[TomNZ]

Hi all!

I’ve had my 2020 mid trim e-niro here in NZ for some months and 5000km, mostly charged from trickle wall plug. I’ve noticed a few times that the percentage battery used seems higher than the reported efficiency. As an example, yesterday’s 31km drive took me from 95% to 85% (~95kph, with a couple of bikes on the back)

Reported History for the day: 31km 14.7kwh/100km

Two calculations for the range based on this drive yield quite different results.

1. Using observed % battery usage: 31/(95-85)*100 = 310km
2. Using reported efficiency and expected battery capacity: 64/14.7*100 = 435km

Wow, massively different! Any thoughts on what could be causing this and any next steps? I’ve taken it to the dealer once and they fobbed me off.

Cheers,
Tom

 

[Greg Elmassian]

All the numbers are estimates, never accurate. You will get a "feel" for your range, and I use the battery % as a rough guide.

This must be the most common source of angst with new owners.

My experience is that 10% difference in battery charge is in the margin of error.

Greg

 

[KiwiME]

... I’ve taken it to the dealer once and they fobbed me off.

They won't have any idea as what you're even talking about. But, welcome fellow kiwi. I have a technically-similar Kona and I'll suggest that my dashboard history numbers tend to be erroneous at low km values much as you've described. I don't know why that is and I pretty much don't care, nor do I use the GoM.

I also tend to stick to relating kms driven to SoC drop as those two sources of data are more direct and consistent. I'll allow about 21% SoC loss per 100km as rule for planning trips. Obviously the bikes on the back are going to increase that, as does lower ambient temperatures and an increase in destination elevation. Once you get a feeling for the numbers you'll have more data to base estimates on. I do Napier-Taupo return quite a lot and always record the outcome and it's often 5% off due to differing AC use and/or speed.

If you're a technical type I would recommend learning about Torque Pro to log data that can be plotted in Excel.

 

[KiwiME]

Two calculations for the range based on this drive yield quite different results.
1. Using observed % battery usage: 31/(95-85)*100 = 310km
2. Using reported efficiency and expected battery capacity: 64/14.7*100 = 435km

Tom, looking at my own numbers with 3 people in the Kona, speeds 80-100 km/h, ambient 20-25°C and assuming SoC linearly represents 64kWh as you've done:
Napier-Taupo 95 to 64% no AC, + 357m elev, 143.7 km, usage is 13.8 kWh/100km while the dash readout reported 14.5
Taupo-Napier 62 to 29% AC on, - 357m elev, 141.5 km, usage is 14.9 kWh/100km, dash reporting 13.3

The numbers all over the place and the linearity of SoC is uncertain. The percent usage both ways was quite different this time than numerous nearly-identical previous trips, noting that I've recently had a BMS update that erases all historical battery use data.
The best quality data from the car IMO is the odometer and the change in CEC and CED values that you can read with several OBD-II smartphone apps.

 

[Richard Wood]

Hi all!

I’ve had my 2020 mid trim e-niro here in NZ for some months and 5000km, mostly charged from trickle wall plug. I’ve noticed a few times that the percentage battery used seems higher than the reported efficiency. As an example, yesterday’s 31km drive took me from 95% to 85% (~95kph, with a couple of bikes on the back)

Reported History for the day: 31km 14.7kwh/100km

Two calculations for the range based on this drive yield quite different results.

1. Using observed % battery usage: 31/(95-85)*100 = 310km
2. Using reported efficiency and expected battery capacity: 64/14.7*100 = 435km

Wow, massively different! Any thoughts on what could be causing this and any next steps? I’ve taken it to the dealer once and they fobbed me off.

Cheers,
Tom

You're falling into the same trap as me (and no doubt others) in thinking that the SOC % reflects kWh, i.e. 10% = 6.4kWh. The puzzling discrepancy led me to discover that it's amp-hours (180Ah total) and the battery voltage drops from around 405V at 100% to maybe 320V at 0 (never taken it that low). Only 40-50% = 6.4kWh (nominal 356V). The top 10% is roughly 7.2kWh and the bottom 10% maybe only 5.9kWh. You'll find that the reported figures are pretty accurate. Here's my plot of the battery curve: Battery graph.pdf

 

[Todor]

You're falling into the same trap as me (and no doubt others) in thinking that the SOC % reflects kWh, i.e. 10% = 6.4kWh. The puzzling discrepancy led me to discover that it's amp-hours (180Ah total) and the battery voltage drops from around 405V at 100% to maybe 320V at 0 (never taken it that low). Only 40-50% = 6.4kWh (nominal 356V). The top 10% is roughly 7.2kWh and the bottom 10% maybe only 5.9kWh. You'll find that the reported figures are pretty accurate. Here's my plot of the battery curve: Battery graph.pdf

Never though about it like that, but now it make sense. The battery on the dash has 18 segments, i.e. each segment is 10Ah.

 

[Richard Wood]

Never though about it like that, but now it make sense. The battery on the dash has 18 segments, i.e. each segment is 10Ah.

Ha, I hadn't twigged that.

 

[DragonQ]

Does this mean that when the car reports 25% SoC, for example, it actually needs more than 75% x 64kWh = 48kWh to reach 100% SoC?

 

[Richard Wood]

Does this mean that when the car reports 25% SoC, for example, it actually needs more than 75% x 64kWh = 48kWh to reach 100% SoC?

By my calculations you'd need to add approximately 50.3kWh to the battery, which, allowing for typical charging loss, would mean maybe 59kWh ex grid or 8.5h on a 7kW home charger. It would be interesting to see what the car estimates for charge time.

 

[DragonQ]

By my calculations you'd need to add approximately 50.3kWh to the battery, which, allowing for typical charging loss, would mean maybe 59kWh ex grid or 8.5h on a 7kW home charger. It would be interesting to see what the car estimates for charge time.

Hmm that's interesting because I'm trying to track how much energy I'm using to charge my car and the other day it took 49 kWh (as reported by my charge point) to get from 27% to 100%. That doesn't add up with your calculations unless there was no loss at all from the grid (impossible).

 

[Richard Wood]

Hmm that's interesting because I'm trying to track how much energy I'm using to charge my car and the other day it took 49 kWh (as reported by my charge point) to get from 27% to 100%. That doesn't add up with your calculations unless there was no loss at all from the grid (impossible).

I must admit I've had some puzzling figures, too. Something funny seems to happen at the high SoC states when charging. Mid-range charges all make sense: kWh ex grid > kWh to battery > estimated battery energy kWh increase. I recently charged from 85% to 100% -- theoretically 10.8kWh but the BMS recorded only 9.9kWh fed to battery and only 24.7Ah as opposed to 27Ah for 15% (read using Soul EV Spy). A similar charge was 79.5% to 100% (BMS records every 0.5%), which should have been 14.4kWh and the BMS said only 13.2kWh to battery and only 32.9Ah as opposed to 37-ish theoretically. This is odd as the discharge situation seems to make sense at any SoC difference..

This whole voyage of discovery was sparked by my wife doing 100 miles on the top 30% of the battery and the car reporting only 4.7 miles/kWh and me driving the same leg on the return with the battery getting low (40% left, easy peasy), getting similar miles/kWh but, alarmingly, using around 35% of the battery. 3x6.4kWh =19.2kWh should have been 5.2 miles/kWh. Why such a huge discrepancy? However, adding up my theoretical top 30% energy after discovering the Ah correlation and the way the voltage climbs makes 21.21kWh,, giving 4.71 miles/kWh, i.e. agreeing with the car's calculation.

 

[KiwiME]

You're falling into the same trap as me (and no doubt others) in thinking that the SOC % reflects kWh, i.e. 10% = 6.4kWh.

In the case of the 64kWh Kona my data shows that SoC does indeed reflect battery energy output, at least down to 30%. It's certainly clear from the schematics that the BMS's instrumentation is capable of this so I don't know why it would do otherwise. It's the product of volts and amps delivered to the motor windings as energy that gets you down the road, not just one or the other.

The data in the graph below was collected July 2020 (winter) from the same round trip I mentioned in post #4. The net battery output values are the difference of energy odometers CED and CEC (offset to zero at the start of the trip), where CEC is the result of regen. I've laid a linear trendline as a reference over the 2793 data points. The stepping is primarily due to the 0.5% resolution of SoC.

I'm not entirely clear as to how all this works but I believe that the SoC display is calculated primarily from these instantaneous energy values adjusted by efficiency information derived from the accumulated historical data and the initial factory SoC. Because I've ignored those adjustments the curve has a slight distortion due to the trip destination being at some elevation. Regen would be significantly higher on the return leg.

I'll just add that the slope of the trendline inverted and multiplied by 100 is 63.6.