Range Reduced on Dual Motor Configs?

[AMPUP]

An electric motor offers nearly zero resistance when spinning. So 2 motors don't really hurt compared to one. When cruising, you might use 20 hp. 2 motors producing 10 hp each uses nearly identical energy as 20 from a single motor. The extra motor requires a tiny amount of extra energy - so small as to be zero in any calculation.

Now - extra weight is worth something but 2% comes to mind (far more in city, close to zero in highway). Then you have the improved efficiency by gearing the 2 motors differently so one is optimal at 30 mph and one is optimal at 60 mph. This is perhaps worth 5% in overall efficiency.

So net is +3% give or take - with up to 5% in highway and 1% in city. Just rough numbers. Remember the new EPA sticker probably has 21 inch rims instead of 19 with the efficiency numbers.

So it isn't just software, it is optimizing 2 systems for 2 different speeds. This trumps any loses in efficiency.

An ICE system is a totally different beast as it just adds weight and extra resistance with no improvements to counter it.

The whole concept of 2 motors is like having a 2 speed transmission instead of one. In an ICE world, that would represent 20% improved efficiency (if not more).

Source ? Or speculation?

 

[jweinstein]

242/265 = 91%, which is pretty close to averaging the 100% and 80% charge states. Just a thought...

 

[gpetti]

Source ? Or speculation?

The different gearing is speculation but what is self evidently true is that AWD on an ICE requires a mechanical connection between front and rear wheels with all the weight and drag that implies. i know there are various clever technologies to optimize this but the spec sheets always show less efficiency for the same level of power. The tesla situation only needs an electrical connection and then pure electrical signaling to tell the motor what to do and when. What Elon referred to as digital versus analog - an expression that I initially disagreed with but I've warmed to now. If you compare like with like, I.e. The 85D with the 85 I think you'll see that the numbers reflect this efficiency in that the AWD is actually more efficient (which is why we are speculating about gearing). In the case of the P85D we have over 200 HP more performance so the numbers can't be used to evaluate electric AWD efficiency as easily. Nonetheless I think early indications are that our highway range will be equivalent or better which is remarkable.

- - - Updated - - -

Just a thought... We are doing much speculation based on a window sticker that was posted in a demo car. Is it possible that the numbers are not fully fleshed out yet?

- - - Updated - - -

Those numbers are measured from the wall, and therefore include charging losses.

The fact that the totals are so different even though the efficiency numbers are the same is a big clue to what is going on here - quite clearly the new rating is based on having less energy on the battery pack - which since the pack is the same size must be a product of the rules about 'normal' and 'range' charging.
Walter

I think youre right, the fact that these totals don't line up indicates there is some variable that has changed here that we are not privvy to. These stickers don't make sense if the same range formula was used and the same level of charge. There isn't much more we can do with this information, until we see a real sticker and try these cars in the wild.

 

[breser]

It's in the document I posted before, energy usage numbers include charging losses. I could go find it again but I remember reading that the energy usage needed to measured in a place where it would include charging losses.

 

[David_Cary]

Obviously some of that is educated guesses.

Some things are obvious of course and so well established that sources would be hard to come by. Elon gave the 3% number and it makes sense to me. The weight one is pretty easy to find and you can of course use EVtripplanner and add in the weight of the 2nd motor/inverter. The weight change is speculation but 250 pounds has been thrown around and it sounds a bit high but reasonable.

2 motors at different gear ratios is common sense. It is true that EVs have a very flar efficiency curve over a broad range but if you look at the curves, there is an optimum - and the difference is up to 5% between optimum and worst. But of course, I don't have the efficiency curve for the Model S motor - so it is speculation that it is similar to other electric motors.

Is it speculation that a 2 speed transmission is more efficient than one? On an ICE vehicle - that is pretty common sense. It is true on an EV that the tranny losses balance the efficiency gains (if we were talking in fact about a transmission) but given those loses don't exist - it is a well established fact that EV motors still have optimal RPMs.

So sure - some speculation. Mostly well established fact.

 

[CalDreamin]

The P85 and P85D are both EPA rated at 89 MPGe (combined city/highway). So the AWD penalty is insignificant as Tesla said, and there's no significant weight or tire efficiency penalty either. They also have the same battery. Seems like the EPA range of these cars should be the same within rounding error -- 1% percent or so. Yet the EPA range is 265 miles for the P85 and 242 miles for the P85D. That doesn't add up.

 

[breser]

The P85 and P85D are both EPA rated at 89 MPGe (combined city/highway). So the AWD penalty is insignificant as Tesla said, and there's no significant weight or tire efficiency penalty either. They also have the same battery. Seems like the EPA range of these cars should be the same within rounding error -- 1% percent or so. Yet the EPA range is 265 miles for the P85 and 242 miles for the P85D. That doesn't add up.

Sure it does. Please go back and look at the labels you just reposted. The P85+ is rated at 88 city, 90 highway, 89 combined. The P85D is rated at 86 city, 94 highway, 89 combined. A significant portion of the testing is at lower speeds (as has been pointed out here previously). The P85D does worse in the city and better at highway speeds speeds. They both happen to have the same combined ratings, but that doesn't mean they should have the same range.

 

[dennis]

From SAE document J1634 OCT2012 "Battery Electric Vehicle Energy Consumption and Range Test Procedure" (thanks to breser for finding this document):

Full charge is to be established using the manufacturer’s recommended charging procedure and appropriate equipment. If the vehicle is equipped with a charger, that charger shall be used. Otherwise, the vehicle shall be charged using an external charger recommended by the vehicle manufacturer. If multiple charging power levels are available, the vehicle shall be recharged at the power level recommended by the manufacturer. If not specified by the manufacturer, the recharging power level expected to be most widely utilized by end users shall be selected.

Note that there is no mention of 80% and 100% and averaging them. But there is a requirement to recharge to the manufacturer recommended or most widely used level. So the loss in range in the P85D could be explained by Tesla now using 90% (the most widely used charge level and where the slider is set on delivery). But if that is the case, why is the 2014 Model S 85 rated at 265 miles? This new version of the spec applies to all MY 2013 and later cars.

I believe Tesla could justify that 100% is the manufacturer recommended charge level for long distance travel. So that would get us back to there being an 8.5% drop in range for the P85D when driven under the EPA 5-cycle test, with its heavy emphasis on city driving where the P85D does worse as explained above.
I had hoped that reading the SAE spec was going to clear this up for all of us. It didn't.

 

[gpetti]

Sure it does. Please go back and look at the labels you just reposted. The P85+ is rated at 88 city, 80 highway, 89 combined. The P85D is rated at 86 city, 94 highway, 89 combined. A significant portion of the testing is at lower speeds (as has been pointed out here previously). The P85D does worse in the city and better at highway speeds speeds. They both happen to have the same combined ratings, but that doesn't mean they should have the same range.

i guess if one assumes that they estimate the range using the combined number or the KWh per 100 miles then it doesn't add up. The bottom line is that we are missing some piece of information. If their testing used a different combined or consumption number that would explain it, or if they are using a different SOC than was used on the earlier sticker, either of those would explain it so we will have to wait and see what a real sticker says, and if that is the same the hopefully at some point Tesla can explain this.

 

[breser]

i guess if one assumes that they estimate the range using the combined number or the KWh per 100 miles then it doesn't add up. The bottom line is that we are missing some piece of information. If their testing used a different combined or consumption number that would explain it, or if they are using a different SOC than was used on the earlier sticker, either of those would explain it so we will have to wait and see what a real sticker says, and if that is the same the hopefully at some point Tesla can explain this.

I'm not sure what the method for calculating the combined number but I'm pretty sure the combined number is not used to calculate the range.

- - - Updated - - -

From SAE document J1634 OCT2012 "Battery Electric Vehicle Energy Consumption and Range Test Procedure" (thanks to breser for finding this document):

Full charge is to be established using the manufacturer’s recommended charging procedure and appropriate equipment. If the vehicle is equipped with a charger, that charger shall be used. Otherwise, the vehicle shall be charged using an external charger recommended by the vehicle manufacturer. If multiple charging power levels are available, the vehicle shall be recharged at the power level recommended by the manufacturer. If not specified by the manufacturer, the recharging power level expected to be most widely utilized by end users shall be selected.

Note that there is no mention of 80% and 100% and averaging them. But there is a requirement to recharge to the manufacturer recommended or most widely used level. So the loss in range in the P85D could be explained by Tesla now using 90% (the most widely used charge level and where the slider is set on delivery). But if that is the case, why is the 2014 Model S 85 rated at 265 miles? This new version of the spec applies to all MY 2013 and later cars.

I believe Tesla could justify that 100% is the manufactured recommended charge level for long distance travel. So that would get us back to there being an 8.5% drop in range for the P85D when driven under the EPA 5-cycle test, with its heavy emphasis on city driving where the P85D does worse as explained above.
I had hoped that reading the SAE spec was going to clear this up for all of us. It didn't.

I'm not surprised. Based on reading a lot of these documents, there are things that the EPA has clearly said are just handled on a case by case basis. The EPA clearly wants to get rid of these ambiguities but it's hard because of the changing nature of technology. If you look back at the CFR link I provided there is a whole series of things that are done differently than the SAE spec. I didn't point this out because none of them seemed relevant to the situation.

I'm much more inclined to believe the range loss is just because the P85D is significantly worse in the city. Probably because it's been geared for performance. I'd bet that the 85D will have improved city performance and highway performance and has something like 275 rated range.

 

[Saghost]

Sure it does. Please go back and look at the labels you just reposted. The P85+ is rated at 88 city, 90 highway, 89 combined. The P85D is rated at 86 city, 94 highway, 89 combined. A significant portion of the testing is at lower speeds (as has been pointed out here previously). The P85D does worse in the city and better at highway speeds speeds. They both happen to have the same combined ratings, but that doesn't mean they should have the same range.

Even if you assume the EPA range was based on 100% city driving, that's only reduce the range from 265 to 259 miles. Any other combination is more favorable to the P85D - and with the same energy available and an all highway basis, it'd be 277 miles.

The differences in efficiency as provided by the MPGe results can not explain the change in rated electric range, no matter how you mix them together. There must be a change in the assumptions/rules about the battery SoC range.
Walter

 

[David_Cary]

I believe the range is 55% city and 45% highway.

 

[breser]

Even if you assume the EPA range was based on 100% city driving, that's only reduce the range from 265 to 259 miles. Any other combination is more favorable to the P85D - and with the same energy available and an all highway basis, it'd be 277 miles.

The differences in efficiency as provided by the MPGe results can not explain the change in rated electric range, no matter how you mix them together. There must be a change in the assumptions/rules about the battery SoC range.
Walter

How are you doing the math to come up with 277 miles?

Trying to reverse the math on the window sticker is a pain. First the biggest problem is the energy numbers are wall to wheel. So without knowing the charging losses they used you can't reverse the math. If you do the simplistic assumption of 85 kWh / 33.7 kWh (assumed electrical equivalent to a gallon of gasoline) then the Model S has about 2.5 (that's rounded down the exact number is slightly higher) gallons of gasoline equivalent in energy. The P85+ would then have 221 miles (2.5*88) of city range and the P85D would have 216 miles (2.5*86) of range. This calculation is obviously wrong because the 85 kWH is not the right number since we need to include the charge losses.

I have no idea what number they're using for energy losses. But let's say it's 20%, which brings up the energy up to 102 kWh. 102 kWh / 33.7 kWh = ~3 gallons of gas (again this result is rounded down). So then the P85+ would have 266 miles (3*88) of range and the P85D would have 260 miles (3*88) of range. So I'm guessing this is roughly what you're doing for the math. My numbers are slightly higher because I'm using the exact number from the MPGe calculation.

If you do the same calculations with a 10% energy loss during charging you come out with 93.5 kWh and then you end up with 244 miles of city range for the P85+ and 238 for the P85D.

If you do it based on the 38 kWh / 100 miles then you come up with 223 miles of range out of 85 kWH or 268 out of 102 kWh or 246 out of 93.5 kWh.

Let's say you try to use the combined numbers and the range to solve for the charge loss. (85 x (1+x))/33.7 * 89 = 242 (or 265 in the case of the P85+). x solves for about 7% for the P85D or 18% for the P85+. But that's not necessarily right because I don't think they calculate the range but rather test it.

You can say this looks fairly ok (the P85D numbers still look a bit weird) but the other numbers could be close depending on what efficiency you're assuming. But there's still an assumption that we're making that's not accurate. We're assuming that you can run the car down to 0 energy in the battery. That's not the case so even if it's an 85 kWh pack you won't be allowed to use the full 85 kWh.

Another factor that nobody has discussed is that the testing for the S85/P85/etc may have been done back before changes to the firmware on the car. If the newer firmwares restrict more of the battery pack then that could also explain the difference. Without any sort of change to the test procedure. That of course would mean that Tesla had effectively reduced the range of their car themselves.

More than likely the reduced number is not a single thing. Probably a combination of several of these issues. But I'm pretty sure this theory that the EPA changed their testing rules is bogus.

 

[TexasEV]

But I'm pretty sure this theory that the EPA changed their testing rules is bogus.

It is neither a theory nor bogus. Here is a Nissan spokesman explaining how the Leaf range was calculated differently last year:
2013 Nissan Leafs EPA-Estimated Range and MPGe Significantly Improved - HybridCars.com

"It was not Nissan’s choice to list the new model’s range estimate based on a 90-percent charge, but rather, this average was stipulated by the EPA. “For 2013, the EPA elected to use an average of Leaf’s two charging modes to generate the range calculation for the window label an 80 percent ‘Long Life Battery Mode’ and a 100 percent ‘Long Distance Mode.’” wrote Brockman. “In the past, the Monroney label displayed range based only on a fully charged battery.”
The 75 miles based on a “90-percent charge” is thus a mid-point between the Leaf’s two default charge modes of 80 or 100-percent.
This, wrote Brockman, effectively means not only is it difficult to compare the 2013 Leaf to previous Leafs, it also muddies the comparison between 2013 Leafs and other electric cars measured under the old methodology."

Note the Leaf stickers say "When fully charged, vehicle travels about..." but gives the average 90% value!

Tesla avoided this last year because of the firmware update changing to a slider rather than the fixed "normal" setting that could be averaged with 100% range setting. The question is, with Tesla having a new model did EPA require it to calculate range on the average of 80% charge and 100% charge as they do for the Leaf and i3?

 

[breser]

It is neither a theory nor bogus. Here is a Nissan spokesman explaining how the Leaf range was calculated differently last year:
2013 Nissan Leafs EPA-Estimated Range and MPGe Significantly Improved - HybridCars.com

My comment is in relation to Tesla. As long as Tesla has had an EPA rated mileage there has been no change to the testing procedure. The Leaf was out before the electric vehicle regulations had been finalized (2011). Tesla had the final regulations when they came up with their range. It's clear that the EPA took a different interpretation of manufacturer recommended charging level than Nissan. But that's not a change in testing procedure. That's just a matter of the EPA interpreting the testing procedure differently than Nissan did and Nissan having to adjust what they were displaying.

Tesla avoided this last year because of the firmware update changing to a slider rather than the fixed "normal" setting that could be averaged with 100% range setting.

I agree, Tesla avoided the problem by using a different UI.

The question is, with Tesla having a new model did EPA require it to calculate range on the average of 80% charge and 100% charge as they do for the Leaf and i3?

Dennis posted the exact text about the battery charge level in the test procedures. It says full or at the manufacturer recommended level. I'd assume that Tesla had some sort of discussion with the EPA about this based on the UI change. I've already said in my past posts it's possible the EPA has changed their position on Tesla's workaround. Possibly because there's new data available. But that would apply to all the Model S vehicles, not just the P85D.

I just can't buy that there's some new test procedure that applied to the P85D and not the other vehicles Tesla is already delivering. The whole point of the test procedures is to provide a uniform measure that people can compare to get an idea about the comparitive efficiency of various vehicles, including vehicles of the same make. If the EPA changed their position on Tesla's 100% charge for the tests then it would apply to all the vehicles and we'd be seeing the changes across the board. So far that hasn't happened.

If there had been a large, applicable to everyone, change in the test procedure you'd see that in the SAE standard or in the CFR.

 

[uselesslogin]

My comment is in relation to Tesla. As long as Tesla has had an EPA rated mileage there has been no change to the testing procedure. The Leaf was out before the electric vehicle regulations had been finalized (2011). Tesla had the final regulations when they came up with their range. It's clear that the EPA took a different interpretation of manufacturer recommended charging level than Nissan. But that's not a change in testing procedure. That's just a matter of the EPA interpreting the testing procedure differently than Nissan did and Nissan having to adjust what they were displaying.

The regulations finalized in 2011 were not mandatory until the 2013 model year. Tesla announced the results of the EPA test in May of 2012 for the 2012 Model S. We know the 2012 LEAF had a 73 mile mile range and the 2013 LEAF has a 15% better range but the 2013 model has a 75 mile range and then they removed the long life mode and it now has an 84 mile range for the 2014 model year. So, I don't know when the test has to repeated, but Tesla has been using the same numbers since 2012.

EPA Downshifts Its Electric Car Range Formula | PluginCars.com
Fuel Economy and Environment Labels Regulations | OTAQ | US EPA
Model S Efficiency and Range | Blog | Tesla Motors

To me this says that is the issue at hand. I don't get how it is bogus. I'll just leave the EPA quote about the regulations here:

Final Rule for Revisions to the Fuel Economy Label for 2013 Model Year (Published July 6, 2011)
EPA and the National Highway Traffic Safety Administration are revising the fuel economy and environment labels that appear on new vehicles beginning with 2013 models. Automakers may also voluntarily adopt the new labels earlier for model year 2012 vehicles.

 

[RiverBrick]

An Engineer directly involved in designing the dual motors was at the Montreal grand opening today and he will be there again Sunday. Anyone interested in the area should attend his presentations and ask him questions.

From today:

For the P85D, emphasis was on performance for what he called the gen 2 motors. For the standard D, the emphasis was on efficiency. On the standard D the motors are not geared differently. New challenges were to make sure no vibrations from the forward motor be felt in the cabin and that the motor not be shoved into the cabin or the battery in case of collision.

 

[dennis]

The regulations finalized in 2011 were not mandatory until the 2013 model year. Tesla announced the results of the EPA test in May of 2012 for the 2012 Model S. We know the 2012 LEAF had a 73 mile mile range and the 2013 LEAF has a 15% better range but the 2013 model has a 75 mile range and then they removed the long life mode and it now has an 84 mile range for the 2014 model year. So, I don't know when the test has to repeated, but Tesla has been using the same numbers since 2012.

To me this says that is the issue at hand. I don't get how it is bogus.

The fact that the Model S numbers are the same in 2014 leads me to believe that they were not required to rerun the test since the hardware configuration had not changed, or they ran it with the same charge level (100%). As I stated previously, Tesla could claim that 100% is the manufacturer's suggested charging level for long distance driving, which is where range matters.

Here is another theory to possibly explain the 8.5% reduction in P85D range:

The P85 sticker range is the same as the 85. But we all know that the 21" tires on the P85 are less efficient. So either the P85 was considered to be a variant of the S 85 that did not require a separate test or the range on the 2012 and later stickers is some blending of 85 and P85 based on expected sales volume. Then along comes the P85D, which is clearly a different model and needs a new test. The P85D range (as compared to the 85) suffers both the rolling resistance penalty of the 21" tires and the weight penalty in the City portions of the test. This could account for the unexpectedly large discrepancy.

 

[breser]

The regulations finalized in 2011 were not mandatory until the 2013 model year. Tesla announced the results of the EPA test in May of 2012 for the 2012 Model S. We know the 2012 LEAF had a 73 mile mile range and the 2013 LEAF has a 15% better range but the 2013 model has a 75 mile range and then they removed the long life mode and it now has an 84 mile range for the 2014 model year. So, I don't know when the test has to repeated, but Tesla has been using the same numbers since 2012.

The idea that a regulation taking effect in 2013 (that was available to Tesla when they did the first Model S testing) is the reason for changes to just the P85D's window sticker seems absurd to me. I'd think you'd have to retest mileage for at least every model year but let me go dig around and find out how often the tests have to be performed.

I will admit it's possible the EPA has decided that Tesla needs to use a charge level based on real world charge preferences of owners. But I'd expect to see all the window sticker values to be adjusted as a result not just the P85D.

 

[Andyw2100]

I say this with all sincerity, and in admiration for those of you who have been discussing this in great depth: there is little doubt in my mind that if those in charge of testing at the EPA put as much thought into designing fair and accurate testing as you guys have put into trying to figure out just what has been done here with the new Tesla EPA ratings for the P85D, etc. and what this all means, we'd have much better standards and much more transparent and meaningful EPA ratings.