Degradation models for lithium batteries

[wigging]

Does PyBaMM have aging/degradation models for lithium batteries? I don't see anything in the documentation about this capability. A paper by Reniers, Mulder, and Howey (see reference below) gives an overview of various degradation models and provides a C++ code called SLIDE (see GitHub link below). Since I prefer Python more than C++, I'm hoping that PyBaMM would have similar features as SLIDE.

1. Jorn M. Reniers, Grietus Mulder, and David A. Howey. "Review and Performance Comparison of Mechanical-Chemical Degradation Models for Lithium-Ion Batteries." Journal of The Electrochemical Society, vol. 166, no. 14, pp. A3189-A3200, 2019. https://doi.org/10.1149/2.0281914jes
2. SLIDE. A C++ code that simulates degradation of lithium ion cells. https://github.com/davidhowey/SLIDE

[valentinsulzer]

You can do degradation modeling by changing the options you pass to the model https://pybamm.readthedocs.io/en/latest/source/models/base_models/base_battery_model.html#pybamm.BatteryModelOptions

See the example notebooks for some examples. If there is a specific degradation model you are after and it isn't there you can open a new issue

[wigging]

Thank you for the comment. I'll try the example notebooks and let you know if I still have questions.

[alihassan90]

@tinosulzer i cant access this link. can you repaste?

[brosaplanella]

This is the updated link: https://pybamm.readthedocs.io/en/latest/source/api/models/base_models/base_battery_model.html#pybamm.BatteryModelOptions

[NLithium]

I will continue on this topic. I want to set up own submodel for calculating SEI formation and SEI reformation, but I have troubles correctly implementing it into PyBaMM. For example, how to correctly implement this submodel in PyBaMM - for example the SEI formation:

where resistance is defined as

and exchange current density is defined as:

equations are based on Journal of The Electrochemical Society, 164 (12) E287-E294 (2017).

[valentinsulzer]

Chapter 5 for SEI refs:
Marquis - 2020 - Long-Term Degradation of Lithium-ion Batteries.pdf

[valentinsulzer]

To change without editing source code you need to write your own SEI class that implements all the same methods as the sei growth class in pybamm, then do

model = pybamm.lithium_ion.DFN(build=false)
model.submodels["sei"] = SEI(model.param)
model.build_model()

See https://github.com/pybamm-team/PyBaMM/blob/develop/examples/notebooks/models/using-submodels.ipynb

[NLithium]

@tinosulzer Thank you very much for your answer!
Currently I am still working with source code, because we are looking for implementation of specific equations.

[Brushnie33]

@tinosulzer Hello, I was trying to execute the EC-reaction limited SEI model by writing it as a class, and then use the following lines:
model = pybamm.lithium_ion.DFN(build=False)
model.submodels["Negative sei"] = SEI_new_option(model.param, "negative","full electrode", model.options)
model.build_model()

I tried to compare the standalone class and the EC-reaction limited SEI model in the existing SEI growth model, the SEI capacity loss and thickness match with each other, however, I couldn't get the total lithium lost correct:

Do you have any suggestions on how to modify the codes?
I copied class BaseInterface(pybamm.BaseSubModel) and BaseModel(BaseInterface) from the source codes, and then define SEI_new_option class for the EC-reaction limited model, do I need to add SEI_new_option in the "SEI" section in base_battery_model.py source codes?

[Brushnie33]

I added the SEI_new_option in the "SEI" section in base_battery_model.py and now it works properly.

[NLithium]

Additionally, I have a follow up questions.

1. I'm attempting to implement the "not equal" operator '>' in the parameter function. The idea is to adjust the plating/stripping equation according on the potential sign.
   

Phi is variables["Negative electrode surface potential difference"]
How can I get the actual value of phi(number) from symbols in a parameter function?
Or there is another way to adjust some equation based on the potential sign with boundary condition?

2. I wanted to run a case with experiment using Chen2020 & Mohtat2020 parameters with okane2020_Li_plating parameters

$$experiment = pybamm.Experiment( ["Charge at C/2 until 4.2 V", "Hold at 4.2 V until C/10", "Rest for 5 minutes", "Discharge at C/2 until 2.8 V", "Rest for 5 minutes"] * 10$$ $$model = pybamm.lithium_ion.DFN( { "cell geometry": "arbitrary", "SEI": "reaction limited", "SEI film resistance": "average", "SEI porosity change": "true", "lithium plating": "reversible", "lithium plating porosity change": "true" "thermal": "lumped" } )$$

The output I get for ```Chen2020``:
"psetup failed: .../casadi/interfaces/sundials/idas_interface.cpp:849: Calculating Jacobian failed"

The output I get for ```Mohtat2020``:
"Experiment is infeasible: 'event: Minimum voltage' was triggered during 'Charge at 1 C until 4.2 V'. The returned solution only contains the first 0 cycles. Try reducing the current, shortening the time interval, or reducing the period."

How can resolve these issues?

[xcblovecy]

You attemped to implement the "not equal" operator '>' in the parameter function. The idea is to adjust the plating/stripping equation according on the potential sign. Can this approach be implemented?