Real World Comparison of a S85 to an 85D Efficiency

[breser]

In this case, breser just wanted to make a trip to a new supercharger. But instead of just hopping in his car and doing it on his own, he took the time to try to find someone on TMC with a car to compare efficiency results against.

In fairness I said something about wanting to do the test and also wanting to do a road trip to the Coeur d'Alene super charger in two different threads and ThortsMD put the two together and sent me a PM. That PM was a critical part of the whole thing since I'm not sure when/if I would have made it out to Coeur d'Alene and I hadn't bothered to post something in the Northwest forum looking for someone to do the test for. So who knows when this would have been done if ThortsMD hadn't taken the time to contact me.

Unrelated to Andyw2100's post I find the discussion of figuring out the gearing without taking the car apart fascinating. It's far beyond my expertise. I'm a software guy.

 

[dirkhh]

This may seem like a trivial issue, but I don't think it is. Because it's not just about this thread. In my mind, this is about what happens the next time someone is thinking about starting a thread like breser's. Does it get started, adding to the knowledge base for all of us, or does the person who might have started it think, "Why bother? I don't need the aggravation."

I'd urge everyone to think carefully about what they're saying when they comment in these kinds of threads.

I'm torn. I cannot help but think critically when I see data collection. Data are a wonderful thing and they are so easily misinterpreted... so I usually appreciate when people are asking critical questions or are criticizing overly broad interpretations.
But yes, challenge the interpretation, challenge the assumptions, but don't challenge the volunteer work that is being done, don't make people feel unappreciated for their effort. Because even if the results aren't statistically relevant or sufficiently comparable, they still help us learn more about the cars, things that Tesla isn't telling us.
So I think the answer is both. Be respectful and grateful to the people who are volunteering to collect the data, but don't overreach trying to interpret the data beyond what they really show.

 

[breser]

I'm torn. I cannot help but think critically when I see data collection. Data are a wonderful thing and they are so easily misinterpreted... so I usually appreciate when people are asking critical questions or are criticizing overly broad interpretations.
But yes, challenge the interpretation, challenge the assumptions, but don't challenge the volunteer work that is being done, don't make people feel unappreciated for their effort. Because even if the results aren't statistically relevant or sufficiently comparable, they still help us learn more about the cars, things that Tesla isn't telling us.
So I think the answer is both. Be respectful and grateful to the people who are volunteering to collect the data, but don't overreach trying to interpret the data beyond what they really show.

The thing is the critical people aren't even challenging overly broad interpretations. For all practical purposes I said in my OP that I couldn't draw a conclusion. I later stepped that back a bit to say that I did think it was reasonable to say that the 85D isn't significantly worse than an S85. But nobody has really attacked those conclusions. They've attacked the methodology, most often the tire differences. I know the tire differences aren't ideal to the test, I said as much in my original post.

If people want to do better tests then by all means please do. I'll try to do better tests assuming I can get willing participants (sorry I don't have two cars like wk057 does so I can do my own tests whenever I want). For instance next time maybe we'll try to offset passenger/cargo weight with ballast weight. Maybe I'll find someone with Michelin Primacy tires that wants to do a test too. Maybe I'll find a whole lot more people that want to go for a drive and we can have more vehicles in more different modes all running side by side. Next time I'll probably draw up a form to track our results to make sure we're tracking all the same numbers. There are a lot of different ways to improve the tests. But I'm not particularly interested in spending money on tires or swapping tires with another owner. I doubt I'd find another owner that wanted to participate under those sorts of terms.

If you want to make suggestions on how to improve the tests or want to participate in helping improve them then by all means speak up. But the hand wavy, change tires response is just an easy and lazy response to make. The first time it was made on the thread, it was one thing, but to see it pop up several times over and over again is really tiresome. It's almost like I spent a bunch of time doing the test and posting the results so people can glance over it and then throw out an obvious response.

 

[NOLA_Mike]

The first time it was made on the thread, it was one thing, but to see it pop up several times over and over again is really tiresome.

I see what you did there.

And I like it. LOL

 

[richkae]

Say what? What's misleading about the title? I say it's a real world comparison. Guess what real world comparisons have differences that are hard to control for. If I said it was a well controlled laboratory comparison then I'd agree it was misleading. But I'm awfully offended by the characterization that the title is misleading.

I apologize for calling it misleading. I do not intend to imply any fault on your part, your post completely described the test parameters.

I almost didn't reply to this but I'm having a real hard time believing this first paragraph is posted with the second one in this post. There clearly is at least one other significant difference, an extra passenger.

You then go on to say there are too many variables. How can there only be one unaccounted for variable but then there are too many.

Yes I get it the test isn't a perfect one. You're all free to go do a better test and post it.

Again I apologize. I did not intend to criticize you, I was overzealous in attempting to dissuade others from drawing conclusions from insufficient data.

The extra passenger is a difference, but not one that is easily quantifiable.
Tire manufacturers do controlled rolling resistance tests and publish the specs.
Those tests control tire pressure, age, wear, and wheel alignment of the vehicles, which are unknowns here.

The discussion should be constructive in helping determine what, if any, conclusions can be drawn, and how to perform a more controlled test in the future, if someone else were so inclined.
I regret that I did not do that.

 

[breser]

I apologize for calling it misleading. I do not intend to imply any fault on your part, your post completely described the test parameters.

Again I apologize. I did not intend to criticize you, I was overzealous in attempting to dissuade others from drawing conclusions from insufficient data.

The extra passenger is a difference, but not one that is easily quantifiable.
Tire manufacturers do controlled rolling resistance tests and publish the specs.
Those tests control tire pressure, age, wear, and wheel alignment of the vehicles, which are unknowns here.

The discussion should be constructive in helping determine what, if any, conclusions can be drawn, and how to perform a more controlled test in the future, if someone else were so inclined.
I regret that I did not do that.

Thanks for the apology, I really do appreciate it. Seeing your post I realize now that you're probably just trying to discourage people running around insisting that the 85D is less efficient than the S85 and that Tesla is lying about it and then pointing at my test. I agree, my test does not support such a conclusion.

 

[dhanson865]

Tire manufacturers do controlled rolling resistance tests and publish the specs.

Please do provide a URL for rolling resistance tests by tire brand, type, size, etc.

I've done lots of research on this in the past and have found the public data spotty at best.

 

[ThortsMD]

Thank you, Andyw2100, for that. This attempt was something of a "back of the napkin" attempt to get additional data to chew on. The effort was certainly not submitted for anyone, especially the OP, to get choked on.

It was a 14 hour drive for me, from Seattle. We had beautiful weather, met lots of nice folks interested in electric vehicles, and I had a lovely day with Breser and his fiancé (thanks for the Model S cookies!). We are a community of passionate like-minded folks, by and large. Share the excitement. Drop the dour.

I think it's both interesting and sad that when a guy like breser (or wk057) goes out of his way to try to post data for something like the above he's likely to get criticized for the test not being perfectly controlled.

Switch tires and do it again? Are you kidding?

In this case, breser just wanted to make a trip to a new supercharger. But instead of just hopping in his car and doing it on his own, he took the time to try to find someone on TMC with a car to compare efficiency results against. Then the two of them took the time to try to control for all the variables they reasonably could control for, record as much data as they could reasonably record, and then breser took the time to write it all up. And while many people are appropriately appreciative some are being, in my opinion, ridiculously critical.

This may seem like a trivial issue, but I don't think it is. Because it's not just about this thread. In my mind, this is about what happens the next time someone is thinking about starting a thread like breser's. Does it get started, adding to the knowledge base for all of us, or does the person who might have started it think, "Why bother? I don't need the aggravation."

I'd urge everyone to think carefully about what they're saying when they comment in these kinds of threads.

 

[Matias]

Say what? What's misleading about the title? I say it's a real world comparison. Guess what real world comparisons have differences that are hard to control for. If I said it was a well controlled laboratory comparison then I'd agree it was misleading. But I'm awfully offended by the characterization that the title is misleading.

I agree. Lets not shoot the messenger.

- - - Updated - - -

Please do provide a URL for rolling resistance tests by tire brand, type, size, etc.

I've done lots of research on this in the past and have found the public data spotty at best.

Car magazines do those tests. For instance Finnish Tekniikan Maailma just did profound summer tire comparison, which included rolling resistance measurements. But of course it is behind pay wall. Those tests are expensive to do correctly, I don't think anyone will publish results for free.

 

[3mp_kwh]

I'm certainly not going to spend money on tires just for this experiment. If someone else with Michelin Primacy tires with an S85 wants to do another test I'm happy to do it with them. But we know that tires are less efficient when new. Jerome told someone as much when the P85D's were new and the economy test procedure requires wear on the cars for the same reason. Comparing efficiencies overall the D was 6% less efficient. Considering the extra cost of an additional passenger and tires, I don't think we can assume that D is worse.

Tesla jack pads are really easy to work with. A tire swap is what I meant. That would be really simple to do. Again, not looking to be ungrateful, but these "5%" efficiency changes from rolling resistance really need to be controlled out. You have consistent, point to point data, showing a watt-hour lag, in the "S-D" vs. "S". Others claim it could be the tires, and not only would it show up if you swapped. It would double. Taking 8 wheels off has to be one of the easiest non-events in how this data could be improved. No pressure

I'm with the wrong crowd, if you guys don't have sockets, electric drills and at least a couple wrenches. I'll grant that the 21mm isn't the most common socket. Six, not twelve point, if you can.

 

[NielsHP]

Following the thought of actually measuring motor currents and frequencies:
Can any of you having access to an actual 85D (mine is still 2 month away) tell, how difficult (and dangerous) it will be to gain access (magnetic) to one of the main wires from inverter to motor, front and back ?

aTdHvAaNnKcSe,
Niels HP

 

[Andyw2100]

Thank you, Andyw2100, for that. This attempt was something of a "back of the napkin" attempt to get additional data to chew on. The effort was certainly not submitted for anyone, especially the OP, to get choked on.

No problem!

And sorry I had not initially given you the credit for having sought out breser. (He later corrected that for me.) In my haste to post, I didn't go back and check the history of the thing.

 

[Lloyd]

I had a ranger out yesterday. While upgrading my firmware to 6.1 xxx.179. He stated that most of the tweaks in this revision were for the D models to improve efficiency, and that I with a P+ would only see minor bug fixes.

 

[breser]

A tire swap is what I meant.

I still think you're going to be hard pressed to find volunteers for that. I don't think even wk057 has done that in his P85 vs P85D comparisons despite owning both of the vehicles. You really think that relative strangers are going to want to swap wheels/tires for a test? If someone wants to do it I'd applaud their effort, but it's going far beyond my intentions of testing.

- - - Updated - - -

I had a ranger out yesterday. While upgrading my firmware to 6.1 xxx.179. He stated that most of the tweaks in this revision were for the D models to improve efficiency, and that I with a P+ would only see minor bug fixes.

I didn't mention it but my car in the posted test had .179 on it.

 

[tcampos]

This is an interesting thread and breser I thank you for posting the data. I'm a P85D owner and this whole thread takes me back 3 months when we were all trying to figure out why our P85Ds were getting such horrible range compared to P85s or S85s.

I'm disappointed by the results posted here, and unfortunately not surprised. I'm disappointed because whether or not the methodology here was fully accurate, it clearly didn't show any improvement of the 85D over the 85 despite Tesla marketing the 85D as the most efficient vehicle in its fleet. I'm not surprised because when it comes to range, I find that Tesla's claims to be way less than trustworthy. I just don't believe them. Yes, some of that are variables, but compared to performance and safety, their track record here is pretty disappointing. With performance specs, the car seems to meet the claims. Safety wise, it seems to be better than claimed. But range wise, except for Cottonwood's recent post, I've not seen people be surprised by the range of the car (and neither have I).

Lest this become a rant, let me make a more constructive statement - one of the things that I have learned from all of these range posts are what the controllable aspects of driving behavior are and the impact they have on range. So, for those who are taking the time to post and share this data, I thank you for your efforts

 

[dirkhh]

I'm disappointed by the results posted here, and unfortunately not surprised.

This very much reflects my sentiment. I am sure that Tesla can demonstrate a controlled environment (flat, no wind, constant 65mph, something like that) where the D has better efficiency than the corresponding RWD Model S. In general driving conditions I would be very surprised if that was the case. The explanations offered by Tesla are hand waving and magical thinking. The car is heavier. More gears are turning, more energy is needed.
And before the haters come crashing down on me... yada, yada, yada. The car is amazing, it could be far worse, yeah, we know all this. It's just that on some topics Tesla doesn't need the slimy dishonest sales people in the car dealerships - they are doing an excellent job challenging that spirit all by themselves. They finally cleaned up their act on displaying the price on the website. But the statements on how far you can go, how quickly you can charge and how much area is covered by superchargers continue to be just misleading at best and fraudulent at worst.

 

[Andyw2100]

This is an interesting thread and breser I thank you for posting the data. I'm a P85D owner and this whole thread takes me back 3 months when we were all trying to figure out why our P85Ds were getting such horrible range compared to P85s or S85s.

I'm disappointed by the results posted here, and unfortunately not surprised. I'm disappointed because whether or not the methodology here was fully accurate, it clearly didn't show any improvement of the 85D over the 85 despite Tesla marketing the 85D as the most efficient vehicle in its fleet. I'm not surprised because when it comes to range, I find that Tesla's claims to be way less than trustworthy.

This very much reflects my sentiment.

While I agree that Tesla may still be overstating their range claims some (I'll come back to this in a bit) I fear you both may be forgetting one pretty major point here: breser did this drive with range mode off. Tesla has said that most of the efficiency benefits of torque sleep require range mode to be on. So reaching the conclusions that you have, based in any part on this test, in my opinion doesn't make a lot of sense.

The main reason I still believe Tesla is overstating the range claims is that I think they have not been forthcoming with the limitations of torque sleep. I think we are discovering that in cold temperatures torque sleep doesn't function at all. I think we're finding that above some relatively low power-draw, torque sleep doesn't function. So those of us living in cold climates won't see benefits from torque sleep at highway speed when it is cold, meaning that we won't have any extra range from torque sleep when it's cold, meaning we'll have far less range when we really need it.

See this thread for details: Torque sleep not functioning properly on some cars? - Page 7

 

[Coiled]

Tesla has said that most of the efficiency benefits of torque sleep require range mode to be on

Slightly off topic, but it would be informative and relatively easy to compare range mode on and off in an 85D in mostly highway driving. Do the same loop back-to-back, with the only difference being range mode. Preferably 30+ miles of mostly flat highway. If that test has been done in an 85D, I missed it. It would be fascinating to see the results.

I will do the test myself in a couple of months if nobody has by then.

 

[breser]

Slightly off topic, but it would be informative and relatively easy to compare range mode on and off in an 85D in mostly highway driving. Do the same loop back-to-back, with the only difference being range mode. Preferably 30+ miles of mostly flat highway. If that test has been done in an 85D, I missed it. It would be fascinating to see the results.

I know this is not exactly a comparision to an S85 but I didn't really want to start a new thread for this.

Did a 107.5 mile round trip today from my house out to Wallace Falls State Park in Washington. I did this route without range mode today and got roughly rated range. I hope to do the same trip tomorrow but weather conditions are not likely to be similar but I was just doing this route because I decided to go out there and because I wanted to compare our efficiency with what EVTripPlanner says. Cabin temperature was at 70°F the whole way. I figured 400 lbs of cargo between us and some things in the car.

Location	Time	Temperature (°F)	Rated Range (miles)	Energy Used (kWh)	Efficiency (Wh/mile)	Distance
Home (Leaving)	14:36	64	242	0	0	0
Wallace Falls State Park (Arrival)	15:58	72	194	13.5	253	53.5
Wallace Falls State Park (Departure)	17:27	73	193	13.5	253	53.5
Home (Arrival)	18:58	65	135	29.4	274	107.5

This trip is largely 2 lane highway that speeds up to 55 mph for some significant hunks but slows down as low as 25 mph through towns. Meaning it's much closer to what the EPA test cycles look like than my trip to Coeur d'Alene last weekend because I'm not going nearly as fast.

Average Speed from Home to Wallace Falls was 39 mph, Speed multiplier of 0.925 on EVTripPlanner Equals this. Average temperature on this trip was 68. Trip to Wallace Falls is a drop in elevation of 351 feet (per EVTripPlanner).

Average Speed from Wallace Falls to Home was 38 mph, Speed multiplier of 0.95 on EVTripPlanner Equals this. Average temperature on this trip was 69. Trip from Wallace Falls is a gain in elevation of 289 feet (per EVTripPlanner).

The following is a comparison to EVTripPlanner

Results from	Distance	Driving Time	Energy Used (kWh)	Energy Used (RM)	Efficiency (Wh/mile)
Actual to Wallace Falls	53.5	1:34	13.5	48	253
EVTripPlanner (85D 19" tires)	52.8	1:22	10.8	36	205
EVTripPlanner (85 S/P 19" tires)	52.8	1:22	11.5	38	218
Actual from Wallace Falls	54	1:25	15.9	58	294
EVTripPlanner (85D 19" tires)	51.7	1:22	11.6	39	225
EVTripPlanner (85 S/P 19" tires)	51.7	1:22	12.2	41	237

Fair warning that EVTripPlanner for the 85D is still in beta and the results may change. Given that I got roughly rated range and EVTripPlanner was saying I should get more I think it's safe to say that EVTripPlanner is probably presuming range mode.

Something else I forgot to mention so I'm editing it in is we beat Tesla's estimate in the energy app for trip both ways. I didn't think to write down what the estimate leaving was.

 

[vvanders]

I manage to beat the trip meter every time (with range mode).

I'd really like to do a comparison with range mode on but it seems to hard to replicate anything more than 2 cars side by side. Just today on a 25mi trip of 50% 60/70mph @ 67F I saw 330 wh/mi out and 270 back (301 avg total).

I think factors like headwind and tempature can easily eliminate any range gains of the 85D.

We're heading out on a 300mi roundtrip tomorrow and I'll try and take better notes.