reach approximately 100–105 Mt in 2024, marking a 5–7% year-on-year increase. Nevertheless, the majority of hydrogen production still relies on fossil fuels without carbon capture and storage (CCS) technology. Low-emission hydrogen (green and blue hydrogen) accounts for only about 2% of total production [9]. As a result, hydrogen production is a critical step in utilizing hydrogen energy, and its LCA is of great significance to FCEVs. Focusing solely on FCEVs usage while isolating the discussion from hydrogen production methods is an incomplete and analytically misleading approach.

Most LCA studies on FCEVs primarily focus on hydrogen production. Folega et al. [10] analyzed the life cycle greenhouse gas (GHG) emissions of FCEVs using coke oven gas (COG) as a hydrogen source. Yang et al. [11] assessed the impacts of different hydrogen production methods and driving distances on sustainability. Yu et al. [12] calculated energy consumption and emissions for twelve hydrogen pathways using the GREET model for FCEVs, internal combustion engine vehicles, and electric vehicles. Hao et al. [13] analyzed GHG emissions of 19 technological pathways for deploying FCEVs in China. Maniscalco et al. [14] compared emissions from hydrogen production using various sources.

While FCEVs produce no emissions during their operational phase, their vehicle architecture and powertrain configurations contribute to varying levels of manufacturing-phase emissions. Furthermore, these design choices influence hydrogen consumption rates during use, thereby indirectly generating emissions through hydrogen production processes [15]. Research in [16] quantitatively confirms that increased vehicle mass leads to elevated hydrogen demand during operation. Thus, it is an oversimplification to attribute FCEVs’ emissions solely to the hydrogen sourcing stage. Bai et al. [17] conducted a well-to-wheel (WTW) sensitivity analysis to compare the lifecycle costs and carbon emissions of FCEVs, ICEVs, and BEVs; their study primarily focused on three critical parameters—electricity price, hydrogen storage duration, and transportation distance—revealing nuanced trade-offs between technology pathways. Agostini et al. [18] assessed the environmental and cost benefits of auxiliary equipment for FCEVs. Savran et al. [19] analyzed the energy consumption and global warming potential (GWP) of FCEVs and battery electric vehicles (BEVs) across different vehicle sizes during their production and use phases. Chen et al. [20] compared the environmental impacts of various bipolar plates, energy control strategies, and hydrogen production methods. Specific studies have focused on propulsion systems’ lifecycle carbon emissions in FCEVs. Cox et al. [21] assessed the lifecycle environmental burdens and total cost of ownership for 2017 and 2040 passenger vehicles with different propulsion systems. Lu et al. [22] compared the energy efficiency of hydrogen FCEVs to gasoline and battery electric vehicles, noting higher efficiency with renewable hydrogen and lower efficiency with methanol-based hydrogen and grid electricity electrolysis. Hydrogen transportation mode and distance are critical; tube trailers are suitable for medium and short distances, while liquefaction or pipelines are preferred for long distances. Devanathan et al. [23] analyzed carbon emissions from FCEVs power systems using LCA. Moholkar et al. [24] conducted a comparative LCA of FCEVs and BEVs within the Indian context. Their findings reveal that green hydrogen-powered FCEVs achieve the lowest carbon emissions, with blue hydrogen pathways also outperforming BEVs reliant on India’s grid electricity mix.

In summary, most existing LCA studies on FCEVs narrowly focus on evaluating hydrogen production methods or assessing vehicles under fixed hydrogen supply pathways. While some studies explore the sensitivity of key FCEVs components, few have investigated comprehensive LCA spanning the entire supply chain—from hydrogen production, storage, and transportation to the manufacturing and end-of-life management of vehicle bodies and critical components. As one of the new energy vehicles, FCEVs exhibit a distinct characteristic of upstream emission shifting within their supply chains. Additionally, their vehicle architecture and powertrain system designs generate emissions both through direct manufacturing processes and indirectly by influencing hydrogen consumption efficiency.

This study aims to address the research gap through a full industrial chain LCA of FCEVs, examining their energy consumption and emissions across the entire lifecycle under diverse hydrogen sourcing pathways. The primary contribution lies in comprehensively evaluating the energy conservation and emission reduction performance of FCEVs across the supply chain, identifying high-impact leverage points for energy conservation and emission reduction, and providing actionable recommendations for their practical benefits and future development trajectories. The paper explicitly underscores the urgency of this research and outlines the following specific objectives:

• Develop a CEA framework to harmonize multi-scale environmental indicators and systematically evaluate the integrated environmental footprint of products or processes.
• Optimize hydrogen sourcing pathways by analyzing energy efficiency and emission profiles across production methods, with explicit quantification of CCS efficacy in mitigating lifecycle emissions.
• This study evaluates the comprehensive environmental impacts of FCEVs across their entire lifecycle. By employing a Sobol-based sensitivity analysis, the contributions of hydrogen production methods, transportation pathways, and vehicle lifespan to environmental impacts are investigated. This approach identifies key lifecycle stages with the greatest potential for energy conservation and emission reduction.

2. Methodology

2.1. Environmental Indicators

The concept of the carbon footprint was first introduced by Wackernagel [25], who quantified the environmental impact of societal activities using the area of bioproductively active land. Researchers such as Wiedmann et al. [26] defined it in the 1970s as the total CO₂ emissions from all life cycle stages of a product, including direct and indirect emissions. Hammond et al. [27] defined it as the total emissions from an individual or process. Fu et al. [28] developed a model assessing the energy-based costs and carbon footprint across a product’s full lifecycle using carbon dioxide equivalents CO₂ -eq).

Pollutants that contribute to global climate change are diverse, each possessing different lifespans and potencies, thereby influencing their specific effects on the climate. These effects are not solely determined by the quantity of their emissions into the atmosphere. Climate indices such as GWP and CO₂-eq are used to measure or compare the impact of different pollutants on global warming. Both GWP and carbon dioxide equivalents are the most commonly used metrics in climate analysis and policymaking. The GWP of a pollutant refers to its effectiveness as a driver of climate change relative to carbon dioxide over a specific time period. This metric is calculated by comparing the cumulative radiative forcing caused by the pollutant to that caused by an equal mass of carbon dioxide over a designated timeframe.

The impacts of non-CO₂ emissions on global warming are measured using GWP weight values or CO₂-eq. In this paper, the GWP weight values are used to quantify the impact of all emissions, including CO₂, converting them into a single GWP value for global warming, as shown in (1).

$$GW{P}_{value}=\sum _1^n{E}_i·GW{P}_{i,time}$$

(1)

where $GW{P}_{value}$ represents the value of global warming potential. Assuming there are n types of emissions that impact the GWP value, $E_i$ represents emission i, and $GW{P}_{i,time}$ represents the GWP weight value of that emission over a specific time period, denoted as $time$.

It is important to note that if a specific time period is not designated, the default time frame is usually assumed to be 100 years. In the paper, this default assumption will be followed, $GW{P}_i=GW{P}_{i,100}$. The GWP value varies depending on the time period considered, especially when the lifespan of a greenhouse gas differs significantly from that of carbon dioxide.

Within the same timeframe, the values of GWP may also vary slightly due to advances in underlying science, such as changes in the efficacy of gases or their atmospheric residence time. GWP values are typically reported in IPCC assessment reports. The concept of GWP was introduced and refined in the IPCC’s Second Assessment Report (1995) [29]. Since then, the definition of carbon footprint in most studies has expanded beyond merely CO₂ itself to include emissions like CH₄, N₂O, and fluorinated gases that may contribute to global warming.

Table 1. Common values of GWP for methane and nitrous oxide.

IPCC	Time	CO2	CH4	N2O
AR4 2007	GWP20	1	72	289
	GWP100	1	25	298
AR5 2013	GWP20	1	85	264
	GWP100	1	30	265
AR5 2021	GWP20	1	82.5	273
	GWP100	1	29.8	273

[30,31,32] lists the GWP values for common greenhouse gases.

In addition to the widely recognized GWP, numerous other environmental impact indicators are employed to quantify the effects of pollutants on the environment, such as human toxicity potential (HTP), photochemical smog potential (POCP), aerosol formation potential (AFP), and acidification potential (AP). In this paper, the weighting factor for GWP is based on the AR5 2021 standards, while the weighting factors for all environmental indicators are as shown in

Table 2. Weight coefficient of common pollutants in various environmental indicators.

Environmental Indicators	Emissions	Weight Coefficient
Global Warming Potential (100)	CO2	1
	CH4	29.8
	N2O	273
Human toxicity potential	CO	0.012
	${\mathrm{S}\mathrm{O}}_x$	0.780
Photochemical smog potential	CH4	0.006
	${\mathrm{N}\mathrm{O}}_x$	0.028
Aerosol formation potential	PM10	1.000
	PM2.5	1.000
Acidification potential	${\mathrm{S}\mathrm{O}}_x$	2.000
	${\mathrm{N}\mathrm{O}}_x$	0.700

[12].

In this study, the lifecycle emissions of hydrogen fuel cell vehicles are evaluated using five environmental indicators: GWP, HTP, POCP, AFP, and AP. For these environmental indicators, each is defined similarly to the GWP value as follows:

$$Indicato{r}_{value}=\sum _1^n{E}_i·Indicato{r}_i$$

(2)

where $Indicato{r}_{value}$ represents the value of an environmental indicator, $E_i$ represents emission i, and $Indicato{r}_i$ represents the corresponding weight value for the environmental indicator.

2.2. Life Cycle Assessment

International standards ISO 14040 [33] and ISO 14044 [34] explicitly divide process-based life cycle assessment into four main phases: goal and scope determination, life cycle inventory analysis, life cycle impact assessment and result analysis, as show in Figure 1 [35]. Next, this paper will evaluate FCEVs around these four steps of LCA.

The primary objective of this study is to construct a comprehensive LCA of FCEVs from a full supply chain perspective, based on the database provided by GREET [36]. This assessment comprehensively considers both the fuel cycle and the vehicle cycle. Among these, the fuel cycle encompasses a total of 18 hydrogen fuel pathways. In this study, hydrogen is maintained in the gaseous state across all scenarios. The analysis encompasses nine hydrogen production pathways: coal gasification (without CCS), coal gasification with carbon capture and storage (CCS), steam methane reforming (SMR) of natural gas (without CCS), SMR with CCS, coke oven gas reforming, nuclear-powered thermochemical water splitting, water electrolysis using wind/solar electricity, hydropower-based electrolysis, and grid-powered electrolysis (global average electricity mix). Two transportation modes are evaluated: pipeline transmission and high-pressure gaseous tube trailer delivery, with detailed paths provided in

Table 3. Hydrogen fuel production pathways.

Path	Hydrogen State	Hydrogen Source	Hydrogen Transport
1	Gaseous	Coal	Tube trailer
2	Gaseous	Coal	Pipeline
3	Gaseous	Coal (with CCS 1)	Tube trailer
4	Gaseous	Coal (with CCS)	Pipeline
5	Gaseous	Natural Gas	Tube trailer
6	Gaseous	Natural Gas	Pipeline
7	Gaseous	Natural Gas (with CCS)	Tube trailer
8	Gaseous	Natural Gas (with CCS)	Pipeline
9	Gaseous	Coke Oven Gas	Tube trailer
10	Gaseous	Coke Oven Gas	Pipeline
11	Gaseous	Nuclear Energy (Thermo-Chemical Cracking of Water)	Tube trailer
12	Gaseous	Nuclear Energy (Thermo-Chemical Cracking of Water)	Pipeline
13	Gaseous	PEM Electrolysis (Solar or Wind)	Tube trailer
14	Gaseous	PEM Electrolysis (Solar or Wind)	Pipeline
15	Gaseous	PEM Electrolysis (Hydro)	Tube trailer
16	Gaseous	PEM Electrolysis (Hydro)	Pipeline
17	Gaseous	PEM Electrolysis (World Average Mix)	Tube trailer
18	Gaseous	PEM Electrolysis (World Average Mix)	Pipeline

¹ CCS: Carbon Capture and Storage.

.

In Table 3, CCS stands for “Carbon Capture and Storage.” It is a technology aimed at reducing carbon dioxide emissions and consists of three main steps: capture, transport, and storage. First, CO₂ is separated from industrial or energy-related sources (such as coal-fired power plants and industrial processes). Next, the captured CO₂ is transported via pipelines, ships, or other means to the storage site. Finally, the CO₂ is injected into underground geological formations (such as depleted oil and gas fields or deep saline aquifers) for long-term storage, preventing its release into the atmosphere. CCS technology is considered a crucial measure for combating climate change and reducing greenhouse gas emissions, particularly in high-emission industrial sectors.

In this study, paths 9 and 10 utilize nuclear energy for hydrogen production through thermochemical water splitting. Paths 13 to 18 are based on PEM (proton exchange membrane) electrolysis for hydrogen production, with the main differences being the sources of electricity used. Three types of electricity sources are considered in this study: solar or wind power (solar or wind), hydropower (hydro), and the world average electricity mix (world average mix). The world average electricity mix refers to the current average cost and emissions of electricity generation globally.

As previously mentioned, this paper evaluates GWP, HTP, POCP, AFP, and AP as assessment indicators. Concurrently, it delineates the system boundaries for the LCA of FCEVs based on the entire industry chain. The entire industry chain process is divided into the fuel cycle and the vehicle cycle. The inputs of the entire system include primary energy sources, water, metal raw materials, and other raw materials, while the outputs are pollutants and other waste discharged into the environment. The system boundary is shown in Figure 2. It should be noted that this study traces the industrial chain only up to the primary energy source. It does not consider the energy consumption and emissions associated with equipment, facilities, and maintenance, such as those from solar panels, wind turbines, and vehicle maintenance during their usage period.

In the fuel cycle, this paper explores 18 hydrogen production pathways. Each pathway includes the refining and processing of primary energy, hydrogen production, hydrogen purification, hydrogen storage, hydrogen transportation, and hydrogen refueling. Moreover, each hydrogen production method is paired with two different hydrogen transportation methods: tube trailers and pipeline transport, with hydrogen being in a high-pressure gaseous state. In this context, producing and transporting 1 kg of hydrogen over a distance of 1000 km is defined as the functional unit.

The vehicle cycle is primarily calculated based on the “Passenger Cars 1: FCV (Conventional Material)” data provided by GREET. This paper includes vehicle production, vehicle use, and vehicle disposal within the scope of the study.

The full life cycle analysis of the vehicle itself is conducted using data provided by GREET2 2022 as the standard. The total emissions of each vehicle are distributed over its lifespan and analyzed per kilometer of usage. Below, the composition and detailed data for “Passenger Cars 1: FCV (Conventional Material)” from GREET are provided in

Table 4. Detailed parameters of FCV (Conventional Material).

Vehicle Parameter	Detail
Lifetime VMT of a Vehicle	278,662.7 km
Total Vehicle Weight	1627.03 kg
Fuel Cell Stack Size (PEMFC)	102 kW
Battery Size 1	38 kW
Battery Weight (Li-Ion)	21.95 kg
Battery Weight (Lead-Acid)	10.02 kg
Fluids Weight 2	25.23 kg
PEMFCs System (including BOP)	130.02 kg
Fuel Cell Onboard Storage	136.09 kg
Transmission System	44.17 kg
Chassis (w/o battery)	368.46 kg
Traction Motor	66.25 kg
Electronic Controller	57.76 kg
Body 3	767.13 kg

¹ Battery size: peak battery power (high power applications). ² Fluids: brake fluid, transmission fluid, powertrain coolant, windshield fluid, adhesives. ³ Body: including BIW, interior, exterior, and glass.

. The vehicle utilizes a 102 kW proton exchange membrane fuel cells (PEMFCs) and a 38 kW power battery as power outputs. In this context, lead-acid batteries supply power to the vehicle’s low-voltage modules, such as onboard electronics, safety systems, and control units, while lithium-ion (Li-Ion) batteries serve as the high-voltage traction battery, working in tandem with the PEMFCs system to propel the vehicle. Since lead-acid batteries are standard equipment and their capacity settings are rarely debated, this study will directly utilize the default lead–acid battery data from the GREET database for calculations. Throughout the vehicle’s lifespan, the fluids used include brake fluid, transmission fluid, powertrain coolant, windshield fluid, and adhesives. The body includes BIW (body-in-white), interior, exterior, and glass.

To complete the full industry chain assessment of the vehicle cycle, this paper first delineates the system boundaries of the vehicle cycle, as shown in Figure 2c. The vehicle cycle is subdivided into the following stages: raw material acquisition, processing and manufacturing, vehicle production, vehicle operation, and disposal and recycling. Additionally, life cycle accounting based on material segmentation requires an inventory of the materials needed for the production of components in FCEVs, as shown in

Table 5. The proportion of materials used in the production of each component of FCEVs.

Material	Fraction (%)	Material	Fraction (%)
Body		Traction Motor
Steel	65.2777	Steel	50.48
Wrought Aluminum	3.0699	Stainless Steel	1.2
Cast Aluminum	0.1646	Cast Aluminum	32.2
Copper/Brass	1.8986	Copper/Brass	10.6
Zinc	0	Nd(Dy)FeB magnet	2.81
Magnesium	0	Phenolic resin	0.05
Glass Fiber-Reinforced Plastic	0.8491	Enamel	0.47
Glass	4.4419	Nickel	0.03
Carbon Fiber-Reinforced Plastic	0	PET	0.32
Average Plastic	21.6747	PBT	0.29
Rubber	1.7678	Mica	0.05
Others	0.8556	Fiberglass	0.01
		Silicone	0.07
Powertrain System (including BOP)		Epoxy resin	0.86
Steel	18.7109	Nylon	0.01
Stainless Steel	31.3479	Methacrylate ester resin	0.13
Cast iron	0.0764	Paint/Varnish	0.39
Wrought Aluminum	16.7949	Zinc	0.02
Cast Aluminum	0	Others	0.01
Copper/Brass	1.7275	Electronic Controller
Magnesium	0	Steel	2.75
Glass Fiber-Reinforced Plastic	2.5917	Cast Aluminum	53.27
Average Plastic	16.6177	Copper/Brass	26.46
Rubber	6.4616	Rubber	1.06
Carbon Fiber-Reinforced Plastic	0	Average Plastic	0.96
PFSA	0.6198	Alumina	0.26
Carbon Paper	2.1401	Epoxy resin	0.26
PTFE	2.5651	Fiberglass	0.67
Carbon & PFSA Suspension	0	Gold	0
Platinum	0.0177	Nickel	0.16
Carbon	0.2728	Nylon	0.07
Nickel	0.0019	PET	3.02
Silicon	0.0540	Polypropylene (PP)	4.33
Others	0	Polyurethane	2.15
Transmission System/Gearbox		Zinc	1.11
Steel	60.5	Zinc oxide	0.02
Copper	18.9	Others	3.45
Cast Iron	0	Fuel Cell Onboard Storage
Magnesium	0	Steel	9.1529
Wrought Aluminum	20	Stainless Steel	8.0818
Cast Aluminum	0	Carbon Fiber-Reinforced Plastic	65.6280
Carbon Fiber-Reinforced Plastic	0	Glass Fiber-Reinforced Plastic	4.4791
Average Plastic	0.2	Wrought Aluminum	0
Rubber	0	Copper	0
Others	0.4	Average Plastics	7.7897
Chassis (w/o battery)		Rubber	0
Steel	66.8037	Nickel	0
Cast Iron	0	Silicon	0.9737
Wrought Aluminum	1.5068	Others	3.8948
Cast Aluminum	18.5998
Copper/Brass	1.3234
Zinc	0
Magnesium	0.1005
Glass Fiber-Reinforced Plastic	0.1546
Average Plastic	2.3483
Rubber	9.1540
Others	0.0087

. In the paper, the primary raw materials associated with each key component are represented as weight percentages of this component mass. This material composition profile serves as the foundational input for subsequent LCA calculations. In particular, the upstream energy inputs (e.g., coal, natural gas, petroleum) and raw materials (e.g., steel, glass, plastics) involved in the lifecycle of FCEVs are excluded from detailed analysis. Their associated energy consumption and emissions are therefore derived from default production processes aligned with industry-average datasets, without scenario-specific adjustments. Based on data from the GREET database, the full life cycle emissions of the entire vehicle cycle can be calculated.

3. Results and Discussion

The entire life cycle of FCEVs is divided into the fuel cycle and the vehicle cycle for discussion, the life cycle process model for FCEVs is represented in Figure 3. Relative contributions are analyzed for each key stage within these two cycles to evaluate the main factors affecting energy consumption and emissions.

3.1. Fuel Cycle

Based on the data provided by the GREET model, each hydrogen energy acquisition path is calculated and accumulated according to its corresponding production, storage, and transportation data. The calculation process is shown in (3).

$$E{C}_{path,n}=m_{H_2}·(E{C}_{P,i}+E{C}_{S,j}+E{C}_{T,k}·d)$$

(3)

where $E{C}_{path,n}$ represents the total energy consumed by path n, $m_{H_2}$ represents the mass of hydrogen in the functional unit, which is 1 kg in this paper, $E{C}_{P,i}$ represents the energy consumption per unit mass of the i-th hydrogen production method, $E{C}_{S,j}$ represents the energy consumption per unit mass of the j-th storage method, $E{C}_{T,k}$ represents the energy consumption per unit mass per unit distance of the k-th transportation method, and d represents the transportation distance in the functional unit, which is 1000 km in this paper. It should be noted that in the subsequent calculation of emissions from the hydrogen energy acquisition path, the calculation method is the same as Equation (3), so it will not be repeated here.

The energy consumption of the hydrogen fuel cycle is evaluated using three indicators: total energy consumption, fossil energy consumption, and non-fossil energy consumption. The comparative results are shown in Figure 4. In the figure, green represents non-fossil energy consumption, orange represents fossil energy consumption, and their sum constitutes the total energy consumption.

As shown in Figure 4, by comparing the methods of direct hydrogen production from fossil energy (path 1 to path 10), it is evident that both total energy consumption and fossil energy consumption are the lowest for path 6. This indicates that among the fossil energy hydrogen production processes, hydrogen production using natural gas has the lowest energy consumption. It is also apparent that while CCS technology can effectively reduce carbon emissions, its energy consumption is not negligible when compared to paths without CCS. According to the IEA’s report, approximately 50% of global hydrogen production comes from natural gas, 19% from coal, and the remaining 4% from water electrolysis and other methods. However, in China, coal is the main source of hydrogen production [37], about 60%. Therefore, in terms of hydrogen production from fossil energy, China should strive to reduce coal-based hydrogen production and increase the proportion of hydrogen production from natural gas. By comparing the methods of hydrogen production from clean energy and water (path 11 to path 18) as shown in Figure 4, it is evident that path 12 has the lowest total energy consumption. This indicates that among the clean energy sources, hydrogen production through nuclear thermochemical water splitting is the most energy-efficient. Comparing path 13 to path 16, the energy consumption differences between hydrogen production using solar or wind power and hydropower are minimal, with hydropower having lower fossil energy consumption. However, path 17 and path 18 show that using the world average electricity mix for hydrogen production results in the highest energy consumption and fossil energy consumption among all paths. This is because coal-fired power accounts for the majority of the world average electricity mix, approximately 36%, followed by natural gas at about 23%, hydropower at around 16%, nuclear power at about 10%, wind power at approximately 7.6%, solar power at around 4.5%, and other sources such as biomass and geothermal accounting for about 4.5% [38]. Therefore, fossil fuels still dominate power generation. Additionally, due to current technological limitations, the efficiency of water electrolysis for hydrogen production is not high. Consequently, paths 17 and 18 have the highest energy consumption. This suggests that water electrolysis for hydrogen production is only reasonable when paired with appropriate power generation methods; otherwise, its energy consumption is even higher than that of traditional fossil fuel-based hydrogen production. Furthermore, comparing the two different transportation methods, pipeline and tube-trailer, it is found that pipeline has lower energy consumption. In summary, path 17 has the highest total energy consumption and fossil energy consumption, path 12 has the lowest total energy consumption, and path 16 has the lowest fossil energy consumption. The total energy consumption for all paths, ranked from highest to lowest, is as follows: 17 > 18 > 3 > 1 > 13 > 7 > 15 > 14 > 4 > 2 > 8 > 16 > 9 > 5 > 10 > 6 > 11 > 12.

Next, taking paths 6, 12, and 17 as examples to explain the relative contributions of hydrogen fuel production, transportation, and compression (storage), as shown in Figure 5. In this figure, red represents the energy consumption for hydrogen production, green represents the energy consumption for transportation, and blue represents the energy consumption for compression. It is evident that in paths 6 and 12, the proportion of energy consumption for transportation is almost negligible, indicating that pipeline transportation has the smallest relative contribution in the hydrogen fuel cycle. In contrast, energy consumption significantly increases when using tube trailers compared to pipelines. Given the substantial differences in hydrogen resource distribution between the eastern and western regions of China, transportation distances are considerable. Therefore, similar to the application of natural gas in China, pipeline transportation is the most promising method for hydrogen transportation. Moreover, the figure shows that production has the highest relative contribution in the fuel cycle. Therefore, within the fuel cycle, the greatest potential for energy conservation lies in the selection and optimization of hydrogen production methods.

In the second section, this study defines the evaluation criteria for emissions, which are GWP, HTP, POCP, AFP, and AP. Next, in this subsection, the emission levels of the hydrogen fuel cycle will be assessed based on these five environmental evaluation indicators, and the pollution contributions of the key stages will be analyzed.

As shown in Figure 6, the GWP of the hydrogen fuel cycle for 18 pathways is illustrated, along with the contributions of three stages of the hydrogen fuel cycle to GWP. From the figure, it can be observed that path 16 has the lowest GWP. It is important to note that in paths 15 and 16, hydrogen is produced using hydropower, and the hydrogen produced at the central plant (the hydropower source) undergoes an initial compression. This compression consumes electricity based on hydropower, which does not generate emissions. Emissions occur only during the second compression at the refueling station. Therefore, the emissions from compression in paths 15 and 16 are 0.5788 kg · CO₂-eq/(kg · H₂), which is significantly lower compared to 1.4661 kg · CO₂-eq/(kg · H₂) in other pathways. At the same time, the GWP emissions for paths 17 and 18 remain the highest, even significantly higher than those for pathways using fossil fuels. This indicates that water electrolysis, if not paired with clean energy sources for electricity generation, has notably poor practical value when conducting a full life cycle assessment.

When evaluating other environmental indicators, AP, AFP, POCP, and HTP are displayed in a unified stacked bar chart, as shown in Figure 7. It can be observed that AP and HTP have relatively higher contributions in the hydrogen fuel cycle. This is because, aside from CO₂, ${\mathrm{S}\mathrm{O}}_x$ is a major emission, leading to higher values for the two indicators dependent on ${\mathrm{S}\mathrm{O}}_x$. The GWP evaluation uses CO₂-eq as the basis, so the unit for GWP is kg, while the evaluation units for the other four indicators are grams. However, this does not mean that GWP has the greatest environmental impact. Therefore, when assessing the combined environmental impact of these five indicators, standardization is necessary. However, in different application scenarios, the importance of each environmental indicator varies. Therefore, weight W is assigned to each indicator, as shown in (4).

$$CEA={\displaystyle \frac{{\displaystyle \sum _1^n}{W}_i·Mapminmax\left(Indicato{r}_{value,i}\right)}{n}}$$

(4)

where $CEA$ represents the comprehensive emission assessment, $W_i$ is the weight of the i-th environmental indicator, and $Mapminmax$ is the normalization function that uses the maximum and minimum values of the array for standardization. $Indicato{r}_{value,i}$ is the array of the i-th environmental indicator. The weight $W_i$ can be set according to the focus of the environmental assessment. The weights $W_i$ in this equation are set to 1. The comprehensive environmental assessment is calculated as shown in

Table 6. Energy consumption and emissions of different hydrogen energy acquisition pathways.

Path	EC (MJ)	GWP (kg)	HTP (g)	POCP (g)	AFP (g)	AP (g)	CEA	LCA Result
1	261	22.14	3.17	0.14	2.35	7.94	0.3	1.95
2	233	20.05	3.02	0.061	2.14	7.69	0.25	2.18
3	265	6.47	3.28	0.15	2.39	8.24	0.22	2.1
4	237	4.38	3.13	0.065	2.18	7.98	0.16	2.39
5	202	12.96	2.08	0.25	0.92	5.06	0.23	2.39
6	174	10.87	1.93	0.17	0.71	4.81	0.18	2.74
7	260	6.96	2.79	0.34	1.32	6.8	0.26	2.04
8	232	4.88	2.64	0.26	1.11	6.54	0.2	2.31
9	207	13.27	2.67	0.31	1.51	6.48	0.28	2.23
10	179	11.18	2.52	0.23	1.3	6.23	0.23	2.53
11	174	3.76	0.96	0.12	0.5	2.31	0.09	3.13
12	147	1.67	0.81	0.04	0.29	2.06	0.04	3.74
13	261	3.56	0.88	0.11	0.46	2.12	0.08	2.49
14	234	1.47	0.73	0.03	0.25	1.86	0.03	2.86
15	254	2.67	0.44	0.09	0.31	0.99	0.06	2.6
16	227	0.58	0.29	0.006	0.098	0.74	0	3.05
17	498	34.11	22.06	0.49	6.88	55.58	1	0.91
18	470	32.02	21.92	0.37	6.66	55.33	0.93	0.94

. From the Table 6, it can be observed that the highest total emissions (CEA) occur in path 17, while the lowest occur in path 16. The order from highest to lowest is as follows: 17 > 18 > 1 > 9 > 7 > 2 > 10 > 5 > 8 > 3 > 6 > 4 > 11 > 13 > 15 > 12 > 14 > 16.

In a full life cycle assessment, it is clear that a comprehensive evaluation of both emissions and energy consumption is necessary. For instance, in energy consumption statistics, path 12 has the lowest total energy consumption, while in emission statistics, path 16 has the lowest total emissions. Therefore, determining how to calculate the LCA for the hydrogen cycle is a critical issue.

As shown in (5), after normalizing the energy consumption, it is multiplied by its respective weight and added to the normalized CEA, yielding the LCA result for the hydrogen fuel cycle. In this paper, the weights $W_1$ and $W_2$ are both 0.5.

$$\begin{array}{c}\hfill LCA={\displaystyle \frac{1}{W_1·Mapminmax\left(Energyconsumption\right)+W_2·CEA+0.1}}\end{array}$$

(5)

The LCA is calculated in Table 6. As shown in Figure 8, clearly, the higher the $LCA$ value, the better the evaluation of the path, and the lower the $LCA$ value, the worse the evaluation. As shown in Figure 8, path 12 has the best evaluation in the hydrogen fuel cycle, with the lowest energy consumption and emissions. Compared to other paths, it performs significantly better. Therefore, from the perspective of life cycle assessment, using nuclear thermochemical water splitting for hydrogen production is the optimal solution. The worst evaluation is for path 17, which has the highest energy consumption and emissions. As illustrated in Figure 4, hydrogen production via renewable energy pathways (e.g., P11–P16) exhibits no significant advantage in energy consumption compared to conventional fossil fuel-based pathways (e.g., coal gasification, natural gas SMR). However, Figure 6 and Figure 7 demonstrate that the emissions of these renewable pathways (P11–P16) are drastically lower than those of fossil-based routes. This divergence explains why renewable pathways achieve superior lifecycle environmental performance despite comparable energy demands. In particular, although water electrolysis itself does not generate high energy consumption and emissions, using grid power sources makes this hydrogen production path non-competitive. A comparative analysis of pathways 1 vs. 3 and 5 vs. 7 reveals that CCS demonstrates a positive impact on coal-based hydrogen production but has the opposite effect on natural gas SMR. As evidenced by Figure 4, Figure 6 and Figure 7, integrating CCS into coal gasification significantly reduces the GWP with only marginal increases in energy consumption, while other emissions remain stable. In contrast, applying CCS to natural gas-based SMR not only incurs higher energy penalties but also amplifies non-GHG emissions, despite achieving comparable GWP reductions. These findings suggest that CCS is strongly recommended for coal-based hydrogen production, whereas its adoption in natural gas SMR requires careful trade-off analyses contingent on regional priorities.

Beyond evaluating the lifecycle energy consumption and emissions of different hydrogen production methods, this study synthesizes prior research to present the technological maturity and application status of existing hydrogen production pathways. Zheng et al. [39], using Guangdong Province, China as a case study, assessed the technological maturity of current hydrogen production processes. They reported maturity levels of 7–9 (the highest is 9 and the lowest is 1) for fossil-based routes (e.g., industrial by-product hydrogen from coke oven gas and coal-to-hydrogen), while green hydrogen pathways represented by wind, hydro, and PV-powered electrolysis achieved maturity levels of 8–9. Under enhanced policy frameworks, industrial by-product hydrogen production is projected to peak by 2030, with coal-based and hydro-based hydrogen production anticipated to decline post-2025. Concurrently, wind- and solar-powered hydrogen production is expected to expand substantially through policy incentives, industrial stimulation, and technological advancements. Safari et al. [40] identified SMR as the most energy-efficient and cost-competitive hydrogen production method currently, despite its high global warming potential (GWP). In contrast, nuclear-powered hydrogen production, while cleaner, more sustainable, and possessing greater scalability potential, remains uncompetitive against SMR on cost grounds.

When integrated with present findings, fossil-based hydrogen production is confirmed as the prevailing approach. Although associated with significant emissions, low cost is maintained as its core competitive advantage. Under projected policy support and technological progression, renewable hydrogen production methods—distinguished by greater emission reduction potential and sustainability—are positioned for extensive long-term development.

3.2. Vehicle Cycle

In the vehicle cycle, this study first analyzes the energy consumption and emissions of key vehicle components based on Table 4 and Table 5. The relative contributions of various parts of FCEVs to energy consumption and environmental indicators are calculated and shown in Figure 9. From the figure, it is evident that in the production of FCEVs, the “Body”, “PEMFCs System”, “Chassis”, and “Hydrogen Storage System” contribute the most to energy consumption and emissions. Therefore, if lightweight design is needed for FCEVs, focusing on these four components will have the most significant impact on energy saving and emission reduction.

Next, the energy consumption and emissions from assembly, disposal and recycling (ADR), battery, and fluid production of a hydrogen fuel cell vehicle over its lifetime are calculated, as shown in

Table 7. Energy consumption and emissions of vehicle cycle over a lifetime.

Indicator	Unit	Component	ADR	Battery	Fluid	Total
EC	MJ	112,501.04	15,210.74	3627.57	2979.11	134,318.5
GWP	kg	6846.22	949.5	233.86	97.49	8127.06
HTP	g	11,441.49	332.38	511.79	147.23	12,432.9
POCP	g	297.76	37.36	11.41	6.18	352.7
AFP	g	3631.72	225.65	142.62	29.17	4029.17
AP	g	33,536.41	1424.25	1471.5	458.15	36,890.31
CEA	–	0.88	0.06	0.02	0	1
LCA Result	–	1.04	5.56	9.09	10	0.91

.

3.3. LCA Result

Based on data provided by GREET, “Passenger Cars 1: FCV (Conventional Material)” has an average energy consumption of 1195.60 kJ/km during operation, which translates to approximately 9.97 g of hydrogen per kilometer (based on the lower heating value of H2, as fuel cells do not recover the latent heat of water vapor). In this study, the total lifetime mileage of an FCEV is 278,662.7 km, meaning that over its lifetime, an FCEV consumes approximately 2778.44 kg of hydrogen. Since no additional emissions are generated during vehicle operation, only energy consumption is considered.

The energy consumption and emissions of the fuel cycle for an FCEV over its lifetime result from hydrogen production, storage, and transportation. During operation, the vehicle itself only consumes hydrogen and does not produce other emissions. Therefore, the energy consumption and emissions of hydrogen production from different pathways are added to the energy consumption and emissions of the vehicle cycle to obtain the total energy consumption and emissions of FCEVs for 18 hydrogen production pathways. Based on Equation (5), the LCA result of FCEVs over its lifetime is calculated. Furthermore, the environmental impacts of conventional ICEVs are calculated based on the default energy consumption and emission profiles from the GREET database. These results serve as a baseline for comparative analysis against FCEVs under different hydrogen production and transportation pathways, as shown in

Table 8. LCA result of FCEV over a lifetime.

Category	EC (MJ)	GWP (kg)	HTP (g)	POCP (g)	AFP (g)	AP (g)	CEA	LCA Result
FCEV1	858,490.3	69,623.69	21,251.7	742.68	10,564.48	58,970.15	0.33	2.33
FCEV2	781,717	63,844.51	19,825.23	522.14	9978.04	58,244.3	0.27	2.78
FCEV3	870,637.1	26,099.14	20,548.05	769.47	10,670.6	59,798.06	0.24	2.56
FCEV4	792,830.7	20,295.21	19,831.37	533.3	10,087.01	59,071.5	0.18	3.13
FCEV5	695,752.3	44,139.88	17,167.09	1047.31	6587.33	50,963.7	0.26	3.23
FCEV6	617,771	38,328.32	16,054.25	825.03	6001.86	50,260.95	0.2	4.17
FCEV7	856,712.9	27,453.19	18,991.29	1297.37	7693.89	55,776.18	0.28	2.5
FCEV8	778,938.9	21,681.32	19,771	1075.09	7112.22	55,057.04	0.23	2.94
FCEV9	608,489.5	44,984.76	19,768.31	1213.62	8227.26	54,891.96	0.31	3.45
FCEV10	631,827.2	39,175.75	19,438.42	990.73	7645.24	54,293.79	0.26	3.57
FCEV11	617,772	18,580.76	3900.26	686.11	5418.39	43,312.45	0.09	5.56
FCEV12	542,753.1	12,775.42	3496.44	463.84	4834.94	42,611.73	0.04	8.33
FCEV13	858,490.3	18,009.19	3694.9	658.33	5307.24	44,283.33	0.08	3.33
FCEV14	784,495	12,208.82	3279.23	436.05	4725.78	43,558.3	0.03	4.17
FCEV15	840,034.2	15,553.26	2465.41	602.76	4890.48	39,640.95	0.06	3.57
FCEV16	764,884.3	9747.2	2057.64	369.37	4301.46	38,945.35	0	4.76
FCEV17	151,7970	102,919.4	73,705.68	1714.14	231,35.31	191,165.6	1	0.91
FCEV18	1,435,185	97,058.02	73,374.18	1379.72	22,545.88	190,548.2	0.93	0.98
ICEV	1,095,552	79,631.04	22,778.57	1766.477	8180.757	71,262.75	0.49	1.59

. As shown in Figure 10, the EC, GWP, HTP, POCP, AFP, AP, and LCA Result are represented in a radar chart to clearly demonstrate the comprehensive environmental assessment across different pathways. To emphasize pathways with lower environmental impacts, the radar chart metrics—EC, GWP, HTP, POCP, and AP—are represented using reciprocal values (1/x). This transformation amplifies the visual prominence of smaller environmental impact values, making sustainable pathways more distinguishable. Since the energy consumption and emissions from the vehicle cycle are held constant across all scenarios, the LCA rankings of each pathway are determined solely by the environmental performance of the hydrogen fuel cycle. From worst to best, the pathways are categorized as follows: 17 < 18 < ICEV < 1 < 7 < 3 < 2 < 8 < 4 < 5 < 13 < 9 < 15 < 10 < 6 < 14 < 16 < 11 < 12. In Figure 10, the LCA results for ICEVs are represented by thick gray-black lines. Notably, most FCEV pathways exhibit superior lifecycle performance compared to ICEVs, with the exception of FCEV_P17 and FCEV_P18. This supports the conclusion that FCEVs generally achieve better energy conservation and emission reduction performance than ICEVs from a full lifecycle perspective. However, when hydrogen is sourced from grid-powered electrolysis (e.g., FCEV_P17/P18), its environmental performance is significantly worse than that of ICEVs. These findings underscore the necessity of comprehensive industrial chain assessments in evaluating new energy vehicles. Analyses limited to the operational phase may yield misleading conclusions, as they ignore upstream/downstream emission hotspots.

Additionally, let us take the representative example of hydrogen production from natural gas, specifically path 5, to explore the proportion of energy consumption and emissions from the fuel cycle and vehicle cycle in a full supply chain LCA. The results are illustrated in Figure 11. It is evident that, for both energy consumption and emissions, the fuel cycle accounts for the majority. The analysis reveals significant disparities in the contributions of the fuel cycle and vehicle cycle across environmental impact categories. While the fuel cycle dominates most metrics—accounting for 66–82% of total impacts in EC, GWP and POCP—the vehicle cycle exhibits pronounced influence on HTP, AFP and AP, contributing 61–72% to these categories. It is critical to recognize that the vehicle cycle’s relatively minor contribution to global warming potential (GWP, 18%) does not inherently reflect a low overall environmental burden. Significant adverse impacts persist in other key environmental categories, such as respirable particulate matter emissions (e.g., vehicle painting processes), human toxicity (e.g., battery production and disposal), and acidification potential (e.g., sulfur oxides generated during steel smelting and component manufacturing). As clearly demonstrated in Figure 11, a full industrial chain lifecycle assessment of FCEVs is indispensable. Within the FCEVs lifecycle, the fuel cycle and vehicle cycle make distinct and often disproportionate contributions to different environmental impact categories. For instance, the fuel cycle dominates GWP (82% contribution) due to emissions from hydrogen production, while the vehicle cycle significantly influences AP (72%) through energy-intensive material processing. Consequently, isolated assessments of individual supply chain segments or single environmental metrics yield incomplete or misleading conclusions, as they fail to capture systemic trade-offs between carbon footprint, resource toxicity, and ecosystem impacts. This finding underscores the necessity of adopting multi-criteria, full industrial chain-based frameworks to holistically evaluate and optimize the sustainability of FCEVs.

4. Sensitivity Analysis

The Sobol sensitivity analysis method, originally proposed by Sobol et al. [41], was employed to quantify the influence of key parameters on the LCA outcomes of FCEVs. Utilizing the SALib Python library, this method calculates first-order (main effect) and total-order (main effect and interaction effects) Sobol indices [42], which decompose the variance of LCA outputs into contributions from individual inputs and their interactions. In this study, the key input parameters and outputs are summarized in

Table 9. The parameters and indicators of sensitivity analysis.

Parameters			Indicators
Hydrogen Production [0, 8]	Transport Method [0, 1]	Vehicle Lifetime [100,000, 500,000]	EC	GWP	HTP	POCP	AFP	AP

. The input parameter “Hydrogen Production” represents nine hydrogen production methods and is defined as an integer-type categorical variable with a range of [0, 8]. The “Transport Method” parameter indicates hydrogen transportation methods, also a categorical variable with a range of [0, 1]. The “Vehicle Lifetime” parameter refers to the operational lifespan of FCEVs, spanning from 100,000 km to 500,000 km. The output variables include six environmental impact categories: EC, GWP, HTP, POCP, AFP, and AP. These metrics collectively evaluate the lifecycle environmental performance of FCEVs pathways under varying hydrogen production, transportation, and vehicle durability scenarios.

The sensitivity analysis results are presented in

Table 10. The result of Sobol sensitivity analysis.

Indicators	Parameters	S1	S1 Conf	ST	ST Conf
EC	Hydrogen Production	0.4201	0.0242	0.4826	0.0208
	Transport Method	0.0112	0.0034	0.0123	0.0005
	Vehicle Lifetime	0.5054	0.0215	0.5694	0.0261
GWP	Hydrogen Production	0.7517	0.0283	0.8627	0.0309
	Transport Method	0.0076	0.0026	0.0088	0.0004
	Vehicle Lifetime	0.1263	0.0152	0.2402	0.0125
HTP	Hydrogen Production	0.8218	0.0464	0.9452	0.0368
	Transport Method	0.0001	0.0003	0.0001	0
	Vehicle Lifetime	0.0545	0.0125	0.1768	0.0152
POCP	Hydrogen Production	0.7213	0.0325	0.8292	0.0295
	Transport Method	0.0146	0.0038	0.0161	0.0008
	Vehicle Lifetime	0.1562	0.016	0.2651	0.0174
AFP	Hydrogen Production	0.7816	0.0424	0.8984	0.0351
	Transport Method	0.0025	0.0018	0.0027	0.0002
	Vehicle Lifetime	0.0984	0.0135	0.2153	0.0146
AP	Hydrogen Production	0.808	0.043	0.9293	0.0372
	Transport Method	0.0024	0.0015	0.0026	0.0001
	Vehicle Lifetime	0.0683	0.0143	0.1886	0.015

, where S1 (first-order Sobol indices) quantify the individual contribution of each input parameter to the output variance, and ST (total-order Sobol indices) represent the total influence, including both individual and interactive effects. The confidence intervals (conf) for these indices are also provided. Figure 12 visualizes these findings, revealing that hydrogen production is the dominant driver across all environmental impact categories except EC. Specifically, it accounts for more than 70% of the variance in GWP, HTP, POCP, AFP, and AP (ST = 0.83–0.95), while vehicle lifetime dominates EC (ST = 0.5694). Hydrogen Production exhibits strong interaction effects (ST - S1 > 0.1) in all impact categories, peaking at 0.1234 for HTP. Vehicle lifetime shows notable interactions in GWP and AP (ST - S1 > 0.11). In contrast, transport method has negligible influence (ST < 0.02 across all metrics, with confidence intervals near zero). For EC, both vehicle lifetime (ST = 0.5694) and hydrogen production (ST = 0.4826) are significant, reflecting the inherent trade-off between prolonged vehicle operation and cumulative fuel demand. However, excluding EC, hydrogen production overwhelmingly dictates environmental performance, followed distantly by Vehicle lifetime, while transport method remains irrelevant (<2% contribution). The lifecycle optimization priority for FCEVs follows the hierarchy hydrogen production > vehicle lifetime > transport method.

5. Conclusions

This study aims to analyze the energy consumption and emissions of FCEVs over their lifetime based on the entire supply chain. The research is mainly divided into two parts: fuel cycle and vehicle cycle.

In the fuel cycle, this study analyzed a total of 18 hydrogen fuel cycle pathways, with the best LCA evaluation being pathway 12, which combines nuclear thermochemical water splitting with pipeline trailer transport. Although proton exchange membrane (PEM) water electrolysis has gained attention in recent years due to reduced fossil fuel usage, its overall energy consumption shows no significant advantage compared to traditional fossil fuel-based hydrogen production. It is noteworthy that despite its high energy consumption, PEM water electrolysis demonstrates significant emission reduction benefits. However, path 17 has the highest emissions and the worst LCA evaluation, further confirming that, from a life cycle assessment perspective, water electrolysis for hydrogen production is not feasible unless paired with clean energy.

In the vehicle cycle study, this paper analyzed the energy consumption and emissions in the production processes of FCEV components and found that the “body”, “powertrain system”, “chassis”, and “hydrogen storage system” contribute the most to the vehicle cycle. Therefore, future research of vehicle cycle should focus on lightweight design improvements for these four parts. Additionally, the study calculated the energy consumption and emissions of the battery, fluids, and vehicle assembly, disposal, and ADR throughout the vehicle’s entire life cycle. The total energy consumption and emissions of the vehicle cycle were then determined, with component production and ADR showing higher energy consumption and emissions. Although the battery’s energy consumption is relatively low, its emissions are comparatively high.

This study conducted a life cycle assessment of fuel cell vehicles based on the full supply chain. Using the natural gas-based hydrogen production pathway (path 5) as an example, the contributions of the fuel cycle and vehicle cycle to the total life cycle results were analyzed. The results show that, for both energy consumption and emissions, the fuel cycle dominates. Although hydrogen is regarded as the “ultimate energy source”, its energy consumption and emissions remain relatively high. Furthermore, in the entire supply chain life cycle assessment, this study found that compared with ICEVs, fuel cell vehicles still demonstrated significant energy consumption and emissions, but direct electrolysis using grid electricity was significantly inferior to ICEVs. This emphasizes the significance of comprehensive life cycle assessment in new energy research. Finally, this paper conducts a sensitivity analysis on hydrogen production methods, transport methods, and the lifetime of FCEVs. The results show that among the LCA of FCEVs, the hydrogen production method is the most influential, followed by the lifespan, and the influence of the transportation method is relatively small compared to these two. In future work, how to design PEMFCs for energy conservation and emission reduction based on the results of life cycle assessment is an important research direction.