A study aimed at assessing the potential impact of vehicle electrification on grid infrastructure and road-traffic green house emissions

Applied Energy 120 (2014) 31–40

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

A study aimed at assessing the potential impact of vehicle electrification on grid infrastructure and road-traffic green house emissions Marco Sorrentino ⇑, Gianfranco Rizzo, Luca Sorrentino Department of Industrial Engineering, University of Salerno, Fisciano, Salerno 84084, Italy

h i g h l i g h t s  Modular computational structure for extensive analyses of car electrification impact.  Simplified market model to assess the CO2 reduction potentialities of PEVs.  Combining macro- with micro-level analyses of car electrification impact.  Developing suitable methodologies to support energy planning for transportation.  Worldwide analysis of electricity mix influence on CO2 reduction potential of electrified cars.

a r t i c l e

i n f o

Article history: Received 14 May 2013 Received in revised form 23 December 2013 Accepted 12 January 2014 Available online 7 February 2014 Keywords: Plug-in hybrid vehicle Car electrification Electric vehicle Grid decarbonization Greenhouse gas emissions

a b s t r a c t In the current paper a thorough analysis is conducted to assess, on one hand, the impact of vehicle electrification on electric grids and their related infrastructures, and, on the other, its potential contribution to GHG emission reduction. Such an analysis covers the timeframe 2011–2050, thus allowing to assess if the environment friendliness of both PHEV and BEV will be enough contributing, particularly towards the fulfillment of the objectives recently established both by official agreements among governments and research consortia (e.g. the International Energy Agency) as well. The expected time evolution of both PHEV and BEV private car fleets is modeled through a simplified market penetration model, along with the associated contribution in terms of well to tank and tank to wheel GHG emissions, thus providing the needed input data to the scenario analysis. Particularly, a longitudinal vehicle model is adopted to accurately estimate electric vehicle energy consumptions and related GHG emissions as a function of powertrain configuration, dimensions and mass. The analysis was run on several countries, thus providing useful outcomes to assess the suitability of given energy mix to fully exploit vehicle electrification. Such indications will therefore be useful to determine to which extent progressive decarbonization of current grids is required to meet the GHG reduction target by 2050. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction In the last years, several political, social and economic issues have been contributing to strengthen the willingness to achieve a new, more environmentally friendly and sustainable mobility paradigm worldwide, particularly aiming at meeting actual mobility demand without constraining development expectations of future generations [3]. The most pressing arguments toward the finding of new solutions for personal mobility mainly include: Fossil fuels depletion; CO2-related greenhouse effects, with dangerous and maybe dramatic impact on global warming and climate changes; ⇑ Corresponding author. Tel.: +39 089964100; fax: +39 089964037. E-mail address: [email protected] (M. Sorrentino). http://dx.doi.org/10.1016/j.apenergy.2014.01.040 0306-2619/Ó 2014 Elsevier Ltd. All rights reserved.

worldwide increasing demand for personal mobility, especially in growing countries such as the BRICs. In this context, the electrification of automobile represents today a major research track for both the industry and academia towards a sustainable mobility. The use of electric energy as an energy carrier for passenger cars has the potential to decrease pollutant emissions in urban areas and, according to the adopted generation mix, also to reduce greenhouse gas (GHG) emissions from transportation. Moreover, new development opportunities are emerging, based on the growing interdependency between the transportation sector and stationary electric power generation, the latter also being increasingly based on renewable sources [4]. Such an aspect is of particular relevance today due to the upcoming introduction of plug-in hybrid electric vehicles (PHEVs) and battery electric vehicles (BEVs), particularly

32

M. Sorrentino et al. / Applied Energy 120 (2014) 31–40

Nomenclature BEV BRIC CC CD CS EG EM GHG HEV I ICE IEA Mln NA NEDC NPI PEV PHEV SOC SOCup SOClo UNISA Symbols A CO2,grid Cr Cx Dyear EC FE

battery electric vehicle Brazil, Russia, India, China conventional car charge depleting charge sustaining electric generator electric motor green house gas hybrid electric vehicle inverter internal combustion engine International Energy Agency million not available new European driving cycle non plug-in vehicles (i.e. CC + HEV) plug-in hybrid-electric & pure-electric vehicle (i.e. PHEV & BEV) plug-in hybrid electric vehicle state of charge maximum SOC in charge sustaining operation minimum SOC in charge sustaining operation university of Salerno

2

frontal area (m ) CO2 emissions yielded by electrical grid (g/W h) rolling resistance wind resistance annual distance (km) electricity consumption (W h/km) fuel economy (km/l)

with the prospective diffusion of vehicle-to-grid (V2G) technologies [5,6]. Therefore, the urgent need to transform policy statements into concrete action has prompted the International Energy Agency (IEA) to develop a series of roadmaps, for some of the most important technologies, aimed at outlining the correct developmental paths to be followed to achieve a substantial reduction of energy-related CO2 emissions [7]. These roadmaps also include special focuses on developing technologies in emerging economies, thus enhancing international collaboration with the final aim of strengthening the efforts toward the targeted global reduction in greenhouse gas emissions. The target set by the IEA ‘‘blue-map’’ scenario [7] corresponds to an overall reduction of 50% of global energy-related CO2 emissions by 2050, as compared to current levels. In this context, the transport sector is requested to contribute by guaranteeing 2050 CO2 emissions be 30% lower than current value. On the other hand, the roadmaps traced for plug-in vehicles (PEVs) market are functional to effectively achieving blue-map objectives. Achieving such results requires that battery electric vehicles (BEV) and plug-in hybrid electric vehicle (PHEV) technologies for passenger cars evolve rapidly over time, with very aggressive rates of market penetration. Particularly, the PEV rate of penetration is influenced by several factors: technology supplier, vehicle offers, charging infrastructure features and costs, increase in maximum grid power (i.e. to ensure PEV, and especially BEV, be charged when needed) and customers demand. Since government policies strongly influence these factors, their role will be crucial to guarantee rapid electrification of the fleet [8,9]. Therefore the next decade will be a key

M Mbody Mc mEG mEM mgear mICE NB Ncar P t Tch,BEV Tch,PHEV UF v

mass (kg) vehicle body mass (kg) mass of a single battery cell (kg) EG unit mass, equal to 0.83 (kg/kW) [1] EM+Inverter unit mass, equal to 1 (kg/kW) [1] gearbox unit mass, equal to 0.48 (kg/kW) [2] internal combustion engine unit mass, assumed equal to 2 (kg/kW), on average, for SI engines [2] number of battery cells private car fleet population (mln) power (W) time (s) BEV charging time (h) PHEV charging time (h) utility factor vehicle speed (m/s)

Greek symbols a road grade (rad) g efficiency q ambient air density (kg/m3) qPtW power to weight ratio Subscript B f r T

battery final road transmission

Superscript * nominal value

‘‘make or break’’ period for BEVs and PHEVs: governments, the automotive industry, electric utilities and other stakeholders must work together to implement the vehicles and infrastructure in a coordinated manner and, furthermore, to ensure that the car customers will be ready to buy. The hope is that the IEA roadmaps will add further attention and urgency to the international debate on the importance of PEV propulsion as a viable technical solution to well-known environmental and energy saving issues. The current worldwide road transport scenario is largely based on vehicles powered by internal combustion engines, which rely almost exclusively on petroleum-based fuels and are responsible for over 74% of the overall GHG amount emitted by the entire transport sector [10]. The use of electric energy as a fuel for passenger cars has the potential, on one hand, to contribute toward the reduction of corresponding GHG emissions and, on the other, to create new development opportunities based on the growing interdependency between transport sector and stationary electric power generation [11,12]. Such an aspect is particularly relevant today due to the expected fast increase of PHEVs [13,14] and BEVs [15] into the market. Abundant literature is now available on the potential benefits associated to vehicles’ electrification on worldwide mobility. Many researchers concentrated their efforts on verifying technical and economical feasibility in the short to long term scenarios, mainly focusing on analyzing [16–19] and eventually improving [11,20,21] the vehicle–grid interaction. Others focused on the social [9,22] and environmental [23] impact of car electrification, as well as on the psychological barriers [8] to be overcome to make electric vehicles deployment really and effectively take place in the near

33

M. Sorrentino et al. / Applied Energy 120 (2014) 31–40

future. Nevertheless, as outlined in [12], it is also important to analyze the generation mix influence on real-world GHG reduction effects linked to electric vehicle diffusion. Therefore main objective of this study is to investigate the impact played by different energy mixes, as well as of expected economic growth, on car electrification benefits in terms of CO2 reduction potential in the timeframe 2011–2050 for several countries. It is also worth remarking at this point how the methodologies, here proposed to deal with both macro- and micro-aspects, were conceived in such a way as to develop a modular computational structure, which is suitable both for preliminary assessments and successive upgrading of numerical analyses’ level of detail. Particularly, the results achieved and discussed in the current study, which are based on simple models of macro- and micro-aspects, are to be considered as a preliminary starting point, to be referred to when aiming at carrying out new, more in-depth analyses, which in such a case will have to rely on the integration, within the proposed modular computational structure, either of more advanced predictors (i.e. with increased statistical features) or more up-to-date information and data. In the latter case, the proposed modeling structure is potentially upgradable, in such a way as to enable the monitoring of time evolution impact of vehicle electrification growth, both in terms of new grid infrastructure requirements and CO2 emission reduction. The paper is structured as follows: the simplified approach followed to describe the first macro-aspect accounted for in this analysis, namely the expected penetration of PEV in the investigated timeframe is presented and discussed in Section 2; Sections 3–5 focus on the micro-aspects analyzed in this study, namely the modeling approaches followed to describe the BEV and PHEV cars representative of the selected fleet (i.e. private cars) and estimate the associated fuel saving contributions; then, Section 6 outlines how the other relevant macro-aspect, i.e. the estimation of the time varying impact of car progressive electrification on grid infrastructure and CO2 emission levels, is treated in this paper. Finally, the results obtained through the simultaneous processing of the above-introduced macro- and micro-aspects models are presented and discussed. 2. Market penetration of electrified vehicles In order to predict the rate of PEV market penetration, the consistency of the private car fleet was evaluated via a simplified market penetration model, which is based both on current data and forecasts for future growth. In particular, the evolution of the fleet operating in Italy, which currently consists of about 37 million vehicles, is estimated at 41 million for 2035 [24,25]. Thus, the fleet evolution over the timeframe 2011–2050 was estimated by linearly extrapolating the above-mentioned data. In accordance with what was already proposed by other authors in the literature [25], the latter assumptions were also applied to other European countries, namely Germany and France, and to USA and Japan, to whom the analyses the current paper focuses on were extended. Specifically, the same time evolution of Italy was considered, but of course starting from the respective initial population of private cars, as described in Table 1. As for France and Japan, due to much slower or even null population growth observed in recent years, a Table 1 Private car fleet populations [24,26–28]. Country

Ncar (2011) (mln)

Projected Ncar (2050) (mln)

Italy Germany USA France(1)|France(2) Japan(1)|Japan(2) China

37 42 137 32,5|32,5 41|41 60

43 49 160 38|32,5 48|41 216

further case (labeled as France(2) and Japan(2) in Table 1) was assumed, i.e. corresponding to zero increase of private car fleet dimension over the timeframe 2011–2050. On the other hand, a different approach had to be followed for China. The strong socio-economic development that is affecting such country is significantly influencing the growth of the private car fleet, which is expected to almost quadruple by 2050 with respect to the current status (see Table 1) [26–28]. Once the overall population of private car fleet is known, the evolution of number of PEVs can be estimated as a function of time and rate of market penetration. The latter can be evaluated starting from the actual share of such vehicles within the global fleet. Although several automakers have already introduced some plug-in vehicles into the market, their incidence in the entire fleet is very limited. This allowed to assume the initial percentage in 2011 equal to zero. In order to obtain a forecast scenario, the percentages addressed by [24] for 2030 were assumed to derive a linear evolution over time of both PHEV and BEV share, as follows:

%PHEVðnÞ ¼ %PHEVref  %BEVðnÞ ¼ %BEVref 

n  n0 nref  n0

n  n0 nref  n0

ð1Þ

ð2Þ

where %PHEV(n) and %BEV(n) represent the share of BEVs and PHEVs at year n within the private car fleet, respectively; n0 is the base year (i.e. 2011), while nref is the reference year (i.e. 2030) at which %PHEVref and %BEVref equal 13% and 7%, respectively [24]. It is worth remarking here that the same percentage evolution over the timeframe 2011–2050 was assumed for all analyzed countries. Such a choice is justifiable recalling once again the fact that as for 2011, the percentage of HEV and PEV cars is negligible worldwide, whereas the projections over the timeframe 2011–2050 are based on the general provisions given by the IEA in the roadmap described in [2]. One point to be carefully taken into account is related to available lithium resources. Indeed the significant increase of PEVs projected by the IEA in [4] could be hindered by the likely to happen lithium shortage worldwide. Therefore, the validity of the road-map defined by IEA, as well as the outcomes of the results presented in the current paper, will rely, on one hand, on proper use (e.g. battery recycling [29]) and commercialization of lithium resources [30] and, on the other, on the development of non-lithium battery technologies that still meet specification requirements of PEVs. As for the simplicity of the proposed market penetration model, it is worth noting how assuming similar trends for most of the analyzed countries allows a useful comparison of current generation mix effect on CO2 reduction obtainable with increased vehicle electrification. Moreover, the macro-aspect associated to number of private car fleet evolution over time is surely a very challenging task, which requires considerable research efforts on its own. Within this study, it was decided to keep it as simple as possible in order to mostly concentrate on the mutual interrelation between such a macro-aspect and the other micro- (i.e. assessment of PEV fuel saving potential) and macro- (i.e. effect of improving generation mixes worldwide) aspects, presented and discussed in the following sections. In such a way, it is possible to demonstrate, within this study, the suitability of the proposed modular computational structure to run more advanced and future monitoring analyses, to be based on new sub-models of both micro- and macro-aspects, as well as on new information and data on technological and socio-economic trends. 3. Vehicle mass model In order to suitably evaluate and compare fuel economy and electricity consumption of PHEVs and BEVs, a parametric model

34

M. Sorrentino et al. / Applied Energy 120 (2014) 31–40

4. Longitudinal vehicle model Currently achievable fuel economies in this paper are evaluated by means of a backward longitudinal vehicle model developed in MatlabÒ environment [31,32], whose basic equations are presented below. Road load is estimated as follows:

Pr ¼ M PHEV  g  v  ½C r cosðaÞ þ sinðaÞ þ 0:5qC x Av 3 þ M eff 

dv v dt

ð7Þ

where v is the vehicle speed and the term Meff equals 1.1 times the PHEV mass MPHEV to suitably account for rotational inertia. For nonnegative Pr values, the mechanical power requested to the EM is (see Figs. 1 and 2):

Fig. 1. Schematic representation of a typical series PHEV powertrain.

PEM ¼

Pr

if P r P 0

gT

ð8Þ

PEM can also be expressed as a function of the power supplied by electric generator and battery, as follows:

PEM ¼ gEM ðPICE gEG þ PB Þ if Pr P 0

ð9Þ

On the other hand, when Pr < 0, the regenerative braking mode is active, resulting in the following expression for the electrical energy delivered by the EM:

PEM ¼ Pr gT gEM

where M and m terms correspond to mass and unit mass values, respectively. Then, either BEV or PHEV mass can be determined adding the respective hybridization devices (see Figs. 1 and 2) and by imposing that the electrified vehicle achieves similar performance to the original conventional car (see Eq. (6)), as follows:

qPtW;i ¼ Mi ¼

PICE;CC M CC

PEM;i

ð5Þ

1 0.8

ð6Þ

0.6 0.4

0 -0.2 -0.4 -0.6

Normalized EM efficiency map 0.83443 0.88962 0.94481

0.83443 0.88962 0.94481

0.9 0.8

0.94481 0.94481 4 81 0.94 89 62 0.8

0.7

0.88962 0.94481 0.60.60.70.7 0.94 481 0.88962 0.94481 13 68 247 09 24 0.880.61367 0.8344 962 6329 0.50 3 0.44 0.83443 0.77924 0.77924 0.72405 0.72405 0.66886 0.61367 0.66886 0.3929 81 73 0.3377 0.50329 0.55848 0.55848 0.50329 1 0.2825 0.4481 0.39291 0.39291 0.4481 286 50.39291 0.28253 0.33772 0.28253 0.33772 0.22734 0.17215 0.22734 0.17215 0.22734 0.17215 0.17215 0.17215 0.17215 0.22734 0.22734 0.22734 0.28253 0.28253 0.28253 0.33772 0.33772 0.4481 0.33772 0.39291 0.39291 0.4481 0.55848 0.50329 0.55848 0.61367 0.66886 0.66886 0.72405 0.61367 0.61367 0.66886 0.72405 0.83443 0.72405 0.77924 0.83443 0.77924 0.88962 8 48 240.50329 962 0.88 7943 0.734 0.94481 0.55 1 0.9448 0.8

0.2

3 34 4 0.8 24 0.779 05 0.724 0.668 86

where the coefficient k (see Eq. (4)) equals 1 and 0 in case the subscript i (see Eqs. (4)–(6)) corresponds to PHEV and BEV, respectively. For further details about the above described weight model and the way the resulting system of equation is solved, the reader is addressed to a previous publication [31]. It is finally worth remarking here that the availability of the above vehicle mass model will allow to take into account the effects of different sizing of powertrain components on vehicle mass, such as batteries of different technology and capacity. Moreover, it allows easily extending the analyses presented hereinafter to different vehicle fleets, such as Light-, Medium- and Heavy-Duty trucks.

4 43 0. 83

43 0.834

qPtW;i

ð4Þ

0.61367

M i ¼ M body þ k  PICE;i ðmICE þ mEG Þ þ PEM;i mEM þ M c NB;i

ð11Þ

0.889 62 0. 94481

ð3Þ

if Pr < 0

where gEG is electric generator efficiency, here assumed equal to 0.93 [1]. BEV was modeled hereinafter extending Eqs. (7)–(11) to the powertrain shown in Fig. 2. Such extension is simply obtained by letting the PICE terms vanish in Eqs. (9) and (11). It is worth remarking here that EM and ICE efficiency are computed by means of the normalized maps shown in Figs. 3 and 4, respectively. Particularly, EM efficiency was derived by the model library proposed in [33], whereas ICE efficiency was inferred from experiments performed at UNISA engine test bench on a Fiat 1.2 l SI engine. Inverter efficiency was included in the just described EM normalized map. On the other hand, battery performance and efficiency (i.e. open circuit voltage and internal resistance estima-

Normalized motor torque

M body ¼ M CC  PICE;CC  ðmICE þ mgear Þ

PB ¼ PEM  PICE gEG

-0.8 -1 0

0.1

0.2

0. 88 9 62 0.94481

was set up to assess the impact of electrification on vehicle mass. By referring to the vehicle architectures shown in Figs. 1 and 2, the mass of an electrified vehicle can be obtained by adding the weights of the major vehicle components to the vehicle body mass. The latter is derived from the mass of the reference conventional car by subtracting the contributions due to the original gearbox and ICE from the original mass of the conventional car, as follows:

ð10Þ

where gT and gEM are the transmission and electric motor efficiency, respectively. During regenerative braking, PHEV battery can be charged by ICE also; thus, the following equation holds for negative Pr values:

0.6 0. 6163 67 0.728 86 0 0.774 9 25 4

Fig. 2. Schematic representation of a typical BEV powertrain.

if Pr < 0

44 81962 0.90. 88

81 44 0.9

0.5

0.4 0.3

0.88962 0.83443

0.4

0.5

0.94481 0.88962 0.94481 0.94481

0.94481

0.94481

0.3

0.6

0.88962 0.83443

0.6

0.7

0.8

0.9

0.2 1

Normalized motor speed Fig. 3. Normalized EM map. Maximum efficiency is lower than 0.9 and higher than 0.85.

35

M. Sorrentino et al. / Applied Energy 120 (2014) 31–40

0.3 0.2 0.1 0

0.1

0.908 0.8 12 62 1 8

4 06 0.95

0.816 25 0.77031 0.724 37 7843 0.6320.6 49 0. 58655 0.540 61 0.4 94 0. 448767 0.40248 0.356 86

0.2

0.3

0.4

0.5

0.9

5 18 62 62 81 31 0.8 0. 770 7 . 0 724 3 3 0. 78 4 0.6 2 49 55 0.630.58640 61 0.54 67 0.498 74 0.404.402886 6 0.35

0.6

0.7

0.8

0.8 0.7

0.63 2 0.58 49 0.54 6 55 0 61

0. 77 0.7 0 3 2 1 0.6 4 37 78 43 0.5 8 0.56 55 0.4 40 94 61 0 67 0..3402 8 0.02.3516086 64 9 00.1.21998 2 73 104

0.9 08 1 2

Normalized engine torque

0.4

0.9 08 12 0.9 54 06

6 54 0 0.9

0.5

5 62 81 0.

0.6

12 0.908

0.7

18 0.862

0.8

12 08 0.9

55 0.67843 86 0.5 49 32 0. 72437 0.6 0 31 0.77 25 16 0.8 8 62 1 0.8

0.9

0. 86 21 8

Normalized ICE efficiency map 1

0.6 0.5 0.4 0.3 0.2

8 64 9 0. 2

0.9

0.1 1

Pr

Normalized engine speed Fig. 4. Normalized ICE map. Maximum efficiency is lower than 0.35 and higher than 0.3. Fig. 5. Qualitative description of thermostatic management for a series PHEV.

tion as a function of SOC) were modeled via the lithium battery sub-model developed in [31], to which the reader is addressed for further details, and referring to the electrified vehicles-oriented battery specifications detailed in [34]. The above-mentioned normalized map are converted in dimensional efficiency map, as a function of ICE and EM power installed on PHEV and BEV cars (see Table 3), prior to be used to perform numerical estimation of electricity and fuel consumption, as detailed later on in Section 5. For instance, once the nominal mechanical power and motor speed range are known for the EM, the efficiency map shown on Fig. 3 can be exploited as follows: multiplying maximum speed by 0.25, the x-axis value, at which maximum power is reached in the constant power efficiency plot region, can be computed; then by dividing maximum power by such a speed value, the maximum torque results, thus allowing to completely de-normalize the efficiency map shown in Fig. 3. It is finally worth mentioning here how normalized efficiency models are used widespreadly in the literature dealing with hybrid vehicles modeling and energy management. Particularly, the contributions of Guzzella and Sciarretta [35] and Rizzoni et al. [36] clearly demonstrate how normalized models allow performing reliable comparison of powertrains with different degree of hybridization. Particularly, normalized efficiency approach enables appropriately evaluating the benefits achievable by vehicle hybridization, such as ICE downsizing, and, moreover, suitably weighing these benefits against drawbacks, such as increased weight due to vehicle electrification. 4.1. Vehicle energy management BEVs rely on the sole battery contribution for electric propulsion, thus requiring trivial energy management strategies. On the other hand, the availability on PHEVs of both an electric generator and a battery pack makes the energy management task absolutely non-trivial, thus motivating several researchers active in the automotive field to propose highly advanced control strategies to maximize fuel and electricity savings [37,38]. Usually [13], the PHEV battery capacity is firstly exploited to run the PHEV powertrain as a pure electric vehicle (see Fig. 5), thus operating the battery pack in charge depleting (CD) mode. Then, once the minimum state of charge SOClo has been reached, the PHEV control strategy switches to the charge-sustaining (CS) task, through which the targeted final state of charge SOCf is reached at the end of the cycle. Specifically in this paper, the adoption of a series architecture for PHEVs (see Fig. 1) allows adopting a simple thermostatic strategy to accomplish the CS control objective [39]. Thermostatic CS control was actually proven to be highly effective as compared to ideal fuel economies, which are achievable by means of dynamic

programming optimization techniques based on a priori knowledge of driving schedule [31]. In the current work, it was imposed that the ICE is always operated at the maximum efficiency point, which for the particular engine considered hereinafter (see Fig. 4) corresponds to 50% of its nominal power. This assumption is reasonably near to the optimum operation in terms of fuel consumption. In a previous contribution [40] the authors themselves proved that this is not necessarily the best strategy, and that further energy savings can be achieved by optimizing engine operation point as a function of vehicle power demand. It is also worth remarking here how the adoption of a sub-optimal energy management strategy, such as the thermostatic CS control, inevitably leads to a slight underestimation of potential benefits achievable by an increasing PEV market penetration, of the order of 3% lower fuel savings with respect to what could be achieved via more advanced PEV control strategies [31]. Fig. 5 also provides a qualitative description of thermostatic control strategy in CS mode. Following the initial charge depletion from fully charged (i.e. SOC = 1) down to SOCf, the battery is initially further depleted until battery SOC becomes lower than a given threshold SOClo = SOCf  0.01. At this point ICE-EG is turned on at the assigned power level and switches off when the upper threshold SOCup = SOCf + 0.01 is reached. The procedure is repeated until the end of the driving cycle, thus ensuring charge-sustaining operation of the battery pack be guaranteed over the entire trip. The simple thermostatic strategy shown in Fig. 5 can be further improved by optimizing the parameters SOClo and SOCup, as well as the ICE-EG power level, depending on the specific features of the current driving cycle. The authors themselves demonstrated the attainable improvements, highlighting the significant benefits that can be achieved even if no a priori knowledge of driving cycle features is available [40]. 5. Estimation of electricity and fuel consumption for electrified vehicles The longitudinal vehicle model described in Section 4 was applied to estimate currently achievable fuel economy (FE) and electricity consumption (EC) for PHEV and BEV, respectively. The simulations were performed on the NEDC driving cycle (see Fig. 6), by considering the vehicle specifications given in Table 2. It is worth pointing out here that the data shown in Table 2 refer to a medium-size passenger car with performance, mass and comfort similar to the most relevant models of both BEV [15] and PHEV [13], which are currently available or ready to appear on the car market. Although it corresponds to a significant simplification,

36

M. Sorrentino et al. / Applied Energy 120 (2014) 31–40

Vehicle speed

1

140

0.9 120

0.8

SOC [/]

[km/h]

100 80 60

BEV in CD mode PHEV in CD mode PHEV in CS mode

0.6 0.5

40

0.4

20 0

0.7

0.3 0

200

400

600

800

1000

1200

0

200

400

600

800

1000

1200

Time [s]

Time [s]

Fig. 7. SOC trajectories simulated for PHEV (CS and CD mode) and BEV (CD only) on the NEDC cycle. The upper curve refers to BEV in CD mode.

Fig. 6. NEDC driving cycle.

Table 2 Specifications of the base CC vehicle. Nominal ICE power (kW) Fuel Length (l) (m) Width (w) (m) Height (h) (m) Mass (kg) Cx Cr FE (NEDC) (km/l)

75 Gasoline 4.44 1.77 1.55 1298 0.33 0.01 14.84

Table 3 Main technical features and performance of electrified vehicles. PHEV

BEV

Nominal ICE/EG power (kW)

20

0

Fuel EM nominal power (kW) Nb

Gasoline 76 45

81.2 65

Battery technology Battery pack energy (kW h) Battery energy density (W h/kg) Battery power density (W/kg) Vehicle mass (kg) FE in CS mode (NEDC) (km/l) EC in CD mode (NEDC) (W h/km)

Lithium-manganese [34] 6.7 28 90 132 760 382 1319 1406 19.6 NA 148 152

such a selection of vehicle specifications can be considered generic enough to be worldwide representative of the private cars fleet to be referred to in the scenario analyses performed in Section 7. Moreover, It has to be considered that, due to the increasing globalization of the automotive market, of the growing urbanization and of the more compelling constraints in terms of emissions and CO2 reduction, the differences among different countries in terms of car size and characteristics are lower than before, and it is reasonable to expect that they will be exhibit a reducing trend in next years. Table 3 lists the main features of PHEV and BEV cars here assumed, as they resulted from the application of the mass model described in Section 3 to the hybridization/electrification of the reference conventional car (see Table 2). The comparison between PHEV and BEV specifications indicates how the higher battery energy density assumed for BEV architecture allows achieving, assuming that the car is always driven on NEDC-like routes, an autonomy as high as 184 km. Table 3 also reports the PHEV fuel

economy estimated on the NEDC cycle, which is, as expected, significantly higher than conventional car value (see Table 2). On the other hand, the slightly lower PHEV mass does not provide such significant benefits in CD mode as compared to BEV, as indicated by the electricity consumption values listed in Table 3. Fig. 7 shows the SOC trajectories simulated in case of CS and CD operation and in CD only mode for PHEV and BEV, respectively. As expected, in CD mode the smaller PHEV battery pack results in a deeper SOC depletion as compared to BEV. On the other hand, the CS transient simulated for the PHEV demonstrates the effectiveness of the thermostatic control strategy described in the previous section to guarantee that final SOC falls close to the initial one. The simulated fuel economies and electrical ranges listed in Table 3 well agree with data published for similar electrified vehicles [15,41], thus proving, on one hand, the validity of the modeling approach proposed in Sections 3 and 4 and, on the other, establishing a reliable benchmark for the comprehensive scenario analysis presented and discussed in the subsequent sections. 6. Modeling of vehicle electrification impact 6.1. Impact on grid infrastructure The electrification of the private car fleet will certainly have an impact on the electric grid due to the need of recharging the batteries. This aspect is crucial because it entails increasing installed capacity so as to ensure meeting the demand arising from the additional plug-in fleet. This power increase, if needed, should be made in view of required decarbonization of the production mix: the adoption of technologies that offer reduced CO2 emissions should be preferred to maximize car electrification benefits. In order to deeply investigate such a key and strategic aspect, it is mandatory to calculate the energy needs and the minimal additional gridpower to be installed to ensure PHEVs and BEVs be run in charge depleting mode up to their respective distances, evaluated as described in the next sections. It is worth noting here that second level analyses also could be required to perform highly accurate assessment of effective car electrification impact on grid infrastructure, such as to account for the time varying issues associated to PEV charging [42,43]. In particular, the impact of uncontrolled PEV and PHEV recharging on electrical transformers life can be very negative, and intelligent control strategies have to be adopted in a ‘‘Smart Grid’’ context [44,45]. Nevertheless, in this specific work the overall impact was examined on a daily time-base, thus allowing to reduce the level of complexity in the scenario analyses performed in Section 7.

M. Sorrentino et al. / Applied Energy 120 (2014) 31–40

6.1.1. BEV impact For BEVs, it should be considered that autonomy only depends on their EC (see Table 2). Therefore, the daily energy needed to recharge the entire BEV fleet can be estimated as:

EBEV ðnÞ ¼ ABEV  ECBEV ðnÞ  ½%PEVðnÞ  Ncar ðnÞ

ð12Þ

where ABEV is the BEV autonomy to be guaranteed on a daily base, assumed equal to 100 km. It is worth mentioning here that such a large autonomy was assumed as a conservative value, in order to guarantee battery recharging always be possible for all drivers, even when they occasionally drive longer than average daily distance. As it will be recalled in the results section, the average daily distance covered by a passenger car is assumed much lower (i.e. 41 km day1, which corresponds to 15,000 km year1) than the above conservative value (i.e. 100 km). In order to account for BEV technology improvement, it was assumed that ECBEV(n) will linearly reach an overall 10% reduction over the timeframe 2011–2050, thus varying from the value reported in Table 3 for 2011 down to ECBEV(2050) = 137 W h/km. The improvement in BEV electricity consumption was conservatively limited to 10% to account for the high uncertainty on future share of different battery technologies used on such vehicles. As a matter of fact, higher energetic performance guaranteed by Li-Ion batteries could be sacrificed to cost-effectiveness and availability of lithium reserves worldwide [30]. The BEV power demand (i.e. PBEV(n)) can therefore be easily calculated dividing EBEV(n) by the recharging time Tch,BEV, here assumed equal to 1 h. The latter value was selected by conservatively referring to a recent press-release issued by a relevant manufacturer of vehicles’ battery chargers [46], which claims its technology ensures BEV battery to be fully charged in about 30 min. As for 2011, the 1 h charging time assumption results in a power demand for BEV recharging equal to 15 kW per car. It is worth mentioning here that the daily amount of energy to be recharged for BEV (i.e. 15 kW h) does not necessarily correspond to the battery pack of the exemplary BEV vehicle specified in Table 3, which was designed in such a way as to grant maximum autonomy reaches up to about 180 km, which of course differs from the average value (i.e. 100 km) here assumed to estimate average grid power requirement for a single BEV. 6.1.2. PHEV impact As sketched in Fig. 5, PHEV autonomy, differently than BEV, can rely on the sole fuel availability to be run in CS mode after battery depletion. Therefore, the power requested to the grid for PHEVs recharging has to be estimated as a function of the so called Utility Factor (UF), which is defined as [47]:

UF ¼

DCD DCD ¼ Dyear DCD þ DCS

ð13Þ

where Dyear is the overall annual distance and DCD and DCS are the average annual distances driven in CD and CS mode, respectively. Therefore the daily energy requested to the grid by the entire PHEV fleet can be computed as:

EPHEV ðnÞ ¼

Dyear  UF  ECPHEV ðnÞ  ½%PHEVðnÞ  Ncar ðnÞ 365

ð14Þ

It is worth noting that the evolution in time of ECPHEV(n) (see Eq. (14)) was modeled applying the same procedure adopted for ECBEV(n) (see Section 6.2), thus assuming again 10% technology improvement. Finally, the overall PHEV power demand can be found via the following relationship:

PPHEV ðnÞ ¼

EPHEV ðnÞ T ch;PHEV

ð15Þ

where the charging time Tch,PHEV was assumed equal to 7 h. The latter assumption is a consequence of the fact that PHEV can also rely

37

on CS operation only; therefore it is not mandatory to recharge the battery during the day at advanced and possibly more expensive fast recharging station, as it is instead required for BEVs in case longer daily distance than average are to be covered. Moreover, considering the Italian one as an exemplary case, reducing nominal household power allows significantly lowering fixed costs on the electric bill, which in turn results in lower electricity cost for those drivers that have the opportunity to recharge their PHEV cars at home during the night.

6.2. Impact on CO2 emissions The evolution of CO2 emissions associated to private car fleet operation was predicted as a function of expected market penetration of PEVs (see Section 2). Of course also the contribution coming from conventional cars has to be taken into account over the timeframe 2011–2050. In this study, it was assumed that from now to 2050 an increasing part of non-plug-in (NPI) cars will consist of the more environmentally friendly HEVs. In order to account for such a positive impact on overall CO2 emitted by non plug-in cars (i.e. conventional cars + HEVs), it was assumed that the NPI fuel economy (FENPI) will increase linearly from the value given in Table 2 for conventional cars up to FENPI(2050) = 18.6 km/l. The latter value results from the following two assumptions: (i) in 2050 NPI fleet will likely exhibit an equal share between conventional and HEV cars [7], whose current FE values set to 14.84 (see Table 2) and 21.5 [48], respectively; (ii) technology improvement of NPI cars will reach 5% (i.e. in terms of fuel economy) over the entire timeframe 2011–2050. Therefore, the contribution per driven km of a single NPI vehicle is estimated as:

CO2;NPI

h g i CO2;f ¼ km FENPI ðnÞ

ð16Þ

where CO2,f is the CO2 emission associated to fuel consumption, here assumed equal to 2302.5 gCO2 per liter of consumed gasoline. It is worth mentioning here that the analysis was limited to gasoline as only fuel for sake of simplicity. On the other hand, the CO2 emitted per km by a single BEV is only due to electricity generation:

CO2;BEV

h g i ¼ CO2;grid  ECBEV ðnÞ km

ð17Þ

A different approach is required for PHEVs, as a consequence of their dependence both on grid electricity and fuel. Particularly, the CO2 emitted by PHEVs consists of two parts, associated to CD and CS operation, respectively. Thus, the contribution of a single PHEV per driven km can be estimated as a function of UF, as follows:

CO2;PHEV

h g i ¼ UF  CO2;CD þ ð1  UFÞ  CO2;CS km ¼ UF  CO2;grid  ECPHEV ðnÞ þ ð1  UFÞ 

CO2;f FEPHEV ðnÞ

ð18Þ

where the time evolution of FEPHEV(n) was modeled similarly to the PEV EC trends (see Sections 6.1.1 and 6.1.2), i.e. assuming a linear improvement of PHEV technology by 10% over the timeframe 2011–2050, thus resulting in an increase from the value given in Table 3 for 2011 up to FEPHEV(2050) = 21.5 km/l. When comparing the assumptions made above on FENPI and FEPHEV to the forecast performed by the IEA about 2050 average fuel economy of lightduty-vehicles (i.e. 20 km/l as reported in [49]), a good agreement is obtained considering that number of NPI cars in 2050 will be approximately equal to the number of PHEVs [7].

M. Sorrentino et al. / Applied Energy 120 (2014) 31–40

The parametric analysis shown in this section aims to assess the impact of expected car electrification, in the timeframe 2011– 2050, on both grid infrastructure and CO2 emissions associated to Italian private cars fleet. Specifically, the objective was to extensively evaluate how either changes in electric cars penetration or energy mix improvements will affect the achievement of the CO2 reduction goals addressed by the IEA blue-map [7]. To this scope, several scenarios (see Tables 4 and 5) were simulated, by applying the mathematical models presented in Sections 2–6 to the PEV powertrains detailed in Table 3 and considering Dyear and UF equal to 15,000 km and 0.6 [50], respectively. Table 4 provides details about the three energy mixes assumed in the scenarios 1–6 defined in Table 5. Particularly, Mix 1 and Mix 2 correspond to the realistic and highly optimistic improvements in terms of CO2 reduction that can be achieved in Italy in the next years, respectively [51]. Fig. 8 shows that, assuming the expected 2050 market penetration, only scenario 3 allows achieving the targeted 30% reduction in annual CO2 with respect to actual status [7]. On the other hand, in case no decarbonization (i.e. scenario 1 in Table 5) or smooth decarbonization (scenario 2) is performed, the 30% CO2 reduction will not be achieved, despite the significant impact on grid infrastructure in terms of increased maximum power demand due to car electrification, as shown in Fig. 9. In scenarios 4–6 a different analysis was performed, namely the models described in Sections 2–6 were deployed to estimate the necessary 2050 PEV market penetration to meet the blue-map goals [7]. Particularly, an increase in market penetration, as it happens in scenario 4 and 5 (see Table 5), can significantly contribute to achieving the 30% reduction of annual CO2 by 2050 (see Fig. 8), but again with a significant impact on grid infrastructure, as shown in Fig. 9. Finally, Figs. 8 and 9 indicate that high decarbonization (i.e. scenario 6 in Table 5) could allow reducing PEV market penetration, which in turn results in lower impact on grid infrastructure as compared to other scenarios, while still meeting the 2050 CO2 reduction target. It is worth noting here that the required power for PEV recharging is very high when compared to current installed power (see Fig. 9), thus confirming the high relevance of finding appropriate solution to the complex interaction among grid infrastructure and private electric mobility [12,16].

Table 5 Scenarios simulated for the Italian private cars fleet. Scenario

2050 Vehicle distribution

1 2 3 4 5 6

Mix

CC + HEV (%)

BEV (%)

PHEV (%)

61 61 61 29 53 65

14 14 14 26 17 13

25 25 25 45 30 22

2011 1 2 2011 1 2

CO 2 emissions trends

90 85 80 75

[Mt/year]

7. Scenario analysis on the Italian generation mix

70 65 60 55 50

NO car electrification Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5 Scenario 6 Blue-map target

45 2010

2020

2030

2040

2050

Year Fig. 8. Simulated CO2 trends for the Italian private cars fleet in the timeframe 2011–2050.

Grid power demand for PHEV&BEV 200 180

Scenarios 1, 2 and 3 Scenario 4

160 140

Scenario 5 Scenario 6 2011 nominal grid power in Italy

120

[GW]

38

100 80

7.1. Extension to other countries

60

This section focuses on the extension of the analysis carried-out for Italian private cars sector to other countries, specifically those listed in Table 6. The latter table also provides the CO2 emissions associated to each country’s electrical grid, whereas the growing trends of passenger cars expected in the timeframe 2011–2050 were already discussed in Section 2. Fig. 8 shows the results obtained for all countries except China, for which a more extended analysis was performed as illustrated in Table 7 and Fig. 11. Particularly, Fig. 10 shows the percentage of CO2 reduction that can be potentially achieved in the timeframe 2011–2050. As expected, the very low CO2 impact of French electric grid (see Table 6) allows meeting the blue-map goal without modifying current generation mix. On the other hand, the other

Table 4 Italian generation mixes [51]. Energy sources

Fossil (%)

Renewables (%)

Nuclear (%)

CO2,grid (g/W h)

Mix 2011 Mix 1 Mix 2

69 40 10

20 33 90

11 27 0

0.53 0.32 0.08

40 20 0 2010

2015

2020

2025

2030

2035

2040

2045

2050

Year Fig. 9. Simulated electric power demand trends for the Italian private cars fleet in the timeframe 2011–2050.

countries do need to upgrade their respective generation mix to achieve 100% matching of blue-map goal by 2050. It is interesting to note that Germany and Japan(1) exhibit the same trend of CO2 reduction percentage (see Fig. 10), which is a consequence of the same CO2,grid value and similar growing trends in number of passenger cars characterizing these countries (see Section 2). In the case no increase in private car population will take place in Japan (i.e. Japan(2) in Table 1), the percentage of target matching will increase over 100%, thus making it possible to meet 2050 CO2 reduction target without acting on current Japan energy mix. Of course, in the case of France, a similar trend in car fleet evolution will not provide any benefit (see France(2) in Fig. 10), as the sole PEV

M. Sorrentino et al. / Applied Energy 120 (2014) 31–40 Table 6 Scenarios simulated for private cars fleets of other countries. Country

2050 Vehicle distribution

Germany USA France Japan

CO2,grid (g/W h) (Mix 2011 [51])

CC + HEV (%)

BEV (%)

PHEV (%)

61

14

25

0.45 0.53 0.1 0.45

Table 7 Scenarios simulated for Chinese private cars fleet. Scenario

7 8 9

2050 Vehicle distribution CC + HEV (%)

BEV (%)

PHEV (%)

61 0 0

14 37 37

25 63 63

Car increase rate (mln/year)

CO2,grid (g/W h)

4 4 3

0.85 0.085 0.085

39

penetration into the market already ensures fully meeting 2050 CO2 reduction target (see France(1) curve in Fig. 10), as discussed above. As mentioned above, the case of China was dealt with by extending the analysis to cleaner generation mixes, which can be potentially achieved in the long-term perspective [7]. Moreover, it was considered the case that future private mobility in China will consist of PHEV and BEV only, as described in Table 7. Fig. 11 shows that in scenario 7 (i.e. 2011 generation mix, as indicated in Table 7) CO2 emissions associated to private cars are expected to dramatically increase, due both to the much higher demand for mobility, as compared to other countries, and to the currently highly carbon-intensive Chinese grid. In scenario 8, a much cleaner but potentially achievable generation mix [7], along with the assumption that all private cars will be electrified in 2050, are still insufficient to meet the blue-map goal for China. Only the additional assumption of a reduced rate of private cars increase (i.e. scenario 9 in Table 7) allows meeting such a goal, as highlighted by the scenario 9 trend shown in Fig. 11.

Percentage of blue-map goal matching in terms of CO2 reduction

8. Conclusions

150 Germany and Japan(1) USA France(1) France(2)

100

[%]

Japan(2)

50

0 2010

2020

2030

2040

2050

Year Fig. 10. Simulated trends of CO2 reduction for other countries. Note than 100% corresponds to full accomplishment of 2050 goal as addressed by IEA blue-map [7]. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Percentage of blue-map goal matching in terms of CO2 reduction

100 0 -100

[%]

-200 -300 -400 -500 -600 2010

Scenario 7 Scenario 8 Scenario 9

2020

2030

2040

2050

Year Fig. 11. Simulated trends of CO2 reduction for China. Note than 100% corresponds to full accomplishment of 2050 goals as addressed by IEA blue-map [7] (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.).

The paper presented a thorough analysis aimed at demonstrating the high potentials offered by plug-in cars, i.e. both PHEV and BEV, to substantially decrease CO2 emissions toward the achievement of the 2050 reduction goals indicated by the IEA blue-map [7] for the private cars sector. A suitable modeling structure was conceived, in such a way as to couple the estimation of actual and future trends in tank to wheel and well to tank emissions to expected growing demand for electric cars from now to 2050. Moreover, some hypothesis and simplifying assumptions were introduced, to enable the development of synthetic scenario analyses aimed at performing preliminary assessment of the impact of the grid-technology interaction on effective potential of car electrification. However, simulation studies have shown that the effects of most of the parameters on model outputs are fairly linear, for moderate differences around the nominal values. Sensitivity studies can therefore be performed in order to check the impact of the assumptions on the results. The analysis outcomes highlighted how regional features of current generation mixes, along with expected market penetration of PEVs, are insufficient to fully accomplish the 2050 targets, except for the specific case of France. More particularly, the extensive scenario analysis conducted for the Italian private cars indicated the strong need of coupling car electrification process to highly effective energy policies, the latter to be mainly aimed at significantly decarbonizing actual electrical grid [52]. Moreover, the impact on grid infrastructure played by plug-in recharging emerges as a non-trivial issue to be solved, with the final aim of making car electrification contribution to GHG effect reduction not only effective, but even realistically achievable. With respect to this matter, a significant counter-measure could rely on the adoption worldwide of really smart grids, which should be particularly capable of reducing the impact of plug-in vehicles on grid infrastructure by optimizing the interaction between their charging phases and vehicle-to-grid (V2G) tasks [53]. An additional counter-measure may also be represented by the progressive decentralization of electricity production, as suggested in previous researches conducted on the topic [54–56]. Finally, the analysis conducted on the specific case of China highlighted the further need of regulating future demand of mobility for emerging countries, in view of guaranteeing the achievement of a really sustainable mobility in the wider context of sustainable development.

40

M. Sorrentino et al. / Applied Energy 120 (2014) 31–40

References [1] Electrical and Electronics Technical Team Roadmap; 2010. . [2] Private communication with a FIAT Elasis research center Engineer working on SI engines design and testing; 2006. [3] Kyoto Protocol; 1997. . [4] Liu W, Hu W, Lund H, Chen Z. Electric vehicles and large-scale integration of wind power – the case of Inner Mongolia in China. Appl Energy 2013;104:445–56. [5] Khayyam H, Abawajy J, Javadi B, Goscinski A, Stojcevski A, Bab-Hadiashar A. Intelligent battery energy management and control for vehicle-to-grid via cloud computing network. Appl Energy 2013;111:971–81. [6] Fernandes C, Frías P, Latorre JM. Impact of vehicle-to-grid on power system operation costs: the Spanish case study. Appl Energy 2012;96:194–202. [7] IEA. Technology roadmap: electric and plug-in hybrid electric vehicles; 2011. . [8] Zhang Y, Yu Y, Zou B. Analyzing public awareness and acceptance of alternative fuel vehicles in China: the case of EV. Energy Policy 2011;39:7015–24. [9] Skerlos SJ, Winebrake JJ. Targeting plug-in hybrid electric vehicle policies to increase social benefits. Energy Policy 2010;38:705–8. [10] Kahn Ribeiro S, Kobayashi S, Beuthe M, Gasca J, Greene D, Lee DS, et al., 2007. Transport and its infrastructure. In: Metz B, Davidson OR, Bosch PR, Dave R, Meyer LA, editors. Climate change 2007: mitigation. Contribution of working Group III to the fourth assessment report of the intergovernmental panel on climate change. Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press. [11] Amoroso FA, Cappuccino G. Impact of charging efficiency variations on the effectiveness of variable-rate-based charging strategies for electric vehicles. J Power Sources 2011;196:9574–8. [12] Juul N, Meibom P. Road transport and power system scenarios for Northern Europe in 2030. Appl Energy 2012;92:573–82. [13] Bullis K. GM’s new electric vehicle. . [14] Borroz T. Toyota promises an ‘Affordable’ Plug-In Prius in 2011. . [15] http://www.nissanusa.com/leaf-electric-car/index#/ leaf-electric-car/index. [16] Lindly JK, Haskew TA. Impact of electric vehicles on electric power generation and global environmental change. Adv Environ Res 2002;6:291–302. [17] Van Vliet O, Brouwerb AS, Kuramochib T, van den Broekb M, Faaij A. Energy use, cost and CO2 emissions of electric cars. J Power Sources 2011;196:2298–310. [18] Hadley SW, Tsvetkova A. Potential impacts of plug-in hybrid electric vehicles on regional power generation. Electricity J 2009;22:56–68. [19] Sioshansi R, Fagiani R, Marano V. Cost and emissions impacts of plug-in hybrid vehicles on the Ohio power system. Energy Policy 2010;38:6703–12. [20] Razeghi G, Brown T, Samuelsen GS. The impact of plug-in vehicles on greenhouse gas and criteria pollutants emissions in an urban air shed using a spatially and temporally resolved dispatch model. J Power Sources 2011;196:10387–94. [21] Gong Q, Midlam-Mohler S, Marano V, Rizzoni G. PEV charging impact on residential distribution transformer life. In: Proceedings of energy tech 2011, Cleveland (MI-USA); May 25–26, 2011. p. 1–6. [22] Saarenpää J, Kolehmainen M, Niska H. Geodemographic analysis and estimation of early plug-in hybrid electric vehicle adoption. Appl Energy 2013;107:456–64. [23] Cai H, Hu X, Xu M. Impact of emerging clean vehicle system on water stress. Appl Energy 2013;111:644–51. [24] Alessandrini S, Girardi P, Pirovano G, Riva G, Gianinoni I. Valutazione dell’impatto dell’inserimento di veicoli elettrici stradali plug-in sull’ambiente (qualità dell’aria) alla scala locale ed a quella regionale. ENEA-Ricerca sul Sistema Elettrico (ERSE), Milano; 2010 [in Italian]. [25] Bodek K, Heywood J. Europe’s Evolving passenger vehicle fleet: fuel use and GHG emissions scenarios through 2035. Laboratory for Energy and Environment Massachusetts Institute of Technology, Publication No. LFEE 2008-03 RP. . [26] http://data.worldbank.org/indicator/IS.VEH.PCAR.P3. [27] http://www.usatoday.com/money/autos/2010-01-08-china-auto-sales_N.htm. [28] http://www.chinadaily.com.cn/bizchina/2010-12/08/ content_11668665.htm. [29] Tong SJ, Same A, Kootstra MA, Park JW. Off-grid photovoltaic vehicle charge using second life lithium batteries: an experimental and numerical investigation. Appl Energy 2013;104:740–50. [30] Grosjean C, Miranda PH, Perrin M, Poggi P. Assessment of world lithium resources and consequences of their geographic distribution on the expected

[31]

[32]

[33]

[34]

[35] [36] [37]

[38]

[39] [40]

[41] [42]

[43] [44]

[45]

[46]

[47] [48]

[49]

[50]

[51]

[52] [53]

[54]

[55]

[56]

development of the electric vehicle industry. Renew Sust Energy Rev 2012;16:1735–44. Sorrentino M, Rizzo G, Vasca F. An energetic comparison for hybrid vehicles ranging from low to high degree of hybridization. SAE Paper 2011-24-0086. In: Proceedings of the 10th international conference on engines & vehicles, Capri, Italy; September 11–15, 2011. Sorrentino M, Arsie I, Di Martino R, Rizzo G. On the use of genetic algorithm to optimize the on-board energy management of a hybrid solar vehicle. Oil Gas Sci Technol – Rev l’IFP 2010;65:133–43. Rousseau A, Sharer P, Ahluwalia R. Energy storage requirements for fuel cell vehicles. SAE paper 2004-01-1302. In: Proceedings of the SAE 2004 world congress & exhibition, Detroit (MI-USA); March 2004. Nelson P, Amine K, Rousseau A. Advanced lithium-ion batteries for plug-in hybrid-electric vehicles; 2012. . Guzzella L, Sciarretta A. Vehicle propulsion systems. Heidelberg New York Dordrecht London: Springer; 2013. p. 89–92. Rizzoni G, Guzzella L, Baumann BM. Unified modeling of hybrid electric vehicle drivetrains. IEEE Trans Mechatronics 1999;4:246–57. Lee TK, Bareket Z, Gordon T, Filipi Z. Stochastic modeling for studies of real-world phev usage: driving schedules and daily temporal distributions. IEEE Trans Veh Technol [in press]. http://dx.doi.org/10.1109/TVT.2011. 2181191. Tulpule P, Marano V, Rizzoni G. Energy management for plug-in hybrid electric vehicles using equivalent consumption minimisation strategy. Int J Electric Hybrid Veh 2010;2:329–50. Barsali S, Miulli C, Possenti A. A control strategy to minimize fuel consumption of series hybrid electric vehicles. IEEE Trans Energy Convers 2004;19:187–95. Sorrentino M, Rizzo G, Arsie I. Analysis of a rule-based control strategy for onboard energy management of series hybrid vehicles. Control Eng Pract 2011;19:1433–41. http://en.wikipedia.org/wiki/Chevrolet_Volt. Su W, Chow MY. Computational intelligence-based energy management for a large-scale PHEV/PEV enabled municipal parking deck. Appl Energy 2012;96:171–82. Schroeder A. Modeling storage and demand management in power distribution grids. Appl Energy 2011;88:4700–12. Marano V, Muratori M, Rizzo G, Rizzoni G. Sustainable mobility: from fossil fuels to renewable energy, opportunities and challenges for the automotive industry. In: Proc. 7th IFAC symposium on advances in automotive, control, Tokyo; September 4–7, 2013. Gong Q, Midlam-Mohler S, Marano V, Rizzoni G. Study of PEV charging on residential distribution transformer life. IEEE Trans Smart Grid 2012;3:404–12. AeroVinment. Press Releases: EV Charging, AeroVironment Introduces New Easy-to-Install and Moveable Home Electric Vehicle Charging Station. . Bradley TH, Quinn CW. Analysis of plug in hybrid electric vehicle utility factors. J Power Sources 2010;195:5399–408. Brooker A, Thornton M, Rugh M. Technology improvement pathways to costeffective vehicle electrification. Conference Paper NREL/CP-540-47454 2010. . IEA. Thecnology roadmap: fuel economy of road vehicles; 2012. . Kelly JC, MacDonald JS, Keoleian GA. Time-dependent plug-in hybrid electric vehicle charging based on national driving patterns and demographics. Appl Energy 2012;94:395–405. De Nigris M. Impatto sul sistema elettrico della potenziale diffusione di veicoli elettrici. ENEA-Ricerca sul Sistema Elettrico (ERSE), Milano; 2010 [in Italian]. . Doucette RT, McCulloch MD. Modeling the prospects of plug-in hybrid electric vehicles to reduce CO2 emissions. Appl Energy 2011;88:2315–23. Sousa T, Morais H, Soares J, Vale Z. Day-ahead resource scheduling in smart grids considering vehicle-to-grid and network constraints. Appl Energy 2012;96:183–93. Wang J, Conejo AJ, Wang C, Yan J. Smart grids, renewable energy integration, and climate change mitigation – future electric energy systems. Appl Energy 2012;96:1–3. Brouwer AS, Kuramochi T, van den Broek M, Faaij A. Fulfilling the electricity demand of electric vehicles in the long term future: an evaluation of centralized and decentralized power supply systems. Appl Energy 2013;107:33–51. Jorge RS, Hertwich EG. Environmental evaluation of power transmission in Norway. Appl Energy 2013;101:513–20.