User avatar
TonyWilliams
Posts: 10090
Joined: Sat Feb 19, 2011 1:48 am
Location: San Diego
Contact: Website

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Sat Sep 21, 2013 10:05 pm

dm33 wrote:
TickTock wrote:Depends on the app. I integrate V*I*delta_T to compute the kWh during driving for efficiency computations. For remaining energy in the battery, gids remains the best indicator. Multiply by .075 to get kWh.
Thanks. That doesn't seem to be what the Android Leaf app is doing. For my car it shows 260 GIDs and 20.8kwh at a recent 100% charge.
260*.075 = 19.5 kwh.
I had seen a GID described as 80 watt hours. That if I use 0.08 then it does work out,
260*.080 = 20.8 kwh.
Which value is correct?
Both are correct. 300 Gid * 80 = 24kWh

Of that 24kWh capacity, 22.48kWh is pumped into the battery, and about 0.48kWh of the 22.48kWh is held in the battery at the bottom of the SOC, leaving 22kWh pumped in the battery that is "available" to use. Unfortunately, there are losses not accounted for with the 80 kWh calculation when you pull the energy out.

The usable value is closer to 75 per Gid, not 80. Actually, for that brand new battery with 21.3kWh usable at room temperature, it is about 77.45 per Gid.

A fully charged battery at 281 Gid, minus the 6-ish Gid remaining at Turtle leaves 275 Gid used multiplied by 77.45 per Gid = 21.3kWh (actually measured in a government test at room temperature). So, I just want to be clear that there are temperature related adjustments to capacity. We see Nissan's programming for these adjustments in the winter as the Gid count actually goes UP, even though the battery capacity goes down with cold.

LEAF battery capacities

Energy from the wall from dead to 100%: 25.414 kWh
Energy from the onboard charger to battery: 22.031 kWh (86.6% charger efficiency)
Energy from the battery during discharge: 21.381 kWh (our generic "21kWh" useable at 70F)

http://www1.eere.energy.gov/vehiclesand ... 2012_o.pdf" onclick="window.open(this.href);return false;

http://www1.eere.energy.gov/vehiclesand ... 2012_p.pdf" onclick="window.open(this.href);return false;



New 70F/20C fully charged battery
12 Fuel bars segments showing

-------------------KWH-------Gid------SOC%
Rated capacity: --24 ------- 300 ---- 100
Max possible. : ---24 ------ 300 ----- 100
Stored energy: --22.5 ----- 281 ----- 95 (4.1v per cell, 393.5 volts)
Usable energy: --21.0 ----- 281 ----- 95 (4.1v per cell, 393.5 volts)
Depleted cutoff: --0.5 ------ 6 -------- 2 (zero fuel bars remaining, 300v)


The battery temperature correction factors that I use are actually used in the LEAF Spy app for the DTE data. The app also uses the 0.48kWh stored at the bottom of the charge for DTE calculations, instead of counting GIDs to zero for range (which it obviously will never reach).

The battery temperature formula is 1% loss per 4F degrees below 70F and gains 1% per 8F above 70F.

Therefore, 281 GIDs at 30F battery temperature will have about 19kWh of usable energy, which is 69 watts per Gid. So, 75 is as good as any singular value, but obviously it needs adjustments for temperature, and the wattHour value per Gid can vary significantly.

User avatar
surfingslovak
Vendor
Posts: 3809
Joined: Mon Jun 13, 2011 1:35 pm

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Sat Sep 21, 2013 10:41 pm

Great post, Tony! Almost makes me feel nostalgic for early 2011.

dm33
Posts: 690
Joined: Sat May 18, 2013 4:43 am
Delivery Date: 31 May 2013
Location: Raleigh, NC

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Sat Sep 21, 2013 11:28 pm

TonyWilliams wrote:
dm33 wrote: Which value is correct?
Both are correct. 300 Gid * 80 = 24kWh
Whoa. Lots of numbers and values.

To put it a different way.... when the car reports 4.1 mi/kwh, which kWh are we talking about. I assume not from the wall. I would think the value someone wants to know while they're looking at an app in the car is the kWh that related to the car reported efficiency (mi/kwh) so you can estimate range.

Is it 84 mile range at 62mph at 4.1mi/kwh? 20.48 kWh. Figuring 281 Gids, thats 73 watts/Gid.

What would a user want to see for kWh displayed in an app. 80watts/kwh to be consistent w the Android Leaf app? Some other value? Are Gids better for a user to use to guestimate range rather than kWh?

And to make sure I understand.... there is no kWh measurement reported by the Leaf. Only Gids...

User avatar
surfingslovak
Vendor
Posts: 3809
Joined: Mon Jun 13, 2011 1:35 pm

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Sat Sep 21, 2013 11:40 pm

It's been suggested to Jim to implement a battery efficiency factor and temperature compensation. I'm not sure if he has gotten around to that yet. To be clear, GIDs are 80 kWh stored energy, full point. To account for losses and other factors, the LEAF monitors pack voltage and substracts from the GID count as needed. This can be observed sometimes when you turn the LEAF off and then on only to see a battery SOC bar disappear. That's because the LEAF adjusted the GID count to more adequately reflect currently available energy in the battery. Of course, none of that is taken into account if the battery simply multiplies GIDs by 80 Wh. This has to be properly understood to account for the different numbers and expectations. I wish Jim had listened early on, but then he was willing to compromise a lot, and go down a path others wouldn't. Phil's LeafScan, the gold standard of meters last year, was supposed to work much the same way. I'm hopeful that all this will be addressed more adequately soon.

Turbo3
Gold Member
Posts: 2011
Joined: Mon Jul 19, 2010 8:34 pm
Delivery Date: 12 May 2011
Leaf Number: 002191
Location: San Jose, CA

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Sun Sep 22, 2013 12:15 am

Leaf Spy implements Tony's full adjustments for both battery temp and ambient temp.

To use go to Settings/DTE panel and check the "Alt DTE calc' check box.

User avatar
surfingslovak
Vendor
Posts: 3809
Joined: Mon Jun 13, 2011 1:35 pm

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Sun Sep 22, 2013 12:17 am

Turbo3 wrote:Leaf Spy implements Tony's full adjustments for both battery temp and ambient temp.
Great to hear, Jim! Thanks for confirming that.

User avatar
garygid
Gold Member
Posts: 12464
Joined: Wed Apr 21, 2010 8:10 am
Delivery Date: 29 Mar 2011
Leaf Number: 000855
Location: Laguna Hills, Orange Co, CA

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Sun Sep 22, 2013 7:52 am

To account for losses and other factors, the LEAF monitors pack voltage
and substracts from the GID count as needed. This can be observed sometimes
when you turn the LEAF off and then on only to see a battery SOC bar disappear.
Minor point of information:

The vanishing battery fullness Bar has little to do with any adjustment of the
GIDs value. With NO change in the GID value, a bar will often "vanish" due to
the hysteresis in managing the Bars in the display.

While driving, a Bar will persist (continue to be displayed) until the Bar is
empty, then it disappears. From 11.0 down to 10.01 the 11th Bar shows
(while displaying a generally declining battery pack "fullness").

However, when starting up the car (like when charging), a bar is displayed
only if it is half full, or more. So, from 10.01 to 10.49 the 11th Bar is NOT
displayed.

Thus, driving along with 8 Bars showing, and the "fullness" is 7.3 (anywhere
between 7.01 and 7.49). Stop, turn the car Off and back On, and...
PRESTO, the 8th bar is gone!

Nothing is different with the capacity, the fullness, or the GIDs, but
the Bar is gone. Why, because that is the way Nissan designed
the Fullness-Bar meter to work, with about a half-Bar of hysteresis. :D
See SOC/GID-Meter and CAN-Do Info
2010 Prius, now for sale
2011 LEAF, sold in 2015
2018 Tesla Model 3
2014 Tesla S, Model 3 in 2019
PU: SDG&E
Solar PV: 33 x 225W -> 7 kW max AC
To Sell: X-treme 5000Li EV motorcycle

User avatar
JPWhite
Gold Member
Posts: 1784
Joined: Sun Jul 31, 2011 3:41 pm
Delivery Date: 30 Jul 2011
Leaf Number: 5734
Location: Hendersonville TN
Contact: Website Yahoo Messenger AOL

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Sun Sep 22, 2013 11:14 am

garygid wrote:
To account for losses and other factors, the LEAF monitors pack voltage
and substracts from the GID count as needed. This can be observed sometimes
when you turn the LEAF off and then on only to see a battery SOC bar disappear.
Minor point of information:

The vanishing battery fullness Bar has little to do with any adjustment of the
GIDs value. With NO change in the GID value, a bar will often "vanish" due to
the hysteresis in managing the Bars in the display.

While driving, a Bar will persist (continue to be displayed) until the Bar is
empty, then it disappears. From 11.0 down to 10.01 the 11th Bar shows
(while displaying a generally declining battery pack "fullness").

However, when starting up the car (like when charging), a bar is displayed
only if it is half full, or more. So, from 10.01 to 10.49 the 11th Bar is NOT
displayed.

Thus, driving along with 8 Bars showing, and the "fullness" is 7.3 (anywhere
between 7.01 and 7.49). Stop, turn the car Off and back On, and...
PRESTO, the 8th bar is gone!

Nothing is different with the capacity, the fullness, or the GIDs, but
the Bar is gone. Why, because that is the way Nissan designed
the Fullness-Bar meter to work, with about a half-Bar of hysteresis. :D
My casual observations align well with Gary's analysis.
--
JP White
http://jpwhitenissanleaf.com
2011 Blue SL-e, 132,400 Miles.
Lost 5 Capacity bars
7/18/13 (29,206), 8/25/14 (51,728), 7/12/15 (71.108), 5/12/16 (88,362), 10/17/16 (96,532)
New Battery 12/3/16 (98,956)
2018 Model 3 20,000 Miles.

User avatar
TonyWilliams
Posts: 10090
Joined: Sat Feb 19, 2011 1:48 am
Location: San Diego
Contact: Website

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Sun Sep 22, 2013 11:35 am

dm33 wrote:
TonyWilliams wrote:
dm33 wrote: Which value is correct?
Both are correct. 300 Gid * 80 = 24kWh
Whoa. Lots of numbers and values.
Yes, lots of numbers. I have spent a lot of time actually driving electric powered cars, and knowing how far they go is important to me, given their limited range and limited public recharging infrastructure. Twice, I have driven between Canada and Mexico, last year in a LEAF and this year in a Toyota Rav4 EV.
To put it a different way.... when the car reports 4.1 mi/kwh, which kWh are we talking about. I assume not from the wall. I would think the value someone wants to know while they're looking at an app in the car is the kWh that related to the car reported efficiency (mi/kwh) so you can estimate range.
The kWh calculated with a Gid=80Wh value with no adjustments for energy that is never accessible at the bottom of the SOC%, nor compensated for battery temperature has marginal value for range calculations. The 4.1 miles per kWh that you reference needs to be calculated:

4.1 * (battery usable stored energy) = range autonomy in miles

The usable stored energy is simply:

(Gid = "281" - 6) * 77 = 21.1kWh for that fully charged, new condition 70F battery, creating:

4.1 * 21.1 = 86.5 miles range autonomy at 70F battery temperature

With adjustments to the battery capacity for temperature, now that same battery at 30F degrees holds:

21.1 * (70F - 30F = 40F delta / 4F =10% loss) = about 19kWh usable, therefore:

4.1 * 19 = 77.9 miles
Is it 84 mile range at 62mph at 4.1mi/kwh? 20.48 kWh. Figuring 281 Gids, thats 73 watts/Gid.

As previously stated, the usable Gid value is variable, primarily with temperature. Internal battery resistance with degradation is another factor.

What would a user want to see for kWh displayed in an app. 80watts/kwh to be consistent w the Android Leaf app? Some other value? Are Gids better for a user to use to guestimate range rather than kWh?
Gids are absolutely the best tool so far to estimate range. If the kWh displayed in the LEAF Spy app showed:

[(Gid - 6) * 0.0775] * [100% - (70F - battery temperature = temp delta... if positive value, divide by 4... if negative value, divide by 8)] = usable kWh compatible with dash economy reading.

So, either 75 to 77.5 watthours per Gid is a good baseline.

Then, you could always look at your dash miles/kWh economy information and multiply it by the displayed kWh and get the range.

Let's try is with my old, degraded yet fully charged LEAF in an Edmonton, BC typical January day @ -10F degrees:

[(230 - 6) * 0.0775] * [100% - (70F - (-10F) = 80 / 4)] = 13.89kWh usable * 4.1 = 56.9 miles autonomy


I would recommend that the displayed kWh match the DTE data, instead of being calculated differently when using my formulas for compensation.
And to make sure I understand.... there is no kWh measurement reported by the Leaf. Only Gids...
And amphours (Ah) and voltage. I believe the current LEAF Spy app display of kWh is merely Ah * nominal voltage (sorry, I'm not sure of the exact method of calculation since if it's not useful data for range estimation, I really don't care how it's calculated).
Last edited by TonyWilliams on Sun Sep 22, 2013 12:06 pm, edited 3 times in total.

User avatar
surfingslovak
Vendor
Posts: 3809
Joined: Mon Jun 13, 2011 1:35 pm

Re: Active EV-CAN sampling: cell voltages, pack temperatures

Sun Sep 22, 2013 11:44 am

TonyWilliams wrote:I believe the current LEAF Spy app display of kWh is merely Ah * nominal voltage (sorry, I'm not sure of the exact method of calculation since if it's not useful data for range estimation, I really don't care how it's calculated).
That has changed. Since Jim was able to get real GIDs off the CAN bus, he is using 80 Wh per GID to calculate the total stored energy. I did not use the app recently to tell you if battery charging and discharging efficiency is taken into account, but Jim confirmed above that he has implemented temperature correction. He also subtracts the fixed reserve at the bottom. I believe that it's user adjustable, and one can select how much reserve should be loped off the indicated stored energy. I think Jim needs to be lobbied to changed the GID multiplier from 80 to something else. Perhaps this could be exposed as an adjustable value in the preferences as well? That could be more easily understood than a battery efficiency factor, which was proposed earlier.

Return to “LEAF CANBus”