Impact and opportunity analysis of the development of the electric mobility at a national level

Transcription

1 Master Thesis 2015 Master in Energy Management and Sustainability Impact and opportunity analysis of the development of the electric mobility at a national level A project carried out at the Energy Center EPFL Author : Cihan Cavdarli Supervisor : Dr. François Vuille Director of Development Energy Center Lausanne, 19th of June 2015

2 Abstract Electric vehicles (EVs) provide a range of environmental (e.g. decrease in CO2e emissions, reduced pollution of urban areas, reduced noise pollution) and socio-economic benefits (lower total cost of ownership of vehicle, improved balance of payments), and may contribute increase supply security at regional level. However they may however also feature some drawbacks such as reduced fiscal revenues, job losses, infrastructure deployment,? An original spreadsheet model to assess the different impacts of electric vehicles has been developed. Over 40 input parameters are used in the model, as it delivers results in the form of 10 key output indicators. Some are of economic or strategic nature (total cost of ownership (TCO) of electric vehicles, fiscal revenues to the national government, level of energy independence, infrastructure cost and employment level), while the others are environmental indicators (level of CO 2 eq emissions, primary energy consumption, electricity demand, level of urban pollution). In a general consideration, the positive impacts of electric mobility development tends to overweight its negative impacts. This conclusion is reported to be relatively robust to variabilities of input parameters. Only in few extreme cases, the positive impact may be questionable. Keywords: Electric vehicle, Total cost of ownership, Fiscal analysis, Climate change, Marginal electricity, Energy independence, Swiss Energy Strategy 2050

3 Contents 1 Introduction 1 2 State of the art 2 3 Model Model description Input parameters Energy prices & electricity mix Vehicle parameters Vehicle market Tax and subsidies Free input parameters Output indicators Environmental impact Economic impact Methodological description Number of ICEVs, composition of the fleet and fuel consumption Number of EVs and electricity demand for vehicles Primary energy use Energy independence for private vehicles CO 2 eq emissions of ICEVs CO 2 eq emissions of EVs PM & NO x emissions in urban areas Total cost of ownership Public sector impact Private sector impact General assumptions Model scope Case study: Switzerland Introduction Description of the scenarii Summary of the scenarii Reference situation: Today Scenario Scenario Results & Discussion Electricity demand & primary energy consumption Energy independence Climate change: CO 2 eq emissions Urban pollution Economic impact for the end user Economic impact for the country I

4 5 Conclusion Impact analysis Added value of the present work Recommendations for further study II

5 List of Figures 3.1 Flowchart of the model Historical oil prices - [9] ICEV Emissions reduction cost curves - [30] Cost of Li-ion battery packs. Nykvist and Nilsson (2015) [23] Electricity production and demand - Today Monthly electricity production by source in 2035 High scenario - Swiss Energyscope calculator Monthly electricity production by source in 2035 Low scenario - Swiss Energyscope calculator Monthly electricity production by source in 2050 High scenario - Swiss Energyscope calculator Monthly electricity production by source in 2050 Low scenario - Swiss Energyscope calculator Electricity demand by EVs Primary energy consumption - Today Primary energy consumption Primary energy consumption (a) Primary energy consumption (b) Share of energy importation for passenger cars CO 2 eq emissions over total life-time CO 2 eq emissions CO 2 eq emissions share for private cars Urban pollution indicators Total cost of ownership - Detailed costs for today Purchase cost - Today TCO - Expected Temporal Evolution TCO dependence on oil price Governmental fiscal revenues from private vehicles Electric utility balance - Today Employment Summary of the output indicators for each scenario III

6 List of Tables 3.1 ICEV input parameters EV input parameters Carbon intensity of Li-ion batteries EV charging pattern - Share of charging EVs during time interval EU limit emissions standards for diesel passenger cars, mg/km EU limit emissions standards for gasoline passenger cars, mg/km Assumed efficiency for non-fossil electricity generation Source: OECD/IEA General input parameters Energy and fuel prices - Today Source: Average among various sources Gasoline ICEVs characteristics Fleet repartition by engine generation Diesel ICEVs characteristics EVs characteristics CO 2 eq emissions for indigenous and import electricity production - Today Life-cycle GHG emissions and efficiencies for various technologies Sources: Swiss EnergyScope, European Environment Agency Energy and fuel prices Source: Prognose Life-cycle GHG emissions and efficiencies for various technologies Sources: Swiss EnergyScope, European Environment Agency Energy and fuel prices Source: Prognose IV

7 Chapter 1 Introduction Electric vehicles may, under some conditions, bring environmental benefits (e.g. decrease in CO 2 eq emissions, reduced pollution of urban areas, reduced noise pollution, synergy with renewable production,...), socioeconomic (e.g. total cost of ownership (TCO), balance of payments of energy,...), and strategic (e.g. energy independence, energy supply security). Yet it may also have some disadvantages such as reduced fiscal revenues, job losses, infrastructure deployment costs, technology lock-in... A lot of region specific parameters must be taken in account when considering the relevance of the electric vehicle market development. These parameters may be macroeconomic (e.g. international emission reduction policies) and uncontrollable on a regional level, or they may be socio-cultural (e.g. reticence to change, "range anxiety",...) that does not follow simple economic logic. Finally, the insufficient development of the electric vehicle market may have negative impacts especially concerning the return on investment of stakeholders that are key to this development. This study aims to evaluate quantitatively the advantages and drawbacks of the the electric vehicle market development on a regional level. To achieve that goal, a parametric model will be developed and applied to the case of Switzerland. With the help of key output indicators, the relevance of the electric vehicle development will be assessed. This report starts by presenting the state of the art in chapter 2. Then, chapter 3 describes the developed model. Chapter 4 discusses the Swiss case study and the results of the model. Finally, chapter 5 proposes a conclusion of the study. 1

8 Chapter 2 State of the art Previous publications study the impacts of electric vehicles on different metrics. However, theses study are often limited to one impact indicator such as economic or environmental. A review of this literature is provided in what follows. Binesh et al. conducted a study on the impact of electric vehicles on global economics. They focused on the study of two global economic impact: microeconomic and macroeconomic [4]. However, their approach is more a qualitative estimation based on a limited parameters. An environmental impact approach was proposed by Granovskii et al. in 2006 [13]. Their results are used to compare four types of private vehicles on economic and environmental level. This study differs from current work as it does not asses the impacts of these different vehicles but only provides a comparison between them. An analysis implying the EU CO 2 reduction goal for 2050 was performed by McKinsey & Co. in 2010 [21]. This analysis is however mainly scenario based and aims to compare and analyse under which conditions the stated goals may be reached. Another environmental analysis was provided by Hawkins et al. in 2013 [14]. Nonetheless their study also limits itself to a life-cycle comparison between conventional and electric vehicles. No economic assessment is considered. The most relevant and similar work was provided by Kampmann et al. in 2011 for the European Commission [16]. They study on a European level economic and environmental impacts. However, their study is oriented on a policy analysis with a recommendation focus. Therefore it does not aim to balance the relevance of the electric vehicles in a particular region. 2

9 Chapter 3 Model 3.1 Model description The constructed model aims to assess the potential impacts of a market uptake of the electric vehicles. The goal is to have an holistic approach of this problematic and provide quantitative results that can be used for a global analysis. In order to achieve this, the model considers several fixed and variable input parameters, calculation modules and output indicators. In figure 3.1, a graphic representation of the model and its modules is represented. The pink blocks represent the main input parameters, the blue blocks intermediary calculation steps and the green blocks the main output variables. A detailed description of these blocks and the methodology behind the calculations are provided in the following sections. Figure 3.1: Flowchart of the model 3

10 3.2 Input parameters The model considers over 40 input parameters to estimate the impact of electric vehicles. These parameters includes both fixed and variable parameters. By variable parameter, it is understood a parameter that may change according to estimations and/or future projections. The value of these parameters is presented in the description of each case study. Fixed parameters are not influenced by the choice of a particular scenario. The most relevant input parameters are detailed in the following sections Energy prices & electricity mix The energy prices are variable parameters that have a major implications for all economic calculations. The energy forms involved in this study include fossil fuels (i.e. gasoline and diesel) and electricity. Another key variable input parameter is related to the marginal electricity generation. Fossil fuel prices The price of oil, raw element for fossil fuels used in transport sector, significantly grew during the last decades (figure 3.2). Also because of its volatility, the fuel prices are very sensitive elements determining the total cost of ownership (TCO) of a vehicle powered by an internal combustion engine. As a variable parameter, the fossil fuel prices have to be defined for each scenario. The projection of this price is out of the scope of this study thus values are based on external sources. A sensibility analysis on the TCO is performed as explained in section In that case, oil price is a free variable parameter. Figure 3.2: Historical oil prices - [9] 4

11 Electricity prices Electricity prices for the end users varies significantly from one region to another. Even within the same country, the pricing may differ between localities and between end customers. The cost of the electricity, which impact its final price to end customers, depends on its source but also on the market demand at the time. In 2012 coal represented 27.4% of the European Union electricity production, natural gas accounted for 18.7%, renewables (including hydro power) for 24.2%, nuclear 26.8% petroleum for 2.2%[11]. Since the electricity production from oil related products (oil and natural gas) represents a small share in the total production, the electricity price is assumed to be decorrelated from oil price variations. Usually, residents have a fixed rate depending on the amount of electricity used. However it is possible to imagine a fully liberalised market that allows all consumers to have access to spot price market. For simplicity and also because fully liberalised markets are not significantly spread at the time of this study, the model considers only single price rate for the consumer. This price is be defined in the description of the case study scenarii. Marginal electricity generation Grid operations involve to meet the instantaneous consumption by using a set of power plants available for power generation. The baseload is generally fulfilled by large coal and nuclear power plants which are designed to operate continuously at low cost. Hydro power plants and natural gas power plants are usually used to meet the peak demand due to their flexibility and/or high operating cost. New renewable energy technologies increases the complexity of grid operations and the determination of the marginal production becomes harder to establish. The marginal production is distinct from the average production, especially in terms of GHG emissions. These two considerations may lead lead to different results. To have a clear view on the electricity usage and the associated emissions, the marginal electricity production linked to the EV market uptake has to be defined. The marginal production depends on mainly on two factors: the volume of the demand and its timing. 5

12 3.2.2 Vehicle parameters Internal combustion engine vehicle parameters The parameters for the internal combustion engine vehicle (ICEV) are can be divided in two main categories: economic and energetic. These parameters are summarized in table 3.1. Economic Vehicle purchase cost Maintenance cost Operation cost Insurance cost Tax and subsidies Energetic Fuel consumption CO 2 e emissions Engine generation Petroleum refinery and transport efficiency Vehicle production carbon intensity Table 3.1: ICEV input parameters The vehicle purchase cost is the actual market price that has to be paid by the customer to purchase the car. In the model, the price of the official car dealer in the studied country is considered. The maintenance costs represent the additional cost to keep the vehicle functioning. It includes annual services and eventual repairs, tires and brakes, and other miscellaneous costs. These cost are specific to the studied country and thus have to be determined in consequence in the case studies. The operation cost includes in particular the cost of the fuel to operate the car. It depends on the fuel consumption of the car, the fuel price described earlier and the mileage. The insurance cost is estimated by selecting the same conditions for all the vehicles. The minimal rate is chosen, ideally, using a comparator. Finally, the tax and subsidies are described in section The fuel consumption and the CO 2 eq emissions are generally given by the manufacturer from the typeapproval test. However, it was reported that these values are generally below the figures observed in real world. The real-world values for CO 2 eq emissions were observed to be up to 50% higher for gasoline vehicles and up to 60% higher for diesel vehicles in 2012 [17]. Therefore, the manufacturer values have to be adjusted using these considerations. The engine generation determines the fuel consumption and pollutant emissions rate in function of the year it was put in circulation. This parameter is detailed in section The petroleum refinery and transport efficiency relates the energy losses during the transformation of crude petrol to gasoline and diesel. This value was reported to be 90.6% on average for US refineries and is assumed to be an average value for all refineries [28]. The production of the vehicle itself already generate CO 2 eq emissions. In order to estimate the amplitude of these emissions, a CO 2 eq/weight model have been developed [32]. It is assumed that the production CO 2 eq are only related to the weight of the vehicle, even though technology development such as hybridisation may influence it. ICEV emissions reduction cost curves The European Commission introduced a new legislation to limit CO 2 eq emissions of new passenger cars. According to this, new cars new cars fleet average emissions have to be 95 gco 2 eq/km by 2021 (Regulation (EU) No 333/2014). It is expected that the limitation goes even lower by 2035 and 2050, however the values are extremely hard to estimate. To achieve this limitation, car manufacturers have to put extra effort on the manufacturing process and push further technologies like hybridisation. Therefore, extra manufacturing costs that will undoubtedly impact the price for the end consumer. Cost curves have been established to estimate the increase in manufacturing costs in function of the emission reduction [30]. Using these curves (figure 3.3) and assuming some future emission reduction targets, future purchase price of ICEVs are estimated. 6

13 Figure 3.3: ICEV Emissions reduction cost curves - [30] Electric vehicle parameters Similarly to the ICEV, the input parameters for the electric vehicle (EV) is distributed between economic and energetic categories. The list of these parameters is given in table 3.2. The EV parameters are similar the the ICEV parameter. The main differences are developed in the following paragraphs. Economic Vehicle purchase cost Battery cost Charging station cost Maintenance cost Operation cost Insurance cost Tax and subsidies Energetic Electricity consumption Battery charge-discharge efficiency Electricity transport efficiency Battery and vehicle production energy intensity Table 3.2: EV input parameters At the time of the study, the battery packs represent the main additional cost for electric vehicles. Do to relative low manufacturing volume, the Li-ion battery packs for EVs do currently not take fully advantage of economies of scale. However, this situation may change in a short time scale according to expert statements and recent studies [23]. This special parameter is detailed in the next part. It is assumed that each electric car requires the purchase of a private charging station. The cost of these stations depends on the region studied. It is assumed that the private charging stations have the same life-time as electric vehicles. The EVs are assumed to have less maintenance requirements due to their simpler power train, mechanically speaking. The electric engines are constituted of less moving mechanical parts compared to standard ICE engines. And due to their power characteristics, the transmission is also simpler since there is no need for a gearbox. Moreover, thanks to the regenerative braking of electric motors, the brake pads are subjected to less deterioration. All these factors combined, the maintenance and costs for EVs are supposed to be 40% lower than ICEV maintenance cost. The operation cost of EVs is very similar to the operation cost of ICEVs. The manufacturer electricity 7

14 Study (Chemistry) Notter et al Majeau-Bettez et al (NCM) Majeau-Bettez et al (LFP) GHG emissions [gco 2 eq/w h] Table 3.3: Carbon intensity of Li-ion batteries 18:00-00:00 00:00-06:00 06:00-18:00 Charging profile 60% 20% 20% Charging profile 2 20% 60% 20% Table 3.4: EV charging pattern - Share of charging EVs during time interval consumption data is adjusted with the charge-discharge efficiency of the batteries to have the real world consumption. The electricity consumption is generally provided by the car manufacturer. However, these values are often under estimated compared to real world consumptions. Thus, these values are adjusted with real world measurements and also considering battery efficiencies. During the charge and the discharge of the battery packs, ohmic occurs due to the internal resistance of the batteries. Due to these losses, the round-trip efficiencies of Li-ion batteries is on average 90% [15], [34]. Significant energy losses occur between the generation and the the consumer s plug. Theses losses have to be considered to determine the real electricity demand of EVs. However, the losses are dependent on the power network configuration and may vary between regions. Thus this parameter is defined specifically for the case study. A major environmental concern of electric vehicles is the carbon intensity of the battery packs. Several studies gathered data on the GHG emissions of Li-ion battery packs during its production. These emissions are extremely variable depending on the manufacturing process and the chemistry variations. Table 3.3 summarise the principal values found in the literature [22], [19]. Due to the lack of information on that particular subject, it is assumed that the carbon intensity of battery packs do not improve in the future. The vehicle production carbon intensity is assumed to be the same as ICEV. However, it is important to note that this assumption may be unfair for EVs, especially in future scenarii where the hybridisation process of ICEV will probably come with more CO 2 eq intensive production phase. Battery pack cost curves The purchase price is a major factor in the realisation of a mass-market electric vehicle. Currently, battery pack costs represent a significant share of the total vehicle. For the vehicles presented in the case study in chapter 4, this battery pack cost represent on average 20% of the total sell price. A recent study present a systematic review of the cost of the battery packs to EV manufacturers [23]. It has been reported that the cost of Li-ion battery packs continue its fall and past projected were significantly higher than costs among market leaders. Figure 3.4 shows the data gathered by this study. Charging pattern The charging pattern defines the timing and the quantity of the power demand by EVs. This parameter plays an important aspect on the while considering the marginal electricity production presented in Recent statistics suggest that the highest electricity demand from EV charging occurs during evening hours [20]. However few statistical data samples are available at the time of this study. The patterns chosen for this study is presented in table 3.4. The main used in the simulations is the charging profile 1. The charging profile 2 is used analyse its impact on key indicators such as the CO 2 eq. The charging profile presents the case where the majority of EVs are plugged in for charging right after working hours. This profile has the drawback to coincide with the commonly seen evening peak in electricity demand. The model considers only V0G chargers to simplify the simulations. V0G chargers act as simple 8

15 Figure 3.4: Cost of Li-ion battery packs. Nykvist and Nilsson (2015) [23] plugs with no communication capabilities with the grid. Other charging technologies are presented although not integrated to the model [12]. V1G: Also called smart charger. There is a real-time communication with the grid. Thus the charger activates itself when the grid needs/allows it. Also includes timed charging and controllable load. V2G: The most complex scheme. Like the V1G but also allows power transmission to the grid. Thus acting as backup source, grid ancillary service and load shifting for renewable power generation. V2B: Same as V2G but limited to one building. Therefore it does not provide grid ancillary service Vehicle market Fleet size, replacement and growth rate The fleet size modelling is beyond the scope of this model, thus available data from the literature is directly used as input for this parameter. The replacement and growth rate are determined using historical data and future projections. These parameters are very important for the fiscal calculations (see sections and 3.3.2). Fleet age & engine generation The vehicle fleet is generally constituted of several vehicle age categories. It is assumed that these categories have different CO 2 eq, NO x and particulate matter (PM) emissions rates. In the model, the European Emissions Standards, presented in table 3.6 for gasoline vehicles and table 3.5 for diesel vehicles, are used as template for this categorisation. The CO 2 eq emissions for each vehicle category is determined using historical emissions rates of the vehicles considered in the study. Since such data is not available for NO x and PM emissions, the limit values specified in the Euro standards are considered as the average emission rates of each category. Fleet substitution pattern The model assumes that the oldest vehicles will be replaced first. Also current trends are studied to determine the share of gasoline and diesel that are replaced by electric vehicles. 9

16 Norm Euro 2 Euro 3 Euro 4 Euro 5 Euro 6 Date Nitrogen Oxide (NO x ) Particulate matter (PM) Table 3.5: EU limit emissions standards for diesel passenger cars, mg/km Norm Euro 2 Euro 3 Euro 4 Euro 5 Euro 6 Date Nitrogen Oxide (NO x ) Particulate matter (PM) Table 3.6: EU limit emissions standards for gasoline passenger cars, mg/km Tax and subsidies Value added tax The value added tax (VAT) is a consumption tax that is part of the purchase cost for practically all goods. Its rate varies from one country to another, for example, the VAT rate ranges from 15% to 27% in the European Union. Depending on the studied region, this tax represents a substantial part of the governments income (e.g. 37% of the total fiscal revenues in Switzerland in 2012 where the VAT rate is 8% for non-food goods [2][1]). Therefore this tax can not be neglected while realising a fiscal analysis. In the scope of this study, the VAT applies to all new vehicles, obtained after determining the vehicle fleet growth rate and replacement rate, and to the energy consumed by these private passenger cars. Mineral oil tax The mineral oil tax represents the tax on oil products used as fuel for ICEVs. In many countries, the mineral oil tax represent a significant share of the fiscal revenues. In 2012, it represented 8.6% of the total fiscal revenues in Switzerland [2]. Due to its non negligible share in the fiscal revenues, the governments may launch new fiscal policies to recover the losses due to a market uptake from EVs. One possibility is to directly transfer this tax to EVs. This possibility is present in the model in order to consider the potential tax transfer on EVS. For this study, the tax amount is calculated in order to correspond to have zero fiscal losses on the fuel tax in the case of 100% BEV market penetration. In practice, the tax is computed on a kw h basis. Moreover, to have more flexibility in the modelling, the tax amount may be adjusted on the whole market penetration range. Another solution, that is not modelled here, is to integrate the tax in the purchase price of the vehicle. This solution may be more manageable for policy makers since the electricity used for mobility may be hard to differentiate from home or office electricity usage. Vehicle and road tax The vehicle and road tax is the tax levied on personal vehicles using the public road. Every country has its own way to define this tax, sometimes with differences within the country. CO 2 tax With the introduction of the UN Kyoto protocol, the signatory parties tried to set up mechanisms to reach the target emission values. One of these mechanism is a tax sanctioning the greenhouse gas (GHG) emissions. The CO 2 tax, sometimes called carbon tax is a form of tax levied on the carbon content of fossil fuels. The implementation of a carbon tax addresses the question of GHG emitters not facing the full costs of their 10

17 actions. Several countries have adopted carbon taxes, yet the major part of these taxes are collected on energy products (e.g. fossil fuels) rather than on real CO 2 emissions. At the moment of this study, no international homogenisation have been implemented, every country has its own set of policies regarding an eventual carbon tax. Government subsidies At the time of this study, the EVs vehicles may benefit from significant subsidy from the government. These subsidies have been introduced in order to reduce the total cost of ownership and accelerate the market penetration. The amount and the form of government incentives varies highly depending on the region Free input parameters EV Market penetration The market penetration of the EVs is the main free parameter for the constructed model. Its range goes from 0%, in which case all the personal transport vehicles are ICEVs, to 100% in which case the personal vehicle fleet is exclusively composed of EVs. The major part of the outputs are represented in function of this parameter. The main purpose of the model being the analysis of market uptake scenarios, letting this parameter free makes possible to compare several outputs and compare their behaviour. Oil price Gasoline and diesel prices are directly linked to the crude oil price. The future projection of the oil price is uncertain and subject to debate. As shown in figure 3.2, the oil price grew steadily during the last decade but the volatility remained high. That are the reasons to let the oil price as free parameter for some sensitivity analysis like on the total cost of ownership. 3.3 Output indicators Environmental impact In this section, the life-cycle impact of standard internal combustion engine vehicles and battery electric vehicles are compared and analysed. This analysis is then used in the model to assess the environmental impact of these technologies. Such a comparison and analysis turns to be extremely important and proved to be essential during the implementation of an effective energy policy. Private vehicles and transportation account for a significant share of the greenhouse gas (GHG) emissions all over the world. For the particular case of Switzerland, the share of CO 2 eq emissions attributable to transportation was 37% of the total CO 2 eq emissions in From this amount, 66% was attributable to private vehicles [27]. Vehicle powered with alternative fuels, such as hybrid or full electric vehicles, are highly promoted in order to mitigate these emissions. To compare the emissions and the energy consumption of power trains under study, a "well-to-wheels" analysis is necessary. Such an analysis considers both "well-to-tank" consumption/emission and "tank-towheels" consumption/emission. It appears that the majority of the energy consumption and the emission occur from tank-to-wheels for internal combustion engine vehicles. Whereas for the full electric battery vehicles, the emissions occur entirely from well-to-tank (during the electricity production) since no tailpipe emissions are present [3]. In the following parts, several aspects of the emissions and the energy consumption are presented. The environmental impacts are divided in three categories allowing to study this impact over the entire life-cycle of the vehicles. 11

18 Production phase The CO 2 eq emitted during the production phase of the vehicles were reported to be responsible for more than 20% of the total life-cycle emissions of ICE vehicles in some cases [30]. The production carbon footprint is even greater for EVs considering the significant amount of CO 2 eq produced during the battery manufacturing. Previous studies estimated the share of the CO 2 eq emitted during the production of an EV, including the battery packs, being 50% of its total life-cycle emissions [14]. Usage phase For an ICEV, the CO 2 eq emission during its life-cycle represent the most important impact. Unlike ICEVs, the electric vehicles have zero tailpipe emissions. However, that does not necessarily mean that the EVs are totally emission free during their use phase. To have a fair comparison, the CO 2 eq emissions produced during the electricity production have to be considered. In this study, the EV CO 2 eq emissions are based on the marginal electricity production as explained in section Since the values of the marginal production may vary in function of the charging pattern, the two cases presented in section are studied. Two kind of non-co 2 emissions are studied for metropolitan areas: atmospheric particulate matter (PM) and nitrogen oxide NO x emissions. These pollutions studied in urban areas because the density of the traffic amplifies the effect of the pollution. A consequence of this choice is that since EVs have zero tailpipe emissions, these type of pollutions become negligible for this type of vehicle. Particulate matters have impact on climate but is also a direct threat to human health. Two types of PM are generally distinguished depending on their diameter: P M 2.5 (2.5 µm) and P M 10 (10 µm). The PM generated from vehicles have several sources. Generally it is admitted that the most important part of the PM production occurs during the fuel combustion process of ICEV. Yet, other mechanisms are also involved in the PM generation process such as: tyre wear, brake wear, clutch wear, road surface wear and corrosion of other vehicle components [18]. In this study it is assumed that non-exhaust PM emissions are equal for ICEVs and EVs, thus these will not be considered in the calculation. However, it is necessary to note that non-exhaust PM production of EVs may be lower than ICEVs since the brakes a less employed due to the regenerative braking of electric motors. Nitrogen oxides generally regroup nitric oxide (NO) and nitrogen dioxide NO 2. These chemical components are formed during fuel combustion at high temperatures. It was reported in the literature that overall NO x emissions for the transport sector may increase with the adoption of EVs [16]. However this conclusion is highly dependent on the electricity generation mix and is only true if the generation mix is highly dependent on coal power plants. Noise pollution is another aspect related to urban transport pollution. This aspect is not considered in this study although it can have a significant importance. Besides the environmental concerns, energetic aspects of the use phase of vehicles is also considered. Especially, the raise in electricity demand due to a market uptake from EVs and the primary energy consumption are studied. The study of the primary energy consumption offers the opportunity to calculate and compare the total resources used for an application after having considered all the transformation losses. In the scope of this study, it represents a quantitative way to compare the energy consumption of ICEVs and EVs. End-of-life While the end-of-life is limited to recycling for the ICEV and EV main parts (e.g. chassis, power train), two possibilities are offered for EV batteries. The first one considers standard recycling for the batteries arrived at their end of life-time. The second possibility considers a second-life for these batteries. The recycling CO 2 eq emissions are assumed to be included in the production phase since old parts are assumed to be used as raw material during new car manufacturing process. Most of EV producers give a warranty on the battery maximum state-of-charge (SoC). If this value goes below a certain level (typically 75% for most manufacturers), the entire battery pack is replaced free of charge. In this study, the warranty-time is considered as the battery expected life-time and the minimum guaranteed 12

19 (SoC) as the limit at which the customer change the battery pack of his EV regardless if it is covered by the warranty. Yet, the minimum guaranteed SoC is usually high enough to imagine stationary applications where weight and volume are not subject to high restrictions as for mobile applications. Therefore, the environmental impact of the production may be partly transferred to the second-life of the battery. It is assumed that the battery is operational until the maximum SoC drops to 25% of the original amount. Thus, considering a second-life, only 1/3 of the CO 2 eq emissions for the battery production have to be considered in the EV CO 2 eq footprint analysis Economic impact Total Cost of Ownership The total cost of ownership (TCO) is a way to compare the total cost for the consumer in a transparent way. Theses costs are computed over the complete lifetime of the vehicle and enable the possibility to compare different models in a quantitative manner. The TCO information is more and more used for marketing purposes in the industry. Existing studies have shown that electric cars are currently more expensive if government subsidies are not considered [35], [6]. Moreover, it has been observed that the TCO information may influence the consumer choice even if the capital cost seems to be higher [8]. The TCO does not remain static in a temporal point of view. It is be affected by several parameter changes such as energy prices, engine efficiencies, government incentives,... The present study considers several parameters. As explained in section 3.2.2, the expected efficiency increase for ICEVs impacts negatively the capital cost but positively the operation cost from a consumer point of view. And the expected decrease of battery prices impacts positively the capital cost for the consumer. Public economy The output on public economy is the impact on the fiscal revenues for the government. The fiscal revenues include all the taxes and subsidies presented in part Its purpose is to assess quantitatively the impact of an EV market takeover on the government revenues. Two cases are considered: with and without the transfer of the mineral oil tax on EVs. Private economy The public economy includes the infrastructure cost and the impact on the employment. The infrastructure cost considers only the public charging points. These costs are assumed to be covered by the private economy. The model develop the worst case scenario for electric utilities in which these actors have to support the complete financing of the public charging infrastructure. The impact on employment is assumed to be driven by two main element. The creation of jobs is assumed to be mainly on the infrastructure development side such as private and public charging stations and EV promotion during the early stage of the market uptake. Whereas job losses may occur on the maintenance part of the vehicle supply chain, since the EVs are assumed to have 40% less maintenance needs compared to ICEVs. Other part of the supply chain like production, support (e.g. gas stations/charging points) and endof-life are assumed to be affected only by a shift from ICEV to EV. Therefore in a employment creation/loss point of view, the impact is assumed to be null. Energy independence A market uptake of electric vehicles will result in a transfer from oil use for transport to electricity use. Therefore, there may be a change in the share of energy import for the studied region. This shift is a priori uncertain and depends on several factor e.g. share of imported oil for transport, share of imported electricity and share of primary energy source to produce electricity. 13

20 3.4 Methodological description To link the input parameters to output parameters, a spreadsheet model based on Microsoft Excel has been developed. The methodology and the calculations for the different blocks are detailed in what follows Number of ICEVs, composition of the fleet and fuel consumption The passenger vehicle fleet size is determined using statistical data. If statistical data allows it, the fleet is separated in categories depending on the engine generation as presented in section For each engine generation, previous models of the cars used to build the today s average vehicle are gathered to determine the parameters for each generation. In particular, the fuel consumption and the CO 2 eq emissions are determined. The model is configured to replace the oldest generation of each type, gasoline or diesel, first. The vehicle replacement ratio what determines which type, gasoline or diesel, is replaced in priority is also determined using statistical data. The fleet replacement calculations are made for each 1% EV market uptake. The fuel consumption is simply calculated by multiplying the number or cars of each category by its respective fuel consumption and the annual mileage. This relation is represented in the following equations. V Gasoline = D V Diesel = D Variable name Units Description 6 N Euro i, G C Euro i, G (3.1) i=2 6 N Euro i, D C Euro i, D (3.2) i=2 V Gasoline m 3 /year Total volume of gasoline consumed by ICEVs V Diesel m 3 /year Total volume of gasoline consumed by ICEVs N Euro i, G Number of gasoline vehicle of generation i {2, 3, 4, 5, 6} N Euro i, D Number of diesel vehicle of generation i {2, 3, 4, 5, 6} C Euro i, G l/km Fuel consumption of gasoline vehicle of generation i {2, 3, 4, 5, 6} C Euro i, D l/km Fuel consumption of diesel vehicle of generation i {2, 3, 4, 5, 6} D km/year Annual mileage Number of EVs and electricity demand for vehicles As in the case of ICEVs, the number of electric vehicle in the fleet is calculated for each 1% step of the market uptake by EVs. Therefore, the electricity demand for EVs is linearly increasing with the growth of its market share. Marginal electricity production The model do not define dynamically the marginal electricity production source. Instead, this information has to be entered before running the calculations. The marginal electricity production has to be defined at an hourly resolution for each month of the year and for all levels of EV market share. Section 4.2 present a more detailed explanation for the method used in the case study. Total electricity demand by source Using the marginal electricity production information, the electricity demand by source is calculated for each 1% step growth of the EV market share. The model computes first the electricity production for the marginal increase, then sums it to have the cumulative production by source. 14

21 3.4.3 Primary energy use While being quite straight forward for ICEVs, the primary energy use calculation is more delicate and requires more attention. The calculation steps for each vehicle type is explained in what follows. The ICEV primary energy consumption includes refinery losses and vehicle consumption. The equation is then: 1 P E ICEV = (V Gasoline k Gasoline + V Diesel k Diesel ) (3.3) η Refinery Variable name Units Value Description P E ICEV MJ/year - Primary energy consumption of ICEVs k Gasoline MJ/l 32.6 Gasoline energy intensity k Diesel MJ/l 35.9 Diesel energy intensity The EV primary energy consumption calculation involves more steps. After assessing the vehicle consumption, battery charge and discharge efficiencies are considered. Then the electricity transportation efficiency is also and finally the generation efficiencies of the power plants are considered. For practical purposes, the efficiency values in table 3.7 are used to compute the "primary energy" from non-fossil electricity generations. The efficiencies of thermal power plants are specified in each case study since in depend on the technology improvements for future scenarios. Nuclear power 0.33 Hydroelectric 1 Wind 1 Solar 1 Geothermal 0.1 Table 3.7: Assumed efficiency for non-fossil electricity generation Source: OECD/IEA The power production of each power plant type is computed in the electricity demand module of the model as detailed in section From there the primary energy use of each power plant is computed and summed to gather the total primary energy use. P E EV = k Electricity i W i /η i (3.4) i {Nuclear, Coal, Natural gas, Oil, Cogeneration, Hydro power, Photovoltaic, Wind, Geothermal, Import} Variable name Units Value Description P E EV MJ/year - Primary energy consumption of EVs W i GW h - Amount of electricity produced by power plant i η i - - Efficiency of the power plant i 15

22 3.4.4 Energy independence for private vehicles For the energy independence, two cases are considered for the electricity. In the first case, only the distinction between indigenous and import electricity is made. The second case considers also the origin of the fuel to generate the indigenous electricity. In the first case, the share of indigenous and import oil for the country is multiplied with the total fuel consumption of ICEVs. The electricity consumption data is already determined by the electricity demand module detailed in The share of the import is then simply calculated using the following equation. W Import share = ICEV Fuel Import + EV Electricity Import W Transport (3.5) Variable name Units Description W Import share - Share of the imported energy for transport sector ICEV Fuel Import MJ/year Yearly amount of imported fuel for ICEVs EV Electricity Import MJ/year Yearly amount of electricity imported for EVs W Transport MJ/year Yearly amount of energy used in transport sector In the second case, the fuel that is imported for generating indigenous electricity (e.g. natural gas) is also considered as import. Thus on variable is added to the above equation. W Import share = ICEV Fuel Import + EV Electricity Import + Electricity Fuel Import W Transport (3.6) Variable name Units Description Electricity Fuel Import MJ/year Yearly amount of fuel imported for electricity production CO 2 eq emissions of ICEVs Production: CO 2 eq production from vehicle As stated in the previous part, the CO 2 eq emissions produced during the manufacturing process of the vehicle is estimated through its weight. The value used for this estimation is: 1.32kgCO 2 eq/kg vehicle [32]. It is important to note that this value does not consider the probable increase of manufacturing carbon intensity of ICEV due to the hybridisation complexity. Usage phase: CO 2 eq emissions from fuel usage The CO 2 eq emissions mainly depend on the engine generation presented in section The number of vehicle of each category is multiplied by their respective average emission rate and the annual mileage. M CO2eq, ICEV use = D 6 N Euro i, G CO 2 eq Euro i, G + N Euro i, D CO 2 eq Euro i, D (3.7) i=2 Variable name Units Description M CO2eq t/year Yearly mass of CO 2 eq emissions from ICEVs CO 2 eq Euro i, G g/km CO 2 eq emissions of Gasoline vehicle of generation i {2, 3, 4, 5, 6} CO 2 eq Euro i, D g/km CO 2 eq emissions of diesel vehicle of generation i {2, 3, 4, 5, 6} D km/year Annual mileage 16

23 3.4.6 CO 2 eq emissions of EVs Production: CO 2 eq production from vehicle & battery production The CO e q emissions produced during the vehicle manufacturing process (excluding the batteries) is estimated according the same process as ICEVs. The CO 2 eq emitted during the battery production for EVs depends on the volume of battery and its chemistry. The figures presented in for the CO 2 eq intensity of the batteries are used to perform the calculations. The total value is annualized with the expected life-time of the batteries. The following equation is satisfied. M CO2eq, battery production = Variable name Units Description 1 L Battery N EV B CO 2 eq Battery (3.8) M CO2eq, battery production t/year Yearly mass of CO 2 eq emissions from EV battery production N EV Number of EVs B kw h Average size of EV batteries CO 2 eq Battery gco 2 eq/w h CO 2 eq intensity of battery production L Battery years Expected life-time of the EV batteries Usage phase: CO 2 eq emissions from electricity production The CO 2 eq emissions from electricity production is also calculated using the marginal production principle. As in the case of the total electricity demand by source, the model first calculates the marginal CO 2 eq for each step of the market uptake by EVs and then sums it to have the cumulative production for a given market share. The CO 2 eq emissions correspond to life-cycle emissions for each sources. Since the actual per kw h values depends on the technology and its future development, these are defined for each scenario in the case study PM & NO x emissions in urban areas The calculations involved in the determination of PM and NO x emissions are similar to those for the ICEV CO 2 eq emissions. The number of vehicles of each category is multiplied by the corresponding emission standard and the annual mileage. Proceeding like this, the values obtained correspond to the total and limit emissions. M PM = D M NOx = D 6 N Euro i, G P M Euro i, G + N Euro i, D P M Euro i, D (3.9) i=2 6 N Euro i, G NO x, Euro i, G + N Euro i, D NO x, Euro i, D (3.10) i=2 Variable name Units Description M PM t/year Yearly mass of PM emissions from ICEVs M NOx t/year Yearly mass of NO x emissions from ICEVs P M Euro i, G mg/km PM emissions of gasoline vehicle of generation i {2, 3, 4, 5, 6} P M Euro i, D mg/km PM emissions of diesel vehicle of generation i {2, 3, 4, 5, 6} NO x, Euro i, G mg/km NO x emissions of gasoline vehicle of generation i {2, 3, 4, 5, 6} NO x, Euro i, D mg/km NO x emissions of diesel vehicle of generation i {2, 3, 4, 5, 6} D km/year Annual mileage 17

24 3.4.8 Total cost of ownership The total cost of ownership is considers the capital costs and the net present value (NPV) of the yearly costs. As presented in section 3.2.2, the capital costs include the vehicle purchase cost (ICEV and EV), battery cost (only EV), charging terminal (only EV) and the eventual government subsidies. The vehicle purchase cost and the battery cost are adjusted for future estimation as explained in section The yearly costs include the maintenance costs, the operation costs, the insurance costs and the tax & subsidies. The battery replacement after its expected life-time is also considered. The cost of this replacement is the estimated cost of the batteries at the moment when the change is required. The annual costs are then adjusted for the inflation. The following equation summarise the operation. LT Vehicle T CO = V C + BC + CT C S + t=0 M t + O t + V T t + I t + T t (1 + i) t + BR (1 + i) LT Battery (3.11) Public sector impact Fiscal revenues Variable name Units Description T CO CHF Total cost of ownership of one vehicle V C CHF Vehicle purchase cost BC CHF Battery purchase cost CT C CHF Charging terminal purchase cost M t CHF/year Annual maintenance cost O t CHF/year Annual operation cost V T t CHF/year Annual vehicle taxes I t CHF/year Annual insurance costs T t CHF/year Annual other taxes (e.g. mineral oil) BR CHF Battery replacement cost LT Vehicle years Expected life-time of the vehicles LT Battery years Expected life-time of the batteries i - Annual interest rate The fiscal revenues are composed of the value-added tax from new vehicles and from the energy expenditure, vehicle tax, the mineral oil tax, an eventual CO 2 tax and the subsidies (negatively). The VAT income from vehicle sales is obtained using the replacement rate and the new car share found thanks to historical data Private sector impact Employment The employment creation or loss is estimated by essentially two parameters. First the employment losses due to lower maintenance requirements of EVs is estimated. Secondly the employment creation thanks charging points installation and maintenance is considered. NJC = N Charging terminals (T Installation + T Maintenance ) W Installer EV Penetration N Tech (1 EV Maintenance rate ) (3.12) 18

25 Variable name Units Description N JC - Net job creation N Charging terminals Number of charging points T Installation hours Installation duration of one charging point T Maintenance hours Maintenance duration of one charging point W Installer hours Annual working time of installers EV Penetration % EV market penetration rate N Tech Total number of ICEV technicians EV Maintenance rate % EV maintenance needs compared to ICEV Infrastructure cost & electric utility balance The infrastructure cost includes the cost of deployment of public charging points. It is supposed that the maximum number of public charging points is reached by 25% market penetration. The electric utility balance is estimated considering as income the electricity sales to EVs, after deduction for tax and transport fees, and as expenses the spot price of the electricity. 3.5 General assumptions In this section are recalled all the general assumptions made in the construction of the model and the related section with the details. The electricity price is decorrelated from oil price fluctuations. No differentiated electricity rate is available to the customer for the end customer. The CO 2 eq emissions produced during the manufacturing process of the vehicles (excluding batteries) are on depends only on the weight. ICEVs and EVs (excluding batteries) have the same weight-to-co 2 eq proportion. Every EV needs a private charging point, this private charging point has the same expected life-time as the EVs. Maintenance costs of EVs are 40% lower than ICEV maintenance cost. The ICEV PM and NO x emissions correspond to the Euro emissions standards of each engine generation, these standards do not limit further the emissions in the future. The CO 2 eq emissions of recycling is assumed to be included in the manufacturing process A second-life as stationary storage may be considered for EV batteries The end users change the battery of his EV at the end of the warranty time, the maximum SoC of the battery at that moment is 75% of the original maximum SoC. 19

26 3.6 Model scope The aim of the model is to provide quantitative information on electric mobility impact on the following indicators: Additional electricity demand from EVs Primary energy consumption by private vehicles Energy independence for private vehicles CO 2 eq emissions of private vehicles and its share in the total CO 2 eq emissions of the country Urban pollution assessment with estimation for PM and NO x emissions Total cost of ownership of ICEVs and EVs Government fiscal revenues from private vehicles Electric utility balance estimation & infrastructure cost Employment The impacts on the power grid are not modelled in this study. Nor are the noise pollution impacts in urban areas. The model has not as a purpose to forecast EV market penetration. 20

27 Chapter 4 Case study: Switzerland 4.1 Introduction In 2011, soon after the Fukushima Nuclear power plant incident, the Swiss government decided to increasingly remove the nuclear power plants from its electricity production resources. This was not the first step towards the so-called energy transition in Switzerland, some policy measures were already adopted to face the climate change problematic. The post-fukushima decisions came with a substantial amount of challenges and questions, at a politico-economic level as well as a technical level. In 2013, associated as reference scenario for what follows, the transport sector accounted for 35% of the final energy consumption [24]. From that energy, 95% is from oil products, 3.6% from electricity and the remaining from other agents [24]. These numbers put a strong emphasis on the dependence of the Swiss transport sector on imported oil. Moreover, the transport sector emitted 36% of the total CO 2 eq emissions of the country in 2012, private cars accounted for 66% of these emissions [27]. From these figures, it is clear that the transport sector has to be revised in the scope of the energy transition. This case study will focus on the impact of the EVs market uptake on the forecast scenarii and analyse the synergy with those. 4.2 Description of the scenarii This study focuses on two different scenarii, each of these scenarii are studied for future projections in 2035 and These scenarii were prepared by Prognose AG [29]. The details on these sub-scenarii are presented on the following sections. The marginal electricity productions are estimated using the Swiss EnergyScope calculator [10]. This calculator, developed in-house at the EPFL, allows the user to choose a scenario and change key parameters (e.g. market share of EVs) and observe the impact on several indicators. In the particular case of this study, the electricity demand and production variations are observed for an EV market penetration rate ranging from 0% to 100%. Using this output, the marginal electricity production as a function of the EV market penetration is estimated for each scenario. Table 4.1 summarise the general input parameters for the case study. 21

28 Input parameter Unit Value Source VAT rate % 8% [1] Discount rate % 0% Swiss Central Bank Number of ICEV technician [31] Total number of worker [7] Number of vehicles OFS Replacement rate % 6% OFS Share of new cars % 2% OFS Price of private charging terminal CHF Alpiq catalog Price of public charging terminal CHF Alpiq catalog Price of fast charging terminal CHF Alpiq catalog Number of private charging points per EV 1 Own estimate Time needed for the installation of one charging point hours 4 Interview based estimate Annual workload of an electrical technician hours/year 1776 Own estimate Annual mileage of private vehicles km/year Own assumption Table 4.1: General input parameters Summary of the scenarii The following table presents the a summary of the different scenarii and their respective characteristics. Name Carbon intensity of electricity generation ICEV average CO 2 eq emission rate 2013 Low 165 gco 2 eq/km 2035 Low Low 95 gco 2 eq/km 2035 High High 95 gco 2 eq/km 2050 Low (a) Low 75 gco 2 eq/km 2050 High (a) High 75 gco 2 eq/km 2050 Low (b) Low 60 gco 2 eq/km 2050 High (b) High 60 gco 2 eq/km Reference situation: Today The reference situation is based on the latest data available. The model results for this scenario serve as base for the interpretation and comparison of the model results of the future projected scenarii. The energy prices for the reference scenario are summarised in table Type Units Price Electricity CHF/kW h Gasoline CHF/l 1.48 Diesel CHF/l 1.55 Natural Gas CHF/kW h Table 4.2: Energy and fuel prices - Today Source: Average among various sources Car fleet In order to have a meaningful representation of the car fleet, the characteristics of five different cars and their parameters are gathered and used to "build" an average car. To make the comparison between EVs and ICEVs as meaningful and fair as possible, all the parameters are averaged over a range of vehicles of 22

29 Model Ford Focus VW Golf Renault Clio BMW 118i Mitsubishi Space Star Average ICEV Price [CHF ] Weight [kg] Power [kw ] Fuel consumption [l/100km] CO 2 eq emissions [g/km] Table 4.4: Gasoline ICEVs characteristics comparable size and from different brands. For simplicity, several assumptions are made concerning the composition of the car fleet. In particular, the following: 1. The car technologies are supposed to be either gasoline, diesel or full-electric vehicles (simply electric vehicle or EV in what follows). The ICEV vehicles also includes the hybrid vehicles. It is assumed that this technology is an evolution of the basic gasoline or diesel powered ICEVs. 2. The fleet size is supposed to be constant in all scenarii. This is justified by the fact that the current growth rate of 1.5% (OFS 2014) per year is not sustainable in the long run. Simple calculations indicate that, with the current growth rate, more than "160 km of cars" is added on the roads each year. Which confirms the relevance of this assumption. For this case study, fifteen commercialised vehicles are used to determine three average vehicles that are representative of each category. The vehicle characteristics of gasoline ICEVs are presented in table 4.4, the diesel ICEVs in table 4.5 and the EVs in table 4.6. The engine generation fleet repartition is provided in table 4.3 Engine generation Gasoline share Diesel share Euro 2 14% 1% Euro 3 19% 4% Euro 4 16% 7% Euro 5 22% 12% Euro 6 3% 2% Total 74% 26% Table 4.3: Fleet repartition by engine generation The maintenance costs are estimated using the figures provided by the Turing Club Suisse [33]. maintenance costs are composed of: The Service & repair: 9.5% of the catalogue price per year Tire: 6.6% of the catalogue price per km Miscellaneous: 390 CHF/year The insurance costs for each vehicle is determined using the Comparis.ch comparator [5]. The vehicles taxes correspond to the vehicle tax rate in canton Vaud. This canton gives a 75% discount on the tax of electric vehicles. Electricity production and demand The current electricity production in Switzerland is very poor in carbon intensity, 57.9% of the production take its source from hydro power, 36.4% is generated by nuclear power plants and 5.7% is generated by other sources like thermal power plants, wind and solar [25]. The net production (after deduction of pumped storage) is reported to be 66.2 GW h whereas the total consumption including the transmission and distribution losses were 63.8 GW h. 23

30 Model Ford Focus VW Golf Renault Clio BMW 118d Nissan Note Average ICEV Price [CHF ] Weight [kg] Power [kw ] Fuel consumption [l/100km] CO 2 eq emissions [g/km] Table 4.5: Diesel ICEVs characteristics Model Nissan Leaf VW egolf Renault Zoe BMW i3 Mitsubishi i-miev Average EV Price (excluding batteries) [CHF ] Battery cost [CHF ] Weight [kg] Power [kw ] Electricity consumption [kw h/100km] Battery size [kw h] Table 4.6: EVs characteristics Even though its electricity production closely meets its consumption in energy terms, Switzerland is very active in the import-export business with its neighbours. The reasons behind this behaviour are economic and technical. As shown in figure 4.1a, the production does not meet the consumption during the winter whereas the demand is lower than production during the summer. Similarly, figure 4.1b shows the daily balances. This imposes the consideration of the European grid carbon intensity while calculating the marginal CO 2 eq emissions. In this particular scenario, due to limited production sources, only the distinction between indigenous and import production is made for the marginal electricity production. It is supposed that the average grid carbon intensities for indigenous and import electricity is constant and not sensitive to the EV market penetration parameter. Table 4.7 summarise the marginal CO 2 eq emissions for this scenario. Indigenous [gco 2 eq/kw h] Import [gco 2 eq/kw h] Table 4.7: CO 2 eq emissions for indigenous and import electricity production - Today (a) Monthly electricity production and reference demand curve by source today - Swiss Energyscope calculator (b) Daily electricity production and demand curves by source today - OFEN [25] Figure 4.1: Electricity production and demand - Today 24

31 4.2.3 Scenario 2035 The first horizon of this study is fixed to This year represent a major milestone in the Swiss energy transition since the all the nuclear power plants are expected to be decommissioned by Therefore, these sources, that generated almost 37% of the total production of the country in 2013, have to be replaced. Two variations of textscprognose for the electricity production are considered: High (C) and Low (E). These variations are detailed in what follows. The electricity generation technology is also assumed to evolve, in particular improving the CO 2 eq of the production and the efficiency of thermal power plants. Table 4.8 summarise these emissions and efficiencies for different technologies for Source Life-cycle emissions [tco 2 eq/gw h] Efficiency Coal % Natural Gas % Oil % Cogeneration % Hydro Power % Photovoltaic % Wind 9 100% Geothermal % Import (EU27 average) % Table 4.8: Life-cycle GHG emissions and efficiencies for various technologies Sources: Swiss EnergyScope, European Environment Agency As stated in the previous section, the vehicle fleet is assumed not to evolve. However the fleet average emission rate is supposed to be 95 gco 2 eq/km. This hypotheses is justified by the European legislation limiting the carbon emissions of new private cars to this value starting from The average fleet age being 8.3 years in 2014 [26], the average value 95 gco 2 eq/km for the fleet is assumed to be reached by The evolution of the energy prices is reported in table 4.9, moreover the subsidies (i.e. vehicle tax reductions) for EVs are assumed to cease by Type Units Price Electricity CHF/kW h Gasoline CHF/l 2.36 Diesel CHF/l 2.53 Natural Gas CHF/kW h Table 4.9: Energy and fuel prices Source: Prognose 25

32 Electricity production: Prognose 2035 High (C) The variant 2035 High (C) represents the most carbon intensive variation to the power shortage issue in the future. Most of the electricity demand is meet with indigenous production. In practice, supposing that the nuclear power plants are not replaced at the end of their service life as announced, new combined-cycle gas plants are constructed to compensate the production shortage. Figure 4.2 shows the monthly electricity production by source. It is observed that the major part of the production is assessed to be from hydro power plants or natural gas plants. Thus this scenario implies a high carbon intensity for the electricity production. That characteristic will highly impact the CO 2 eq emissions of the electric vehicle. Figure 4.2: Monthly electricity production by source in 2035 High scenario - Swiss Energyscope calculator Electricity production: Prognose 2035 Low (E) The variant 2035 Low (E) represents the variation with the most renewable implementation. Actually, in this case, no thermal power plants are constructed to face the electricity demand. One exception to consider is the combined heat and power (CHP) plants. If the indigenous generation is not sufficient, the electricity is imported from the European electricity market. This is expected o be the case for most of the year except during the summer (figure 4.3). Figure 4.3: Monthly electricity production by source in 2035 Low scenario - Swiss Energyscope calculator 26

33 4.2.4 Scenario 2050 The second horizon is fixed to The significance of this year lie in the fact that it also represents the final milestone in the "Energy Strategy 2050" of the Swiss Federal Government. As for the scenario 2035, two variations for the electricity production are considered as a continuation of those introduced in the previous section. New life-cycle CO 2 eq emissions and efficiencies are considered for electricity generation technologies. These new values are summarized in table Source Life-cycle emissions [tco 2 eq/gw h] Efficiency Coal % Natural Gas % Oil % Cogeneration % Hydro Power % Photovoltaic % Wind % Geothermal % Import (EU27 average) % Table 4.10: Life-cycle GHG emissions and efficiencies for various technologies Sources: Swiss EnergyScope, European Environment Agency Two values are considered for the average CO 2 eq emissions of the vehicle fleet, 75 gco 2 eq/km and 60 gco 2 eq/km. These two cases are studied since limited information is available concerning that matter for this time horizon. The energy prices are also updated for this scenario, new values are summarized in table Type Units Price Electricity CHF/kW h Gasoline CHF/l 2.57 Diesel CHF/l 2.74 Natural Gas CHF/kW h Table 4.11: Energy and fuel prices Source: Prognose 27

34 Electricity production: Prognose 2050 High (C) Even in 2050, in the High (C) variant of the electricity production, a significant share of the electricity production is done through natural gas power plants. These power plants represent 36% of the total generation, whereas hydro power plants represent 47% of the total generation (figure 4.4). Therefore this scenario still remains very carbon intensive. Figure 4.4: Monthly electricity production by source in 2050 High scenario - Swiss Energyscope calculator Electricity production: Prognose 2050 Low (E) On the contrary of the variant Low (E) for 2035, the 2050 version do not consider any import for the base electricity demand. This situation is achieved by both reducing the electricity demand and increasing the production capacity of renewable sources. The production even exceeds the demand in summer months due to the increase in hydro and photovoltaic production (figure 4.5). This variation is expected to be the most interesting in terms of carbon emissions and energy independence. Figure 4.5: Monthly electricity production by source in 2050 Low scenario - Swiss Energyscope calculator 28

35 4.3 Results & Discussion Electricity demand & primary energy consumption Electricity demand As expected, the electricity demand of EVs increases linearly with the penetration rate. At its highest, the additional demand increases by 14 T W h/year. This amount corresponds to the real demand from source, i.e. it includes the electricity consumption by EVs but also the transmission and distribution losses. It is observed that, at its maximum value, the electricity demand of EVs increases the base electricity demand of the country by 19% to 24% depending on the scenario. Figure 4.6: Electricity demand by EVs 29

36 Primary energy consumption The evolution of the primary energy consumption by passenger cars is particularly interesting for future scenarios. It is essential to remind that the primary energy consumption here represents only the amount of energy consumed during the use phase of the vehicles. The production energy of the vehicles does not take part of these calculations. It is also important to note that the baseline primary energy consumption changes across the scenarii due to the efficiency improvement if ICEVs. Starting from T J/year today, going down to T J/year in 2035 and to T J/year T J/year in 2050 for case (a) and (b) respectively. Figure 4.7: Primary energy consumption - Today As of today, it is seen that the primary energy consumption of passengers decreases with the increase of EVs penetration rate. That is explained due to the very low efficiency of current ICEV and the high electricity generation efficiency for the marginal production, which is mainly hydro power. The primary energy consumption decreases by 46% for a full penetration of EVs

37 (a) Primary energy consumption High (b) Primary energy consumption Low Figure 4.8: Primary energy consumption An interesting result is observed for the scenario 2035 High (figure 4.8a). It is seen that the decrease in primary energy consumption is less important. Several inflexion points are present while looking to the relative difference in primary energy consumption curve. These inflexion points correspond to a change in the marginal electricity production. The primary energy consumption difference curve has a lower slope compared to lower penetration rates. This indicates that the EV efficiency, including the whole electricity generation and transportation chain, is lowering closer to the ICEV efficiency, including refinery efficiency. A similar behaviour is observed for the scenario 2035 Low (figure 4.8b). In that case, the primary energy consumption decreases slightly more. A decrease of more than 23% is observed compared to the baseline with 0% EV penetration. 31

38 (a) Primary energy consumption High (a) (b) Primary energy consumption Low (a) Figure 4.9: Primary energy consumption (a) In 2050, the primary energy consumption of EVs is highly sensitive to the energy scenario. For the (a) case of the scenario 2050 (ICEV fleet average emissions 75gCO 2 eq/km), only a slight improvement is observed for the High scenario over the whole range of EV market penetration (figure 4.9a). Moreover, an optimum point at 87% EV penetration rate is observed for which only a 3.5% decrease in primary energy consumption compared to the 0% EV penetration baseline is observed. For scenario 2050 Low (a), a clear advantage in terms of primary energy consumption is observed for the EVs. At its highest penetration rate, the primary energy consumption decreases of about 22% (figure 4.9b). The justification of this difference between these scenarii comes directly from the marginal electricity production difference. For the 2050 High scenario, the marginal electricity production for EVs is mainly from natural gas or import. Whereas the marginal electricity production in 2050 Low scenario is mainly cogeneration, hydro power and photovoltaic. Thus the overall electricity generation efficiency is much higher in the latter scenario. 32

39 (a) Primary energy consumption High (b) (b) Primary energy consumption Low (b) Figure 4.10: Primary energy consumption (b) The situation is however different for the scenario 2050 (b), for which the ICEVs are expected to have higher efficiencies (ICEV fleet average emissions 60gCO 2 eq/km). This consideration impacts highly the change in primary energy consumption with the introduction of EVs. In the scenario 2050 High (b), the EVs clearly loose their efficiency advantage, increasing the primary energy consumption up to 20% for a complete market penetration (figure 4.10a). In this particular scenario, the overall efficiency of the ICEVs is better than the overall efficiency of EVs. The 2050 Low (b) scenario presents also an optimum point in terms of primary energy consumption. This point is situated at 87% EV market penetration and the corresponding decrease in primary energy consumption is 3.5% (figure 4.10b). For higher penetration rate, the primary energy consumption increases due to the imported electricity, which has a lower generation efficiency. 33

40 4.3.2 Energy independence (a) Share of final energy importation for passenger cars (b) Share of primary energy importation for passenger cars Figure 4.11: Share of energy importation for passenger cars Two variety of energy independence are studied: final energy independence (figure 4.11a) and primary energy independence (figure 4.11b). It is observed that this distinction leads to greatly different results. From these variations, the primary energy independence represents the real energy independence of a country. The final energy independence is reported for analytic purposes. The final energy independence improves for all scenarii even if slight differences remains on the magnitude. It is observed that the worst case scenario is 2035 Low, in which case the final energy import represent 26% of the final energy consumption from passenger cars for an EV market penetration of 100%. The best case scenario is 2050 Low (both variations), where only 1% of the final energy is imported supposing a full penetration of EVs. As announced previously, the primary energy independence leads to different results. In these estimations, it is assumed that the natural gas used in natural gas power plants and the cogeneration plants is completely imported. However, a small share of this gas may be indigenous in the future. Since the size of the exploitable reserves remains uncertain at the time of this study, this situation is not considered. In 2050 High scenarios, the import share drops only to 98% of the total primary energy consumed for passenger cars. The best case scenario is the 2050 Low, where the import share drops down to 35%. These 34

41 results are explained by the fact that in 2050 High scenario, the marginal electricity production is mainly from natural gas power plants of cogeneration power plants, and both of these plants use imported natural gas as fuel. In contrast, in scenario 2050 Low, a non negligible share of the marginal electricity is produced through photovoltaic and hydro power plants, which produce completely indigenous electricity. In 2013, the energy independence of Switzerland was only about 22%. This same indicator shows that the energy independence, considering only private cars, may be significantly improved with the increase of EV market penetration. However, this conclusion is valid only under some circumstances. The energy independence shows an interesting improvement if the energy mix comprises a large share of renewable energy sources, as it is the case in both 2035 and 2050 Low scenarii. In terms of energy supply security, the increase of energy independence could be interpreted as an improvement of energy security. Nevertheless, this is not exact true due to the difficult storage of electricity. As presented in the previous paragraph, the energy independence improves with the integration of renewable energy source in the mix. A key characteristic and drawback of these sources, especially for wind and solar power, is that the production is variable and non-controllable. Therefore, the security of supply is guaranteed but methods to efficiently store and manage this production are needed. It has been observed that the energy independence would not be substantially improved with the market penetration of EVs in High scenarii. The security of supply is also uncertain due to the limited storage capacity of natural gas. Therefore, the gain in energy security supply is not clear. Lastly, the worst cases in terms of energy supply security are observed for the scenario 2035 Low and 2035 High (is today s situation is put a part). The share of final energy import is represents respectively 26% and 17% of the total final energy consumed to power private vehicles with an EV market penetration rate of 100%. In other words, that means up to 26% of the electricity that powers all the private cars in the country is imported. This represents a huge risk in energy supply security due to the non-storable nature of electricity. In conclusion for that part, it is observed that the energy independence increases in all scenarii with the increase of EV market penetration. In worst case, it remains unchanged compared to today. The same goes for the energy supply security, except in particular cases discussed in the previous paragraph, which is improved with the introduction of EVs. An interesting fact to underline is that the improvements in both energy independence and energy supply security are achieved only by variations on the demand side. 35

42 4.3.3 Climate change: CO 2 eq emissions (a) CO 2 eq emissions over total life-time - Today (b) CO 2 eq emissions over total life-time (c) CO 2 eq emissions over total life-time Figure 4.12: CO 2 eq emissions over total life-time The life-time CO 2 eq emissions are highly sensitive on the scenario, for both EV and ICEV. This behaviour is expected for ICEVs since the largest part of the CO 2 eq emissions occur during the use phase of the vehicle. In the case of the EVs, the CO 2 eq emissions produced during the operation of the vehicles becomes predominant in scenarii with carbon intensive electricity generation (i.e High and 2050 High). 36

43 The results for today shows a clear advantage for the EV in terms of CO 2 eq emissions predominantly due to the poor efficiency of ICEVs (figure 4.12a). Even without considering a second-life for he batteries, the life-time CO 2 eq emissions of the EVs remain significantly below the CO 2 eq emissions of ICEVs, no matter what the battery chemistry is considered. The results for the 2035 scenarii show a considerable improvement for the ICEVs. The worst case scenario in terms of electricity carbon intensity (High) combined to the most carbon intensive battery chemistry (LFP) leads to a life-time CO 2 eq emissions very close to the average diesel ICEV of 2035, 13.7 tco 2 eq and 14.3 tco 2 eq respectively (figure 4.12b). However, it is necessary to underling that the present results do not consider any improvement for the carbon intensity of EV battery packs. Since it is very plausible that improvements will occur for that matter, the results here represent the worst-case scenario for battery production. The impact of efficiency improvement of ICEVs is even more pronounced in 2050 scenarii. In that case the worst case scenario for EVs for both battery production and electricity generation leads to a higher life-time CO 2 eq emissions compared to ICEVs (figure 4.12c). Nevertheless, these results do not consider the inevitable increase in ICEV production CO 2 eq. To achieve such efficiencies, the ICEVs must go through the hybridisation of the power train. Proceeding that way, the production CO 2 eq of ICEVs would increase significantly lowering or even suppressing its advantage against EVs. (a) CO 2 eq emissions - Today (b) CO 2 eq emissions High (b) (c) CO 2 eq emissions Low (b) Figure 4.13: CO 2 eq emissions The annual fleet CO 2 eq emissions of ICEVs and EVs follow logically the same behaviour presented previously with the life-time CO 2 eq emissions (figure 4.13). Only the extreme case results are presented for this indicator. The clear advantage on a fleet level is clearly observed for today s situation. In the future, this advantage is lowered in carbon intensive electricity generation scenarios, whereas this advantage is conserved for low carbon intensive electricity generation scenarios. As mentioned previously, these results dot not consider any improvement in the carbon intensity of EV battery packs and do not consider higher CO 2 eq emissions due to the hybridisation of ICEV for the production phase in the future. Moreover, it has been observed that the change in charging pattern as explained in do not have a significant impact on the CO 2 eq results. 37

44 (a) CO 2 eq emissions share for private cars - Today (b) CO 2 eq emissions share for private cars High (b)(c) CO 2 eq emissions share for private cars Low (b) Figure 4.14: CO 2 eq emissions share for private cars An interesting metric is the share of the total CO 2 eq emissions that can be imputable to private vehicles (figure 4.14). This share is 25% today in a configuration without EVs in the fleet. This share could possibly reduced down to 10% in worst case situation today and below 5% in the best-case situation considering a second-life for the EV batteries (figure 4.14a). The share of CO 2 eq emissions imputable to private vehicles decreases in carbon intensive scenarii even before the introduction of EVs thanks to the efficiency improvement of ICEVs. The situation is different for Low scenarii where the total CO 2 eq emissions is expected to decrease. Therefore the share imputable to private vehicles increases in these scenarii. Even though the EVs have the potential to reduce significantly CO 2 eq emissions today, the situation is highly dependent on the energy mix in the future. In relative terms, the potential of CO 2 eq reduction through private cars remains limited two conditions come together, carbon intensive electricity mix and efficient ICEVs. On the contrary, if the electricity mix is poor in carbon and the private vehicles account for notable share in total CO 2 eq emissions, the reduction potential through private vehicles becomes highly attractive. These situations are observed in all Low scenarii and also in current situation. It is important to note that the share of CO 2 eq emissions are higher in Low scenarii due to the projected emissions decrease in other sectors. In comparison, in High scenarii, the CO 2 eq emissions of other sectors are not expected to experience the same decrease. Thus the share attributable to private cars is less important in any case. 38

45 4.3.4 Urban pollution (a) Vehicle fleet tailpipe PM emissions (b) Vehicle fleet tailpipe NO x emissions Figure 4.15: Urban pollution indicators The urban pollution is represented by two quantities, the particulate matter emissions and the NO x emissions. As explained in the methodology, only tailpipe emissions are considered since the aim is to measure the urban pollution. Two situation are compared, today and future scenarii. No distinction has been made between the future scenarii since it is assumed that the emission standards will not be lowered further. The beneficial impact of EVs on these emissions is emphasized for the actual situation due to the composition of the fleet that still includes old generation engines. It is observed that half of the PM emissions can be cut with 35% EV market penetration (figure 4.15a). The results for the NO x emissions shows the same behaviour, the NO x emissions can be lowered by 50% with 31% EV market penetration (figure 4.15b). Since no engine generation category has been differentiated for future scenarii, the impact of EV market penetration is purely linear. 39

46 Figure 4.16: Total cost of ownership - Detailed costs for today Figure 4.17: Purchase cost - Today Economic impact for the end user Total cost of ownership The model results for the total cost of ownership (TCO) shows that the average electric vehicle is slightly more expensive than the average gasoline vehicle, but slightly cheaper than the average diesel vehicle (figure 4.16). As expected, the capital cost is higher for the EV mainly due to the cost of the battery but also because of the private charging point (figure 4.17). The business model used by Renault for the Zoe (also proposed by Nissan for the Leaf) model seems interesting since it helps to reduce the capital costs, costs to which the customer may be more sensitive. Comparatively, it is observed that for all vehicles, the most important part of the TCO is related to the vehicle maintenance costs. Even though it was assumed that the service and reparation costs of EVs are 40% less compared to the ICEVs, it represents a non-negligible share of the TCO because of the high catalogue price of EVs. Nevertheless, it is observed that, as expected, the operation costs are lower for the EV compared to both gasoline and diesel vehicle. The difference is significant if the mineral oil tax (MOT) is not transferred to EVs. However, the TCO of the EV is above both gasoline and diesel vehicle if the MOT is completely transferred. It is seen that the reduction in battery pack costs are playing a significant role in the competitiveness of the EVs (figure 4.18). The results show that by 2020, the TCO of the EV is expected to be at the same level of an average gasoline ICEV even for the worst case estimation of battery pack costs. Yet, the break-even point with the diesel ICEV is in 2035 for the worst case estimation if the MOT transfer is considered. Notably, it is observed that the TCO of gasoline vehicles is expected to decrease in the future. It suggest that the increase in efficiency is counterbalancing the extra costs due to additional manufacturing costs and the increasing price of fuel. This effect is particularly pronounced for the most efficent car in 2050 (curve 40