2016 Spark EV battery size/usable capacity....what the

[Zoomit]

MYSTERY SOLVED!! Well, maybe.

I got in contact with one of the moderators on the gm-volt.com forum (he works for GM...probably some kind of engineer), and he told me the 2015 and 2016 Spark EV batteries have 19.44 kWh total energy. He said the press releases of 18.4 and 19 kWh were probably "typos".....so GM never bothered to correct them? Wtf?

He also said under ideal conditions, the battery can exceed 20 kWh under ideal charging conditions.

So assuming the 15/16 Spark batteries have about 18.5 kWh usable, that still means almost 95% of the total battery capacity is utilized.

I'd like corroborating evidence on the 19.44 kWh figure, as I don't believe they just had typos and let them continue for 2 years.

Nevertheless, be skeptical of the "energy details" display data, as it varies day-to-day. I purposefully ran the car to empty this afternoon to get this image. The car was unable to move when I took this picture and the battery was effectively dead. For reference, the car is a '15 with just under 13,000 miles.

 

[MrDRMorgan]

MYSTERY SOLVED!! Well, maybe.

I got in contact with one of the moderators on the gm-volt.com forum (he works for GM...probably some kind of engineer), and he told me the 2015 and 2016 Spark EV batteries have 19.44 kWh total energy. He said the press releases of 18.4 and 19 kWh were probably "typos".....so GM never bothered to correct them? Wtf?

He also said under ideal conditions, the battery can exceed 20 kWh under ideal charging conditions.

So assuming the 15/16 Spark batteries have about 18.5 kWh usable, that still means almost 95% of the total battery capacity is utilized.

I'd like corroborating evidence on the 19.44 kWh figure, as I don't believe they just had typos and let them continue for 2 years.

Nevertheless, be skeptical of the "energy details" display data, as it varies day-to-day. I purposefully ran the car to empty this afternoon to get this image. The car was unable to move when I took this picture and the battery was effectively dead. For reference, the car is a '15 with just under 13,000 miles.

Your display data yields a battery capacity calculation of 17.68 kWh (17.5 kWh / 99%). My average battery capacity calculation for 26 charging sessions over the last 4 months is 17.61 kWh. My 2015 Spark EV 2LT had 10,412 miles on the ODO as of the last charging session on June 14, 2016. Looks like very good correlation.

 

[Zoomit]

To be a little pedantic about it, you cannot divide by .99 to get 17.68. The battery was dead. There is likely rounding error in the three numbers you're adding to get 99%. In reality it was at 100% since the car wouldn't move. Getting rid of that uncertainty is why I took it to the dead point instead of extrapolating from a partial charge state.

I think all we can say is that day, that battery put out 17.5 +/-0.05 kWh.

Still pretty close correlation with your numbers.

 

[bro1999]

MYSTERY SOLVED!! Well, maybe.

I got in contact with one of the moderators on the gm-volt.com forum (he works for GM...probably some kind of engineer), and he told me the 2015 and 2016 Spark EV batteries have 19.44 kWh total energy. He said the press releases of 18.4 and 19 kWh were probably "typos".....so GM never bothered to correct them? Wtf?

He also said under ideal conditions, the battery can exceed 20 kWh under ideal charging conditions.

So assuming the 15/16 Spark batteries have about 18.5 kWh usable, that still means almost 95% of the total battery capacity is utilized.

I'd like corroborating evidence on the 19.44 kWh figure, as I don't believe they just had typos and let them continue for 2 years.

Nevertheless, be skeptical of the "energy details" display data, as it varies day-to-day. I purposefully ran the car to empty this afternoon to get this image. The car was unable to move when I took this picture and the battery was effectively dead. For reference, the car is a '15 with just under 13,000 miles.

Your display data yields a battery capacity calculation of 17.68 kWh (17.5 kWh / 99%). My average battery capacity calculation for 26 charging sessions over the last 4 months is 17.61 kWh. My 2015 Spark EV 2LT had 10,412 miles on the ODO as of the last charging session on June 14, 2016. Looks like very good correlation.

Here's the full I received from that mod:
"Both 2015 and 16 have 19.44kWh total energy (I believe merely a "typo" in the original 2015 spec sheet)

This is derived by a 54Ah 2P capacity multiplied by an open terminal voltage of 360V (nominal)

In fact it can exceed 20kWh when charging voltages above nominal approaching 400V under ideal conditions"

 

[bro1999]

MYSTERY SOLVED!! Well, maybe.

I got in contact with one of the moderators on the gm-volt.com forum (he works for GM...probably some kind of engineer), and he told me the 2015 and 2016 Spark EV batteries have 19.44 kWh total energy. He said the press releases of 18.4 and 19 kWh were probably "typos".....so GM never bothered to correct them? Wtf?

He also said under ideal conditions, the battery can exceed 20 kWh under ideal charging conditions.

So assuming the 15/16 Spark batteries have about 18.5 kWh usable, that still means almost 95% of the total battery capacity is utilized.

I'd like corroborating evidence on the 19.44 kWh figure, as I don't believe they just had typos and let them continue for 2 years.

Nevertheless, be skeptical of the "energy details" display data, as it varies day-to-day. I purposefully ran the car to empty this afternoon to get this image. The car was unable to move when I took this picture and the battery was effectively dead. For reference, the car is a '15 with just under 13,000 miles.

Depending on how GM programmed the energy display to round (round up .5 and higher? Or just truncate the decimal, with 89.9% still displaying as 89%, etc...). If the energy usage % are merely truncated, that means there could be up to almost 3% margin of error (if each of the % readings is actually at x.9%, with the .9 being dropped).

Combined with the fact the energy usage display is merely an estimate (though pretty accurate one) of energy used, I think the total margin of error could up around +/- 5%.

 

[Zoomit]

Here's the full I received from that mod:
"Both 2015 and 16 have 19.44kWh total energy (I believe merely a "typo" in the original 2015 spec sheet)

This is derived by a 54Ah 2P capacity multiplied by an open terminal voltage of 360V (nominal)

In fact it can exceed 20kWh when charging voltages above nominal approaching 400V under ideal conditions"

You may be on to something. This seems plausible. The only other similar reference I can find is in this article comment from 2014:

"Actually the Spark EV pack is 54Ah using a 2P96S configuration."
http://insideevs.com/2015-chevy-spark-ev-highlighting-changes/#comment-478095.

I don't know what the "2P96S configuration" refers to, but it's awfully specific.

The 2016 Spark EV press release only says "19 kWh", which of course makes more sense if the battery was 19.44 instead of 18.4.

 

[MrDRMorgan]

To be a little pedantic about it, you cannot divide by .99 to get 17.68. The battery was dead. There is likely rounding error in the three numbers you're adding to get 99%. In reality it was at 100% since the car wouldn't move. Getting rid of that uncertainty is why I took it to the dead point instead of extrapolating from a partial charge state.

I think all we can say is that day, that battery put out 17.5 +/-0.05 kWh.

Still pretty close correlation with your numbers.

I completely agree with you but I can only use the battery usage information available to me which, in this case, is the car's display. My interest is not the battery's "actual" capacity. Rather, my interest is to learn the rate of battery capacity degradation over time and/or mileage. Knowing the rate of battery degradation will allow me to determine how many years I can expect the car to meet my driving range needs before I have to replace the battery or the car. Kevin's graph shown in an earlier post in this section shows graphically what I mean.

 

[Zoomit]

I completely agree with you but I can only use the battery usage information available to me which, in this case, is the car's display. My interest is not the battery's "actual" capacity. Rather, my interest is to learn the rate of battery capacity degradation over time and/or mileage. Knowing the rate of battery degradation will allow me to determine how many years I can expect the car to meet my driving range needs before I have to replace the battery or the car. Kevin's graph shown in an earlier post in this section shows graphically what I mean.

Ah right, in that case, we're forced to always assume the sum of the three numbers represents the battery discharge state. While we recognize there's rounding error inherent in that sum, we take the sum as truth to be able to create a long-term trend.

As bro1999 stated above, that error could be +3%/-0% if the data is truncated, or +/-1.5% if rounded. But those extreme errors are statically unlikely since they are the sum of three independent calculations. (One could be high by 0.5% and be canceled out by another that's low by 0.5%.) We can expect the error of the sum to "regress to the mean" for a large sample size. If the display rounds the data, the error goes to zero. If the display truncates, the error creates a +1.5% bias. That's a 0.27 kWh difference.

So in summary, if you collect a lot of data, a rounding error goes away. If they truncated the display data and you collect a lot of data, you need to bias the results up by 0.27. Either way, if you're just looking for the long-term trend, you can ignore these errors if you just collect a lot of data!

 

[SparkevBlogspot]

2P = 2 parallel, 96S = 96 series. It's how the cells are arranged. 96*2 = 192 cells total, but 96 in series.

Since each cell is 3.7 V nominal, but could get close to 4 volts (or even more), total voltage could be 355 to 384 volts. I've actually seen it close to 400 volts at San Clemente EVPump DCFC, which would mean 4.1 volts per cell.

I don't know if 18.4 kW or 19.44 kW or whatever is nominal or peak. Assuming AH rating is constant, we can get various values, all of them "correct" in their own context.

355*54 = 19.2 kWh (nominal)
400*54 = 21.6 kWh ("peak")

Based on what I saw at DCFC close to 400 volts with 2015 SparkEV, and assuming 54 AH is correct, 2015 SparkEV has close to 21.6 kWh at peak, 19.2 kWh nominal, 18.4 kWh peak usable, 16.5 kWh extrapolating from DCFC reported.

 

[Zoomit]

Based on what I saw at DCFC close to 400 volts with 2015 SparkEV, and assuming 54 AH is correct, 2015 SparkEV has close to 21.6 kWh at peak, 19.2 kWh nominal, 18.4 kWh peak usable, 16.5 kWh extrapolating from DCFC reported.

Good info, thank you.

I followed everything until this: "18.4 kWh peak usable, 16.5 kWh extrapolating from DCFC reported."

Where did you get 18.4? and how is that different than 16.5?

 

[MrDRMorgan]

I completely agree with you but I can only use the battery usage information available to me which, in this case, is the car's display. My interest is not the battery's "actual" capacity. Rather, my interest is to learn the rate of battery capacity degradation over time and/or mileage. Knowing the rate of battery degradation will allow me to determine how many years I can expect the car to meet my driving range needs before I have to replace the battery or the car. Kevin's graph shown in an earlier post in this section shows graphically what I mean.

Ah right, in that case, we're forced to always assume the sum of the three numbers represents the battery discharge state. While we recognize there's rounding error inherent in that sum, we take the sum as truth to be able to create a long-term trend.

As bro1999 stated above, that error could be +3%/-0% if the data is truncated, or +/-1.5% if rounded. But those extreme errors are statically unlikely since they are the sum of three independent calculations. (One could be high by 0.5% and be canceled out by another that's low by 0.5%.) We can expect the error of the sum to "regress to the mean" for a large sample size. If the display rounds the data, the error goes to zero. If the display truncates, the error creates a +1.5% bias. That's a 0.27 kWh difference.

So in summary, if you collect a lot of data, a rounding error goes away. If they truncated the display data and you collect a lot of data, you need to bias the results up by 0.27. Either way, if you're just looking for the long-term trend, you can ignore these errors if you just collect a lot of data!

EXACTLY! I really like learning all I can about the Spark EV but, in the end, it comes down to how far and how long can I drive this car before I incur a major expense like having to replace the battery - if it is even available.

 

[SparkevBlogspot]

Where did you get 18.4? and how is that different than 16.5?

18.4 kWh is based on various forum posts as well as my experiment. I got bit over 18 kWh when I was measuring this in early days, so I figure the peak must be 18.4 kWh that many places show. Origin is probably Chevy.

16.5 kWh is based on extrapolating DCFC. Just yesterday, it charged from 13% to 85% at 12.09 kWh. Doing the math shows about 16.8 kWh for 100%. This is roughly what I got when the car was new from time to time, although I also got as high as 18.42 kWh. Not sure why it varied so much. Then this is conservative low end estimate of battery capacity when using DCFC. Since GOM is about the same at given % as when the car was new, I suspect the battery did not degrade (much), but just measurement or per-charging incidence difference. But just in case, I'm going to pay closer attention to DCFC capacity.

A note about peak at close to 400 volts: that's only when charging. Once the battery gets off the charger, it drops the voltage pretty quickly, as you can see from other rechargeable batteries. Then that "peak" is not really meaningful, and nominal would be more accurate figure.

 

[Zoomit]

18.4 kWh is based on various forum posts as well as my experiment. I got bit over 18 kWh when I was measuring this in early days, so I figure the peak must be 18.4 kWh that many places show. Origin is probably Chevy.

Ok, thanks. My data averages between 17.5-18.0 kWh for usable battery capacity, even including when new. Individual data points range from about 16.5 to 18.3 kWh.

I'm going to collect a few more months of data, to make sure I isolate any seasonal variation, then will write it up.

 

[MrDRMorgan]

18.4 kWh is based on various forum posts as well as my experiment. I got bit over 18 kWh when I was measuring this in early days, so I figure the peak must be 18.4 kWh that many places show. Origin is probably Chevy.

Ok, thanks. My data averages between 17.5-18.0 kWh for usable battery capacity, even including when new. Individual data points range from about 16.5 to 18.3 kWh.

I'm going to collect a few more months of data, to make sure I isolate any seasonal variation, then will write it up.

I leased my 2015 Spark EV 2LT at the end of May in 2015. During June and July last year (2015) I collected data from 11 charging sessions (1000 miles total). The average battery capacity calculation was 18.61 kWh. During Mar - Jun of this year (2016), I collected data for 27 charging sessions (2000 miles total). My average battery capacity for 2016 is 17.62 kWh. Currently, my ODO shows 10454 miles.

 

[SparkevBlogspot]

I added to my blog post with new findings. Scroll down to "Update 2016-06-23" for latest data.

http://sparkev.blogspot.com/2015/05/battery-capacity-estimate.html

It seems 2015/2016 SparkEV would be bit over 20 kWh if 54 Ah comment is true.