Are we sure that capacity is actually being lost?

OK. I don't want to insult anybody, but I am getting a little confused when I read posts from the likes of Ingineer and garygid whom I consider to be well informed, that imply (to me) the following:

1. 100% charge with the timer will not balance the battery cells. - I say this from reading statements about needing to leave the car plugged in for a few hours after reaching 100%. Since the timer will stop at 100%, that means that all charging is stopped with the timer and leaving the car plugged in 3hrs afterwards doesn't do anything. So only a manual (timer override) charge would rebalance the cells.

2. 100% charge isn't the full capacity when it first completes. - I gather this from Ingineer's 11:54pm post http://www.mynissanleaf.com/viewtopic.php?f=30&t=8802&start=530. Reading this tells me that there is a potential 10-15% SoC gained during that additional rebalance period after the charge has first stopped. So again using a timer would means that even if your cells were fairly well balanced, not getting 12bars isn't necessarily a true loss of capacity.

Could it be that the 11bar numbers at 100% that are being reported are really just an issue with the charger firmware not performing rebalances and the display not truly a SoC or actual capacity? Even with a GIDMeter, since the firmware may be stopping the charge before the battery is 'full', the GID reading can be deceptive. It would seem that without a LEAFSCAN we're kind of guessing.

If one is getting the 11bars after a 100% timer charge, there are several reasons that this could be happening (temperature, out of balance cells, etc), and we might be worrying unnecessarily. However if one has 11bars after two or three consecutive manual (not timer) 100% charges where the LEAF was plugged for 3hrs after it reached 100%, then I guess there is reason that people are freaking out...

Please don't flame me, I guess I'm just hoping that there is a simple answer to this issue...

>>>> not getting 12bars isn't necessarily a true loss of capacity. <<<<

I'm too lazy to go through your whole post, but this seems the MOST wrong.

Loss of capacity is just that. That means the range / autonomy is reduced. Keeping 12 battery capacity bars doesn't mean that you have 100% capacity. Does that make sense to you?

There is not 10-15% regained by balancing, at least I've never witnessed a battery so far out of balance that this happened.

There are a lot of perfectionists that are bothered by the possibility that their packs could be "out of balance", but all this will normally affect is a small amount of usable capacity until it is corrected. Balancing is corrected when it needs to be by the battery ECU, as it will never allow it to get far enough out of balance to cause any significant problems. (unless of course one or more defective cells prevent it from achieving proper balancing.)

I have seen balancing occur on timer charges, so stating that balancing will not occur is false.

The assumption is that because charging can re-start many hours after it "finishes" to top off a balance operation, that preventing this by timer or power disconnection will somehow prevent balancing from occurring. All it will do is mean you don't have as much capacity as you would if you hadn't prevented the extra charging.

The balancing system works only by discharge, or shunting. It doesn't ever add charge.

-Phil

padamson1 said:

1. 100% charge with the timer will not balance the battery cells. - I say this from reading statements about needing to leave the car plugged in for a few hours after reaching 100%. Since the timer will stop at 100%, that means that all charging is stopped with the timer and leaving the car plugged in 3hrs afterwards doesn't do anything. So only a manual (timer override) charge would rebalance the cells.

Click to expand...

What timer are you talking about? If you mean a timer in an EVSE then, yes, once it cuts off the battery can't rebalance, or at least can't top off after it does. But I can't think of any reason why that would be a limit if you are using the timer in the car.

padamson1 said:

2. 100% charge isn't the full capacity when it first completes. - I gather this from Ingineer's 11:54pm post http://www.mynissanleaf.com/viewtopic.php?f=30&t=8802&start=530. Reading this tells me that there is a potential 10-15% SoC gained during that additional rebalance period after the charge has first stopped.

Click to expand...

I'm quite certain Ingineer had a typo in the post you referenced. He said "A charge to 100% stops at 80% SoC (real SoC, not what Gidmeters report), whereas a "full" charge usually is stopped from 94-95% SoC." That first 100% was supposed to be 80%. He was trying to say that a so-called "80%" charge controlled by the car's timer is closer to "full" than you would expect, and is more than 80% of the number of Gids you get if you charge to full.

padamson1 said:

Could it be that the 11bar numbers at 100% that are being reported are really just an issue with the charger firmware not performing rebalances and the display not truly a SoC or actual capacity? Even with a GIDMeter, since the firmware may be stopping the charge before the battery is 'full', the GID reading can be deceptive.

Click to expand...

This almost sounds as if you are confusing the two separate gauges. The 11-bar concerns are with the capacity gauge, the 12 short bars (with the last two red) at the far right. 11-bar capacity folks still see 12 available charge bars, the longer blue and white ones wrapped around the GuessOMeter, when they have charged their battery fully.

Ray

planet4ever said:

I'm quite certain Ingineer had a typo in the post you referenced. He said "A charge to 100% stops at 80% SoC (real SoC, not what Gidmeters report), whereas a "full" charge usually is stopped from 94-95% SoC." That first 100% was supposed to be 80%. He was trying to say that a so-called "80%" charge controlled by the car's timer is closer to "full" than you would expect, and is more than 80% of the number of Gids you get if you charge to full.

Click to expand...

Thanks Ray, I meant to say "A charge set to 80% stops at 80% SoC". Since full is typically not more than 95% true SoC, the difference is only ~15% rather than the 20% expected.

-Phil

Ingineer said:

planet4ever said:

I'm quite certain Ingineer had a typo in the post you referenced. He said "A charge to 100% stops at 80% SoC (real SoC, not what Gidmeters report), whereas a "full" charge usually is stopped from 94-95% SoC." That first 100% was supposed to be 80%. He was trying to say that a so-called "80%" charge controlled by the car's timer is closer to "full" than you would expect, and is more than 80% of the number of Gids you get if you charge to full.

Click to expand...

Thanks Ray, I meant to say "A charge set to 80% stops at 80% SoC". Since full is typically not more than 95% true SoC, the difference is only ~15% rather than the 20% expected.

-Phil

Click to expand...

Thanx planet4ever. That typo by Ingineer was in fact what really got me confused ('100% charge stops at 80%'). I guess I should just responded directly to that post, but I still got good clarifications for my other confusion too (as well as losing any credit on the subject :roll: ). For instance I wasn't aware inner gauge was at 12 and outer at 11
(http://www.mynissanleaf.com/wiki/in...loss_in_16k_miles_7165397296_41cfcf8d33_b.jpg) when people reported the capacity loss, I thought both were at 11bars.

Maybe I would've been clearer if I just said "It appears that the ECU firmware could stop 100% charge timer before the battery was really at full capacity (because of temperature, cell imbalance, etc), whereas a charge with timer override will run for a few hours past the first shutdown of the onboard charger (if it remains connected) allowing the battery to be truly full. Could this issue just be a firmware bug when the onboard timer is used and not actual capacity loss?"

Thanx everybody for straightening me out.

planet4ever said:

What timer are you talking about? If you mean a timer in an EVSE then, yes, once it cuts off the battery can't rebalance, or at least can't top off after it does. But I can't think of any reason why that would be a limit if you are using the timer in the car.

Click to expand...

Actually, he's probably correct. If you're using a regular start time/stop time timer, the car won't restart charging unless you either hit the start time again, or reinsert the plug while inside the start/stop interval. Which would explain why I've never observed it on my car. Of course, as Ingineer pointed out, whether the car actually tops itself off or not, any balancing should occur anyway.

davewill said:

planet4ever said:

What timer are you talking about? If you mean a timer in an EVSE then, yes, once it cuts off the battery can't rebalance, or at least can't top off after it does. But I can't think of any reason why that would be a limit if you are using the timer in the car.

Click to expand...

Actually, he's probably correct. If you're using a regular start time/stop time timer, the car won't restart charging unless you either hit the start time again, or reinsert the plug while inside the start/stop interval. Which would explain why I've never observed it on my car. Of course, as Ingineer pointed out, whether the car actually tops itself off or not, any balancing should occur anyway.

Click to expand...

And regardless of the top off charge, it needs the TIME post 100% charge. That top off charge just signals the end of the process (well, actually, I've seen it happen more than once after a 100% charge).

TonyWilliams said:

And regardless of the top off charge, it needs the TIME post 100% charge. That top off charge just signals the end of the process (well, actually, I've seen it happen more than once after a 100% charge).

Click to expand...

Yep - charging to 100% on L2 and then immediately driving off doesn't allow for much time for cells to equalize.

One rudimentary method to encourage balancing would be to charge to 100% on L1 rather than L2 - the slow charge will allow more time to discharge any high voltage cells before charging stops.

padamson1 said:

Are we sure that capacity is actually being lost?

Click to expand...

Yes I am certain my 13 month old LEAF with 17,000+ miles will not go as far as it used to.
Since new some charging days are better than others. I assume the better days are after the rebalance.

It's simple, if you are trying to do a capacity test, whether via reading full Gids or by attempting a range test, you must have a full charge.

Since "full" is highly variable depending on temperature and how balanced your pack is, it's NOT ACCURATE to do one "full" charge on a pack that routinely gets 80% charges and use that as a benchmark! (especially if attempted on a hot pack!)

I usually charge to 80%, but on occasion, I need more range, so I'll bump it to "full" right before my trip. Rarely does this give me close to 280 Gids!

Now if I drive it, then do another full charge or 2, I start seeing the "full" be around my normal 280. I'm in the mild San Francisco Bay Area, so I never really have to worry about temperature, but if you were in Phoenix and your pack is hot during the full charge, it's going to stop prematurely! (Cell Dv/Dt)

If you are trying to assess capacity in this way, then you absolutely need to allow a complete balance cycle (or 2) and it needs to be done in cooler ambient.

Hopefully when I get LEAFSCAN into everyone's hands we can put these capacity worries to rest, as everyone will have easy access to the real capacity figures from the battery ECU, as well as actual pack temperatures.

-Phil

So I guess here lies the question:

1. It's clear there's temperature affects GID reads - higher temps = lower GID readings.
2. In theory higher temperature should mean that the battery can hold more energy.
3. People in hot climates with low 100% GID readings appear to show the same voltage (~393V) when charged to 100%
4. Each GID is supposed to represent 80 Wh of capacity.

2, 3 and 4 seem to contradict each other.

If the BMS is restricting SOC due to high pack temperatures, one would expect 100% voltage to be lower than normal, no?

drees said:

If the BMS is restricting SOC due to high pack temperatures, one would expect 100% voltage to be lower than normal, no?

Click to expand...

Yes, absolutely. I've been watching these figures as well. The real question to me is, how much of a difference do 10 Gids make in terms of voltage? I haven't measured it, but I believe that it's likely around 20 mV on pack level near full charge. I can confirm however, at least from the small data sample I pulled from the forum, that there is some correlation between Gids and temperature. It's difficult to track though, especially since we can't monitor pack temperature without Leafscan.

drees said:

1. It's clear there's temperature affects GID reads - higher temps = lower GID readings.
2. In theory higher temperature should mean that the battery can hold more energy.
3. People in hot climates with low 100% GID readings appear to show the same voltage (~393V) when charged to 100%
4. Each GID is supposed to represent 80 Wh of capacity.

2, 3 and 4 seem to contradict each other

Click to expand...

Since gids are simply watt-hours, (when multiplied by 80) so it isn't that Gids are "being lost", it's that the BMS is simply halting charge SOONER. (The top gid count never goes as high as it was before)

-Phil

Ingineer said:

Since gids are simply watt-hours, (when multiplied by 80) so it isn't that Gids are "being lost", it's that the BMS is simply halting charge SOONER. (The top gid count never goes as high as it was before)

-Phil

Click to expand...

Question - How or when does the Gids get "adjusted" for any drift? For example, if gids are just measured watt-hours going in and out, then the number could eventually drift off of reality. There are probably several ways of taking measurements and correcting this. It could do it after a 100% charge and measure the open circuit voltage to determine the state of charge. If this were true then the folks who have seen a capacity loss would still keep seeing 281 gid for full charge. It could also be done on every charge by measuring the open circuit voltage at the start of the charge (just prior) and then just counting up until the charge ends. This scenario probably more closely matches what people are seeing, but I was wondering if you knew for sure how it was being handled. This second method would also provide more accurate information if the start of the charge were delayed longer after the end of a driving event. OCV would have "settled" given more time after use.

Second Question - How is capacity determined? Again there are several ways this could be performed. One way would be to see what the OCV is at the start of the drive. Measure the amp-hours removed from the pack and then at the end of the drive measure the OCV. Compare the capacity between the two OCV values to the amp-hours removed and you can see if you are using fewer amp-hours from between the two points than from a table that contains data from a "perfect" battery. Do you know for sure how it is being done?

If charging gets more efficient when the battery is hot, we might see lower GID counts while the usable capacity remains the same. This might explain how I got 288 gids on a QC session in mid-winter. Colder battery => higher internal resistance => more charging loss => more gids needed to charge to 100%...

Ingineer said:

Since gids are simply watt-hours, (when multiplied by 80) so it isn't that Gids are "being lost", it's that the BMS is simply halting charge SOONER. (The top gid count never goes as high as it was before)

Click to expand...

Right - but my point is that if BMS is halting charger sooner, there should be a noticeable reduction in battery pack voltage that goes along with the lower GID count, but so far the data hasn't shown that...

drees said:

Ingineer said:

Since gids are simply watt-hours, (when multiplied by 80) so it isn't that Gids are "being lost", it's that the BMS is simply halting charge SOONER. (The top gid count never goes as high as it was before)

Click to expand...

Right - but my point is that if BMS is halting charger sooner, there should be a noticeable reduction in battery pack voltage that goes along with the lower GID count, but so far the data hasn't shown that...

Click to expand...

read my possible scenario two posts up. Adjusting gid to the OCV just prior to charging and then counting up until you reach full (determined by voltage and current), would give you less gid if your capacity was diminished. I'm not sure if this is how it is done, just asking Phil in case he knows for sure how it is calculated.

drees said:

Ingineer said:

Since gids are simply watt-hours, (when multiplied by 80) so it isn't that Gids are "being lost", it's that the BMS is simply halting charge SOONER. (The top gid count never goes as high as it was before)

Click to expand...

Right - but my point is that if BMS is halting charger sooner, there should be a noticeable reduction in battery pack voltage that goes along with the lower GID count, but so far the data hasn't shown that...

Click to expand...

Not necessarily, as there is a voltage temperature coefficient. This means that the OCV/SoC varies depending on temperature. So temperature must be taken into account when attempting to calculate SoC from OCV. There are numerous factors, one is that the battery ECU wants to reduce stress to an already hot pack, so even though it could charge to the same SoC, (this would show up as a higher voltage) it does not.

There's also an internal resistance temperature coefficient to take into consideration.

-Phil

palmermd said:

Read my possible scenario two posts up. Adjusting gid to the OCV just prior to charging and then counting up until you reach full (determined by voltage and current), would give you less gid if your capacity was diminished. I'm not sure if this is how it is done, just asking Phil in case he knows for sure how it is calculated.

Click to expand...

Another possibility is that we are in fact getting lower pack voltage, by about 0.5V. This is at least the resolution the display on Gary's meter can provide. I believe that this delta would cover a substantial difference in Gids.