Hey folks. This is merely one (the first!) of a series of threads where we can discuss the various effects of battery pack balance. This one will be focused on range/capacity fluctuations.
Likely, this is the least important of all such related topics. Thus, this will be the lightest of all those threads, and probably the only one with pictures!
Context
First, this is all merely informational and there is little you can do with this information. It is just information, and isn't significant enough to warrant a change to your charging habits despite any takeaways.
This is the often-cited post regarding pack imbalance, containing an excerpt from a leaked internal document: Battery Management System - What I Learned At Tesla Service Center. I'll draw some conclusions & assumptions from it below, these will probably be discussed elsewhere. But most importantly, we must recognise that any interpretation of those documents may be incorrect (or updates may have changed it).
I will take two meanings for imbalance, though they tell a similar story:
Based on a lot of data with my own car, I observed that changes to our charge set-point (e.g. charge to 80% one week, start charging to 90% next week or perhaps 100% for a single day) will result in small changes in estimated capacity.
Further related observation is that on trips, my estimated capacity changes more so than usual. This is partly due to my charging habits temporarily changing a lot. After the trip, my numbers settle back to what they were before within a week.
I therefore speculate that to some degree, along with Assumption #2, the dynamic balance point has a daily influence on true capacity (which is then accurately reported by the BMS as fluctuations in daily range).
Explanation: Example Scenario
I present an exaggerated scenario: A battery pack with 4 series bricks with rather large (but managed) imbalances between them. The graphs below are in terms of energy and not voltage to simply things (and related more to what we see in the car).
When charging to 80%, let's say it aims to have the packs balanced at this point. The length of the bars represent their respective energy capacities, with the lighter top region representing the "top 20%" for that brick. When charged and balanced to 80%, these should all line up as below.
There are a few further things to note here:
Something interesting happens when we charge and balance at 100% though.
(unlike the other images, this one has the "top" portion of the charge chopped off)
This has some very interesting takeaways:
For one, the above is a theoretical exercise with extremely exaggerated numbers and a lot of hand-waving to demonstrate only the effect in energy imbalances between the bricks. I really must emphasise the huge amount of hand-waving. There's a lot, and even with that the numbers above are just generalisations.
But given that,
A Note on Other Batteries
Generally, packs are ultimately "top-balanced", i.e. balanced at 100% SoC. This extracts the most energy out of the whole pack, since voltage is highest at 100% - more time drawing power at higher voltages results in more energy. The more calculus explanation is simply "area under the curve", if you prefer. These would likely fluctuate a bit less, since they're being handled very consistently.
Model 3 cannot do this, because unlike pretty much every other battery in someone's life, it doesn't get routinely charged to 100%. But it would be disadvantageous to change at whatever percent it rests at - if it was low, well, see previous paragraph. So the thought is that it keeps a "top-ish" balance, wherever you happen to charge to routinely.
And I will sign off every balancing thread with this: If you are worrying about imbalance, don't. This is all merely informational and there is little you can do about it. The car is taking car of it for your charging habits. Tesla has replaced batteries under warranty prior to the 30% threshold for severely imbalanced packs - you would personally know if this was going to happen, because the car tells you and suddenly you're booking an approved warranty replacement of the battery.
Likely, this is the least important of all such related topics. Thus, this will be the lightest of all those threads, and probably the only one with pictures!
Context
First, this is all merely informational and there is little you can do with this information. It is just information, and isn't significant enough to warrant a change to your charging habits despite any takeaways.
This is the often-cited post regarding pack imbalance, containing an excerpt from a leaked internal document: Battery Management System - What I Learned At Tesla Service Center. I'll draw some conclusions & assumptions from it below, these will probably be discussed elsewhere. But most importantly, we must recognise that any interpretation of those documents may be incorrect (or updates may have changed it).
I will take two meanings for imbalance, though they tell a similar story:
- Primarily: Energy capacity differences between groups of parallel cells (called "bricks").
- Secondarily: Voltage differences between bricks (primarily due to capacity differences)
- We understand and trust that balancing occurs above 4.0V/cell (roughly 73%) when the battery is at rest (not providing 12V, e.g. for Sentry Mode, not being charged, and not running climate control in any way)
- As a result, Model 3 contains a sort-of "top balanced" battery, but crucially, the balance point varies.
- Balance impacts estimated range, since it modifies the usable energy (one brick will reach unsafe levels before the others - the stronger the imbalance, the more this is true).
- I'm a dude on the internet. While I am always right because of this, I am simultaneously also always wrong. The truth is somewhere in between. I've learned a lot here, and I'm just trying to share some of that. I'll likely need to correct some of this in the future as I learn more, from both this community and my personal never-ending dive into this car.
Based on a lot of data with my own car, I observed that changes to our charge set-point (e.g. charge to 80% one week, start charging to 90% next week or perhaps 100% for a single day) will result in small changes in estimated capacity.
Further related observation is that on trips, my estimated capacity changes more so than usual. This is partly due to my charging habits temporarily changing a lot. After the trip, my numbers settle back to what they were before within a week.
I therefore speculate that to some degree, along with Assumption #2, the dynamic balance point has a daily influence on true capacity (which is then accurately reported by the BMS as fluctuations in daily range).
Explanation: Example Scenario
I present an exaggerated scenario: A battery pack with 4 series bricks with rather large (but managed) imbalances between them. The graphs below are in terms of energy and not voltage to simply things (and related more to what we see in the car).
When charging to 80%, let's say it aims to have the packs balanced at this point. The length of the bars represent their respective energy capacities, with the lighter top region representing the "top 20%" for that brick. When charged and balanced to 80%, these should all line up as below.
There are a few further things to note here:
- The "smallest" brick, Brick 2, is determining the bottom limit. 0% is reached when the first brick hits 0%, so it limits the bottom of the charge even though others have more energy.
- Using 80-0% will consume "68%" of the original rated range, with the combination of reduced brick capacities (see legend) and their imbalances (mostly Brick 2 ending the cycle early).
- This maximises energy starting at 80% (but missing out slightly for a 100% charge, since it's imbalanced a bit at the top)
Something interesting happens when we charge and balance at 100% though.
- Brick 2, still being the smallest one, still limits the discharge.
- It is now balanced at 100%, but if we discharged to 80%, there would be a small imbalance between the bricks.
- The imbalance at the bottom is greater than the 80% case.
- This maximises energy starting at 100%, but will result in less available energy for any lower portion of the battery (see next picture!)
(unlike the other images, this one has the "top" portion of the charge chopped off)
This has some very interesting takeaways:
- The bottom limitation is actually like that of the 100% balance, assuming it hasn't reached balance at 80% yet (this takes some time at rest to change the balance point!)
- Since the battery was balanced to extract the most energy at 100%, starting from only 80% means it's already imbalanced and we'll extract less energy out of it as a result.
- Eventually, given time to do so, the battery will balance to the 80% point again.
For one, the above is a theoretical exercise with extremely exaggerated numbers and a lot of hand-waving to demonstrate only the effect in energy imbalances between the bricks. I really must emphasise the huge amount of hand-waving. There's a lot, and even with that the numbers above are just generalisations.
But given that,
- Maximum range for a pack with imbalances is only achievable by consistently charging to high SoC (so that it balances to a higher setpoint)
- Some range recovery is potentially available if you start charging to 90% instead of 80%, for example. This has been noted in various cases on this forum, and this is a simple explanation of why that could be the case.
- Since balance is a moving target, that appears to leave some room for fluctuations in usable energy. This could explain at least some of the fluctuations in range on a day-to-day basis, especially since they are generally small.
- This is not universal. Battery packs are incredibly complicated. There are something like 4600 individual cells in a Model 3 LR battery pack, and each one is its own incredibly complex electrochemical system. There will be variances and factors beyond what is discussed here, some of which may give results opposite to what has been stated here.
- Since this may explain some fluctuation, as I often state, I believe the BMS to actually be reporting rated range (capacity) very accurately. It may seem crazy that it's varying +-1% or so from week to week, but perhaps that is it simply factoring in many things you do not actively see.
- In the higher (e.g. 100%) balance setpoints, the imbalance on the low end of the battery is even higher. Voltage imbalance has other side effects (e.g. additional stress on those cells), especially with larger imbalances. Like usual, even if a higher SoC would net a higher range consistently, it may not be best for battery health.
- This corresponds extremely strongly with my own findings, though I recognise the confirmation bias there.
- Charging habits are important if you are tracking "range at 100%", capacity over time, etc. Varying charge set-points will result in truly varying usable energy amounts.
- Charging to "just" 80% is still absolutely fine. The range difference you might get by charging to 90% instead, for example, should be miniscule. If it's large, you likely have a larger imbalance problem and are well on your way to a battery replacement (hopefully under warranty).
A Note on Other Batteries
Generally, packs are ultimately "top-balanced", i.e. balanced at 100% SoC. This extracts the most energy out of the whole pack, since voltage is highest at 100% - more time drawing power at higher voltages results in more energy. The more calculus explanation is simply "area under the curve", if you prefer. These would likely fluctuate a bit less, since they're being handled very consistently.
Model 3 cannot do this, because unlike pretty much every other battery in someone's life, it doesn't get routinely charged to 100%. But it would be disadvantageous to change at whatever percent it rests at - if it was low, well, see previous paragraph. So the thought is that it keeps a "top-ish" balance, wherever you happen to charge to routinely.
And I will sign off every balancing thread with this: If you are worrying about imbalance, don't. This is all merely informational and there is little you can do about it. The car is taking car of it for your charging habits. Tesla has replaced batteries under warranty prior to the 30% threshold for severely imbalanced packs - you would personally know if this was going to happen, because the car tells you and suddenly you're booking an approved warranty replacement of the battery.
