Rated miles vs Wh/mi

[hiroshiy]

The data I've reported here is mostly with AC turned off (and heated seats off etc), because I've been trying to achieve actual miles traveled > rated miles consumed. I am definitely displaying rated miles, not ideal miles. I'm in Maryland and the temps have been fairly mild lately in the low 80's (F) so it's hot not cold, so there is no way I would turn any heat on.

I can understand how AC, heat etc while parked can be excluded from kWh trip measurement (because it doesn't count any energy towards a trip while parked or off), but I'm skeptical that accessory energy isn't counted while the car is driving. For one thing, when I do turn the AC on while driving, I always see higher average Wh/mi values and people on the forums routinely report higher Wh/mi figures for winter driving when they are using heat. For another, Tesla would have to separately meter the power going to the drive train in order to exclude all the accessories etc. In doing so, they would create customer confusion (like mine) regarding why their energy consumption is so low, yet their rated miles consumed is so high, so it doesn't make sense to intentionally exclude accessory power while driving from the trip power meter.

My bad, I couldn't edit my previous post, but yes you're correct that power consumption include everything consumed while the car is moving.

So since mine is a Japanese car and RM = 320Wh/m I can easily achieve RM and can go further than display. But I sometimes go lower than 290Wh/m using surface streets.

It could be that latest 70D uses lower RM number like 260 as it is very efficient; lighter battery with dual motor and improved aero with facelift. I wonder why Tesla does such confusing changes. It sure looks that the car can go further with lower Wh/m constant for RM.

 

[Boatguy]

The data I've reported here is mostly with AC turned off (and heated seats off etc), because I've been trying to achieve actual miles traveled > rated miles consumed. I am definitely displaying rated miles, not ideal miles. I'm in Maryland and the temps have been fairly mild lately in the low 80's (F) so it's hot not cold, so there is no way I would turn any heat on.

I can understand how AC, heat etc while parked can be excluded from kWh trip measurement (because it doesn't count any energy towards a trip while parked or off), but I'm skeptical that accessory energy isn't counted while the car is driving.

Accessory power used while driving is definitely included in the Trip, Energy and other Wh/mi computations.

Looking back over this thread, if you're reported Wh/mi during a period of continuous driving is less than 290 Wh/mi, and the decline in RM during the trip is more than the actual miles driven, then then something is wrong and you should contact Tesla. You can reach them either at [email protected] or by calling your local service center.

 

[ran349]

Accessory power used while driving is definitely included in the Trip, Energy and other Wh/mi computations.

Looking back over this thread, if you're reported Wh/mi during a period of continuous driving is less than 290 Wh/mi, and the decline in RM during the trip is more than the actual miles driven, then then something is wrong and you should contact Tesla. You can reach them either at [email protected] or by calling your local service center.

If you looked at my posts, you will see that with the same car, I get similar numbers to hacer; 260 Wh/mi to achieve rated miles. I don't believe there is anything wrong with his car. For me, with 260 Wh/mi x 240 miles( rated range), this equals about 62.5 kWh of the 70 kWh battery used. This is about right since there is always a reserve of power remaining even if you go down to zero on you rated miles.

 

[ran349]

Those attachments in my post were unintentional. They need an edit feature here.

 

[Rocky_H]

I just detailed my car today and I had the AC running for about two hours straight with the car unplugged. I lost about 15-20 rated miles but the kWh used since last charge did not increase. This confirms that power used to cool is not accounted for in the calculation, as I suspected.

You observed the effect, but that's not quite the cause. It's not that cooling energy isn't tracked; it's that energy use is not tracked while the car is not in a drive gear.

 

[hacer]

I previously made an erroneous reference to the dashed line on the energy display showing the rated Wh/mi. It's actually the solid line which shows the Rated Wh/mi, the dashed line is the average Wh/mi for the selected period.

If you manipulate your driving (pretty easy to do over 5mi) so that the dashed line is on top of the solid line, the Average Wh/mi number displayed is the Rated Wh/mi being used for RM calculations.

So my solid "rated" line seems to be at 300 Wh/mi; when my average Wh/mi is 300 Wh/mi the dashed line falls on top of that solid line and the predict range exactly matches the rated range. In the photo below the Wh/mi was 299 and the dashed line is ever so slightly below the solid one. It would be completely, totally impossible to achieve that predicted range at that high a consumption rate. So if nothing else is wrong, the mileage predictor is totally whacked. Another data point is that the window sticker EPA rated Wh/mi is 330 but that is at the AC source. A 91% efficient charger would yield an on-board 300 Wh/mi but my measured efficiency was 83% which would give a 274 Wh/mi on-board which even that wouldn't get rated miles. All of my actual driving points to somewhere between 245 and 252 Wh/mi to get actual miles driven = rated miles consumed. Since my data continues to fall on the straight lines I posted earlier, the driveable capacity is 60 kWh, more than you'd expect from a 60 kWh battery, but generously low for a 70 kWh battery.

I did contact my service center about this and they emailed me some boilerplate about why its impossible to calculate battery capacity.

Finally it's worth noting that all of this has been with the battery in the range of 40% to 80% charge, maybe things will change when the voltage starts to change at the low end of the charge, or if I range charge it, so I'm going to make a longish practice trip before I attempt Canada, because having predicted range so ridiculously above actual range is very range-anxiety provoking for attempting any long stretch.

 

[ran349]

"... the mileage predictor is totally whacked."

I would agree with that statement from my own experience, as I described in a previous post. The trip planner may be useful since it does account for elevation and speed but you will have to have apply a correction factor to make up for the prediction error. Also, what are your tire pressures? That may make some difference in your mileage. I didn't check my tire pressures for the first few months, and when I did, they were several PSI below nominal.

 

[hiroshiy]

I'm getting persuaded by a hypothesis that Tesla uses different RM constant for each models and their energy graph is still using old 300Wh/m constant whereas in fact actual RM calculations are based on model-by-model constant. 85/90D=290, 85/90 single=300, 70D=260 etc??

 

[Boatguy]

I'm getting persuaded by a hypothesis that Tesla uses different RM constant for each models and their energy graph is still using old 300Wh/m constant whereas in fact actual RM calculations are based on model-by-model constant. 85/90D=290, 85/90 single=300, 70D=260 etc??

The definitely use different Wh/mi assumptions for different cars as you indicate. My S90D energy graph and rated/projected calculations are all based on 290Wh/mi.

The big difference is that my RM drops based on actual consumption relative to 290, whereas @hacer is seeing his drop relative to about 250.

 

[Boatguy]

I did contact my service center about this and they emailed me some boilerplate about why its impossible to calculate battery capacity.

That's not acceptable. The calculations should be accurate with SOC in the 40-80% range, SOC is not the issue. Something is wrong and Tesla needs to address it. I suggest you setup a simple test case.

Reset Trip A
Photograph the Units settings to show it is on Rated and not Ideal
Photograph your odometer
Photograph your dash showing the RM
Drive 20-30 miles any way you like, don't worry about hitting a specific Wh/mi
Photograph the odometer
Photograph Trip A
Photograph the dash showing the RM
Do the math to calculate what the consumed RM (i.e., Delta RM = Actual Wh/mi / 300 * miles driven)

Send all that to Tesla and demand an explanation. And for good measure, post it here.

 

[Rocky_H]

That's not acceptable. The calculations should be accurate with SOC in the 40-80% range, SOC is not the issue. Something is wrong and Tesla needs to address it. I suggest you setup a simple test case.

Reset Trip A
Photograph the Units settings to show it is on Rated and not Ideal
Photograph your odometer
Photograph your dash showing the RM
Drive 20-30 miles any way you like, don't worry about hitting a specific Wh/mi
Photograph the odometer
Photograph Trip A
Photograph the dash showing the RM
Do the math to calculate what the consumed RM (i.e., Delta RM = Actual Wh/mi / 300 * miles driven)

Send all that to Tesla and demand an explanation. And for good measure, post it here.

The calculation should really be a bit more simple than that. Get the number for the kwh used for the drive and the number of actual miles driven. Divide. Doing a longer distance will have a bit more stable data to be a bit more accurate.

*EDIT* As corrected below, this should have been rated miles consumed, not actual miles driven.

 

[ran349]

The calculation should really be a bit more simple than that. Get the number for the kwh used for the drive and the number of actual miles driven. Divide. Doing a longer distance will have a bit more stable data to be a bit more accurate.

What you described is how the driver's Wh/mi is calculated. But you don't need to do that because it is shown on the trip odometer display and therefore should match that, subject to rounding uncertainties. To calculate your actual rated Wh/mi, you take kWh driven divided by rated miles consumed. That is what hacer is doing.

 

[Rocky_H]

What you described is how the driver's Wh/mi is calculated. But you don't need to do that because it is shown on the trip odometer display and therefore should match that, subject to rounding uncertainties. To calculate your actual rated Wh/mi, you take kWh driven divided by rated miles consumed. That is what hacer is doing.

Oh, yes, you're right. I did mean rated miles consumed, not actual miles. Oops.

 

[Boatguy]

Yesterday I ran my own test on the issue raised by @hacer. I drove up to play golf with some friends, about 50mi each way. Here are the results:

Outbound
RM: 58

Trip A reported:
Odometer miles: 54.5
Wh/mi (per Trip A): 293
Power used: 16 kWh

Implicit Wh/RM = 276

5 RM lost wil parked for 5hrs (no HVAC, just sitting).

Return
RM: 55

Trip A reported:
52.1 mi
287 Wh/mi
14.9kWh

Implicit Wh/RM = 271

Trip total
Odometer miles: 106.6
RM: 118

First of all, the RM used in both directions was more than should have been used based on 1 RM = 290Wh. Outbound should have been 55 and the return should have been 52. Second, how did I lose 5RM while the car was parked for 5hrs? There was no HVAC running (the car was sitting outside in about 85F). Overall the trip used about 10% more RM than it should have based on average consumption of 290 Wh/mi for the trip. I was losing RM at an implied rate of around 274 Wh/mi.

There is room for some rounding error in there, but overall the trip consumed 10% more RM than it should have, the equivalent of losing 10% of the range of the car and certainly makes the case for having at least a 20% buffer when trip planning. I got home with an SOC of about 10% (27RM).

I'll report this to ServiceHelpNA, but I'm not expecting anything more than form letter response.

 

[msnow]

Yesterday I ran my own test on the issue raised by @hacer. I drove up to play golf with some friends, about 50mi each way. Here are the results:

Outbound
RM: 58

Trip A reported:
Odometer miles: 54.5
Wh/mi (per Trip A): 293
Power used: 16 kWh

Implicit Wh/RM = 276

5 RM lost wil parked for 5hrs (no HVAC, just sitting).

Return
RM: 55

Trip A reported:
52.1 mi
287 Wh/mi
14.9kWh

Implicit Wh/RM = 271

Trip total
Odometer miles: 106.6
RM: 118

First of all, the RM used in both directions was more than should have been used based on 1 RM = 290Wh. Outbound should have been 55 and the return should have been 52. Second, how did I lose 5RM while the car was parked for 5hrs? There was no HVAC running (the car was sitting outside in about 85F). Overall the trip used about 10% more RM than it should have based on average consumption of 290 Wh/mi for the trip. I was losing RM at an implied rate of around 274 Wh/mi.

There is room for some rounding error in there, but overall the trip consumed 10% more RM than it should have, the equivalent of losing 10% of the range of the car and certainly makes the case for having at least a 20% buffer when trip planning. I got home with an SOC of about 10% (27RM).

I'll report this to ServiceHelpNA, but I'm not expecting anything more than form letter response.

You could easily lose 2-3 miles just with vampire loss. That wouldn't be unusual.

 

[Boatguy]

You could easily lose 2-3 miles just with vampire loss. That wouldn't be unusual.

In 5 hrs? I've seen my car lose 3-4 in 24hrs, but not 5 in 5 hrs.

 

[msnow]

In 5 hrs? I've seen my car lose 3-4 in 24hrs, but not 5 in 5 hrs.

Agree its excessive but a couple of miles could be explained that way.

 

[Boatguy]

Another test yesterday afternoon / evening on a round trip to a party.

RM: 131
Trip A: 120.7 / 35.8kWh / 297Wh/mi
Implied Wh/RM = 35,800 / 131 = 273Wh/RM

To refresh memories, 290Wh/mi is the solid line on the S90D energy display where RM = Projected Miles. But like the previous trip, the actual RM used was higher than what would be expected based on 290Wh/mi.

Expected RM based on 290Wh/mi = (297/290) * 121 = 124

I'm going to measure this on a few more trips, but my interim conclusions are:

1) The actual Wh/RM (based on Trip A reported kWh) is in the range of 271 - 276, a fairly narrow range given the rounding error inherent in the miles reported.

2) Combined with my recent drop in total RM, my 100% charge of 281RM implies a usable battery pack of just 281 * 273 = 76.7kWh.

3) Real range with a nominal consumption of 290Wh/mi = 273 / 290 * 281 = 264mi.
Real range with my current average consumption of 310Wh/mi = 273 / 310 * 281 = 247mi.

Substantially less range than the 294mi in the Design Center!

 

[Boatguy]

I realized that the above Wh/RM nomenclature may be confusing. What the calculations are telling me is that consuming ~ 270 Wh will cost me 1 Rated Mile. So if my consumption per mile driven is 310 Wh, then I will consume 1.15RM for every mile on the odometer.

It also means the Energy display of Projected versus Rated miles is at best in error, and at worst intentionally misleading.

 

[hacer]

Sunday I decided to take my son to Hershey Park which is 110 miles from home. So this was my first test at a long drive and first time using a supercharger. My results were different than I've seen before, and I feel better about my car's long range capabilities. Here is the relevant trip data:

Starting: Charged to 85%, 203 RM which is higher than I've ever charged before. Began the trip immediately after charging complete
Arrived at Hershey park: 110.7 actual odometer miles, 28.5 kWh used, 257 Wh/mi trip average. 103 RM used, 42% battery SOC. That works out to 276 Wh/RM, considerably higher than I've ever calculated before. The weather conditions were pretty ideal. The temperature was 69 F at the start of the drive and reached 75 by the time we arrived. Cabin AC was set to 71 for the whole trip and I was driving reasonably fast. Part of the really low Wh/mi is probably because I followed a large truck going 70 MPH on I-83 for about 25 miles with TACC set to 75 and distance of 5 lengths which I think provided good drafting. We were parked at the amusement park for about 6.5 hours, then drove to the Harrisburg supercharger.

Arrived at the supercharger: 122.9 odometer miles since leaving home 32.2 kWh used, 262 Wh/mi trip average. 118 RM used and I foolishly forgot to record battery SOC (I was excited to start supercharging). That works out to 272 Wh/RM still much higher than all driving done before this day. I charged the car to 95% (230 RM) and it went surprisingly quickly. It was projected to take an hour but finished in 50 minutes.

The drive home from the supercharger was 99.5 odometer miles and consumed 27.2 kWh for an average of 273 Wh/mi. 98 RM were consumed for 277 Wh/RM, a new record. Ending SOC was 56%.

The RM vs SOC points still fall on the straight lines, but the change in kWh vs change in SOC yield a decidedly different slope from my previous measurements. If I adjust the position for the different starting SOC, the kWh consumed points don't fall on the straight lines I had before.

Many things changed for this data, first was that my car had a software update the day before. The release notes describe only changes to Summon, but who knows what else may be in there. Second I charged the car to higher levels than ever before and discharged it to lower levels than ever before.

I think that when the battery was previously kept in a narrow range where the battery voltage doesn't change much with SOC that forces the SOC estimator to be a strict coulomb counter which is why the kWh vs SOC was so perfectly linear. It seems like the software was using a conservative (from one point of view anyway) slope so that it takes away more range than you'd expect while in the middle SOC range until it actually sees some voltage change where it then adapts the SOC value. That would explain the new points not falling on the straight line.

The consumption graph continues to display the "rated" line at 300 Wh/RM, and still seems to follow the formula:
estimated range = RM * 300 / (average Wh/mi). I think that is still a problem if my "true" Wh/RM is 272. The trip consumption which displays SOC seems to be more reliable. I haven't see that curve previously because I hadn't used a navigation destination.

Finally from this I think it is hard to pin down a car's Wh/RM from measurements.