Assessing the economic and environmental impacts of battery leasing and selling models for electric vehicle fleets: A study on customer and company implications

.


Introduction
Climate change mitigation and the decarbonization of the economy stand as humanity's most pressing challenges.There is a consensus that achieving full decarbonization requires a deep transformation in energy production, consumption patterns, resource utilization, and environmental interactions (IPCC, 2022a).
However, some sectors, such as transportation, energy-intensive industries, and mining, pose significant difficulties in achieving decarbonization due to technical and economic barriers, as well as a resistance to changing consumption patterns and institutional frameworks (Ritchie and Roser, 2022).Among these sectors, road transport contributes a significant portion of global CO 2 emissions, with petrol and diesel combustion in internal combustion engines (ICEs) being the primary source (IEA, 2022a).
The Intergovernmental Panel on Climate Change (IPCC) affirms that replacing ICE vehicles with battery electric vehicles (BEVs) powered by low-emission electricity offers the greatest potential for decarbonizing land-based transport (IPCC, 2022a;IPCC, 2022b).The accelerated adoption of BEVs is driven by the maturity and commercial availability of lithium-ion batteries, coupled with their decreasing costs (IEA, 2022a;IEA, 2022b).Known for their high energy density and small size, lithium-ion batteries have surpassed previous battery technologies in terms of power output, charge retention, and reduced environmental toxicity (Li, et al., 2018;Manthiram, 2017;Kim et al., 2019).
Over the past 25 years, batteries have become 5%-10% more efficient and 15%-18% cheaper every year (IEA, 2022a;IEA, 2022b).This exponential growth has led to a remarkable increase in the global stock of BEVs, which grew over 1000-fold between 2010 and 2021 (IEA, 2022a).Driven by increasingly stringent climate policies, this growth trend is expected to continue into the future.Projections by the International Energy Agency (IEA) suggest a potential 12-and 21-fold increase in electric vehicle (EV) stocks between 2021 and 2030 (IEA, 2022a).
Despite this promising future, various challenges remain to be addressed, including one) safety issues; two) issues related to the availability of minerals and their social and policy impacts; three) environmental issues related to mineral extraction; four) infrastructure issues; and five) disposal issues.
Safety concerns primarily revolve around thermal runaway, fire risk and hazards associated with batteries under extreme conditions, demanding further research and development (R&D) (Sun, et al., 2020;Christensen et al., 2021).
Lithium, cobalt, and nickel are critical metals for manufacturing lithium-ion batteries (IEA, 2022a).While these metals are abundant, the increasing demand has strained existing mines, resulting in supply limitations and substantial price increases (IEA, 2022a).Furthermore, the concentration of the electric mobility supply chain in a few companies has led to social and policy issues, including child labor, poor working conditions, violence, governance concerns, corruption, and political instability (Huber and Steininger, 2022).
The sustainability of the extraction of minerals and raw materials has also raised environmental concerns (IPCC, 2022b).The mining of materials to produce lithium-ion batteries poses a high potential for soil, water and air contamination (Winslow, et al., 2018;Mrozik et al., 2013) as well as harmful effects on human health (Awasthi, et al., 2018).Moreover, the mining of lithium, cobalt and nickel occurs in dry areas and, at the same time, is a very water-intensive activity, resulting in a substantial negative effect on the water supplies and ecosystems in those areas (Sakunai, et al., 2021).
Another challenge lies in the development of charging infrastructure.The expansion of EV adoption demands a robust network of charging stations, including both public and private charging points, to ensure convenient access for consumers and to eliminate range anxiety (IEA, 2022a).The installation of charging infrastructure requires substantial investments in grid upgrades, energy storage systems, and efficient load management, which should be integrated with renewable energy generation to maximize environmental benefits (IEA, 2022a).
Finally, the end-of-life management of lithium-ion batteries requires appropriate recycling and disposal systems to minimize environmental harm and maximize resource recovery.Currently, the recycling rates for lithium-ion batteries are relatively low, with only a fraction of the materials being recovered (Christensen, et al., 2021).As the volume of retired batteries increases, developing efficient recycling technologies and establishing circular economy practices will become imperative (Olsson, et al., 2018;Wralsen et al., 2021).
To address these challenges, circular business models (CBMs) have emerged as potential solutions (Picatoste, et al., 2022).Through a comprehensive literature review, two predominant types of CBMs for BEVs have been identified (see Section 2.2 for detailed information): i) repurposing batteries following their initial use in BEVs, coupled with recycling, and ii) battery or vehicle leasing.Notably, however, while existing research has extensively addressed various facets of battery repurposing and recycling, limited attention has been paid to comprehensively studying the economic and environmental ramifications of battery leasing.
Battery leasing offers an alternative ownership model where consumers lease the battery separately from the vehicle, paying for the use of the battery and the energy it stores, while the vehicle itself is owned outright or leased separately (Picatoste, et al., 2022).This approach shifts the responsibility for battery ownership, maintenance, and end-of-life management from the consumer to the manufacturer or service provider.
The implementation of battery leasing can potentially yield some advantages.First, it can reduce the upfront cost of EVs, making them more affordable and accessible to a broader range of consumers (Sadek, 2012;Li and Ouyang, 2011).Second, battery leasing provides flexibility and potential cost savings, as consumers can upgrade to newer and more efficient battery technologies without having to replace the entire vehicle (Christensen, et al., 2012;Yang et al., 2017).Third, it enables manufacturers to maintain control over the battery's life cycle, facilitating efficient recycling and resource recovery (Liao, et al., 2018;Huang et al., 2021).
Despite the potential benefits, comprehensive studies assessing the economic and environmental impacts of battery leasing compared to traditional selling models are lacking.This research aims to fill this gap.To do so, the key research objective of this study is to assess the economic and environmental impacts of the CBM of battery leasing for BEVs and to compare them with those of the linear model of selling and buying batteries.
The novelty of this paper lies, first, in the comparison of circular and linear business models for BEVs and, second, in the combination of methods to evaluate their economic and environmental impacts.To provide a better representation of reality, instead of relying on the single "average" profile used in previous studies, a dedicated battery fleet model that incorporates various driving profiles is built.This detailed approach enhances the comprehension of how mileage, failures and replacements evolve over time for a fleet of users.This realistic representation is essential for assessing the economic and environmental impacts of business models, which depend on customer preferences and decisions.The study combines a net present value (NPV) analysis to evaluate the economic aspects and a cradle-to-grave life cycle assessment (LCA) to quantify the environmental impacts associated with both business models.Moreover, the NPV is analyzed from not only the manufacturer perspective but also the customer perspective.
By shedding light on the economic viability and environmental impacts of battery leasing, this study aims to provide valuable insights for policymakers, battery manufacturers, vehicle manufacturers, and other stakeholders involved in the transition to a sustainable and decarbonized transport sector.The proposed methods may be relevant for other business models targeting a substantial user base in the BEV domain, such as battery-as-a-service (BaaS) programs and battery swapping.Moreover, the findings can contribute to the ongoing discourse on sustainable mobility and serve as a basis for designing effective policies and business strategies to accelerate the adoption of EVs and promote a circular economy in the transportation sector.
This paper is structured as follows: Section 2 presents a review of the literature on CBMs for BEVs; Section 3 describes the methodology adopted in this paper; Section 4 sets out the main results; Section 5 discusses the results and situates them in a broader context; and Section 6 concludes the study.

Literature review
In this section, first, an overview of battery leasing and its supply chain ecosystem is presented.Second, a review of the literature on CBMs for BEVs is conducted.Finally, an overview of battery leasing initiatives is provided.

Battery leasing
Battery leasing is a business model where instead of purchasing batteries outright, customers can lease them for a specific period and pay a fee to use them.This approach helps to reduce the upfront cost of EVs, making them more affordable for customers (Li and Ouyang, 2011;Huang et al., 2021).The leasing agreement typically includes provisions for battery maintenance, replacement if the capacity deteriorates below a certain threshold, and sometimes even roadside assistance (Liao, et al., 2018).
Battery leasing can be considered a CBM, primarily due to its focus on maximizing resource utilization and waste reduction (Wralsen, et al., 2021;Sopha et al., 2022).First, in battery leasing, the same battery can M. Gonzalez-Salazar et al. be used by multiple users throughout its lifespan, which maximizes resource utilization and reduces the need for new battery production.Second, it reduces waste.When a leased battery reaches the end of its life or its capacity deteriorates significantly, the responsibility for recycling or appropriate disposal lies with the leasing company rather than the end user.This aspect ensures that batteries are managed appropriately, reducing the environmental impact associated with battery waste.
The battery leasing supply chain ecosystem works by establishing collaborations and partnerships between different actors, including battery manufacturers, leasing companies, vehicle manufacturers, customers, recycling facilities and regulatory bodies (see Fig. 1).Battery manufacturers produce batteries and supply them to leasing companies or vehicle manufacturers.Leasing companies enter into agreements with customers, providing batteries, maintenance, and replacement services.Customers are individuals or organizations that lease batteries.They pay a fee to use the batteries and enjoy the benefits of reduced upfront costs and maintenance responsibilities.When batteries reach the end of their useful life, leasing companies or vehicle manufacturers centralize the retirement of batteries and coordinate their disposal in recycling facilities.These entities ensure appropriate battery recycling or disposal to recover valuable materials and minimize environmental impact.Finally, regulatory bodies such as governments and regulatory agencies set guidelines, regulations, and standards related to battery leasing, recycling, and disposal, ensuring environmental sustainability and safety.

Circular business models for BEVs
This section presents a comprehensive review of the literature on CBMs for BEVs.An overview is presented in Table 1.Two main types of CBMs for BEVs are identified across studies: i) repurposing batteries after their first use in BEVs either in BEVs in a different market or in another application (e.g., energy storage), followed by recycling, and ii) battery or vehicle leasing.The majority of studies (~60%) have focused on CBMs dealing with the second use of batteries and recycling, while the rest (~40%) have investigated CBMs dealing with battery or vehicle leasing.Eighty percent of studies on CBMs dealing with repurposing batteries were written in the last 3 years, whereas 90% of studies dealing with leasing batteries were written before 2019.Two major types of analysis can be found across these studies: economic analysis and environmental analysis.Economic analysis includes business model analysis, profitability analysis and cost analysis.Environmental analysis mainly includes LCA and carbon footprint analysis.One-third of studies on battery repurposing and recycling have evaluated environmental impacts, while two-thirds have focused only on economic analysis.On the other hand, all studies on leasing have focused on economic analysis, and none have evaluated the environmental impact.
This literature review suggests that there is a lack of studies simultaneously evaluating the economic and environmental impacts of CBMs dealing with battery or vehicle leasing (Bohnsack, et al., 2014).A closer look at the environmental impacts of BEVs shows that batteries and their energy consumption during operation represent nearly 80% of the life cycle environmental impact of BEVs (Picatoste, et al., 2022).This figure suggests that to reduce the environmental impact of BEVs, it could be more effective to first address CBMs dealing with battery leasing and then address those dealing with vehicle leasing.
The benefits of battery leasing found in the literature include 1) making BEVs more affordable, as users do not need to pay for the battery upfront; 2) facilitating flexibility for users; 3) improving the lifetime management of batteries; 4) shifting the uncertainty of battery reliability from the user to the company, which increases residual car value; 5) accelerating the standardization of battery manufacturing and the charging interface; 6) promoting the repurposing and recycling of batteries, as battery leasing provides stable and concentrated sources of retired batteries; and 7) extending the battery lifespan by centralizing the maintenance of batteries.
In summary, little research has simultaneously examined the economic and environmental impacts of CBMs dealing with battery leasing despite their potential benefits.

Battery leasing initiatives
Table 2 provides a concise but not exhaustive overview of recent attempts by various companies to implement battery leasing for BEVs, showcasing different levels of success (more details can be found in (Cui, et al., 2023)).Noteworthy cases include Renault's Z.E.Flex, Nissan's LEAF Flex, Better Place, and Tesla.Renault initially introduced battery leasing with Z.E.Flex, reducing the upfront costs and offering flexibility; however, it eventually phased out this option.Nissan's LEAF Flex allowed battery leasing, but the company later transitioned to selling vehicles with batteries included.Better Place aimed to revolutionize EVs through battery swapping but faced challenges in infrastructure and demand, resulting in bankruptcy.Tesla introduced battery swapping convenience but discontinued the service due to low customer demand.The factors contributing to their lack of success included financial constraints, limited customer adoption, infrastructure limitations, high costs, and inadequate customer demand (Neil, 2022;Cui et al., 2023).
On the other hand, initiatives such as Mahindra Electric's NEMO Life, NIO's BaaS program, and Renault's Dacia Spring have experienced M. Gonzalez-Salazar et al. more success.Mahindra Electric's leasing of batteries through NEMO Life reduced the upfront costs and improved accessibility.NIO, a Chinese EV manufacturer, offers a BaaS program where customers can purchase the vehicle without the battery and subscribe to a separate battery pack monthly.This approach has gained popularity in China and has contributed to the company's success in the EV market.Renault's Dacia Spring uses battery leasing to enhance affordability and encourage EV adoption.The key factors underlying the success of these initiatives include timely implementation, customer-centric approaches, and effective pricing strategies (Cui, et al., 2023;Huang et al., 2021).
Notably, across these initiatives, lithium-ion batteries employing various chemistries, such as nickel manganese cobalt oxide (NMC), iron phosphate, and nickel cobalt aluminum oxide (NCA), are the most prevalent.Moreover, the battery capacities range from 15 to 100 kW h, effectively catering to different segments, including small vehicles, midsize vehicles, and SUVs.

Materials and methods
A discussion of the proposed methods for economic and environmental assessment and a comparison of these methods to state-of-the-art methods are presented in this section.

Discussion of existing methods for economic and environmental assessment
An important challenge is the development and acceptance of methods to simultaneously evaluate the economic and environmental impacts of products or services.Various methods have been used to separately evaluate economic and environmental impacts, and they are widely acknowledged by the scientific and industrial communities (Böckin, et al., 2022).Methods to evaluate economic performance include the profit margin method, which evaluates the profitability of products or services as functions of revenues and costs (Vogtlander, et al., 2017); the NPV method, which has been used to determine whether a new investment is profitable or not (Berk and DeMarzo, 2017); and the Business Model Canvas (Osterwalder and Pigneur, 2010) and its extension to circularity (Lewandowski, 2016), which have been used to identify new business models and value propositions.On the environmental side, LCA has been widely used to quantify the environmental impact of services and products, and its use is expected to continuously grow in the future to test and implement environmental or climate policies (Sala, et al., 2021;Sanyé-Mengual and Sala, 2022).
Combinations of these tools have been explored in prior research, as described in (Pieroni, et al., 2019).For example, a method called business model life cycle assessment (BM-LCA), combining conventional LCA analysis with a profit margin analysis, was recently proposed by (Böckin, et al., 2022;Goffetti et al., 2022).The method shows that Can innovative business models overcome resistance to electric vehicles?Better Place and battery electric cars in Denmark (Christensen, et al., 2021) 2012 Economic Battery leasing Urban electric vehicles: a contemporary business case (Sadek, 2012) 2012 Economic Battery leasing Business models for sustainable technologies: Exploring business model evolution in the case of electric vehicles ( Bohnsack, et al., 2014) 2014 Economic Battery and vehicle leasing Battery ownership model medium duty HEV battery leasing & standardization (Kelly, et al., 2015) 2015 Economic Battery leasing Optimal planning of swapping/charging station network with customer satisfaction (Yang, et al., 2017) 2017 Economic Battery leasing Competitiveness and sustainability effects of cars and their business models in Swedish small town regions (Nurhadi, et al., 2017) 2017

Economic and environmental
Vehicle leasing A consumer-oriented total cost of ownership model for different vehicle types in Germany (Letmathe and Suares, 2017) 2017 Economic Battery leasing The impact of business models on electric vehicle adoption: A latent transition analysis approach (Liao, et al., 2018) 2018 Economic Battery and vehicle leasing Circular business models for extended EV battery life (Olsson, et al., 2018) 2018 Economic Repurposing Towards sustainable business models for electric vehicle battery second use: A critical review (Reinhardt, et al., 2019) 2019 Economic and environmental

Repurposing
Towards a circular and low-carbon economy: Insights from the transitioning to electric vehicles and net zero economy (Bonsu, 2020) 2020 Economic Repurposing Environmental potential of reusing, renting, and sharing consumer products: Systematic analysis approach ( Amasawa, et al., 2020) 2020 Environmental Vehicle leasing Buy, lease, or share?Consumer preferences for innovative business models in the market for electric vehicles (Huang, et al., 2021) 2021 Economic Battery and vehicle leasing Circular business models for lithium-ion batteries -Stakeholders, barriers, and drivers (Wralsen, et al., 2021) 2021 Economic Repurposing Integration of energy flow modeling in life cycle assessment of electric vehicle battery repurposing: Evaluation of multi-use cases and comparison of circular business models (Schulz-Mönninghoff, et al., 2021) 2021 Environmental Repurposing Circular business models for electric vehicle lithium-ion batteries: An analysis of current practices of vehicle manufacturers and policies in the EU (Albertsen, et al., 2021) 2021 Economic Repurposing Enablers and barriers for circular business models: an empirical analysis in the Italian automotive industry (Urbinati, et al., 2017) 2021 Economic Repurposing Circularity and life cycle environmental impact assessment of batteries for electric vehicles: Industrial challenges, best practices and research guidelines (Picatoste, et al., 2022) 2022 Environmental Repurposing Mapping a circular business opportunity in electric vehicle battery value chain: A multi-stakeholder framework to create a win-win-win situation (Chirumalla, et al., 2022) 2022 Economic Repurposing Challenges and recent developments in supply and value chains of electric vehicle batteries: A sustainability perspective (Rajaeifar, et al., 2022) 2022 Environmental Repurposing Barriers and enablers of circular economy implementation for electric-vehicle batteries: from systematic literature review to conceptual framework (Sopha, et al., 2022) 2022

Repurposing
Prioritising low-risk and high-potential circular economy strategies for decarbonization: A meta-analysis on consumer-oriented product-service systems (Koide, et al., 2022) 2022 Environmental Vehicle leasing M. Gonzalez-Salazar et al.
economic and environmental activities may be dissociated, and that decision-making might be enhanced by taking economic and environmental impacts into account.However, it also presents certain shortcomings.First, most importantly, the profit margin analysis does not consider the time value of money.Second, it ignores or downplays the effects of depreciation and taxes over time.
To overcome these shortcomings, some studies propose the NPV as a method to evaluate economic performance and LCA to evaluate environmental impacts.There are numerous examples in different fields, e. g., (Li, et al., 2019;Sattari et al., 2022;Tantisattayakul et al., 2018).While the NPV method considers costs and revenues in the same way as profit margin analysis, it considers the time value of money.This aspect, coupled with the fact that it considers the effects of present and future investments, depreciation, and taxes, makes the NPV method suitable for evaluating activities lasting several years.Given the robustness of the LCA-NPV combination and the numerous studies applying it to different fields, it is down-selected and adapted for this analysis.

Proposed method
The LCA-NPV approach is adapted to analyze the economic and environmental performance of business models based on battery leasing and selling for EVs.An overview of the proposed method is shown in Fig. 2. For the economic evaluation, a fleet model and an NPV model are built and linked.These models use various technical, economic, and financial assumptions from state-of-the-art battery EVs.The NPV is evaluated from not only the perspective of the company manufacturing batteries but also the perspective of the customers using them.To conduct the environmental evaluation, i.e., the LCA, the results from the fleet model (i.e., the number of batteries and replacements needed) and the inventory of the energy and materials that are required across all the stages of the life cycle are used.Details about the NPV and LCA methods, as well as the cases analyzed, are presented in the following sections.

Economic evaluation
This subsection explains in detail the business, fleet and NPV models used, the prices and costs assumed, and the uncertainty quantification.

Business models
A company manufacturing lithium-ion batteries for EVs can apply different business models to offer batteries to users, which might imply different economic and environmental impacts.For simplicity, it is assumed here that companies manufacturing lithium-ion batteries are also companies manufacturing EVs.Two main business models are proposed for the analysis; see Fig. 3.
The refence case is a business model in which a company manufactures batteries and sells them directly to EV users (i.e., customers).In this case, customers retain ownership of the batteries, use them, perform maintenance if needed and dispose of them at the end of their life.While customers can sell their retired batteries for second life uses at a recovery price, which somewhat alleviates their upfront investment in the batteries, such selling is not an economic goal for the manufacturing company.Based on (Abdelbaky, et al., 2021), it is assumed that a small portion of manufactured batteries (2%) do not comply with the warranty.This assumption has two consequences.First, additional units need to be manufactured and provided to customers to comply with the warranty, with no additional income.Second, malfunctioning manufactured units have reached their end of life.
The second case is a business model in which a company leases batteries to users on an annual basis.This means that the company retains ownership of all batteries and rents them to users in exchange for a constant annual fee.The company ensures the proper functioning of batteries by replacing and repairing malfunctioning units.It is expected that a portion of the fleet of batteries is continuously out of operation because of malfunctioning and repairs.Information on this fleet overcapacity is scarce, but values ranging from 0.06% for car-sharing services in Munich (Germany) (Sohl, 2019) to 0.5% for electric scooter sharing in Berlin (Germany) (Severengiz, et al., 2020) were found in the literature.To be conservative, it is assumed that a 0.5% additional fleet is always needed to meet demand.When batteries retire, the company sells them for second life uses and receives additional income.As in the reference case, it is assumed that 2% of manufactured batteries cannot comply with the warranty and that additional units need to be manufactured.

Fleet model
The following subsections make explicit assumptions concerning driver profiles, fleet size, battery characteristics, and fleet simulation.

Mileage and fleet size.
The analysis is focused on Germany, a country that produces EVs and that has high e-mobility and BEV aspirations (Secinaro et al., 2020;Ziegler and Abdelkafi, 2022).A fleet size of 10,000 BEVs and five typical annual mileage driver profiles are considered.The typical driver profiles are outlined in Table 3, and they are taken from VuMA Touchpoints, a German national market survey (ARD-RMS-ZDF, 2022).These driver profiles are not specific to any particular type of vehicle; hence, it is assumed that they can be applicable to BEVs, as suggested in a study on the total cost of vehicle ownership in Germany (Letmathe and Suares, 2017).

Battery characteristics.
A lithium nickel manganese cobalt oxide (NMC) battery is considered, as this type of battery represents the most common battery chemistry in BEVs today (IEA, 2022a).To be consistent with the environmental analysis, the characteristics of the NMC battery shown in the Product Environmental Footprint Categorical Rules (PEFCR) for rechargeable batteries guideline (RECHARGE batteries, 2020) are assumed.The PEFCR do not disclose the chemistry of batteries, and they provide general information, such as the weight (225 kg) and lifetime energy use (8000 kW h) of batteries.However, other key characteristics are missing, and therefore, some assumptions are made.First, an energy density of 162 W h/kg is assumed, which is state of the art (Wassiliadis, et al., 2022) and translates into a battery capacity of 36.5 kW h.Second, for the business model of selling batteries, it is assumed that the company offers a warranty to customers of 160,000 km and 8 years at a depth of discharge (DoD) of 80% (Wassiliadis, et al., 2022;Wicki et al., 2022).Third, regarding the lifetime of NMC batteries, the most recent reviews of performance and LCAs for batteries (Peters, et al., 2017;Aichberger and Jungmeier, 2020) are used here; they suggest 180,000 km, 12 years and 1850 cycles for a DoD of 80%.

Strategies to manage battery mileage.
In the business model of selling batteries, a battery is used by a single user (i.e., the buyer) throughout its entire lifetime.The battery's lifetime depends on various parameters but primarily on the user's mileage.Therefore, users with a high annual mileage will likely reach the battery's retirement point during the warranty period of 8 years.This means that the company will manufacture and provide a new battery to these customers without receiving any income.
For the case of leasing batteries, batteries are distributed among different users every year.This redistribution of batteries across users reduces the risk that some batteries are excessively used by drivers with a high annual mileage.However, it also increases the mileage of batteries underused by drivers with a low annual mileage.Two ways to redistribute batteries on an annual basis across users are tested (see Fig. 4).The first strategy is a random distribution.In this strategy, once batteries have been used by customers during the first year, they are randomly distributed to new users in the second year and so on.The key advantage of this strategy is that it is simple, requires little effort and is effective.The second strategy is a smart distribution.This strategy aims to minimize the number of batteries with extremely low or extremely high mileage by selecting the sequence of users based on their driver profile.For example, if a user with a high annual mileage uses a battery in year one, then the battery is followed in year two by a user with very  low annual mileage.This strategy requires more management effort but also a higher utilization than the random strategy.

Fleet simulation.
To comprehensively assess the performance of the entire fleet of BEVs across various business models, a robust fleet model is developed in Microsoft Excel and @Risk version 8.2.2.This model evaluates the mileage and lifespan of each individual battery throughout the analysis period, predicting the optimal time for replacement.The assessment of mileage and lifespan is contingent upon the specific business model employed.For instance, in the business model of selling batteries, where each battery is assigned to a single user throughout its lifespan, the allocation of batteries within the fleet is determined based on the distribution of driver profiles, as illustrated in Table 3.The number of batteries assigned to each driver type x is mathematically determined by the following equation: where s x is the total number of batteries driven by driver type x and k x is the share of the driver profile shown in Table 3.The fleet model estimates the mileage of each battery over the years as a linear function of the annual mileage for the different driving profiles using Eq. ( 2): where m x is the annual mileage for the type of driver x and M x is the overall mileage over the period of analysis, i.e., 12 years.The lifespan of a battery L x used by driver type x (in years) is then a function of the lifetime mileage expectancy of 180,000 km and the annual mileage: If lifespan L x is lower than 8 years, which is the warranty period, then the battery must be replaced.
In the case of the battery leasing business model, where batteries are utilized by multiple users over their lifespan, a distinct methodology is adopted to ensure efficient allocation.Specifically, for the leasing scenario employing a smart distribution, the allocation process follows the following guidelines.Initially, in the first year, batteries are assigned based on the distribution of driver profiles (Table 3), similar to the battery selling model.However, starting from the second year, an optimization algorithm is employed.This algorithm is designed to minimize the number of batteries surpassing the designated lifetime mileage of 180,000 km.It works as follows: if a battery is driven by a high-mileage user in the first year, the algorithm allocates it to a low-mileage user in the subsequent year and so forth.The aim is to keep the mileage within the expected range (refer to Fig. 4 for a visual representation).This optimization algorithm is developed within the @Risk version 8.2.2 framework.
In the leasing scenario with a random distribution, the allocation of batteries follows a randomized approach.Initially, in the first year, batteries are assigned based on the distribution of driver profiles, similar to the previous cases.However, starting from the second year, a Monte Carlo simulation dominates the allocation process (as illustrated in Fig. 5).During this simulation, batteries are redistributed among different drivers in a random manner, guided by a probability distribution that encapsulates the driving profiles outlined in Table 3. Notably, the Monte Carlo simulation is repeated 100 times through bootstrapping, and the results are subsequently averaged to ensure robustness.

Prices and costs
The cost of producing NMC batteries, as well as their market price and the recoverable value of batteries that reach their end of life, is fundamental for economic analysis.Regarding the cost and price of NMC batteries, this analysis is based on the database of battery cost forecasting available in (Mauler, et al., 2021).The database values are averaged, and a curve with exponential decay following the equation below is assumed: where C(t) is the price/cost at time t, C o is the price/cost at time t = 0 (in this case, 2022), and τ is the exponential decay constant.Exponential decay constants are estimated using the values for 2022 and 2030 and are 4.9% for the battery's cost and 4.2% for the price.The assumed values are shown in Fig. 6.
To estimate the recoverable value of batteries, values from prior studies are collected, adjusted for inflation using World Bank data (World Bank, 2022) and averaged; see Table 5 in the Appendix.An exponential decay function with a decay constant equal to that of the battery's cost (i.e., 4.9%) is assumed.
For battery leasing, an essential economic aspect is the annual cost of maintaining each battery.There is a limited literature solely on the expense of maintaining batteries.Studies on the total cost of EV ownership include the costs of maintenance and repairs for the entire EV but not for the battery alone, e.g., (Liu, et al., 2021;Lebeau et al., 2019).Two studies are found that disaggregated the maintenance costs for batteries alone.The first study by the National Renewable Energy Laboratory (NREL) and Ricardo discloses an annual constant cost of $100/battery (Kelly, et al., 2015).The second study by the Argonne National Laboratory (Burnham, et al., 2021) discloses a cost of $495 for a service interval of 80,000 km.As the assumed lifetime is 180,000 km, this cost doubles, i.e., $990 over the lifetime.For simplicity, this cost is annualized, and depending on the lifetime in years, it ranges between $83/battery and $124/battery.Finally, assuming an exchange rate in which 1 U.S. dollar equals 1 euro, the average of these three numbers is used, i.e., €102/battery.

NPV calculationthe company perspective
The NPV is the difference between the present value of the benefits and the present value of the costs of a project or an investment (Berk and DeMarzo, 2017).First, the NPV from the perspective of the company manufacturing batteries is evaluated.It is determined as the present value of all project cash flows using the following equation, as described in (Berk and DeMarzo, 2017): where N t is the net cash flow, i.e., cash inflowcash outflow; i is the Fig. 6.Assumed NMC battery price, cost, and recoverable value (Mauler, et al., 2021).
discount rate, which here is assumed to be 10%; and t is the period of the cash flow in years.The total number of periods N is as long as the assumed lifespan of batteries, i.e., 12 years.All cash flows are then converted into the present value, which is assumed to be for 2022.Cash flows that are included in the analysis include sales, capital costs, operation, and maintenance (O&M) costs, manufacturing costs, depreciation, and income tax (30% in Germany).The types of cash flows differ for the two different business models.For the business model of selling batteries, the company incurs costs for manufacturing batteries and receives income from selling them, and the gross input is then taxed.For leasing batteries, the company incurs costs for manufacturing batteries, and these costs are accounted for as capital expenditures, as the company retains ownership of the batteries.Hence, batteries are company assets that depreciate, which impacts the earnings before interest and taxes (EBIT) and the taxes that the company should pay.The company should also cover the costs of maintaining and repairing batteries.In return, the company receives income from leasing batteries with a constant annual fee.Additionally, the company receives further income from selling batteries that reached their end of life for recycling or other purposes.While the manufacturing and maintenance costs can be found in the prior literature, the leasing fees need to be estimated.To evaluate the leasing fees, it is assumed that leasing batteries should equal the same NPV as selling batteries, following the approach suggested in (Böckin, et al., 2022).Batteries that do not reach their expected lifetime can then be sold for other uses.Here, the price at which they can be sold SP t in year t is a function of the market price MP t , the remaining lifetime of the battery (LM is the lifetime mileage expectance of the battery, i.e., 180,000 km, and m t is the mileage of the battery in year t) and an assumed factor of 80%: Table 4 shows an overview of the assumptions for evaluating the NPV from the perspective of the manufacturing company.

NPV calculationthe customer perspective
The NPV from the customer perspective was used to evaluate what option could be more cost-effective.Every cash flow made by customers to acquire or lease batteries is evaluated.Then, Eq. ( 5) is used to evaluate the NPV.Next, the NPV is divided by the lifetime mileage in kilometers.For customers buying the battery, the income received by selling it once it reaches its end of life is included.An important distinction is that the discount rate for a customer is not the same as that for a manufacturing company.The discount rate for customers is assumed to be the current U.S. treasury yield for 10 years, i.e., 3.82% (Yahoo, 2023).

Uncertainty quantification
To determine the uncertainty in the NPV, a Monte Carlo simulation is performed.The purpose of this study is not to offer a comprehensive and detailed evaluation of prediction accuracy but, rather, to quantify the degree of uncertainty.The NPV depends on different variables that involve uncertainties, including 1) the price, cost, and recoverable value of batteries; 2) the exponential decay constant for the price and cost of batteries; 3) the discount and tax rates; 4) the percentage of failed batteries; 5) the maintenance cost per battery for the leasing options; and 6) the fleet overcapacity for the leasing options.
For each of these variables, a probability density function and a correlation matrix are defined; see Tables 6 and 7 in the Appendix.As the uncertainty of many of these variables is unknown, uniform distributions for these variables are selected.These distributions require minimum and maximum limits and can provide a first order of magnitude of the uncertainty and the key influencing factors.Monte Carlo analysis is performed in @Risk version 8.2.2 for Microsoft Excel using a Latin hypercube sampling method with a Mersenne twister generator and 1 million trials.

Environmental evaluation
For the environmental evaluation, a cradle-to-grave LCA for each of the business models and scenarios described above is conducted.The analysis follows the ISO 14044 standard, according to which the structure of LCA includes four parts, namely, goal and scope definition, inventory analysis, impact assessment and interpretation.

Goal and scope definition
This study aims to quantify and compare the cradle-to-grave environmental impact of two business models, namely, NMC battery selling and NMC battery leasing.This goal is different from conventional LCAs, which typically focus on the environmental impact of products based on their function.All the lifetime stages of NMC batteries for the two business models are considered, including raw material extraction, manufacturing, distribution, use and end of life (see Fig. 7).Two main data sources are used to evaluate the environmental impacts, the first at a battery scale and the second at a fleet scale.First, at a battery scale, for each battery, the LCA model follows the PEFCR guidelines for rechargeable batteries (RECHARGE batteries, 2020).One important reason for following the PEFCR guidelines is that they might become the EU standard for conducting LCAs on lithium-ion batteries, as pointed out by (Yudhistira, et al., 2022).Second, at a fleet scale, the fleet model is used to evaluate the number of batteries needed in each business model.The environmental impacts are evaluated for the conditions in Germany and for a 12-year period, which is the same period of analysis for the NPV.
M. Gonzalez-Salazar et al. by all users throughout the 12-year period is adopted as the functional unit (FU).This amount is calculated using the following equation: where M i is the annual mileage for driver type i and x i is the number of type i drivers, with both values being derived from Table 3.The calculated FU is 1715.97 million kilometers traveled.

Considerations at the battery level.
At the battery level, cradleto-gate inventories are taken from the Product Environmental Footprint (PEF) database (RECHARGE batteries, 2020), which was developed by Ecoinvent.As mentioned above, the PEFCR provide only limited information about battery characteristics and uses.While they disclose the weight (225 kg) and lifetime energy use (8000 kW h) of batteries, other key features are not provided.For example, the PEFCR do not disclose the chemistry of the battery, nor do they disclose its energy density, lifetime mileage, number of cycles, efficiency, or DoD.Therefore, some assumptions are made.
First, an energy density of 162 W h/kg is assumed, which is measured from state-of-the-art batteries in (Wassiliadis, et al., 2022).Using this value, a battery capacity of 36.5 kW h is estimated.Second, a lifetime energy use of 8000 kW h was used in the PEFCR study, which was estimated as the lifetime energy losses to the battery and charger.While this method is commonly used in the literature (e.g.(Zackrisson, et al., 2010;Cusenza et al., 2019),), the assumptions used in the PEFCR study are unknown.An alternative method is to evaluate energy use as a function of the lifetime mileage of the battery in kilometers and the average powertrain efficiency in kWh/km.A lifetime mileage of 180, 000 km is assumed, as it is the most common value according to recent reviews of LCA for batteries (Peters, et al., 2017;Aichberger and Jungmeier, 2020).An average powertrain efficiency of 0.138 kW h/km is assumed, which is the measured value for the World Harmonized Light-Duty Test Procedure (WLTP) according to (Wassiliadis, et al., 2022).The resulting lifetime energy use based on this second method is 24,840 kW h.For comparison, two scenarios with different lifetime energy uses are considered: a) a reference case using 8000 kW h, as in the PEFCR study, and b) 24,840 kW h, an "alternative" case calculated with the second method described above.

Considerations at the fleet level.
The fleet model is used to determine the number of batteries required to serve the demand of 10,000 users of BEVs with the driving profiles shown in Table 3.The fleet model calculates the mileage of the batteries and evaluates when a battery needs to be replaced.It also determines the number of batteries that fail to comply with the warranty and those that break and are not available.These amounts are used following a conventional LCA methodology for building a life cycle inventory (LCI).The LCIs evaluated with the fleet model are shown in Section 4.1: Results.
An important question is how to consider the environmental impact of batteries that after the 12-year period have not achieved their lifetime.Here, the approach suggested in (Wilson, et al., 2021) is followed.This approach allocates inputs and outputs based on the fraction of the expected lifetime after the 12-year period of analysis.For example, if after 12 years the mileage of a particular battery is 90,000 km, i.e., 50% of its lifetime mileage expectancy (180,000 km), then its environmental impacts are 50% of those of a battery reaching its lifetime in 12 years.The allocation factor is defined using the following equation: where m t=12 is the mileage of the battery after 12 years of operation and 180,000 km is the lifetime mileage expectancy.
Another important consideration at the fleet level is the environmental impact associated with the maintenance or repair of batteries.While information on the environmental impact of maintenance at the vehicle level can be found in the literature (e.g. ( de Souza, et al., 2018;Xiong et al., 2019),), data on maintenance for the battery alone are lacking.There is a very limited amount of publicly available information on repairs for batteries and their environmental impact; to the best of the authors' knowledge, only one study (Granehed and Lövstedt, 2022) discloses such information.Due to this lack of information, which has already been noted in (Picatoste, et al., 2022), some assumptions are made.In the study conducted by Granehed & Lövstedt, three typical repair cases are identified, namely, a common case, a min case and a max case.However, no frequency of repairs is provided for these cases.It is here assumed that 75% common repairs, 12.5% min repairs and 12.5% max repairs can be expected on an annual basis for fleet overcapacity (i.e., 50 batteries) in the leasing scenarios.

Life cycle impact assessment and results interpretation
In this analysis, the guidelines of the PEFCR for rechargeable batteries (RECHARGE batteries, 2020) regarding the impact assessment categories are followed.The assessment method used is the environmental footprint (EF) methodology and the midpoint indicator, which is employed in the PEFCR.For simplicity, three impact categories are selected for comparison: climate change, resource use (fossils) and resource use (mineral and metals).
LCA analyses were carried out with the open-source software OpenLCA developed by GreenDelta, 2023a,b).The PEF database for OpenLCA and the LCIA method v.2.0.2, both openly available in Fig. 7. System boundary for the LCA study.
M. Gonzalez-Salazar et al.OpenLCA Nexus (GreenDelta, 2023a,b), were used.The results are extracted and visualized in Microsoft Excel and Tableau and presented in the next section.

Results
The following subsections reveal the results of the fleet analysis and the economic and environmental assessments.

Fleet analysis
When a BEV is driven by a single user, for example, in the case of battery selling, the mileage and lifespan after 12 years of operation depend on the driver type and annual mileage.This relationship is illustrated in Fig. 8.The linear nature of the model means that a higher annual mileage results in greater total mileage but a shorter lifespan after 12 years.For example, batteries with annual mileages below 15,000 km will not reach the expected lifetime mileage of 180,000 km, while batteries with annual mileages above 15,000 km will exceed the guaranteed mileage and require replacements (represented by the red line in Fig. 8).
To evaluate the environmental impact of batteries that do not reach the expected lifetime mileage, an allocation factor is used.This factor, calculated using Eq. ( 8), compares the battery mileage to the lifetime mileage expectancy (represented by the green line in Fig. 8).Importantly, this allocation factor can also be applied to battery leasing business models.
In contrast, when a battery is leased, its mileage depends on a series of drivers with different annual mileages.Consequently, not all batteries have the same lifespan over time.Analyzing the mileage distribution across the battery fleet provides valuable insights.Fig. 9 presents this distribution for different business models, revealing that battery leasing significantly reduces the spread in mileage compared to battery selling.The coefficient of variation (i.e., the ratio of the standard deviation to the mean), a commonly used measure of the distribution spread, indicates that selling batteries results in the widest range (50% coefficient of variation), while battery leasing with random and smart distributions shows narrower ranges of 14% and 7%, respectively.A narrower spread implies fewer batteries with extremely high or low mileages.For instance, in a 12-year period, 90% of batteries sold will fall within the mileage range of 72,000 to 168,000 km (see Fig. 10).This range narrows to 103,000-171,000 km for battery leasing with a random distribution and 141,000-177,000 km for battery leasing with a smart distribution.
Notably, while the spreads narrow for the leasing cases, their median mileages increase compared to the selling case.The median mileage rises from 120,000 km in the case of selling to 137,500 km and 165,000 km for battery leasing with random and smart distributions, respectively.Higher median mileages indicate lower lifespans and higher allocation factors (as shown in Fig. 10).
Batteries are replaced once they reach the lifetime mileage expectancy of 180,000 km (represented by the red lines in Fig. 9).In a 12-year period, the number of required replacements ranges from 2595 units for battery selling to 2464 units for leasing with a random distribution.Notably, no batteries exceed the lifetime mileage expectancy during the 12-year period in the case of leasing with a smart distribution.
The total number of batteries originally produced and replaced is shown in Fig. 11 for the different business models with and without allocation.When comparing values with and without allocation, flows without allocation are used for economic analysis, while flows with allocation are used for environmental analysis.In addition to the originally manufactured and replaced batteries, there are additional flows of batteries, such as replacements for failed or repaired batteries; detailed information is available in Figs. 17 and 18 of the Appendix.
Since some batteries can last longer than 12 years, they will have a smaller environmental impact than those lasting exactly 12 years.Therefore, the allocation factor described above is applied to the battery flows.The overall number of batteries used is similar across the three business cases, with 9533 units for battery selling, 9492 units for battery leasing with a random distribution, and 9633 for battery leasing with a smart distribution.
The similar number of batteries being used across different business models can be attributed to two opposing causes.On the one hand, leasing reduces the risk that some batteries will be excessively used by drivers with a high annual mileage, resulting in fewer replacements compared to battery selling.On the other hand, leasing also reduces extremely low mileages, which leads to a lower lifespan and higher allocation factor for this specific group of batteries compared to battery selling.These two factors counterbalance each other and ultimately result in a similar number of batteries being used across different business models.

Economic evaluation
A comparison of the NPV from the manufacturing company  perspective is shown in Fig. 12 for selling batteries (left) and leasing batteries with a random distribution (right).As discussed above, the models are set up such that the NPV in both cases is the same.In the case of selling batteries, a gross margin results from subtracting the costs of manufacturing from revenues.The NPV is then obtained after taxing this gross margin, which amounts to €6.4 million.In the case of leasing batteries, the leasing fees paid by users are estimated such that the NPV remains the same as that in the selling case.Retaining the ownership of batteries is important from a financial perspective, as it allows the depreciation of batteries.A company leasing batteries incurs two types of costs: i) the costs of manufacturing batteries and ii) the costs of maintaining and repairing batteries.This means that the overall costs in the leasing case are higher than those in the selling case.To obtain the same NPV as in the selling case, higher revenues and higher taxes are expected.In the case that the company leasing batteries is not the same as the company manufacturing them, which has not been analyzed here, even higher costs could be expected.
The uncertainty in the estimation of the NPV and the leasing fees are shown in Fig. 13.The NPV is estimated to be €6.4 million with a confidence interval of ±5.5% at 90% certainty.The leasing fees are estimated to be €0.0585/km± 10% for a random distribution and €0.0560/km ± 11% for a smart distribution at 90% certainty.These results indicate that leasing with a smart distribution is always more cost-effective than leasing with a random distribution.
A comparison of the NPV (€/km) from the customer perspective and under different business models is shown in Fig. 14 as a function of the annual mileage.The thick lines represent the mean values, and the thin lines represent the confidence interval at 90% certainty.While the NPV for the leasing cases remains constant regardless of the annual mileage, for the selling cases, it is inversely proportional to the mileage.This result means that if a customer buys a battery, the more the battery is used, the lower the NPV per lifetime mileage.Buying the battery is the most cost-effective option for customers driving more than 10,000 km/ year.On the other hand, leasing the battery is the best option for customers driving less than 10,000 km/year.These results indicate that a manufacturing company that aims to lease batteries and keep the same profitability as when selling them will translate its increased costs into leasing fees that are not always beneficial for customers.

Environmental evaluation
The environmental impacts at the battery level (i.e., per kilogram of battery) in the PEFCR and the "alternative" cases are shown in Fig. 15.In the PEFCR case, climate change, fossil fuel resources, and mineral and battery resources account for 18.9 kg CO 2 eq./kg battery, 315.8 MJ/kg battery, and 242.6 kg CO 2 eq./kg battery, respectively.Raw material and manufacturing contribute to most of the climate change and fossil fuel use impacts and the entirety of the mineral and metals use impacts in the PEFCR case.In the "alternative" scenario, the energy use is nearly three times higher than that in the PEFCR case.This results in climate change and fossil fuel impacts that are 50% higher than those in the PEFCR case per kg of battery.The increase in energy use in the "alternative" scenario implies only a marginal increase in the use of minerals and metals, as they are not heavily used in the use stage.
The environmental impacts at the fleet level considering the FU of 1715.97 million kilometers traveled by all users throughout the 12-year period are shown in Fig. 16.The results show that the environmental impacts of leasing batteries are nearly the same as those of selling batteries in all categories and cases.The climate change impact of selling batteries ranges from 0.0239 kg CO 2 eq./km traveled in the PEFCR case to 0.0351 in the "alternative" case.The percentage differences between the two leasing cases and the selling case are ±1% in the PEFCR and the "alternative" cases, which can be considered negligible.The use of fossil fuels in the case of selling batteries ranges from 0.4 to 0.6 MJ/km traveled in the PEFCR and "alternative" cases, respectively.No substantial differences are expected here in the leasing cases.Finally, the use of minerals and metals in the case of selling batteries ranges between 0.309 and 0.312 kg Sb eq./km traveled.Here, again, the differences between leasing and selling are marginal.
Importantly, as the environmental impacts of the two leasing options are nearly identical, cost-effectiveness may be a crucial factor in motivating a potential selection.In fact, if one of the two options should be selected, a smart distribution may be preferred because it offers the same environmental impacts as a random distribution but with substantially lower leasing fees.

Discussion
In this section, first, the main results of this investigation are presented.Second, the associated policy implications are discussed.Third, the limitations of this study are acknowledged, and future recommendations are made.

Discussion of the results
Compared to selling batteries, leasing batteries has significant economic and environmental impacts.From the manufacturer perspective, leasing batteries can be as equally profitable as selling them, as both options can yield a positive NPV.However, companies that lease batteries will incur higher expenses, including production costs, battery maintenance, and repair charges, than companies that sell batteries.
To obtain the same NPV, companies that lease may need to generate higher revenues and pay more taxes than those that sell.This means they will have to charge leasing fees to customers, which may not always be economically beneficial for customers.In fact, buying the battery is the most cost-effective choice for customers who drive over 10,000 km/ year, i.e., the majority.
These results contrast with those of a study conducted by (Nurhadi, et al., 2017), which predicted similar trends for leasing and buying batteries between 2000 and 15,000 km/year in Sweden.They also differ from the results of a study on the total cost of ownership in Germany published in 2017 b y (Letmathe and Suares, 2017), which suggested that leasing could be more economical for an annual mileage of 10,000 km/year.One potential reason for this discrepancy lies in the assumed price of the battery in the German study (€373/kWh, adjusted for inflation), which was more than double the price used in this analysis (€161/kWh).The economic advantage of buying batteries is expected to continue in the future as battery prices continue to decline.
On the environmental front, the findings suggest that leasing batteries offers no substantial environmental benefits compared to selling them.The environmental impacts of leasing and selling batteries are essentially identical, with differences expected to be below 1%.Despite the advantages mentioned in the literature, such as improved lifetime management, the promotion of repurposing and recycling, and centralized maintenance, the results suggest that leasing batteries does not significantly reduce environmental impacts.This finding aligns with the results of recent environmental studies on leasing EVs conducted by (Koide, et al., 2022;Amasawa et al., 2020).While these studies focus on vehicle leasing rather than battery leasing, they are relevant because batteries account for 80% of the life cycle environmental impacts of EVs (Picatoste, et al., 2022).Both studies concluded that leasing EVs offers negligible environmental benefits compared to buying or sharing them.

Policy implications
The findings of this study may provide insights to policymakers in the field BEV business models.The policy implications include the following.Policymakers may consider developing awareness campaigns or providing information to potential EV users to help them make informed decisions.• Further R&D: The results of this study highlight the need for continued R&D of battery technologies, pricing models, and environmental impact assessments.Policymakers may allocate resources to support research efforts that aim to improve the economic and environmental performance of battery leasing models or explore alternative business models that could offer more benefits.

Limitations of this study and future recommendations
It is important to acknowledge the limitations of this study.First, it focuses on a specific type of battery (NMC) for the small vehicle segment under driving profiles and conditions specific to Germany.By utilizing  fleet models to evaluate the mileage and lifespan of batteries, this study makes an improvement over previous approaches that relied on a single "average" driving profile.However, this approach involves making assumptions about modeling options and inputs.While the fleet model mainly concentrated on the impact of users' mileage on battery lifetime, important factors such as driving patterns, DoD, cycling frequency, and calendar aging were not considered.Additionally, this study did not account for the aging, maintenance, and retirement of vehicles themselves, excluding the battery.Therefore, a notable limitation is the lack of real-world data to calibrate the fleet models.
On the economic side, this study relied on publicly available information on current and future battery prices and costs.However, there is considerable uncertainty associated with these economic parameters.On the environmental side, a significant limitation is that primary data were not used for estimations.Instead, the analysis primarily relied on secondary data from other authors and simulation results, introducing uncertainties.Two uncertainties arise: the lifetime energy consumption, which could range from 8000 to 25,000 kW h/kg-battery, and the environmental impact of battery repairs, for which limited information is available.
These limitations highlight two important points.First, the results contain inherent uncertainties that cannot be effectively reduced without access to real-world data.Second, the findings cannot be easily generalized to other battery characteristics and regions.While economic data and proposed models could be adapted to other countries or regions (e.g., the European Union or the United States) without massive efforts, data for LCA could not be easily extrapolated for other chemistries or regions.
Consequently, further research is necessary to provide more conclusive arguments.The most critical recommendation for future studies is to develop and calibrate models using real-world data to address the uncertainties described above.However, obtaining such data may be challenging, as companies exploring novel business models may be hesitant to share critical information.Additionally, future studies should delve into the impact of user decisions, including driving patterns, cycling frequency, and DoD, on the battery lifespan and the influence of user decisions on overall fleet behavior.To enhance the breadth and validity of the results of this study across different conditions and contexts, it is crucial to evaluate various battery chemistries, vehicle types, and countries or regions, particularly from an environmental perspective.

Conclusions
This paper investigates the economic and environmental impacts of battery leasing for EVs as a CBM, and it compares these impacts to those of the linear model of selling-buying batteries.Unlike previous research that focused on either economic or environmental aspects or that focused solely on vehicle leasing, this analysis considers both aspects of battery leasing.
To provide a deeper understanding of the subject, instead of relying on the single "average" profile used in previous studies, a dedicated battery fleet model that incorporates various driving profiles is built.This detailed approach enhances comprehension of how mileage evolves over time and its implications for cost structures and cost-effectiveness.Compared to using a single "average" profile, modeling a large number of customers with different driving profiles better reflects real-world scenarios.This realistic representation is essential for assessing the life cycle impacts of business models, as such representation depends on customer preferences and decisions regarding driving distances.
The findings demonstrate that leasing batteries can be equally as profitable as selling them, with a positive NPV of €6.4 million with a confidence interval of ±5.5% at 90% certainty for a fleet of 10,000 BEVs.However, companies that lease batteries may require higher revenues and face increased tax obligations to obtain the same NPV as selling.To generate additional income, companies will need to charge customers fees, which may not always be economically beneficial for them.Interestingly, purchasing the battery proves to be the most costeffective option for customers driving more than 10,000 km/year.From an environmental perspective, the LCA results indicate that leasing batteries offers no substantial environmental benefits compared to selling them.In a broader context, this research reveals that leasing batteries effectively reduces the need for battery replacements.However, there is no evidence to suggest that leasing batteries significantly extends their lifespan, which is a crucial factor that strongly influences their environmental impacts.These findings suggest that future research should focus on exploring other CBMs that have the potential to increase the lifetime of batteries (e.g., through repurposing) or reduce the demand for raw materials (e.g., recycling) while simultaneously delivering added value to customers.These insights hold significance not only for battery manufacturers and EV investors but also for policymakers aiming to achieve climate targets.Moreover, the proposed methods may be relevant for other business models targeting a substantial user base in the BEV domain, such as BaaS programs and battery swapping.
It is important to acknowledge the limitations of this study, which concentrated on NMC batteries for vehicles in the small vehicle segment and conducted analysis based on driving profiles and conditions specific to Germany.Therefore, further research considering different battery chemistries and vehicle types and other countries or regions is necessary to provide more conclusive arguments.Nonetheless, as battery prices continue to decline, it is reasonable to expect the economic advantage of purchasing batteries over leasing them to persist in the future.Such a scenario will benefit customers not only by providing the perception of more affordable products but also by offering enhanced quality and longer warranty periods.However, such a scenario will also pose challenges on the environmental side, as the responsibility to dispose of batteries will primarily lie with consumers.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Table 6
Assumptions for quantifying the uncertainty in the NPV in the reference scenario.The limits are taken from (Mauler, et al., 2021) for the reference or best cost estimates of two studies (Nykvist, et al., 2019;He et al., 2020).The base value is the average.Cost of NMC batteries (€/kWh) Uniform 125.0 136.0 130.5 The limits are taken from (Mauler, et al., 2021) for the reference or best cost estimates of two studies (Schmidt, et al., 2019;Hsieh et al., 2019).The base value is the average.The limits are taken from the values of the "minimum" and "maximum" scenarios of studies (Nykvist, et al., 2019;He et al., 2020).Exponential decay constant for the cost Uniform 0.046 0.090 0.049 The limits are taken from the values of the "minimum" and "maximum" scenarios of studies (Schmidt, et al., 2019;Hsieh et al., 2019).Discount rate for companies (%) Uniform 5 10 10 The discount rates vary for energy-related technologies between 5% and 10% (Timilsina, 2021).

Table 7
Correlation between variables for uncertainty propagation.

Fig. 4 .
Fig. 4. Strategies to manage the lifetime of batteries.

Fig. 5 .
Fig. 5. Monte Carlo simulation to evaluate the mileage of battery leasing with a random distribution.

Fig. 8 .
Fig.8.Battery lifespan (years) and allocation factor as functions of mileage after 12 years of operation.

Fig. 9 .
Fig. 9. Distribution of the cumulative mileage for the battery fleet over the years disaggregated by business model.The red lines represent replacements, i.e., batteries that exceed the lifetime mileage expectancy of 180,000 km.Whiskers show quartiles.(For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 11 .
Fig. 11.Number of batteries originally produced and replaced under different business models with and without allocation.

Fig. 12 .
Fig. 12. NPV in million euro in the business case of selling batteries (left) and leasing batteries with a random distribution (right).

Fig. 13 .
Fig. 13.Uncertainty in the NPV from the manufacturing company perspective (left).Uncertainty in the leasing fees in €/km under different distribution strategies (right).

M
.Gonzalez-Salazar et al.   • Investment incentives: Policymakers may choose to prioritize incentives aimed to improve the quality and lifespan of batteries and reduce the cost for users.Incentives could include funding R&D and creating favorable conditions for battery manufacturers and associated industries.• Incentivizing sustainable practices: As this study implies that the traditional model of selling batteries could be more beneficial than the business model of leasing batteries, the responsibility for battery disposal could primarily lie with consumers.Policymakers could incentivize and educate consumers to properly dispose of batteries, and they could also prioritize the development of infrastructure to support battery recycling and disposal.Doing so could include establishing battery recycling centers and implementing appropriate regulations for the responsible disposal of batteries.• Consumer awareness and education: The findings of this study highlight the importance of educating consumers about the economic and environmental impacts of different business models.

Fig. 14 .
Fig.14.NPV per lifetime mileage (€/km) from the customer perspective and under different business models.Thin lines represent a confidence interval of 90% probability.Note: the values for leasing (random) are very close to those for leasing (smart) and, for better visualization, are not shown.

Fig. 15 .
Fig. 15.Environmental impacts for a single battery in the PEFCR and "alternative" scenarios.Climate change (left) in kg of CO 2 equivalent per kg of battery.Resource use and fossil fuels (middle) in MJ per kg of battery.Resource use, minerals, and metals (right) in mg of antimony equivalent per kg of battery.

Fig. 17 .
Fig. 17.Inventories of batteries in different business models with and without allocation.

Fig. 18 .
Fig. 18.Number of batteries manufactured, replaced, and used in different business models with and without allocation.

Table 1
Overview of studies investigating circular business models in the context of battery electric vehicles.

Table 2
Examples of battery leasing initiatives.

Table 4
Overview of assumptions for evaluating the NPV for the manufacturing company.
The limits and base values are taken from Table5.