Comparative Life Cycle Analysis of Battery Electric Vehicle and Fuel Cell Electric Vehicle for Last-Mile Transportation

Abstract

This study investigates whether Battery Electric Vehicles (BEVs) or Fuel Cell Electric Vehicles (FCEVs) represent the superior alternative to conventional vehicles for last-mile delivery, with a particular focus on large enterprises that prioritize both economic feasibility and environmental performance. Life Cycle Assessment and Life Cycle Cost methodologies are applied to evaluate both technologies across the full cradle-to-grave life cycle within a unified framework. The functional unit is defined as one kilometer traveled by a BEV or FCEV in last-mile transportation, and the system boundary includes vehicle manufacturing, operation, maintenance, and end-of-life treatment. The environmental impacts are assessed using the ReCiPe 2016 Midpoint (H) method implemented in OpenLCA 2.0.4, and normalization follows the standards provided by the official ReCiPe 2016 framework. The East China Power Grid serves as the baseline electricity mix for the operational stage. Regarding GHG emissions, FCEVs demonstrate a 12.36% reduction in carbon dioxide (CO₂) emissions compared to BEVs. This reduction is particularly significant during the operational phase, where FCEVs can lower CO₂ emissions by 53.51% per vehicle relative to BEVs, largely due to hydrogen energy’s higher efficiency and durability. In terms of economic costs, BEVs hold a slight advantage over FCEVs, costing approximately 0.8 RMB/km/car less. However, during the manufacturing phase, FCEVs present greater environmental challenges. It is recommended that companies fully consider which environmental issues they wish to make a greater contribution to when selecting vehicle types. This study provides insight and implications for large companies with financial viability concerns about environmental impact regarding selecting the two types of vehicles for last-mile transportation. The conclusions offer guidance for companies assessing which vehicle technology better aligns with their long-term operational and sustainability priorities. It can also help relevant practitioners and researchers to develop solutions to last-mile transportation from the perspective of different enterprise sizes.

1. Introduction

Global carbon emissions reached a new peak of 37.4 Gt in 2023 [1]. Transport and logistics accounted for 24% of global emissions, with 78% coming from road transport [2]. Much of this has been pinned on the rapid development of business-to-customer (B2C). Compared with self-pickup in the offline market, logistics service has also become an indispensable part of customers evaluating goods. Hence, last-mile transportation has emerged as a significant issue in the logistics sector [3]. As the e-commerce and logistics industries continue to expand, the frequency of last-mile transportation and carbon dioxide (CO₂) emissions will also rise, further exacerbating greenhouse gas (GHG) emissions from road transport. Therefore, decarbonizing last-mile transportation is an urgent priority. The traditional modes of last-mile transportation are mainly vans, light diesel trucks, bicycles, and tricycles. The latter two are suitable for areas with higher population densities and have the advantage of being small and easy to maneuver, making deliveries convenient when roads are narrow and housing areas are compact. However, their disadvantages include the inability to transport a significant amount of goods at once and increased travel time in larger delivery areas. Consequently, the traditional way is to use vans or light diesel trucks for last-mile transportation. Light diesel trucks are more favorable for short-distance deliveries than larger vans. However, the use of diesel trucks for deliveries leads to significant GHG emissions, thus presenting environmental challenges associated with these traditional vehicles [4].

The issue is further complicated as global warming becomes a critical concern, prompting national carbon emissions policies to restrict companies from pursuing low-cost strategies at the expense of sustainability. As a result, companies are compelled to transform their logistics and delivery processes, and they need to look for more sustainable delivery methods to meet this challenge. The emergence of Battery Electrical Vehicle (BEV) and Fuel Cell Electrical Vehicle (FCEV) technologies has become an alternative to Conventional Vehicles (CVs) such as light diesel trucks for last-mile transportation. Both of them do not emit any CO₂ during operation. Recent studies have predicted that FCEV trucks could replace 25% of the fleet by 2050 [5]. Although both BEVs and FCEVs are competitive alternatives to CVs, they will be more expensive than CVs in terms of cost. Although BEVs have a cost advantage over FCEVs, FCEVs will contribute more in terms of environmental benefits. The high cost remains the primary barrier for companies when considering FCEVs. However, larger companies with capacity for long-term growth are more inclined to invest in decarbonization strategies and can accept the slower rate of return that comes with such long-term investment [6]. As a result, major players with long-term development prospects might explore whether choosing FCEVs can overcome the disadvantages of high investment costs in the long run and bring more sustainable economic and environmental benefits to the company. Therefore, the research question is whether BEVs and FCEVs are viable alternatives to CVs, and more critically, which is the superior choice for last-mile delivery, especially for large enterprises.

China is selected as the study region not only due to data availability but also because it plays a central role in the global transition toward low-carbon transportation. China hosts the world’s largest e-commerce and logistics market, which generates substantial last-mile delivery demand [7]. It also leads global electric vehicle production and adoption, making it an important context for evaluating BEV deployment [8]. In addition, China is the world’s largest hydrogen producer and is expanding renewable energy capacity, creating favorable conditions for future FCEV development [9,10]. These characteristics provide a strong analytical basis for selecting China as the focus of this study.

To address the motivation mentioned above, this study sets out the following research objectives. The first objective is to assess and compare the two vehicle technologies in terms of their economic and environmental performance using LCA and LCC across the full cradle-to-grave life cycle. The second objective is to quantify the differences in life cycle greenhouse gas emissions and operational costs between BEVs and FCEVs based on the defined functional unit of one kilometer traveled and the system boundaries established for the LCA and LCC models. The third objective is to examine the distribution of environmental impacts in order to identify the specific impact areas in which each technology performs better or worse. These objectives directly inform the methodological design by determining the need for an integrated LCA–LCC framework, structuring the subsequent modeling of environmental impacts and life cycle costs, thereby ensuring a clear alignment between the study’s motivation, analytical choices, and intended contribution to last-mile logistics. The target audience for this study includes, but is not limited to, governments, policymakers, logistics and transport companies, vehicle manufacturers, and researchers.

There have been some discussions in the existing literature about the recommended vehicles for last-mile transportation. Refs. [11,12,13] believed that cargo bikes could provide a more cost-effective and environmentally friendly last-mile delivery service than conventional lorries, in terms of their fixed costs and energy consumption. In addition, ref. [14] found that using smart bikes instead of traditional vehicles may reduce costs by 87–90 percent. Additionally, some studies believe that drones have advantages in terms of efficiency and environmental benefits for last-mile deliveries. However, as drones are limited by technology costs, operations and maintenance, weather challenges, and other factors, a hybrid truck plus drone transport fit is currently required to be more effective in reducing delivery times and improving environmental benefits [15,16]. Therefore, as the most highly regarded BEVs and FCEVs, academics believe they are the most suitable alternatives to replace conventional vehicles. Reference [5] suggested that finding alternatives to the internal combustion engine is crucial to reducing carbon footprints, and they propose that FCEVs and BEVs will replace CVs in the future. Similarly, ref. [17] evaluated the cost of electric Light Commercial Vehicles (eLCVs) and CVs by using the Total Cost of Ownership (TCO) model. They concluded that eLCVs are more cost-effective than CVs due to the competitive pricing of eLCVs. Reference [18] found through cost-based life cycle assessment that FCEVs have environmental and economic advantages over CVs in long-distance transportation.

Current research in terms of last-mile delivery focuses on equipment innovation and route optimization. References [19,20] claimed that the innovations are mainly in drone transport, robotics, smart parcel stations, crowdsourcing, and automated vehicles. In terms of reducing carbon emissions in the last mile, ref. [21] explored drone transport to address route optimization challenges and ecological issues. Reference [22] used the simulation of predicted vehicle in the heavy congestion operating environment to optimize traffic flow. The existing literature is relatively limited for applying BEVs and FCEVs in last-mile transportation. However, since vehicular transport is still the dominant means of transport in urban distribution, the research on these two emerging vehicle technologies deserves more effort. The characteristic that distinguishes purely electric cars from other vehicles (diesel or hybrid) is that they run entirely on the power stored in the battery to propel the electric motor [23]. Therefore, they do not emit any CO₂ during operation. In addition to environmental motivations, many governments worldwide are promoting BEVs and FCEVs as part of broader industrial and technological strategies aimed at strengthening competitiveness in the global automotive sector. Recent studies show that national policies in both developed and developing economies increasingly link low-carbon mobility transitions with goals related to industrial upgrading, supply chain leadership, and long-term economic positioning in emerging clean-technology markets. Such policies reflect a recognition that the shift toward electrified and hydrogen-powered transport is not only a response to climate and energy challenges but also a strategic opportunity to shape future manufacturing capabilities and innovation trajectories [24,25]. This broader policy-industrial landscape underscores the significance of evaluating BEVs and FCEVs within an evolving technological and economic environment.

Although pure electric vehicles are more environmentally friendly, they also suffer from weak range, long charging time, and low penetration of charging infrastructure [26]. According to [27], electric vehicles are a promising alternative to conventional internal combustion engines for passenger and freight transport. eLCVs can contribute to cleaner urban goods distribution. Electric trucks emit 42–61 percent less greenhouse gases and consume 32–54 percent less energy than diesel trucks [17]. Therefore, eLCVs are generally chosen as alternatives to regular light trucks for last-mile issues.

A highly anticipated alternative is the Hydrogen Fuel Cell Vehicle (HFCV), in which hydrogen fuel and fuel cells replace batteries. It takes compressed hydrogen from an onboard storage tank, mixes it with the atmosphere, and generates direct current electricity to drive an electric motor [28]. FCEVs are environmentally friendly, emitting only water and involving no heat or combustion. Because hydrogen fuel contains no carbon, it also emits no CO₂, carbon monoxide, or hydrocarbons. According to the power type of fuel cell, it can be divided into three types: Compensation, Hybrid, and Full Power. It uses the fuel cell system as the primary power source and is equipped with a power battery as an auxiliary power source, which can play the role of energy recovery [29]. The most common fuel cell type for FCEVs is the Proton Exchange Membrane Fuel Cell (PEMFC), which is particularly suitable for automotive applications due to its fast start-up capability and high-power density, making it particularly suitable for such scenarios where quick response operation is required [30]. However, the critical point for FCEVs regarding environmental protection is how hydrogen is produced, which is still dominated by Steam Methane Reforming (SMR), or gray hydrogen, which accounts for 50 percent of the overall hydrogen [31]. Moreover, this way of making hydrogen still releases CO₂ into the atmosphere. However, due to high cost, the green hydrogen produced by renewables accounts for only a small portion of hydrogen production. Still, it is already being invested heavily and is expected to be able to reduce the price significantly in the next decade [32]. As hydrogen costs are likely to decrease by 43% gradually, fuel cell vehicles are rapidly developing with a hydrogen demand of about 60,000 tons [33]. Therefore, although there are still challenges with FCEVs today, with the gradual advancement of green hydrogen fuel technology, hydrogen plays a role in the energy transition of the transport sector. Recent studies illustrate that the carbon intensity of hydrogen differs substantially depending on the production route and logistics, with evidence showing that supply chain configuration, transportation distance, and production energy mix collectively determine the life cycle impacts of hydrogen used in mobility applications [34,35]. Emerging green hydrogen pathways have been shown to significantly reduce upstream emissions compared with conventional methods, highlighting the importance of considering hydrogen source variability in FCEV assessments [34]. In addition, China, as the world’s largest hydrogen producer, will achieve a hydrogen production of 25 Mt in 2030, a year-on-year increase of 13.6%, accounting for one-third of the global hydrogen production [18]. The decarbonization trajectory of China’s regional electricity grids will also significantly influence the long-term environmental performance of BEVs, as declining grid carbon intensity directly reduces operational emissions in the use phase [36,37]. These findings underscore the need to consider evolving energy systems when comparing life cycle results for BEVs and FCEVs in China.

Research on BEVs and FCEVs in the existing literature has largely focused on daily travel or public transportation. When it comes to commercial applications, studies typically discuss overall transit rather than delving into specifics for each transportation segment. Some scholars believe that BEVs would be a better choice. Compared to CVs, BEVs have some emission reduction benefits and are more mature than FCEVs in terms of construction and development of charging infrastructure [38]. Reference [39] conducted a sensitivity analysis based on a cost-based life cycle assessment and concluded that increasing the mileage and reducing fuel costs are the key improvement points that can increase the benefits of FCEVs and BEVs in China. Reference [40] simulated urban freight transportation and concluded that BEVs have a cost advantage over FCEVs, as the price and unstable source of hydrogen are still factors to be considered. Therefore, because of the relatively low cost of BEVs and their effectiveness in reducing carbon footprint, many believe that BEVs will become mainstream commercial transportation in the future [40]. Recent comparative life cycle assessments provide updated evidence that the relative environmental performance of BEVs and FCEVs varies across vehicle segments, life cycle stages, and regional energy conditions, illustrating that no single technology performs best under all scenarios [41,42,43]. Systematic reviews further emphasize that material production, particularly for batteries, and the treatment of end-of-life processes remain critical contributors to BEV impacts [44]. Despite FCEVs’ advantages over BEVs in terms of carbon emissions and the effectiveness of decarbonization, their high cost has hindered broader consideration. However, with the gradual stabilization and development of hydrogen battery technology, FCEVs are expected to become competitive with BEVs in terms of capital cost, sustainability, and reducing range anxiety [45]. Moreover, ref. [46] found that the industry is still undecided about which of the two types of vehicles is the industry’s future. Different companies will choose different types of vehicles to expand their business territory based on their unique business models. Therefore, BEVs and FCEVs are still attractive for logistics companies to study in the last mile. The literature review reveals that research focuses on the entire transportation industry, there is a lack of specific studies on the relationship between different business sizes and vehicle choices. Reference [47] evaluated the choice between BEVs and FCEVs for the last-mile problem in the Spanish market, mainly through a TCO model. However, the article’s focus on cost raises the question of whether the competitiveness of FCEVs could be enhanced by considering the assessment of long-term environmental benefits. Since FCEVs have higher greenhouse gas emission reduction benefits, as mentioned by [40], it is hypothesized that balancing the carbon reduction benefits with the acquisition costs of FCEVs may be more favorable for companies that can grow in the long term. Reference [6] also analyzed the sustainability management of companies of different sizes. They found that larger companies have more spare resources to use more effective tools in reducing emissions. Incorporating dynamic modeling elements, such as battery aging, technology learning, evolving grid structures, and hydrogen supply chain variability into vehicle LCAs is essential, as these factors can meaningfully shift comparative outcomes between BEVs and FCEVs [34,42,44].

To synthesize the current state of knowledge and to highlight methodological differences across existing assessments,

Table 1. Summary of representative LCA studies comparing BEVs and FCEVs.

Study	Scope & Vehicle Type	Functional Unit	System Boundary	LCIA Method	Key Findings
Syré et al. (2024) [41]	Comparative LCA of BEV and FCEV passenger vehicles	1 km	Cradle-to-grave	ReCiPe 2016 Midpoint	FCEVs show higher toxicity-related impacts; hydrogen source strongly affects results.
Alexander & Abraham (2024) [42]	Meta-review of LCA for electrified vehicles	Various	Cradle-to-gate, cradle-to-grave, and well-to-wheel	Multiple (ReCiPe, ILCD, TRACI)	BEV impacts depend heavily on battery chemistry and electricity mix; harmonization needed.
Oluwalana & Grzesik (2025) [44]	Systematic review of EV LCA	Not Specified	Cradle-to-gate, cradle-to-grave, and well-to-wheel	Various	Highlights inconsistency in LCA practices; emphasizes importance of regionalized LCAs.
Šimaitis et al. (2025) [43]	BEV carbon footprints under decarbonization pathways	1 vehicle km	Cradle-to-grave	Carbon footprinting	BEVs consistently lowest carbon emissions across future grid scenarios.
Vu & Chang (2025) [34]	LCA of green hydrogen supply chain	1 kg hydrogen	Cradle-to-gate	ReCiPe/CML	Green hydrogen reduces GHGs but has upstream burdens from renewable infrastructure.
Eltohamy et al. (2024) [35]	Review of LCA applications in electric mobility	Not Specified	Cradle-to-grave and well-to-wheel	Various	Identifies gaps in current LCA practice, including battery aging and system boundaries.
Zhang et al. (2024) [36]	Carbon intensity of Chinese electricity	1 kWh	Cradle-to-grave	CO2-eq accounting	Regional variation significantly affects BEV LCA outcomes.
Tang et al. (2022) [37]	BEV vs. ICEV LCA in China	1 km	Cradle-to-grave	ReCiPe Midpoint	BEVs reduce GHGs but increase mineral depletion and toxicity impacts.
Winkler et al. (2022) [40]	FCEV vs. diesel/BEV for urban freight	1 km	Well-to-wheel	GHG footprint	FCEVs benefit from renewable hydrogen but remain expensive.
Lombardi et al. (2017) [48]	LCA of multiple powertrain options	1 km	Cradle-to-grave	ILCD	FCEVs have high resource and toxicity impacts; BEVs improve with grid decarbonization.

Source: compiled based on [34,35,36,37,40,41,42,43,44,48].

summarizes representative LCA studies focusing on BEVs and FCEVs, including their scopes, functional units, system boundaries, assessment methods, and key findings.

Although existing studies provide valuable insights into the relative cost and environmental performance of BEVs and FCEVs, they tend to generalize results across the entire transportation sector and do not reflect the distinct operational conditions of companies of different sizes. Current LCA and LCC analyses often examine these vehicle technologies independently or emphasize only one dimension of performance, which limits their usefulness for organizations that must evaluate both long-term environmental outcomes and financial investment strategies. Moreover, most assessments do not employ a unified LCA–LCC framework that compares BEVs and FCEVs across multiple environmental impact categories while simultaneously accounting for lifetime cost implications. As a result, there remains a lack of research that specifically examines how large enterprises should evaluate the tradeoffs between BEVs and FCEVs for last-mile delivery, since they possess greater capacity to adopt emerging technologies.

Building on these insights, this study extends previous comparative analyses of BEVs and FCEVs by introducing a large-company perspective. The study focuses on large-scale companies that aim to sustain long-term operations by achieving both economic and ecological benefits. Given the dual focus of this study on both environmental performance and economic feasibility, a life cycle-based analytical framework is particularly suitable for assessing the trade-offs between BEVs and FCEVs. Previous research has primarily applied the Life Cycle Assessment (LCA) and Life Cycle Cost (LCC) framework to individual users or public transportation cases, without considering how firm heterogeneity influences decision-making. Large enterprises, however, differ substantially from smaller entities in investment horizon, risk tolerance, and sustainability orientation. They are also the first to be targeted by policy reforms and often subject to stricter decarbonization requirements, which motivate them to pursue long-term emission reductions despite higher initial costs. Moreover, to align the analysis with firms’ long-term investment horizons, this study employs a total discounted cost of ownership (TDCO) approach, which is an enhanced form of the traditional TCO model that integrates discount and inflation rates to account for the time value of money [33]. The TDCO approach provides a more realistic assessment of economic feasibility from a long-term perspective, particularly relevant for large firms with greater financial resilience and stronger sustainability commitments. Through this integration, the research not only extends the analytical reach of existing TCO-based studies but also enhances the policy relevance of life cycle analysis by illustrating how firm-scale differences in financial flexibility and carbon accountability can reshape the comparative evaluation of BEVs and FCEVs. This perspective suggests that policies promoting low-carbon logistics should account for firm-scale heterogeneity, prioritizing large enterprises as early adopters while designing differentiated support mechanisms for smaller firms.

The originality of this research lies in integrating LCA and LCC within a unified analytical framework to evaluate BEVs and FCEVs simultaneously from both environmental and economic perspectives. In contrast to previous studies that examine these technologies in general transportation contexts, this study focuses specifically on large enterprises engaged in last-mile delivery, a segment where long-term operational strategy and sustainability investment capacity play a crucial role. Furthermore, by assessing multiple environmental impact categories using the ReCiPe 2016 Midpoint (H) method, the study provides a more comprehensive understanding of the tradeoffs between BEVs and FCEVs.

To address these gaps, the present study analyzes BEVs and FCEVs using an integrated LCA–LCC framework tailored to the operational conditions of large enterprises engaged in last-mile delivery. This study applies Life Cycle Assessment (LCA) and Life Cycle Cost (LCC) to determine whether BEVs or FCEVs are better options for last-mile transportation for large companies. The study focuses on large-scale companies that seek to maintain long-term operations through sustainable economic and ecological benefits. The rest of this article is structured as follows: Section 2 presents the methodology used to evaluate BEVs and FCEVs in terms of economic and environmental aspects; Section 3 gives results and preliminary insights on three subsections on BEVs, FCEVs and a comprehensive comparison of both vehicles; Section 4 is an in-depth analysis of the results. It also lists the study’s contributions; Section 5 summarizes this study and indicates the limitations and implications for future research directions.

2. Materials and Methods

2.1. Life Cycle Cost (LCC)

LCC assessments typically evaluate direct, indirect, internal, and external costs and revenues generated throughout a defined life cycle [49]. The TCO minimization method should be applied to determine the most cost-effective solution for LCC. TCO refers to the total costs incurred over the entire life cycle of owning or using an asset. This method compares and evaluates different product choices [50]. References [17,40] studies, used a TCO model-based approach to compare different vehicle types. However, according to the aim of this study, it combined the total discounted cost of ownership (TDCO) model, which is an optimized TCO model proposed by [47], and the parameters in [27] as shown in Equations (1) and (2). This methodology incorporates the time value of money by factoring in the discount and inflation rates to minimize TCO. Therefore, it can satisfy part of the research objectives and answer the research question, considering economic feasibility from a long-term perspective.

LCC_ij = TDCO_ij/(n_i × AAM_i)

(1)

where:

• LCC_ij denotes the LCC in the i-th vehicle model;
• n_i is the lifetime of the i-th vehicle model;
• TDCO_ij denotes the Total discounted costs of ownership of the i-th vehicle model in the j-th city;
• AAM_i is the average annual mileage of the i-th vehicle model.

TDCO_ij = MGP_i − GS_ij + LF + PSI_i + PT_i + ∑k − 1n [(MFV_ijk + MFSI_ijk + EUF_ijk)/(1 + r)^{(k − 1)}] − [RCi/(1 + r)^{(ni − 1)}]

(2)

where:

• MGP_i denotes the manufacturer guidance price of the i-th vehicle model;
• GS_ij denotes the government subsidy of the i-th vehicle model in the j-th city;
• LF denotes the license fee of a vehicle;
• PSI_i denotes the price of supply infrastructure of the i-th vehicle model;
• PT_i denotes the purchase tax of the i-th vehicle model;
• MFV_ijk is the maintenance fee of the i-th vehicle model in the j-th city in the k-th year;
• MFSI_ijk is the maintenance fee of supply infrastructure for the i-th vehicle model in the j-th city in the k-th year;
• EUF_ijk is the energy use fee of the i-th vehicle model in the j-th city in the k-th year,
• r is the discount rate of economic cost;
• RC_i is the residual cost of the i-th vehicle model.

Detailed data of LCC modeling information are presented in

Table A1. The required data of LCC.

Parameters	NISSAN e-NV200	Vivaro-e HYDROGEN
i	BEV	FCEV
j	Shanghai
ni	10
AAMi	15,000
MGPi	247,676.52 RMB (€31,995)	566,644.54 RMB (€74,595)
GSij	0
LF	0
PSIi	60,000	0
PTi	0
MFVijk	3270	3420
MFSIijk	11,500	0
EUFijk	7200	9750
r	7%
RCi	104,130.56	246,143.846

Sources: Authors compiled based on [13,36,37,38,39].

. The vehicles used in this project are the BEV Nissan e-NV200 and the FCEV Vivaro-e HYDROGEN, respectively, with guide prices sourced from [17,51], which already account for the conversion from euros to yuan. The overall scenario is based on Shanghai, so it is informed by the latest subsidy policy for new energy vehicles (including BEVs and FCEVs) released by [52]. It is known that the Shanghai government currently does not provide financial subsidies directly to enterprises, but rather converts them into a number of free licenses. Due to the support of the policy, they also do not need to pay the purchase tax. The maintenance fee for both vehicles and the depreciation rates for residual cost are taken from a TCO report of Automotive Data of China Co., Ltd. (Tianjin, China) [53]. There are no charging facilities for FCEVs because they are generally funded by the government due to technical and financial constraints. A direct current charging station has been chosen for BEVs because of its high efficiency. The supply infrastructure and energy costs are taken from the average market price. The discount rate of economic cost is taken from the general level of the new energy industry [54].

2.2. Life Cycle Assessment (LCA)

LCA is an evaluation method that assesses a product’s environmental issues and potential environmental impacts throughout its life cycle [55]. It is used to analyze the process of a product, from raw material acquisition to production, use, and end-of-life. LCA contains four stages [55]:

(a)

Goal and scope definition—identify the system boundary and scope of the study based on the needs of the study;

(b)

Life Inventory Analysis (LCI)—collect the required data and develop a list of relevant inputs and outputs;

(c)

Life Cycle Impact Assessment (LCIA)—select the appropriate LCIA method based on the research needs and assess the product’s environmental impacts through different environmental impact indicators;

(d)

Life Cycle Interpretation—summarize and make recommendations.

This study is conducted using the LCA software OpenLCA version 2.0.4 and follows the ISO 14044 standard for life cycle assessment [55]. This study is based on one vehicle and analyzes its entire life cycle (from cradle to grave). The functional unit is one kilometer traveled by a BEV or FCEV in last-mile logistics. It includes the life cycle’s production, use (including maintenance), and end-of-life (including recycling) [56].

Table 2. Scope of this study.

Life Cycle Stage	BEV	FCEV
Manufacturing	Glider and powertrain (electric motor, battery pack, power electronics).	Glider and powertrain (fuel cell stacks, tanks, balance of plant (BOP), electric motor).
Operation	Maintenance, electricity consumption, charger.	Maintenance, hydrogen consumption.
End of Life (Disposal)	Considering the used glider, powertrain, and charger.	Considering the used glider and powertrain system.

Source: Authors.

shows the scope definition. In the manufacturing segment, the components of both vehicles are considered in terms of the glider and powertrain [57,58]. In the operation stage, it considers the BEV’s maintenance, electricity consumption, and charger. In addition to the FCEV, charger facilities are generally government-funded and set up in public places. In the end-of-life stage, it considers all the physical components of both vehicles [59,60]. For this purpose and scope, Figure 1 shows the system boundaries for this study.

The following details the assumptions used for this study:

• This study is based on Eastern China and uses the East China Power Grid (ECPG) as the grid source.
• Both vehicle types are light-duty commercial vehicles, which are suitable for last-mile delivery [17].
• The batteries and vehicle have the same lifespan of 10 years [61].
• The kilometers per year for a vehicle are 15,000, suitable for both types [61].
• This study considers gliders and powertrain systems as two parts of vehicle production [58].
• The disposal phase considers that, ideally, all materials can be recycled.

2.3. Life Cycle Inventory (LCI)

The data in the LCI list are derived from secondary data collected from the existing literature, official websites, and the Ecoinvent 3.9.1 Database. The database reflects the electricity market structures and background as of 2020, based on national and international statistical sources, particularly for China [61].

In this study, most of the datasets were selected from the global (GLO) and Rest of the world (RoW) boundaries. The aim is to enable the model to capture a wide range of supply chain activities across different regions, enhancing the model’s universality and ensuring comparability with the research. To capture regional characteristics, the electricity input for the use phase was modeled using data from the East China Power Grid (CN-ECGC), reflecting the specific emissions of the local power mix.

In line with the guidance provided by the Ecoinvent Association, this study adopts the allocation method, as cut-off by the classification system model [61]. This approach assigns waste treatment impacts to the waste generator, allows recyclable materials to enter subsequent systems burden-free, and handles co-products through allocation.

2.3.1. Vehicles Data

This study focuses on two representative Light Commercial Vehicles (LCVs) commonly used for last-mile transportation: the Nissan e-NV200 as a typical BEV and a Vivaro-e HYDROGEN-based FCEV. LCVs are widely adopted in last-mile logistics due to their compact size and high cargo capacity. The Nissan e-NV200, equipped with a 40 kWh battery pack, is a well-established model that offers a sufficient driving range and is particularly suitable for navigating narrow urban roads with frequent stops. For FCEV modeling, the Vivaro-e HYDROGEN was chosen as a reference vehicle. Since commercially available FCEVs suitable for last-mile delivery are limited and often hybridized, a full-system hydrogen fuel cell configuration was assumed in this study to enable a direct and fair comparison with the BEV. The key technical specifications used for the life cycle inventory modeling are summarized in

Table 3. Key technical specifications of BEV and FCEV.

Category	Parameter	BEV: Nissan e-NV200	FCEV: Vivaro-e HYDROGEN	Unit
Weights	Curb weight	1558	1226	kg
	GVW	2220	3075	kg
	Payload	662	1100	kg
Powertrain	Electric motor type	AC synchronous	AC synchronous	kW
	Energy source	Electricity	Electricity	–
Battery/Fuel cell	Battery/Fuel cell capacity	40	45	kWh
	Battery/Fuel cell type	Laminated lithium ion	PEMFC	–

Sources: compiled based on [61,62,63].

.

2.3.2. Inventory Data for Manufacturing

Table 4 presents the list of BEV manufacturing models. Some data are derived from literature assumptions and calculations:

• The glider occupies 41% of the curb weight [51].
• e-LCV was used with NMC111 batteries for city transportation [59].

Table 4. BEV modeling list for the manufacturing phase.

Components	Weight (kg)	Processes Used
Glider	910.2	market for glider, passenger car|glider, passenger car|Cutoff, U-GLO
Battery	276.5	market for battery, Li-ion, NMC111, rechargeable, prismatic|Cutoff, U-GLO
Electric motor	262.44	market for electric motor, vehicle|electric motor, vehicle|Cutoff, U-GLO
Power electronics	108.87	Refer to [50]

Sources: compiled based on [61,64,65].

Table 4. BEV modeling list for the manufacturing phase.

Components	Weight (kg)	Processes Used
Glider	910.2	market for glider, passenger car|glider, passenger car|Cutoff, U-GLO
Battery	276.5	market for battery, Li-ion, NMC111, rechargeable, prismatic|Cutoff, U-GLO
Electric motor	262.44	market for electric motor, vehicle|electric motor, vehicle|Cutoff, U-GLO
Power electronics	108.87	Refer to [50]

Sources: compiled based on [61,64,65].

For the FCEV configuration, the selection of all components is primarily informed by [30].

Table 5. FCEV modeling list for the manufacturing phase.

Components	Weight (kg)	Processes Used
Glider	784.6784	market for glider, passenger car|glider, passenger car|Cutoff, U-GLO
Fuel cell stacks	197.1135	market for fuel cell, stack polymer electrolyte membrane, 2 kW electrical, future|Cutoff, U-GLO
Electric motor	49.9863	market for electric motor, vehicle|electric motor, vehicle|Cutoff, U-GLO
Tank	157.5	Refer to [53]
BOP	36.7818	Refer to [53]

Sources: compiled based on [61,66,67].

presents the detailed manufacturing inventory for the FCEV. The component masses were determined according to the proportions illustrated in Figure 2. These revised shares are adapted from the hybrid vehicle structure proposed by [66], combined with the technical specifications of the Vivaro-e HYDROGEN, to ensure their applicability to a full FCEV system. Notably, the Vivaro-e HYDROGEN is equipped with three 700-bar hydrogen tanks. Therefore, the mass distribution was adjusted by referring to the modeling of two 700-bar tanks in Toyota vehicles and Opel technical data [67,68].

2.3.3. Inventory Data for Operation

The modeling of the BEV’s inputs is derived from [61]. It is believed that the environmental burden caused by BEV is mainly due to the grid structure, and the East China Grid, which is more developed in B2C, is chosen as the grid background here. Secondly, BEVs do not produce exhaust emissions, but they cause non-exhaust emissions due to brake, road, and tire wear resulting from vehicle operation. This study is based on the [63] model, as shown in

Table 6. Non-exhaust emissions.

Category	Emission (kg/vehicle·m)
Brake wear emissions	8.9 × 10−10
Road wear emissions	9.79 × 10−9
Tire wear emissions	5.73 × 10−8

Sources: compiled based on [61,63].

.

The modeling of the FCEV’s inputs is derived from [61]. It is believed that the environmental burden associated with FCEV is primarily due to the hydrogen production method. Therefore, the process of ‘market for hydrogen, gaseous|Cutoff, U–GLO’ in the Ecoinvent database has been chosen to represent a mixture of commercially available methods of manufacturing hydrogen and to account for the losses incurred during transportation. The hydrogen consumption is 0.85 kg/100 km [66]. The production efficiency of hydrogen is 85% [48]. Secondly, FCEVs also do not produce exhaust emissions; however, they do cause non-exhaust emissions, similar to those of BEVs, which are also included in the output, as noted in [63].

2.3.4. Inventory Data for End-of-Life

The end-of-life data for the Nissan e-NV200 are based on ten years of operation, prorated without regard to mass loss incurred during that time. This phase was based on the end-of-life modeling of the eLCV by [69], and the processes used are from [61]. One of the conventional end-of-life recycling methods for the NMC111 is hydrometallurgy, as noted by [59], which is also applicable to this study. The remaining automobile parts are scrapped by ‘wasted electric and electronic equipment,’ referring to the treatment of electronic materials in [69].

For FCEV, the calculation is based on ten years of operation, prorated without regard to mass loss incurred during that period. This phase was based on the end-of-life modeling presented in [60], and the processes used are described in [61]. All components are categorized into two main blocks: gliders and powertrain systems, for the purpose of recycling.

2.4. Life Cycle Impact Assessment (LCIA)

This study selects ReCiPe 2016 Midpoint (H), which provides a more comprehensive life cycle environmental impact assessment, and its scope of application is more suitable for China [70]. The ReCiPe 2016 method, as provided by the LCIA and OpenLCA software, is used in this study. The standards applicable to ReCiPe 2016, as listed on its official website, are used to derive the normalization. This process can be divided into two steps:

First, the results of characterization will be calculated by the software. The OpenLCA converts the LCI data (relevant emissions and resource usage) into potential levels of impact for different environmental impact categories, such as water eutrophication and marine toxicity. The results for each impact category are usually expressed in terms of specific impact units. For example, global warming is usually expressed in terms of CO₂ equivalent quantities. The impact scores, expressed in terms of the respective impact units for each environmental category, are the results of characterization. These results can facilitate a comparative analysis of two vehicles under the same environmental impact category.

Second, the values of normalization can be calculated. The results of normalization are the results of characterization divided by the standardized value of a specific rating scale. Since each unit is not the same across environmental impact categories, it is difficult to compare the degree of environmental impact between different environmental issues. Therefore, the environmental rating scale applicable to the ReCiPe 2016 methodology is used here to provide a relative scale by which different environmental impacts can be more easily compared. It helps us to see the relative importance of each environmental impact category and derive a weighting of the environmental impacts that need attention.

While global warming potential (GWP) is commonly regarded as the most influential category in vehicle LCA studies due to its direct link with climate change, this study also recognizes the relevance of other midpoint categories, especially those where BEVs or FCEVs may exhibit distinct disadvantages under current technological conditions. As noted in recent research [71], including qualitative assessments of electric vehicles, broader categories such as freshwater ecotoxicity, human carcinogenic toxicity, and marine ecotoxicity can offer important insights beyond GWP alone. Incorporating both quantitative results and qualitative interpretation therefore provides a more comprehensive understanding of how environmental burdens are distributed across impact categories and supports a more holistic comparison between BEVs and FCEVs.

3. Results

3.1. Life Cycle Cost Results

Table 7. The LCC of a BEV and an FCEV.

Vehicle Type	BEV (per Vehicle)	FCEV (per Vehicle)
One-time costs (RMB)	307,676.52	566,644.54
PV annuity (RMB)	154,308.09	92,500.57
PV residual (RMB)	56,640.13	133,885.94
TDCO (RMB)	405,344.48	525,259.17
LCC (RMB/km)	2.7	3.5

Note: All values are in RMB per vehicle, except for the last row (RMB/km).

presents the process data, as well as the final data, for both BEVs and FCEVs. Based on the modeling equations, a BEV’s average cost per kilometer over its life cycle (10 years) is approximately 2.7 RMB. Similarly, the cost per kilometer for an FCEV over the same life cycle is 3.5 RMB. Overall, an FCEV costs 0.8 RMB more per kilometer than a BEV, representing more than one-third of the BEV’s total cost per kilometer. It also shows that BEVs have an advantage in one-time payment costs, while FCEVs have an advantage in long-term operating costs. According to [18] research, the hydrogen price is the dominant driver of life cycle cost in commercial fuel cell vehicles, placing FCEVs within an approximate range of 3–5.6 RMB/km. The cost levels estimated for last-mile delivery vehicles in the present assessment fall within this interval and exhibit a comparable cost structure. Although the two analyses employ different modeling frameworks and functional indicators, both identify upstream energy pathways as a key factor shaping the environmental and economic performance of FCEVs.

3.2. LCA Results of the BEV

The results in Figure 3 show the percentage contribution of each life cycle stage to all environmental impact categories for BEVs. The manufacturing stage dominates across most categories, including global warming, mineral resource scarcity, water consumption, and terrestrial acidification, accounting for 60–90% of the total impact. This is primarily due to the production of batteries, glider fabrication, and electric motor manufacturing, which involve substantial material and energy inputs. The operation stage makes notable contributions in categories such as global warming and fossil resource scarcity, primarily due to electricity use from fossil fuels like coal and natural gas, which release CO₂, CH₄, and N₂O. End-of-life processes generally contribute a relatively small share across all impact categories (typically under 10%), but their role is non-negligible, particularly in categories such as marine ecotoxicity, marine eutrophication, and freshwater ecotoxicity.

The characterization and normalization results for different impact categories are listed in

Table A2. Life cycle environmental impact of a battery electric vehicle.

Impact Category	Characterization Results	Normalization Results
Fine particulate matter formation	5.51 × 101 kg PM2.5 eq	2.16 × 100
Fossil resource scarcity	5.91 × 103 kg oil eq	None
Freshwater ecotoxicity	6.90 × 103 kg 1,4-DCB	2.74 × 102
Freshwater eutrophication	1.13 × 101 kg P eq	1.73 × 101
Global warming	2.46 × 104 kg CO2 eq	3.07 × 100
Human carcinogenic toxicity	4.49 × 103 kg 1,4-DCB	4.36 × 102
Human non-carcinogenic toxicity	8.74 × 104 kg 1,4-DCB	2.80 × 100
Ionizing radiation	1.97 × 103 kBq Co-60 eq	4.11 × 100
Land use	5.31 × 102 m2a crop eq	8.61 × 10−2
Marine ecotoxicity	8.83 × 103 kg 1,4-DCB	2.03 × 102
Marine eutrophication	1.94 × 100 kg N eq	4.21 × 10−1
Mineral resource scarcity	7.67 × 102 kg Cu eq	6.39 × 10−3
Ozone formation, Human health	8.98 × 101 kg NOx eq	4.37 × 100
Ozone formation, Terrestrial ecosystems	1.07 × 102 kg NOx eq	6.05 × 100
Stratospheric ozone depletion	8.69 × 10−3 kg CFC11 eq	1.45 × 10−1
Terrestrial acidification	1.24 × 102 kg SO2 eq	3.02 × 100
Terrestrial ecotoxicity	4.49 × 105 kg 1,4-DCB	2.95 × 101
Water consumption	5.49 × 102 m3	2.06 × 100

. Figure 4 shows the bar chart of normalization results. The ReCiPe 2016 does not list the reference value of ‘fossil resource scarcity’ to calculate the normalized value; therefore, it has not been included in the table and chart. Figure 4 indicates that BEVs have substantial impacts in five environmental categories: human carcinogenic toxicity, freshwater ecotoxicity, marine ecotoxicity, terrestrial ecotoxicity, and freshwater eutrophication. The column values represent the ratio of the characterized results to the corresponding reference values for each impact category. Given its prominence, human carcinogenic toxicity is further examined in this section. However, the normalized value for global warming (3.07) appears lower than those of the other major categories. This may be explained by the use of a full life cycle perspective. As shown in later results, the manufacturing phase contributes most of the BEV’s environmental burden. Impacts in other categories may dominate during different life cycle stages, which can reduce the relative weight of global warming in the normalized results. If only the wheel-to-wheel stage had been considered, the severity of global warming’s impact might have been greater. Moreover, this study is based on last-mile transportation as a background, which is highly related to the greenhouse effect. Therefore, global warming is also included in the subsequent comparative analysis.

Figure 5 illustrates the contribution of individual sections of a BEV to human carcinogenic toxicity. The manufacturing phase of BEVs is the largest contributor to human carcinogenic toxicity, accounting for almost 90% of the total. In contrast, the end-of-life and operational phases are less significant. The glider remains the largest contributor among the three stages, accounting for 52% of the total. The components of the Li-ion battery and electric motor are also the second-tier contributors.

The analysis of global warming impact categories, as shown in Figure 6, reveals that the manufacturing phase is the predominant source of GHG emissions, accounting for 52.63%, with gliders and lithium batteries being key contributors. Despite the near equivalence in their emission contributions, manufacturing 1 kg of glider material results in 6.50 kg of CO₂ emissions, whereas 1 kg of lithium batteries yields 18.86 kg, which suggests that lithium batteries are a critical focal point for reducing GHG emissions. Surprisingly, the operational phase of BEV also contributes nearly half of the GHG emissions.

To further examine the sources of GHG emissions during the BEV’s operation phase, the overall contribution of the operation phase is set to 100% here, and the contribution value of each process in

Table 8. The contribution of BEV operation processes to global warming.

Process	Total Result [kg CO2 eq]	Contribution (%)
market for electricity, low voltage|CN-ECGC	8871.969676	82.02%
market for maintenance	1765.841874	16.32%
market for charger	179.116101	1.66%

refers to the weight of the process to the GHG emissions of the operation phase. Table 8 shows that electricity consumption during BEV operation accounts for 82% of the contribution, while maintenance throughout the whole life cycle accounts for 16% of the GHG emissions.

Data analysis reveals that electricity’s impact on GHG emissions during the vehicle operation phase constitutes 36.12%, as shown in Figure 7. Because of ECPG’s significant reliance on hard coal for power generation, particularly in the Jiangsu region, which has the highest GHG emissions. This reflects the heavy reliance of the East China Power Grid on non-renewable energy sources, despite the absence of tailpipe emissions during BEV operation.

3.3. LCA Results of the FCEV

Figure 8 shows the percentage contribution of each life cycle stage for FCEVs across the selected environmental impact categories. The manufacturing stage of FCEV shows significant contributions in several environmental impact categories, usually exceeding 90%. This dominance primarily stems from the fuel cell system, particularly its auxiliaries, including the compressor, pump, and electronic units, as well as the membrane electrode assembly, which contains precious metals. The materials and processes of the fuel cell system have high energy demands, high resource intensity, and complex supply chains, which make it a significant source of environmental burdens in particulate matter formation, toxicity-related impacts, and ecotoxicity categories. The operation stage of the FCEV shows significant contributions in several impact categories, especially those related to hydrogen supply. In the categories of fossil resource scarcity and ozone formation (human health), the operation stage contributes about 27–54%, mainly driven by hydrogen production. Based on the hydrogen dataset used in this study, the environmental impacts reflect a mixed supply dominated by fossil-based hydrogen production. Processes related to hydrogen production require large amounts of fossil energy and emit nitrogen oxides and volatile organic compounds. It highlights the importance of clean hydrogen production routes in reducing environmental impacts. The contribution of the EOL stage of FCEV to the overall impact is very small, usually less than 5%. These contributions primarily stem from waste treatment, resulting in a relatively small number of emissions.

The characterization and normalization results of the FCEV are shown in

Table A3. Life cycle environmental impact of a fuel cell electric vehicle.

Name	Characterization Results	Normalization Results
Fine particulate matter formation	9.62 × 101 kg PM2.5 eq	3.76 × 100
Fossil resource scarcity	9.95 × 103 kg oil eq	None
Freshwater ecotoxicity	1.62 × 104 kg 1,4-DCB	6.44 × 102
Freshwater eutrophication	2.24 × 101 kg P eq	3.45 × 101
Global warming	2.15 × 104 kg CO2 eq	2.69 × 100
Human carcinogenic toxicity	5.81 × 103 kg 1,4-DCB	5.64 × 102
Human non-carcinogenic toxicity	2.43 × 105 kg 1,4-DCB	7.78 × 100
Ionizing radiation	1.66 × 103 kBq Co-60 eq	3.46 × 100
Land use	7.24 × 102 m2a crop eq	1.17 × 10−1
Marine ecotoxicity	2.06 × 104 kg 1,4-DCB	4.74 × 102
Marine eutrophication	2.24 × 100 kg N eq	4.85 × 10−1
Mineral resource scarcity	9.55 × 102 kg Cu eq	7.95 × 10−3
Ozone formation, Human health	1.00 × 102 kg NOx eq	4.86 × 100
Ozone formation, Terrestrial ecosystems	1.14 × 102 kg NOx eq	6.42 × 100
Stratospheric ozone depletion	2.02 × 10−2 kg CFC11 eq	3.37 × 10−1
Terrestrial acidification	2.82 × 102 kg SO2 eq	6.88 × 100
Terrestrial ecotoxicity	1.21 × 106 kg 1,4-DCB	7.96 × 101
Water consumption	2.79 × 102 m3	1.05 × 100

. Figure 9 shows that the FCEV significantly affects five environmental problems, from top to bottom: freshwater ecotoxicity, human carcinogenic toxicity, marine ecotoxicity, terrestrial ecotoxicity, and freshwater eutrophication. For the same reason as the BEV, it also does not contain the impact category of fossil resource scarcity in the figure and table. Although global warming appears less prominent than other categories, its normalized value is still 2.69 times the reference value. However, for a better comparison and for the same reasons as for BEV, the overall contribution to global warming from the various stages of an FCEV is also further interpreted. In contrast, the pronounced impacts of FCEV in freshwater and marine ecotoxicity warrant further explanation at the supply chain and component levels. Contribution analysis was therefore conducted to identify the underlying drivers of these toxicity-related categories.

Figure 10 presents the contribution of individual sections of an FCEV to freshwater ecotoxicity. The manufacturing phase of the FCEV is the largest contributor to human carcinogenic toxicity, accounting for almost 95% of the total. In contrast, the end-of-life and operational phases are less significant. The BOP is the largest contributor among the three stages, contributing to 75% of the total.

Figure 11 illustrates the contribution of each stage of FCEV operation to the greenhouse effect, where the manufacturing stage accounts for nearly 80% of GHG emissions. In this stage, the glider and balance of plant (BOP) each contribute slightly more than 20%. Following the BEV approach, it is found that 1 kg of gliders contributes 6.5 kg of CO₂, while 1 kg of BOP contributes 122 kg of CO₂, which is 19 times more than that of gliders.

Table 9. The contribution of FCEV operation to global warming.

	Contribution (%)	Process	Total Result [kg CO2-Eq]
Total	100.00	FCEV operation	4688.417094
	80.34	Global market for hydrogen, gaseous	4053.63916
	19.36	Global market for maintenance	975.237873
Market for hydrogen, gaseous	79.26	RoW	3202.097066
	16.12	Europe without Switzerland	652.7096729
	1.97	Brazil (BR)	80.09754581
	1.97	India (IN)	79.73154527
	0.50	Sourth Africa (ZA)	20.42132035
	0.34	Peru (PE)	13.81578171
	0.12	Colombia (CO)	4.76622555

shows that the overall market data for hydrogen (including sales, production, and transport) accounted for 80% of the operation phase of the FCEV. The analysis of the hydrogen market data reveals that the rest of the world (RoW) (including the US and China) is the main source of market activity.

To further analyze the contribution of hydrogen market activity, Figure 12 examines the extent of GHG emissions from processes within the RoW, revealing that its two largest contributions are both from the petroleum market.

Both freshwater and marine ecotoxicity show a highly consistent pattern in the hotspot analysis. More than 90% of the impacts in the two categories are from the manufacturing stage, with BOP components representing the dominant share. The major source of ecotoxicity in both categories is the copper supply chain. Large amounts of copper used in multi-conductor cables result in significant contributions from copper cathode production, including the smelting of sulfide ores, electrorefining, and the management of sulfidic tailings. These processes release Cu²⁺, Zn²⁺, and Ag⁺ ions into aquatic environments, and these ions have very high characterization factors under the ReCiPe 2016 method. As a result, long-term heavy-metal leaching from copper mine tailings is the primary mechanism driving ecotoxicity in both freshwater and marine systems. This consistent pattern indicates that the unfavorable ecotoxicity performance of FCEVs in these categories is linked to current copper mining and refining practices. The impacts are therefore considered transitional. Improvements in copper recycling, enhanced tailings and wastewater management, and potential material substitution offer clear pathways for reducing ecotoxicity burdens in future FCEV production.

3.4. Comparative LCA Results

Figure 13 compares the contribution of BEVs and FCEVs in each impact category. The contribution of the larger vehicle is set at 100% for a more visual comparison. As shown in Figure 13, FCEVs exhibit higher overall contributions across most impact categories, while BEVs show greater impacts in greenhouse gas emissions, ionizing radiation, and water consumption. The overall characterization results are presented in

Table A4. The impact assessment results of BEV and FCEV (characterization).

Impact Categories	Unit	BEV_eLCV_Entire	FCEV_eLCV_Entire
Fine particulate matter formation	kg PM2.5 eq	55.13575	96.23729
Fossil resource scarcity	kg oil eq	5910.29779	9952.38284
Freshwater ecotoxicity	kg 1,4-DCB	6898.83269	1.62 × 104
Freshwater eutrophication	kg P eq	11.26494	22.37123
Global warming	kg CO2 eq	2.46 × 104	2.15 × 104
Human carcinogenic toxicity	kg 1,4-DCB	4486.25078	5811.18921
Human non-carcinogenic toxicity	kg 1,4-DCB	8.74 × 104	2.43 × 105
Ionizing radiation	kBq Co-60 eq	1971.51711	1658.47966
Land use	m2a crop eq	530.90106	724.10385
Marine ecotoxicity	kg 1,4-DCB	8827.90555	2.06 × 104
Marine eutrophication	kg N eq	1.94417	2.23652
Mineral resource scarcity	kg Cu eq	767.17251	954.93133
Ozone formation, Human health	kg NOx eq	89.79472	100.09817
Ozone formation, Terrestrial ecosystems	kg NOx eq	107.40992	113.82396
Stratospheric ozone depletion	kg CFC11 eq	0.00869	0.02024
Terrestrial acidification	kg SO2 eq	123.70375	281.86969
Terrestrial ecotoxicity	kg 1,4-DCB	4.49 × 105	1.21 × 106
Water consumption	m3	549.15705	278.69992

.

The normalization results for these two vehicles are listed in

Table A5. The normalization results of BEV and FCEV (unit: times of reference value).

Impact Category	BEV	FCEV
Fine particulate matter formation	2.16 × 100	3.76 × 100
Freshwater ecotoxicity	2.74 × 102	6.44 × 102
Freshwater eutrophication	1.73 × 101	3.45 × 101
Global warming	3.07 × 100	2.69 × 100
Human carcinogenic toxicity	4.36 × 102	5.64 × 102
Human non-carcinogenic toxicity	2.80 × 100	7.78 × 100
Ionizing radiation	4.11 × 100	3.46 × 100
Land use	8.61 × 10−2	1.17 × 10−1
Marine ecotoxicity	2.03 × 102	4.74 × 102
Marine eutrophication	4.21 × 10−1	4.85 × 10−1
Mineral resource scarcity	6.39 × 10−3	7.95 × 10−3
Ozone formation, Human health	4.37 × 100	4.86 × 100
Ozone formation, Terrestrial ecosystems	6.05 × 100	6.42 × 100
Stratospheric ozone depletion	1.45 × 10−1	3.37 × 10−1
Terrestrial acidification	3.02 × 100	6.88 × 100
Terrestrial ecotoxicity	2.95 × 101	7.96 × 101
Water consumption	2.06 × 100	1.05 × 100

. Figure 14 shows that both BEVs and FCEVs have extremely high environmental impacts on human carcinogenic toxicity, marine ecotoxicity, and freshwater ecotoxicity. Among these categories with significant impacts, FCEVs all have greater contributions than BEVs.

However, the total CO₂ emissions per vehicle are 2.46 × 10⁴ kg for the BEV and 2.15 × 10⁴ kg for the FCEV. Regarding the greenhouse effect and decarbonization, an FCEV can reduce carbon emissions by 3035.80 kg, as shown in Figure 15 and Table A4. Considering only the operation phase, FCEVs can lower CO₂ emissions by 53.51% per vehicle.

To further clarify the underlying drivers and support future research,

Table 10. The top 5 contributions of BEV and FCEV processes to global warming.

BEV		FCEV
Contribution (kg CO2 eq)	Process	Contribution (kg CO2 eq)	Process
2280	hard coal mine operation and hard coal preparation|CN	1950	natural gas venting from petroleum/natural gas production|GLO
2210	electricity production, hard coal|high voltage|CN-JS	1220	electricity production, hard coal, conventional|high voltage|ZA
1450	electricity production, hard coal|high voltage|CN-AH	1050	hydrogen production, gaseous, petroleum refinery operation|RoW
1080	electricity production, hard coal|high voltage|CN-ZJ	683	treatment of residue from shredder fraction from manual dismantling|RoW
819	electricity production, hard coal|high voltage|CN-FJ	681	hard coal mine operation and hard coal preparation|CN

lists the top five contributions to the grids for BEVs and FCEVs. Overall, most of the contributions for both vehicles are from the operation stage of the energy manufacturing structure. Both power generation and hydrogen production still rely on fossil energy sources, which cause extremely high GHG emissions.

4. Discussion

4.1. Economic Analysis

Economically, the cost of BEVs is more advantageous. To facilitate an economically interpretable comparison, life cycle cost intensity is considered on a per-kilometer basis, with values of 2.7 RMB per kilometer for the BEV and 3.5 RMB per kilometer for the FCEV. According to the LCC results, an average FCEV costs 0.8 RMB/km more than a BEV. This is based on the study by [47], which assumed a fleet of 200 vehicles. Therefore, on average, operating a fleet of FCEVs costs 160 RMB/km more compared to a fleet of BEVs. While these costs may seem prohibitively high for a company, it does not necessarily mean that FCEVs are an inferior choice [40,47]. There are three dimensions to support it. The primary reason for the higher costs associated with FCEVs is their elevated initial purchase price, which is 2.3 times that of BEVs. This disparity is largely attributed to the relative immaturity of FCEV technology compared to the more developed BEV technology [72]. It is important to note that comparing these two technologies may not be entirely fair, given the nascent state of FCEVs in the market. Thus, despite the current cost challenges, FCEVs still hold the potential for future adoption as their technology advances. The second reason is that FCEVs incur significantly lower annual operating costs than BEVs, with savings amounting to 60,000 RMB per year per car. This cost advantage primarily stems from the government-funded hydrogen supply facility for FCEVs, which substantially reduces expenditures in this area compared to BEVs. This cost efficiency highlights the potential long-term economic benefits of lower overheads and maintenance costs. Thus, for companies aiming for long-term sustainability, FCEVs represent a more suitable option for annual operating expenses. The third reason is the steady positive policy support for FCEVs. Governments, such as those in Shanghai, are continually launching policies to support both types of new energy vehicles [52]. BEVs are the first to be marketed as new energy vehicles, and the government has provided more subsidies for these types of vehicles to make them widely available in the market. However, China’s government policies have begun to show a neutral stance towards both vehicle types, following the widespread adoption of BEVs in the marketplace [73]. Notably, recent policy adjustments indicate a gradual reduction in support for BEVs, paralleled by pilot initiatives to integrate FCEVs in urban transportation systems in select Chinese cities [74]. Nevertheless, the governments are also stepping up financial support for the FCEV-related industry and setting up a blueprint for the expected number of launches [75]. This shift highlights a potential governmental inclination to promote FCEVs, similar to their earlier endorsement of BEVs. Therefore, enhancements in government incentives related to the purchase prices of FCEVs and subsidies for hydrogen fuel are poised to provide a significant impetus for companies to pivot towards FCEVs. As technology advances, improvements in vehicle longevity are anticipated, which will likely lead to a decrease in annual operating costs. Consequently, FCEVs are expected to gain popularity, driven by their economic advantages over time.

Ultimately, FCEVs have a longer-term advantage in mitigating the potential economic loss associated with carbon emissions. The Social Cost of Carbon (SCC) concept is introduced here to quantify the potential economic loss associated with emitting one kilogram of CO₂. According to [76], the potential economic cost of 1 kg of CO₂ emissions is between $0.177 and $0.805. Then, according to the results of this study, an FCEV can reduce an average of 10,434.0446 RMB, and a fleet can reduce 20.7 million RMB. Compared to the studies by [76,77], it can be inferred that the SCC has exhibited an increasing trend over the years. Thus, perhaps in the future, the potential cost losses caused by CO₂ emissions will be much greater than they are currently. Therefore, from the point of view of SCC, FCEV is also more suitable for large companies that have long-term interests in planning.

4.2. Environmental Analysis

This study concludes that FCEVs are better suited for Chinese enterprises that typically engage in last-mile logistics and distribution. However, BEVs may be a better choice for some enterprises with special environmental protection requirements (they may be logistics enterprises or customers of the former). Here are two key dimensions that help explain the causes of the greenhouse effect and other environmental issues. Firstly, according to the results, FCEVs have an advantage over BEVs in terms of greenhouse gas emission reduction, which aligns with the general consensus in the literature review. Under the functional unit assumption, when life cycle GHG emissions are expressed on a per-kilometer basis, the BEV and FCEV exhibit GHG intensities of 164 gCO₂/km and 143 gCO₂/km. These reflect the operational differences in GHG performance between the two technologies. On average, an FCEV can reduce 3035.80 kg of CO₂ emissions compared to a BEV. Additionally, based on a fleet of 200 vehicles, the fleet can reduce CO₂ emissions by 607,160 kg. However, since the technology of FCEVs is not yet mature, according to the results of this study, FCEVs have a greater impact than BEVs on most environmental impact categories, such as freshwater ecotoxicity and marine ecotoxicity. The technology of FCEVs is also developing at a rapid pace and will become increasingly mature [78]. In the future, FCEVs are expected to perform better in other impact categories.

4.3. Last-Mile Problem Analysis

The last-mile problem in parcel delivery is exacerbated by the complexity of routing, which is caused by the varied residential locations. This complexity often results in couriers spending excessive time searching for efficient traffic routes, thereby increasing delivery times and reducing overall efficiency. FCEVs offer a compelling solution to this challenge due to their superior range compared to BEVs. This advantage ensures that couriers are less likely to encounter situations where the vehicle runs out of power in complex terrain, thereby maintaining high operational efficiency. Furthermore, FCEVs can be recharged much more quickly than BEVs. For instance, the reference FCEV model of Vivaro-e HYDROGEN by [69] can achieve a full charge within 3 min, significantly enhancing the productivity of delivery operations by reducing downtime. This rapid refueling capability is critical in logistics, where time efficiency directly correlates with service quality and operational costs.

However, the high cost of FCEVs compared to BEVs is a deterrent to businesses, which is similar to the cost of last-mile transport. From a fleet operation perspective, such cost differences become particularly critical when expressed on a per-kilometer basis. Even modest unit cost gaps can accumulate into substantial expenditures for large logistics fleets. Whereas, with anticipated government support, reductions in initial purchase costs, and lower average annual operating expenses, large companies are increasingly likely to invest in FCEVs. In this context, policy-driven measures are being implemented to reduce operational barriers for FCEV. These measures include the expansion of hydrogen refueling networks and city cluster demonstration programs targeting large-scale logistics applications, which improve refueling accessibility and station utilization.

For last-mile logistics scenarios, coordinating large fleets along shared urban delivery routes has the potential to increase the utilization rate of existing hydrogen refueling stations [79]. This improved utilization can lower the effective cost of hydrogen supply and narrow the cost gap between FCEVs and BEVs. Economic incentives and operational efficiencies are gradually making FCEVs a potentially attractive choice for addressing the logistical challenges of last-mile parcel delivery.

4.4. Research Significance and Contribution

This study addresses the research question of large companies’ choice of vehicles for the last mile, where the preference is for FCEVs. Firstly, it evaluates BEVs and FCEVs economically and environmentally using both LCC and LCA methodologies, presents the results, and provides analysis. Secondly, this research found that an FCEV can eliminate 3035.80 kg of CO₂ emissions compared to a BEV; however, the LCC of the former is 0.8 km/RMB higher than that of the latter. Thirdly, this work found that FCEVs generally have a greater impact than BEVs on most environmental categories, and that both vehicles are categorized in the impact category of freshwater ecotoxicity, marine ecotoxicity, and human carcinogenic toxicity.

This study fills the research gap stemming from the lack of studies on the relationship between specific size companies and segmented supply chain logistics and enriches the existing literature in the field of last-mile logistics, particularly for new energy vehicles. In addition to the fact that this study examines the choice of BEVs and FCEVs in the last mile from a new perspective, another great contribution is that this study is different from the wide range of comparisons between BEVs and FCEVs in the existing literature, which generally agree that FCEVs are more environmentally efficient. This study found that FCEVs, although better at reducing the greenhouse effect, did not perform as well as BEVs in most environmental impact categories due to the limitations of current technology. Our work provides an analysis and discussion of different environmental impact categories as well.

5. Conclusions

In conclusion, this study provides a comprehensive comparison of BEVs and FCEVs in terms of life cycle economic costs and environmental impacts. The main findings are as follows:

• An average FCEV costs 0.8 RMB/km more than a BEV. However, FCEVs incur significantly lower annual operating costs than BEVs, with an annal saving of 60,000 RMB per vehicle.
• An FCEV can eliminate 3035.80 kg of CO₂ emissions compared to a BEV throughout the entire life cycle.
• Both vehicle types exhibit high impacts in terms of human carcinogenic toxicity, marine ecotoxicity, and freshwater ecotoxicity, with FCEVs showing higher contributions in these categories.

These findings indicate that the relative advantages of BEVs and FCEVs are not uniform across environmental or economic dimensions. Rather than identifying a single superior technology, the results highlight the importance of considering specific impact categories, life cycle stages, and operational contexts when evaluating both vehicle types. Although BEVs have a cost advantage at the current stage, policy support and lower long-term operating costs position FCEVs as a promising option for large companies with long planning horizons. FCEVs also demonstrate stronger performance in reducing life cycle CO₂ emissions. However, their higher impacts in categories such as freshwater ecotoxicity, marine ecotoxicity, and human carcinogenic toxicity reflect the early-stage maturity of fuel cell manufacturing and the carbon intensity of current hydrogen production pathways. These results underscore that decarbonization benefits alone do not capture the full environmental tradeoff between the two technologies.

These comparative patterns highlight that the relative performance of BEVs and FCEVs depends strongly on contextual conditions. BEVs tend to perform better in regions with lower grid carbon intensity or when vehicles operate over shorter lifetime distances, where electricity-related emissions dominate. In contrast, FCEVs show increasing advantages under longer operational lifetimes or when hydrogen is produced from low-carbon sources, conditions under which their CO₂ reduction benefits become more pronounced. These findings indicate that neither technology is universally superior; rather, their performance varies with energy pathways and operational requirements. This study also assesses the operational suitability of BEVs and FCEVs for last-mile delivery, considering factors such as range and charging or refueling requirements. The comparative results suggest that FCEVs may offer long-term advantages for large enterprises, although the extent of these benefits remains sensitive to future developments in technology and energy systems.

Despite the comprehensive cradle-to-grave scope of this study, several limitations should be acknowledged. The analyses rely on current assumptions about the East China Power Grid and hydrogen production pathways, both of which are evolving and may influence future comparative results. In addition, detailed manufacturing and end-of-life inventory data remain limited for certain components, especially fuel cell systems and large battery packs. To address these limitations, future research could incorporate scenario-based analyses of renewable hydrogen pathways, apply regionalized LCA data reflecting differences, and explore optimized hybrid fleet configurations that combine BEVs and FCEVs according to route characteristics and operational needs. Further work could also benefit from updated inventories to capture changes in technology maturity over time.