Battery Life Impact of Vehicle-to-Grid Application of Electric Vehicles

Transcription

1 EVS28 KINTEX, Korea, May 3-6, 2015 Battery Life Impact of Vehicle-to-Grid Application of Electric Vehicles Hajo Ribberink 1, Ken Darcovich 2, Fleurine Pincet 3 1 (corresponding author) Natural Resources Canada, 1 Haanel Dr., Ottawa, Ontario, Canada, Hajo.Ribberink@nrcan.gc.ca 2 Energy, Mining and Environment Portfolio, National Research Council of Canada, Ottawa, Ontario, Canada, K1A 0R6 3 ICAM-Lille, 6 Rue Auber, Lille, France. Abstract In recent years, electric vehicles (EVs) have successfully been gaining a market foothold in many countries around the world. The number of plug-in vehicles is forecast to grow steadily to over 10 million vehicles before In many countries, concerns about the environmental impact of fossil fuel-based electricity production have resulted in a shift towards more sustainable power generation technologies. Electricity production from solar and wind, however, is intrinsically unsteady. Costly provisions therefore need to be made within electricity grids to ensure back-up power is available on demand. The power stored in the batteries of the growing fleets of EVs could potentially be a cost effective alternative to conventional spinning reserves through Vehicle-to-Grid (V2G) applications. Concerns arise over the use of EVs for V2G purposes; as such use is expected to negatively impact EV battery life. Data on additional battery degradation due to V2G is presently quite scarce. A detailed simulation model was developed to explore the various processes that impact EV battery life. The model was used in a simulation study to determine the battery life impact of various V2G scenarios in comparison to a base case of regular driving and charging. Scenarios with different driving styles, various charge levels and different V2G events were evaluated. The impact of fast charging of EVs on battery life was also considered. The study concluded that aggressive driving and fast charging have a great impact on EV battery life. A similar level of battery degradation was found for intense participation in V2G services, fully discharging the battery on a daily basis. However, less intense use of the EV battery can still provide useful V2G services with acceptable battery degradation. Manufacturers of EVs are therefore encouraged to implement V2G capability in their EVs. Keywords: Electric vehicle, Battery life, Vehicle-to-Grid, Simulation EVS28 International Electric Vehicle Symposium and Exhibition 1

2 1 Introduction In recent years, electric vehicles (EVs) have successfully been gaining a market foothold in many countries around the world. The number of plug-in vehicles is forecast to grow steadily and the size of the worldwide fleet of EVs is expected to exceed ten million vehicles before In many countries, concerns about the environmental impact of fossil fuel-based electricity production have resulted in a shift towards more sustainable power generation technologies and a historically unparalleled increase in the use of solar and wind power for electricity production. Power generation from such renewables however, is intrinsically unsteady. Costly provisions therefore need to be made within electricity grids to ensure back-up power is available on demand, at times when output from these renewable sources is insufficient. The power stored in the batteries of the growing fleets of EVs could potentially be a cost effective alternative to conventional spinning reserves through Vehicle-to-Grid (V2G) applications. However, concerns have arisen over the use of EVs for V2G purposes; as such use is expected to negatively impact EV battery life. Data on additional battery degradation due to V2G is presently quite scarce. A detailed simulation model was therefore developed to explore the various processes that impact EV battery life. The battery life model was first used to determine the battery degradation from regular driving and recharging. Subsequently, the model was applied to evaluate the additional battery degradation of EVs used in a various V2G scenarios. This paper has the following structure: The battery life model is described in Section 2. Section 3 presents battery degradation results for EVs used in regular driving and recharging, followed by results for EVs involved in various V2G scenarios. Section 4 gives a discussion of the simulation results and conclusions are described in Section 5. 2 Battery Life Model Two different modelling procedures were required for this project. The first was a simplified fundamentals-based electrochemical model of the operation of a single Li-ion cell. The output of this first model was collected for a range of currents and condensed into a second model, an engineering type empirical model to represent the function of the total electric vehicle Li-ion battery pack. The models are summarized below, full details may be found in [1]. 2.1 Model description Li-ion battery model Based on measured charge and discharge data, and a set of material parameters, the fundamental electrochemical single particle model (SPM) [1, 2] was used to produce a number of charge and discharge curves at currents ranging from very small currents to very large currents that are beyond practical ranges for experimental tests. Charge and discharge curves from the SPM then formed the basis of an empirical representation of the Li-ion battery in an equivalent resistance type (Gao) model [3]. This approach uses a reference curve based on data fit to a polynomial curve for battery depth of discharge (DOD) as a function of open-circuit voltage (OCV), as well as current and temperature for the given battery chemistry. The Gao model is then used in vehicle use scenarios to track the operational state of the Li-ion battery. In this project, a constant battery temperature of 30 C was assumed. In the Gao model, the voltage during discharging is given by: V (t, DOD, I(t) ) = OCV (DOD) - I(t)R int (DOD) (1) where I(t) is the current and R is the internal int resistance. This equation describes that under load, there is a shift of the voltage as a function of DOD, equal to the internal resistance multiplied by the current. The energy use simulation advances in discrete time-steps in a quasi-steady-state manner within the time interval. Second-order polynomials for OCV(DOD) and R (DOD) allow the voltage to int be determined. At a given DOD, an imposed power load P, where P = IV, determines a required current I(t) based on the cell voltage. Applying this current over a time step will cause a small voltage drop, and correspondingly advances the DOD via: I( t) t DOD (2) cap ref In this case, cap is the battery capacity and α ref and β are Gao model parameters [3]. EVS28 International Electric Vehicle Symposium and Exhibition 2

3 2.1.2 Capacity fade model Use of a battery will gradually degrade its available capacity. This capacity loss occurs from low-rate irreversible chemical side reactions. The side reactions proceed, even when the battery is not in use, accounting for the calendar life of a battery. The end-of-life for a battery is typically specified as the point at which the battery capacity falls below 75% of its original value [4]. Battery degradation accelerates in proportion to applied current. Two components of capacity fade must therefore be accounted for; cycling capacity fade and calendar fade. Cycling capacity fade measurements for the cells modelled here were taken from the literature [5]. At a reference current of C/3, a nominal capacity fade of A h/cycle ( cap REF ) was fade rate reported. Capacity fade is also known to be a strong function of the level of applied current, and the battery DOD, especially when there are rapidly varying power loads that may oscillate from discharging to charging modes [1]. In terms of the DOD, the general expression for the cycling contribution to capacity fade for each time step, can be stated as REF DOD capfade capfade rate CUR DOD F [ I( t)] F [DOD] (3) Here, F CUR [ I( t)] is a ratio of the extent of fades measured as a function of applied current to the C/3 reference value, and F DOD [DOD ] is a factor applied to reflect the extent of fade attributable to the instantaneous DOD state. Explicitly, from experimental measurements, F [ I( t)] exp ( I( t)) (4) CUR For the determination of F DOD [DOD ], a set of experimental measurements were made where cells were cycled over very narrow DOD ranges for extended periods to estimate capacity fade rates as a function of DOD. Data from these tests are shown in Figure 1. F [DOD] can be expressed as DOD 2 FDOD [DOD] DOD DOD (5) The values obtained from Eq. 5 have been normalized so that evaluating them across the entire DOD range results in an amount of capacity fade equal to cap REF. As the capacity fade is fade rate updated dynamically in the simulation, a running total over time of the cumulative capacity fade is also tracked. Figure 1: Capacity fade rates for LiNMC cells versus DOD. Data taken at 30 C, results after 300 equivalent cycles. The markedly higher capacity fade rates observable at values of DOD lower than about 0.25 are known to occur [6]. A recent study [6] shows dramatic vehicle battery lifetime loss at high currents when lower DOD recharge values are permitted. To account for calendar fade ( cal, in A h), fade curve-fitting data from [5] provided the following expression, calfade E 07 NDAY E 03 (6) In Eq. 6, N DAY is the total number of days the cell has been in existence. In simulations, the calendar fade is determined once daily to update the total cumulative capacity fade at the end of each day. 2.2 Model Validation A representative electric vehicle was chosen for the simulation on the basis of the scope and quality of drive cycle battery discharge data and battery recharge data available. The electric vehicle battery had conventional LiNMC cathode material, with 28 kwh of discharge capacity. As simulations were run on the cell level, potentials, states of charge, and vehicle power loads were all normalized to single cell values. The voltage where full recharge was considered complete according to the vehicle specifications EVS28 International Electric Vehicle Symposium and Exhibition 3

4 Figure 2: High temporal resolution measured and simulated battery pack potentials for the test vehicle over a short period of an LA4 drive cycle. was 4.02 V, which corresponds to a DOD of The simulation was then validated against dynamometer data provided by Environment Canada for the LA4 (city) Drive Cycle. Overall, the average error between measured and simulated data was very small, only 0.15%. The largest errors occur during battery relaxation phase, due to time lag and high internal resistance at low currents. In Figure 2, a portion of the simulated validation data is shown in great detail in high temporal resolution in order to demonstrate the dynamic response of the simulation to rapidly changing load demands. It can be seen in Figure 2 that when there are very sudden drops in the current demand on the battery, there is a more rapid relaxation back to resting potentials for the simulated values, but in general the correspondence is quite close, generally within 0.01 V. evaluate the impact of the type of driving on the EV battery life. Figures 3 to 5 display the speed profiles of respectively the city drive cycle ( LA4, also called UDDS drive cycle), the highway drive cycle ( HWY ), and the drive cycle representing aggressive driving ( US06 ). The city drive cycle covers a distance of 12.0 km, the highway drive cycle 16.5 km, and the aggressive driving drive cycle 12.9 km. More information on these standard drive cycles can be found in [8]. 3 Results of Battery Life Simulation 3.1 Base Cases of Regular Driving and Recharging Any use of a battery has impact on its life. The regular use of electric vehicles therefore already has an influence on the life of the EV battery. The use of EV batteries in various scenarios of regular driving and recharging was simulated to establish base cases for comparison to the scenarios in which the electric vehicles would also be involved in V2G activities Driving characteristics There is a great variation in the way and extent that vehicles are driven. In this study, standard drive cycles representing city driving, highway driving and aggressive driving were used to Figure 3: Speed profile of city driving test cycle (LA4/UDDS) Figure 4: Speed profile of highway driving test cycle EVS28 International Electric Vehicle Symposium and Exhibition 4

5 start of the driving on each new day, the EV was made to continue at the point of the drive cycle where it ended the day before, to ensure that all sections of the drive cycles would get equal coverage. Figure 5: Speed profile of US06 aggressive driving test cycle Charging characteristics Electric vehicles can be recharged using different charge levels. Depending on the charge level, the time to fully recharge the battery will vary, as will the efficiency of the charging process. Table 1 presents the (AC) charge levels and associated charging efficiencies used. Most recharging of electric vehicles at home will be done using Level 1, Level 2a or Level 2b. Fast charging at Level 3a or Level 3b is often limited to other locations. However, regular charging at these power levels was included in this study to investigate the impact of regular fast charging on EV battery life. Table 1: Electric vehicle charge and discharge levels and charger efficiencies Charge/Discharge level kw (AC) Charger efficiency L L2a L2b L3a L3b The charging efficiencies for each charge level were based on ballpark figures known for these conditions. However, actual measured data were available for charging at 3.3 kw (L2a). It was decided to use these more detailed results, despite the fact that they are a bit off from the trend of the other charge levels. The impact of this slightly higher efficiency on the battery life results was expected to be insignificant. The AC power levels at which the EVs would feed power back into the grid were assumed to be the equal to the AC charge levels. The same applied to the discharge efficiencies Daily EV usage schedule In each scenario, the electric vehicle was driven 50 km per day and then fully recharged. At the Battery life impact of regular driving and recharging First the calendar life of the EV battery was determined using the battery life model. Without any usage of the battery, the EV battery would slowly degrade and after years the battery capacity would have dropped to 75% of its original capacity. At this point in time, the battery is assumed to have reached its end of life in the context of being useful in an electric vehicle. In reality, this does not mean that the EV batteries will have to be discarded or recycled. There are good opportunities for secondary usage of EV batteries possible, for instance in less demanding stationary electricity storage applications. Base case scenarios were run to investigate the additional battery degradation from regular driving and recharging. 15 scenarios were run combining each of the three different drive cycles and the five charge levels. Figure 6 displays the gradual decrease in EV battery capacity over time for the different drive cycles and recharging at the lowest charge level (1.3 kw). City and highway driving have an approximate equal effect on battery life, reducing its useful life to just over 11.5 years. Aggressive driving results in a much faster degradation of the battery and an additional reduction in battery life of almost three years in comparison to the other types of driving. Figure 6: EV battery life for various drive cycles for Level 1 (1.3 kw) charge and discharge power Figure 7 presents similar results, but now the EV was always charged at the highest power level (50 kw). Using fast charging will more quickly degrade the EV battery. The battery life for scenarios of city and highway driving decreased by about 1.2 years to approximately 10.5 years in EVS28 International Electric Vehicle Symposium and Exhibition 5

6 comparison to using the lowest charging power. The aggressive driving scenario showed an additional 1.1 year reduction in battery life compared to recharging at the 1.3 kw level, with the EV battery already reaching the end-of-life point of 75% of the original capacity within 8 years. Figure 8: EV battery life for the (33%/33%/33%) mixed drive cycle for various levels of charge and discharge power Figure 7: EV battery life for various drive cycles for Level 3b (50 kw) charge and discharge power The battery life results for all 15 scenarios are summarized in Table 2. Table 2: EV battery life (in years) for various drive cycles and levels of charge and discharge power Charge LA4 HWY US06 level L L2a L2b L3a L3b The drive cycles for city driving and highway driving were created in Current real world traffic is significantly more aggressive than 40 years ago. To use a more representative drive cycle for current day driving, a mixed drive cycle was defined. In the mixed drive cycle, the EV would drive one third of its daily 50 kilometres using the city driving cycle, one third following the highway drive cycle, and it would use the aggressive driving drive cycle for the last third of its daily kilometres. Figure 8 displays the battery life results for using this mixed drive cycle and for all charge levels. The numeric results are included in Table 3. The expected battery life for scenarios using the mixed drive cycle falls in between the results for the city/highway drive cycles and the aggressive drive cycle and ranges from years for Level 1 charging to 9.44 years for Level 3b charging. Table 3: EV battery life (years) for the mixed drive cycle and various levels of charge and discharge power Charge level Mixed drive cycle (33% LA4/33% HWY/33% US06) L L2a L2b L3a L3b Summary of base case results Drive style and charge level both have a significant impact on the EV battery life. While daily charging at Level 2 (3.3 kw or 6.6 kw) does not substantially reduce the EV battery life compared to charging at Level 1 (1.3 kw), this cannot be said of fast charging. Daily fast charging at a power level of 50 kw reduces battery life by years. The impact of drive style is even greater. Aggressive driving can reduce EV battery life by years compared to a more delicate driving style. 3.2 V2G Scenarios When electric vehicles will not only be used for regular driving and recharging, but also for Vehicle-to-grid (V2G) activities, additional wear on the EV battery is to be expected. A large number of V2G scenarios were simulated to investigate their specific impact on EV battery life Description of V2G scenarios The simulated V2G scenarios varied in Charge level: L1 (1.3kW), L2b (6.6kW), and L3b (50kW) were used. EVS28 International Electric Vehicle Symposium and Exhibition 6

7 Timing of the V2G event: Scenarios were evaluated in which after the driving the EV battery was first fully recharged before the V2G event would occur ( after recharging scenarios). Other scenarios assumed the V2G event to occur directly after the driving ( after driving scenarios), i.e. before the battery had been recharged to make up for power used for driving. In both cases, the battery would be completely recharged after the V2G event to have a full battery for next day s driving. Frequency of V2G events: Monthly, weekly, or daily. Discharge power used: Full discharge power (100%), but also 50%, and 25% of the maximum power per charge level. Total amount of energy fed back into the grid: Scenarios were evaluated with various levels of battery depletion (10%, 25%, 50%, or 100% of original battery capacity). All V2G scenarios simulated used the mixed drive cycle as defined in section A total of over 400 scenarios were run. This paper can only present part of all results Importance of timing of V2G event Before presenting detailed results for the V2G scenarios, it is important to emphasize the effect that the timing of the V2G event has on the ability of the EV to feed power back into the electrical grid. Figure 9 displays for both the after recharging and after driving scenarios the fraction of the battery capacity that is available for V2G services over the life time of the battery. Figure 9: EV battery capacity available for V2G for the scenarios of V2G after recharging and V2G after driving (i.e. before recharging). It is clear from Figure 9 that when the battery is recharged first, it can provide significantly more energy for V2G. The example given in this figure is for the scenario of daily V2G activity in which the EV battery is fully discharged at the L1 charge level. In this example, the total amount of electricity supplied for V2G in the after driving scenario over the life of the EV battery is only 43% of the amount provided in the after recharging scenario Impact of level of battery depletion and the frequency of V2G events on EV battery life Simulations were run to determine the impact of the level of battery depletion (the amount of energy taken from the battery) and the frequency of V2G events on the EV battery life. Figure 10 displays the battery life results for various levels of battery depletion, for monthly, weekly and daily V2G events for after recharging and for after driving scenarios and for three different charge levels. Because the after driving scenarios have less electricity to provide for V2G, the graphs on the right in Figure 10 only show results up to a 50% battery depletion used for V2G. All graphs in Figure 10 show a larger reduction in battery life with increasing levels of battery depletion. For the after recharging scenarios, the influence of monthly V2G usage of EV batteries is fairly minor. EV battery life is reduced by only two months for power levels up to 6.6 kw and by three months when using fast charging. Weekly V2G activity decreased EV battery life by 0.6 to 1.0 years depending on the charge level. The impact of daily V2G activity is much more severe, decreasing EV battery life by 3.2, 3.4 and 4.4 years for L1, L2b and L3b power levels, respectively. When the V2G events are timed before the recharge of the battery, the impact of the additional use of the EV battery on its life is generally lower than for the after recharging scenarios. The lower total amount of electricity available for V2G in the after driving scenarios results in a smaller decrease in EV battery life of approximate 40% - 70% of that for similar after recharging scenarios. The decrease in battery life for the after driving scenarios ranges up to 2.2 years for daily V2G at 50 kw. Figure 10 shows that the frequency of the V2G events has a high impact on EV battery life. The shortest EV battery life was found for the combination of daily V2G activity fully depleting the battery (after recharging) and daily fast charging. The battery life in this scenario was dramatically reduced to just over five years. EVS28 International Electric Vehicle Symposium and Exhibition 7

8 a) Level 1 (1.3 kw) d) Level 1 (1.3 kw) b) Level 2b (6.6 kw) e) Level 2b (6.6 kw) c) Level 3b (50 kw) f) Level 3b (50 kw) Figure 10: EV battery life for scenarios of different levels of battery depletion, frequency of V2G events, and level of (dis)charge power. Figures a c on the left present results for scenarios in which V2G events occurred after recharging of the EV battery, figures d f on the right display results for scenarios in which V2G events happened directly after driving. a) After recharging b) After driving Figure 11: EV battery life for daily V2G events, various levels of battery depletion, different discharge power levels and for different timing scenarios EVS28 International Electric Vehicle Symposium and Exhibition 8

9 a) After recharging a) After driving Figure 12: EV battery life for daily depleting the EV battery using various percentages of L3b discharge power and for different timing scenarios and various levels of battery depletion Impact of the charge level of V2G events on EV battery life The impact of the (dis)charge level on the additional battery degradation due to daily V2G events is presented in Figure 11 for scenarios of various battery depletion levels. The results do not show a big difference between using L1 and L2b, while daily use of fast charging causes much greater battery degradation. The scenarios for V2G after driving show a less severe impact compared to the after recharging scenarios Impact of the percentage of V2G discharge power on EV battery life The battery capacity that is available for V2G can be fed back into the grid at the full discharge level, or at a reduced charge level for a longer period of time. Figure 12 displays the results for fully depleting the EV battery daily for different (dis)charge powers (100%, 50%, or 25% of the nominal 50 kw L3b discharge power). The graphs clearly show that the influence of the specific discharge power level used is less than the impact of the amount of battery capacity discharged on a daily basis Battery life impact of the specific state of charge of the battery at the time of the V2G event Figures all show that there is an increase in the impact of V2G on the EV battery life when larger amounts of energy are taken from the battery (larger battery depletion). However, the impact on battery life is not only determined by the amount of energy taken from the battery, but also by the specific state of charge of the battery during the V2G activity. To illustrate this, Table 4 compares the results of two sets of scenarios of V2G events, both for V2G after the recharging of the EV battery. For all charge levels, the end of life for the EV battery is reached sooner with a daily V2G event that depletes only 10% of the battery, than for a weekly event of fully discharging the battery. This seems counterintuitive, as the total amount of energy of seven daily events of 10% battery depletion is clearly less than one full battery discharge. This unexpected result can be explained by looking at Figure 1, which shows that the biggest battery degradation occurs at low levels of DOD, i.e. when the battery is almost full. The daily 10% battery depletion events cycle the battery in the high-impact area of the curve, while a full discharge of the battery results in a smaller average impact. This is also the reason for the relatively larger reduction in battery life for low levels of battery depletion in all graphs in Figures Table 4: EV battery life (years) for scenarios of different combinations of V2G frequency and level of battery depletion Charge level No V2G Weekly V2G, 100% depletion Daily V2G, 10% depletion L L2b L3b This effect can also be seen by comparing the results for similar scenarios of daily discharge of 10% of the original battery capacity (see Table 5). The scenarios, in which the V2G events occur before the battery has been recharged, show a significantly lower decrease in battery life than the scenarios for V2G after recharging of the battery. After driving, the battery is roughly depleted for about 50%. This battery state of charge corresponds to the area in Figure 1 with the lowest specific battery degradation, in contrast EVS28 International Electric Vehicle Symposium and Exhibition 9

10 to the area of highest impact for the after recharging scenarios. Table 5: EV battery life (years) for scenarios of daily V2G events of 10% battery depletion Charge level No V2G After driving After recharging L L2b L3b Discussion This study started off to research the battery life impact of V2G applications on EV batteries. First the influence of regular driving and recharging was investigated. While a delicate drive style would result in an EV battery life of about 11.5 years, aggressive driving would substantially decrease EV battery life by about three years. Recharging at power levels up to 6.6 kw (L2b) was found to hardly impact battery life, while daily fast charging would further reduce the battery life by about one year. The impact of V2G activities on the life of the EV battery was found to depend strongly on the frequency of the V2G events and the amount of energy taken from the battery. Monthly V2G events had only a minor influence on EV battery life. Participation in V2G on a weekly basis would reduce EV battery life by up to one year. A more significant reduction of EV battery ( years) was seen for daily V2G events, which would supply a full battery capacity to the grid. When the V2G events would occur before the EV battery had been recharged, less energy would be available for V2G and the battery life reduction would only be around 40-70% of the decrease in life for similar scenarios with full battery depletion. The battery life is impacted more by the amount of energy transferred from the EV battery, than by the charge level at which this amount of energy is supplied to the grid. The specific state of charge of the EV battery also has a large influence on the reduction in battery life for V2G events in which the battery would only be partially depleted. When a battery is full, the charging and discharging will have a much bigger impact on the life of the battery than when the battery is at a medium state of charge (a battery that is half empty). What do the results of this study mean for the potential of V2G to support the electrical grid? The introduction of V2G will likely start in the commercial market segment. Companies that have both high peak loads of electricity and their own fleet of electric vehicles may be able to significantly reduce capacity charges on their electricity bill by using the electricity stored in the batteries of their fleet of EVs. This study has shown that battery life impacts of V2G activities are more limited for vehicles utilizing charge powers up to 6.6 kw and should not be a showstopper for this application. Delivery trucks that would use fast charge connections may experience higher loss of battery life. The economics of using EVs for V2G, however, will vary per application and will have to be evaluated based upon the specific conditions for each application. A potentially larger market will consist of private individuals owning an electric vehicle. Charge levels for these vehicles mostly do not exceed the 6.6 kw level. Acceptable battery life impacts for this charge level seem possible for participation in V2G events on a regular basis. Most electric vehicle OEMs have been very hesitant to allow the use of EV batteries for V2G activities. However, the impact of the most extreme V2G case (fully depleting the battery for V2G every day) results in about the same additional loss of battery life as when EV owners would practice aggressive driving. As the OEMs do not limit EV owners in how to drive their vehicles, OEMs are therefore encouraged to also embrace V2G and make their EVs V2G capable. 5 Conclusions Aggressive driving has a strong impact on EV battery life. An aggressive drive style reduces EV battery life by three years. Regular charging at levels up to 6.6 kw has a minimal effect on EV battery life, while fast charging reduces battery life by about one year. The frequency and amount of energy transferred from the EV battery to the grid have the largest impact on EV battery life. The extreme case of daily discharging the full EV battery for V2G results in a year reduction of EV battery life. Less intense use of the EV battery can provide useful V2G services with acceptable battery degradation. The specific battery degradation of a small discharge is bigger for full batteries than for half-empty batteries. Shallow cycling of batteries is best done at medium state of charge. Participating in V2G before the EV EVS28 International Electric Vehicle Symposium and Exhibition 10

11 battery has been recharged results in a lower impact on battery life than after recharging. EV manufacturers are encouraged to make their EVs V2G capable, as the maximum impact of daily V2G activities is comparable to that of aggressive driving. Nomenclature α β cal cap DOD F I N OCV P R V Gao model parameters Gao model parameters calendar battery capacity Depth of discharge Ratio of the extend of capacity fade Current Number of days the battery exists Open circuit voltage Power (load on the battery) Resistance Voltage subscripts CUR current DOD depth of discharge DAY days fade fade fade rate rate of fade int internal ref / REF reference t time Acknowledgments The authors would like to thank Aaron Loiselle- Lapointe from Environment Canada for providing EV test data for calibrating the battery model. Funding for this work was provided by Natural Resources Canada through the Program of Energy Research and Development. References [1] K. Darcovich, B. Kenney, D.D. MacNeil, M. Armstrong, Control strategies and cycling demands for Li-ion storage batteries in residential microcogeneration systems, in press, 2014, Applied Energy, 141 (2015) [2] K. Darcovich, E.R. Henquin, B. Kenney,I.J. Davidson, N. Saldanha, I. Beausoleil-Morrison, Higher capacity lithium ion battery chemistries for improved residential energy storage with micro-cogeneration, Applied Energy, 111 (2013), [3] L. Gao, S. Liu, R. Dougal, Dynamic Lithium-Ion Battery Model for System Simulation, IEEE Trans. Components and Packaging Technologies. 2002; 25: [4] B.G. Pollet, I. Staffell, J.L. Shang, Current status of hybrid, battery and fuel cell electric vehicles: From electrochemistry to market prospects, Electrochimica Acta, 84 (1) 2012, [5] F.R. Kalhammer, B.M. Kopf, D.H. Swan, V.P. Roan, M.P. Walsh, Status and prospects for zero emissions vehicle technology, Report of the ARB Independent Expert Panel, [6] K. Darcovich, S. Recoskie, D.D. MacNeil, J. Puch, Partial depth of discharge and state of charge functionality as related to capacity fade in lithium ion batteries, in press, 2015, J. Appl. Electrochem. [7] A. Marongiu, M. Roscher, D.U. Sauer, Influence of the vehicle-to-grid strategy on the aging behavior of lithium battery electric vehicles, Applied Energy, 137 (2015), [8] Environmental Protection Agency, Dynamometer drive schedules, m, accessed January Authors Hajo Ribberink has a M.Sc. degree in Applied Physics. He uses modelling and simulation to assess new and innovative technologies for the production or use of electricity and/or heat. He leads CanmetENERGY s research in the field of integration aspects of electric vehicles and the electrical grid. Ken Darcovich completed a PhD in Chemical Engineering. His research at NRCC features transport-based multiphysics simulation, and has been applied to several technology areas, most recently for lithium-ion batteries. Fleurine Pincet will complete the Diplôme d ingénieur at ICAM-Lille in EVS28 International Electric Vehicle Symposium and Exhibition 11