Jean-Michel Vinassa

Impact of Calendar Life and Cycling Ageing on Supercapacitor Performance

Supercapacitors (SCs), which are also known as ultracapacitors, are commonly based on porous activated carbon electrodes and electrostatic charge-storage mechanisms. Carbon electrodes are supposed to be chemically and electrochemically inert, and the electrostatic nature of the charge-storage mechanism is highly reversible. These properties should assure that SCs have an infinite shelf life. However, in practice, SC cells exhibit performance fading when they are used for months. The purpose of this paper is to compare the performance fading of SCs in two types of tests. The first accelerated test is the calendar life test, which evaluates cell degradation when cells are used at elevated voltages and temperatures; these degradations are observed in cells during rest periods. The second test is the power cycling test, which corresponds to degradations during the actual use of SCs as peak power sources in hybrid electric applications. In this paper, we use an impedance model to monitor changes in cell behavior. This analytical impedance model is based not only on physical quantities, such as double-layer capacitance, specific area, and electrolyte conductivity, but on elementary electrode dispersion as well. This model is validated in the frequency domain using the initial state of the tested SC for different temperature and voltage configurations. Then, we present and specify two types of tests. A periodic characterization based on impedance spectroscopy is done to quantify impedance changes during the ageing tests. The obtained results confirm that the impedance real part is increasing and that the capacitance is decreasing during the two ageing tests; however, the way they change is different. Indeed, the shape of the impedance real part is conserved in the calendar life test, whereas it is distorted in the power cycling test. This difference will be explained in terms of the evolution of the impedance model parameters. This paper confirms that the two ageing tests affect the SC performances differently.

Embedded Fractional Nonlinear Supercapacitor Model and Its Parametric Estimation Method

This paper deals with a supercapacitor (SC) onboard model and its online identification procedure for embedded applications. To take into account its nonlinear behavior, the strategy used in this paper consists, as the first step, in the application of the porous electrode theory to the SC and the approximation of the resulting model. A set of fractional linear systems which are represented by a differential equation involving fractional derivatives is then obtained. Each element of this set represents the behavior of the SC only around one operating voltage. A global nonlinear model is then deduced through an integration method. An online identification procedure has been developed for this nonlinear model. This time identification is based on the mean least square method. Its time behavior has been compared with that of an SC cell for a specific current profile with different levels.

Strategy for designing accelerated aging tests to evaluate IGBT power modules lifetime in real operation mode

Accelerated aging tests are widely used in electronics industry to validate the technological choices of packaging, especially when significant lifetime is required. The thermal coefficient of expansion mismatch between the soldered materials results in crack propagation in the solder joints, which irreversibly increases the thermal resistance of the assembly and progressively leads to the failure of the module. Accelerated testing is very expensive and can last a long time, so that such experiments must be optimized. In addition, thermomechanical behavior of the assembly under realistic operating conditions for long times can be studied in shorter times thanks to finite element simulations. Non-linear finite element simulations have been carried out to make a correlation between accelerated aging tests and real operation. Stress-strain history under representative thermal loading has been determined in both cases. The energy dissipated in solder joints has been calculated and has been used as an estimator of the damage in the solder joint. Finally, thermal fatigue experiments with representative samples have allowed validating the previous results. This study is a step toward the understanding of the correlation between accelerated testing and actual operating conditions.

Adaptive voltage estimation for EV Li-ion cell based on artificial neural networks state-of-charge meter

This paper reports some results relating to adaptive cell modeling from neural network state-of-charge (SOC) estimation in a full-electric-vehicle (EV) application. The cells in question are commercialized ones, Lithium-ion Polymer based, with a nominal capacity of about 100 Ah and dedicated to energy applications. Using a recurrent neural network, we developed a SOC predictor that takes into account operational conditions. More importantly, the predictor allows very precise SOC estimation, therefore allowing the vehicle controller to confidently use the battery packs full operating range without problem of over- or under-charging cells. In this work, the estimated SOC values helped to estimate the parameters of an adaptive-dynamic battery model using RLS algorithm with time-dependent forgetting factor. Simulation results confirmed the accuracy of the terminal voltage estimation of the battery.

Ultracapacitors self discharge modelling using a physical description of porous electrode impedance

The main goal of this paper is to establish a link between the ultracapacitors impedance at low frequency and slow phenomena such as self-discharge and charge recovery. Usually these phenomena are studied in the time domain. However, additional information can be extracted from specific characterization methods in the frequency domain like impedance spectroscopy. In this way, we intend to introduce an analytical method to characterize the ultracapacitors only by using impedance spectroscopy results and so to simplify some ultracapacitors tests like power cycling and calendar life tests. Firstly, we introduce a complete impedance model of ultracapacitor, which is based on the main physical behaviour in the porous electrodes and at the electrode- electrolyte interface. A specific identification method is used in order to determine the model parameters from the impedance spectroscopy results. This model has a good accuracy, especially in the low investigated frequency range [1 mHz,100 Hz]. Then, we focus only in the dynamic behaviour in the low frequency range using some justified simplifications. The link with voltage decay is discussed and validated by using experimental results of self-discharge tests. Then a mathematical relation is introduced in order to predict the self-discharge rate from impedance spectroscopy results. Finally, a time domain identification procedure of the ultracapacitors model is introduced. It allows obtaining an accurate model of the dynamic behavior without using the impedance spectroscopy. The proposed model has been validated thanks to experimental results. The validation tests have been performed using a large cell 2700F/2.7 V for a significant test duration. The results show a good agreement between measured and simulated results during the relaxation periods over few hours.

Real-time SOC and SOH estimation for EV Li-ion cell using online parameters identification

In this paper, we present a real-time adaptive estimation of the state of charge (SOC) and the voltage of a high-energy-density lithium-ion cell used in electric vehicle (EV). We developed a simple and linear-recursive battery SOC and voltage models that proved their efficiency regarding the comparison between simulation results and real data from power cycling with ECE 15 European driving cycle tests. Based on the on-line estimation of battery model parameters, a recursive least squared algorithm (RLS) with a time-variant forgetting factor is used to describe the battery dynamic behavior which can vary within each experiment. The estimations have shown very good performances, the maximum and the mean relative modeling error do not exceed 1% for both of the estimators.

Study of ultracapacitors dynamic behaviour using impedance frequency analysis on a specific test bench

This paper presents a study on the ultracapacitors dynamic behaviour. Frequency analysis is chosen as the tool to perform this investigation. First, the impedance spectroscopy principles are presented. Then the experimental results, obtained by applying impedance spectroscopy with high current levels, are presented and discussed. Ultracapacitors show a particular behaviour which leads to it specific model composed by a small serial inductance associated with a four RC-branches network.

Switching performances comparison of 1200 V punch-through and nonpunch-through IGBTs under hard-switching at high temperature

Switching performances comparison of 1200 V punch-through (PT) and nonpunch-through (NPT) IGBTs under hard-switching at high temperature has been studied in detail with the aid of extensive measurements at various temperatures, gate resistances and clamping voltages. The test circuit allows the characterisation of the switching losses of the IGBTs under hard-switching independently of those due to the freewheeling diode. We show that these losses are quite similar to each other during turn-on, but more important for PT-IGBT during turn-off. Moreover, NPT-IGBT is less sensitive to temperature and clamping voltage variations than PT-IGBT.

Ultracapacitors modeling improvement using an experimental characterization based on step and frequency responses

This paper deals with ultracapacitors modeling that has been improved for high power source purpose such as in hybrid and electric vehicles applications. An experimental method based on the combination of step and frequency response has been implemented and used in a dedicated test bench. First, a critical overview of available ultracapacitors models is made and a new approach is proposed. Then, after constant current and impedance spectroscopy tests, the experimental results are presented and discussed. Finally, the exploitation of the tests results leads to the identification of a proposed model parameters. This electric circuit is validated by comparing simulation to experimental curves.

Control of a hybrid Energy Storage System using a three level neutral point clamped converter

The increasing penetration level of the Renewable Energy Sources (RES) and their variability can compromise the proper operation of an electrical grid. The microgrid is being analysed as a solution to obtain a high penetration of RES in a controlled way. Due to the stochastic nature of the RES, the Energy Storage Systems must be used to maintain the energy and power balance between generation and demand. The limits of the storage technologies that are nowadays available make it necessary to associate more than one storage technology creating a Hybrid Energy Storage System (HESS) in order to satisfy the specifications of a microgrid application: high energy autonomy, power capacity and fast response. This paper analyses by means of simulations and experimentally the feasibility of a promising topology based on a Three Level Neutral Point Clamped (3LNPC) converter. The reduced power losses, improved THD and the fact of being an integrated solution are the advantages of this system. An islanding operation case study where a HESS formed by a SuperCapacitor (SC) and a Vanadium Redox Battery (VRB) supplies power to a resistive AC load that is suddenly reduced is analysed by means of simulations and experimentally. The simulation and experimental results prove the feasibility of the presented topology and of the control system to divide the power between the SC (fast power variations) and the VRB (low frequency part of the power).