P85D range and highway battery performance

[stevezzzz]

Here are the assumptions I used in EV Trip Planner:

19" wheels
Speed factor 1.0
Climate 66
OAT 50
Payload 500lb

It estimated 265 Rated miles for this uphill, 251 mile leg. We used 233 Rated miles and could have arrived with perhaps ten miles more if we'd kept the speeds down over the last 50 miles. In fact, I believe we could have made Cordes Junction if we'd stuck with strict hypermiling.

 

[MarcG]

Well done Steve! Will you be adding all your trip data to the Google spreadsheet?
P85D Uninterrupted Hwy Trip Data - Google Sheets

It seems that these days you and I are mostly adding data, but it's nice to see anyway. Thanks in advance!

 

[Andyw2100]

Here are the assumptions I used in EV Trip Planner:

19" wheels
Speed factor 1.0
Climate 66
OAT 50
Payload 500lb

It estimated 265 Rated miles for this uphill, 251 mile leg. We used 233 Rated miles and could have arrived with perhaps ten miles more if we'd kept the speeds down over the last 50 miles. In fact, I believe we could have made Cordes Junction if we'd stuck with strict hypermiling.

So, in the other thread I had asked you if you'd include your data, as it compared to EV Trip Planner, in that EV Trip Planner comparison thread ( Comparing P85D Torque Sleep efficiency (versions .139 and .140) to EV Trip Planner ). But seeing the above, I realize that unless you can use the appropriate speed factors in EV Trip Planner, that data probably wouldn't be all that helpful. You had written:

Long story short: we made Flagstaff, without stopping to charge, with 20 Rated miles remaining, at an average of 277 Wh/mi. Speeds were mainly 55-59 in TACC (which felt like standing still on the straight, lightly-used roads where most people blew by us doing 80);

What I'm hoping to do in that thread ( Comparing P85D Torque Sleep efficiency (versions .139 and .140) to EV Trip Planner )is have people post their actual data as it compares to what EV Trip Planner predicts for what they actually did, so that we can use that as a benchmark. So, for example, we could say that you beat EV Trip Planner by x%. But to do that, everyone has to make the EV Trip Planner inputs match up with what actually happened, and in your case using a factor of 1, to indicate travelling the speed of traffic, wouldn't do that.

So if there's any way you'd be able to go back (or for the legs you are still taking now) record the data in a way that would be useful for the benchmarking too, that would be great. It seems like you are one of the people really getting fantastic efficiency, so it would be great to be able to get that value for x for you for some of the legs of your trip.

As a comparison, I'm still generally either seeing efficiency a little worse than EV Trip Planner would predict, or on some trips a little better. I'm working on pinning down what is going on, but my theory is that something is wrong. So having examples of P85Ds that are beating EV Trip Planner by 10% or more, with all relevant factors entered correctly would be really interesting to see!

Thanks!

 

[stevezzzz]

I see what you're asking. But I use EV Trip Planner a little differently: for planning I like to set it up for speed factor 1.0 and input my best guess as to the OAT, climate settings and payload. That way I know, based on my experience over several long road trips in my S85 and now in my P85D, whether I'm going to have to watch my consumption like a hawk, or not.

When there's any question of whether I have adequate range I'll start out slower than I predict will be necessary to complete the leg. As things go along I'll adjust my speed, up or down, depending on what I encounter. For instance, during yesterday's Blanding-Flagstaff run I started out driving at 59 mph (with a posted speed limit of 65 mph). Fifty miles into the trip we recognized that a significant headwind had developed and slowed to 55 mph. Later, when the opportunity arose to draft a big pickup towing a boxy, enclosed trailer, I allowed TACC to hold a safe following distance and watched as the speed varied from 60-70 mph. When the pickup turned off, I slowed to 55 again. Finally, when it became clear that we had Flagstaff made with a comfortable margin, I picked up the speed to 69 (PSL+4) for the rest of the trip and watched as my margin eroded from a predicted 12% of the pack to 7% by the time we rolled into Flagstaff. All the while we employed other hypermiling techniques, like limiting the max power draw during short uphills and allowing our speed to decay, so long as we weren't impeding traffic.

Climate control is another wild card: with the sun shining and temps in the forties, I couldn't leave it completely off without roasting slowly, but neither did I really need heat or A/C. What I did was to set the thermostat to 64, turn off the A/C, and use the fan speed to regulate the cabin temperature: low fan speed allowed the temperature to climb, high fan speed cooled us down.

So, what would you have me enter in EV Trip Planner to adequately model what in real life is a very fluid situation? I may have averaged about .95 (which is 59 mph in this case, yielding an EV Trip Planner estimate of 254 Rated miles required), but how does that capture the moment-by-moment nature of decision-making when driving for efficiency?

 

[Andyw2100]

So, what would you have me enter in EV Trip Planner to adequately model what in real life is a very fluid situation? I may have averaged about .95 (which is 59 mph in this case, yielding an EV Trip Planner estimate of 254 Rated miles required), but how does that capture the moment-by-moment nature of decision-making when driving for efficiency?

It sounds like this trip / leg probably was just too complicated a trip / leg to be a good candidate for a good benchmark example. Perhaps if there's a shorter leg at some point, where you don't wind up needing to hypermile, or where you don't make many adjustments, we might use that leg.

For this leg, as a rough approximation, if we go with your actual 233 rated miles used, and the 254 estimated, that would mean that you beat EV Trip Planner's estimate by 21/254, or a little over 8%. That's completely in line with what people who are seeing better efficiency have been doing.

I really appreciate the effort!

Thanks!

 

[ArtInCT]

Andy:
Last time I looked, EV Trip Planner did not yet have the P85D or 85D as a Model S choice type.
Are you doing the data logging to help EV Trip Planner develope some P85D performance characteristics?

 

[Andyw2100]

Andy:
Last time I looked, EV Trip Planner did not yet have the P85D or 85D as a Model S choice type.
Are you doing the data logging to help EV Trip Planner develope some P85D performance characteristics?

No, not at all.

I'm trying to get some good benchmark data so that people here can have some idea of where they stand in comparison to others.

In my case, I am pretty sure I am not getting the efficiency I should be getting with torque sleep. But with weather conditions and all the other variables in play, it's really hard to say that with any certainty. If I keep logging a bunch of my trips, and they all wind up, say, between 3% better than and 5 percent worse than EV Trip Planner would predict, that tells me something about the kind of efficiency I am getting. But that doesn't tell me anything about how I stack up against other P85Ds.

So if lots of people all offer up some data, we'll really have something.

From the data we have, it seems, in rough terms, the people who are pretty certain they saw an efficiency improvement with .139 or .140 are beating EV Trip planner by somewhere in the neighborhood of 10%, give or take.

All of us are using just a Model S in EV Trip Planner, (hopefully with the correct wheel size) as you are correct, it does not yet offer an option for a P85D. I think it is actually listed as a Model S, and then it also says something like S/P, but basically there is just one Model S with 21" wheels, and one with 19" wheels, for now.

Having a good amount of data will allow everyone to see where they stand in relation to others.

 

[MarcG]

In my case, I am pretty sure I am not getting the efficiency I should be getting with torque sleep. But with weather conditions and all the other variables in play, it's really hard to say that with any certainty.

I think you nailed it. Rain, in my case, has been the single worst contributor to raising consumption for all the runs that I've compared so far.

Once the weather warms up and dries up in your area, I think you'll see substantial improvements. Hope it's not a long winter for you... it's already summer for us out here in NorCal

 

[ArtInCT]

Ok, I get the drift now.... little snow joke there....

There are many challenges you all face... the data elements entered into EV Trip Planner are for each individual driver and their environment and payload. They will not be the same.

Additionally, driving styles will probably have a real lot to do with outcome. I can say this with real world experience. Frankly I can drive my wife's Prius and get about 50 MPG day in and day out.... she has a hard time getting over 43 MPG just about ever. She drives the Prius like a normal ICE car, giving it no chance to ever get into EV assist mode.

My wife and I are at opposite ends of the style spectrum.

EV Trip Planner has no data element or operand for driving style...

I am a pretty firm beliver in the fact that Tesla really needs to implement some sort of indicator that informs the dual motor driver that Torque Sleep is invoked. Such an indicator, preferably in the driver's display. would quickly give feedback to the driver what driving techniques are best to invoke TS.

Probably, the right hand power gauge needle just needs be glowing, or some ZZZzzz's indicator showing or some such.

The Tesla implementaion of TS appear to be so good and seemless that many drivers are not aware when it is working or when it is not and what driving techniques and behaviors allow quickest entry and the ability to maintain TS.

I feel that all P85D and 85D owners should ask for such and indicator to be implemented as soon as possible.

 

[gg_got_a_tesla]

Didn't take detailed notes like others here but, have done exceedingly well since getting .139; about 823 mostly-highway miles - including a trip up to Squaw Valley over the long weekend - at 323 Wh/mile. Range mode On, Sport mode. Normal AC use throughout.

 

[Andyw2100]

I think you nailed it. Rain, in my case, has been the single worst contributor to raising consumption for all the runs that I've compared so far.

Once the weather warms up and dries up in your area, I think you'll see substantial improvements. Hope it's not a long winter for you... it's already summer for us out here in NorCal

The problem is though I don't have as good data from before .139, the data I do have, and the feel for it that I have would seem to indicate that there has been no significant change. The weather could be dampening, pun intended, efficiency all around, but torque sleep should still be improving it over what it was before I had torque sleep, and I see little or no evidence of that.

There are many challenges you all face... the data elements entered into EV Trip Planner are for each individual driver and their environment and payload. They will not be the same.

Additionally, driving styles will probably have a real lot to do with outcome.

You can enter payload into EV Trip Planner, and I have been. As for driving style, I agree that adjusting for that is an issue. But some of the trips I'm tracking for my wife take place at 6:00 or 6:30 in the morning, with almost no traffic, and the TACC set on the highway. Her driving style wouldn't really come into play there. She's got 30 minutes straight of the car deciding whether to torque sleep or not based purely on the terrain and the fact that she set the speed to 65 or 68. That 30 minutes is about half of each trip I'm charting. I assure you that there's not enough wrong she could be doing in the other half of each trip to account for not seeing the effects of torque sleep. And again, I'm comparing, at least anecdotally, trips where she was the driver before and after. So the driving style is a constant.

As for driving style making the benchmarking less valid, I agree, to a point, but I still think the benchmarking will help draw a big picture. Perhaps someone beating EV Trip Planner by 9 percent increase would really be beating it by 12 percent if they drove more efficiently, but if we gather enough numbers, the extremes should cancel each other out.

I am a pretty firm beliver in the fact that Tesla really needs to implement some sort of indicator that informs the dual motor driver that Torque Sleep is invoked. Such an indicator, preferably in the driver's display. would quickly give feedback to the driver what driving techniques are best to invoke TS.

Probably, the right hand power gauge needle just needs be glowing, or some ZZZzzz's indicator showing or some such.

The Tesla implementaion of TS appear to be so good and seemless that many drivers are not aware when it is working or when it is not and what driving techniques and behaviors allow quickest entry and the ability to maintain TS.

I feel that all P85D and 85D owners should ask for such and indicator to be implemented as soon as possible.

I certainly agree with that. I've been trying for almost two weeks now to get some information about my car and torque sleep, and I'm having a heck of a time getting any where. If I could see when it was active I wouldn't be doing this, provided it is, in fact, active.

I think part of the problem I'm having getting answer is I happened to get lucky enough to have the rep I initially contacted be someone who was just getting trained. I've been uncharacteristically patient so far, but my patience is wearing thin, and in the next few days I'm going to get a little more assertive in trying to pin down some answers. The only information supplied by an engineer was that torque sleep was active, but the numbers supplied at the same time as the answer don't support that.

 

[billarnett]

...
I am a pretty firm beliver in the fact that Tesla really needs to implement some sort of indicator that informs the dual motor driver that Torque Sleep is invoked. Such an indicator, preferably in the driver's display. would quickly give feedback to the driver what driving techniques are best to invoke TS.

Probably, the right hand power gauge needle just needs be glowing, or some ZZZzzz's indicator showing or some such...

How about splitting the power meter bar into two parts one for each motor as a sort of stacked bar graph? Make the two parts slightly different colors. When the rear motor is "asleep" you would just see the one bar for the front motor.

 

[Saghost]

How about splitting the power meter bar into two parts one for each motor as a sort of stacked bar graph? Make the two parts slightly different colors. When the rear motor is "asleep" you would just see the one bar for the front motor.

I like this idea, but if they do it they should use either three or four colors - front motor, rear motor, and everything else - which will presumably be driven by HVAC.

(The fourth color would be an explicit HVAC number, but I'm having trouble coming up with any other multi-kilowatt potential loads in the car, so it is likely a waste.)
Walter

 

[billarnett]

Good idea to show "everything else". Is there currently a way to see that?

 

[Saghost]

Good idea to show "everything else". Is there currently a way to see that?

I was assuming it is all in the dash gauge, but maybe that's a question to verify with current owners:

When you initially do a cold start in this winter weather, after you turn the car and HVAC on but before shifting into D and starting to move, does the power gauge show 6-8 kW, or 0?

 

[billarnett]

I was assuming it is all in the dash gauge, but maybe that's a question to verify with current owners:

When you initially do a cold start in this winter weather, after you turn the car and HVAC on but before shifting into D and starting to move, does the power gauge show 6-8 kW, or 0?

I tried turning on all the heaters and lights while sitting in,my garage: the power level showed zero :-(

 

[stevezzzz]

I tried turning on all the heaters and lights while sitting in,my garage: the power level showed zero :-(

Were you still plugged in, perhaps? I definitely see the heater load on the meter when stopped at a traffic signal.

 

[andrewket]

Were you still plugged in, perhaps? I definitely see the heater load on the meter when stopped at a traffic signal.

Same here. Lights etc draw so little you won't see it. I don't recall if you're plugged into shore power running the HVAC if the energy use is shown or not.

 

[electrish]

The problem is though I don't have as good data from before .139, the data I do have, and the feel for it that I have would seem to indicate that there has been no significant change.

I have been quietly commiserating with you.
I also have been on .139 forever, never got .140 and energy efficiency was never great, albeit weather has been cold and icy here.

Today, however, I received the .167 update and I feel that this has been a big improvement for me.

Also, the "insane" setting allows range mode now (I don't think that was there before).

So I hope you'll see some improved energy efficiency with this update!

 

[MarcG]

Also, the "insane" setting allows range mode now (I don't think that was there before).

I can confirm Range Mode has been untied from SPORT mode, meaning it can be set to ON while still in INSANE mode, since 2.2.139.

As for 2.2.167, I have yet to test it out as I just upgraded this afternoon after taking a long trip out of town. Once I get a chance to do my usual runs I'll be able to tell whether or not it improves efficiency over .139/140.

Will keep you all posted, as usual.