Analysis and Modeling of Value Creation Opportunities and Governing Factors for Electric Vehicle Proliferation

Abstract

This research presents a comprehensive analysis of electric vehicle (EV) proliferation factors and various monetary and non-monetary value streams emerging in the EV domain. A comprehensive mathematical model is implemented to study EV proliferation and the resulting market share applicable to any geography and jurisdictional regime. Further, a novel framework is presented to analyze the interdependency between EV proliferation factors and value streams. The proposed model and framework can be leveraged to quantifiably evaluate the timeline available for grid operators to accommodate EV growth while utilizing those as Distributed Energy Resources (DERs) to improve grid reliability, commercial value, and environmental benefits. Compared to the previous studies, the analysis indicated that if all the factors which impact EV proliferation are addressed simultaneously, EV market share can surpass the internal combustion engine vehicle (ICV) in as quickly as 15–20 years. The study also highlighted the importance of policy making around EVs, which can offset EV market share by up to 10% between two countries following similar sustainability goals. Therefore, the study also helps aid decision making around policies and technology investments by public and private sector organizations in the space of EV.

1. Introduction

Electric Vehicles (EVs) have been gaining traction due to their multiple environmental and societal benefits. One of the most important benefits is the higher energy efficiency (3 to 5 times) over their traditional counterparts, the Internal Combustion Engine Vehicles (ICVs) [1]. Due to zero-emission, battery EVs (BEVs) produce no tailpipe emissions, and in general, plug-in hybrid EVs (PHEVs), which partially run by electricity, also generate lesser tailpipe emissions than ICVs [2]. EVs can help significantly reduce overall greenhouse gas (GHG) emissions when combined with the increase in low-carbon electricity generation. Lately, EV charging stations or supply equipments (EVSEs) are being deployed worldwide, supporting the case for EVs across multiple transit modes such as shared transportation (buses, taxis) [3,4], light-motor vehicles (cars, two/three-wheelers) [1], human-operated vehicles (e-Rickshaws) [5], as well as heavy-duty vehicles for short-range urban deliveries [1,6]. Furthermore, EV manufacturers worldwide have increased their investments, resulting in various EV models offered today, offering broader choices to consumers across various segments. Notwithstanding, favorable and effective policies are crucial to faster EV proliferation by lowering the upfront investment cost gap, promoting charging infrastructure, and ensuring a smooth integration of EV charging demands into power systems and the overall grid [1].

Overall, the proliferation of Distributed Energy Resources (DERs), such as solar photovoltaic (PV), battery storage, and EVs, is an emerging challenge to the power grid. However, unlike solar PV, EVs cause more significant challenges [7] due to being mobile and mainly creating clustered density of load at public charging, workplace charging, and multi-unit residential charging facilities [8]. Moreover, as commercial EV fleets emerge, the charging facilities will likely be concentrated at seaports, airports, and other retail and wholesale transportation distribution centers, further increasing the problems emerging due to load density. Recent trends indicate an increased focus on installing fast or higher-speed charging stations to promote EV proliferation, which may pose an increasingly severe threat to grid stability if not planned carefully [1,9]. Accelerated EV deployment may soon exceed the capacity of existing electrical infrastructure, which has already started showing early signs of weaknesses [7,10]. A typical infrastructure upgrade or new capacity deployment takes many months to a few years due to the time-consuming processes involving planning, regulatory approvals, project commissioning, and execution. Therefore, consideration of opportunities to mitigate costs and incorporate load management strategies to minimize on-peak charging is the need of the hour.

While the rapid proliferation of EVs challenges the grid’s reliability, it allows for a bottom-up approach where EVs enable new energy services to allow consumers to buy and sell energy. Connected devices now enable communication and control, which was impossible in the past. To some extent, advancements in technology, including two-way near real-time communication and the solid-state power conversion electronics of EV supply equipment (EVSE) or EV chargers, are together helping mitigate the potential impact of new loads and provide valuable grid services. Utilities have already started to leverage this as an opportunity to support essential grid services such as congestion management and voltage regulation [1,11].

Utilities have already started to leverage EVs to support essential grid services such as congestion management and voltage regulation [1,11]. Managed EV charging, vehicle-to-grid (V2G), and vehicle-to-home (V2H) services have started helping solve issues presented by EVs at the distribution network level. At the same time, EV charging facilities have started installing solar photovoltaic and energy storage systems, further enhancing their capability to support grid operators [2]. In addition, market operators, utilities, and aggregators have started creating new value streams leveraging the consolidated capacities of controllable and dispatchable EV resources. A value stream can be an initiative generating monetary, non-monetary, or both types of value to one or more beneficiaries, such as system operators, utilities, aggregators, end-consumers, and the environment.

This paper proposes a dynamic model to study EV growth across different jurisdictions and how EVs could be leveraged to create new value streams for different beneficiaries. The main contributions of this paper are as follows:

• Identified factors impacting EV proliferation and an approach to analyzing those factors for EV relative to ICV through a critical review of existing studies (Section 2), followed by the review of EV initiatives that helped create a list of value streams (Section 3).
• Proposed a dynamic model (Section 4) to analyze EV proliferation across any jurisdictional boundary influenced by the applicable regulatory regime, technological and economic factors, and situations such as a pandemic.
• Analyzed and presented (Section 5) the impact of the following on EV proliferation: EV technology advancement timelines, situations such as the COVID-19 pandemic, and all other factors considered as part of the model (Section 4) development.
• Defined a novel framework (Section 5.4) to analyze the impact of EV proliferation factors on value streams and help market operators, grid operators, and aggregators secure timelines for tackling EV proliferation challenges and convert them to opportunities.
• Defined a reusable approach to examine the impact on EV proliferation due to policy changes across different jurisdictional regimes and countries (Section 5.5).

The remainder of this paper is structured as follows. Section 2 reviews the literature regarding factors influencing EV proliferation, while Section 3 studies associated value streams. A dynamic model to study EV proliferation relative to ICVs is presented in Section 4. Section 5 presents a novel framework as scenario simulations of the proposed model to assess the interdependencies of value streams and EV proliferation factors along with probable EV growth patterns under multiple scenarios. The final section derives conclusions from the simulation results, their practical implications, and the overall literature contribution to future research directions.

2. Factors Impacting EV Proliferation

Several studies have analyzed the factors behind EV proliferation and have also forecasted growth patterns based on those factors. Based on the studies, public and private sectors industry initiatives, as well as regulatory directions, each factor can be classified under different categories [12,13,14,15,16,17,18,19,20,21]. The first category covers technological factors accounting to those which are related to battery, EV raw material, or overall EV advancements. The second category covers jurisdictional policies typically directed toward supporting environmental and sustainability goals. On the other hand, economic factors are direct drivers of EV purchase decision making, including the purchase cost, purchase tax, and relative operational cost of EVs versus ICVs. Lastly, some new factors have recently emerged, such as due to the COVID-19 pandemic, and are covered in the others category. The following sections describe each category and associated factors.

2.1. Technological Factors

Among different technological factors impacting a vehicle buyer’s decision to purchase an EV, the availability of EV charging infrastructure, predominantly driven by public charging stations across most geographies, is the most prominent influencer [12,16,18]. In countries such as India, with currently small EV charging networks, it is acting as a significantly adverse influencer, and hence the government is investing in deploying more EVSEs [22]. Despite the government’s efforts towards defining EV favorable policies, Ref. [23] highlighted a need for more aggressive push-through policy mandates by reducing taxes on EVs (BEV, PHEV) while simultaneously imposing higher taxes on ICVs. Furthermore, EV proliferation in the Indian market faces unique challenges, such as a limited local supply chain of powertrain and battery pack assembly leading to high EV and EVSE prices, as well as high real-estate acquisition price to install EVSE, resulting in insufficient EVSE availability as discussed by [24]. On the contrary, EV infrastructure has been growing for several years in the USA and China at a tremendous pace [9,12,22]. The number of EVSEs in the USA has grown from 75,000 to 100,000 between 2020 and 2021, while in China, those numbers are 210,000 today [25]. However, the number of gasoline stations in the USA has only grown from 111,100 in 2016 [26] to 115,000 in 2020 [27]. These trends highlight strong support for EV proliferation in some geographies while discouraging using ICVs.

Lately, the availability of fast-charging technology has been gaining traction. As per a recent report from the National Renewable Energy Laboratory (NREL), about 12.0% and 49.6% of the Level 2 and DC fast EVSE, respectively, required to meet projected demand in 2030, have been installed as of Q1 2020 in USA [9]. Moreover, there are 13,627 public and workplace DC fast-charging EVSEs and 71,975 public and workplace Level-2 EVSEs available in the United States. However, a similar EV fast charger proliferation level is required to help EVs penetrate globally. In addition to fast chargers, improvements in battery technology have collectively contributed to a significant reduction in time to charge an EV fully. Recently, StoreDot [28], a battery manufacturing firm, has released mass production-ready batteries that can fully charge in around five minutes and has successfully demonstrated it for smaller batteries in phones, drones, and electric scooters [29]. The firm highlighted that using silicon in place of graphite for battery electrodes has primarily led to this development and helped bring the cost closer to existing Li-Ion batteries, further helping obtain investments from major automotive manufacturers globally.

EV infrastructure interoperability challenges play a significant role when an EV buyer cannot charge the vehicle using an EVSE from another make when needed. Some EVSE manufacturers focus on interoperability [30,31] while some [32] do not, consequently impacting EV growth. According to a recent Natural Resources Defense Council report, the EV charging infrastructure interoperability is one of the most critical factors towards EV proliferation and is increasingly more significant in large markets such as the USA, China, and India [22]. Another critical factor influencing EV buyer’s decisions is EV driving range on a full charge [1,12]. While EV manufacturers are working on improving the driving range, Ref. [33] has provided a unique model for predicting EV driving range under the influence of factors such as days, temperature, and the depth of discharge (DOD) of a battery pack.

EV proliferation is continuously being impacted by the recent incidents of safety concerns around EVs [34], primarily due to batteries catching fire caused by overheating or poor health. Notable recent efforts to help address these issues include identifying determinants impacting battery health [35] and the creation of tools to measure battery health over time [36]. While recent advancements in battery technology have minimized these risks, it is vital to address EV safety issues in totality to sustain EV growth. Recent efforts by Tesla and Volkswagen have proven to be successful in recent years [37,38,39]. EVs appeared as catalysts in the overall industrial development due to them being the potential enabler of cost reduction in battery and copper-based technologies [2]. Battery cost, typically accounting for up to 30% of EV cost [40], has fallen 87% since 2010 and is expected to drop another 60% by 2030 [41]. Furthermore, demand for copper is forecasted to rise nine-fold by 2027, given that it is the second most crucial component constituting the majority of equipment costs in EVs and EVSEs [42]. Consequently, similar to batteries, copper prices need to drop in coming years, as highlighted by global auto manufacturers [43].

2.2. Jurisdictional Policies to Support Environmental Goals

Jurisdictional directions significantly impact a vehicle buyer’s decision to purchase an EV. In the last few years, governments across the globe have started intervening in setting the targets for GHG emissions, forcing vehicle manufacturers to shift their strategy towards making more EVs over ICVs, and avoid fines with reputational setbacks [1,22,44]. One such example is the European Union’s regulation on CO₂ emission performance standards for new passenger cars and vans [45], which could bring penalties for up to 50% of vehicle manufacturers [12]. GHG emission targets across jurisdictions could vary, where some have been more stringent over others, as highlighted by online resources such as the climate action tracker [46]. In the USA, the Center for Climate and Energy Solutions [47] provides a granular view of GHG emissions across states to help establish GHG mandates.

Another recent trend is imposing fines, punitive taxes, or complete bans on older ICVs to address air pollution concerns [48]. Similarly, some jurisdictions have started providing EV-related privileges (EVP) and exemptions (EVE) to encourage EV adoption. Examples of privileges may include having EV dedicated lanes and allowing high occupancy vehicle (HOV) lanes for EVs with one occupant [49,50], unlike the rules in general for vehicles to carry at least two or more passengers. Ref. [51] highlighted the benefits of exemptions to promote customer adoption of EVs. Furthermore, introducing self-driving features is increasingly encouraging EV adoption, where emerging ideas suggest dedicated lanes for autonomous cars, especially at toll booths [52]. Recent studies point towards this direction where [53] studied the impact of dedicated lanes for autonomous vehicles on traffic flow throughput, while [54] presented a conceptual framework to design and operate dedicated lanes for connected and automated vehicles on motorways. As of November 2020, at least 45 states in the USA, along with the District of Columbia, offer incentives to support the deployment of EVs or alternative fuel vehicles and supporting infrastructure, either through state legislation or private utility incentives [55]. Legislative incentives include measures that provide HOV lane exemptions, financial incentives for purchasing EVs and EVSEs, vehicle inspections or emissions test exemptions, parking incentives, and utility rate reductions.

In countries such as Japan, the lower availability of EV models over ICVs is one of the unique factors that is negatively impacting EV growth [1]. While vehicle manufacturers globally have been investing in all types of EVs, including PHEVs and BEVs, Japan’s unique focus thus far towards self-charging hybrid EVs has slowed the growth of overall EV proliferation [56]. It is mainly because the country’s largest manufacturers have been left behind in the latest technological shift in the automobile sector while keeping their focus toward self-charging hybrid vehicles [57]. However, recently, Japan rolled out policies for vehicles and chargers to achieve its target of “next-generation vehicles” sales to account for 50–70% of the total vehicles in the country by 2030 [1,58].

2.3. Economic Factors and Related Policies

Multiple economic factors play an essential role in decision making while buying a vehicle [1,16,51]. Certain municipalities across the globe have launched programs to reduce costs for vehicle charging in public charging stations and, in some cases, eliminate parking charges for EVs [1,55]. Financial incentives in the form of cash subsidies and reduced insurance on EVs have encouraged EV adoption, while certain jurisdictions have maintained or increased taxes on ICVs to encourage buyers to own an EV. Ref. [51] demonstrated the importance of vehicle purchase tax and carbon tax for EV adoption in the short and long terms. Most people look for an overall lower cost of ownership of vehicles, which includes one-time purchase cost, tax, and the operational cost of the vehicle, covering cost/km(or mile) and maintenance cost over the vehicle lifespan. Ref. [16] proposed a dynamic model of EV adoption, which helps calculate the overall life-cycle costs of EVs and ICVs as a result of operational and one-time costs for each.

Multiple studies have assessed the impact of policies on EV market share where a unique model developed by [59] predicts PHEV market share under alternative policy settings. Different subsidy options are analyzed for high- versus low-income consumers to examine their impact and help identify strategies to maximize EV proliferation. Further, the authors, through a different study [60], argued that PHEV can be made more economical by employing focused incentives based on factors such as household income, vehicle disposal, geography, and vehicle travel usage. In addition, they challenged the existing policies, such as in the USA, which offer more significant subsidies for PHEVs with larger batteries, and argued that offering similar incentives regardless of the battery size could result in higher EV proliferation. Ref. [61] adopted a more practical approach to conduct an econometric study of purchase incentives by analyzing actual data on PEV sales from 32 European countries between 2010 and 2017. Their study concluded that factors such as household incomes, fuel prices, and supporting financial incentives impact PEV sales and thus can facilitate their diffusion. A similar study conducted by [62] examined the impact of purchase incentive policies on EV proliferation and found that up to 35% of electric vehicle sales could be attributed to the purchase incentives.

2.4. Other Factors

EV transition is expected to be slower in nations with lower per capita income where high population and cultural differences regarding mobility models could be additional factors impacting transition to EVs [12,22]. For example, India, dominated by mass and low-cost mobility models, is a region that EV manufacturers have not penetrated so far because of comparatively higher EV prices than ICVs. However, fleet owners globally, including the developing nations, have been investing in buying EVs due to lower operating costs and leveraging government incentives [1,63]. Fleet sales represent a significant proportion of all cars sold globally and are an essential driver for overall EV sales [13].

In light of COVID-19, investments in fleets were initially stalled for a few months as corporates reduced their expenditure and prioritized other investments [1]. However, fleet investments started rising again as the world is approaching normal, and the forecasts show it in favor of EVs [64,65]. Moreover, trends show an overall decline in sales in the automotive industry, where, in some geographies, close to 50% of prospective vehicle buyers now plan to keep their existing vehicles for longer than initially intended [1]. However, the same reports highlighted comparatively less impact on EV sales than on overall automotive sales. In addition, since the COVID-19 pandemic has hit the world, people’s driving behaviors have changed globally [66]. A significant population requires short-distance trips to a grocery store, taking away the current concerns for EV owners, range anxiety, and long wait times to charge their vehicle [12]. Moreover, people prefer charging their vehicles at home versus going to a gas station to refill the vehicle due to pandemic-related safety concerns. Ref. [67] studied in detail and provided evidence of significantly different post-pandemic travel behavior compared to pre-pandemic. Overall, these factors collectively influence a vehicle buyer’s decision to favor EVs over ICVs.

Lately, transactive energy initiatives have been gaining widespread adoption, both due to jurisdictional mandates as well as public–private partnerships [68]. Multiple pilot projects [69,70,71] have demonstrated transactive energy systems leveraging EVs could generate monetary value by generating revenue and infrastructure deferral. Simultaneously, they can provide non-monetary benefits such as increasing grid reliability and environmental and social benefits. Simultaneously, multiple academic studies [72,73,74] understood the aspects of transactive energy and proposed models to understand and demonstrate their role in supporting electricity networks. Ref. [75] highlighted the importance of EVs’ transaction behavior and their interactions with buildings in establishing a sustainable transactive energy community from physical energy space, data cyberspace, and human social space perspectives. Others have proposed models to leverage EVs for energy participation by adopting transactive energy concepts [76,77]. Overall, these approaches have successfully demonstrated the importance of DERs, specifically EVs, in supporting the grid and generating new value opportunities and are discussed in the upcoming sections.

2.5. Related Work

The previous sections highlighted existing work which identified and, in some cases, examined the factors which impact EV proliferation to some degree. For example, to calculate the vehicle purchase price, most of the studies highlighted it to be low, whereas [16] went a step further to highlight the importance of lower purchase tax especially. In another example of obtaining the operational cost of an EV, Refs. [14,17] focused only on EV Efficiency while [15,16] also considered vehicle maintenance and electricity price as the dominant factors. While each of these studies presented a unique approach to examining these factors, they all lack the breadth of analysis that could impact the EV market share. For instance, Ref. [13] specifically highlighted charging time as the most critical factor for improvements around battery technology. This stems from the fact that new factors, such as due to COVID-19 impacts, and transactive energy initiatives, emerge over time while some existing ones lose their impact due to technological advancements, policy changes, and other jurisdiction-specific impacts.

This paper provides a more comprehensive approach than previous related work and considers all the following factors collectively:

• Number of EVSE (N_EVSE), which includes EV charging infrastructure and also studying it against the number of gas stations.
• Improvement in battery technology (IBT), including lower charging time and battery density.
• EV raw material cost (CRM) to consider price drop over time.
• EV driving range (EVDR).
• Safety concerns related to EVs (SCEV).
• EV charging infrastructure interoperability (CII) across different EV models.
• Purchase cost (PC) for the vehicle, which includes purchase price, tax, related subsidies, and economic exemptions.
• Operational cost (OC) for the vehicle, which includes electricity and gas prices and their increase over time, EV and ICV efficiencies, average vehicle life and yearly mileage.
• EV-related privileges (EVP) such as dedicated lanes, HOV access, and others.
• Exemptions for EV (EVE).
• Available EV models (AEVM).
• GHG emissions targets (ET) as directed by regulators to motivate automaker investments in EVs and vehicle buyers’ purchase decisions towards an EV.
• Short-distance trips (SDT) more common since the COVID-19 pandemic.
• Refueling convenience (RC), which comes along with EV also charging at home versus ICV only at public gas stations.
• EV usage in vehicle fleets to become mainstream (FM).
• Transactive energy initiatives (TEI) increasingly leveraging EVs, making them more mainstream.

Furthermore, the following

Table 1. Related work which identified or analyzed EV proliferation factors.

Factors	[12]	[13]	[14]	[15]	[16]	[17]	[18]	[19]	[20]	[21]	Proposed Study
NEVSE		✓	✓	✓	✓		✓		✓		✓
IBT		✓							✓		✓
CRM								✓	✓		✓
EVDR			✓		✓	✓		✓	✓		✓
SCEV					✓						✓
CII							✓				✓
PC	✓		✓	✓	✓	✓	✓	✓	✓	✓	✓
OC			✓	✓	✓	✓	✓	✓	✓	✓	✓
EVP	✓			✓			✓				✓
EVE				✓			✓				✓
AEVM	✓				✓						✓
ET	✓		✓				✓			✓	✓
SDT						✓					✓
RC											✓
FM							✓				✓
TEI										✓	✓

summarizes the existing literature that identified different factors and, in some cases, provided the approach to studying EV proliferation through those factors.

3. Value Streams within the EV Domain

Recent research studies [78,79,80,81], pilot initiatives [69,70,71], and industry solutions [32,82] have demonstrated the capabilities of EV chargers. They have also highlighted the exponential increase in EV proliferation and indicated that the value obtained from the EV-based ecosystem solutions would increase at a similar or even higher rate. This section studies EV proliferation’s impact on the utility grid and value streams arising in this domain. It further examines the EV proliferation from an economic perspective where multiple ecosystem solutions, such as in the space of EV charging management, and parking lot energy management, could generate new monetary opportunities for existing and new businesses. In addition, large organizations relying on vehicle fleets for their day-to-day operations could increasingly benefit from EVs due to their significantly lower operational cost than ICVs. Lastly, household consumers could save a lot of dollars on rising fuel prices by steering toward more economical EV commute and ownership options.

3.1. EV Charging Management (EVCM)

A rapid increase in EV adoption is leading to peak-demand problems for the grid [83,84], which are becoming more common in recent years, where the clustered density of EVs is further aggravating those issues [8]. Multiple studies [80,81] demonstrated the benefits of managed EV charging approaches to address peak-demand-related issues encountered at the transformer and feeder level. If not managed carefully, these issues at scale can increase grid congestion beyond the feeder level up to the substations, leading to network instability, infrastructure/equipment failures, and outages.

An additional strategy by utilities to manage grid congestion and loading issues is by encouraging DER owners to participate in time-of-use (TOU) pricing programs. DER owners could be incentivized for their active participation by allowing utilities to control their EV charging rate to promote participation in these programs. Pilot programs like the one by Colorado in partnership with Xcel Energy [69] and the other by Chicago with ComEd [70] are good examples of DER participation. Similarly, studies [71,78,79], are selected examples of EV energy participation in the last decade. TOU pricing programs offered by utilities to EV owners can provide significant monetary benefits in the long run by offering a suite of services such as registration of EV assets, operational and compliance management, and data analytics, thus encouraging participation by EV owners [69,70,71,78,79].

3.2. EV Fleet Management and Optimization (EVFMO)

A more recent group of beneficiaries in the EV space are the fleet operators managing EV charging for themselves, or on behalf of fleet owners, by optimizing charging schedules for the monetary, grid, and environmental benefits. EV charging schedule optimization can be performed in multiple ways, where the rudimentary strategy is to perform it without grid participation. It involves generating an EV charging schedule for the Fleet Operators (FOs) without coordination with the network or market operator. As described by [80,85], this approach focuses on maximizing FO profits by predicting the energy requirements of their fleet without considering load impacts on the grid.

While the above strategy does not help resolve grid congestion, a network/market operator-coordinated EV fleet management approach can do so. Refs. [80,85] have highlighted the benefits of this approach where it could be further implemented as a planned or an online coordinated system. A planned coordination approach involves generating charging schedules based on network information sharing at specific times of the day. It helps resolve congestion during peak hours and optimizes well between FO profits and minimizing distribution grid congestion. Despite having no real-time communication between the FOs and the distribution grid for congestion management, DSOs can override the charging schedule to mitigate network congestion. However, a more sophisticated online coordinated market-based approach involves scheduling in real-time with the network/market operators. The charging schedule is based on the network state at any given point in time. Moreover, the network operator can override the charging schedule to manage grid stability, which FOs can further leverage to improve their predictive modeling for scheduling optimization.

3.3. Vehicle to Grid, Home, and Vehicle (V2G/H/V)

Multiple research studies [86,87] have highlighted the benefits of using energy stored in an EV battery for participation in demand response, energy arbitrage, or vehicle-to-grid (V2G) services. Although this approach has not yet seen widespread adoption, it helps enhance the value obtained from an EV when it is not on-road but connected to a bi-directional EV charger. Similarly, Refs. [86,87] have highlighted the benefits of vehicle-to-home (V2H) techniques where energy stored in EV batteries can be used to fulfill load requirements within a residential or an industrial premise. However, similar to V2G, this approach can rapidly deteriorate EV battery lifespan due to multiple charge/discharge cycles throughout the day and is recommended only for emergencies. Leveraging the concept of V2G/H, peer-to-peer (P2P), or vehicle-to-vehicle (V2V) EV charging has recently been seen as an appealing business model while the associated infrastructure is maturing. Studies [85,88,89] have presented two-way energy trading scenarios where bi-directional EV charging infrastructure can significantly eliminate concerns about EV range.

Today, a typical bidirectional EV charger is significantly more expensive (more than six times) than its unidirectional counterparts [90]. However, there have been efforts by automakers such as Tesla [91], Volkswagen [92], and third parties such as Quasar [93] to make them cheaper and more affordable for large-scale adoption. On the contrary, current industry regulations not favoring the standardization of bi-directional EV charging stations are negatively impacting its proliferation and becoming a viable industry business model [92]. At the time of writing this literature, the growth opportunities in the space of V2G or V2H are limited. However, once bi-directional EV charging is widely adopted, additional opportunities in the space of P2P or V2V charging will likely achieve a more significant number of real-world applications.

3.4. Parking Lot Energy Management (PEM)

This approach involves real-time scheduling of EV charging stations where the parking-lot schedule operator (PSO) will work with DSOs who can override the charging schedule to manage grid stability. PSOs need to utilize real-time scheduling and revise charging schemes where the objective is to avoid energy imbalance while participating in the power market. Up-to-date information about network congestion helps PSOs conduct real-time risk analysis and define charging schedules. In a typical scenario, EVs charge according to the schedule. However, if the grid’s regular technical operation is compromised, PSO management can be overridden by the DSO operation, such as in a load-shedding scheme. Contrary to the online approach described by [80] for EV fleet charging optimization, an online approach for a parking lot is different, where it has little or no authority over the charging patterns of incoming vehicles [94], thus making it more challenging. Multiple techniques have been proposed in this space, where some focused on non-utility coordinated offline approaches [95,96,97,98], while others such as [94,96,99] devised a more sophisticated grid-coordinated approach.

3.5. Distribution Grid Monitoring for EV Loads (DGMEV)

This approach involves understanding EV load patterns within the distribution network to understand charging behavior for supporting the short and longer-term distribution network planning processes and conducting more accurate risk analysis. To implement the above approach, near real-time communication with the utility’s control room and tighter integration with the planning and risk analysis processes are required. In addition, as DER penetration is increasing, seamless integration of control room technologies with DER management solutions is required to obtain the distribution network’s state at any given time and precisely manage it too. Primarily, the focus thus far has been on understanding the EV charging behavior across studies [100,101,102] and reports [103,104]. To the author’s knowledge, no real-world implementations exist where utilities have integrated this approach into the distribution network at a large scale.

3.6. Research Gaps

It is important to highlight that, to the best of the authors’ knowledge, none of the existing work thus far provides an approach to support the following objectives:

• Identify a comprehensive approach to understanding all the factors which impact EV proliferation and consequently impact EV market share with respect to their traditional counterparts i.e., ICVs.
• Study and categorize different value streams in the EV domain.
• Provide an approach to analyze EV proliferation in the light of different EV value streams.
• Deliver a flexible approach to analyze EV proliferation to aid in decision making for investment in technology advancements or informing policy changes for a regulatory regime.

This paper provides a comprehensive approach to achieving the above-mentioned objectives. While Section 2 examined existing literature to identify the factors which impact EV proliferation, Section 3 presented a survey across research and industry initiatives to generate the list of value streams. The following Section 4 and Section 5 help achieve the remainder of the objectives by providing a flexible approach to analyze EV proliferation under different scenarios, which could help in decision making around investments and policy making for EVs.

4. Dynamic Modeling of EV Adoption

This section presents a model which helps study EV proliferation by leveraging all the contributing factors presented in Section 2. Each factor influences a vehicle buyer’s decision to choose an EV versus an ICV, impacting the total market share of EVs at any given time. For simplicity, the proposed model assumes the availability of only two types of vehicles in the market, i.e., ICV and EV, and ignores other types, such as fuel-cell vehicles. Moreover, the modeling of each contributing factor reflects one possible approach where it could be further refined to more accurately represent individual cases. Figure 1 presents a high-level view of the model, and Figure 2 represents its MATLAB implementation, highlighting EV proliferation factors, their collective impact on informing relative EV proliferation rate (REVPR), and consequently overall EV market share (EVMS) at any given point in time.

4.1. Calculating REVPR Using EV Proliferation Factors

For economic factors, a vehicle’s purchase cost (PC) is primarily a combination of the vehicle purchase price and taxes imposed at the time of purchase. Therefore, the purchase cost of a vehicle can be depicted as follows:

$$\mathrm{P}{\mathrm{C}}_{\mathrm{V}\mathrm{e}\mathrm{h}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}}=\mathrm{P}\mathrm{r}\mathrm{i}\mathrm{c}{\mathrm{e}}_{\mathrm{V}\mathrm{e}\mathrm{h}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}}+\mathrm{T}\mathrm{a}{\mathrm{x}}_{\mathrm{V}\mathrm{e}\mathrm{h}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}}$$

(1)

where Vehicle ∈ {EV, ICV}.

Similarly, a vehicle’s operational cost (OC) comprises mileage and maintenance costs. For an ICV, the mileage cost can be calculated using the cost of gasoline consumed and vehicle efficiency, whereas, for an EV, the electricity consumed towards charging the vehicle is considered along with vehicle efficiency. The following set of considerations are made to simplify the model:

• One vehicle per vehicle owner or buyer;
• A fixed vehicle mileage per year (VMPY);
• Same average vehicle life (AVL) for both ICV and EV;
• Constant efficiencies for ICV (Eff_ICV) and EV (Eff_EV);
• A fixed rate of increase for gasoline (RIGP) and electricity (RIEP) prices;
• A fixed maintenance cost per year for ICVs (CMPY_ICV) and EVs (CMPY_EV) respectively;
• Zero cost for accidental damages.

Therefore, considering AVL, VMPY, Gasoline Price (GP) per gallon, Electricity price (EP) per unit, RIGP, RIEP, Eff_ICV, and Eff_EV, the operational costs for an ICV and EV can be formulated as:

$$\begin{array}{c}\hfill \mathrm{O}{\mathrm{C}}_{\mathrm{I}\mathrm{C}\mathrm{V}}=\mathrm{A}\mathrm{V}\mathrm{L}*\mathrm{V}\mathrm{M}\mathrm{P}\mathrm{Y}*\mathrm{G}\mathrm{P}\left(\mathrm{t}\right)\ast (1+(\mathrm{R}\mathrm{I}\mathrm{G}\mathrm{P}*\mathrm{t}))*\mathrm{E}\mathrm{f}{\mathrm{f}}_{\mathrm{I}\mathrm{C}\mathrm{V}}\end{array}$$

(2)

$$\begin{array}{c}\hfill \mathrm{O}{\mathrm{C}}_{\mathrm{E}\mathrm{V}}=\mathrm{A}\mathrm{V}\mathrm{L}*\mathrm{V}\mathrm{M}\mathrm{P}\mathrm{Y}*\mathrm{E}\mathrm{P}\left(\mathrm{t}\right)\ast (1+(\mathrm{R}\mathrm{I}\mathrm{E}\mathrm{P}*\mathrm{t}))*\mathrm{E}\mathrm{f}{\mathrm{f}}_{\mathrm{E}\mathrm{V}}\end{array}$$

(3)

The total cost of ownership of a vehicle (TCO) over its lifetime can be calculated as

$$\begin{array}{c}\hfill \mathrm{T}\mathrm{C}{\mathrm{O}}_{\mathrm{V}\mathrm{e}\mathrm{h}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}}=\mathrm{P}{\mathrm{C}}_{\mathrm{V}\mathrm{e}\mathrm{h}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}}+\mathrm{O}{\mathrm{C}}_{\mathrm{V}\mathrm{e}\mathrm{h}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}}\end{array}$$

(4)

where the Vehicle could be either EV or ICV.

Therefore, the relative total cost of ownership for EV with respect to ICV (RTCO_EV) can be depicted as:

$$\begin{array}{c}\hfill \mathrm{R}\mathrm{T}\mathrm{C}{\mathrm{O}}_{\mathrm{E}\mathrm{V}}={\displaystyle \frac{\mathrm{T}\mathrm{O}{\mathrm{C}}_{\mathrm{E}\mathrm{V}}}{\mathrm{T}\mathrm{O}{\mathrm{C}}_{\mathrm{I}\mathrm{C}\mathrm{V}}}}\end{array}$$

(5)

Considering technological factors, the relative rate of EVSE growth compared to the number of gas stations can be represented by the following equation:

$$\mathrm{E}\mathrm{V}\mathrm{S}\mathrm{E}\left(\mathrm{t}\right)=({(1+R)}^t\ast N_{EVSE}\left(0\right))/N_{GS}\left(0\right)$$

(6)

where R = rate of increase for EV Charging Stations, t = projection years, N_EVSE(0) = number of EV charging stations at the beginning of the simulation, and N_GS(0) = the number of gas stations at the beginning of the simulation. This equation can be used to study relative EV infrastructure for any geography.

As discussed, EV Driving Range (EVDR) is expected to improve in the coming years and become insignificant in decision-making criteria towards buying an EV over ICV. Hence, it could be modeled for a linear rise from 0 to 1 in ST_DR (Saturation Time for Driving Range) number of years in a specific geography. Similarly, improvements in battery technology (IBT), which helps in reducing the time required to charge an EV fully, EV charging infrastructure interoperability (CII), availability of EV models over ICVs (AEVM), EV fleet to become mainstream (FM), and the number of Transactive Energy initiatives globally (TEI) are all expected to improve significantly in the coming years and consequently become insignificant in decision-making criteria towards buying an EV over ICV. Additionally, safety concerns with EV (SCEV) predominantly due to battery technology and costs for raw materials (CRM) such as battery and copper used in the motors are all expected to decrease significantly in the coming years and similarly become insignificant in decision-making criteria.

Therefore, all the above-mentioned factors influencing a vehicle buyer’s decision over time will collectively steer them toward buying an EV. Hence without loss of generality, their effect can be modeled as a linear rise from 0 to 1 in their respective saturation times (ST) measured in the number of years for a particular geography and represented as follows:

$$\begin{array}{c}\hfill \mathrm{X}\left(\mathrm{t}\right)=\left\{\begin{array}{c}1,t>S{T}_{\mathrm{X}}\hfill \\ t/S{T}_{\mathrm{X}},0<t<S{T}_{\mathrm{X}}\hfill \\ 0,t=0\hfill \end{array}\right.\end{array}$$

(7)

where X ∈ {EVDR, IBT, CII, AEVM, FM, TEI, SCEV, CRM}

On the other hand, automobile manufacturers have been striving to meet GHG emission targets (ET) as established by different jurisdictional mandates such as EU’s Climate action plan [45], Zero-Emission Vehicle Program in California, USA [105], and China’s New Energy Vehicle Credits regulation [106]. However, from an EV buyer’s perspective, it only influences the decision making of those who are willing to or can afford to pay higher prices in favor of EVs, especially when there is a cheaper ICV counterpart available today. This subset of the population’s size and percentage could also vary depending on the per-capita income across different geographies. On the other hand, EV-related privileges (EVP) and exemptions (EVE) discussed before are temporary measures that will only be available until EVs become mainstream and may not significantly influence decision making for a vehicle buyer after that period.

Due to COVID-19, people’s preferences have changed recently, as highlighted by some recent studies [12,107,108]. Many people now perform short-distance trips (SDT) due to workplace closures and remote-working arrangements. Due to pandemic-related safety reasons, refueling convenience (RC) has also become a new priority where; the people who own EVs increasingly prefer to charge at home versus going to public charging stations. To some degree, the safety concerns also promote EV proliferation as ICV owners do not have that choice and mostly have to go to a public gas station to refill their vehicles. If the pandemic sustains, ref. [109] suggests that these factors may increasingly influence the vehicle buyer’s decision going forward. Irrespective of whether the pandemic factors sustain or not, their impact will only influence a subset of the overall vehicle buyers, and hence they can be mathematically represented as:

$$\begin{array}{c}\hfill \mathrm{Y}\left(\mathrm{t}\right)={\displaystyle \frac{\%\hspace{0.17em}\mathrm{o}\mathrm{f}\hspace{0.17em}\mathrm{v}\mathrm{e}\mathrm{h}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}\hspace{0.17em}\mathrm{b}\mathrm{u}\mathrm{y}\mathrm{e}\mathrm{r}\mathrm{s}\hspace{0.17em}\mathrm{c}\mathrm{a}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g}\hspace{0.17em}\mathrm{a}\mathrm{b}\mathrm{o}\mathrm{u}\mathrm{t}\hspace{0.17em}\left(\mathrm{Y}\right)}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{n}\hspace{0.17em}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\hspace{0.17em}\mathrm{o}\mathrm{f}\hspace{0.17em}\mathrm{v}\mathrm{e}\mathrm{h}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}\hspace{0.17em}\mathrm{b}\mathrm{u}\mathrm{y}\mathrm{e}\mathrm{r}\mathrm{s}}}\end{array}$$

(8)

where Y ∈ {ET, EVP, EVE, SDT, RC}.

As highlighted in Figure 1, the above set of factors represented by Equations (5)–(8) together helps define the relative EV proliferation rate (REVPR) with respect to ICVs at any given point in time. REVPR is calculated as a weighted sum of these factors where each factor’s weight could be assigned on a scale of 0 to 1, representing 1 as high, 0.66 as a medium, 0.33 as low, and 0 as insignificant based on the geographical applicability of each factor. The weighted sum method is widely adopted and is applicable for scenarios where it is necessary to calculate a composite objective function that combines multiple objective functions into one scalar [110]. Therefore, the REVPR can be defined as the following:

$$\mathrm{R}\mathrm{E}\mathrm{V}\mathrm{P}\mathrm{R}\left(t\right)=w_1{f}_1\left(x\right)+w_2{f}_2\left(x\right)+....+w_m{f}_m\left(x\right)$$

(9)

where, f_i(x) represents individual factors impacting EV proliferation and w_i represents their corresponding weight, satisfying

$$\sum _{i=1}^M{w}_i=1,w_i\in (0,1)$$

(10)

4.2. Calculating EVMS Using REVPR

The REVPR defined using the above approach signifies a vehicle buyer’s decision making in choosing an EV over an ICV, where a higher value of REVPR represents a higher probability of choosing an EV. REVPR influences the decision making of both first-time vehicle buyers (FTB) and vehicle repurchasers (VR). Therefore, at any given point in time (t), the number of EV buyers (EVB) and ICV buyers (ICVB) can be represented as:

$$\begin{array}{c}\hfill \mathrm{E}\mathrm{V}\mathrm{B}\left(t\right)=\mathrm{R}\mathrm{E}\mathrm{V}\mathrm{P}\mathrm{R}\left(t\right)\ast \left(\mathrm{F}\mathrm{T}\mathrm{V}\mathrm{B}\right(t)+\mathrm{V}\mathrm{R}(t\left)\right)\end{array}$$

(11)

$$\begin{array}{c}\hfill \mathrm{I}\mathrm{C}\mathrm{V}\mathrm{B}\left(t\right)=(1-\mathrm{R}\mathrm{E}\mathrm{V}\mathrm{P}\mathrm{R}(t\left)\right)\ast \left(\mathrm{F}\mathrm{T}\mathrm{V}\mathrm{B}\right(t)+\mathrm{V}\mathrm{R}(t\left)\right)\end{array}$$

(12)

where FTVB can be calculated using the growth rate of new vehicle buyers per year (GRNVPY) over an initial set of vehicle buyers:

$$\begin{array}{c}\hfill \mathrm{F}\mathrm{T}\mathrm{V}\mathrm{B}\left(t\right)={\int }_0^t\mathrm{N}\mathrm{e}\mathrm{w}\hspace{0.17em}\mathrm{V}\mathrm{e}\mathrm{h}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}\hspace{0.17em}\mathrm{B}\mathrm{u}\mathrm{y}\mathrm{e}\mathrm{r}\mathrm{s}\left(0\right)\\ \hfill {\ast (1+\mathrm{G}\mathrm{r}\mathrm{o}\mathrm{w}\mathrm{t}\mathrm{h}\hspace{0.17em}\mathrm{R}\mathrm{a}\mathrm{t}\mathrm{e})}^t+\mathrm{F}\mathrm{T}\mathrm{V}\mathrm{B}\left(0\right)\end{array}$$

(13)

and VR can be considered as a combination of EV owners (EVO) and ICV owners (ICVO) whose vehicle’s average lifespan (AVL) is completed and are going to buy a new vehicle:

$$\begin{array}{c}\hfill \mathrm{V}\mathrm{R}\left(t\right)=(\mathrm{E}\mathrm{V}\mathrm{O}+\mathrm{I}\mathrm{C}\mathrm{V}\mathrm{O})\hspace{0.17em}\mathrm{w}\mathrm{h}\mathrm{e}\mathrm{r}\mathrm{e}\hspace{0.17em}\mathrm{A}\mathrm{V}\mathrm{L}\ge 10\hspace{0.17em}\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{s}\end{array}$$

(14)

Consequently, the total number of EVO and ICVO can be calculated using the following:

$$\begin{array}{c}\hfill \mathrm{E}\mathrm{V}\mathrm{O}\left(t\right)={\int }_0^t\mathrm{E}\mathrm{V}\mathrm{B}\left(t\right)dt+\mathrm{E}\mathrm{V}\mathrm{O}\left(0\right)\end{array}$$

(15)

Similarly, the total number of ICV owners (ICVO) can be calculated using the following:

$$\begin{array}{c}\hfill \mathrm{I}\mathrm{C}\mathrm{V}\mathrm{O}\left(t\right)={\int }_0^t\mathrm{I}\mathrm{C}\mathrm{V}\mathrm{B}\left(t\right)dt+\mathrm{I}\mathrm{C}\mathrm{V}\mathrm{O}\left(0\right)\end{array}$$

(16)

Finally, the EV market share (EVMS) can be obtained using the following:

$$\mathrm{E}\mathrm{V}\mathrm{M}\mathrm{S}\left(t\right)={\displaystyle \frac{\mathrm{E}\mathrm{V}\mathrm{O}\left(\mathrm{t}\right)}{\mathrm{E}\mathrm{V}\mathrm{O}\left(\mathrm{t}\right)+\mathrm{I}\mathrm{C}\mathrm{V}\mathrm{O}\left(\mathrm{t}\right)}}$$

(17)

4.3. Quantifying Value Streams Using REVPR

At a high level, the concept of value streams is qualitative. However, as it helps in EV-related investments and policy-making decisions, it is essential to quantify value streams for efficient decision making and course correction where needed. One way to quantify value stream is by assessing it against a “desired” EV proliferation that is required for a value stream to be considered “effectively” realized based on a policy or business decision. This paper employs binary classification to quantify the “effectiveness” of the value stream, which further helps assess the efficacy of a previous decision as good or bad or make new decisions regarding policies and investments. Therefore, the relationship between a value stream and REVPR can be represented as follows:

$$\begin{array}{c}\hfill \mathrm{V}\mathrm{S}\left(\mathrm{t}\right)=\left\{\begin{array}{c}1,REVPR\left(t\right)\ge REVP{R}_{\mathrm{d}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{r}\mathrm{e}\mathrm{d}}\hfill \\ 0,REVPR<REVP{R}_{\mathrm{d}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{r}\mathrm{e}\mathrm{d}}\hfill \end{array}\right.\end{array}$$

(18)

where VS is a value stream such as EVCM, EVFMO, and others as described in Section 3, REVPR(t) is the relative EV proliferation rate at a given point in time t, and REVPR_desired is assigned based on a policy or business decision, for a value stream to be considered “effective”.

It is essential to highlight that applicable domain knowledge must be leveraged to determine the most influential factors which realize the specific value stream. For example, EVFMO will realize a quicker “effectiveness” by favorable policy decisions and higher investments to support EV fleet proliferation and will translate into selecting a lower value for ST_FM. Moreover, it is noteworthy that there is no perfect approach for selecting an ST value, and it requires qualitative estimation based on domain knowledge and related efforts toward policies and investments. A lower ST value would require immediate efforts toward policy making and related investments. On the other hand, a higher ST would translate into relaxing a specific area of policy or investment. To comprehend this mathematically, the REVPR Equation (9) is impacted by the X(t) Equation (7), which is a function of ST. Therefore, the value stream to the ST relationship can be logically represented as follows:

$$\begin{array}{c}\hfill \mathrm{VS}\left(\mathrm{t}\right):REVPR⟶S{T}_{\mathrm{X}}\end{array}$$

(19)

5. Model Validation and Simulation

The model presented in the previous section is simulated for multiple scenarios which analyze the relative EV proliferation and total market share across jurisdictions to study the timelines for each EV value stream. The base set of values for all factors, as highlighted in

Table 2. Base values for scenario analysis.

Factor	Value
AVL (years)	10
GRNVPY(0) in 2020	0.018
VMPY (miles)	15,000
ICV Cost (USD) in 2020	16,000
EV Cost (USD) in 2020	35,000
TaxICV (%)	10
TaxEV (%)	10
CMPYICV in USD	500
CMPYEV in USD	170
Gas Stations in 2020: NGS(0)	590,000
EVSE in 2020: NEVSE(0)	78,500
New Vehicle Buyers in 2020: NVB(0)	17,000,000
ICV in 2020: TINICV(0)	287,300,000
EV in 2020: TINEV(0)	1,800,000
GP(0) in 2020 (USD/gallon):	2.419
EP(0) in 2020 (USD/unit):	0.2
RIGP (%)	0.0223
RIEP (%)	0.018
EffICV	0.08
EffEV	0.195
STIBT, CRM, EVDR, SCEV in years	10
STFM in years	15
STCII, AEVM, TEI in years	20
STCAP, ETC, SDT, CAH (%)	50
STET (%)	100
wIBT, CRM, CHI, SCEV, ECO	1
wEVDR, CII, AEVM	0.66
wCAP, ETC, SDT, CAH, FM, TEI, ET	0.33

, are obtained from multiple sources [1,9,25,26,27,42,43,111,112,113] where, without loss of generality, the geography is selected as USA and the base year for simulation as 2020. The model can be applied across any jurisdictional and regulatory boundary to study EV proliferation as long as the supported data is available. Going forward, as new factors emerge, the model will require updates to include those as part of the analysis.

5.1. Scenario 1: Comparative Analysis to Assess the Impact of Additional Factors on EV Proliferation and Market Share with Respect to the Previous Related Work

This scenario focuses on assessing the contribution of this work by considering additional EV proliferation factors not covered in previous studies, as highlighted in Table 1. To realize that, the base model factors are simulated against the scenario where the specific factors based on Table 1 were considered for respective studies and the rest were eliminated by setting their weight to zero. Figure 3 and Figure 4 present the projections for REVPR and EVMS. It is evident that the REVPR and EVMS generated from the previous studies differ significantly among themselves, highlighting the importance of additional factors considered in this paper and the importance of a more comprehensive approach. Based on the values for the base scenario, the time when the EV proliferation will surpass the total number of ICV in the specific geography, i.e., EV(#) > ICV(#), is around 20 years, whereas, when specific factors are ignored as per previous works of literature, it is more than 30 years. However, it should be noted that the importance of this analysis could be understood by changing the values of different factors and weights to inform decision making around technology decisions, investments, and policy making. A separate scenario is presented later to analyze this aspect further.

5.2. Scenario 2: Analyzing the Impact of Technology Improvements on EV Proliferation

The focus of this scenario is to assess the importance of technological factors as recent EV technology improvements impact multiple factors in our study, including IBT, CRM, EVDR, CII, and SCEV. The technological improvement projections depend on jurisdictional policies and geopolitical forces and may vary once they change. Therefore, it is vital to understand the impact of technology improvement variations on EV market share over time. As highlighted in Figure 5, the quicker technology improvements, such as in less than five years, can help EVs capture the majority market share in just fifteen years from now.

5.3. Scenario 3: Approach to Assess the Impact of the COVID-19 Pandemic

A recent NREL report [9] studied the challenges due to the COVID-19 pandemic, including the resulting restrictions and economic downturn, and concluded that the EV charging industry did not get impacted as severely as other areas of the energy sector [114]. This scenario focuses on assessing this impact by evaluating the EVMS projections considering with and without pandemic factors. As represented by the EVMS values in Figure 6, the impact of the pandemic alone may not be noticeable soon, but its gradual impact over a more extended period will demonstrate a significant change in people’s preferences towards buying EVs over ICVs.

5.4. Scenario 4: Demonstrating Framework’s Applicability to Realize Value Streams through EV Proliferation Factors

This scenario highlights the EV proliferation framework and model capabilities by demonstrating the interdependency between the EV proliferation factors and the value streams through simulations. A reference to the latest US administration goals of 50% EV sales by 2030 [115] is made to quantify value streams’ “effectiveness” aspect and assign a realistic value to REVPR_desired. In other words, as REVPR represents EV sales over ICV in our model, setting REVPR_desired = 50 will help us consider each value stream to be “effectively” realized, as described in Section 4.3. Moreover, leveraging the example from Section 4.3, a ST_FM = 10 will help ensure EV in fleets will have achieved its saturation point in 10 years (2020–2030) and will no longer act as a dominant factor in an EV buying decision. Therefore, assigning values of REVPR = 50 and ST_FM = 10 could be utilized to adjust the remainder of the values as provided against EVFMO in

Table 3. Value-stream mapping with EV proliferation factors aid in analyzing the timelines to be met for each factor influencing individual value stream.

Value Stream	Tech Factors (STIBT, CRM, EVDR, CII, SCEV)	STAEVM	STFM	STTEI	TaxEV
EVCM	10 years	20 years	15 years	10 years	10%
EVFMO	10 years	10 years	10 years	10 years	0%
V2(G/H/V)	5 years	20 years	15 years	10 years	0%
PEM	10 years except STCII = 5 years	20 years	15 years	10 years	10%
DGMEV	10 years	20 years	15 years	5 years	10%

. A visual representation of this example is provided in Figure 7.

Please note that since each EV proliferation factor impacts individual value streams differently, Table 3 represents one possible illustration of the same. As highlighted in Section 4.3, each ST is directly governed by decision making on technological investment, policy directives, and other aspects such as innovation breakthroughs. Therefore, it is evident that based on the domain knowledge, certain factors and corresponding values of ST can be made larger while others smaller on a case-by-case basis. For example, increased investment by EV manufacturers to expand the number of EV models will lower the ST_AEVM, directly impacting associated value streams and making them become “effective” sooner. In another case, a policy directive to subsidize fleet operators may result in a sharp increase in the number of Fleet EVs, in turn reducing the ST_FM, thus rapidly generating a high value for the EVFMO value stream. In a third case, a sudden innovation breakthrough may help reduce the battery or copper prices, in turn lowering the ST_CRM and creating high-value across multiple value streams. Therefore, changing the policy or investment for a particular aspect will impact the corresponding ST, in turn helping achieve the desired value of REVPR in the desired time period.

The mapping in Table 3 represents an illustrative example that helps analyze how EV proliferation factors could work differently across each value stream to achieve the same goal of minimum 50% EV sales, as shown in Figure 7.

5.5. Scenario 5: Comparative Analysis to Understand Model Applicability across Different Countries and Jurisdictional Policy Regimes

This scenario highlights the model’s applicability across different countries, where individual policies and locational decisions may impact EV proliferation differently. To keep the focus on understanding policy impacts on economic and other related factors, the following assumptions have been made to forecast the EV market share:

• The selection of countries, such as Norway, Sweden, and the Netherlands, is based on the understanding that they have similar environmental sustainability focus, which drives policy directives, consequently impacting EV proliferation.
• The USA’s inclusion in this scenario study enables the comparison of a different jurisdictional regime partially driven by state policies and not just federal directives, as in the case of the other three countries.
• Values for the factors typically not impacted by government policies are kept the same across countries. Those include AVL, VMPY, CMPY_ICV, and CMPY_EV.
• Based on the assumption that geo-political reasons such as energy (fuel and electricity) prices have a similar global impact, factor values such as RIGP (%) and RIEP (%) are kept the same across countries.
• Assuming technology advancements propagated similarly across all countries, values for Eff_ICV and Eff_EV are kept the same across each country.
• To ensure fair comparison across all the countries, parameters as ST_{IBT, CRM, EVDR, SCEV}, ST_FM, ST_{CII, AEVM, TEI}, ST_{CAP, ETC, SDT, CAH, ET}, and w_{IBT, CRM, CHI, SCEV, ECO}, w_EVDR, w_CII, w_AEVM, w_CAP, w_{ETC, SDT, CAH, FM, TEI, ET} are also assigned equal values across countries.

Considering the above assumptions, the rest of the values highlighted in

Table 4. Values across different countries.

Factor	USA	Norway	Sweden	Netherlands
AVL (years)	10	10	10	10
GRNVPY(0) in 2021	0.02	0.024	0.018	0.1399
VMPY (miles)	15,000	15,000	15,000	15,000
ICV Cost (EUR) in 2021	15,095	40,000	35,000	33,000
EV Cost (EUR) in 2021	33,000	39,000	57,000	52,000
TaxICV (%)	10	30	21	25
TaxEV (%)	10	5	22	16
CMPYICV in EUR	400	400	400	400
CMPYEV in EUR	140	140	140	140
Gas Stations in 2021: NGS(0)	620,000	1600	2900	4100
EVSE in 2021: NEVSE(0)	113,600	19,300	16,335	99,500
New Vehicle Buyers in 2021: NVB(0)	14,900,000	176,276	301,006	320,000
ICV in 2021: TINICV(0)	287,300,000	5,354,451	171,573	224,000
EV in 2021: TINEV(0)	667,731	647,000	129,433	96,000
GP(0) in 2021 (EUR/gallon):	2.28	7.23	6.367	6.783
EP(0) in 2021 (EUR/unit):	0.19	0.13	0.25	0.32
RIGP (%)	0.0223	0.0223	0.0223	0.0223
RIEP (%)	0.018	0.018	0.018	0.018
EffICV	0.08	0.08	0.08	0.08
EffEV	0.195	0.195	0.195	0.195
STIBT, CRM, EVDR, SCEV in years	10	10	10	10
STFM in years	15	15	15	15
STCII, AEVM, TEI in years	20	20	20	20
STCAP, ETC, SDT, CAH, ET (%)	100	100	100	100
wIBT, CRM, CHI, SCEV, ECO, wEVDR, CII, AEVM, CAP, wETC, SDT, CAH, FM, TEI, ET	1	1	1	1

are obtained from different sources [116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134]. These factors have a direct impact due to a policy directive, investment decisions, or other policy-related aspects and are, therefore, critical in producing the results in the scenario analysis.

The values in Table 4 are leveraged to analyze the policy impacts and individual decisions made across different countries. As depicted in Figure 8, Norway is already leading the pack in terms of EV market share due to its increased focus on policies such as a low tax on EVs as opposed to Sweden and Netherlands. Other factors, such as relatively higher gas prices and much lower electricity prices than its neighbors, also support high EV proliferation in Norway. The USA, on the other hand, is the smallest in EVMS% compared to the other three countries, given its vast network of gas stations and ICV numbers, despite being one of the largest uptakers of EVs globally.

5.6. Research Summary and Limitations

The framework and the implemented model demonstrate their applicability for EV manufacturers to focus their investment in specific areas, represented by each factor to cater to different EV market segments such as fleet owners or households. Similarly, the framework can be leveraged by utilities and grid operators to obtain the timelines required for infrastructure advancements and ensure grid reliability before significant EV proliferation starts overloading the grid. It is important to note that the goal of providing the EV proliferation factors to value stream mapping is to highlight the interdependencies between those where policy or investment decisions may impact a factor, consequently impacting one or more value streams. On the other hand, a policy change may make a value stream very lucrative for investors to increase investments, further driving the growth of an EV proliferation factor.

At the same time, it is crucial to understand that the weight and value of each factor will change over time within a jurisdiction and across different jurisdictions, consequently impacting value streams differently. More importantly, as various scenario analyses showcased framework flexibility, the results presented here are entirely based on the model equations and input data. Therefore, to apply the model accurately and obtain meaningful results, careful attention must be paid to the following:

• Assessment of each factor and its weight based on the scenario;
• Accuracy of the input data for independent variables;
• Models coefficient tuning across different equations;
• Where applicable, assessment and tuning of model equations in cases where a different relationship is observed between the dependent and independent variables

Future research directions could include evolving the model for identifying additional EV proliferation factors and related EV value streams and understanding accessory interdependencies among those.

6. Conclusions

This paper provides multiple research contributions which differentiate it from the existing literature. Some of the notable contributions of this research are:

• Identification and classification of different EV proliferation factors influencing a vehicle buyer to buy an EV versus ICV. Through analysis, it was found that the EV market share can surpass the ICVs in as quickly as 15–20 years if all the factors which impact EV proliferation are addressed through policy making and investments simultaneously, whereas focusing on a few, such as in the previous studies, will lead to slower EV adoption.
• An exhaustive survey of monetary and non-monetary value streams emerging due to EV proliferation could benefit electricity grid operators at both distribution and wholesale levels, new and existing businesses, including those who rely heavily on vehicle fleets, as well as consumers who are currently facing challenges due to high gas prices globally.
• A mathematical model to study relative EV proliferation versus ICV, followed by a reusable framework to assess the impact of EV proliferation and ensure timely decision making towards grid and EV investments and a more informed data-driven policy making. The analysis indicated how significantly the policies could impact EV market share of up to 10% or more where individual European countries have made different decisions, even while being part of the same European Union mandate towards sustainability [135].
• The scenarios studied in this paper have been carried out to validate the proposed model across different countries. The results uncovered that concentrated effort globally is required to support EV proliferation, as discrete efforts will sometimes take twice the time to reach a desired EV market share. Other scenario analyses highlighted that the factors such as technology improvements, and more recent ones such as people’s driving behavior change (largely due to COVID-19) could play a major role in EV proliferation and, therefore, cannot be ignored during key decision making around investments and policies.

Finally, the paper highlights important caveats essential to consider while leveraging the model or the framework for current or future studies. Lastly, as new EV proliferation factors emerge or existing ones change their current impact, the paper highlights the limitations of our approach to help further evolve the mathematical model and framework presented in this literature.