Incremental Capacity Curve Peak Points-Based Regression Analysis for the State-of-Health Prediction of a Retired LiNiCoAlO2 Series/Parallel Configured Battery Pack
Round 1
Reviewer 1 Report
The paper proposes a methodology to divine the state-of-health (SOH) of every cell in retired battery packs regardless of how they are used. The methodology is based on the analysis of the change of a single feature on the incremental capacity (IC) curve. I find a number of flaws with this proposed methodology.
First, IC curves from the retired cells are compared to data from cells aged using CC-CV cycles. However, cells age differently depending on the cycle conditions. Moreover, degradation due to cycle aging can be drastically dissimilar to degradation from calendar aging, which is not accounted for in this analysis. The difference in degradation would manifest themselves in the IC curves making the data from the CC-CV cycled cell ill-suited for application as a standard for comparison. Read articles listed below:
Dubarry, G. Baure, A. Devie, 2018, "Durability and Reliability of EV Batteries under Electric Utility Grid Operations: Path Dependence of Battery Degradation", J. Electrochem. Soc., 165, (5) A773-A783, 10.1149/2.0421805jes.
Keil, P.; Jossen, A., Calendar Aging of NCA Lithium-Ion Batteries Investigated by Differential Voltage Analysis and Coulomb Tracking. Journal of the Electrochemical Society 2017, 164 (1), A6066-A6074.
Keil, A. Jossen, 2016, "Charging protocols for lithium-ion batteries and their impact on cycle life—An experimental study with different 18650 high-power cells", Journal of Energy Storage, 6 125-141, 10.1016/j.est.2016.02.005.
Second, the methodology is confined to the initial quasi-linear capacity loss regime and is incapable of predicting any acceleration in degradation. This second stage of aging is noted in several articles including the ones below. Any SOH determination model is trivial if it is only valid in the initial aging stage. In this stage of degradation, the accuracy calculated is unsurprising and insufficient for model verification. The model would only be novel if it were able to predict changes beyond the initial trend.
X.-G. Yang, Y. Leng, G. Zhang, S. Ge, C.-Y. Wang, 2017, "Modeling of lithium plating induced aging of lithium-ion batteries: Transition from linear to nonlinear aging", J. Power Sources, 360 28-40, 10.1016/j.jpowsour.2017.05.110.
F. Schuster, T. Bach, E. Fleder, J. Müller, M. Brand, G. Sextl, A. Jossen, 2015, "Nonlinear aging characteristics of lithium-ion cells under different operational conditions", Journal of Energy Storage, 1 44-53, 10.1016/j.est.2015.05.003.
Third, IC changes do not necessarily correlate with capacity loss. The publications below were able to correlate the changes to portions of the IC curve to individual or combined degradation mechanisms. Some of which did not affect capacity loss initially. No persuasive evidence is presented that the particular change in the IC curve dubbed “Peak point 2“ is an indicator of the SOH of the cell. References 18 and 25 are the same paper and in it a different battery system is analyzed. Direct correlations between the observations from that paper and this manuscript cannot be drawn.
Gao, S. Yang, J. Jiang, C. Zhang, W. Zhang, X. Zhou, 2019, "The Mechanism and Characterization of Accelerated Capacity Deterioration for Lithium-Ion Battery with Li(NiMnCo) O2 Cathode", J. Electrochem. Soc., 166, (8) A1623-A1635, 10.1149/2.1001908jes.
Baure, M. Dubarry, 2019, "Synthetic vs. Real Driving Cycles: A Comparison of Electric Vehicle Battery Degradation", Batteries, 5, (2), 10.3390/batteries5020042.
For these reasons, I reject this manuscript for publication. The methodology is flawed and nothing of consequence can be deemed from the analysis. A more rigorous examination of the incremental curves is needed to correctly estimate the remaining useful life of retired cells.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments:
The authors present the estimation method of state-of-health for a retired series/parallel battery pack by using a regression analysis model. The error range of their method for the capacity fading model was as small as nearly 1%. I think their study is interesting, but more reasonable experimental results should be provided to support their claims. Overall, the manuscript could be published to Electronics after major revision. Comments are listed below:
1) It is well known that C-rate affects the degradation rate of lithium-ion batteries. The authors tested at 1 C-rate only. To verify their regression analysis method, other C-rate conditions should be studied.
2) The main mechanism of charge storage in lithium-ion batteries is ion intercalation (i.e., redox reaction). They should use redox reaction related peaks (Peak1, Peak3, and Peak5) to correctly verify their method rather than peak2 and peak4 which are related to a minor contribution of charge storage mechanism, capacitive storage.
3) They should correct the notation used in the manuscript. For example, they claimed five peaks in Figure 10 but that’s not correct. In general, peak2 and peak4 are not peaks but valleys. Only Peak1, Peak3, and Peak5 are the peaks related to redox reactions.
Other minor comments:
1) The style of figures is inconsistent. For example, the shape of Figure 8 is little collapsed whereas Figure 9 is not.
2) In line 203, tau (τ) should be italicized.
3) In line 262, r should be italicized. Please make sure all symbols italicized.
4) In line 304, R2 should be corrected to R2.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Still assumes the degradation of the cell remains the same no matter the usage and SoH value would only apply to that degradation.
The grammar used in the paper makes it very difficult to read.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
The authors have clearly addressed all comments. Thank you.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 3
Reviewer 1 Report
The paper is acceptable and further warrants investigation of ICA methods for use as predictors of a cell's RUL.
The grammar is improved, though it seems as if each sentence was checked separate from the paragraph as a whole. At times, the paper lacks the flow needed to illustrate the author's point clearly and succinctly. There is a difference between being grammatically correct and being readable. One hopes an expansion of the writer's vocabulary will help in this regard.
