The mystery of the missing kWh or why you can't trust the state of charge percentage

[Newton]

Actual is based on the full battery size of 143kWh, including the buffers. Displayed is the usable amount.

They cross around 35% due to the top and bottom buffers.

That is what I believed, but the data does not seem to support that. If you take the kWh and divide it by the battery percentage at any given time you will obtain the battery size. If this were true, I should be seeing a constant size (whatever the true battery size is) with this calculation. Instead it is all over the place. In a way this is good because it indicates that the truck's idea of the battery size is based on some measurable parameters rather than just forcing it but it shows that these measurements are not particularly consistant.

Sponsored

 

[RickLightning]

That is what I believed, but the data does not seem to support that. If you take the kWh and divide it by the battery percentage at any given time you will obtain the battery size. If this were true, I should be seeing a constant size (whatever the true battery size is) with this calculation. Instead it is all over the place. In a way this is good because it indicates that the truck's idea of the battery size is based on some measurable parameters rather than just forcing it but it shows that these measurements are not particularly consistant.

No, you cannot do that math.

First, the amount of energy (kWh) varies by temperature. Battery needs to be at 77 degrees for that number to be accurate.

Second, there's uncertainty at the top and bottom of the packs due to accumulated errors over time.

When I have checked with CarScanner, I have never seen more than 127.xx. After I get it back from getting a new battery array, I intend to check it when it is 77 degrees and see what it now shows.

To be accurate, you would want to charge to say 90% at 77 degrees, take a reading, use the truck, and charge to 90% again at 77 degrees, and repeat.

 

[mr.Magoo]

I don't charge mine to 100% very often, but this is the highest "remaining capacity" I have recorded.

 

[Maxx]

I don't charge mine to 100% very often, but this is the highest "remaining capacity" I have recorded.

‘Wow your actual vs real SOC at 9:37 looks very different than this. That green line should be almost parallel to the yellow line in that short period of time.

 

[mr.Magoo]

‘Wow your actual vs real SOC at 9:37 looks very different than this. That green line should be almost parallel to the yellow line in that short period of time.

I wouldn't make too much of that, I don't have the graph available to double check right now (it resides in the black box in my truck) but I believe they're on different axis since they're different units plus it might not be a linear scale (I'm using logarithmic on some) to begin with.

 

[SpaceEVDriver]

That is what I believed, but the data does not seem to support that. If you take the kWh and divide it by the battery percentage at any given time you will obtain the battery size. If this were true, I should be seeing a constant size (whatever the true battery size is) with this calculation.

No. Well, sort-of yes, but not exactly. The misconception I think you're running into is that there is no "true battery size." The available energy contained by the battery and usable by the vehicle is not a fixed value. Sometimes the battery "size" might be 150 kWh, sometimes it's 143 kWh, sometimes it might be 100 kWh. A large part of that fluctuation is temperature dependent and another part of it is (dis)charge rate and yet another part is battery age.

We don't know the specific formulation Ford uses for its battery SoC and SoCD estimations, but one common approach to estimate SoC is in the images I've attached (for a completely different lithium cell). That equation looks linear, but it's not (voltages are temperature dependent and this is a piecewise linear approximation to a nonlinear curve). Because of the larger spread of acceptable kWh values for SoC, this equation would be used with different parameters for V_low, V_high, SoC_low, and SoC_high for SoC and SoCD.

(Note that the x-axis on this curve is backwards from what you might expect--this is how much energy has been extracted from the cell, not how much is remaining.)

 

[flyct]

Boy, what a bunch of geeks we are.

 

[RickLightning]

Boy, what a bunch of geeks we are.

 

[TheWoo]

Boy, what a bunch of geeks we are.

And I'm here for it. I've learned so much just by being curious about EVs. Glad there are lots of smart people who also like to share in ways we can understand.

Sponsored