WetEV
Posts: 2256
Joined: Fri May 04, 2012 8:25 am
Delivery Date: 16 Feb 2014
Location: Near Seattle, WA

Re: Living the Leaf Life Below 40% SOC and pulse driving - battery protection strategy

Tue Aug 07, 2018 8:09 am

Silverfish wrote:
WetEV wrote:Leafspy just reports on what the BMS estimates the battery capacity is. The BMS can be fooled. Fooling the BMS doesn't improve the battery. Resetting the BMS doesn't improve capacity. Recharge and range tests confirm this.

DCQC and high discharge and regeneration probably wear the battery faster.


@WetEV, how can you tell the difference between fooling the BMS and doing things that are actually good for your battery? I've read that the LeafSpy battery stats are kind of meaningless over the short term, like a few months, but I'd think that over 15 months, they would at least mean something.

In the 2+ months I've had my car, I have noticed that babying the battery makes the LeafSpy stats go down, and getting the car to really stretch its legs -- longer distances, freeway driving, deeper discharges and recharges -- makes them go up a bit. I get that those numbers may not be accurate, but in that case, how can I measure and determine what's good for my car?


What follows is my mostly opinion.

I think that the battery true capacity is mostly declining with time. Faster for deep discharge/charge cycles, faster for high rates of charge/discharge, faster at high levels of charge or low levels of charge and faster at higher temperatures. Slightly faster for the first 5% of decline, then slower, then faster somewhere below 70% capacity.

I think that most variation of the BMS statistics that LeafSpy reports is just inaccuracies in the estimate, and not real changes. I don't have measurements accurate enough to prove this. In a few months last fall, I measured my LEAF's battery capacity with a L2 recharge test at temperature of 20C, then tried to increase the BMS estimate by DCQC, faster driving, deeper discharge and a few other tricks, then to tried to reduce the BMS estimate by slow L2 charge (8 Amp L2), local driving only, keeping charge near 50%, shallow charge cycles. The result was that the BMS estimate can be 12% or more larger capacity than real, and perhaps 2% to 3% percent less than real. This is making the assumption that the real capacity doesn't change much over a few months, and variations in driving/charging don't actually change real capacity.

So how could I take measurements with the BMS that vary +12% to -3% from real capacity and find the trend of declining capacity which is about 2.5% to 3% a year in my case? With at least a few months every year of consistent driving in consistent temperatures and such and a few years of records, I think this is a realistic goal, however isn't easy as your life changes, and as year to year weather may be different. This gets harder on a shorter term basis, but on a year to year period might be fairly accurate.

So how could find the smaller variations in that small trend caused by different charging rates, driving styles, depth of charge/discharge cycles and so on? What makes this problem even worse is that the BMS software has been updated at least on some cars, the battery chemistry has changed over the years and so forth... I don't think there is an easy way to use BMS statistics to really tell what's good for the car. It seems to me that it would take years and multiple cars to get to a believable answer. It would be easier to use a more accurate way of measuring battery capacity, and multiple cars driven in the same conditions with different charge rates, depth of charge/discharge cycles, etc. But with one car? You are looking at a tiny needle in a large haystack. The haystack is fuzzed up because of BMS inaccuracy. I don't think you can tell the difference between "fooling the BMS and doing things that are actually good for your battery", at least with one car and a reasonable amount of time.
WetEV
#49
Most everything around here is wet during the rainy season. And the rainy season is long.
2012 Leaf SL Red (Totaled)
2014 Leaf SL Red

Silverfish
Posts: 27
Joined: Thu Jun 21, 2018 10:32 am
Delivery Date: 21 Jun 2018

Re: Living the Leaf Life Below 40% SOC and pulse driving - battery protection strategy

Sun Aug 12, 2018 11:33 am

WetEV wrote:
Silverfish wrote:
WetEV wrote:Leafspy just reports on what the BMS estimates the battery capacity is. The BMS can be fooled. Fooling the BMS doesn't improve the battery. Resetting the BMS doesn't improve capacity. Recharge and range tests confirm this.

DCQC and high discharge and regeneration probably wear the battery faster.


@WetEV, how can you tell the difference between fooling the BMS and doing things that are actually good for your battery? I've read that the LeafSpy battery stats are kind of meaningless over the short term, like a few months, but I'd think that over 15 months, they would at least mean something.

In the 2+ months I've had my car, I have noticed that babying the battery makes the LeafSpy stats go down, and getting the car to really stretch its legs -- longer distances, freeway driving, deeper discharges and recharges -- makes them go up a bit. I get that those numbers may not be accurate, but in that case, how can I measure and determine what's good for my car?


What follows is my mostly opinion.

I think that the battery true capacity is mostly declining with time. Faster for deep discharge/charge cycles, faster for high rates of charge/discharge, faster at high levels of charge or low levels of charge and faster at higher temperatures. Slightly faster for the first 5% of decline, then slower, then faster somewhere below 70% capacity.

I think that most variation of the BMS statistics that LeafSpy reports is just inaccuracies in the estimate, and not real changes. I don't have measurements accurate enough to prove this. In a few months last fall, I measured my LEAF's battery capacity with a L2 recharge test at temperature of 20C, then tried to increase the BMS estimate by DCQC, faster driving, deeper discharge and a few other tricks, then to tried to reduce the BMS estimate by slow L2 charge (8 Amp L2), local driving only, keeping charge near 50%, shallow charge cycles. The result was that the BMS estimate can be 12% or more larger capacity than real, and perhaps 2% to 3% percent less than real. This is making the assumption that the real capacity doesn't change much over a few months, and variations in driving/charging don't actually change real capacity.

So how could I take measurements with the BMS that vary +12% to -3% from real capacity and find the trend of declining capacity which is about 2.5% to 3% a year in my case? With at least a few months every year of consistent driving in consistent temperatures and such and a few years of records, I think this is a realistic goal, however isn't easy as your life changes, and as year to year weather may be different. This gets harder on a shorter term basis, but on a year to year period might be fairly accurate.

So how could find the smaller variations in that small trend caused by different charging rates, driving styles, depth of charge/discharge cycles and so on? What makes this problem even worse is that the BMS software has been updated at least on some cars, the battery chemistry has changed over the years and so forth... I don't think there is an easy way to use BMS statistics to really tell what's good for the car. It seems to me that it would take years and multiple cars to get to a believable answer. It would be easier to use a more accurate way of measuring battery capacity, and multiple cars driven in the same conditions with different charge rates, depth of charge/discharge cycles, etc. But with one car? You are looking at a tiny needle in a large haystack. The haystack is fuzzed up because of BMS inaccuracy. I don't think you can tell the difference between "fooling the BMS and doing things that are actually good for your battery", at least with one car and a reasonable amount of time.


Thanks.

At this point, I'm mostly tracking the car's reported battery percentage and the trip meter, to get a sense of how many miles per battery percent the car can go. There's a lot of variation depending on driving conditions, but the averages don't seem to have decreased over the last couple of months, even though the LeafSpy stats have.

I have to admit, though, that it is unsettling to see those LeafSpy numbers go steadily down. I guess I'll have to just trust that babying the battery is my best option and start ignoring LeafSpy.

jim0266
Posts: 68
Joined: Fri Mar 17, 2017 10:14 am
Delivery Date: 17 Mar 2017
Leaf Number: 330446
Location: Akron, Oh

Re: Living the Leaf Life Below 40% SOC and pulse driving - battery protection strategy

Sun Aug 12, 2018 8:12 pm

Ran across this article of a gentleman in NZ doing the same thing with his 2012 Leaf by keeping the SOC below 40% as he also drives short distances.

Early adopters as teachers: a case of extraordinary care of a Leaf battery
https://flipthefleet.org/2018/early-adopters-as-teachers-a-case-of-extraordinary-care-of-a-leaf-battery/

"The goal was to minimise battery degradation using optimum charging practices, identified as shallow depth-of-discharge, frequent part-charging, and as far as possible to keep charge level at or below a level considered best practice for storage (30-40%). Web discussion papers suggest that li-ion batteries are most stable at 3.92 cell voltage; a SoC level of 39.0% gives a maximum cell voltage of 3.925v for the Leaf under discussion."

Silverfish
Posts: 27
Joined: Thu Jun 21, 2018 10:32 am
Delivery Date: 21 Jun 2018

Re: Living the Leaf Life Below 40% SOC and pulse driving - battery protection strategy

Mon Aug 13, 2018 10:06 am

jim0266 wrote:Ran across this article of a gentleman in NZ doing the same thing with his 2012 Leaf by keeping the SOC below 40% as he also drives short distances.

Early adopters as teachers: a case of extraordinary care of a Leaf battery
https://flipthefleet.org/2018/early-adopters-as-teachers-a-case-of-extraordinary-care-of-a-leaf-battery/

"The goal was to minimise battery degradation using optimum charging practices, identified as shallow depth-of-discharge, frequent part-charging, and as far as possible to keep charge level at or below a level considered best practice for storage (30-40%). Web discussion papers suggest that li-ion batteries are most stable at 3.92 cell voltage; a SoC level of 39.0% gives a maximum cell voltage of 3.925v for the Leaf under discussion."


Interesting. Good to have another data point.

hieronymous
Posts: 35
Joined: Tue Sep 16, 2014 8:54 pm
Delivery Date: 30 Jul 2014
Leaf Number: 200222
Location: Auckland, New Zealand

Re: Living the Leaf Life Below 40% SOC and pulse driving - battery protection strategy

Tue Aug 14, 2018 9:59 pm

jim0266 wrote:Ran across this article of a gentleman in NZ doing the same thing with his 2012 Leaf by keeping the SOC below 40% as he also drives short distances.

Early adopters as teachers: a case of extraordinary care of a Leaf battery
https://flipthefleet.org/2018/early-adopters-as-teachers-a-case-of-extraordinary-care-of-a-leaf-battery/

"The goal was to minimise battery degradation using optimum charging practices, identified as shallow depth-of-discharge, frequent part-charging, and as far as possible to keep charge level at or below a level considered best practice for storage (30-40%). Web discussion papers suggest that li-ion batteries are most stable at 3.92 cell voltage; a SoC level of 39.0% gives a maximum cell voltage of 3.925v for the Leaf under discussion."


Hi Jim - thanks for the FtF contact re above. Thought any discussion may possibly benefit others here.

My thinking now is that the BMS is designed not to function below 40% SoC, a Nissan mechanism to allow batteries stockpiled for the production line, and transported subsequently, to remain at 100% until sale time perhaps many months ahead. My Leaf was stockpiled for 26 months before I bought it as a new car, with SOH at 96%. No doubt calendar ageing is still occurring, but the stable numbers have a "feel good" factor. The main benefit is the stability offered by the 3.9v cell level.

On your longer trips I think the 100% charge is the key to your better metrics; the battery becomes optimally balanced for a while, returning the best range and boosting range related data. A spell under 40% tends to do the opposite, with an increasing cell voltage delta (not that it matters for short trips).

So, for long battery life: choose a temperate climate to live in (my annual battery temperature range is about 13-25C), keep the battery below 40% as much as possible (2 charge bars or less), charge only as much as required, and only when needed (the car timers can make this easy), and build in one regular 100% charge/fairly full discharge cycle every month or two, to maintain reasonable pack balance.

Just my 2c worth :)
Local use only: max. charge 39.0% SoC
Built 05/12, Nissan stockpiled 26 months
07/14: 57km, SOH 96%, AHr 63.37 (new)
07/15: 4032km, 93%, 61.41
07/16: 6628km, 95%, 62.90 (P3227)
07/17: 8742km, 95%, 62.90
07/18: 11355km, 94.8%, 62.22

jim0266
Posts: 68
Joined: Fri Mar 17, 2017 10:14 am
Delivery Date: 17 Mar 2017
Leaf Number: 330446
Location: Akron, Oh

Re: Living the Leaf Life Below 40% SOC and pulse driving - battery protection strategy

Thu Aug 16, 2018 8:55 am

Thanks hieronymous for weighing in. I do agree about the feel good factor. It's a grand experiment.

I also sent you a PM.

jim0266
Posts: 68
Joined: Fri Mar 17, 2017 10:14 am
Delivery Date: 17 Mar 2017
Leaf Number: 330446
Location: Akron, Oh

Re: Living the Leaf Life Below 40% SOC and pulse driving - battery protection strategy

Mon Aug 27, 2018 10:45 am

Update:

Finally made a long 108 mile RT pulse-driving run (lunch with mom :D ). I L2 charged to 100% (97.1% actual) at home. LeafSpy reported 275 GIDS on a full charge. This was the first time in 34 days I’ve charged beyond 40%. I drove 406 miles in between these dates.

Starting stats:
Hx 95.18%, SOH 97.05%, Ahr 63.47

The battery lost Hx values all the way to my destination. I arrived at my destination around 44% SOC and L2 charged back to 89%. Around 77% SOC on the way home pulse driving started to increase the Hx values. Hx dropped as low as 92.58% before rising.

Ending stats:
Hx 93.90%, SOH 95.53%, Ahr 62.48

Over the last month if I had gone above 40% SOC I would have had to work to keep these numbers.

As I make longer runs over time it will be interesting to see if the pack is more or less responsive to my pulse driving. Even when I was pulse driving on a regular basis I could never predict if the pack would be willing to accept the shocks or be recalcitrant like on this run. The first long trip I made after keeping <40% for 3 weeks the pack went from 93% to 97% SOH, where it stayed until I made this trip.

My 14 Leaf now has 21,420 miles.

I've also started to track my GIDS to look for changes over time. Recent readings end of July to date:

SOC GIDS

39 86
39.1 84
39.1 85
39.1 88
39.2 86
39.2 82
39.6 86
39.7 87
39.8 87
39.8 86
39.8 87
39.8 87
39.8 88
39.8 86
39.9 90
39.9 90
39.9 84
39.9 88
40 88
40 90

Return to “General / Main Owners Forum”