12% capacity loss in 9 months is "normal"

[TickTock]

After reading through this thread I am having a thought… a charger problem??? I am wondering if this could be an overlooked cause of your lower state of charge.

Do you have the QC port? Could you do a QC and then read your gids? I believe a QC completely bypasses the onboard charger.

Would someone with a bit more experience or knowledge with the communication between the BCM and charger comment about the possibility of a charging issue?

Good idea but no luck. I went to the QC at Riverside today. Blink charger said it charged me to 97% (I asked for 100%) but I still ended up with only 238 gids (85%). kWh is energy entering the battery (integral of volts*amps*delta_time). I forgot to label my starting gids (130).

 

[DaveEV]

Good idea but no luck. I went to the QC at Riverside today. Blink charger said it charged me to 97% (I asked for 100%) but I still ended up with only 238 gids (85%). kWh is energy entering the battery (integral of volts*amps*delta_time). I forgot to label my starting gids (130).

Interesting chart. I don't recall seeing any other QC charges nearly immediately hitting 392V. What did the charger say?

At 130 GIDs that should be around 5 bars? Resting voltage was around 370-375V before starting to charge?

The battery pack certainly looks full from the V/A plot.

It looks to me that your pack has higher internal resistance than expected. I would very much like to see a similar plot of data from a "normal" car along with charges from a lower SOC.

 

[Stoaty]

As the 80% and 100% are voltage based, the degradation can be extrapolated, to some extent, to the rest of the SOC.

Yes, but why extrapolate when we know the number of Gids at a full charge? That would be a lot more accurate.

 

[gbarry42]

The bars are adjusted so that a battery with 50% degradation still shows 12 charge bar, but each represents a lower energy.

I thought it was mentioned several times that bars would stop lighting up, so essentially leave you with 6 bars with 50% degradation...

Okay, some things to clear up.

There are two sets of "bars". The long ones show your SOC and all 12 should light when the battery has been charged and can't be charged further. The little segments to the right of those are the ones that show battery health, and will disappear as the battery ages. So, you can see, that if you just make a statement about "bars", half the people will agree and the other half will get confused. Unless it's obvious from context. But there will always be a few that think you're talking about the "other" ones.

The thing I was saying, that vegastar disagreed with, was that the SOC bars wouldn't go to 12 if the battery didn't get charged all the way. But, again, I don't know what's happening in this case, and so we'll get a different result depending on what the cause is. If the charger stops early on a good battery, we should see less than 12 bars. But if the car is reading the charge state incorrectly in the first place, then it can show the bars incorrectly as well. And then, if the battery really is bad, 12 bars would mean the battery really is "full", though only as full as it can be in its degraded state. There aren't any right answers here...yet. All we can do is match the theories to the behaviors we are seeing.

 

[surfingslovak]

As the 80% and 100% are voltage based, the degradation can be extrapolated, to some extent, to the rest of the SOC.

Yes, but why extrapolate when we know the number of Gids at a full charge? That would be a lot more accurate.

Most owners don't have Gary's meter. Perhaps there is some simple metric, which would not require running the car from full to turtle. If recording the time and energy required to charge from 80% to full was sufficient and worked reliably, I could see some folks use that approach.

 

[klapauzius]

I tried the 80% to 100% with the blink, but it stopped after exactly 65 minutes.
I let the pack cool down a bit and started again and it seems that it would put some extra juice in. I ran for over 25 minutes now, bringing the total charge time up to 1:31 h.

So that method might cause of lot of initial 'capacity' anxiety and a flood unreasonable complaints ?

First run from 80 to "100%" : 65 minutes and 3.85 kWh from the wall
2nd run from "100%" to "really 100%" 26 minutes and 0.89 kWh from the wall.
total = 4.74 kWH from the wall.
I guess that means 4.22 went into the 3.3 kW onboard charger and if this was all linear, that would come out
to a total pack capacity of 21.13 kWH.

Anyway, I am waiting for ingineers leafscan and hopefully then this can turn into a more exact science.

Just for completeness sake, I almost never charge to 100%, so presumably the pack might have been unbalanced to begin with, necessitating the 2nd 26 minute run.

 

[surfingslovak]

So that method might cause of lot of initial 'capacity' anxiety and a flood unreasonable complaints.

Good point! Thank you for trying it out...

 

[planet4ever]

After reading through this thread I am having a thought… a charger problem??? I am wondering if this could be an overlooked cause of your lower state of charge. Do you have the QC port? Could you do a QC and then read your gids? I believe a QC completely bypasses the onboard charger.

Good idea but no luck. I went to the QC at Riverside today. Blink charger said it charged me to 97% (I asked for 100%) but I still ended up with only 238 gids (85%).

I'm out of my depth here, but let me try anyway. I believe jamesanne's statement is not precisely true. For QC the power certainly doesn't go through the on-board charger, but the Service Manual shows that control of the charging does still go through it. i.e. it is apparently the on-board charger which converts data from the LBC (aka BCM) into CHAdeMo commands to the external charger. So if the charger is misinterpreting battery data, or getting bad data, for L2, it could also be doing the same for QC.

The thing I was saying, that vegastar disagreed with, was that the SOC bars wouldn't go to 12 if the battery didn't get charged all the way. But, again, I don't know what's happening in this case, and so we'll get a different result depending on what the cause is. If the charger stops early on a good battery, we should see less than 12 bars. But if the car is reading the charge state incorrectly in the first place, then it can show the bars incorrectly as well. And then, if the battery really is bad, 12 bars would mean the battery really is "full", though only as full as it can be in its degraded state.

These are key points, and well put. The SOC bars won't (in theory) go to 12 until the battery is almost completely topped off. If it is not in fact that close to full, and TickTock is seeing 12 bars, then whatever is misleading the charger is also misleading the Combination Meter.

This is a crazy idea, but could TickTock's car have an incorrect mapping table down at the LBC or VCM level? It would seem that it would have to be something that affects Phil's real SOC without affecting Gary's GIDs. One hour charging from 80% to full sounds extremely fishy to me.

Ray

P.S. Incidentally, to underscore what gbarry42 said in response to IBELEAF, the Service Manual has three full pages of glass-half-full pictures to emphasize that all 12 available charge bars can be lit up even when the battery is 50% degraded.

 

[LEAFfan]

Good idea but no luck. I went to the QC at Riverside today. Blink charger said it charged me to 97% (I asked for 100%) but I still ended up with only 238 gids (85%). kWh is energy entering the battery (integral of volts*amps*delta_time). I forgot to label my starting gids (130).

Riverview Toyota? I went to ECOtality today for a QC, and after it said 98% (started with 10% SoC), I had twelve full bars, but only 87% SoC (usually get 94%). This is only the second time it has showed below 90% after a '100%' QC. On the way home though, I broke an all time record and recorded 14 miles (city streets) on the 12th bar. My gauge used to show 5-8% above the CW's bar reading.

 

[TonyWilliams]

Good idea but no luck. I went to the QC at Riverside today. Blink charger said it charged me to 97% (I asked for 100%) but I still ended up with only 238 gids (85%). kWh is energy entering the battery (integral of volts*amps*delta_time). I forgot to label my starting gids (130).

This test is flawed, as it has been observed several times (myself included) that a 100% QC doesn't get to 100% Gid (my last one was 244, and I drove the car 100 miles back home on the freeway).

It appears that the Gids can't keep up to a full charge on a quick charge.

 

[vegastar]

The thing I was saying, that vegastar disagreed with, was that the SOC bars wouldn't go to 12 if the battery didn't get charged all the way. But, again, I don't know what's happening in this case, and so we'll get a different result depending on what the cause is. If the charger stops early on a good battery, we should see less than 12 bars. But if the car is reading the charge state incorrectly in the first place, then it can show the bars incorrectly as well. And then, if the battery really is bad, 12 bars would mean the battery really is "full", though only as full as it can be in its degraded state.

These are key points, and well put. The SOC bars won't (in theory) go to 12 until the battery is almost completely topped off. If it is not in fact that close to full, and TickTock is seeing 12 bars, then whatever is misleading the charger is also misleading the Combination Meter.

This is a crazy idea, but could TickTock's car have an incorrect mapping table down at the LBC or VCM level? It would seem that it would have to be something that affects Phil's real SOC without affecting Gary's GIDs. One hour charging from 80% to full sounds extremely fishy to me.

Ray

P.S. Incidentally, to underscore what gbarry42 said in response to IBELEAF, the Service Manual has three full pages of glass-half-full pictures to emphasize that all 12 available charge bars can be lit up even when the battery is 50% degraded.

AFAIK TickTock's car scales the SOC bars accordingly to a lower capacity gids count. The transition from 12 bars to 11 bars happens at a lower gid count than a normal car, and so on. The BMS knows it has a lower capacity battery and scales the SOC bars accordingly. The only way I can think that the charger would not fill the battery completely would be a faulty voltage sensor, so 394V would be in reality 390V or less, and of course the battery would not charge to full. But I think that is highly unlikely.

What is strange is that he still has all the 12 capacity bars, even though his battery has almost 20% less capacity. It could be his battery has less capacity from the beginning and the capacity bars are relative to the new capacity of the battery and not an absolute capacity. Of course this would be unaceptable, as the car, when new, must be able to do 100 miles in the LA4 cicle as advertised.

 

[DaveEV]

Just for completeness sake, I almost never charge to 100%, so presumably the pack might have been unbalanced to begin with, necessitating the 2nd 26 minute run.

Quite possible. How long did you wait before finishing off the charge?

even though his battery has almost 20% less capacity.

It's pretty clear that his capacity loss (if going by GID count on a full L2 charge) is less than 10% and not 20%.

TickTock - out of curiosity - have you tried any full charges on 120V?

 

[TickTock]

Continuing on the charger angle, now that it getting noticed, how long does it take to charge from 80 percent to 100? Anything shorter than 90 minutes would be suspect.

Last night it took 65 minutes to bring an 80% charge to 100%. Carwings predicted 90.

I did the 80 to 100 test again last night. This time it took 73 minutes and took my gids from 205 to 246. Wall meter said I used 4.25kW and the evcan log says 3.85kW of that went into the battery. However, I am somewhat discouraged that the last two 100% charges only got me to 246 (even the one before the QC).

my charge log

 

[Stoaty]

However, I am somewhat discouraged that the last two 100% charges only got me to 246 (even the one before the QC).

Yes, that would be discouraging. It works out to a 12.5% loss of capacity.

 

[TickTock]

This sounds like material for an article on autobloggreen or by Chelsea Sexton. I think your best weapon against Nissan is the threat of negative publicity. I would even go so far as to alert Nissan USA (e.g., Mark Perry) that this is what you plan to do if you don't get satisfaction. Your car is clearly an outlier, and Nissan is trying to keep from doing any kind of repair of marginal battery packs, even when the problem occurs in the first year. While you don't want to hurt the Leaf, facts is facts. Nissan needs to put up or shut up.

Any suggestions on how to escalate? (i.e. Mark Perry's contact info?) I don't think I will get anywhere through the dealer...

 

[EdmondLeaf]

I tried the 80% to 100% with the blink, but it stopped after exactly 65 minutes.
I let the pack cool down a bit and started again and it seems that it would put some extra juice in. I ran for over 25 minutes now, bringing the total charge time up to 1:31 h.

So that method might cause of lot of initial 'capacity' anxiety and a flood unreasonable complaints ?

First run from 80 to "100%" : 65 minutes and 3.85 kWh from the wall
2nd run from "100%" to "really 100%" 26 minutes and 0.89 kWh from the wall.
total = 4.74 kWH from the wall.
I guess that means 4.22 went into the 3.3 kW onboard charger and if this was all linear, that would come out
to a total pack capacity of 21.13 kWH.

Anyway, I am waiting for ingineers leafscan and hopefully then this can turn into a more exact science.

Just for completeness sake, I almost never charge to 100%, so presumably the pack might have been unbalanced to begin with, necessitating the 2nd 26 minute run.

I do not know if we getting anywhere with this 80 to 100% comparison. For my car it was first 100% ever, 76 min and 5.03 kWH from the wall (smart meter), on those 2 bars I did today 24.3 miles average speed 40mph and 6.4mkWh. I do not have Gary's meter.

 

[edatoakrun]

TickTock,

Please reconsider the suggestion made by myself, and several similar ones by others, many pages of comments back:

IMO, the most accurate and comprehensive calculation of battery capacity, can (presently) be accomplished by range tests, from 100% charge, to the lowest level your situation allows...

http://www.mynissanleaf.com/viewtopic.php?f=31&t=8331&start=40" onclick="window.open(this.href);return false;

Did you post results of a capacity test, that I missed?

 

[TickTock]

TickTock,

Please reconsider the suggestion made by myself, and several similar ones by others, many pages of comments back:

IMO, the most accurate and comprehensive calculation of battery capacity, can (presently) be accomplished by range tests, from 100% charge, to the lowest level your situation allows...

http://www.mynissanleaf.com/viewtopic.php?f=31&t=8331&start=40" onclick="window.open(this.href);return false;

Did you post results of a capacity test, that I missed?

I disagree with this. Such a test adds more random variables to the measurement (wind, tire pressure, grade, condition of the road, driving style, etc.). The ultimate means to characterize capacity (until we get the LeafScan) is:

• 1) to use Gary's box and log a charge from turtle to 100% and integrate the power going into the battery.
  2) Next best is to multiply gids times .08. If you take the QC plot I posted above, and plot (gids-130)*80, the curve follows the energy entering the battery curve almost exactly. I've confirmed and reconfirmed over and over. A gid really is a pretty good measure of available charge in the battery.
  3) Next best is to install a wall-meter and measure the power from the grid from turtle to 100% (only random variable there is the charging efficiency).
  4) Next best is to measure using the utility meter (means turning off power to all the other circuit breakers in the house for the duration of the test and reading the meter before and after).

I've done 1-3. Problem is I only have a handful of quantified datapoints from other Leafs.

Now, maybe if we did a side-by-side test (i.e. driving both leafs in the same lane on CC at the same time but far enough apart that neither drafts the other) from 100% to turtle (thus eliminating the GOM from the equation) that would be an interesting datapoint.

Since everyone has at least a utility meter, I've been trying to get at least that datapoint for comparison from as many people who care to particepate:

Collecting data:Off-the-wall power for turtle to 100% charge

 

[EVDRIVER]

This sounds like material for an article on autobloggreen or by Chelsea Sexton. I think your best weapon against Nissan is the threat of negative publicity. I would even go so far as to alert Nissan USA (e.g., Mark Perry) that this is what you plan to do if you don't get satisfaction. Your car is clearly an outlier, and Nissan is trying to keep from doing any kind of repair of marginal battery packs, even when the problem occurs in the first year. While you don't want to hurt the Leaf, facts is facts. Nissan needs to put up or shut up.

Any suggestions on how to escalate? (i.e. Mark Perry's contact info?) I don't think I will get anywhere through the dealer...

As I mentioned earlier. Contact the Nissan regional rep. Call the dealer and ask for that persons name and number. I have done this with other auto makers with very good results when I had a paint issue and went to arbitration and received a nice settlement. Unfortunately you have no warranty coverage on this but escalating it is a good idea, I would not waste time with dealers.

 

[surfingslovak]

I've done 1-3. Problem is I only have a handful of quantified datapoints from other Leafs.

Luke, if you wanted to do a range test, you will see some variability, that's for sure. I have done it twice, in similar weather conditions and using the same route. The first result returned 21 kWh, and the second result returned 20.3 kWh of usable battery capacity. We have a lot of smart people here, and this is just my personal opinion, but considering that you are using a home-built device to look for a CAN bus value someone has recommended, and project loss of use from that, you might not find many open ears. I agree that Nissan should investigate this more thoroughly, and use adequate test instruments, but please look at this from the other side as well.