Ok, so our theory now appears to have to some credibility. I think it is unfortunate that Tesla does not balance after every charge. Over time, cells will become out of balance and this will accelerate permanent degradation.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Ok, so our theory now appears to have to some credibility. I think it is unfortunate that Tesla does not balance after every charge. Over time, cells will become out of balance and this will accelerate permanent degradation.
Why would an out of balance state accelerate permanent degradation given that most of us cycle on a very small part of total capacity? I would have assumed that the BMS would not allow the cells in an unbalanced pack to get too high or too low - hence the reduced usable range. Why would that accelerate permanent degradation in a measurable way? How far out of balance must it go before this kind of effect is an issue?
Instead, I think they need to allow for pack balancing at any arbitrary charge level. After all, a range charge isn't a 100% charge anyways, so it's already an arbitrary charge level.
Just because the BMS is programmed to prevent "bricking" of a specific cell does not shield it from permanent degradation. Granted, the cell must have a meaningfully lower voltage than other cells in the pack, but I see that as a significant possibly if one were never to range charge the car.
Yes, Tesla should allow balancing at any charge limit SOC and is something for which I have expressed support for quite some time.
Yes, Tesla should allow balancing at any charge limit SOC and is something for which I have expressed support for quite some time.
The best way to think about this is to remember that the SOC shown is the average of all the sheets. That means that as they drift apart some will drift higher and some will be lower. Without a BMS that preforms balancing, while you might use only the SOC of 55-45% some cell sheets would eventually be cycled 90-100% and 0-10%. Now a better management system would reduce the available capacity in that case to only 10% to not allow the system to over or undercharge those cells. This MAY be what some are seeing. Because this type of reduction means that some cell sheets are being held at higher and lower points, they will see accelerated wear. The amount of acceleration depends on exact Voltage levels of the sheet, and the temperature. Generally there is a cubic factor in this type of degradation. The best SOC thread has some good papers on those calculations.
I agree that they need to add back low SOC balancing, as it use to be there!
I agree that they need to add back low SOC balancing, as it use to be there!
djp
Model 3 Performance
Incorrect. On the Roadster the displayed SOC is based on the lowest brick, and charging stops when the highest brick reaches the target SOC. This is why range is reduced by an out of balance pack.
I don't buy the argument that an out of balance pack leads to faster degradation since the SOC is always constrained by the highest and lowest brick. If anything the rest of the pack is being run through more shallow cycles, which should help battery life.
Almost 23,000 miles - we range charge 1 to 3 times a week. (This week 3 times - all at home - 40 amps. 2 stops at SC's but only to add - most there was to 210, short stints).
Anyway our range charges have been 250 to 255 for as long as we remember. Varies in that area and I always schedule it so that I can see it on completion, so there was no Vamp drain before I checked. Other times we charge to about 190 - 210 all at home.
If there has been degradation, we haven't noticed (other than the software reason)
'A' Battery vintage late April '13 on the sunny slopes.
Anyway our range charges have been 250 to 255 for as long as we remember. Varies in that area and I always schedule it so that I can see it on completion, so there was no Vamp drain before I checked. Other times we charge to about 190 - 210 all at home.
If there has been degradation, we haven't noticed (other than the software reason)
'A' Battery vintage late April '13 on the sunny slopes.
djp
Model 3 Performance
Also keep in mind that there are two effects at play here:
1. Pack balancing
2. Algorithm to estimate capacity (CAC)
It's possible that the range loss that people are seeing is mostly related to the capacity estimate algorithm, not pack balancing.
On the Roadster, CAC is recalculated when you do a single drive (single key turn) from above 85% SOC (Standard charge) to below 25%, then recharge back to 85%.
We know the Model S does a similar capacity estimate, but not what the exact triggers are. It's not a bad assumption to use the Roadster levels as a best guess. For those who aren't seeing an improvement from balancing, try doing a long drive on a single key turn from >90% to <25% then recharge to 90%.
This might also explain why owners who exercise the full range of their battery have little to no range loss. The losses could have nothing to do with balancing, and only be an effect of the CAC algorithm losing track of the battery's capacity when doing shallow charges.
1. Pack balancing
2. Algorithm to estimate capacity (CAC)
It's possible that the range loss that people are seeing is mostly related to the capacity estimate algorithm, not pack balancing.
On the Roadster, CAC is recalculated when you do a single drive (single key turn) from above 85% SOC (Standard charge) to below 25%, then recharge back to 85%.
We know the Model S does a similar capacity estimate, but not what the exact triggers are. It's not a bad assumption to use the Roadster levels as a best guess. For those who aren't seeing an improvement from balancing, try doing a long drive on a single key turn from >90% to <25% then recharge to 90%.
This might also explain why owners who exercise the full range of their battery have little to no range loss. The losses could have nothing to do with balancing, and only be an effect of the CAC algorithm losing track of the battery's capacity when doing shallow charges.
While this makes sense, I believe apacheguy is saying that individual cells within each "brick" can be out of balance even though each brick is being properly managed. The bricks are as granular as the BMS is able to get, as it does not manage individual cells within each brick. This is confusing enough to give me a headache, so I just charge to what I need and trust that Tesla knows what it's doing.
djp
Model 3 Performance
The cells within the brick are wired in parallel so it's not possible for them to be out of balance. Any excess voltage on one cell would bleed over to the other cells. The BMS only needs to balance at the brick level, since those are the groups that are wired in series, and can't automatically balance relative to each other.
Is that actually true? In the Prius, measuring individual cells in a module shows different levels and the group is degraded when one cell in the module is below a certain threashold (and it appears to always be once cell that's the cause of a degraded module). That's why it's possible to recondition a Prius battery with just a few cells. Now the cells are much larger in the Prius than in the Model S which allows for a greater variation and, of course, the chemistry is entirely different. I'd suspect that the smaller the cells in a module, the less variation there will be between them.
djp
Model 3 Performance
Cells wired in parallel will act as a single cell. One individual cell that is degraded will bring down the capacity of the brick, so yes it's possible to restore the brick's capacity by replacing that cell, but that's different than balancing. All the cells in a brick are guaranteed to be at the same voltage because they're wired in parallel.
As a confirmation of the excellent points made here, I just did a full range charge on the Roadster, 237 miles, for a trip to The City (180 miles roundtrip) which I try to do on this car every 5 weeks on average. Came back with 20 miles remaining due to full winter conditions. Under better weather I stop the range charge at 220 miles. The car just turned 10k miles but these numbers are the same as when I bought it @3k miles, i.e. if I ever got a slightly higher full range charge it would quickly drop down after sitting for 10 minutes or so (phantom charge).
--
--
You're doing fairly well. Judging by your VIN I'd wager a guess that you have a B pack. Also interesting is that you had not performed a range charge in quite some time. This suggests that the pack was most likely slightly out of balance, and yet you zoomed right up to 256 on the first try. I've repeatedly clocked in at 244-245 miles with between 22-25 K on the odo. Mine still has an A...
glhs272
Unnamed plug faced villian
Decreasing rated range? Nope. @17,475miles I can still reach 209 rated miles on a range charge, same as new. I refuse to baby my pack. I charge to 90% daily and am not shy about range charging.
When I did this range charge it stayed at 209 for a while. When it finally finished, as soon as I pulled the plug, got back in and put it in drive the rated range jumped to 212
When I did this range charge it stayed at 209 for a while. When it finally finished, as soon as I pulled the plug, got back in and put it in drive the rated range jumped to 212
hans
P631
I am curious what 209 rated miles range looks like in ideal miles. Just trying to prove/disprove a theory I have.
gg_got_a_tesla
Model S: VIN 65513, Model 3: VIN 1913
Whoa, glhs272, those are remarkable numbers for a 60 kWh car at those many miles! Looks like you consume a lot of power too looking at your Wh/mile numbers.
I guess the key takeaway from your stats is the bit about "not being shy about range charging". That must be the trick. While everyone and their mother recommend not doing so often to preserve long term battery life, it's probably the only way to consistently get full rated range.
I guess the key takeaway from your stats is the bit about "not being shy about range charging". That must be the trick. While everyone and their mother recommend not doing so often to preserve long term battery life, it's probably the only way to consistently get full rated range.
glhs272
Unnamed plug faced villian
Forgot to check. I have heard that this number is more reliable indicator of battery capacity, but is otherwise useless. I did achieve these numbers after a seemingly effective balancing session the previous night (left the car all night at 100% charge). This was with a warm battery from a drive from my house to the Mauston supercharger.
Similar threads
- Question
