Increasing trend in the 'Total lithium in negative electrode [mol]' during aging simulations across cycles

[Hanyu-B]

Dear Pybamm Team,

I have noticed a case wheretotal lithium in negative electrode [mol]' shows an increasing trend during aging simulation over time, so I am just wondering whether this is a bug or just something underlaying different aging results. Also, should we focus on the difference between the max and min of this value as the total cyclable lithium, which means it is determined by the stochemistry of the cell?

Details regarding simulation: 1C CCCV cycling aging for about 1500 cycles with the following aging models enabled:

   model2 = pybamm.lithium_ion.DFN({
        "SEI": SEI_growth_option[1], #ec reaction limited 
        "SEI on cracks": "true",
        "SEI film resistance":"distributed",
        "SEI porosity change":"true",
        "lithium plating":plating_options[1], #irreversible one 
        "lithium plating porosity change":"true",
        "loss of active material": "stress and reaction-driven",
        "particle": "Fickian diffusion",
        "particle mechanics": ("swelling and cracking", "swelling only"),
        "thermal": "lumped",
    }) 

Solver settings: default 'safe mode' with adjustment of tolerance level from 1e-6 to 1e-4.

Figure of results:

Different color plots are just different cycling conditions where the orange one is the 1 C CCCV protocol plot.
Thanks in advance for your time and consideration (˶˚ ᗨ ˚˶)!

Thank and Best regards,
Hanyu(˶˚ ᗨ ˚˶)