[help] Rapid battery discharge, "limited motor power", and sporadic power output at temp. < 30F and > 40% motor power.

[SageBrush]

Re. the suggestions to try and raise codes when at a lower SOC, can someone explain why it might throw codes at low SOC, but not at e.g. 75% SOC even though the underlying cause is supposedly the same?

The problem is differences in capacity between the cells, resulting in differences in voltage as they discharge. When the cells have lots of charge, those differing voltages are still close together. As the bad cells near empty, though, while the good cells stay well charged, the voltage differences increase, because discharged cells have substantially lower voltages than charged ones. Below 20% charge the imbalances start to near the trigger point for the battery management system (BMS) to sound the alarm (DTCs) and act to protect the pack from permanent damage by reducing power first, then going to Turtle mode, then shutting the car down.

To add to this a bit --

Cell voltages drop as load is applied
Imagine one cell at 3.2 volts and another at 2.8 volts.
A load is applied that drops each cell by 0.4 volts. The healthy cell drops to 2.8v -- no problem
The bad cell drops to 2.4V -- TURTLE

If the pack was well charged the healthy cell would be at 4.1 volts and the unhealthy cell at perhaps 3.9 volts. The load would drop the cell voltages to e.g 3.7 and 3.5 volts, respectively. Neither voltage drop would present a problem.

 

[GerryAZ]

Just to be clear, turtle mode (which Nissan calls power limited mode) will not set DTC's unless there is another issue. Power limited mode is caused by the voltage of the lowest cell reaching the low voltage threshold. In your case, turtle mode cleared once you stopped because the voltage recovered after the load was reduced. The state of charge (SOC) just needs to be above the very low battery warning (VLBW) level to clear the turtle indicator and allow the car to be driven normally. The double circles becoming single circles happens just before turtle mode--there is normally not much time after the first circle becomes single before power limited mode (which we call turtle mode on the forum because the amber indicator looks like a turtle). Even discharging to shutdown (when the main contactor opens to isolate the traction battery) will normally not set DTC's unless there are other issues. Obviously, the car must be charged above VLBW after shutdown before it can be driven if the battery is truly discharged. In your case, I don't know if the car would recover from shutdown by just turning it off and back on because the SOC is still at about 3 bars, but suspect it would recover. This means that having a dealer tech or other representative see it happen in a test drive and formally document it while the car is still under warranty is critical. I suggest that you limp back to the dealer in turtle mode and then floor the accelerator to cause shutdown to give more emphasis to the safety aspect.

 

[DaveinOlyWA]

Re. the suggestions to try and raise codes when at a lower SOC, can someone explain why it might throw codes at low SOC, but not at e.g. 75% SOC even though the underlying cause is supposedly the same?

The problem is differences in capacity between the cells, resulting in differences in voltage as they discharge. When the cells have lots of charge, those differing voltages are still close together. As the bad cells near empty, though, while the good cells stay well charged, the voltage differences increase, because discharged cells have substantially lower voltages than charged ones. Below 20% charge the imbalances start to near the trigger point for the battery management system (BMS) to sound the alarm (DTCs) and act to protect the pack from permanent damage by reducing power first, then going to Turtle mode, then shutting the car down.

To add to this a bit --

Cell voltages drop as load is applied
Imagine one cell at 3.2 volts and another at 2.8 volts.
A load is applied that drops each cell by 0.4 volts. The healthy cell drops to 2.8v -- no problem
The bad cell drops to 2.4V -- TURTLE

If the pack was well charged the healthy cell would be at 4.1 volts and the unhealthy cell at perhaps 3.9 volts. The load would drop the cell voltages to e.g 3.7 and 3.5 volts, respectively. Neither voltage drop would present a problem.

Mostly right. A bad cell is bad because its degradation is greater than the pack. So it charges faster, discharges faster because its essentially a smaller cell now.

When charging, full is full and its voltage controlled so the bad cell will hit 4.1 volts faster than the rest of the pack which means you now have unusable capacity because the good cells cannot be fully charged which is the key difference from the degradation warranty.

Same happens when discharging except in reverse so the bad cell hits cutoff voltage quickly while the good cells still have something to give so more unusable capacity. This is the reason you see the reduced power mode but as soon as you let off the power, the good cells brings the bad cell back into balance. This means full power returns and SOC goes up.

As far as triggering Turtle at 75% SOC? Its possible with enough power demand. Eventually that is where that cell will be. Its basically degrading at a MUCH greater rate than the rest of the pack. The bad cell is running from 100% to cutoff several times while the good cells are running from probably 80% to 40% and that range will shrink as the bad cell continues to degrade.

Either way, LEAF Spy will immediately reveal the cell. Get Pro and you will have logs of each cell pair voltage measurement in 10 second increments (actually "appears" to be around 9.something seconds...)

 

[LeftieBiker]

Exactly. Voltage sag under load is more associated with lead-acid batteries, but lithium batteries experience it as well. ALL batteries currently in use do, as best I can tell. In the electric bicycle and scooter forums we see this kind of thing all the time, because of the uneven quality of the cells used to build those smaller packs.

Gerry slipped. I thought that Turtle mode would set at least non-crippling DTCs. My bad. So the OP needs to get some serious voltage sag in the bad cell(s) to trigger a DTC for high cell voltage imbalance. The best way to do that is still to apply higher loads with low SOC.

 

[GerryAZ]

To give an idea of what the cell voltage display looks like with weak cells, here are a couple LEAF Spy screenshots from my 2015 a few days before I traded for the 2019.

 

[LeftieBiker]

Hang in there ! We all fully expect you to get a 40 kWh battery by the time the fat lady sings.

This raises an interesting question. If Nissan has any 24kwh "Lizard" packs left in stock then that is what the OP will get - not a 40kwh pack, which Nissan is now only using to replace the 30kwh "Lettuce Pack" when they are defective or fail. But what if they don't? So far I don't think that Nissan has done any 24kwh to 40kwh Leaf pack upgrades. We all know that it's possible, as the 2016-2017 Leaf is otherwise the same car as the 2015 24kwh Leaf. It will be interesting to find out...

 

[SageBrush]

Hang in there ! We all fully expect you to get a 40 kWh battery by the time the fat lady sings.

This raises an interesting question. If Nissan has any 24kwh "Lizard" packs left in stock then that is what the OP will get - not a 40kwh pack, which Nissan is now only using to replace the 30kwh "Lettuce Pack" when they are defective or fail. But what if they don't? So far I don't think that Nissan has done any 24kwh to 40kwh Leaf pack upgrades. We all know that it's possible, as the 2016-2017 Leaf is otherwise the same car as the 2015 24kwh Leaf. It will be interesting to find out...

^^ This is right. I was thinking of bad 30 kWh packs

 

[SageBrush]

To give an idea of what the cell voltage display looks like with weak cells, here are a couple LEAF Spy screenshots

Did DTCs log ?

It sounds like the specific conditions to get a DTC from a weak cell is still not well understood. Are we trying to reach a threshold delta voltage between cells ? If that is the case, perhaps OP should find the pack state where the weak cell(s) are past the knee in the cell voltage discharge curve but the rest of the pack has not, and can still take a load without a marked voltage drop. Then an applied large load would most affect the weak cell in terms of voltage drop.

 

[Dala]

This raises an interesting question. If Nissan has any 24kwh "Lizard" packs left in stock then that is what the OP will get - not a 40kwh pack, which Nissan is now only using to replace the 30kwh "Lettuce Pack" when they are defective or fail. But what if they don't? So far I don't think that Nissan has done any 24kwh to 40kwh Leaf pack upgrades. We all know that it's possible, as the 2016-2017 Leaf is otherwise the same car as the 2015 24kwh Leaf. It will be interesting to find out...

It is not possible to do the OEM retrofit of 40kWh to 2011-2015 24kWh Leaf yet. Nissan made many communication changes, so in order to do an OEM upgrade without any CAN-translators, you will need to flash the VCM in the car to be able to handle new things such as LB_MAXGIDS, Chargetimer differences, 0x5BC formatting, so many things that will diff otherwise, especially for the 2011-2013 ZE0 Leaf! Will be interesting for sure to see if Nissan ever releases a new VCM software.

 

[GerryAZ]

To give an idea of what the cell voltage display looks like with weak cells, here are a couple LEAF Spy screenshots

Did DTCs log ?

It sounds like the specific conditions to get a DTC from a weak cell is still not well understood. Are we trying to reach a threshold delta voltage between cells ? If that is the case, perhaps OP should find the pack state where the weak cell(s) are past the knee in the cell voltage discharge curve but the rest of the pack has not, and can still take a load without a marked voltage drop. Then an applied large load would most affect the weak cell in terms of voltage drop.

There were no DTC's set associated with the screenshots I posted previously. I had fairly rapid drops in SOC during highway driving toward the end with the 2015 from a few weak cells, but never got to the point of having DTC's set. I fully intended to drive it to 100k miles hoping for DTC's and also planned to formally complain to Nissan about the sudden loss of SOC (similar to OP's video, but not as pronounced) in attempt to have the battery repaired/replaced under warranty. The lure of the new Plus model with twice the horsepower, more than 2-1/2 times the original battery capacity, 100k mile battery capacity warranty, and $7,500 tax credit caused me to prematurely end the hot climate science experiment on the 2015 and start a new one with the 2019.

There have been cases noted in other threads on this forum with early 2018's setting DTC's and shutting down due to failed or weak cells, but we don't have enough information to know what it takes to set DTC's due solely to a few weak cells.

 

[LeftieBiker]

This raises an interesting question. If Nissan has any 24kwh "Lizard" packs left in stock then that is what the OP will get - not a 40kwh pack, which Nissan is now only using to replace the 30kwh "Lettuce Pack" when they are defective or fail. But what if they don't? So far I don't think that Nissan has done any 24kwh to 40kwh Leaf pack upgrades. We all know that it's possible, as the 2016-2017 Leaf is otherwise the same car as the 2015 24kwh Leaf. It will be interesting to find out...

It is not possible to do the OEM retrofit of 40kWh to 2011-2015 24kWh Leaf yet. Nissan made many communication changes, so in order to do an OEM upgrade without any CAN-translators, you will need to flash the VCM in the car to be able to handle new things such as LB_MAXGIDS, Chargetimer differences, 0x5BC formatting, so many things that will diff otherwise, especially for the 2011-2013 ZE0 Leaf! Will be interesting for sure to see if Nissan ever releases a new VCM software.

That's a shame. I thought that I had read about a conversion or two. I guess not...

 

[SageBrush]

To give an idea of what the cell voltage display looks like with weak cells, here are a couple LEAF Spy screenshots

Did DTCs log ?

It sounds like the specific conditions to get a DTC from a weak cell is still not well understood. Are we trying to reach a threshold delta voltage between cells ? If that is the case, perhaps OP should find the pack state where the weak cell(s) are past the knee in the cell voltage discharge curve but the rest of the pack has not, and can still take a load without a marked voltage drop. Then an applied large load would most affect the weak cell in terms of voltage drop.

There were no DTC's set associated with the screenshots I posted previously. I had fairly rapid drops in SOC during highway driving toward the end with the 2015 from a few weak cells, but never got to the point of having DTC's set. I fully intended to drive it to 100k miles hoping for DTC's and also planned to formally complain to Nissan about the sudden loss of SOC (similar to OP's video, but not as pronounced) in attempt to have the battery repaired/replaced under warranty. The lure of the new Plus model with twice the horsepower, more than 2-1/2 times the original battery capacity, 100k mile battery capacity warranty, and $7,500 tax credit caused me to prematurely end the hot climate science experiment on the 2015 and start a new one with the 2019.

There have been cases noted in other threads on this forum with early 2018's setting DTC's and shutting down due to failed or weak cells, but we don't have enough information to know what it takes to set DTC's due solely to a few weak cells.

Some stress testing is in order then. LeafSpy will help define DTC conditions.

I note that in the two LSpy screen captures you posted, the one at 20% SoC reports multiple weak cells while the screen at 15% SoC reports 'all cells OK' despite the much larger delta V. I wonder if the LSpy author knows why this is the case. I'm tempted to guess that larger delta V is allowed as the pack SoC drops but it would be nice to know the specifics.

By the way, it makes sense that 'power limited' is not a DTC per se since it will happen to any pack that has a sufficiently low SoC. The DTC algorithm has to differentiate between a low *pack* SoC and a low *cell* SoC in the context of a not-low pack SoC. If it is time based then starting from a low but not too low SoC and gunning the go pedal might expose the weak cell through a DTC.

Other thoughts ?

 

[alozzy]

That's a shame. I thought that I had read about a conversion or two. I guess not...

At least there's an upgrade path to larger packs, using the Muxsan CAN bridge and similar devices. If people like Emile can reverse engineer solutions, Nissan can definitely retrofit larger packs, if they want to. I don't think they'll do that...

It wouldn't surprise me if Nissan has made an arrangement with Envision (who bought AESC), to supply new 24 kWh packs for warranty claim purposes, through to the end of this year (when 2016 LEAFs with 24 kWh packs are no longer covered by capacity warranty coverage).

For the more rare defects, they'll likely swap bad cells until the 8 year warranty expires on 2016s.

Hopefully I'm wrong though

 

[alozzy]

I wonder how many 24 kWh/30 kWh battery pack replacements Nissan would have to do, if they offered to warranty claimants the option of trading in their current LEAFs for a new LEAF or Ariya, along with a discount incentive that's equivalent to Nissan's cost on a replacement pack?

Nissan could harvest the old pack modules and resell them as there's a healthy market for solar projects and such.

Image owing an older LEAF, with a heavily degraded pack and, rather than getting a replacement battery under warranty, you can opt for a new Ariya with a $10k discount!

FWIW, the discount could be based on the size of the original pack (i.e. $5k for a 24 kWh, $8k for a 30 kWh, $10k for a 40 kWh).

I can dream

 

[twilsonco]

Thanks for the ongoing discussion all, (though I realize you enjoy this stuff as much as I do)

I love how much more freely I can imagine the dynamics of cell voltage and resistance as a function of temperature, load, and SOC as a result of your collective insight. You might be interested in a response over on speakev where I crossposted (https://www.speakev.com/threads/help-rapid-battery-discharge-limited-motor-power-and-sporadic-power-output-at-temp-30f-and-40-motor-power.156115/post-2950977) DBMandrake points out that the bad cells have not only had their capacity degrade, but their resistance as well (likely both), more than the other healthy cells.

I'll be sure to update as things progress.
Currently, they still have my car to look at other issues—which I realize should each be their own post, now that I see how insanely helpful this forum is.

My updated plan of action is then similar, but prepended with one step:

• When I pick up the car, register a complaint with the manager (general or service?) that my car is experiencing a failure accompanied by a significant safety risk, and that I feel the techs aren't listening to or addressing my concerns to my satisfaction.

Otherwise, I'll perform rapid acceleration and charging tests at various SOCs to both isolate the problem and (hopefully) generate DTCs to the satisfaction of the techs and warranty folks.

 

[twilsonco]

This is also evident in severely diminished DC charging speeds the one time I did a full DC recharge from about 12% to 90% SOC on a 50kW ChargePoint charger. I expected it to hum along at close to 50kW (or whatever the max supported by my 24kWh Leaf) until it got to around 80%, then decrease the power logarithmically-ish until it trickled to full. Instead, it rose to about 35kW for a minute or so, and immediately started the logarithmic-ish throttling as shown below:

Contrast this to when I charge at 6.6kW (from aobut 30% to full) and it follows my expectations:

So this adds to the growing set of evidence that my issue is caused by one or more bad cells. Can't wait to get at this with Leaf Spy Pro.

I've noticed that the car always charges in about 60% of its estimated time. That's also explained by the bad cells' reduced capacity; they fill first spike voltage once full, making the car think it's at full SOC.

 

[SageBrush]

I've noticed that the car always charges in about 60% of its estimated time. That's also explained by the bad cells' reduced capacity; they fill first spike voltage once full, making the car think it's at full SOC.

Right! At least if you are thinking of SoC as a percentage.

You are going to *love* LeafSpy
I agree, nice post by Simon over at SpeakEV

 

[twilsonco]

...severely diminished DC charging speeds...

Actually, I see that that DC charging behavior is qualitatively similar to the 24kWh curve here (https://support.fastned.nl/hc/en-gb/articles/204784998-Charging-with-a-Nissan-Leaf-e-or-e-NV200). Here the x axis is SOC, but in my plots it was time, so you have to do a little mental math to imagine how, in my plot, when the charging rate lowers, it would decrease the speed at which you go from left to right in the power vs SOC plot. This would have the effect of making the curve in my DC charging plot more shallow, much like the 24kWh curve here. Add in the fact that my battery's at 7 bars, and it makes the DC charging on my car seem pretty normal.

 

[LeftieBiker]

When I pick up the car, register a complaint with the manager (general or service?) that my car is experiencing a failure accompanied by a significant safety risk, and that I feel the techs aren't listening to or addressing my concerns to my satisfaction.

That isn't the problem. The problem is that Nissan has constrained them to only address battery issues that generate DTCs or result in 4 or more lost capacity bars. They may be happy with this, but there is no point in antagonizing them for following Nissan policy. Just get those codes generated.

 

[twilsonco]

I was thinking in terms of getting documentation going on their end, per @SageBrush and others' advice.

I'm not sure what the content of such a complaint should be. My inclination is to compile a complete document with evidence and analysis to support my argument, but I'm not a certified Nissan EV tech so that won't mean anything, so I figure a general complaint as above would be enough to get the documentation ball rolling.