Errors about long cycle simulation

[YouHyin]

I created a function to compute the cycle life using the Prada2013 parameter sets.

import pybamm
import numpy as np
from set_parameters import set_parameters

pybamm.set_logging_level("NOTICE")
parameter_values=set_parameters()
# Calculate stoichiometries at 100% SOC
parameter_values.set_initial_stoichiometries(1);
options = {
    "SEI": "ec reaction limited",
    "lithium plating": "irreversible",
    "thermal" : "lumped",
    }
spme = pybamm.lithium_ion.SPMe(options)
N_cycles=100

def compute_cycle_life(c1,c2,c3,soc1,soc2,soc3) :

    time_1=soc1*3600/c1
    time_2=soc2*3600/c2
    time_3=soc3*3600/c3

    experiment = pybamm.Experiment([
                                       (f"Discharge at 4C until 2.0V",
                                        f"Hold at 2.0V until C/50",
                                        "Rest for 5 minutes",
                                        f"Charge at {c1} C for {time_1} seconds",
                                        "Rest for 5 minutes",
                                        f"Charge at {c2} C for {time_2} seconds",
                                        "Rest for 5 minutes",
                                        f"Charge at {c2} C for {time_2} seconds",
                                        "Rest for 5 minutes",
                                        f"Hold at 3.6V until C/50",

                                        )
                                   ] * N_cycles,
                                   termination="95% capacity"

                                   )
    sim = pybamm.Simulation(spme, experiment=experiment, parameter_values=parameter_values)
    #solver = pybamm.CasadiSolver(atol=1e-6, rtol=1e-6, mode="fast")
    sol=sim.solve()
    cycle_life=sol.summary_variables["Cycle number"]

    return cycle_life

print(compute_cycle_life(3.4,5,4.8,0.2,0.3,0.5))

However, when I ran this code, I got this error.
pybamm.expression_tree.exceptions.SolverError: Could not solve for summary variables, run sim.solve(calc_esoh=False)` to skip this step

So I changed it to sim.solve(calc_esoh=False) and now I get this error and can't run it.
raise KeyError(f"'{key}' not found. Best matches are {best_matches}")
KeyError: "'Capacity [A.h]' not found. Best matches are ['Measured capacity [A.h]', 'Throughput capacity [A.h]', 'Loss of capacity to SEI [A.h]']"

What should I do to solve this situation??
I would be very grateful if you could let me know.

[brosaplanella]

The second error is because in order to access the summary_variables you need to have calc_esoh=True. For the first error it's hard to tell, could you share a minimum working example?

[brosaplanella]

This example would not run as we do not have the set_parameters function. A minimum working example should be a single script with the minimum code to make the error appear (i.e. you should check which lines are not necessary to reproduce the error and remove it).

[Sparsh-Proj]

We are facing a similar issue. With "calc_esoh=False", we can not access the summary variables. But we'd like to get summary variables for plotting purposes.
With "calc_esoh=True", (Which is the default value, nonetheless), the code fails and throws a SolverError as follows:
SolverError: Could not solve for summary variables, run `sim.solve(calc_esoh=False)` to skip this step

Link to minimum working code examples' google colab: https://colab.research.google.com/drive/11xUA7_o9avRH6ka3RgN1qEQAnAuqPtUj?usp=sharing

I've attached the same code below as well:

%pip install pybamm -q

import pybamm

pybamm.set_logging_level("NOTICE")

model = pybamm.lithium_ion.DFN()

params = pybamm.ParameterValues("Prada2013")

exp = pybamm.Experiment(
    [
        ("Discharge at C/10 for 10 hours or until 2.8 V",
        "Rest for 1 hour",
        "Charge at 5 A until 3.5 V",
        "Hold at 3.5 V until 50 mA",
        "Rest for 1 hour"),
    ] * 10
)

sim = pybamm.Simulation(
                          model, 
                          experiment = exp, 
                          parameter_values = params
                       )

# sol = sim.solve()

# Throws following error: SolverError: Could not solve for summary variables, run `sim.solve(calc_esoh=False)` to skip this step

# Bypassing the error:

sol = sim.solve(calc_esoh = False)

pybamm.plot_summary_variables(sol)

# Throws following error: KeyError: "'Capacity [A.h]' not found. Best matches are ['Measured capacity [A.h]', 'Throughput capacity [A.h]', 'Loss of capacity to SEI [A.h]']"

[brosaplanella]

I think the issues are closely related to the Prada parameter set. Some of the values did not match the paper, and this has now been fixed in the develop branch (#3096). You can try to install from source and see if it works.

[YouHyin]

Hi, I saw (#3096) and I don't quite understand: are you asking me to install pybamm version 23.5?? If so, can I install it via this command? (conda install -c conda-forge pybamm=23.5) I'm having trouble installing it, how can I fix it? Thank you very much.

[brosaplanella]

I am not familiar with conda (I use pip), but I guess the instructions here should work: https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-pkgs.html#updating-packages