Separate Voltages for events and setting initial SOC

[js1tr3]

Description

It appears the parameters
Lower voltage cut-off [V] 2.8
Upper voltage cut-off [V] 4.2

are used both to set the initial SOC and for simulation events, as a result if the voltage limits for discharge are changed, this impacts the initial SOC.

Motivation

It would be good to have a sperate variable for the simulation event termination and calculation of SOC/initial stoichiometry.
The present implementation is breaking my use case of comparing the model to different discharge voltages, as the initial SOC changes also.

Possible Implementation

No response

Additional context

No response

[valentinsulzer]

Hi Jason :) You can change the values in the experiment specification

pybamm.Experiment(
    [
        "Discharge at C/5 for 10 hours or until 3 V",
        "Rest for 1 hour",
        "Charge at 1 A until 4.2 V",
        "Hold at 4.2 V until 10 mA",
        "Rest for 1 hour",
    ]
)

without affecting the voltage limits for SOC, is this what you need?

[js1tr3]

Hi Tino, thanks for the rapid response! I'm actually using a recorded current trace, and feeding in using the old interpolant method. I'll give the new experiment using drive cycles a go. I just pulled upstream changes and broke a few things on my branch... so im trying to catch up on parameter changes. What I want to do is charge to 110% SOC or over discharge assuming the ocp curve is well defined and stoichiometries are nonzero (within limits) this means the voltage events should be outside the min/max voltages used for calculating SOC... if that makes sense... happy to talk through my idea if you have time. Best -J

…

On Wed, Dec 14, 2022, 9:48 PM Valentin Sulzer ***@***.***> wrote: Hi Jason :) You can change the values in the experiment specification pybamm.Experiment( [ "Discharge at C/5 for 10 hours or until 3 V", "Rest for 1 hour", "Charge at 1 A until 4.2 V", "Hold at 4.2 V until 10 mA", "Rest for 1 hour", ] ) without affecting the voltage limits for SOC, is this what you need? — Reply to this email directly, view it on GitHub <#2562 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAE2YEPGCSBWZVMNWMEXAQTWNKBHBANCNFSM6AAAAAAS7GYB64> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[valentinsulzer]

I see, that makes sense. And even if the SOC is still <100% the voltage might go above the cut-off. So we should add parameters "Open circuit potential at 0% SOC" and "Open circuit potential at 100% SOC", which can be different from the voltage cut-offs.

Could you open a PR to do this? It should be pretty straightforward. Basic steps are:

1. Add these parameters to each parameter set, defaulting to the same value as the cut-offs
2. Add a two new Parameter object to the lithium_ion_parameters.py files (e.g. self.opc_soc_0, self.ocp_soc_100)
3. Replace the cut-off voltage in this line (and other places in that file) with the new target ocp voltage

   PyBaMM/pybamm/models/full_battery_models/lithium_ion/electrode_soh.py

   Line 446 in 1f25a1a

   V_min = parameter_values.evaluate(param.voltage_low_cut_dimensional)

[js1tr3]

I'll take a stab at it!

…

On Thu, Dec 15, 2022, 10:24 AM Valentin Sulzer ***@***.***> wrote: I see, that makes sense. And even if the SOC is still <100% the voltage might go above the cut-off. So we should add parameters "Open circuit potential at 0% SOC" and "Open circuit potential at 100% SOC", which can be different from the voltage cut-offs. Could you open a PR to do this? It should be pretty straightforward. Basic steps are: 1. Add these parameters to each parameter set, defaulting to the same value as the cut-offs 2. Add a two new Parameter object to the lithium_ion_parameters.py files (e.g. self.opc_soc_0, self.ocp_soc_100) 3. Replace the cut-off voltage in this line (and other places in that file) with the new target ocp voltage https://github.com/pybamm-team/PyBaMM/blob/1f25a1a3bbaf739848b67dec3b81562db5ad5c7a/pybamm/models/full_battery_models/lithium_ion/electrode_soh.py#L446 — Reply to this email directly, view it on GitHub <#2562 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAE2YEMCQTAO6JUNAHUWNG3WNMZ2TANCNFSM6AAAAAAS7GYB64> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[js1tr3]

I made the changes you suggested in my repo for testing, but it seems that the voltage is not updated, is it possible that if the initial SOC is still one then the calculation of the initial concentrations is being skipped? https://github.com/js1tr3/PyBaMM/blob/7b943d147f14a51892e29cf1b6e16295ec1358b7/pybamm/simulation.py#L384

…

On Thu, Dec 15, 2022 at 6:39 PM Jason Siegel ***@***.***> wrote: I'll take a stab at it! On Thu, Dec 15, 2022, 10:24 AM Valentin Sulzer ***@***.***> wrote: > I see, that makes sense. And even if the SOC is still <100% the voltage > might go above the cut-off. So we should add parameters "Open circuit > potential at 0% SOC" and "Open circuit potential at 100% SOC", which can be > different from the voltage cut-offs. > > Could you open a PR to do this? It should be pretty straightforward. > Basic steps are: > > 1. Add these parameters to each parameter set, defaulting to the same > value as the cut-offs > 2. Add a two new Parameter object to the lithium_ion_parameters.py > files (e.g. self.opc_soc_0, self.ocp_soc_100) > 3. Replace the cut-off voltage in this line (and other places in that > file) with the new target ocp voltage > https://github.com/pybamm-team/PyBaMM/blob/1f25a1a3bbaf739848b67dec3b81562db5ad5c7a/pybamm/models/full_battery_models/lithium_ion/electrode_soh.py#L446 > > — > Reply to this email directly, view it on GitHub > <#2562 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AAE2YEMCQTAO6JUNAHUWNG3WNMZ2TANCNFSM6AAAAAAS7GYB64> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> >

[valentinsulzer]

Does test_electrode_soh.py work as expected?

[js1tr3]

it does now that I re-wrote the unit test. one complication is that I think the Vmin/Vmax has to be outside the voltage windows for SOC for the solver to converge. probably I will try to add a warning and unit test for that also in the coming week.

…

On Mon, Dec 19, 2022 at 6:14 AM Valentin Sulzer ***@***.***> wrote: Does test_electrode_soh.py work as expected? — Reply to this email directly, view it on GitHub <#2562 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAE2YELMY7L6ZADDCJEP6FDWOA7TBANCNFSM6AAAAAAS7GYB64> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[valentinsulzer]

Can you open a draft pull request so I can check it? It shouldn't make a difference what Vmin and Vmax are for this calculation, if it's based entirely on OCP at 0/100% SOC. Make sure you change voltage_low_cut_dimensional to opc_soc_0_dimensional everywhere relevant.

[rtimms]

@js1tr3 how is this going?

[js1tr3]

@js1tr3 how is this going?

Good, got tied up with teaching this semester, but hope to get back on this shortly. Any suggestions for resolving conflicts in jupyter notebooks?

[rtimms]

No worries, thanks for the update. Conflicts in notebooks can be a bit of a pain. Are they notebooks you have actively changed for this PR? If not, just accept all the incoming changes or check the file out from develop.

[valentinsulzer]

Solved by #2580