Real-World Mobility and Environmental Data for the Assessment of In-Vehicle Battery Capacity Fade

: This work develops scenario-based analyses for predicting in-vehicle performance degradation of automotive traction batteries. It combines recent capacity performance-based models of NCM-LMO Li-ion (Nickel Cobalt Manganese Oxide—Lithium Manganese Oxide) variant batteries with real-world vehicle driving data from different geographical areas of Europe. The analysis addresses different battery and vehicle architectures (PHEVs (Plug-in Hybrid Electric Vehicles) and BEVs (Battery Electric Vehicles)) combined with different recharging strategies and mobility patterns and environmental temperatures. The mobility pattern datasets used in this analysis refer to six European cities and include up to 508,609 private vehicles, corresponding to 1.78 billion GPS records, 9.1 million trips and parking events and a total driven distance of 106.1 million kilometers. The results show the effect that the environmental temperature, the recharging power, and the driven kilometers have on the calendar and cycling aging. The majority of the combinations of the considered vehicle architectures and recharge strategies do not lead to battery capacity drop below 80% of its nominal value in less than ﬁve calendar years for a usage proﬁle of up to 1000 km/month. forward in the topic of in-vehicle battery durability assessment combining calendar and cycle capacity fade models for electriﬁed vehicle traction batteries with large-scale real-world driving data. The presented approach might contribute to determining the key inﬂuencing parameters of the phenomenon in view of new regulatory frameworks.


Introduction
Battery durability is a key element for evaluating the economic, social, and environmental impact of electrified vehicles [1][2][3]. Loss of environmental performance is important in particular because governmental regulatory compliance programs often credit electrified vehicles with a certain level of expected environmental benefit, which might fail to be realized over the life of the vehicle if sufficient battery degradation occurs. In addition to changes in driving range and energy consumption, for hybrid electric vehicles that are equipped with both a conventional and electric powertrain, the criteria for pollutant emissions from the conventional powertrain could be impacted by the degradation of the battery [4]. The problem of establishing battery durability for representative usage scenarios, chemistries, and configurations is extremely complex. Hence, a better understanding of the degradation mechanisms relevant for automotive applications is highly desirable, together with a quantification of the battery life in real-world use conditions. Several studies have been carried out in this respect to determine the key mechanisms responsible for electrochemical aging of battery cells, such as [5,6], and hence, loss of energy capacity and decrease of the output power of the cell [7][8][9].
This work presents in-vehicle battery durability performances of different electrified vehicles and usage conditions. The work relies on performance-based models [10][11][12][13][14][15][16][17] retrieved from literature and implemented in the EU JRC Transport tEchnology and Mobility Assessment (TEMA) platform [18], developed to explore the potential of big data in support to transport policy assessments. The work is an extension of [19], where EU-wide scale mobility driving patterns from conventional fuel vehicles [20] are the inputs for TEMA. They consist of twenty databases of navigation data for a total of 691,751 monitored The results of this work and the flexibility shown by the TEMA battery aging analyses contribute to informing the discussion on the new vehicle global technical regulation (GTR) on in-vehicle battery durability under development within the United Nation Economic Commission for Europe (UN ECE) Electric Vehicles and Environment Informal Working Group (EVE IWG). This group is one of the IWGs of the Working Party on Pollution and Energy (GRPE) [24] subsidiary body of the World Forum for Harmonization of Vehicle Regulations (WP.29) that prepares regulatory proposals on pollution and energy efficiency.

TEMA Platform
JRC TEMA [18] is a modular big data platform designed to reproduce mobility behaviors of vehicles from datasets of trips collected on conventional fuel vehicles by means of GPS [21][22][23]. It is used to quantify the possible impacts of new vehicle technologies on real-world mobility and develop scenarios to assess the impact of policy actions in transport [25][26][27][28][29][30][31][32]. A comprehensive overview of TEMA with applications and results is provided in [18]. The platform has been extended with the calendar and cycle capacity fade models [3]. The advantage of adopting TEMA for estimating the capacity fade of lithium-ion batteries is that the platform allows for combining state-of-the-art cycle and calendar fade models with large-scale real-world driving data representative of thousands of vehicles and millions of kilometers [20,33]. This enables the simulation of a large variety of EV deployment scenarios with different driving styles, recharge patterns, vehicle architectures, and environmental conditions. This work focuses on six province areas: Modena (IT), Florence (IT), Amsterdam (NL), Brussels (BE), Paris (FR), and Luxembourg (LU), mostly or exclusively including passenger vehicles, as reported in Table 1. The results of this work and the flexibility shown by the TEMA battery aging analyses contribute to informing the discussion on the new vehicle global technical regulation (GTR) on in-vehicle battery durability under development within the United Nation Economic Commission for Europe (UN ECE) Electric Vehicles and Environment Informal Working Group (EVE IWG). This group is one of the IWGs of the Working Party on Pollution and Energy (GRPE) [24] subsidiary body of the World Forum for Harmonization of Vehicle Regulations (WP.29) that prepares regulatory proposals on pollution and energy efficiency.

TEMA Platform
JRC TEMA [18] is a modular big data platform designed to reproduce mobility behaviors of vehicles from datasets of trips collected on conventional fuel vehicles by means of GPS [21][22][23]. It is used to quantify the possible impacts of new vehicle technologies on real-world mobility and develop scenarios to assess the impact of policy actions in transport [25][26][27][28][29][30][31][32]. A comprehensive overview of TEMA with applications and results is provided in [18]. The platform has been extended with the calendar and cycle capacity fade models [3]. The advantage of adopting TEMA for estimating the capacity fade of lithium-ion batteries is that the platform allows for combining state-of-the-art cycle and calendar fade models with large-scale real-world driving data representative of thousands of vehicles and millions of kilometers [20,33]. This enables the simulation of a large variety of EV deployment scenarios with different driving styles, recharge patterns, vehicle architectures, and environmental conditions. This work focuses on six province areas: Modena (IT), Florence (IT), Amsterdam (NL), Brussels (BE), Paris (FR), and Luxembourg (LU), mostly or exclusively including passenger vehicles, as reported in Table 1.

Battery Aging Models
Battery degradation is the result of electrochemical aging, a process that degrades the electrical, chemical, and mechanical properties of a cell. Several performance-based models from the literature [11][12][13][14][15][16][17] have been implemented in TEMA. The different combinations of models to calculate the capacity and power fade of three different cell chemistries, LiFePO4, NCM (LiNi 1/3 Co 1/3 Mn 1/3 ) +spinel Mn and NCM-LMO (LiNi 1/3 Co 1/3 Mn 1/3 +LiMn 2 O 4 ) are reported in Table 2. These models are typically calibrated using experimental datasets and despite the fact that their validity is confined to the boundaries of the experimental data used for calibrating the model, performance-based models can provide good results that highlight the direct link between aging and its influencing factors [34]. Details information for all the aging models implemented in TEMA are reported in [3]. TEMA has also been generalized for using fitting coefficients of calendar and cycling aging cell test data with given equations and parameters.
This work focuses on the capacity fade according to Model 5 of Table 2, where the calendar aging of NCM is given by [14,15] and the cycling aging by [16]: where Q loss-calendar is the calendar loss, Q loss-cycle is the cycling loss, T is the cell temperature in K, t is the calendar time in days, Ah is the total ampere-hours exchanged by the battery and SOC min is the minimum State of Charge value reached during the cycle. The total capacity fade is calculated as net of the capacity fade reserve. This is a region of the SOC that is used to balance the loss of capacity of the cell during the first life. The total capacity fade of the vehicle is, therefore, calculated as per Equation (5), which considers calendar aging and cycle aging as additive components of total aging, according to [17,35,36]: The capacity fade models adopted in this work describe cell-level operation phenomena. A series of assumptions, as described in [3], are applied to scale them up at the vehicle-level in order to be combined with activity data within TEMA and simplify a complex phenomenon as the in-vehicle battery durability. TEMA model aging results have been compared with real-world data from the field, showing a good agreement in end of life (EoL) years estimates [3,34].

Reference Vehicles, Battery Architectures, and Recharge Strategies
To assess the battery durability performance, the authors considered in this work five reference battery packs associated with vehicle architectures available on the market. The vehicles are generically labeled as PHEV (Plug-in Hybrid Vehicle) 1, 2, and 3 (PHEV-1, PHEV-2, PHEV-3), and BEV (Battery Electric Vehicle) 1 and 2, (BEV-1, BEV-2). PHEV-1 adopts a T-shaped battery pack of 192 pouch cells (2P-96S electric architecture) for a total of 16 kWh nominal capacity, PHEV-2 adopts a parallelepiped battery pack of 95 prismatic cells (95S electric architecture) of 8.8 kWh nominal capacity, PHEV-3 adopts a parallelepiped battery pack of 80 prismatic cells (80S electric architecture) of 12 kWh nominal capacity, BEV-1 adopts instead a parallelepiped battery of 192 pouch cells (48S-2P-2S electric architecture) for a total of 24 kWh nominal capacity, while BEV-2 has a flat battery pack, sometimes referred to as "skateboard", of 6912 cylindrical cells (16S-72P-6S electric architecture) for a total of 85 kWh nominal capacity. Each battery is assumed to have a beginning of life (BoL) usable energy equal to 75% of the nominal capacity, i.e., 12 kWh for the PHEV-1, 6.6 kWh for the PHEV-2, 9 kWh for the PHEV-3, 18 kWh for BEV-1 and 63.75 kWh for BEV-2. Each battery is considered to reach its EoL when the usable energy becomes equal to 80%. In addition, each battery allows for an energy reserve value equal to 25% of the nominal capacity for PHEVs and 15% of the nominal capacity for BEVs. Table 3 reports the assumed usable energy criteria for the reference vehicles together with their distance-specific energy consumption from real drive-tests [37][38][39]. The presented results should not be taken as definitive predictions for their durability, neither must they be intended to specifically address these vehicles because of their different battery chemistries in respect of those considered in this work and the several assumptions made to address this complex phenomenon. TEMA replicates the driving behavior from the selected datasets in combination with recharge behavioral models, which aim at representing the most likely recharging behaviors, depending on the individual choices of the driver and on the recharge infrastructure available. Among the sixteen strategies of TEMA [19], three are presented in this paper, as reported in Table 4. Strategy 1 (Long-Stop Random AC) requires a vehicle stop longer than 120 min. It is assumed that charging uses conventional Italian recharge infrastructure (i.e., AC, single-phase at 3.3 kW, IEC 62196 Mode 1/2 [25]), and it is representative of a recharge that can take place at home or wherever the vehicle is subject to a long parking event (e.g., offices, shopping malls, airports, or train station parking lots, etc.). The recharge power is scaled down to a constant value of 2 kW to account for the recharging profile (i.e., power modulation applied from the vehicle), and the recharge is subject to a random-generated threshold parameter between 0 and 1. Recharging is assumed to occur if this number is higher than 0.6 (i.e., 40% of the probabilities). This random threshold represents three possible situations: There is no recharge station where the vehicle is parked, the recharge station is not available, or the driver does not connect the vehicle to the recharge station (forgets or chooses not to recharge). Strategy 2 (Short-Stop Random DC) requires a vehicle stop longer than 20 min, (i.e., short-stop) and the recharge is done with high-power DC (55 kW, IEC 62196 Mode 4, [25]). The random threshold is the same as per Strategy 1, and charging is only assumed to occur in 40% of stops. It is representative of the recharge that could take place in parking lots equipped with fast-charging devices. In this case, the recharge power is scaled down to a constant value of 40 kW to account for the recharging profile. Strategy 3 (Night AC) recharges a vehicle when it is parked in a specific time window between 22.00 and 07.00, and the parking event is longer than 4 h. No random threshold is applied. The recharge is applied in AC (3.3 kW, IEC 62196 Mode 1/2, [25]), and the power is scaled down to a constant value of 2 kW, as in Strategy 1. The Night AC (Strategy 3) represents the typical recharge that might happen overnight at home. The charging efficiency has been set equal to 95% [25] for all strategies, according to measured values [38].  Figure 2 depicts the monthly maximum and minimum temperatures in several province areas for the year during which the data have been collected [40]. Moreover, the monthly temperatures referring to Stockholm and Lisbon have been included in the plot being them used for studying the effect of cold and warm environmental temperatures on the in-vehicle battery durability. Analyzing Europe as a whole, in the first week of March 2015, the average minimum temperature was around 0 • C [40] in Amsterdam, Brussels, and Paris. Moreover, the average maximum temperature was 9 • C in Amsterdam, 14 • C in Brussels, 12 • C in Paris. Paris has slightly higher temperatures in comparison to the other provinces.   Figure 2 depicts the monthly maximum and minimum temperatures in several province areas for the year during which the data have been collected [40]. Moreover, the monthly temperatures referring to Stockholm and Lisbon have been included in the plot being them used for studying the effect of cold and warm environmental temperatures on the in-vehicle battery durability. Analyzing Europe as a whole, in the first week of March 2015, the average minimum temperature was around 0 °C [40] in Amsterdam, Brussels, and Paris. Moreover, the average maximum temperature was 9 °C in Amsterdam, 14 °C in Brussels, 12 °C in Paris. Paris has slightly higher temperatures in comparison to the other provinces.      [20]. The top and bottom pictures are complementary. The derived mobility behavior of passenger cars is similar for all databases, periodically repeated in the days of the week, exhibiting three traffic peaks from Monday to Friday, i.e., in the morning (approximately at 7.30), at noon, and in the evening (approximately at 18.30). On the weekend (Saturday and Sunday), the shape of the curves is different, showing mainly two peaks, approximately at 12.00 and at 19.00. Some vehicles are in motion during late Saturday night, although values above 99% of the vehicles are always parked between 1 and 5 o'clock in the morning. The share of the vehicles in motion at the same time never exceeds 11.7% of the complete fleet for Modena and 10.4% for Florence, with a mean value of 4.3% for Modena and 4.5% for Florence. It is observed that Amsterdam shows peak values below 7% while Brussels, Paris, and Luxembourg well below 5% in line with [41,42]. Table 5 reports the main trip indicators for the analyzed data. The averaged trip has a length between 5 and 20 km, the averaged trip duration varies between 10 and 20 min and the trip speed varies between 26 and 54 km/h. The average parking duration lasts between 2 and 12 h. These values are average values on the full sample available in each dataset.  [41,42]. Table 5     Following the approach depicted in [25], the work focuses only on the users that can drive all their trips electric on the analyzed time period. The results show that vehicles with larger battery packs accumulate greater electric kilometers, a higher number of electric trips, and a greater number of recharges per month, but not a high variation of the SOC per recharge due to their high battery capacity, as shown in Figure 4. For the majority of the vehicles, the number of trips ranges between 60 and 120 trips/month (i.e., 2-4 trips/day), with 15-30 recharging events per month in AC and 30-60 in DC [3]. Following the approach depicted in [25], the work focuses only on the users that can drive all their trips electric on the analyzed time period. The results show that vehicles with larger battery packs accumulate greater electric kilometers, a higher number of electric trips, and a greater number of recharges per month, but not a high variation of the SOC per recharge due to their high battery capacity, as shown in Figure 4. For the majority of the vehicles, the number of trips ranges between 60 and 120 trips/month (i.e., 2-4 trips/day), with 15-30 recharging events per month in AC and 30-60 in DC [3].  Figure 5 shows the calendar, the cycling aging, and the calendar plus cycle aging minus the reserve in the function of the years for the several provinces considered in the analyses, given the mobility patterns of each geographical area and temperature, as described above.  Figure 5 shows the calendar, the cycling aging, and the calendar plus cycle aging minus the reserve in the function of the years for the several provinces considered in the analyses, given the mobility patterns of each geographical area and temperature, as described above. Figure 5 refers to the case of BEV-1 and Recharge Strategy 1 as an example. Each scenario is further broken down into five usage bins, representing different categories of users, e.g., from users that drive less than 500 km/month (6000 km/year), to users that drive more than 2000 km/month (24,000 km/year), and they are key elements for classifying the EoL performance. The usage bins corresponding to more than 1500 km/month are not present for most of the provinces, showing that the combination of these specific vehicle characteristics and recharge strategy does not allow driving long distances in the month. This might be possible considering a recharge strategy with a higher frequency of recharge and higher recharging electric power. The Battery Management System (BMS) that regulates the temperature of the battery is assumed active only during battery cycling, i.e., driving or charging the vehicle, and not during the calendar aging, i.e., vehicle parked without being charged. Hence, the environmental temperature of the areas has a larger effect on the calendar aging. The cycling aging is higher in the areas corresponding to higher average driven distance and charging power. Paris and Florence provinces correspond to a higher calendar and cycling aging in respect to the other geographical areas, given the same recharging strategy, due to the average long driving distance and characteristic environmental temperatures. Table 6 presents the EoL estimates in years and the years needed to reach both 100,000 and 160,000 km of cumulative kilometers, per each recharge strategy, vehicle type, and geographical area. The predicted number of years needed to reach 100,000 km and 160,000 km of cumulative kilometers for a specific usage bin is calculated using average km per month of that user scenario. Each value is then colored red if it is below 5.0 years, yellow if it is between 5.0 years and 10.0 years, and green if it is above 10.0 years. The coloring criterion is purely arbitrary, with the sole aim of providing the reader with a simpler visualization of the results [3]. The results show that the users that drive up to 1500 km/month (i.e., first three bins) experience EoL beyond five years. It is interesting to note that through the first two bins, the capacity fade EoL criteria (<80% initial capacity) is normally reached before 100,000 km for BEV-1. Results for users in the third bin have EoL predictions that vary between the years to EoL and kilometers threshold, depending on battery architecture, charging strategy, and vehicle technology. Bins 4 and 5 are considered higher kilometer drivers, and EoL predictions suggest that the kilometer threshold is more likely to be reached before the capacity fade EoL threshold for all the scenarios. Red scenarios are only predicted for kilometers accumulation threshold for the considered battery chemistry. BEV-2 never shows a predicted lifetime below 10 years due to its battery architecture that optimizes the Ah-throughput and its large battery capacity. By comparing recharge strategy #2 (fast charge) with the other strategies (slow charge), aging is predicted to occur slightly more quickly with fast DC charging, though this does not seem to be a dominant effect for BEV-2. Figure 6 shows the calendar aging, the cycling aging, and the calendar plus cycle aging minus the reserve in function of the years of the three reference PHEVs for the province of Modena. High kilometers per month are not possible to be driven in pure electric mode by the PHEVs. Calendar aging is higher than cycling aging. However, for these chemistries and vehicle technologies, capacity fade is shown to become significant after more than 10 years. Aging appears to be higher for PHEV-3. A comparison has been performed with a set of real-world aging data for PHEVs, confirming the dependence on the results of the assumed capacity reserve for PHEVs [43].  Figure 5 refers to the case of BEV-1 and Recharge Strategy 1 as an example. Each scenario is further broken down into five usage bins, representing different categories of users, e.g., from users that drive less than 500 km/month (6000 km/year), to users that drive more than 2000 km/month (24,000 km/year), and they are key elements for classi-  EoL below 5.0 years (purple); EoL above or equal to 5.0 and below 10.0 years (yellow); EoL above or equal to 10.0 years (green). EoL below 5.0 years (purple); EoL above or equal to 5.0 and below 10.0 years (yellow); EoL above or equal to 10.0 years (green).

Capacity Fade Results: Effect on Warm Environmental Temperature
To estimate the effect of warm environmental temperatures on the battery capacity fade, the duty cycle of Modena is combined with the environmental temperature of Lisbon [40]. If a vehicle is driven in a warm environment, such as that of Lisbon in summer, it is assumed that the user operates the air conditioning system to cool the vehicle cabin down. This is reflected in higher energy consumption while driving the vehicle. In this simulation, a constant increase of 15% in the driving energy consumption due to the air conditioning is assumed [38]. The basic scenario (Modena province duty cycle and environmental temperature) is compared to the warm environmental temperature scenario (Modena province duty cycle with the temperature of Lisbon) and its combination with the usage of the Heating, Ventilation, and Air Conditioning (HVAC) system (Modena province duty cycle with HVAC system in operation and the temperature of Lisbon) in Figure 7 for BEV-1 and Recharge Strategy 1. The effect of the warmer temperature is visible in the calendar aging plots (red bars) in each figure in comparison to the base scenario (blue bars) (the BMS is not in operation). The HVAC system operation affects the cycling aging (yellow bars) but in slightly the same way as the environmental temperature during driving and charging. Figure 8 shows the results assuming the same scenario with HVAC system in operation and warm temperature for different geographic areas.

Capacity Fade Results: Effect on Warm Environmental Temperature
To estimate the effect of warm environmental temperatures on the battery capacity fade, the duty cycle of Modena is combined with the environmental temperature of Lisbon [40]. If a vehicle is driven in a warm environment, such as that of Lisbon in summer, it is assumed that the user operates the air conditioning system to cool the vehicle cabin down. This is reflected in higher energy consumption while driving the vehicle. In this simulation, a constant increase of 15% in the driving energy consumption due to the air conditioning is assumed [38]. The basic scenario (Modena province duty cycle and environmental temperature) is compared to the warm environmental temperature scenario (Modena province duty cycle with the temperature of Lisbon) and its combination with the usage of the Heating, Ventilation, and Air Conditioning (HVAC) system (Modena province duty cycle with HVAC system in operation and the temperature of Lisbon) in Figure 7 for BEV-1 and Recharge Strategy 1. The effect of the warmer temperature is visible in the calendar aging plots (red bars) in each figure in comparison to the base scenario (blue bars) (the BMS is not in operation). The HVAC system operation affects the cycling aging (yellow bars) but in slightly the same way as the environmental temperature during driving and charging. Figure 8 shows the results assuming the same scenario with HVAC system in operation and warm temperature for different geographic areas.
the usage of the Heating, Ventilation, and Air Conditioning (HVAC) system (Mo province duty cycle with HVAC system in operation and the temperature of Lisbo Figure 7 for BEV-1 and Recharge Strategy 1. The effect of the warmer temperatu visible in the calendar aging plots (red bars) in each figure in comparison to the scenario (blue bars) (the BMS is not in operation). The HVAC system operation affect cycling aging (yellow bars) but in slightly the same way as the environmental temp ture during driving and charging. Figure 8 shows the results assuming the same scen with HVAC system in operation and warm temperature for different geographic are   Table 7 shows the EoL in years and years needed to reach 100,000 km and 160,00 for the Modena province area with the environmental the temperature of Lisbon both BEV-1 and BEV-2 with Recharge strategy 1 and 2 to be compared to Table general, a lower number of years are needed to reach the EoL.    Table 7 shows the EoL in years and years needed to reach 100,000 km and 160,000 km for the Modena province area with the environmental the temperature of Lisbon and both BEV-1 and BEV-2 with Recharge strategy 1 and 2 to be compared to Table 6. In general, a lower number of years are needed to reach the EoL.   Table 7 shows the EoL in years and years needed to reach 100,000 km and 160,000 km for the Modena province area with the environmental the temperature of Lisbon and both BEV-1 and BEV-2 with Recharge strategy 1 and 2 to be compared to Table 6. In general, a lower number of years are needed to reach the EoL. EoL below 5.0 years (purple); EoL above or equal to 5.0 and below 10.0 years (yellow); EoL above or equal to 10.0 years (green).

Conclusions
The visible effects of electrochemical aging of battery cells are a loss of energy capacity and a decrease of the output power of the cell. An aging model must aim at reproducing these two effects while considering driving cycles, environmental and battery temperature, charging rate and frequency, and parking time. To this purpose, the EU JRC has developed a dedicated in-vehicle battery durability assessment module within its TEMA platform, based on performance-based models and large-scale real-world driving data. JRC TEMA is a modular big data platform designed to reproduce mobility behaviors of vehicles from datasets of navigation system data of conventional fuel vehicles and quantify possible impacts of new vehicle technologies on real-world mobility while supporting transport policy assessment.
The mobility pattern datasets used in this analysis refer to six cities, Modena (IT), Florence (IT), Amsterdam (NL), Brussels (BE), Paris (FR), and Luxembourg (LU), and include up to 508,609 private vehicles, corresponding to 1.78 billion GPS records, 9.1 million trips and parking events and a total driven distance of 106.1 million km. The analysis focuses on calendar and cycle capacity fade of NCM LMO Li-ion variant [14][15][16], PHEV and BEV batteries architectures, combining different recharging strategies of different power and profiles with different driving duty cycles from the six cities mentioned above and related yearly environmental temperatures. In the studied scenarios, the cycling aging is affected more by the distance driven than the temperature, so its contribution is higher in the areas corresponding to higher average driven distance and charging power. The HVAC system operation affects the cycling aging in slightly the same way as the environmental temperature. Additionally, by comparing recharge strategy #2 (fast charge) with strategy #1 and #3 (slow charge), aging is predicted to occur slightly more quickly at first, though this does not seem to be a dominant effect for BEV-2. TEMA model aging results have been compared with real-world data from the field, showing a good agreement in EoL years estimates [3,34].
Despite the assumptions and limitations of the aging model, the results constitute a step forward in the topic of in-vehicle battery durability assessment combining calendar and cycle capacity fade models for electrified vehicle traction batteries with large-scale real-world driving data. The presented approach might contribute to determining the key influencing parameters of the phenomenon in view of new regulatory frameworks.
Author Contributions: Conceptualization, methodology, software, formal analysis, data curation, E.P.; writing-original draft preparation, E.P.; writing-review and editing, G.M. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest.