Investigation of path dependence in commercial lithium-ion cells chosen for plug-in hybrid vehicle duty cycle protocols

doi:10.1016/j.jpowsour.2010.05.058

Journal of Power Sources

Volume 196, Issue 7, 1 April 2011, Pages 3395-3403

https://doi.org/10.1016/j.jpowsour.2010.05.058 Get rights and content

Abstract

There is a growing need to explore path dependence of aging processes in batteries developed for long-term usage, such as lithium-ion cells used in hybrid electric vehicle (HEV) or plug-in hybrid vehicle (PHEV) applications that may then be “retired” to be utilized in grid applications. To better understand the foremost influences on path dependence in the PHEV context, this work aims to bridge the gap between ideal laboratory test conditions and PHEV field conditions by isolating the predominant aging factors in PHEV service, which would include, for example, the nature and frequency of duty cycles, as well as the frequency and severity of thermal cycles. These factors are studied in controlled and repeatable laboratory conditions to facilitate mechanistic evaluation of aging processes. This work is a collaboration between Idaho National Laboratory (INL) and the Hawaii Natural Energy Institute (HNEI). Commercial lithium-ion cells of the Sanyo Y type (18650 configuration) are used in this work covering two initial independent studies of path dependence issues. The first study considers how the magnitude of power pulses and charging rates affect the aging rate, while the second seeks to answer whether thermal cycling has an accelerating effect on cell aging. While this work is in early stages of testing, initial data trends show that cell aging is indeed accelerated under conditions of high discharge pulse power, higher charge rates, and thermal cycling. Such information is useful in developing accurate predictive models for estimating end-of-life conditions.

Introduction

Path dependence is emerging as a premier issue of how electrochemical cells age in conditions that are diverse and variable in the time domain. For example, a laptop battery might be used sporadically, or lithium-ion cells in a vehicle configuration will experience a variable combination of usage and rest periods over a range of temperature and state of charge (SOC). This is complicated by the fact that some aging can actually become worse (or better) when a lithium-ion cell is idle for extended periods under calendar-life (calL) aging, as opposed to cycle-life (cycL) conditions where the cell is used within a predictable schedule. The purpose of this study is to bridge the gap between highly idealized and controlled laboratory test conditions and actual field conditions regarding PHEV applications, so that field-type aging mechanisms can be mimicked and quantified in a repeatable laboratory setting. The main parameters here involve the magnitude of power pulses, frequency of duty cycles, as well as the frequency and severity of thermal cycles, looking at isothermal, mild, and severe scenarios.

Although there is little published literature that covers comprehensive aging and path dependence work, there are articles that address key issues, either singly or in some combination [1], [2], [3], [4], [5], [6], [7], [8], [9], [10] (as a short list). This work advances this body of literature by having a comprehensive and consistent basis for testing and analysis, given the context of a PHEV application.

To date, little is known about lithium-ion aging effects caused by thermal cycling superimposed onto electrochemical cycling, and related path dependence. This scenario is representative of what lithium-ion batteries will experience in vehicle service, where upon the typical start of a HEV/PHEV, the batteries will be cool or cold, will gradually warm up to normal temperature and operate there for a time, then will cool down after the vehicle is turned off. Such thermal cycling will occur thousands of times during the projected life of a HEV/PHEV battery pack. We propose to quantify the effects of thermal cycling on lithium-ion batteries using a representative chemistry that is commercially available (Sanyo Y).

Electrochemical cycling is based on PHEV-relevant cycle-life protocols that are a combination of charge-depleting (CD) and charge-sustaining (CS) modes discussed in the Battery Test Manual for Plug-in Hybrid Electric Vehicles [11]. A realistic duty cycle will involve both CD and CS modes, the proportion of each defined by the severity of the power demands. There are innumerable ways to define and configure such PHEV-type cycling profiles. However, we can at least provide some boundaries to narrow our field of operating conditions. We assume that the cells will start each cycling day at 90% SOC, and that they will not be allowed to go below 35% SOC, with operation around 70% SOC being a nominal condition. The 35, 70, and 90% SOC conditions are also being used to define critical aspects of the related reference performance test (RPT) for this investigation. There are three primary components to the RPT, all assessed at room temperature: (A) static and residual capacity (SRC) over a matrix of current, (B) kinetics and pulse performance testing (PPT) over current for SOCs of interest, and (C) EIS for SOCs of interest. The RPT is performed on all cells every 28-day test interval, as well as a pulse-per-day to provide a quick diagnostic snapshot. Where feasible, we utilize various elements of Diagnostic Testing to characterize performance of the cells and to gain mechanistic-level knowledge regarding both performance features and limitations.

Herein we present the rationale behind the experimental design, early data, and discuss the fundamental modeling tools used to elucidate performance degradation mechanisms. While we recognize that the datasets are underdeveloped due to recent commencement of these tests, we are able to see early-life trends of key performance metrics, which have bearing on upfront issues such as ohmic resistance and the extent of completion of cell formation. The benefit of this work, upon completion, will be the testing approach used to explore aging path dependence, and to provide more realistic and accurate life predictions by accounting for the influence of thermal cycling effects and related path dependence on aging mechanisms.

Section snippets

Test cells

The secondary lithium-ion cells used in this study are Sanyo Y cells of the 18650 configuration that have a nominal capacity rating of 1.9 Ah, and consist of a {LiMn₂O₄ + LiMn_1/3Ni_1/3Co_1/3O₂} cathode of proprietary mixture and a graphite anode. A summary of salient properties is given in Table 1, and a photograph of a portion of these cells in their test fixtures is given in Fig. 1. The capacity and current specifications in Table 1 are provided by the manufacturer, and are presumably at room

Concept of aging path dependence

An understanding of “aging path dependence” is needed to design experiments related thereto and to interpret test data. Path dependence is an important factor for any batteries that will undergo prolonged usage in one or more applications, especially if usage patterns change appreciably during service life. In the context of lithium-ion cells in arbitrary service (vehicle applications, laptops, and consumer electronics), the extent of aging experienced by a cell at a given point in its life

Results and discussion

Work is progressing for PD Studies 1 and 2, which will illuminate fundamental performance limitations, aging mechanisms, and path dependence thereof for cells under these aging conditions. Data analysis is underway for the huge amount of information gathered so far during testing for PD Studies 1 and 2. In particular, pulse-per-day impedances (charge and discharge) and monthly RPT data is being evaluated. Given below is a concise discussion of early data subsets. A more complete discussion and

Conclusions

Through the collaboration of INL and HNEI, a set of path-dependent studies have been initiated that focus on fundamental factors regarding the use of lithium-ion cells in PHEV-type applications. Early performance data over test time indicates that our test cells experience a slightly higher rate of aging at higher magnitude power pulses, higher temperature, and more severe thermal cycling. These results, although somewhat intuitive, allow us to quantify aging effects in the laboratory that have

Acknowledgements

The authors gratefully acknowledge funding provided by the Office of Energy Efficiency and Renewable Energy within the United States Department of Energy, per Contract No. DE-AC07-05ID14517.

References (15)

B.Y. Liaw et al.
J. Power Sources
(2003)
T. Osaka et al.
J. Power Sources
(2003)
K. Asakura et al.
J. Power Sources
(2003)
B.Y. Liaw et al.
J. Power Sources
(2005)
B.Y. Liaw et al.
J. Power Sources
(2007)
M. Dubarry et al.
J. Power Sources
(2007)
M. Dubarry et al.
J. Power Sources
(2007)

There are more references available in the full text version of this article.

Cited by (87)

Temperature-driven path dependence in Li-ion battery cyclic aging
2024, Journal of Power Sources
Operation temperature is known to be one of the most impacting parameters for aging of Li-ion batteries under cycling conditions. While Li plating usually dominates aging at low temperatures, Solid-Electrolyte-Interphase (SEI) growth often dominates aging at elevated temperatures. The longest cycle life for Li-ion batteries is observed at the crossover temperature of both aging mechanisms at around 25 °C. We introduce aging color maps to show changes in the aging rate during battery aging. Reconstructed 3-electrode full cells with Li reference electrode show that the prolonged CV phase during charging at low temperatures counteracts Li plating by an increasing anode potential. Thirty-six temperature variations during cycling between 0 °C and 45 °C were studied to investigate the temperature path dependence of battery aging. The permutability of aging temperatures is analyzed by means of commutators. It turns out that aging at elevated temperatures is only slightly affected by previous aging processes and aging at lower temperatures with Li plating accelerates itself. Previous high temperature cycling decelerates Li plating and shifts its onset to lower temperatures, leading to a new and improved optimum cycling temperature of 20 °C. Differential voltage analysis and Post-Mortem analysis were performed to identify the underlying aging mechanisms.
ICFormer: A Deep Learning model for informed lithium-ion battery diagnosis and early knee detection
2024, Journal of Power Sources
Accurate diagnosis of lithium-ion battery (LIB) degradation is critical for safe and reliable operation in real-world applications. In recent years, data-driven approaches powered by Machine Learning algorithms emerged as a promising solution, among which Deep Learning methods were proven to be effective for various tasks such as State of Charge and State of Health estimation and Remaining Useful Life prediction, however, their application for knee point estimation is still at an early stage. While experimental techniques, were demonstrated to be successful for detecting knees, current methods focus on tracking individual cycles rather than analyzing the overall trajectory of degradation, which could potentially offer better performance. In this paper, we introduce ICFormer, a novel Deep Learning model based on a Transformer encoder that leverages self-attention on the evolution of incremental capacity curves to accurately identify relevant changes in LIB degradation trajectories. The proposed model not only detects knees, but also anticipates them while also outperforming state-of-the-art approaches in diagnosing degradation modes. The effectiveness of the model is validated using both synthetic and experimental data. We demonstrate that the method can provide valuable knowledge on the factors contributing to capacity loss and offer advanced insights for battery management and predictive maintenance strategies.
Two-stage deep learning for online prediction of knee-point in Li-ion battery capacity degradation
2022, Applied Energy
Accurate monitoring of capacity degradation of a lithium-ion battery is important as it enables the user to manage the battery usage for optimal performance/lifetime and to take preemptive measures against any potential explosion or fire. Battery capacity fades gradually through repetitive charging and discharging until it reaches the so called ‘knee-point’, after which it goes through rapid and irreversible deterioration to reach its end-of-life. It is crucial to forecast the knee-point early and accurately for safety and economic use of the battery. Machine learning based methods have been used to predict the knee-point with early cycles cell data. Despite some notable progress made, the existing methods make the unrealistic assumption of constant cycle-to-cycle charge/discharge operation. In this study, a novel two-stage deep learning method is proposed for online knee-point prediction under variable battery usage. A CNN-based model extracts temporal features across past and current cycles to sort out those that should be monitored closely for near-term failures, and then predict the number of cycles left to reach the knee-point for them. The proposed method extracts features from time-series data and thus reflects dynamic changes in battery properties, resulting in improved prediction performance under realistic scenarios.
A novel hybrid-point-line energy management strategy based on multi-objective optimization for range-extended electric vehicle
2022, Energy
To achieve the optimal energy allocation for the auxiliary power unit (APU) and battery of a range-extended electric vehicle, a novel hybrid-point-line energy management strategy (H–P-LEMS) has been proposed from a multi-scale view. First, a multi-objective optimization (MOO) model is established which takes into account energy consumption, emissions and battery life. The barebones multi-objective particle swarm optimization is applied for solving the MOO problem. And a dynamic programming-optimized algorithm is applied to obtain the optimal curve/area of APU to establish objective function of MOO. Then, an adaptive approach uses a fuzzy logic controller with the battery consideration to adjust parameters in real time. Simulation results show that there is a clear conflict that three optimization objectives cannot be optimal at the same time and the final optimization solution with optimal comprehensive evaluation index is selected to evaluate the performance of the proposed methodology. Finally, the simulation and experimental results thoroughly indicate that the proposed H–P-LEMS has better balance than conventional rule-based energy management strategy (EMS). As expected, economy improvement, emission reduction and prolonging the battery service life are kept in balance effectively. And this result can be used to develop EMS to improve comprehensive performance levels.
Rapid failure mode classification and quantification in batteries: A deep learning modeling framework
2022, Energy Storage Materials
Unique, rapid identification and quantification of the dominant aging modes in lithium-ion batteries (LiBs) with early and non-specialized test data is a significant scientific challenge. Leveraging synthetic-data, deep-learning (DL) techniques have great potential to enable fast and robust classification and quantification of battery aging modes that produce different patterns of cell aging. This study presents, for the first time, a synthetic–data-based DL modeling framework for rapid and automatic classification and quantification of battery-aging modes and resultant aging, with experimental validation of the technique. Availing synthetic dQ.dV⁻¹ curves for roughly 26,000 initial conditions and aging modes, the framework classified the dominant aging modes for cells undergoing fast charge in less than 100 cycles. Upon classification, the framework quantified the evolution of the aging modes, which were often not uniform with cycling, for 22 Gr/NMC532 pouch cells tested up to 600 cycles at different charging rates (1C–9C).
State-of-health estimation using a neural network trained on vehicle data
2021, Journal of Power Sources
The validation of battery aging models in automotive applications requires reliable aging data to compare the accuracy of each proposed model. Using a sample of 704 vehicles aged up to eight years under diverse nominal conditions two aging estimation models are proposed. By analyzing relevant automobile battery data a more relevant fit of a semi-empirical holistic model is provided with an Arrhenius temperature dependence and pseudo-Tafel voltage dependence. As a comparison, a neural network capturing the aging behavior using the most correlated variables available in the data-set was also developed. Over 110,000 measurements from seven relevant indicators are available as aging predictors, as well as, highly-accurate capacity measurements which is used as the ground truth capacity targets to train and validate the proposed models. Against these points the Semi-Empirical and Neural Network models achieved a root mean squared error of 3.4%-SOH and 3.0%-SOH, respectively.

View all citing articles on Scopus

View full text

Investigation of path dependence in commercial lithium-ion cells chosen for plug-in hybrid vehicle duty cycle protocols

Abstract

Introduction

Section snippets

Test cells

Concept of aging path dependence

Results and discussion

Conclusions

Acknowledgements

J. Power Sources

J. Power Sources

J. Power Sources

J. Power Sources

J. Power Sources

J. Power Sources

J. Power Sources