LEAFer's LEAF on GIDs

[LEAFer]

My "Week With GIDs" ;-) Or "Giddy Week" if you prefer.

Also cross-posted a link in a couple of other threads to this completed data report. And now it's "Geeks Go Hog Wild With Giddy Data". Have at it and figure out what you can. My apologies in advance as to the length of this post.

Borrowed a GID meter on Wednesday 11/7 and have been recording various daily events, including a Range Test today and a subsequent Turtle-to-100% Fill Up.

Background: This is an early LEAF with a relatively high odometer in the Sacramento Area, which can get hot. It's not Phoenix, but a couple of months this summer were pretty continuously hot. We have GeekEV that has reported a lost a capacity bar in this area, but our LEAF still has all 12. Summer 2011 was relatively mild here. However, after a hot 2012 August & September here it felt like (subjectively) that range was suffering. In October I performed two Turtle-to-100% runs and associated unscientific range report (here and here). The data seemed to imply a 15% range loss and the surprise was how could this LEAF *not* have lost its top capacity bar.

Having access to Tony's GID meter from the BC2BC return tour at the end of June gave the results shown in the signature (259). I have not yet found a post I (may have) made for that measurement. Given the 259 reading it was felt that significant loss due to August & September heat had occurred by the time October rolled around.

So ... here's the Giddy Data ... (large spreadsheet ... also posted on GoogleDocs here) ...

And now to today's unscientific Range Test. Ideally I could duplicate Tony's Phoenix Range Test, but for several reasons that's pretty much impossible without an inordinate amount of time/effort/coordination I don't have right now. Mostly, due to where we live (9 miles from freeway) it's hard to stick to a continuous 100km/h. Secondly I wanted to do an accurate subsequent fill-up to 100%. So, yes, I made some very specific compromises, but nevertheless came up with some useful info to complete the "Giddy Data".

Cold Tire Pressure: 40 psi (rather than 36), but ambient Temps much cooler (low 60s, rather than mid 80s (IIRC)). First 9 miles: 5 stop signs, 2 lights (both hitting red when I got there), but in-between: 62mph. Freeway: pretty much continous 100km/h (plus/minus); most of the time I was watching the geeky sensor data (Vgs=1000), and had to alternate on CruiseControl between 63mph (Vgs 988-992) & 64mph (Vgs=1006-1009). I tried hard not to draft or be too close to vehicles ahead. Negligible wind. Overcast. No HVAC use. NAV on, Radio off. At mile 33 did my freeway turn-around. At mile 52.5 rebooted the NAV (losing sensor data). At 60.8 miles exited freeway. Surface roads with 4 stop signs within a mile. Then back to 63mph on country roads. At 62.3 miles (47gids) LBW. At 64.8 (29 gids; with 7+ miles to go) I finally chickened out ( I am adverse to tows ) and killed the CruiseControl as I watched the GIDs drop quickly; apparently quicker than before. Now set CC=25mph. VLBW at 67.2 miles (24gids). Continued with CC=25mph. At 71.9 miles/16 gids reached my private road turn-off; continued at 25mph on CC. After a mile or so started varying the speed up to 35mph to reach powerlimits (lost power circles). This happened at 74.1miles/10 gids. Shifted to ECO (for the first time) and resumed 25mph. Reached home at 75.3miles/7 gids. With ECO mode still engaged, turned on heater; after 2 minutes turtle popped-up. Power-off.

Until the point of cancelling CruiseControl (2.5 miles past LBW) EE=4.0miles/kWh, Dist=64.8 Gids=29 AvgSpeed=43.6.

(Below spreadsheet is at GoogleDocs here.)

And here are two items from Carwings. I don't have the latest updates, so I believe the data is inaccurate, but I am providing it FWIW. Notice the discrepancies in EE (EnergyEconomy, miles/kWh) between CW and the dash. Also notice the missing miles due to Nav Reboot (I was watching the Sensor data) at 52.5 miles. The GoogleDocs version is here.

Edit 16Nov2012 10:14: The kWh data comes from a CT-500 tied into a ChargePoint Network account and is believed to be trustworthy. Charging typically at 242V (239-244), topping out somewhere between 3.68-3.70kW. The OAT data is from the LEAF and relatively unreliable in cases where the car has been at rest for an extended period of time -- see time stamps recorded and notes in the Comment field.

Edit 17Nov2012 08:10 GID meter is being returned today (Saturday). Final data points above. I will add a few edits to clarify or pontificate ;-)

Edit: 19Nov2012 10:48: Added Carwings Data.

 

[TonyWilliams]

I didn't see the final miles/kWh, or is that "EE" of 4.0?

Until the point of cancelling CruiseControl (2.5 miles past LBW) EE=4.0, Dist=64.8 Gids=29.

It seems unlikely that Gids were 29 at 2.5 miles past 47 Gid.

Ok, 62.3 miles, 100% to 17.4% @ 4.0

62.3 / 4 = 15.5 kWh burned, with 3.1kWh remaining to turtle.

15.5 + 3.1 = 18.6 kWh

18.6 / 21 = 88.5% battery capacity

 

[LEAFer]

I didn't see the final miles/kWh, or is that "EE" of 4.0?

Yes.

It seems unlikely that Gids were 29 at 2.5 miles past 47 Gid.

I know. But that's what I wrote down while the car was slowing from 62 to 25. The Gids were starting to drop awfully fast at that speed. Unfortunately I did not reach down to check the Battery Voltage. But I had about 7 miles to go, and was not in a good area (or mood ) for a tow.

Ok, 62.3 miles, 100% to 17.4% @ 4.0

I did not record %, but at 47 it should be 16.7%, right ? That would change the numbers a little ...

My Turtle Voltage was 303.5 ... more data later this evening.

 

[TonyWilliams]

Ok, 62.3 miles, 100% to 17.4% @ 4.0

I did not record %, but at 47 it should be 16.7%, right ? That would change the numbers a little ...

My Turtle Voltage was 303.5 ... more data later this evening.

Yes, 47 Gid would be 16.7%, however we've been indexing LBW to 17.4%.

What's 0.7% between friends?

 

[LEAFer]

I uploaded the "Giddy Data" ... basically taking readings several times a day before departure and at arrival and after certain charging events (like 80%, but then again when boosted to 100%).

I have not yet created the GoogleDoc from the spreadsheet. Will do that tomorrow or after taking final data Saturday morning.

Geeks: on your marks ... get set ... GO !

 

[DaveinOlyWA]

ok. so you were charging to 256-259 and then did range test with 246? and preheated first? not questioning just verifying my understanding of your data

 

[LEAFer]

ok. so you were charging to 256-259 and then did range test with 246? and preheated first? not questioning just verifying my understanding of your data

Nope. Luck of the draw ... co-incidence that today was to be Range Test ? This morning was the first morning of significantly lower GID (and 9 bars). I did pre-heat, but it was also to see if I could get a higher GID count; unfortunately that did not happen. 100% charge finished at 10:27. And except for an extra TB and extra 0.5V the rest of the results of the "top-off" were the same.

I should add: the pre-heat only ran for 5 minutes (to 11:48), because I started prepping the car for Range Test and possible tow. Checking tire pressures and inflating to 40psi with the compressor in the back, etc. The TimerOverride for 13 minutes of charging did not help the Gid count.

Just finished adding the Turtle-to-100% data in the OP ...

 

[DaveEV]

The data points to somewhere around 16% capacity loss according to my calcs:

75.3 mi / 4.3 mi/kWh = 17.5 kWh usable or 83.4% of 21 kWh

20.763 kWh turtle to 100% from the wall / 25 kWh (assuming 84% charging efficiency) = 84%

The data matches up - good job recording everything. Am surprised you haven't lost the first capacity bar yet.

Also looks like one can assume that if the voltage reads 394V the pack is fully charged and well balanced.

 

[surfingslovak]

Wow, this is fantastic! Good job collecting all the data.

 

[DaveinOlyWA]

The data points to somewhere around 16% capacity loss according to my calcs:

75.3 mi / 4.3 mi/kWh = 17.5 kWh usable or 83.4% of 21 kWh

20.763 kWh turtle to 100% from the wall / 25 kWh (assuming 84% charging efficiency) = 84%

The data matches up - good job recording everything. Am surprised you haven't lost the first capacity bar yet.

Also looks like one can assume that if the voltage reads 394V the pack is fully charged and well balanced.

gid count suggests he has only lost about 10%

i ran a range test last winter and only accessed 18.88 Kwh (100.1 miles @5.3 miles/Kwh...wish i would have recharged to 100% but didnt. only charged 3 hours and had to run off and do something and gas?? not an option!) and when i posted it, the temperature chart said i should lose about 8% so instead of accessing 21.5 Kwh i should have seen 19.78 Kwh so i was short a Kwh or so somewhere.

doesnt mean much but that is still nearly 5%. so am i to think i was 5% down 8 months ago? well, ya...figured after more than 13 months at the time, should have lost something... but have to think the miles/kwh is a bit inflated but also dont think that the gauge is the same across all vehicles. too many people who drive like i do (or at least they claim to) to the same places i do and they get MUCH lower values than i do. they keep asking me how i do it and i have growing evidence that i am not. my car is lying to me... not a big lie mind you... a few tenths here, a few tenths there...

 

[edatoakrun]

Excellent work, LEAFer.

Particularly useful are your reports of recharge times vs "wall", which is from a meter, I assume?

This allows all of us without wall meters an additional reference point to our own recharge times.

I almost wish I had done all my range tests to turtle, rather than VLBW, just so I could see my own battery's future (14,000+ miles so far) more clearly.

Have you seen any increase in gid counts @ "100%" from summer lows that some others have reported?

The data points to somewhere around 16% capacity loss according to my calcs:

75.3 mi / 4.3 mi/kWh = 17.5 kWh usable or 83.4% of 21 kWh

20.763 kWh turtle to 100% from the wall / 25 kWh (assuming 84% charging efficiency) = 84%

The data matches up - good job recording everything. Am surprised you haven't lost the first capacity bar yet.

Also looks like one can assume that if the voltage reads 394V the pack is fully charged and well balanced.

I don't know how you can take m/kWh reports of energy use very seriously after the Phoenix testy (apparently) showed the same very large levels of inaccuracy I believe I have observed since last Summer from range tests my own LEAF.

In terms of the recharge calculation, there are, IMO, still the major variables of charge efficiency and capacity at variable temperature .

If LEAFer's recharge efficiency was within the plausible range, which IMO is probably between about 86% and 90%, from a "100%" L2 charge, his total capacity loss from a nominal "new" ~21 kWh available capacity at his recharge temperature, is within the range of ~10% at 90% efficiency, and ~14% capacity loss at 86% efficiency, correct?

And so, no bar loss would be expected, yet...

Since we don't know how much LEAFer's battery temperature varied (I am presuming it was somewhat lower) from the one that that showed the 21.381 kWh test result, we don't know how valid the 21 kWh nominal capacity I used for in the example above is.

http://www1.eere.energy.gov/vehiclesandfuels/pdfs/merit_review_2012/veh_sys_sim/vss030_lohsebusch_2012_o.pdf" onclick="window.open(this.href);return false;

IMO, if both charges could be confirmed to have occurred at the test temperature which resulted in the 21.381 kWh capacity, the capacity loss range would be more like between ~12% and ~16%.

 

[LEAFer]

I updated the Giddy Log with this morning's data. Apparently a battery balancing "act" was performed between 04:28 - 04:53. These times are from the CT-500. Be aware that they are not sync'ed to real-time, could be off by five minutes, but in this case it really does not matter. For that matter, the times that "Charge Complete" and other CW messages come in at are also not sync'ed to "real time".

Anyway, the ChargePoint Network account has a graph which reports (at 5 minute intervals) the following for this morning:

04:28    0 kW
04:33 2.53 kW
04:38 1.53 kW
04:43 1.21 kW
04:48 0.72 kW
04:53    0 kW

With an additional 0.426kWh added during the balancing act. So the total Turtle-to-100% is now 21.189 kWh. This compares to 20121013=20.722; 20121025=21.049 (see signature).

As to some of the other questions (sorry ... multi-quoting each is a pain):

I did at some point in the past check on the accuracy of the CT-500 by tracking it against the SMUD Utility meter. Within the accuracy provided by it (1 kWh) the results matched, so I am inclined to trust the info provided by the CT-500. Also, in at least one of the October Turtle Runs I recorded amps and volts per the SMUD meter (there's a trick you can use to display this). Our house is typically over-volted and runs in the 242-244V range generally.

The 4.3 miles/kWh: I guess it's as valid (or inaccurate) as any of the LEAF displays. The reason Tony used 4.0 in his first reply to the OP is that is what I recorded at the point of "chickening out" at 64.8 miles. Obviously the significant slow-down to 25mph (despite only an additional 10.5 miles driven) made a big difference (increasing EE from 4 to 4.3). But it's really that sudden fast drop in Gid count that caught me off guard. What I mean to say, either 4.3 or 4.0 (with the respective adjustments) could be used for calculations, but be aware of the conditions and that either number could be suspect.

Efficiency of charging: The CT-500 is Level 2, of course, at relatively high voltage (a little over 240). I hope to have a chance tonight to wrap up data collection with a L1 charge just for comparison sake.

Gid Count versus Turtle Run: Yes, I agree. There seems to be some sort of discrepancy here. The surprise was the high Gid count, even reaching 262, which is higher than end of June. The subjective Range Loss that I attributed to hot August/September certainly does not manifest itself in the Gid count. But it does show up in the Turtle Runs (Turtle or Dead to 100% charging). Too bad I never before did any Turtle Runs (at delivery or summer of 2011, for example).

I had seen many or almost the majority of 80% charges (which we don't do too often) to only 9 bars until November. Then obviously with the exception of Nov15 we are seeing 10 bars again. Why only 9 bars (and 201 gid count) on the day of the Range Test ? Who knows. It recovered again a day later.

A note about "OAT" (Outside Air Temp): these data points are from the LEAF's eyebrow display, and obviously very suspect for most of the data points because the car was at rest and experienced a significant change in actual AirTemp while not sensing it properly. I tried to comment on this in the Comment field. In terms of how battery temperature was affected we need to guess a little and take into account the timestamps I recorded. This allows us to estimate where the battery might be at (after being cold-soaked for example).

Notice how on Nov15 TB=4 rises to TB=5 while just sitting near noon (in the shade with a mostly overcast day). Having just charged from 80% to 100%, and using pre-heat, of course, helped that too. Trick (Nissan) Question: So what was the actual Battery Temp at that point ?

Edit: I updated the OP with an edit at the bottom and an addition "AvgSpeed 43.6" at 64.8 miles, so that the AvgSpeed can be compared to the one recorded at the end of the Range Test.

 

[LEAFer]

GID meter is being returned today (Saturday). Final data points above in the OP on L1 charging with a Kill-A-Watt. I will add a few edits to clarify or pontificate ;-) after our Folsom Meetup.

Keep fingers crossed for no rain 10am - 11am, but bring extra umbrellas anyway

 

[LEAFer]

I added some data from Carwings. I don't have the latest CW updates, so I believe the data is inaccurate, but I am providing it FWIW. Notice the discrepancies in EE (EnergyEconomy, miles/kWh) between CW and the dash. Also notice the missing miles due to Nav Reboot (I was watching the Sensor data) at 52.5 miles.

 

[LEAFfan]

The data points to somewhere around 16% capacity loss according to my calcs:

75.3 mi / 4.3 mi/kWh = 17.5 kWh usable or 83.4% of 21 kWh

20.763 kWh turtle to 100% from the wall / 25 kWh (assuming 84% charging efficiency) = 84%

The data matches up - good job recording everything. Am surprised you haven't lost the first capacity bar yet.

I'm not. If the loss were really 16%, then a bar would be missing. A bar will disappear when the Gid reading hits 80-81%. I've tested many LEAFs, and everyone lost a bar at 80-81%.

 

[DaveEV]

I'm not. If the loss were really 16%, then a bar would be missing. A bar will disappear when the Gid reading hits 80-81%. I've tested many LEAFs, and everyone lost a bar at 80-81%.

You are comparing the results of hot-climate cars with cool-climate cars. While it does seem to take a 80-81% GID reading to trigger the loss of the 12th bar, we all know that cool temperatures result in higher GID counts.

It is my belief that cool-climate cars will have lost more actual capacity than hot-climate cars when the 12th capacity bar is first lost (when tested under the same conditions).

 

[ht2]

I think very good news is last June your 100% GID count was 259 and after hot summer your 100% GID count is still 256-262.
No lost capacity during hot summer in hot(Compare to SF Bay area) Sacramento.

 

[LEAFer]

I think very good news is last June your 100% GID count was 259 and after hot summer your 100% GID count is still 256-262.
No lost capacity during hot summer in hot(Compare to SF Bay area) Sacramento.

It appears that way ... based only on GID count. But the Turtle-to-100% tells a different story. That discrepancy alone (valuable info) is worth something

 

[ht2]

I think very good news is last June your 100% GID count was 259 and after hot summer your 100% GID count is still 256-262.
No lost capacity during hot summer in hot(Compare to SF Bay area) Sacramento.

It appears that way ... based only on GID count. But the Turtle-to-100% tells a different story. That discrepancy alone (valuable info) is worth something

Do you have June's Turtle-to-100% data?
Best guess usable capacity of your battery is 21kWh x 259/281 = 19.35kWh
Turtle-to-100% of 20.7kWh to 21.2kWh looks normal. (~10% loss)

Anyway thanks for good news to all bay area Leaf owner.

 

[LEAFer]

Do you have June's Turtle-to-100% data?

Unfotunately ... no.

I think very good news is last June your 100% GID count was 259 and after hot summer your 100% GID count is still 256-262.
No lost capacity during hot summer in hot(Compare to SF Bay area) Sacramento.

It appears that way ... based only on GID count. But the Turtle-to-100% tells a different story. That discrepancy alone (valuable info) is worth something

Best guess usable capacity of your battery is 21kWh x 259/281 = 19.35kWh
Turtle-to-100% of 20.7kWh to 21.2kWh looks normal. (~10% loss)

I can agree with the 19.35kWh ... since it's simply based on Gid count ... 259/281 = 92.2%.
But the Turtle-to-100% values should really be compensated by the efficiency of L2 charging. The efficiency I accept for that case is 87%. So ...
20.722 * 0.87 = 18.03 --> or 85.8% of 21kWh.
21.049 * 0.87 = 18.31 --> or 87.2% of 21kWh.
21.189 * 0.87 = 18.43 --> or 87.8% of 21kWh.

I know it's just a few percent (92.2% versus the 85.8-87.8), but taking into account the surprising post-summer loss of range (in the October runs) and the L1 charging time estimates (much less than 25 hours), I am sure there's quite some degradation going on, very close in fact to losing the first bar.

Does anyone have a method (without Consult) to tell if our battery pack has a possible bad module or more ? (Recall the sudden drops in Gid count in the Nov 15 Turtle run.)