Here is another recent thread about this:

I think its due to the SOC and mi/kwh displayed being a guess created by a math model based on voltage/current and coulomb counting. And it gets off and then re-adjusts itself. Similar to how after a deep cycle (re-calibration) many people report gaining up to 8kwh of MEC (max. energy capacity) in the HV: showing the BMS has changed its guess about where 0 and 100% are located. And it looks like it is able to do that kind of change at any point both driving or sitting.

Another factor that goes into the guess about SOC is the HV battery temperature. I have seen my SOC drop 4% just sitting between when I get home and its warm 70F, and it cools to 32F. So some adjustments to SOC are happening as the battery warms or less often cools while driving.

 

Thank you for the possible explanation. However, my point is that from 11% to 5%, it seems like the car was overestimating. What if I believed in it and then suddenly the car told me that it's actually out of battery? Would it be that I was an idiot to belive what the car told me initially?

I would be more comfortable if it underestimated but not the other way around.

Furthermore, I don't think I have seen this kind of behavior in my old '21.

 

Would it be that I was an idiot to belive what the car told me initially?

Yes. Never believe anything the car tells you about battery percentage.

 

Re calibrate the BMS. Charge to 100%, drive it down to 5%-10%. then charge back up to 100%. That should help the BMS know how much juice is left. I do not have any issues trusting the SOC. I do charge to 100% about every 3 months or so for long trips.

 

I'm maybe a little more concerned for your battery or BMS behavior than some of the other replies, and don't think your concern about % SOC being inaccurate this close to zero is misplaced.

I'm all on board with the displayed state of charge being a somewhat fluid estimate based on factors other than a direct voltage during operation. However, these estimates are more erroneous at mid-pack states of charge, where voltage drop is relatively flat, and the stakes aren't so high.

As the battery nears 0%, voltage "falls off" more rapidly, so the BMS – even when the pack is under load – should be able to double check itself with respect to how much it calculates should be left in the tank.

To see an immediate revision downward by 6% has me asking the same question as you: how would this have played out if you were white-knuckling it to the next DCFC station 20 miles further up the road.

I've seen my '21 BMS recalculate % SOC . But never by this much, never instantly at shutdown (I believe the car waits for the battery to cool to ambient to get an accurate voltage reading), and never at this low a state of charge – and I've confidently taken this car into single digits on a number of occasions.

 

But never by this much, never instantly at shutdown

Note this case was not at shutdown, it was after they started charging.

 

I get that, yet recalibration should also be happening realtime, especially at the far end when voltage drop is detectable while in use.

The calibration you're talking about is important for MEC and range calculations and needs stable voltages, a baseline pack temperature, and a recent delta between flat and full. But I don't believe any of this should be necessary for the countdown to indicated 0%, which should only get more accurate as the battery gets closer to actual 0%, particularly while the vehicle is in operation.

We should be able to have a solid level of trust that our indicated SOC is, if anything, conservative, including the hidden reserve buffer. A 6% overestimation at this low SOC is something I would find unacceptable and unsettling, and haven't observed in any of my EVs.

 

I get that, yet recalibration should also be happening realtime, especially at the far end when voltage drop is detectable while in use.

Large voltage drop is always happening whenever under load. Modeling the amount of voltage drop is the tricky part of SOC estimates, since it changes with temperature and SOC. Do you have any data to back up "these estimates are more erroneous at mid-pack states of charge " since I have never seen that. The drop gets less predictable at low SOC, and the tricky part is you don't want the estimate to jump around by 6% while driving, or even simply by parking for lunch and come back and it has dropped 6%. Luckily nobody has reported that yet.

 

Modeling the amount of voltage drop is the tricky part of SOC estimates, since it changes with temperature and SOC.

Assuming VW is continuing to use NXP for their BMS, that is their bread and butter: modeling SOC under load conditions. I don't expect it to be a perfect estimate, hence why we get the "revised estimates" after the car has been parked for a bit.

Do you have any data to back up "these estimates are more erroneous at mid-pack states of charge "

Data being just the generic battery discharge curves, where there's a minor cowlick at the 100% end where voltage spikes, and a steep tail-off at the bottom end, usually in the 10% range (which would equate to about 5% indicated in the ID.4, accounting for the bottom buffer). My "erroneous" comment coming from that it's difficult to know 50% from 55% or 45% when the resting voltage is close and flat, so coulomb counting provides some resolution, but not necessarily certainty, but the stake are lower if it's off by a few percent. There's much more certainty at the lower end when measuring the discharge tail, especially after the BMS "learns" the battery characteristics.

 

I would be interested in how the 80/84% above the 11/5% was driven. Was it the last leg of a highway trip? Was it a period of several/many around town trips?

I would definitely go through a calibration and watch for it to happen again.

 

I have the exact same situation with a new 2025 ID.4 Pro S AWD. No matter whether the car has set for some period of time or immediately plugged in. Happens whether Level 1, 2, or 3 charger. If charge is currently 60%, it will immediately drop to around 53%. I charge it to 70%, it'll stop at 70%. After several hours of sitting whether plugged in or not (typically occurs overnight at the house), the SOC will show 77%. The lower the starting SOC, it appears the larger variance. At 24%, it dropped to 14% and then charged to 70%. Couple hours later it was 80%. Took it to dealer yesterday and he duplicated the situation. Tech was a great guy, but I believe this whole SOC calculation is a mystery to everyone at VW. The car is not reporting any faults and he talked to VW Tech support and they said there's nothing wrong. Seems wrong to me. Asked about doing a recalibration, he was totally unfamiliar with it and had never been trained on anything like it. He asked me to try it and let him know. Question on the process. Does it matter what level charger you use to go up to 100% Otherwise, this car has been amazing.

 

Welcome to the forum.

Does it matter what level charger you use to go up to 100%

No. L2 is the best since DCFC becomes too slow for most to want to sit there that long. If you get a $30 dongle and free car scanner app you can instantly see if it improved by watching MEC and deltaV to see your battery health.