Efficiency

[MotMot]

My tour has dropped from 77 to 73.5kwh capacity in 14 months and 15k miles. Quite happy with that tbh. My prev 58kwh dropped to 52 after a year. All from OBD readings.

[CarterHounslow]

The data is definitely massively off now. Normally I've seen somewhere between 2 and 10% difference between VW readings and actuals. However I think the software update has done something very odd.

Yesterday I did 85 miles and used 44% battery, which is about 3.3 miles per kwh. The car was claiming I did 4.1. That's not even close.

[sidehaas]

The data is definitely massively off now. Normally I've seen somewhere between 2 and 10% difference between VW readings and actuals. However I think the software update has done something very odd.

Yesterday I did 85 miles and used 44% battery, which is about 3.3 miles per kwh. The car was claiming I did 4.1. That's not even close.

You can't use SoC to get an accurate estimate over 44% of the battery. If you do it over 90% you'll find it's more accurate. Also remember that after 1-2 years you almost certainly have a true battery capacity somewhere in the range 50-53 kWh than 58 kWh.

[CarterHounslow]

The data is definitely massively off now. Normally I've seen somewhere between 2 and 10% difference between VW readings and actuals. However I think the software update has done something very odd.

Yesterday I did 85 miles and used 44% battery, which is about 3.3 miles per kwh. The car was claiming I did 4.1. That's not even close.

You can't use SoC to get an accurate estimate over 44% of the battery. If you do it over 90% you'll find it's more accurate. Also remember that after 1-2 years you almost certainly have a true battery capacity somewhere in the range 50-53 kWh than 58 kWh.

Not the point. For the car to claim 4.1 over that distance is crazy. Its incredibly misleading. I've never had it be that far out. If I had used up the whole 80% I probably would have only been able to go 150 miles. But it's a very hilly twisty journey, so I would expect around 3.5 miles per kwh max.

It's almost as if it's not taking something into consideration with the calculation, like it's not factoring in air conditioning or something.

[sidehaas]

The data is definitely massively off now. Normally I've seen somewhere between 2 and 10% difference between VW readings and actuals. However I think the software update has done something very odd.

Yesterday I did 85 miles and used 44% battery, which is about 3.3 miles per kwh. The car was claiming I did 4.1. That's not even close.

You can't use SoC to get an accurate estimate over 44% of the battery. If you do it over 90% you'll find it's more accurate. Also remember that after 1-2 years you almost certainly have a true battery capacity somewhere in the range 50-53 kWh than 58 kWh.

Not the point. For the car to claim 4.1 over that distance is crazy. Its incredibly misleading. I've never had it be that far out. If I had used up the whole 80% I probably would have only been able to go 150 miles. But it's a very hilly twisty journey, so I would expect around 3.5 miles per kwh max.

It's almost as if it's not taking something into consideration with the calculation, like it's not factoring in air conditioning or something.

My point is this: you've calculated 3.3mi/kWh by assuming a 58kwh battery and multiplying by 0.44 to estimate kWh used (25.5kwh). Or used Tronity, which does the same thing. It's incorrect.
Its more likely you've done 85 miles at a consumption of 4.1 mi/kWh so you've used 20.7 kWh. The car indicates you have used 44% of the battery. If correct, this implies you have a battery capacity of 47kwh, which does seem very low. However there will be some degradation and the SoC scale itself isn't linear.
Also, if you did your 85 miles over multiple journeys and used any preheating/precooling, them the preheating/precooling will not be included in the consumption figure.

To summarise, you are assuming that the battery's total capacity is fixed and the SoC display is accurate, so that there must be an error in consumption figure. In practice, there are errors in all three, and unless you have been using preheating/precooling, the consumption data will be the most accurate of the three.
You can test your battery capacity if you wish by doing a fuller run over 90% of the battery (so errors in % displayed are smaller) without any preheating/precooling and calculating 4mileage covered divided by displayed consumption.

[CarterHounslow]

You can't use SoC to get an accurate estimate over 44% of the battery. If you do it over 90% you'll find it's more accurate. Also remember that after 1-2 years you almost certainly have a true battery capacity somewhere in the range 50-53 kWh than 58 kWh.

Not the point. For the car to claim 4.1 over that distance is crazy. Its incredibly misleading. I've never had it be that far out. If I had used up the whole 80% I probably would have only been able to go 150 miles. But it's a very hilly twisty journey, so I would expect around 3.5 miles per kwh max.

It's almost as if it's not taking something into consideration with the calculation, like it's not factoring in air conditioning or something.

My point is this: you've calculated 3.3mi/kWh by assuming a 58kwh battery and multiplying by 0.44 to estimate kWh used (25.5kwh). Or used Tronity, which does the same thing. It's incorrect.
Its more likely you've done 85 miles at a consumption of 4.1 mi/kWh so you've used 20.7 kWh. The car indicates you have used 44% of the battery. If correct, this implies you have a battery capacity of 47kwh, which does seem very low. However there will be some degradation and the SoC scale itself isn't linear.
Also, if you did your 85 miles over multiple journeys and used any preheating/precooling, them the preheating/precooling will not be included in the consumption figure.

To summarise, you are assuming that the battery's total capacity is fixed and the SoC display is accurate, so that there must be an error in consumption figure. In practice, there are errors in all three, and unless you have been using preheating/precooling, the consumption data will be the most accurate of the three.
You can test your battery capacity if you wish by doing a fuller run over 90% of the battery (so errors in % displayed are smaller) without any preheating/precooling and calculating 4mileage covered divided by displayed consumption.

With all that said then, how are you meant to even plan a journey? There's no way my battery has degraded 20%, it doesn't fit in line with the figures I'm getting back (for example, there are journeys I did two years ago that took 10% and they still take 10% now). Just guess how far the car goes? Because the consumption data from the car is pretty useless.

[sidehaas]

Not the point. For the car to claim 4.1 over that distance is crazy. Its incredibly misleading. I've never had it be that far out. If I had used up the whole 80% I probably would have only been able to go 150 miles. But it's a very hilly twisty journey, so I would expect around 3.5 miles per kwh max.

It's almost as if it's not taking something into consideration with the calculation, like it's not factoring in air conditioning or something.

My point is this: you've calculated 3.3mi/kWh by assuming a 58kwh battery and multiplying by 0.44 to estimate kWh used (25.5kwh). Or used Tronity, which does the same thing. It's incorrect.
Its more likely you've done 85 miles at a consumption of 4.1 mi/kWh so you've used 20.7 kWh. The car indicates you have used 44% of the battery. If correct, this implies you have a battery capacity of 47kwh, which does seem very low. However there will be some degradation and the SoC scale itself isn't linear.
Also, if you did your 85 miles over multiple journeys and used any preheating/precooling, them the preheating/precooling will not be included in the consumption figure.

To summarise, you are assuming that the battery's total capacity is fixed and the SoC display is accurate, so that there must be an error in consumption figure. In practice, there are errors in all three, and unless you have been using preheating/precooling, the consumption data will be the most accurate of the three.
You can test your battery capacity if you wish by doing a fuller run over 90% of the battery (so errors in % displayed are smaller) without any preheating/precooling and calculating 4mileage covered divided by displayed consumption.

With all that said then, how are you meant to even plan a journey? There's no way my battery has degraded 20%, it doesn't fit in line with the figures I'm getting back (for example, there are journeys I did two years ago that took 10% and they still take 10% now). Just guess how far the car goes? Because the consumption data from the car is pretty useless.

You can plan a journey with knowledge of the consumption and range you usually get on a longer journey of that type eg motorway. The car won't tell you that anyway, unless all your driving is on similar road types so you can just look at the long term data.
If you are at a loss, just assume 100% = 200 motorway miles in summer and 180 motorway miles in winter, it won't be far wrong. Then you can keep a rough eye on consumption and if gets too high to reach your destination, you slow down.

I'm sure it hasn't degraded 20%, that would seem extreme, but I suspect 5-10% is normal after a couple of years. 'Battery Life' on YouTube has done his own measurements of his ID3 degradation by driving down from 100% to below 10% and using the consumption data to derive kWh usage, as I described above. He estimated 8% after 1 year and 11% after 2 years, assuming a 58kwh starting point. After two years he then did an Aviloo test which estimated 7% from 54kwh (ie, remaining capacity around 50kwh) and had an official VW test done by his dealer which said 10%. So his calculations were roughly right. He had done a lot of DC charging, mind.

I've done my own calculations similarly on a recent trip and estimated our car has 53kwh remaining (I drove down from 100% to between 10% and 15% on that journey I think, so not ideal). However if I'd done it over smaller battery % intervals like yours above I could have estimated significantly less. It's really inaccurate unless you measure it over almost the full battery range, because of the SoC prediction error. I suspect if you did the test over a longer journey you'd see a better result and the overall calculations as per your original post would match better.

[MattgID3]

Really pleased with economy yesterday. Although calc suggest real economy 5.2miles kwh based on cars reported 303 mile range. The other possibility is the the car has lost 20 miles battery range if my maths is right about 6.5% capacity so down to 54.23kw.

[CarterHounslow]

My point is this: you've calculated 3.3mi/kWh by assuming a 58kwh battery and multiplying by 0.44 to estimate kWh used (25.5kwh). Or used Tronity, which does the same thing. It's incorrect.
Its more likely you've done 85 miles at a consumption of 4.1 mi/kWh so you've used 20.7 kWh. The car indicates you have used 44% of the battery. If correct, this implies you have a battery capacity of 47kwh, which does seem very low. However there will be some degradation and the SoC scale itself isn't linear.
Also, if you did your 85 miles over multiple journeys and used any preheating/precooling, them the preheating/precooling will not be included in the consumption figure.

To summarise, you are assuming that the battery's total capacity is fixed and the SoC display is accurate, so that there must be an error in consumption figure. In practice, there are errors in all three, and unless you have been using preheating/precooling, the consumption data will be the most accurate of the three.
You can test your battery capacity if you wish by doing a fuller run over 90% of the battery (so errors in % displayed are smaller) without any preheating/precooling and calculating 4mileage covered divided by displayed consumption.

With all that said then, how are you meant to even plan a journey? There's no way my battery has degraded 20%, it doesn't fit in line with the figures I'm getting back (for example, there are journeys I did two years ago that took 10% and they still take 10% now). Just guess how far the car goes? Because the consumption data from the car is pretty useless.

You can plan a journey with knowledge of the consumption and range you usually get on a longer journey of that type eg motorway. The car won't tell you that anyway, unless all your driving is on similar road types so you can just look at the long term data.
If you are at a loss, just assume 100% = 200 motorway miles in summer and 180 motorway miles in winter, it won't be far wrong. Then you can keep a rough eye on consumption and if gets too high to reach your destination, you slow down.

I'm sure it hasn't degraded 20%, that would seem extreme, but I suspect 5-10% is normal after a couple of years. 'Battery Life' on YouTube has done his own measurements of his ID3 degradation by driving down from 100% to below 10% and using the consumption data to derive kWh usage, as I described above. He estimated 8% after 1 year and 11% after 2 years, assuming a 58kwh starting point. After two years he then did an Aviloo test which estimated 7% from 54kwh (ie, remaining capacity around 50kwh) and had an official VW test done by his dealer which said 10%. So his calculations were roughly right. He had done a lot of DC charging, mind.

I've done my own calculations similarly on a recent trip and estimated our car has 53kwh remaining (I drove down from 100% to between 10% and 15% on that journey I think, so not ideal). However if I'd done it over smaller battery % intervals like yours above I could have estimated significantly less. It's really inaccurate unless you measure it over almost the full battery range, because of the SoC prediction error. I suspect if you did the test over a longer journey you'd see a better result and the overall calculations as per your original post would match better.

I totally understand all of that, however, try explaining this to someone who doesn't actually drive an EV and wants to adopt one. They get worried about it running out. If you take my journey the other day, which I want to stress, I was in no way close to running out of charge. I used 44% of my battery, regardless of what miles per kwh the car was telling me I was getting, I only went 85 miles. Yet the car was telling me I had about 90 miles of range remaining. Now it doesn't take a genius to work out, that can't be right because if you do the maths on that, with 36% remaining and a similar level of consumption for the rest of the journey, I would most likely get around 68 miles running it down to 1%. Now that's 153 miles from 80% charge if I had done that, on a warm-ish day, with two people in the car with air con and not even driving quickly.

The average person is not going to get this at all once mass adoption starts happening. They're worried about running out of charge as it is without us having to explain to people above how state of charge is not linear etc, I think we'll lose them within seconds.

The software needs to do a better job of predicted range without us guessing.

[Mellons]

Getting some impressive predictive ranges in 77kW tour 5- for urban driving consistently around 610k and up to and once over 630km for 100% charge. Not sure how accurate these are, but seem consistent. I’m doing normal urban driving in eco mode, keeping an eye on speed etc. and making an effort to drive more smoothly.

[TimF]

IMO, the most useful figures are those based on same state of charge. That is, kWh supplied to restore to the endpoint of the previous charge session divided into the mileage between those two sessions (80% to 80% being normal for me). I assume the car will stop the charging at a nearly identical SoC for those two sessions, irrespective of the actual, hypothetical or estimated 100% and 0% end points of the battery. As several have commented, the longer the run(s) involved, the more useful the mi/kWh figure, as rounding and other errors become smaller.

I'm looking forward to 3.2's dashboard information about efficiency, and hope 3.2 manages to be more consistent than 2.4 in regard to displayed remaining range. Yesterday we had a day out locally, where the total mileage was 22 miles. When we set off, remaining range was stated as 196 miles. At around the 18 mile point, remaining range was 186 miles, and that was displayed until half a mile from home. In that short distance (all 30mph roads, moving freely, no net height gain, no hard braking), the remaining range dropped to 184 miles. Which was still being shown first thing this morning. After a trip to the shops (2.5miles), it's now showing 179 miles.

[MotMot]

Getting some impressive predictive ranges in 77kW tour 5- for urban driving consistently around 610k and up to and once over 630km for 100% charge. Not sure how accurate these are, but seem consistent. I’m doing normal urban driving in eco mode, keeping an eye on speed etc. and making an effort to drive more smoothly.

I’m at about 300 miles (480km) for 100% pretty consistently in my Tour S this summer. That’s usually 50-70% motorway driving….

I’ve done a few ‘economy’ runs on the motorway recently - and the best I’ve had was 4.3m/kWh sat at 60mph with the trucks… 3.7 yesterday with faster (70-80mph) speeds which was decent…

[Mellons]

Getting some impressive predictive ranges in 77kW tour 5- for urban driving consistently around 610k and up to and once over 630km for 100% charge. Not sure how accurate these are, but seem consistent. I’m doing normal urban driving in eco mode, keeping an eye on speed etc. and making an effort to drive more smoothly.

I’m at about 300 miles (480km) for 100% pretty consistently in my Tour S this summer. That’s usually 50-70% motorway driving….

I’ve done a few ‘economy’ runs on the motorway recently - and the best I’ve had was 4.3m/kWh sat at 60mph with the trucks… 3.7 yesterday with faster (70-80mph) speeds which was decent…

That sounds about right- when doing motorway im between 480-500km range

[HeidiFlowerpt Driver]

After my car's software 3.2 update I saw efficiency improvement but I formed no conclusions short term because it might have been the journey types, or the weather. But now I feel confident that the efficiency really has improved. My 1st Edition used to give about 3.5M/kWh, and now it's giving around 4.5M/kWh. That's a big improvement and I'm delighted. On some longer journeys I've had over 5M/kWh and on really short around-town journeys I've seen even better.

[MotMot]

Yes my long term Summer consumption is up (or down) from 3.8 to 4.2.

Let’s see what winter brings.. that’s the real test.

[sidehaas]

Yes my long term Summer consumption is up (or down) from 3.8 to 4.2.

Let’s see what winter brings.. that’s the real test.

My experience with 3.0 OTA is that winter benefit over 2.4 was small, it's been more significant in spring and summer. I still saw some benefit on long journeys but the bulk of short journeys in the cold seemed no better than under similar conditions the winter before (when my car was on 2.3 but never did the whole battery heating when standing thing). I suspect this is because the improvements are in the drivetrain control software and when you start turning up the heating it still eats just as much power as before. No proof of course that 3.0 ota is the same as 3.2 in this regard, but my suspicion is it is.

[MotMot]

Yes when the heater is on more the driving part of the total consumption will be a smaller proportion.

[Mellons]

I know its hot at present, but today achieved equivalent of 640km range at 100% at start of the day (397miles) on my tour 5. Normal urban driving, hilly near me with some twisties and short motorway runs.

[Makro99]

Might be a silly question - does efficiency
Get worse in winter? Mine was
Delivered in august and I got regularly >4.4mpkWh, now can’t get north of 3.5. And I don’t use the heaters ….

[sidehaas]

Might be a silly question - does efficiency
Get worse in winter? Mine was
Delivered in august and I got regularly >4.4mpkWh, now can’t get north of 3.5. And I don’t use the heaters ….

Yes.
Heaters use a lot of energy, especially on short journeys, but even with them off the whole system is less efficient at low temperatures. 3.5 in winter is good.