Parallel pack question: should balancing currents exist during zero-current rest periods?

[dawoomi]

Hi PathSim team,

I'm investigating current sharing and aging divergence in parallel battery packs.

While reviewing the implementation, I noticed that branch currents appear to be calculated as fractions of the external pack current:

I_a = f(...) * I_pack
I_b = g(...) * I_pack

This means that during a rest period:

I_pack = 0

and therefore:

I_a = 0
I_b = 0

even when the parallel cells have different terminal voltages.

My question is:

Is this the intended physical behavior in PathSim?

For a real parallel pack, I would normally expect voltage-driven balancing currents to exist during rest, e.g.

I_pack = 0
I_a = +50 mA
I_b = -50 mA

while net pack current remains zero.

Does PathSim currently support this type of internal balancing current during rest, or is this outside the current modeling scope?

I'm exploring current-sharing, SOC divergence, and aging behavior, so understanding the intended behavior would be very helpful.

Thanks!

[milanofthe]

Hi @dawoomi, I am tagging @DavidMStraub who is the maintainer of the pathsim-batt toolbox.

[DavidMStraub]

Hi @dawoomi can you please share the model you're referring to?

[dawoomi]

Hi, I'm currently testing with a simple 2P pack using PyBaMM's DFN model with the Chen2020 parameter set. While reviewing the implementation, I noticed that branch currents appear to be derived from the external pack current. During a zero-current rest period, this seems to force all branch currents to zero even when parallel cells have different terminal voltages. My question was whether voltage-driven balancing currents between parallel cells during rest are currently supported, or whether this behavior is outside the intended scope of the current parallel-pack model. Thanks!

…

------------------------------------------------------------------- Dawoomi Hong

On Fri, Jul 10, 2026, 12:55 PM David Straub ***@***.***> wrote: Hi @dawoomi <https://github.com/dawoomi> can you please share the model you're referring to? — Reply to this email directly, view it on GitHub <#238?email_source=notifications&email_token=B3DJCDDTQK6V23FGWXIUPFL5EFC4DA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZWGAYDINJQUZZGKYLTN5XKO3LFNZ2GS33OUVSXMZLOOSWGM33PORSXEX3DNRUWG2Y#discussioncomment-17600450>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/B3DJCDHEPKRGA7U56TII33T5EFC4DAVCNFSNUABIKJSXA33TNF2G64TZHM4DENZXHE2DEMBUHNCGS43DOVZXG2LPNY5TCMBUGA3DGMRVUF3AE> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/B3DJCDDSIZOFROSDCUINHEL5EFC4DA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZWGAYDINJQUZZGKYLTN5XKO3LFNZ2GS33OUVSXMZLOOSVGM33PORSXEX3JN5ZQ> and Android <https://github.com/notifications/mobile/android/B3DJCDDR7AX6DYT2VYZQYDT5EFC4DA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZWGAYDINJQUZZGKYLTN5XKO3LFNZ2GS33OUVSXMZLOOSXGM33PORSXEX3BNZSHE33JMQ>. Download it today! You are receiving this because you were mentioned.Message ID: ***@***.***>

[DavidMStraub]

Again - can you please share the model you're referring to? Minimal reproducible example demonstrating the issue?

[dawoomi]

Sure.

The example is a very simple 2P pack:

• Cell A: 5.0 Ah
• Cell B: 4.5 Ah
• PyBaMM DFN model (Chen2020 parameters)
• Parallel connection
• CC charge
• Zero-current rest period

My question comes from reviewing the current-sharing implementation rather than a simulation result.

From the code, branch currents appear to be calculated as fractions of the external pack current.

Therefore during a rest period:

Ipack = 0

which seems to imply:

Ia = 0
Ib = 0

even when the two parallel cells have different terminal voltages.

In a physical parallel pack, I would normally expect voltage-driven balancing currents to exist during rest while the net pack current remains zero.

I'm trying to understand whether this behavior is intentionally outside the scope of the current parallel-pack implementation, or whether balancing currents during rest are expected to be supported.

Thanks.

[DavidMStraub]

You're constantly talking about a parallel-pack implementation but not sharing it. pathsim-batt contains cell models which can be used in a parallel pack implementation but I need to see the model you're talking about.

Let me be crystal clear.

1. I'd be happy to help you find the answer to your question
2. If you don't share a minimal example, this discussion is over

Thank you very much

[dawoomi]

simplified example

cell_a = PyBaMMCell(
model="DFN",
parameter_set="Chen2020",
capacity_ah=5.0
)

cell_b = PyBaMMCell(
model="DFN",
parameter_set="Chen2020",
capacity_ah=4.5
)

pack = ParallelPack(
cells=[cell_a, cell_b]
)

charge

pack.run(current=5.0, duration=3600)

rest

pack.run(current=0.0, duration=1800)

During the rest period:

pack current = 0 A

Cell A voltage = 4.18 V
Cell B voltage = 4.12 V

Expected:
balancing current A -> B

Observed:
Ia = 0
Ib = 0

[DavidMStraub]

Thank you, but this is not a reproducible example. ParallelPack is not a symbol that exists in pathsim-batt or pathsim.

[dawoomi]

Thanks. The previous snippet was only pseudo-code to explain the idea. The actual implementation I'm using is available here: dawoomi/pathsim_playground <https://github.com/dawoomi/pathsim_playground> I'm currently experimenting with a simple 2P pack using PyBaMM DFN (Chen2020) cells with capacity mismatch. My question is specifically about behavior during zero-current rest periods and whether voltage-driven balancing currents between parallel cells are currently expected to occur. If I'm misunderstanding the intended model behavior, I'd appreciate any guidance. Thanks! [image: image.png]

…

On Mon, Jul 13, 2026 at 10:46 AM David Straub ***@***.***> wrote: You're constantly talking about a parallel-pack implementation but not sharing it. pathsim-batt contains cell models which can be used in a parallel pack implementation but I need to see the model you're talking about. Let me be crystal clear. 1. I'd be happy to help you find the answer to your question 2. If you don't share a minimal example, this discussion is over Thank you very much — Reply to this email directly, view it on GitHub <#238?email_source=notifications&email_token=B3DJCDBE63LU2JV762XCXEL5EUOBFA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZWGI2TSMBRUZZGKYLTN5XKO3LFNZ2GS33OUVSXMZLOOSWGM33PORSXEX3DNRUWG2Y#discussioncomment-17625901>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/B3DJCDCW4MTLNN5Y3AEOMI35EUOBFAVCNFSNUABIKJSXA33TNF2G64TZHM4DENZXHE2DEMBUHNCGS43DOVZXG2LPNY5TCMBUGA3DGMRVUF3AE> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/B3DJCDEN5VBZBYJPA3EY3FD5EUOBFA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZWGI2TSMBRUZZGKYLTN5XKO3LFNZ2GS33OUVSXMZLOOSVGM33PORSXEX3JN5ZQ> and Android <https://github.com/notifications/mobile/android/B3DJCDEKZFMHLVH5IULVSVD5EUOBFA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZWGI2TSMBRUZZGKYLTN5XKO3LFNZ2GS33OUVSXMZLOOSXGM33PORSXEX3BNZSHE33JMQ>. Download it today! You are receiving this because you were mentioned.Message ID: ***@***.***>

-- ------------------------------------------------------------------- Dawoomi Hong

[DavidMStraub]

I'll unsubsribe from this thread for now.

[dawoomi]

Apologies, I am new to GitHub.

I realized that I shared my GitHub profile while the repository was still private, so you could not see the implementation.

The repository is public now:

https://github.com/dawoomi/pathsim_playground

The parallel-pack implementation I was referring to is in:

parallel_pack_block.py

I think I found the source of my confusion. My custom ParallelPackBlock computes branch currents as:

Therefore during rest, when i_pack = 0, both branch currents are forced to zero by definition.
So my current implementation is clearly an Ipack-based current splitter, not a physical voltage-driven parallel-network model.
If I want physical voltage-driven charge redistribution between parallel cells during zero-current rest, what is the recommended formulation in PathSim?
Is there an existing block / network representation for this, or would this require implementing a different pack topology?