Real SOC (State of Charge) - Information

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Ingineer

Well-known member
Joined
Oct 15, 2010
Messages
2,742
Location
Berkeley, California
It was recently asked how I derive SOC for LEAFSCAN. This number is simply obtained from the Battery ECU, and is displayed as-is. The various Leaf systems use this as a basis for making operational decisions and displaying information. The SOC is created by coulomb counting, which is basically defined as counting the electrons passing in/out of the battery. Due to the fact that it's not actually possible to do this very accurately in a mass-produced EV, the battery ECU periodically makes adjustments to keep the SOC as accurate as possible. One of the main ways this is done is by checking open-circuit voltage of the pack and applying some math to correct this number for things that will affect it, and using a lookup table of known states. This all working together makes for a pretty accurate SOC calculation. The Leaf's engineering team make use of a hall-effect based current monitor to perform the coulomb counting, and these have a number of issues that make them less accurate than ideal, but they are cheap and good enough.

The reported SOC usable range seems to be from about 95% (fully charged) to 2% when the battery main contactor opens up and disconnects the pack. The first low battery warning occurs at around 18%, and the second one at around 9%. If I remember correctly, Turtle mode comes on at about 5%, and begins limiting power. At about 2%, you are totally dead.

-Phil
 
Ingineer said:
The reported SOC usable range seems to be from about 95% (fully charged) to 2% when the battery main contactor opens up and disconnects the pack. The first low battery warning occurs at around 18%, and the second one at around 9%. If I remember correctly, Turtle mode comes on at about 5%, and begins limiting power. At about 2%, you are totally dead.

-Phil

So, with any of fancy testing, does this look reasonable?

Code:
Gid%       Gid Value  SOC%
100.0%     281          95

91.5%      257

84.0%      236

77.9%      219

70.8%      199

66.2%      186

58.0%      163

50.9%      143

43.4%      122

36.3%      102

31.3%       88

26.0%       73

17.4%       49        18 (LBW)

8.9%        25         9 (VLB)

1.4%        4          5 (Turtle)

???         2          2 (dead)
 
TonyWilliams said:
So, with any of fancy testing, does this look reasonable?
Yes. Just as you see the top-end "Gids" vary from charge to charge, the SOC also varies a bit.

Keep in mind "Gids" are actually the Battery ECU's Watt-hours number (times 80), so as the battery ages, expect this number to drop, while SOC will still rise to ~95%. This means the watt-hours/SOC will slowly lessen. The battery ECU constantly calculates how the battery ages and produces a Loss of Capacity Coefficient, which will be available on LEAFSCAN.

-Phil
 
Based on suggestion, I may add an expanded SOC variable that actually runs from 0% to 100% rather than 2% to 95%. The roughly 7% gap is what Nissan decided should not be used for longest life on the battery pack. Last time I checked, my Leaf was reporting 67.568 amp-hours full capacity.

-Phil
 
Ingineer said:
Based on suggestion, I may add an expanded SOC variable that actually runs from 0% to 100% rather than 2% to 95%. The roughly 7% gap is what Nissan decided should not be used for longest life on the battery pack. Last time I checked, my Leaf was reporting 67.568 amp-hours full capacity.

-Phil

I think that is wise. Otherwise, it will confuse some people.

Then, we can see the battery at 100% every charge (+/-), and watch the Gid % drop over time. With the two pieces of data, you'll have a good understanding of what you have 100% of.

I think it would be a bit messy to have 88% Gid, and 94% SOC to be a fully charged battery.

Will the LEAFSCAN include the Gid, so I don't need two boxes?
 
Ingineer said:
The reported SOC usable range seems to be from about 95% (fully charged) to 2% when the battery main contactor opens up and disconnects the pack. The first low battery warning occurs at around 18%, and the second one at around 9%. If I remember correctly, Tourtle mode comes on at about 5%, and begins limiting power. At about 2%, you are totally dead.
-Phil

Mine shows LBW at 17.4% every time, VLBW at 8.5% every time, and Turtle varies between 2.4% and 1.4%. When I drove fast, it was 2.4, but when I drove slowly Turtle happened always at 1.4.
 
LEAFfan said:
Mine shows LBW at 17.4% every time, VLBW at 8.5% every time, and Turtle varies between 2.4% and 1.4%. When I drove fast, it was 2.4, but when I drove slowly Turtle happened always at 1.4.
Where are you getting these SOC values? Again, Gary's box only shows watt-hours, not SOC.

-Phil
 
I actually prefer this since, having used Gary's meter for a while now, I find the idea of knowing how much actual charge is in the battery more useful to me for on-going range planning than a scaled percentage of an unknown and changing capacity.

Ingineer said:
TonyWilliams said:
Will the LEAFSCAN include the Gid, so I don't need two boxes?
Yes, I can add "Gids" and also the real watt-hours number (gids times 80).
 
To display an approximation of Usable energy, the GID-Meter (SOC-Meter) displays "GIDs/281" as a percentage, with one decimal: 99.7 for 280.

Having a value that always shows 95% when "Full" is misleading, because the top 5% is not usable.

If the "fuel" value shown always goes to 100%, even after the battery loses capacity, that is also misleading for calculating range. If the battery has lost 7.6%, I would want my "full" fuel gauge to show 92.4% (percentage of New-Battery usable "fuel", or energy).

The GIDs might be derived from 80 Wh increments, but the GIDs at the top disappear more quickly, and those at the bottom will occasionally disappear much more quickly, so, at times, like just before the more-rapid decline, the GIDs do NOT do a good job of indicating the remaining usable energy. I suspect the GID value is "adjusted" downward when any cell-pair's voltage drops "early".

The "real" SOC, as a percent of the whole, but unusable and possibly diminished, battery capacity is not useful for most driving purposes, certainly not for estimating Range.

Yes, maybe a 100% (or other constant value) display indicating that the diminished-capacity battery is "as full as it is going to get" is useful for the car during charging, but not really useful while driving, or for planning to drive.

A MUCH more useful figure would be the Pack's usable-energy Capacity as a percentage of New-Battery (and in kWh). Right now, Nissan "hides" that dapacity-loss information until 15% is gone - MUCH too late to modify usage habits to better preserve the Pack Capacity.

IMO, Gary
 
garygid said:
<snip>If the "fuel" value shown always goes to 100%, even after the battery loses capacity, that is also misleading for calculating range. If the battery has lost 7.6%, I would want my "full" fuel gauge to show 92.4% (percentage of New-Battery usable "fuel", or energy).</snip>

+1 Seems like this will go hand-in-hand with Tony's range chart to give really accurate predictive capability ... but not if it changes over time.

(or perhaps I'm confused.....)
 
Ingineer said:
Where are you getting these SOC values? Again, Gary's box only shows watt-hours, not SOC.
Phil, LEAFfan is using a modified, or prototype, ScanGauge in his car. As far as I know, it's reading the same CAN bus code Gary does.
 
The problem with using the watt-hours number is resolution. If you are happy with the ~280 step gauge, then by all means you can continue to use that on LEAFSCAN. However it's not for me, as I love seeing the SOC numbers tick down during my drive, and the 5 digit resolution is enough to even instantly see the effect using the climate control has on remaining energy. The "Gids" are just not a high enough resolution for me!

There are a lot of factors that are needed to "accurately" calculate range. For instance, cooler denser air will increase energy consumption, so you have to take that into account as well. For the same reason that Nissan was unable to create an accurate range estimator, we also have that problem. Seeing exactly how much energy is currently being used at all times and seeing how quickly your actions behind the wheel suck the pack dry allows your wetware upstairs to get a really good feel for range and also makes you a more efficient driver.

If someone can devise a good formula for displaying the usable energy in a higher resolution, I'll be happy to implement it. Maybe we can take the current capacity figure (amp hours with precision to the thousandths) and multiply that by corrected SOC and get a higher resolution watt-hours figure that both represents real energy (that will degrade over time) so you can "feel" it tick down.

-Phil
 
Ingineer said:
Based on suggestion, I may add an expanded SOC variable that actually runs from 0% to 100% rather than 2% to 95%. The roughly 7% gap is what Nissan decided should not be used for longest life on the battery pack. Last time I checked, my Leaf was reporting 67.568 amp-hours full capacity.

-Phil

Phil-

What variability of battery capacity are you seeing, with battery temp changes only?

I don't think capacity loss over time will be too great a problem for a driver watching a % of capacity display, as an indication of available range, since this will occur slowly, over years.

But if battery capacity changes significantly due to battery temp, the % could be misleading for range purposes, from day to day, or even over a single day, even for drivers capable of using "wetware".
 
I believe LeafFan shows the same data that the "SOC-Meter" (GID-Meter) shows:

GID = EV:5BC:D1 and top 2 bits of D2 (of D1-D8)
And, the "soc" or "gid" percent = 100.0% * (GID / 281)

There should be no confusion about where the GIDs and GID% come from, but just a question of what it REALLY means, in spite of the 80 Wh description.

Only a few times have cars reported over 281, sometimes associated with a QC session.

Many cars have reported 281 occasionally, but most report slightly less (oftenvin the mid to high 270's).

A few cars (like TickTock's) report significantly lower values after a "100%" charge.

While watching digits whirl down, it is not possible for most people to get meaningful information, at least quickly. That is why we show significant rate-of-change meters: Wh/minute, Wh/mile, or Miles/kWh.

Hopefully the LS will be able to show both integrated (quantity from a starting-point from a rate) and differentiated (rate from a changing quantity) derived values.

Also, showing 0.001% value changes on a value that is unlikely to be accurate to 1%, would appear to be just a fun game, not really very useful IMO.

But, the ability to modify or ignore the "standard" screens and create one's own screens will allow a LOT of flexibility, especially if the last-seen screen (or a user-selected default screen) is displayed at boot-up.

Please keep up the great work.
 
For my internal range estimate I try to guess the remaining kWH in the pack, which
combined with the Miles/kWh measure on the dash works reasonably well...So if you
can display the kWH remaining, that would be great.

Since you have the remaining Ah it should be possible, although I dont know how much the voltage under load varies and how reliable integration of current amps*voltage would be? Maybe averaging over a few seconds will make it stable enough, but then I guess that is what the built-in SOC meter is doing anyway?

In any case, "remaining kWH", regardless of the battery capacity etc, would be most helpful for me.
 
klapauzius said:
For my internal range estimate I try to guess the remaining kWH in the pack, which
combined with the Miles/kWh measure on the dash works reasonably well...So if you
can display the kWH remaining, that would be great.

Since you have the remaining Ah it should be possible, although I dont know how much the voltage under load varies and how reliable integration of current amps*voltage would be? Maybe averaging over a few seconds will make it stable enough, but then I guess that is what the built-in SOC meter is doing anyway?

In any case, "remaining kWH", regardless of the battery capacity etc, would be most helpful for me.
+1
 
I could easily display GIDs * 80 / 1000 = "kWh" (like "E24.8"),
as a 3rd value in Mode 1, but I think the resulting Energy
value is substantially misleading.

So far I have not seen any evidence that 80 Wh per GID
actually represents the USABLE Energy, but there appears
to be some evidence that it does not accurately do that.

Without knowing how the car measures, corrects, and thus
derives the GIDs, the value should be used with educated caution.

If I put this "E" feature in, I might need to use a multiplier lower than 80, and possibly make it user-adjustable.

Suggested values?
 
A MUCH more useful figure would be the Pack's usable-energy Capacity as a percentage of New-Battery (and in kWh). Right now, Nissan "hides" that dapacity-loss information until 15% is gone - MUCH too late to modify usage habits to better preserve the Pack Capacity.

Gary,

I think Phil said his unit will display the accurate loss of battery capacity (not the dumbed down 15% loss equals a bar).

With that data, shouldn't a remapped SOC at 100% (real 95% SOC) times a ten percent loss of capacity equal about 90% Gid that we expect to see (with the 10% capacity loss)?

I think we'll have enough data now (with battery temp also) to give a real world value of remaining kWh multiplied by miles/kWh equals how far you'll go in existing conditions.

Exciting times,

Tony
 
Back
Top