-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Wiki Sudden Loss Of Range With 2019.16.x Software
- Thread starter Dutchmeeuw
- Start date
For a broader context:
Plenty of posts like this cropping up on user sites for various EV's (not surprisingly)
mybmwi3.com
"A number of owners of 60 Ah i3's (2014-2016) are reporting significant loss of actual range (including me). Whether this is due to actual battery cell degradation or the battery management system decreasing the usable capacity to minimize the rate of degradation in hopes of minimizing the number of battery capacity warranty claims remains unknown. The effect on the driver is the same: less usable range."
(20" wheels, Rex vs. Non-Rex, 2014 vs. 2017 - BMW i3 Forum)
Plenty of posts like this cropping up on user sites for various EV's (not surprisingly)
mybmwi3.com
"A number of owners of 60 Ah i3's (2014-2016) are reporting significant loss of actual range (including me). Whether this is due to actual battery cell degradation or the battery management system decreasing the usable capacity to minimize the rate of degradation in hopes of minimizing the number of battery capacity warranty claims remains unknown. The effect on the driver is the same: less usable range."
(20" wheels, Rex vs. Non-Rex, 2014 vs. 2017 - BMW i3 Forum)
BigNick
Infamous Fat Sweaty Guy
I charged overnight at 5A from about 70% to 85%, continued at 40A to 95%, then lowered the charge to 5A again until the charging stopped, which took more than 5 hours. There was no real balancing taking place. Here is my battery after reaching the cutoff, about 5 minutes after disconnecting from the HPWC.
Knowing it is “bad” to leave the pack sit at 100% for a long time, I immediately drove about 70 to 75 miles, down to about 68%.
It appears balancing was occurring during use, rather than charging. The 20mV reported delta concerns me, however.
My pack is original to the car, 5 years old and 42K miles. Still have about 30K and 2.5 years on the Used warranty. Since buying the car I have lost roughly 3 kWh of the pack’s reported capacity.
I Supercharge very infrequently (575 kWh over the life of the car, about half of which was mine) and normally would charge to 80% daily, which I lowered to 75% after charging to 80% would cause the coolant pump to run after charge completion.
I don’t want to need a battery replacement out of warranty or even in-warranty as it seems replacement packs sometimes aren’t much better than the pack being replaced - a Band-Aid fix.
I’d like to know if my pack has a developing problem, and is there anything I can do to help keep it healthy?
Knowing it is “bad” to leave the pack sit at 100% for a long time, I immediately drove about 70 to 75 miles, down to about 68%.
It appears balancing was occurring during use, rather than charging. The 20mV reported delta concerns me, however.
My pack is original to the car, 5 years old and 42K miles. Still have about 30K and 2.5 years on the Used warranty. Since buying the car I have lost roughly 3 kWh of the pack’s reported capacity.
I Supercharge very infrequently (575 kWh over the life of the car, about half of which was mine) and normally would charge to 80% daily, which I lowered to 75% after charging to 80% would cause the coolant pump to run after charge completion.
I don’t want to need a battery replacement out of warranty or even in-warranty as it seems replacement packs sometimes aren’t much better than the pack being replaced - a Band-Aid fix.
I’d like to know if my pack has a developing problem, and is there anything I can do to help keep it healthy?
Sure looks like a voltage cap to me. If the cap is indirect through some other capping mechanism it's still specific enough to be quantified as a voltage cap.
Really interesting observation. Glad you tracked it. You seem to have a group of modules contributing. I have a single module that has reached a peak of ~14mV delta, so not crazy, but still stands out. I'll do a similar test to see if the imbalance reduces due to usage.
It makes sense for it to work this way. What it means is the lower capacity modules are delivering power from their available capacity at a reduced percentage from higher capacity modules, otherwise the other modules would not "catch up". The end result should be that, as the pack reaches overall capacity exhaustion, all of the modules are close, so they run out at nearly the same time. This prevents stranded capacity being left in the pack at low SOC.
Based on the ongoing discussion, it really appears attempting to balance during charging is futile, since the imbalance reflects physical differences in the module conditions. The goal of charging should be to put as much power into the pack as possible without introducing potential damage to the cells/modules.
The discharge cycle is responsible for getting as much of that stored energy out of the pack as possible. What this means is, at any given instance in time, an unbalanced pack will deliver incrementally less power than if all modules were delivering at peak output. The difference is hopefully negligible, though would be interesting to work out the math. Wondering if the output logic targets module equality right at power exhaustion, or at some point above it.
Following on that part of my comment, it speaks to @David99 pointing out the differences between module/cell degradation and chargegate. With a capped pack the goal of charging may still be to put as much power into the pack as possible, but a different limitation besides module differences has been introduced.
For a moment I thought we might be done on this thread! 
.... and I could get on with life.
I don't know about others, but I find it difficult trying to make sure I have read long, detailed posts (mine included) correctly, and then work out what the poster is saying and if it agrees with other posts / knowledge.
Right now I feel we reached a point where a lot of info came together fairly clearly leaving a reasonable picture of a concensus.
It would be a shame to sink into a mess of noise again imo, and I think owners affected by these issues will keep tacking posts on without having followed the prior details.
Is this inevitable?
.... and I could get on with life. I don't know about others, but I find it difficult trying to make sure I have read long, detailed posts (mine included) correctly, and then work out what the poster is saying and if it agrees with other posts / knowledge.
Right now I feel we reached a point where a lot of info came together fairly clearly leaving a reasonable picture of a concensus.
It would be a shame to sink into a mess of noise again imo, and I think owners affected by these issues will keep tacking posts on without having followed the prior details.
Is this inevitable?
@wk057 said that balancing is separate to charging and my understanding is that there is no mechanism to actively charge or draw power from specific bricks.
@David99 's post covers all the issues, but unless I am misreading, the evidence seems to suggest to me that the bricks showing the highest voltage at the moment charging stops could be weak bricks. The BMS absolutely should be balancing all / most of the time (a small power dissipated over a long period of time is a larger amount of energy). I'm not certain if it actually balances at all SOC, or may be just starts at say 50% SOC.
In any case, if the state of a brick is such that it's voltage rises relatively quickly due to the brick cells not sharing charge evenly (and so you are charging some cells in bricks faster than others) you are creating an imbalance that's too big for the BMS to fix. As @David99 says, once you have a weak brick, there's not much you can do. Such bricks will reach their maximum permitted voltage before other bricks, stopping those bricks from being chargeable as there is only so much the BMS can do to help.
I belive that if @David99 left his car to stand, he would see the voltage of the highest bricks sag while the good, evenly charged bricks stay steady.
@wk057, if any of that is incorrect, please put me right so I don't go screwing the thread up!
@David99 's post covers all the issues, but unless I am misreading, the evidence seems to suggest to me that the bricks showing the highest voltage at the moment charging stops could be weak bricks. The BMS absolutely should be balancing all / most of the time (a small power dissipated over a long period of time is a larger amount of energy). I'm not certain if it actually balances at all SOC, or may be just starts at say 50% SOC.
In any case, if the state of a brick is such that it's voltage rises relatively quickly due to the brick cells not sharing charge evenly (and so you are charging some cells in bricks faster than others) you are creating an imbalance that's too big for the BMS to fix. As @David99 says, once you have a weak brick, there's not much you can do. Such bricks will reach their maximum permitted voltage before other bricks, stopping those bricks from being chargeable as there is only so much the BMS can do to help.
I belive that if @David99 left his car to stand, he would see the voltage of the highest bricks sag while the good, evenly charged bricks stay steady.
@wk057, if any of that is incorrect, please put me right so I don't go screwing the thread up!
Last edited:
Based on @wk057 saying the BMS starts to struggle once there is more than 20mv imbalance, you could be at that threshold. Intentionally trying to bring on failure is likely inadvisable but not really fair for you to nurture the battery just past wty!
Maybe one just has to let things take their own course?
With 40mV imbalance, @David99 says he is seeing noticeable reduction in range and increase in imbalance. Since he is maybe ahead of you, his experience will predict yours.
I can understand the brick voltages rising quicker if cells are not being charged evenly, but what is the equivalent explanation for the bricks that are showing the lower voltage? (at 4.175v full charge)
Last edited:
Which I am taking to be limitation of BMS, possibly further limited by elevated temperature inside the pack (or at least you can help the BMS out by keeping pack temp down) so you have to throttle charging rate.
It really would be interesting to see comparable charge rate graphs for latest packs (the 200kw capable rev g / h or whatever) to see SOC vs charge rate (at range of battery temps).
Has the balancing been beefed up?
Are newer models (3 / Y) better able to maintain more even charge as batteries degrade?
Is the only cost & technology viable option to make cells even better and maintain even tighter control over temperature?
Last edited:
Speculation:
This could be an explanation for keeping cooling fans running after charging.
The idea of running the fans (after charging stops) wouldn't really make sense purely to bring the brick voltages down or somehow stop the battery from spontaneously combusting. You could do that by stopping charging earlier. Likewise, however you discharge the battery, once your weakest brick reaches lowest permitted voltage, your battery can't deliver any energy remaining in the better bricks, so just drawing energy for the sake of it (running cooling) makes no sense.
If we accept that Tesla would be crazy to allow batteries to habitually get to a state where there was any chance of thermal runaway, then the cooling after charging could be more likely not related to the cells themselves, but the rest of the battery. And one function of the battery that had already been flagged as 'amazing how it does such a good job' is the balancing, which also needs to dissipate some heat through pretty modestly spec'd circuits (I'm guessing).
Are you basing that on the fact that brick voltages are apparently more even after use?
I would suggest that you are seeing a module or two of bricks that charged less due to weak cells(s) (and which therefore reached higher voltage first and probably signalled the battery to stop charging) discharging (at least voltage dropping first).
I believe there is no mechanism by which any one part of the battery can be loaded differently from another part except by balancing circuits.
I suggest there is no mechanism in the battery to allow / encourage different power delivery from certain modules. You draw power from the battery and it does what it can - as you might expect, being limited by the weakest link. Once a brick (I think it is a brick rather than module) reaches the minimum permitted cell voltage, there is no way to draw from only the remaining bricks that still have something to give.
When you start charging, that would mean from an stored energy / charge perspective, your pack would be already potentially quite unbalanced.
It would be like the brick's combined internal resistance is higher, even though the mix of cells in the brick are a mix of higher ( and may be lower) ir. So voltage has a bigger hysteresis between charge and discharge. (softness, sponginess, large voltage difference during charge vs discharge).
If the charge imbalance is great enough, then the week brick voltage would rise too quickly, you could exceed max charge rate for some cells in a brick, and the balancing circuit wouldn't be able to bleed off enough power to stop the weak cell prematurely reaching max permitted voltage.
End conjecture.
Last edited:
You may just have a regular old imbalanced pack which appear to become worse if you normally charge to less than 90%.
Many suggest to at least occasionally charge to 90% or higher (if higher, limit time spent at 90%) to help give the BMS enough signal to know that certain bricks need balancing. Balancing will occur at any time once the BMS knows how much energy to bleed off other bricks. You might try discharging to 10-20% and then charging to 90% a few times to see if that helps get the BMS to better balance the pack.
I'm interested what leads you to think that charging to at least 90% helps balance the battery? Tesla allow you to set your charge limit to a wide range of values. Mine is typically at 70%. If that means the bricks don't balance until higher SOC, then presumably the balance would just get worse and worse over time?
My battery seems to drift off to about 15mV difference when left standing and unplugged for a week or so. A slow charge from say 60% SOC to 80% and the voltage difference is back to 4mV.
On the idea of discharging to 10% and recharging, if the balance doesn't improve charging from 50% to 70%, why would it improve charging from say 40% to 60%? Or 20% to 90%? I expect the longer the time charging, the nearer to balance you'd get, but only if you are at least moving in the right direction.
If the car balances at all times, even when not on charge, then even without purposefully charging / discharging, the battery would balance itself. I know for sure that my car balances when charging and drifts off when left standing unplugged for some time.
Last edited:
There is evidence based on TMC commentary and Tesla recommendations (at least in the Model 3 forum) that deep cycles allow the BMS to better estimate range. Specifically 20-90+%. As far as balancing goes, Tesla has stated to numerous people (in the Model 3 forum) that balancing only occurs over 85%. I don’t know if this applies to Model S packs.
I suspect that people are being too loose with terminology. My opinion is that range predictions are not a BMS function. BMS is balancing cells, maintaing battery in good order / optimising energy storage and controlling charging.
Predicting how far the car (or perhaps more accurately finding out how much energy is actually available) will likely use data obtained from BMS but I hardly see it as a BMS function. Yes, I can see why using a large proportion of the maximum energy the battery can store, and taking the battery from high SOC to low SOC will generate data that shows how much energy is actually the battery when it's at highest permitted voltage / SOC. I can't see any real benefit of doing that though. I would rather keep my battery in good shape and not put extra strain charging and discharging to extremes.
Predicting how far the car (or perhaps more accurately finding out how much energy is actually available) will likely use data obtained from BMS but I hardly see it as a BMS function. Yes, I can see why using a large proportion of the maximum energy the battery can store, and taking the battery from high SOC to low SOC will generate data that shows how much energy is actually the battery when it's at highest permitted voltage / SOC. I can't see any real benefit of doing that though. I would rather keep my battery in good shape and not put extra strain charging and discharging to extremes.
I don't KNOW, but that doesn't make sense unless Tesla also confirmed:
You don't need to keep the battery balanced.
Or
You MUST charge above 85% and confirm your battery has balanced.
Or
M3 has completely different system of balancing such that (for example) it can selectively discharge different sections of the battery into the drive motors while driving, then I can see that MAYBE it is less important to bother balancing at lower SOC.
But even then, why would you ever want a battery to be significantly unbalanced at any SOC since that would mean you could end up having to stop driving just because one (imbalanced) section of your battery had reached lowest permitted SOC?
IMO the battery should be reasonably balanced all the the time to be best able to deliver the maximum energy from all the cells in the pack.
Last edited:
BigNick
Infamous Fat Sweaty Guy
I believe that my experience as detailed by the screenshots I posted (sorry for the size, they didn't look so big on my phone) supports @wk057's statement that balancing and charging are separate functions, as the closer to "full" I charged, the imbalance grew to 20 mV.For a moment I thought we might be done on this thread!
I don't know about others, but I find it difficult trying to make sure I have read long, detailed posts (mine included) correctly, and then work out what the poster is saying and if it agrees with other posts / knowledge.
Right now I feel we reached a point where a lot of info came together fairly clearly leaving a reasonable picture of a concensus.
It would be a shame to sink into a mess of noise again imo, and I think owners affected by these issues will keep tacking posts on without having followed the prior details.
Is this inevitable?
My comment about "balancing occurring during use" is speculation based on the displayed data in the graph. However, I don't know if that is a specific algorithm of the BMS, or just "natural" behavior of the higher voltage cells providing more energy during discharge.
Just when I think I know how things work, I realize I know a lot less than I thought.
Similar threads
