OCP hysteresis for a half-cell model

[lefile]

Hi,

tried to simulate the following half-cell model:

half_cell_model = pybamm.lithium_ion.DFN( options={"working electrode": "positive", "open-circuit potential": "current sigmoid"})

but got the error:
UnboundLocalError: cannot access local variable 'ocp_surf' where it is not associated with a value

Is it possible that the "current sigmoid" option does not work for half-cell models?

[lefile]

Any opinion/feedback here?

Best regards!

[agriyakhetarpal]

Hi @lefile, could you please post a full MWE of your code so that someone can help you better? Thanks!

[rtimms]

Without seeing the full code my guess is you need to do half_cell_model = pybamm.lithium_ion.DFN( options={"working electrode": "positive", "open-circuit potential": ("single", "current sigmoid")}) as it is currently trying to use the hysteresis model in both the Li metal and the working electrode.

[lefile]

Hi guys, here is a MWE:

import pybamm

param = pybamm.ParameterValues("Xu2019")

half_cell_model = pybamm.lithium_ion.DFN(
        options={
            "working electrode": "positive",
            "open-circuit potential": ("single", "current sigmoid")
        }
    )

exp = pybamm.Experiment(
        [
            (
                "Discharge at 1C for 2 hours or until 3.0 V",
                "Charge at 1C for 2 hours or until 4.2 V",
            )
        ]
    )

sim = pybamm.Simulation(
    half_cell_model,
    parameter_values=param,
    solver=pybamm.CasadiSolver(mode="safe"),
    experiment=exp
)

sim.solve()

Where Xu2019 is changed in a way that I replaced parameter

"Positive electrode OCP [V]": nmc_ocp_Xu2019'

with

"Positive electrode lithiation OCP [V]": nmc_ocp_Xu2019,
"Positive electrode delithiation OCP [V]": nmc_ocp_Xu2019,

The error I get is
KeyError: "'Positive electrode OCP [V]' not found.

[DrSOKane]

The culprit here is the electrode stats of health (eSoH) model, which does not work properly when there is hysteresis. If you don't need eSoH, just turn it off using sim.solve(calc_esoh=False). If you do need it, you can make up a function for "Positive electrode OCP [V]" but be prepared for errors to creep in, depending on your choice of function.

[rtimms]

To clarify: the eSOH model assumes you are at thermodynamic equilibrium with a single OCP for each electrode using the equations from this paper.

I would recommend setting calc_esoh=False and then computing any state-of-health metrics yourself to account for hysteresis.

[lefile]

Running the MWE from before with sim.solve(calc_esoh=False) still produces the same error.