Where to start with the task of developing a lifecycle test procedure. Battery cell ageing is such a complex subject that it makes it difficult to know where to start. One of the first things to do is to break the problem into Calendar and Cycle Ageing.
The standard method used by industry has two fundamental test regimes as inputs:
- calendar ageing – where the cells are held at different fixed temperatures and SoC.
- cycle ageing – where the cells are cycled through a complete charge discharge at different fixed temperatures.
At regular intervals the capacity of the cells are measured. This data is then used as inputs to the ageing predictions, a simple summation of these two parts gives the overall ageing model [1].
Now we know we are going to be doing a calendar ageing study and a cycle ageing study.
Calendar Ageing
Here the cells are held at a fixed temperature and state of charge (SoC). Then every month the cells are removed from the fixed temperature storage and moved to a standard 25°C facility where the capacity and internal resistance are tested.
It should be noted that calendar ageing increases with temperature and SoC, hence 90% SoC and 50°C is going to age faster than a cell held at 30% SoC and 15°C. This means that you are going to bias the testing towards the higher temperatures and SoC values.
Example test table:
| SOC / Temperature | 25°C | 40°C | 55°C |
| 95% | for each of these you will record capacity vs time | ||
| 65% | |||
| 40% | |||
| 10% |
This post has been built based on the support and sponsorship from: Thermo Fisher Scientific, Eatron Technologies, About:Energy and Quarto Technical Services.
Kailong Liu et al [2] state in their conclusions when looking at calendar ageing modelling: “It should be noted that modelling the battery capacity degradation during storage is a challenging task due to the strong coupling of different ageing mechanisms and their nonlinear dependence on various stress factors such as storage temperatures and SOCs.”
Cycle Ageing
As per the title we are going to cycle the cells and do that for different temperatures, Depth of Discharge and at different discharge rates. In order to reduce the test matrix size you might want to cover different charge rates as a separate item.
After a number of cycles the test will be halted, the cell will be returned to the standard tempoerature and the capacity and internal resistance will be measured. The cell will then be returned tp the cycling regime.
| DoD / Temperature | 10°C | 25°C | 40°C | 50°C |
| 100% | for each of these you will record capacity vs cycle number | |||
| 90% | ||||
| 80% | ||||
| 60% |
| DoD / Temperature | 10°C | 25°C | 40°C | 50°C |
| 100% | ||||
| 90% | ||||
| 80% | ||||
| 60% |
| DoD / Temperature | 10°C | 25°C | 40°C | 50°C |
| 100% | ||||
| 90% | ||||
| 80% | ||||
| 60% |
| DoD / Temperature | 10°C | 25°C | 40°C | 50°C |
| 100% | ||||
| 90% | ||||
| 80% | ||||
| 60% |
With the cycle ageing you can see that you very rapidly have a lot of tests to run. It is very likely that you will want to run the DoD from different start and finidh points, eg: 90% DoD would run between 100% to 10% and 90% to 0%, thus allowing you to determine the impact of operating at the limits of the cell.
1C or 2C or Customer Cycles
The cycling can be standard cycle rates and that makes the data anaysis easier, however, you can cycle the cell over something closer to the usage profile. The difficulty with this is that you will probably want to analyse, simplify and carefully position these cycles with respect to DoD, C-rate and rest periods.
Data Points
However, do think about data quality and hence what data you will require to form the equations and surfaces. Once you have an ageing model you are likely to get asked some difficult questions, such as:
- “if we change our temperature set point for cooling from 30°C to 34°C it increases the vehicle range, but what is the impact on the battery pack ageing?”.
- “If we increase upper SoC from 95% to 96% we gain x miles, but can we still meet the 10 year warranty?”
Do you have the fidelity of data to estimate the impact?
One idea is to use dummy data to initially form the inputs to the model and see if they could work. You can then see the impact of increasing the number of tests on both your data and quality of the model. The number of data channels and controlled temperature environments is probably limited and hence you will want to fully understand these before committing to 2 years of testing.
Hopefully this has given you an overview of how to develop a lifecycle test procedure in it’s simplest form. There are many ways in which other parameters can be added to this, such as cell pressure. However, think carefully before adding more variables as the complexity of the testing will increase significantly.
References
- T. M. N. Bui, M. Sheikh, T. Q. Dinh, A. Gupta, D. W. Widanalage and J. Marco, “A Study of Reduced Battery Degradation Through State-of-Charge Pre-Conditioning for Vehicle-to-Grid Operations,” in IEEE Access, vol. 9, pp. 155871-155896, 2021
- Kailong Liu, T.R. Ashwin, Xiaosong Hu, Mattin Lucu, W. Dhammika Widanage, An evaluation study of different modelling techniques for calendar ageing prediction of lithium-ion batteries, Renewable and Sustainable Energy Reviews, Volume 131, 2020
- Martin Petit, Eric Prada, Valérie Sauvant-Moynot, Development of an empirical aging model for Li-ion batteries and application to assess the impact of Vehicle-to-Grid strategies on battery lifetime, Applied Energy, Volume 172, 2016