Error while trying to run drive cycles long experiment on Li ion SPM model

[alihassan90]

I am getting this error while i try to run the experiment on my Li-ion cell

step 3/5: _Step(current, [[    1     2]
 [    2     3]
 [    3     2]
 ...
 [48059     2]
 [48060     2]
 [48061     3]])
2023-07-06 16:32:27.994 - [WARNING] callbacks.on_experiment_infeasible(231): 

	Experiment is infeasible: 'event: Minimum voltage [V]' was triggered during '_Step(current, [[    1     2]
 [    2     3]
 [    3     2]
 ...
 [48059     2]
 [48060     2]
 [48061     3]])'. The returned solution only contains up to step 3 of cycle 1. 
2023-07-06 16:32:30.464 - [NOTICE] callbacks.on_experiment_end(220): Finish experiment simulation, took 3.981 s

The code looks like this

import pybamm
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np

parameter_values = pybamm.ParameterValues("Mohtat2020")
parameter_values.update({"SEI kinetic rate constant [m.s-1]": 1e-12})
#parameter_values.update({"SEI kinetic rate constant [m.s-1]": 1e-14})
spm = pybamm.lithium_ion.SPM({"thermal": "lumped", "SEI": "reaction limited","SEI film resistance": "distributed", "SEI porosity change": "false"})
# Calculate stoichiometries at 100% SOC

pybamm.set_logging_level("NOTICE")



drive_cycle=pd.read_csv("US06.csv",comment='#',header=None).to_numpy()

def map_drive_cycle(x, min_op_value, max_op_value):
    min_ip_value = x[:, 1].min()
    max_ip_value = x[:, 1].max()
    x[:, 1] = (x[:, 1] - min_ip_value) / (max_ip_value - min_ip_value) * (
        max_op_value - min_op_value
    ) + min_op_value
    return x

drive_cycle_power = map_drive_cycle(drive_cycle, 1.5, 3.5)



experiment = pybamm.Experiment(
    [
        (
             "Discharge at 1 C until 3.0 V",
             "Rest for 1 hour",
             pybamm.step.current(drive_cycle),
             "Charge at 1C until 4.2V",
             "Hold at 4.2V until C/50"
         ),
    ]*5,
)

[brosaplanella]

The Mohtat parameter set has initial conditions for a fully discharged battery. You can either add a charging step at the beginning of your experiment, or pass an initial_soc argument when you solve your simulation.

[alihassan90]

Thank you, it worked.

The paper you tagged above shows how SEI and Loss of Lithium makes the degradation non-linear near the knee point or beyond 80% SOH. However I am only getting a linear degradation no matter how many cycles I run my drive cycles for. Whats the reason for that. Do i need to change the SEI, Lithium plating etc options?

[brosaplanella]

Sorry, I pasted the wrong link. I meant to paste a link of the PyBaMM docs: https://docs.pybamm.org/en/latest/source/api/simulation.html#pybamm.Simulation.solve

I guess the knee is not captured due to the physics/parameters you are using. For example, in the paper above they use SEI + plating, with porosity change. With SEI only (and constant porosity) I don't think it is possible to capture the knee.

[alihassan90]

Any suggestions on how i can inculcate porosity change in my experiment?

"SEI porosity change": "true"
or
"lithium plating porosity change": "true"

will it do the the job? Thanks

[brosaplanella]

That should enable the porosity change, but might not be enough to capture the knee. Note that if you have porosity change for plating, but don't set a plating model, that has no contribution to your model.

[ganmad]

Is there an easier way to add cut-off voltages for each step in drive profile for below command?

experiment = pybamm.Experiment( [pybamm.step.current(drive_cycle_current)] )

This command lets simulation run with a current given on one column for a duration given on another column, for a row. Now, for each step on given row, can I add a cut-off voltage as well?

The alternative method I think is to generate a list of commands for each row, like below:

pybamm.step.current(1, duration="1 hour", termination="2.5 V")

pybamm.step.current(-1, duration="1 hour", termination="4.2 V")

But, can I use the command to run the steps between 2.5V - 4.2V, and if the step crosses the voltage, then move to next step?
Something like experiment = pybamm.Experiment( [pybamm.step.current(drive_cycle_current), cut-off = ["2.5 V", "4.2 V"] )

Thanks

[ganmad]

Agreed. Thanks for taking time to reply. Sorry, I over simplified current profile, but I was asking what could we do if duration and magnitude of current varies frequently as in drive cycle. If the cell can hold the current for the given duration, there won't be a problem running the drive profile on the cell. Otherwise, it wont run for the given duration.

[brosaplanella]

If you run a given drive cycle (e.g. in which current changes every second), then you can use

experiment = pybamm.Experiment([pybamm.step.current(drive_cycle)])

where drive_cycle is a 2-column array (more details in Tutorial 5). If you set termination conditions they will apply to the whole drive cycle.

If you want the drive cycle to continue from the next step, then I think you need to define each individual step as you did above.

[alihassan90]

Hello, Is there a way to include both termination conditions

max,min voltage
&
80% SOH in experiment. Apparently "termination" key word can be used once only.

[alihassan90]

Also this pybamm.Experiment( [pybamm.step.current(drive_cycle_current), termination= ["2.5 V", "4.2 V"] is treating 4.2V as stopping voltage and not maximum voltage. web

Error:

2023-07-26 11:33:06.084 - [NOTICE] callbacks.on_cycle_end(213): Stopping experiment since minimum voltage (3.933 V) is below stopping voltage (4.200 V).
2023-07-26 11:33:06.086 - [NOTICE] callbacks.on_experiment_end(220): Finish experiment simulation, took 25.471 s

[ganmad]

Instead of drive profile, you can create a list of step current commands using for loop for each row in the drive profile and pass them to Experiment class. You can then use termination as 80% capacity as well. Hope this works.

For example: For each row in the drive profile,
pybamm.step.current(1, duration="1 hour", termination="2.5 V")
pybamm.step.current(-1, duration="1 hour", termination="4.2 V")
pybamm.step.current(0, duration="1 hour")
pybamm.step.current(1, duration="1 hour", termination="2.5 V")
pybamm.step.current(-1, duration="1 hour", termination="4.2 V")
pybamm.step.current(0, duration="1 hour",)
pybamm.step.current(1, duration="1 hour", termination="2.5 V")
pybamm.step.current(-1, duration="1 hour", termination="4.2 V")

[ganmad]

Is there a way to add SoC as a cut-off instead of voltage for discharge/charge step?
For example, I wanted to implement:
pybamm.step.current(1, duration="1 hour", termination="5% Capacity")

I wanted to run several of such charge/discharge cycles. I understand termination works for whole cycling experiment, but not sure how to cycle the cell between partial depth of discharge, say 80% to 20% SoC instead of voltage cut-off. Any insights will be helpful. Thanks.

Best regards,
Ganesh

[brosaplanella]

This is not yet implemented, but we are working on it: #2553

[ganmad]

Thanks for update.