EVS28 Experimental behaviour of Li-ion and supercapacitors cells for HEVs under standardized and tailored-life cycle testing

A large European Project, named HCV (Hybrid Commercial Vehicles) started in January 2010 with the participation of 18 European organizations (vehicle manufacturers, components integrators and suppliers, and research organizations) and with the scope to develop and demonstrate the next generation of hybrid heavy duty (HD) commercial vehicles by using various types of storage systems. In this project, seven research (AIT, ENEA, University of Pisa) and industrial organizations (IVECO, Volvo, Magna, DimacRed) from various European countries have been working together to experimentally analyse, with electrical and safety tests, the behaviour of Li-ion and supercapacitor cells and modules to support the design and the optimization of the final storage systems to be installed on different HEV (Hybrid Electric Vehicles): urban buses and commercial vans. This paper summarizes the experimental work carried out at ENEA and is focussed on electrical and safety tests, which fully characterized the selected storage samples according to conventional and testing procedures, tailored on the technical specifications of the HEV under development. Initially, basic characterization testing, together with safety tests, confirmed the technical performances of the two storage technologies, and, subsequently, project-specific testing, including cycle life and accelerating procedures, verified the behaviour in operating conditions, adapted to the selected HEV. The final results substantiated the suitability of the storage systems in powering the commercial hybrid vehicles under development in HCV project, and gave innovative inputs to the definition and validation of mathematical models and control algorithms, not analysed in this paper, to be used in the BMS (battery management systems) for both storage technologies, suitable for thermal management and overall storage control.


Introduction
The hybrid electric vehicles (HEV) can significantly contribute in containing exhaust emissions in urban areas, with improved energy efficiency. In particular, the use of urban buses and commercial vans in hybrid configurations is highly promising in passengers and goods transport in urban areas, because the typical travel missions stabilize the operating conditions of these types of drivetrains around wellcontrolled optimal working points. The current research and development activities on advanced heavy duty (HD) hybrid vehicles (buses, trucks and commercial vans) are aimed in most cases at extremely challenging requirements, with technological and economical breakthroughs and subsequent industrialization concentrated on the drivetrain technologies with the possibilities to use alternative or complementary storage systems to perform key functions (traction assistance to conventional internal combustion engine, regenerative braking and, eventually, pure electric traction mode for a limited range). A large European Project, named HCV (Hybrid Commercial Vehicles) and partially funded by the European Commission (EC), was started in January 2010 and was completed in 2014, with the participation of 18 European organizations (vehicle manufacturers, components integrators and suppliers, research organizations). The HCV project had the scope to demonstrate the status of current HD HEVs in real applications and develop the next generation of cost-effective and more efficient hybrid commercial vehicles by using various types of storage systems, with the final practical objectives to reduce powertrain cost of about 40% and fuel consumption of 30% in a city bus cycle, compared with present hybrid bus technologies. To better assist storage systems suppliers and assemblers and vehicles manufacturers in optimizing components, subsystems and, eventually, complete drivetrains, a dedicated "Energy Storage Systems" SubProject was carried out for the experimental evaluation and in-depth characterization of the technical performances and reliability in normal and extreme conditions of the storage systems in relation to the specific HEV architectures, with focus on two energy storage technologies: lithium-ion batteries and electrochemical supercapacitors (SC) [1,2]. Seven research (AIT -Austrian Institute of Technology, ENEA, University of Pisa) and industrial organizations (IVECO, Volvo, Magna, DimacRed) from three European countries have been working together to experimentally analyse, with electrical and safety tests, the behaviour of Li-ion and supercapacitor cells and modules to support the design and the optimization of the final storage systems to be installed on different HEV (Hybrid Electric Vehicles): urban buses and commercial vans. One of the main objectives of these activities on energy storage has been to verify the basic performances of the samples with electrical, thermal and safety/abuse testing and compare life behaviour in standardized or accelerated duty cycles with respect to the use in driving profiles tailored on the HCV vehicles. ENEA (Italian National Agency for New technologies, Energy and Economic Sustainable Development) has been involved in various phases of the activities by contributing to identifying and defining technical specifications and adapting testing procedures to the specific need of the project and, finally, executing electrical tests on the cells and assembled modules, on which ENEA also executed mechanical tests (vibrations). This paper reports about the main final results achieved by ENEA during basic characterization, life and vibration testing on both storage technologies, based on the testing procedures, tailored on the specific HEV configurations and targeted performances. Initially, the samples under tests and the summary of the testing procedures, reported in details in [3, 4, 5], are described. Subsequently, the main results for each test type are shortly reported and analysed with main conclusions also related to the definition and validation of mathematical models and algorithms for management and control of both storage technologies, not analysed in this paper, but more described in [6, 7, 8].  Table 1 shows basic characteristics of Li and SC cells. During the progress of the project, an industrial assembling process (not part of the project, but as internal industrial innovation in Magna) was developed and applied for assembling Li-ion modules, while for the SC modules an initial generation was improved and a 2 nd and a 3 rd generation were assembled in 2013. A set of 14 Li-ion modules were prepared in the same design chosen for the final system to be installed on the HEV with these final characteristics: 18 cells in series, capacity of 8.5 Ah, voltage of 59.4 V, energy of 505 Wh, peak power of 11.25 kW, and an overall weight with some sensors and connectors of about 10 kg. Similarly, ten SC modules of two different generations were also assembled and made available for testing: one without case and the second one with the case (a third one was also prepared for comparison using a different case material) with the same specifications of the ones used in the final system. The SC modules have these technical specifications: six cells in series, capacitance of 500 F, voltage of 16.2 V, specific energy of 5.96 Wh/kg, specific power of 5.9 kW/kg, and an overall weight with some sensors and connectors of about 4.4 kg.

Electrical testing
The operating data for the definition and validation of mathematical models. To achieve a proper adaptation of testing procedures to the samples' characteristics, the technical specifications of the complete HEV storage systems have been scaled down and adapted to the sample sizes (cells and modules) by using a Battery Size Factor (BSF), intended as an integer number, which is the minimum number of cells expected to be required to meet all the performance and life targets. For example, in the case of the Altra HEV Li-ion ESS (a storage system of 45 kW), the BSF has been roughly rounded to 192, while if the ESS is based on SC the BSF is 144. These BSF are constant for all the tests.

Electrical testing on Li-ion cells
ENEA work has been concentrated on the preparation and execution of the electrical testing on cells with the set-up of a new testing equipment, integrated with a climatic chamber used for controlling the working temperature of the sample. More than 100 Li-ion cells has been characterized and continuously monitored, according to the basic characterization tests to verify technical specifications. Due to the very repeatable technical characteristics of the Li cells, it was not necessary to establish any specific criterion for cell selection from the available set to be used for electrical testing. The Li-ion cells have undergone the following tests: 1

Basic characterization of Li-ion
cells Some examples of the measurements of the main basic characteristics are described hereafter. The capacity determination, after electric formation, was carried out for seven different discharge current rates (from C/2 to 20C, where C is the nominal capacity of the cell). The capacity and the specific energy of each cell at various discharge rates were in line with the manufacturer data, as shown in Table 2 and in  Table 3.

SE*= Specific energy= Wh/kg
The self-discharge test was carried out for two different rest periods (48h and 120h) and two different temperatures. Table 4 summarizes the results of self-discharge tests on some cells by reporting the capacity, measured at the end of the test at RT.  Table 4 confirms a limited self-discharge in all the testing conditions.

Life testing of Li-ion cells
Life testing of Li-ion cells was the core activity of the electrical testing with the scope to verify the behaviour in standard and HCV-specific operating conditions and to estimate degradation of the cell performances (for modelling purposes) and data useful for the energy storage modules/systems integrator and the end users. Life tests were carried out according to the conventional and HCVspecific test sequences: Life cycle test based on HCV profiles, Calendar life in off-load conditions and Accelerated High power battery cycling, with selected parameters (temperature and power profiles) for fast degradation of cell performances. All these tests were periodically interrupted to perform control checks of the basic characteristics of the cells. The HCV-specific Life Cycle Testing was carried out mostly using the proposed HCV commercial van profile (a micro-cycle of 167 s, to be repeated 18 times for one hour work and 180 times for one day work). The power profile of the entire ESS was adequately reduced at cell level using a scale down factor (BSF = Battery Size Factor), based on the number of cells in the fully system, as shown in Figure 1.  The negative effects of storage at higher temperature (60 °C) and low SOC are confirmed by comparing test results. Similar analysis has been carried out also for other key parameters. In the case of the theoretical discharge power in function of SOC, the variation, as shown in Figure 3, increases significantly at low SOC after long storage at higher temperature. Analogously, the experimental data related to the internal resistance (IR, in charge and discharge) in function of SOC confirm the limited variation occurring at 45 °C and a larger fluctuation at low SOC after long storage at the temperature of 60 °C. The Accelerated High power battery cycling was performed on two cells by applying continuously a high power profile (in Figure 4 current and voltage profiles are represented). Three test sequences were repeated on the cells, each lasting 6-7 weeks and starting from an SOC of 55%, as planned, with an initial and periodic parameter check-ups (at the end of each test sequence). Each of the three test sequences was carried out at a different temperature: 23°C, 30 °C and 40 °C. Table 5 summarizes the entire test sequence. At beginning and at the end of each test sequence a parameter check-up has been performed with the acquisition and calculation of the main performance characteristics of each cell under test.  The behaviour of the cells does not change significantly during high power cycle testing at any temperature. These results are confirmed also comparing the other technical characteristics measured or calculated during parameter checkups. Only for exemplification of this behaviour, Figure 5 compares the capacity variations of the two cells before and after high power cycling. Also the attempt to accelerate cell degradation with the temperature has been limitedly effective because the used high temperature of 40 °C has not been able to accelerate significantly ageing of the cells together with the number of high power cycles. For comparison the EIS spectra have been recorded before and after cycling to verify the variations of the cell impedance. Despite of the number of spectra measured and analysed on different cells, the focus has been put on the test and the cell having the major decline during accelerated calendar testing. With these measurements, there was confirmation of a more significant degradation during calendar life at 60 °C. In total, 194 complete standard life cycles have been performed to optimize the standard cycle for the specific testing needs (continuous testing without interruptions) and, then, verify the cell voltage dispersion and equalization needs. Figure  6 presents the discharge capacity measured during life cycling with a standard cycle. It is evident in Figure 6 that there is a significant fluctuation in the capacity values up to the 90 th cycle, due to the continuous corrections on the module voltage to reach a satisfying compromise between the overall capacity delivered by the module with the recommended limit for the voltage of each cell. Life cycle testing has been repeated by applying the ALTRA micro-cycle, adequately scaled down to the module size, by using the same procedure used for the cells. The entire life testing lasted 140 equivalent working days (more than 50% of the targeted 250 working days/year), corresponding to 25,200 micro-cycles and travelled km (each micro-cycle was equivalent to a travelled distance of 1 km). The parameter check-ups results are summarized in Figure 7 that reports the capacity and energy trends at the verification points, measured with a 1C discharge rate. No significant decline of performance has been recorded confirming the high maturity of the cells and the industrial level of the assembly process. The module under test has been also thermally controlled with infrared camera. Figure 8 contains an example of numerous (tens) thermographs taken during life testing. The temperature increase is well below the maximum recommended value of 55 °C, even considering that there is no thermal conditioning system applied to the module. The overall testing analysis on modules gave indications about the modes of cell equalization by using a passive balancing circuit, with possible feedback to the final battery management (BMS) strategy, including thermal control needs.

Electrical testing on SC cells
The test activities have been carried out in the ENEA SC Test Facility, able to perform electrical and thermal testing on single cells and small modules. During the cell testing, the external temperature of the SC cell is measured with a dedicated sensor. Figure 9 displays the test control screen with control data and results elaboration with a continuous vision of the sample under test.

Basic characterization of SC cells
After initial formation cycles (a few complete standard cycles, charge/discharge, repeated to verify nominal working voltage), the capacitance has been measured at room temperature (RT) and at 40 °C. The calculated value at RT of 3020 F confirmed the nominal value of 3000 F. The internal resistance (ESR = Equivalent Series Resistance) was also calculated with a specific test during charge and discharge, giving the following values at RT: ESR(ch)=0.426 mΩ with a current step of 80 A; ESR(disch)=0.416 mΩ with a current step of 60A (the nominal value, indicated by the manufacturer, of ESR (disch) is 0.29 mΩ with a current step of 100A). The fast charge tests have been performed to define the capability in energy storing of the SC in defined current and time ranges. The nominal measured energy content of the cell has been 3.04 Wh. Table 6 summarizes the OCV and energy percentage reached at any step of the profile during fast charge. The fast discharge tests have been also performed. Table 7 summarizes the OCV and energy percentage reached at any step of the test profile during fast discharge. The tests confirmed the excellent behaviour of SC cells during fast charge and discharge rates.
The peak power test was carried out to verify one of the more important features of SC: the high power density. The current and voltage profiles for peak power tests are shown in Figure 10, while Figure 11 summarized the peak power test results. The maximum (specific) peak power reached (for 5s) at a maximum peak current of about 150 A is about 750 W/kg (corresponding to a cell peak power for a continuous period of 5s of 370 W).

Life testing of SC cells
The life cycle tests on the SC cell have been aimed at estimating potential cycle life in HCV application by applying the ALTRA-IVECO driving pattern, referred to the complete vehicle battery power requirement and scaled down from the complete vehicle storage system to the cell level. In addition, in order to accelerate technical performance degradation and better assist modelling definition and validation, the cell has been tested not only at RT but also at the temperature of 40 °C. In total 21,200 micro-cycles have been performed at RT (and 21,200 km travelled). This number of cycles corresponds to about 118 working days of the Altra HCV commercial vehicle: 18 microcycles amount to 1-h cycle; every working day is composed of 10 1-h cycles and, in a year, a duty cycle is completed with 250 working days (45,000 micro-cycles corresponding to 45,000 km of travelled distance in a year). In summary, the cycle life test has simulated about half a year of working time. Figure 12 shows some cycles after the completion of more than 5,000 micro-cycles. The degradation of basic performances, capacitance and ESR (during charge, ESR ch , and during discharge, ESR dch ), during life cycle testing at RT has been very limited and well below the standard end-of-life criteria (loss of 20% of nominal capacitance and doubling, or increase of 100%, of ESR). This conclusion is directly derived from the test results summarised in Figure 13, achieved with periodic parameter check-ups. The SC technical performances are limitedly affected by the more severe micro-cycle derived from HCV application in ALTRA hybrid commercial van, demonstrating a high stability that is expected in relation to the declared cycle life of 1,000,000 of complete charge/discharge cycles (based on a standardized profile at constant current in charge and discharge).
The ALTRA cycle life testing has been repeated by increasing the temperature at 40 °C, as possible accelerating factor (as known in literature, together with working voltage close to the maximum allowed). In total, more than 8,000 micro-cycles (equivalent to an additional working period of about 45 days and about 8,045 km) have been completed. Combined with the cycle life testing at RT, the SC cell has accumulated a total service life of 163 working days (with respect to a planned year operation of 250 days) with two different temperatures corresponding to the minimum and maximum values required by the Altra-IVECO procedure. Periodically, the capacitance has been measured at RT and also at 40 °C to verify degradation. For accumulating as many cycles as possible, it was decided to limit the performance control to capacitance measurements, because the other tests during parameter check-ups required longer time. The final results related to the capacitance measurements at periodic parameter check-ups and the comparison with the same results achieved at RT are summarised in Figure 14. At the end, the test showed a limited degradation of the capacitance as a combined effect of the life cycling and the higher temperature. After a working period of 164 days at two different temperatures, the SC capacitance was reduced of 4.7% (the initial value was 3020 F and the final 2878 F). During all the tests, the SC sample was instrumented with a temperature sensor on the external case to follow all the variations in temperature during any tests: the analysis confirmed, the limited thermal stress (and temperature variation) of the cell case, also due to the not severe working conditions, required by the Altra-IVECO hybrid commercial van application.

Electrical testing on SC modules
ENEA has also carried out basic characterization and life cycle testing on 2 nd Generation SC modules. The Electric Formation and initial  Figure 15 shows the variation of the temperature during normal life testing. The final module capacitance decreased of about 4% at the end of the high temperature cycling, as shown in Figure 16. Finally, thermal analysis with infrared camera was also performed on the SC module at RT with limited increase of temperature, which reached a maximum of 29.1 °C. The reached temperature value even during testing at 40 °C has always been well below the maximum recommended value, giving clear indications about the limited need of thermal management in real operations.

Safety/abuse testing: vibration
ENEA work has been concentrated on the definition and execution of mechanical abuse testing on one Li-ion and two SC modules, with focus on vibration tests. These tests have been carried out with an electro-dynamic shaker in the ENEA Vibration Hall, resulted compatible, after experimental verification, with the selected test procedure.
The complete test was performed as follows: 1. the SOC of the module was put at 80% (HEV application); 2. set the power spectrum density (PSD) vs. frequency as shown in Table 8; 3. test duration of 8 hours for each of the three planes of the module under test.

Conclusions
The experimental characterization of Li-ion and SC cells and modules has been functional to the complete determination of the technical performances and their suitability to the specific applications to the HEV developed in the HCV project. The experimental results substantially have confirmed the general good behaviour of the selected cells and the assembled modules: 1. The characterizations have verified the nominal characteristics of the samples with a negligible spread of values: a confirmation of mature production level. 2. The power capability of Li-ion module is relevant and adequate for the applications. 3. The thermal behaviour of Li modules is considered good because no significant increase in temperature has been measured, even without thermal management (or cooling) system. 4. The cycle life with standard cycles, accelerated cycles and HCV-cycles has been excellent with negligible deterioration of performances for both storage technologies. 5. The thermal behaviour of SC module required some control during charging to limit temperature increase: this result is somehow in contrast with the initial indication of no necessity of cooling system. 6. All the attempts (with different cycles and higher temperatures) to accelerate performance decline of Li and SC module performances have not been effective in giving clear indications on how to model their behaviour and allow for an estimation of SOH (State Of Health).
In conclusions, the complete analysis of the life testing activities at ENEA showed that the Li-ion cells and modules used in the HCV project are quite insensitive to different types of profiles in terms of performance decline, with the only exception of storage at high temperature for the cells.
For the SC cells and modules, the tests on cells showed a limited degradation of the capacitance as a combined effect of the life cycling and of the use of higher working temperature, while minor mechanical damages of the module case during vibration tests had no impact on the electrical performances of the tested module. Ennio Rossi is a mechanical engineer and, since 1994, has been involved in electric and hybrid vehicles, developing and applying test methodologies and carrying out testing on scooters, cars and light commercial vehicles. He is also working on supercapacitor and battery testing and in a program for the realization of hybrid vehicles.

Nomenclature
Vincenzo Sglavo is a mechanical engineer and researcher at ENEA, where he is working in the "Low Environmental Impact Vehicles" Laboratory, especially in the section relating to electrochemical batteries and supercapacitors. His research work is related to storage devices testing, with focus on Lithium-ion and lead-acid batteries, supercapacitors.
Francesco Vellucci is a mechanical engineer and has been working in the ENEA's "Low Environmental Impact Vehicles" Laboratory, especially on electrochemical batteries and supercapacitors. He is involved in preparation of standards in the Italian IEC for electric vehicles and in charge of testing activities on energy storage systems.