Lithium-ion Battery Degradation Research Articles

Onset Potential for Electrolyte Oxidation and Ni-Rich Cathode Degradation in Lithium-Ion Batteries.

High-capacity Ni-rich layered metal oxide cathodes are highly desirable to increase the energy density of lithium-ion batteries. However, these materials suffer from poor cycling performance, which is exacerbated by increased cell voltage. We demonstrate here the detrimental effect of ethylene carbonate (EC), a core component in conventional electrolytes, when NMC811 (LiNi0.8Mn0.1Co0.1O2) is charged above 4.4 V vs Li/Li+—the onset potential for lattice oxygen release. Oxygen loss is enhanced by EC-containing electrolytes—compared to EC-free—and correlates with more electrolyte oxidation/breakdown and cathode surface degradation, which increase concurrently above 4.4 V. In contrast, NMC111 (LiNi0.33Mn0.33Co0.33O2), which does not release oxygen up to 4.6 V, shows a similar extent of degradation irrespective of the electrolyte. This work highlights the incompatibility between conventional EC-based electrolytes and Ni-rich cathodes (more generally, cathodes that release lattice oxygen such as Li-/Mn-rich and disordered rocksalt cathodes) and motivates further work on wider classes of electrolytes and additives.

Lithium-Ion Battery Degradation and Capacity Prediction Model Considering Causal Feature

Accurate life prediction of lithium-ion batteries is essential for the safety and reliability of smart electronic devices, and data-driven methods are one of the mainstream methods nowadays. However, existing prediction methods suffer from the problems such as lack of practical meaning of features and insufficient interpretability. To address this problem, this article proposes a battery degradation and capacity prediction model based on the Granger causality (GC) test and the long short-term memory network. First, initial health indicators are set from the monitoring data of the battery. Second, the vector autoregressive model and the GC test are used to select causal features that are associated with capacity degradation. Then, the impulse response analysis approach is proposed for the first time to analyze the exact influence of the features on capacity degradation and combine the battery aging mechanism, further clarifying the interpretability of the selected features. Finally, using the causal features as model input, a prediction model based on the long short-term memory network is constructed. The experimental results of the two datasets show that the minimum root mean square error is 0.0093 Ah and 0.9635 mAh with the mean relative errors of 0.25% and 0.13%, which verifies the validity and accuracy of the proposed method.

Degradation Modeling for Lithium-Ion Batteries with an Exponential Jump-Diffusion Model

The degradation of Lithium-ion batteries is usually measured by capacity loss. When batteries deteriorate with usage, the capacities would generally have a declining trend. However, occasionally, considerable capacity regeneration may occur during the degradation process. To better capture the coexistence of capacity loss and regeneration, this paper considers a jump-diffusion model with jumps subject to the exponential distribution. For estimation of model parameters, a jump detection test is first adopted to identify jump arrival times and separate observation data into two series, jump series and diffusion series; then, with the help of probabilistic programming, the Markov chain Monte Carlo sampling algorithm is used to estimate the parameters for the jump and diffusion parts of the degradation model, respectively. The distribution functions of failure time and residual useful life are also approximated by the Monte Carlo simulation approach. Simulation results show the feasibility and good performance of the combined estimation method. Finally, real data analysis indicates that the jump-diffusion process model with the combined estimation method could give a more accurate estimation when predicting the failure time of the battery.

Piecewise-linear modelling with automated feature selection for Li-ion battery end-of-life prognosis

The complex nature of lithium-ion battery degradation has led to many machine learning-based approaches for health forecasting being proposed in the literature. However, machine learning using sophisticated models can be computationally expensive, and although linear models are faster they can also be inflexible. Piecewise-linear models offer a compromise—a fast and flexible alternative that is not as computationally expensive as techniques such as neural networks or Gaussian process regression. Here, a piecewise-linear approach for battery health forecasting, including an automated feature selection step, is compared to a Gaussian process regression model and found to perform equally well in terms of the median error on a training dataset, and indeed somewhat better at the 95th percentile of error. The feature selection process demonstrates the benefit of limiting the correlation between inputs. Further trials found that the piecewise-linear approach was robust to changing input size and availability of training data.

Energy efficiency and CO2 emission comparison of alternative powertrain solutions for mining haul truck using integrated design and control optimization

The heavy-duty mining haul truck (MHT) with hundreds tonnage could emit hundreds of times emissions over a passenger vehicle. The heavy burden of emissions and continuous global warming force the mining industry seeking clean powertrain solutions for MHT. This research optimizes several emerging clean powertrain solutions for a typical 240 tonnage MHT using integrated design and control optimization method, and compares the energy efficiency and carbon dioxide (CO2) emission to the benchmark MHT with a diesel engine and diesel-electric drive. The well-established components model and optimal control theory support the integrated optimization under given operation profile. Currently, the battery electric solution could cut 63 percent usage cost incorporated with the lithium-ion battery degradation cost and reduce 21 percent well-to-wheel CO2 emission, making the battery electric solution a competitive candidate. Though the hydrogen fuel cell emits 26 percent less CO2 emission than the diesel counterpart, the high investment cost and degradation cost hinder its wide application. Further improvement of hydrogen production, capital cost and durability of fuel cell, and fast-charging of lithium-ion battery could release huge potential in cleaner MHT. The study provides an approach to quantitatively compare clean powertrain solutions and identify the optimal powertrain parameters for cleaner MHT.

Bayesian transfer learning with active querying for intelligent cross-machine fault prognosis under limited data

Most existing deep learning (DL)-based health prognostic methods assume that the training and testing datasets are from identical machines operating under similar conditions requiring massive labelled condition monitoring (CM) data to guarantee the prediction accuracy and generalisation capacity. However, these strict restrictions significantly hinder the deployment of the dl-based prognostic methods in real industries. In this paper, a Bayesian semi-supervised transfer learning with active querying-based intelligent fault prognostic framework is developed for remaining useful life (RUL) prediction across completely different machines under limited data. The proposed method strategically integrates the advantages of transfer learning (TL) and active learning in the Bayesian deep learning (BDL) framework. In the proposed framework, Bayesian neural networks with Monte Carlo dropout inference are utilised to quantify RUL prediction uncertainty, which is further leveraged to develop an active querying-based training data selection mechanism. Moreover, TL is simultaneously embedded into the BDL framework to relieve data distribution discrepancies existing among the completely different machines. The experimental verifications from open-sourced bearing datasets and lab testing-based lithium-ion battery degradation datasets demonstrate that the proposed framework can effectively and reliably achieve bi-directional transfer fault prognostic tasks under limited labelled CM data in target domain. Finally, generalisation and superiority of the proposed method are also validated by comparing with other state-of-the-art methods.

Operational intra-cycle temporal and current mode effects on battery capacity loss

As part of an on-going research effort towards developing a comprehensive overview of contributing operational factors to lithium ion battery degradation, new analysis and results are presented here which demonstrate the durational effects of battery charge and discharge operations on cell capacity loss. A set of cell cycling tests with a range of different discharge amplitudes provided data for this analysis which led to an advance on how to quantitatively account for the degradation effects in terms of their temporal duration. A factor accounting for operational duration effects usable as a parameter in battery life models was a key result of this project. An additional set of cell cycling experiments was conducted with the aim of decoupling degradation rates attributable to either the discharge or recharge portion of the complete cycle, and some differences were observed and quantified. • Uniquely patterned battery cycling experiments isolated cell degradation factors. • In particular, variable amplitudes and asymmetric cycle forms were tested. • Functionality obtained for a degradation driver based in operational duration. • Decoupling of discharge and recharge effects on degradation; recharge is costlier. • These cycle patterns relate to duty cycles found in the transportation sector.

リチウムイオン蓄電池の充放電エネルギー効率劣化の差電圧指標によるモデル化

リチウムイオン蓄電池の充放電エネルギー効率劣化の差電圧指標によるモデル化

Kalman-Filter-Based Learning of Characteristic Profiles of Lithium-Ion Batteries.

The main analyzed aspect of lithium-ion battery (LIB) degradation so far has been capacity fading. On the other hand, interest in efficiency degradation has also increased in recent years. Battery aggregation, which is expected to absorb the surplus of variable renewable energies such as photovoltaic energy, is affected by efficiency degradation in terms of the decreases in the economic gain and renewable energy use. Reusable LIBs could be used as aggregation components in the future; naturally, the variety of charge–discharge efficiencies might be more complex. To improve the operation efficiency of aggregation, including that obtained using reusable LIBs, we propose the Kalman-filter-based quasi-unsupervised learning of the characteristic profiles of LIBs. This method shows good accuracy in the estimation of charge–discharge energy. It should be emphasized that there are no reports of charge–discharge energy estimation using the Kalman filter. In addition, this study shows that the incorrect open-circuit voltage function for the state of charge, which is assumed in the case of a reused battery, could be applied as the reference for the Kalman filter for LIB state estimation. In summary, it is expected that this diagnosis method could contribute to the economic and renewable energy usage improvement of battery aggregation.

Optimal Charging Protocols to Restrict Lithium-Ion Battery Degradation

Lithium-ion batteries degrade due to several mechanisms that occur during cell operation. These mechanisms can be parasitic reactions that lead to loss of lithium inventory or loss of active material. Some examples of these side reactions include the growth of the solid-electrolyte interphase (SEI) layer, transition metal dissolution and deposition, lithium plating, solvent oxidation, etc. Thermal effects and mechanical fatigue on the electrodes also lead to cell aging and capacity fade.For widespread implementation of lithium-ion batteries for electric vehicles, batteries should last approximately 10 years.1 Fast charging has also been heavily researched for accelerating this prevalent circumstance. However, charging at high currents can severely degrade the cell, affecting the cell’s lifetime. Algorithms that optimize charging can be effective to reduce cell aging. Physics-based battery models can be used to design these algorithms by restricting phenomena that occur in the cell that causes cell aging. For example, lithium plating can be prevented by restricting the overpotential of the reaction.In our previous work, we have shown optimal charging profiles where the intercalation induced stresses, lithium plating and the SEI layer growth was restricted using reformulated physics-based models using model-based control 2 - 3 In this talk, we show different optimal charging protocols for restricting different mechanisms, and analyze the interplay between them to predict optimal charging profiles. References M. T. Lawder, P. W. C. Northrop, and V. R. Subramanian, J. Electrochem. Soc., 161, A2099–A2108 (2014).B. Suthar, P. W. C. Northrop, R. D. Braatz, and V. R. Subramaniand, J. Electrochem. Soc., 161, F3144–F3155 (2014).M. Pathak, D. Sonawane, S. Santhanagopalan, R. D. Braatz, and V. R. Subramanian, ECS Trans., 75, 51–75 (2017).

Lithium-Ion Battery Degradation: How to Diagnose It

Many different degradation mechanisms occur in lithium-ion batteries, all of which interact with one another [1]. However, there are few fewer observable consequences of degradation than there are mechanisms [2]. It is possible to measure the different degradation modes: loss of lithium inventory (LLI), loss of active material (LAM), impedance change and stoichiometric drift [3].It is not always possible to link these observable consequences of degradation to any particular mechanism or combination of mechanisms. Many models of degradation exist [4], but these models have many parameters that cannot be measured directly. A recent modelling study [5] found the number of parameters that the model is sensitive to is greater than the number of observable degradation modes.However, the same model [5], despite including just four degradation mechanisms, found five possible degradation pathways a battery can follow. The model was built so that more mechanisms can easily be added later, so more pathways will be found.In this work, a new approach to diagnosing battery degradation is proposed, based on these pathways. Experimental data for the degradation modes can be identified as being consistent with a particular pathway. Once the correct pathway is found, the parameters that particular pathway is sensitive to can be fit to the data, feeding back into the model.[1] Jacqueline Edge et al., Phys. Chem.: Chem. Phys. vol. 23, pp. 8200-8221, 2021.[2] Christoph Birkl et al., Journal of Power Sources vol. 341, pp. 373-386, 2017.[3] Matthieu Dubarry et al., J. Electrochem. En. Conv. Stor. vol. 17, pp. 044701, 2020.[4] Jorn Reniers et al., J. Electrochem. Soc. vol. 166 pp. A3189-A3200, 2019.[5] Simon O’Kane et al., Phys. Chem.: Chem. Phys., submitted, 2022. https://arxiv.org/abs/2112.02037

Changes in Heat Generation during Degradation of Commercial Lithium-Ion Batteries

The performance, lifetime and safety issues of lithium ion batteries (LIB) are strongly linked to temperature and many studies in the past three decades have been devoted to improving the understanding of thermal behaviour of LIB at cell and pack level1. We present ageing and diagnostic data of a commercial pouch LIB and present a thermal battery model. The thermal battery model is used to discuss changes in different heat generation sources alongside battery ageing measurements. It further allows for a differentiation between temperature induced and C-rate accelerated ageing.A selection of LCO-graphite cells was aged and cycled in the full state-of-charge window at 5, 15, 25, 35 and 45°C. We compare constant current cycling with charge and discharge rates of 1 C, 2 C and 3 C, and constant power discharging. The cells were monitored in detail during the ageing including surface. The performance at lower C-rates, internal resistances and the changes in cell thickness were obtained as functions of changes in state-of-health. Thermal conductivity measurements were performed for the electrode material at end and beginning of life.A 1D analytical thermal model is used for a discussion about the impact of the heat sources, the effect of a change in heat transport with ageing due to decreased thermal conductivities, and the impact of internal temperature gradients on ageing. By utilizing the thermal battery model, it is possible to analyze the effect of the different heat sources on the overall battery temperature as well as degradation dependent changes in heat generation. The heat sources included in the model are the ohmic heat produced, heat production due to overpotential, and reversible entropic heat due to the electrode reactions, where the entropic heat can both act as a heat source and heat sink. The transfer of heat inside the battery is described using Fourier’s second law. These heat and temperature effects allow further to differentiate between temperature induced and C-rate accelerated ageing. Together with evaluating the impact of temperature gradients over the cell compared to the maximum cell temperature, this allows to estimate the impact of an improved thermal management system. References 1 Spitthoff, L.; Burheim, O.S.; Shearing, P.R. Temperature, Ageing and Thermal Management of Lithium-Ion Batteries. Energies 2021, 14, 1248. https://doi.org/10.3390/en14051248

Identification and quantification of ageing mechanisms in Li-ion batteries by Electrochemical impedance spectroscopy.

Transportation sector reported almost a quarter of global CO2 emissions. Thus, efforts to decarbonize this sector are essential for achieving net zero emission goals. Among the actions to mitigate the effects of climate change in favor to decarbonization, lithium-ion electric vehicle market has expanded over the past years because of both scientific advances and encouraging public policies. The analysis of ageing in lithium-ion batteries is essential to ensure optimal performance and determine the end of the useful life for that purpose. Lithium-ion batteries degradation is a complex multi-causal process. Ageing mechanisms could be grouped mainly into three degradation modes: Loss of Conductivity (CL), Loss of Active Material (LAM) and Loss of Lithium Inventory (LLI). Ageing battery process can be evaluated as a state of health (SoH) and tracked based on capacity and power. Although SoH quantifies the battery degradation through the decrease in capacity, its definition does not include an indication of the underlying deterioration mechanisms causing the degradation. Combined with electrochemical impedance spectroscopy (EIS), degradation modes can be identified and quantified non-destructively with the aim of finding a correlation between their evolutions with SoH. This paper proposes a method to identify and measure the ageing mechanisms in commercial 18650 NMC lithium-ion batteries over time using the EIS technique. The EIS spectra were fitted to a proper equivalent electrical circuit and the main mechanisms responsible for the degradation identified through the calculated parameters variations with time. Thus, the increase in Rohm was assigned to CL, Rsei and Rct with LLI and Rw with LAM. Correlation between degradation modes with SoH was also reported.

Probabilistic Modeling of Li-Ion Battery Remaining Useful Life

The estimation of lithium-ion batteries degradation plays an important role for the correct operation of smart grid and electric vehicle applications. In fact, only healthy battery energy storage systems meet minimum performance in terms of supplied voltage and power. Health prognostic is mandatory to ensure safe and reliable operation of batteries, as unsuccessful operation may cause technical/economic detriments or complete failures. The battery remaining useful life (RUL) depends on several random factors and thus it should be probabilistically characterized to allow the application of decision-making processes. In this article, a hybrid methodology is developed to characterize the RUL even with small amount of data, in view of the typical scarcity of data for adequately fitting the RUL probability distributions. Several distributions are considered in the methodology and compared for such purpose: the inverse Burr (IB) distribution, and mixtures of IB with inverse Gaussian and with inverse Weibull distributions. Expectation-maximization algorithms are specifically developed for the parameter estimation of the considered mixture distributions in the proposed methodology. The IB-based distributions are applied on a RUL dataset created by Monte Carlo sampling on an electrochemical battery model fitted upon given charge/discharge cycles. Numerical experiments are reported to assess the proposal with respect to benchmark distributions.

Effects of Zn, Mg and Cu Doping on Oxidation Reaction of Al (111) Surface

Aiming at the performance degradation of lithium-ion batteries due to shell corrosion, the doping of alloy elements Zn, Mg and Cu on Al (111) surface and the effect on oxidation reaction of Al (111) surface were studied by the first-principles calculation method. The results show that Zn, Mg and Cu atoms stably combine with Al atoms, and the surface smoothness is slightly different due to their different radii and electronegativity. The dissociative adsorption of O2 molecules is related to the surface doping atoms and O2 coverage, while the electron tunneling of underlying metal promotes O2 adsorption on the surface. As O2 coverage increases, the O atoms adsorbed on the hcp site gradually migrate to the subsurface layer. Zn, Mg, Cu and vacancy defect hinder the migration of the surrounding O atoms to subsurface layer, resulting in different structures and thicknesses of the oxide film near the doped atoms. At the same time, Zn, Mg, and Cu atoms differ in their ability to gain or lose electrons compared with Al atoms, resulting in their different positions on the surface. In addition, the surface work function rises with the increase of O2 coverage, and Zn and Cu atoms make the work function increase faster. Finally, according to the research results, it can be inferred that Zn and Mg are the unfavorable factors for the oxidation reaction of Al surface.

A hybrid data-driven method for rapid prediction of lithium-ion battery capacity

The quick and accurate prediction of future capacity is a challenging issue in the field of battery health management. To solve this problem, the paper proposes a data-driven model based on advanced machine-learning techniques. First, wavelet transform is used to denoise the initial features, and the relatively important features are selected with the help of random forest. Second, the paper uses the approximate entropy theory to construct multi-time scale sliding window data to solve the problem that the sliding window size can not be determined in traditional methods. Third, in order to overcome the disadvantages of traditional methods, such as the comparative inefficiency in converging and using temporal contextual information, the paper proposes a novel mogrifier long short-term memory network denoted as QGA-ASM-LSTM, which is based on attention mechanism and similarity judgment mechanism, using quantum genetic algorithm to optimize its parameters. Finally, the proposed data-driven method is used to model the degradation of lithium-ion batteries and finish the fast capacity prediction. The results of experiments on different public datasets demonstrate that the performance of QGA-ASM-LSTM exceeds the state-of-the-art models in terms of prediction accuracy and training time.

Li-ion battery individual electrode state of charge and degradation monitoring using battery casing through auto curve matching for standard CCCV charging profile

The internal status, such as the state of charge and degradation of lithium-ion batteries, should be carefully monitored to ensure battery performance and safety. However, this task is challenging due to the weak observability. Estimating the state and degradation of each individual electrode (cathode and anode) is advantageous for battery management systems yet even more challenging. To address these challenges, this paper presents a novel method of using battery casing as a reference electrode to calculate the state of lithiation of both electrodes and the degradation status (including the storage capacity fade of each electrode and the total cyclable lithium in the battery). First, we introduced an innovative approach of using battery casing as the reference electrode and collected experimental data. The reference electrode provides valuable information on the anode and cathode potential, which can be used to extract internal information, including the state of charge of each electrode and its degradation status. Then, we calculated the transferred lithium ions between the electrodes by matching the anode and cathode voltage curves to the standard potential curves of graphite anode and lithium nickel manganese cobalt oxide (NMC) cathode, respectively. Next, we calculated the capacity of the electrodes using the state of lithiation of the electrodes and the total charge transferred. Last, we analyzed the battery degradation status by comparing the capacity fade of individual electrodes to the total charge transferred to the cell. The proposed method can easily calculate each electrode’s capacity, state of lithiation, and remaining cyclable lithium ions in the battery using the curve matching algorithm, which can be used to monitor battery degradation in real-time.

The economics of firm solar power from Li-ion and vanadium flow batteries in California

The cost of providing near 24-7-365 power from solar panels at a commercial facility in South California was modelled to be similar for vanadium flow batteries (VFB) and lithium ion batteries (LIB) at around $0:20/kWh. In hotter locations, LIB economics suffer due to accelerated background cell ageing. Even within South California there was enough variation to affect the economic comparison. Although LIB degradation could be reduced in a hybrid VFB-LIB system, there was negligible benefit to the overall electricity cost.As a result of falling photovoltaic panel costs in the last decade solar power (PV) is now claimed to be the cheapest source of electricity. However, the intermittent nature of supply means that it cannot solve the energy trilemma alone, and a form of backup power is required for reliability. This application is well suited to batteries, but the cost implications of providing high levels of reliability in this way have not been widely studied. In this work, the levelised cost of electricity (LCOE) achievable by optimal combinations of PV and batteries is determined for a large food retailer at a range of self-sufficiency ratios (SSR). Both lithium ion batteries (LIB), vanadium redox flow batteries (VFB) and hybrid systems of the two technologies are modelled. In combination with an over-sized PV array, both systems are capable of providing a SSR of 0.95 for a LCOE of less than $0.22/kWh. The optimal LCOE values overlap across the SSR range for both technologies depending on cost and ambient temperature assumptions. A VFB is more likely to give the lower LCOE at lower SSR, and a LIB is favoured at high SSR as the cycle rate drops as SSR increases. It is also shown that a state of charge (SOC) minimisation strategy has a significant impact on the LIB economics by reducing calendar ageing. Lastly, hybrid systems combining LIB and VFB were modelled, but in no cases showed an improvement over the optimal single choice. The overlap in the LCOE of the two battery types highlights the importance of other considerations, such as sustainability, space requirements and safety.Graphical abstract

Modelling Solvent Consumption from SEI Layer Growth in Lithium-Ion Batteries

Predicting lithium-ion battery (LIB) lifetime is one of the most important challenges holding back the electrification of vehicles, aviation, and the grid. The continuous growth of the solid-electrolyte interface (SEI) is widely accepted as the dominant degradation mechanism for LIBs. SEI growth consumes cyclable lithium and leads to capacity fade and power fade via several pathways. However, SEI growth also consumes electrolyte solvent and may lead to electrolyte dry-out, which has only been modelled in a few papers. These papers showed that the electrolyte dry-out induced a positive feedback loop between loss of active material (LAM) and SEI growth due to the increased interfacial current density, which resulted in capacity drop. This work, however, shows a negative feedback loop between LAM and SEI growth due to the reduced solvent concentration (in our case, EC), which slows down SEI growth. We also show that adding extra electrolyte into LIBs at the beginning of life can greatly improve their service life. This study provides new insights into the degradation of LIBs and a tool for cell developers to design longer lasting batteries.

Fast Modeling of the Capacity Degradation of Lithium-Ion Batteries via a Conditional Temporal Convolutional Encoder–Decoder

Recently, deep neural networks have been successfully applied to battery degradation modeling, and onboard graphic processing units can potentially be used to host these models. However, most existing degradation models rely on recursive operations, which are slow on hardware optimized for vectorized computations. This article proposes a two-encoder, one-decoder conditional temporal convolutional network that uses no recursive operations, thus allowing for the conditional modeling of battery degradation and improving upon the prediction speed of long short-term memory encoder–decoder by a factor of up to 70 for long tests. Datasets containing 115 batteries from multiple sources and 195 000 steps are used to verify the proposed model. The data are augmented by randomly selecting the prediction starting point during training. The proposed model achieves an average prediction error of less than 4% of the maximum available capacity based on large datasets. In addition, prediction results and convolution weight analysis indicate that the model can learn the temporal and conditional dynamics of typical lithium-ion batteries and is robust to the selection of the prediction starting point. We further demonstrate a prediction under hypothetical future conditions and a Monte Carlo prediction under unknown future conditions, and the degradation trends are accurately predicted.