# Forums

## Survey: Rated Range and Average Energy Tesla Model S

In an attempt to answer the question "Is related range related to driving habits?"

We have heard from Tesla employees that there is a correlation, some drivers say no.

Please fill out this four question survey monkey after charging up your car. A summary of the data will be provided to this forum once collected. Thanks.

http://www.surveymonkey.com/s/569L5MJ

TeslaOwnerBlog | 2013年7月20日

sorry for the typo, should have been "is rated range related to driving habits"

tobi_ger | 2013年7月20日

Oh, I misread it first as "driving rabits"... oops.

Bob W | 2013年7月20日

A few weeks ago I posted a similar survey, but it got buried in a different thread, so it only generated about 30 responses so far (most of which I entered myself after looking into various cars at the Menlo Park Coffee and TESLIVE).

Using the Energy App., you can calculate how many kWh the app. believes are available for use (average kWh/mi * projected range, when average selected = total kWh available).

If you divide kWh available by 300 to 309 Wh/mi (it varies from car to car), you will get the EPA rated range number that is displayed on the instrument panel. There is just no apparent correlation with tire size, battery size, Performance/Standard, or driving habits as measured by the trip meter. Why the rated range calculation varies slightly from car to car remains a complete mystery. It would be nice if an engineer at Tesla could explain it.

On theory is that the fixed Wh/mi no. for each car is correlated to all options that affect the weight of the car, such as the pano roof, the twin charger, child seat, etc., but one would think that the battery size and tire size would be far more dominant factors than these other things. Yet there seems to be no correlation whatsoever, so I'd really like to understand how the Model S chooses which fixed number to use to calculated rated range.

TeslaOwnerBlog | 2013年7月21日

Hi Bob,

I never saw your survey. I check out the Tesla Motors Club forum a lot more than this one. I have gotten a reasonable number of responses so far. Perhaps we can merge our data in a week or so? I'd be happy to take your data in for my five questions.

I would love to have a Tesla person from the engineering department answer this question myself. But in the spirit of reverse engineering I'll do this.

I don't expect the options to have much difference as that would be more a minor weight factor. So far I am surprised how consistent the data is for most cars.

I'm "Owner" on the Tesla Motors Club Forum and my blog is teslaowner.wordpress.com

DJay | 2013年7月21日

In May, because of lost rated miles after a Standard Charge, I tried to understand if Rated Range was modified for driving style as some have suggested. I took a slightly different approach based on the theory that Ideal Range would likely not be modified for driving style and only reflect the state of charge of the battery. If this is true then Rated Range as a ratio to Ideal Range would vary by car if driving style is a modifier to the calculation of Rated Range. I took a small survey and asked Tesla the question below followed by their response.

Question to Tesla: my rated miles as a ratio of ideal miles has been running 86.8%. Today I had 236 rated and 272 ideal miles. I looked at the cars on the showroom floor, one had exactly the same ratio, but the other car had a ratio of 88.6% with rated miles of 235 and ideal miles of only 265. I am under the impression the calculation of rated miles changes with driving style and temperature conditions. Do you think it might be possible my ideal range, which I never look at has been constant after a charge, and the rated range, which is adjusting based on driving conditions has adjust downwards?

Answer from Tesla Ownership Experience. The most likely reason for the discrepancy between the vehicles in our showroom is slight variations between software versions. Only projected range is designed to adjust based on driving conditions, Rated range does not have a variable that adjusts. I’m not sure why you Model S originally had such a high estimate on a standard charge in rated range initially, but settling to about 235 miles range sounds very normal.

Interesting is that my currently my ratio has gone up to 88.2% and has been consistence at this ratio since I last recorded this information at the end of May. I'm not sure if this is after the software upgrade, but is clearly a difference from May 2 when I drafted the email to Tesla. So, I ask, does this ratio vary and may account for the feeling of lost Rated Range?

Brian H | 2013年7月21日

Good blog. Lots of material there.

Rod and Barbara | 2013年7月21日

@ TeslaOwnerBlog – Knowing the rated range at the end of a charge cycle is not the same thing as knowing the rated range Wh/M. The rated range at the end of a full non-range mode charge depends on a number or number of things such as:
- What charge level is selected (50% to 100% available in software v4.5)
- How long has it been since the charge cycle completed (vampire load drains miles in software v4.4 and prior, even in software v4.5 I see slight decreases in the rated miles if left plugged in after the charge is complete)
- How new is the battery (batteries lose capacity as they age)
- How close is the battery to a nominal spec battery (manufacturing tolerances likely mean that every battery does not have the identical capacity; through extensive data collection I determined that my Roadster started with a 98% capable battery)

The rated range Wh/M is not dependent on any of those items mentioned above. Bob W’s suggestion for determining the rated range Wh/M is an excellent method. My only addition would be to suggest that the readings be taken while the car is parked. I’ve taken readings while parked and while driving and found the parked readings to be more consistent.

Finally, I think any thread that gathers and presents actual data from vehicles should be a private thread.

@ Bob W – I entered a few data points in your survey. I will enter some more data. I was disappointed that the summary data does not present the calculated Wh/M for each car type. I was unable to use the raw data to make my own calculations since the raw data isn’t correlated to the car type. If what you say about the Wh/M not being correlated to car type, I wonder if the data needs to be tracked by VIN, or some other unique identifier, until we determine what the affecting factor or factors are.

I’m still perplexed that we have essentially the same vehicle and you calculate about 300 Wh/M in your vehicle and I calculate about 308 Wh/M in my vehicle. On the Energy app graph there is a solid gray horizontal line drawn at the rated mile Wh/M. This line is slightly wider than the horizontal grid lines. My rated mile Wh/M line is clearly displayed slightly above the 300 Wh/M grid line. If it were displayed at 300 Wh/M it would overlay the grid line and the only way to know it was there would be to notice that the 300 Wh/M grid line is slightly wider than the other grid lines. Is your rated mile Wh/M line drawn on the Energy app graph at 300 Wh/M or 308 Wh/M?

J.T. | 2013年7月22日

@Rod and Barbara

I'm curious about "vampire loss" while the car is plugged in. I was under the impression that "shore power" would be used when the car was plugged in. Therefore why would there be any loss at all while plugged?

Shesmyne2 | 2013年7月22日

@Jtod
I'm pretty sure that shore power is only in play if you are actively heating/cooling the MS, or if the battery capacity falls below some threshold.

@Rod and Barbara
We see right around 308 as well for Rated.

Still grinning. ;-)

Rod and Barbara | 2013年7月22日

@ jtodtman – In software v4.4 and previous the Model S “woke up” every 24 hours and topped off the charge if the SOC had dropped 3% or more from it’s previous value. The software v4.5 release notes state: “When Model S is plugged in but not actively charging, it will draw energy from the wall rather than using energy stored in the battery, e.g., when you’ve turned on climate control to cool/ heat the cabin or when you’re sitting in your car and using the touchscreen while it’s parked.” That statement seems a bit ambiguous since both examples described involve the owner demanding electrical usage above and beyond the vampire loss while just sitting idle. Based on some limited empirical testing I believe the wall power handles most of the vampire load but not all of it. For example, during a recent out of town trip I left my Model S plugged in for 8.8 days after the charge was complete. Normally the vampire load would have used about 105 rated miles during this time, but when I got home my Model S had only lost 4 rated miles (231 to 227). In addition, the energy draw from wall power to handle the vampire load is not recorded in the kWh Added display on the charging screen. For this reason, I don’t leave my Model S plugged in after the charge is complete because this behavior hides the power drawn to support the vampire load and I cannot accurately calculate the true cost to drive the car.

TeslaOwnerBlog | 2013年7月22日

@Rod and Barbara

My survey asks folks to fill the rated range after a complete non-range charge, software version. I'm not going to consider vampire load as that would be too hard to ask folks to consider. Batteries don't generate oh so fast, but I do have some indications with folks with heavy mileage numbers. They age more due to use than time.

I don't see why this needs to be a private thread.

nrb | 2013年7月22日

I've had my MS60 for a couple of months now and I'm getting about 2/3 of the mileage I'm expecting. I've been driving conservatively and reading about 327Wh/Mi over 750mi. The typical scenario is that I'll drive 20 miles with a couple of stops with no AC during the day, but have the Rated Range displayed on the dash drop by 30mi.

Putting a watt meter on the charger I'm reading for the last 24hrs 40.4mi travelled with the car reporting 12.8kWh consumed. This is with AC on for about 5mi and again during the day. The power consumed by the charger on the other hand is 21.5kWh. The difference is 316Wh/Mi (reported) vs. 532Wh/Mi (measured) which matches the experience driving.

Seems like I have to build in a 66% safety margin on trips, which now makes me wish I had gotten the larger capacity.

The other thing I can't figure out is why the dashboard and main console display different rated ranges (I've set the energy setting to rated and not ideal).

J.T. | 2013年7月23日

Rod, Barbara, shesmyne2

Thanks for the explanation.

Brian H | 2013年7月23日

nrb;
check with TM about that; those charging losses seem excessive. Your hookup may be (dangerously) under-wired.

They don't affect your mileage once charged up, though, just your expenses.

TeslaOwnerBlog | 2013年7月23日

rnb:

I agree with Brian H that seems excessive. But 327 is pretty consistent with the "average" in the survey so far. But the data is also showing several folks achieving 250-300 range. Something I have not done myself, or gotten even close to.

Folks, please fill in your survey data, the more I can collect the more useful the data will be.

And thanks to those who have!

nrb | 2013年7月23日

BrianH: The Tesla mobile adapter is plugged directly into the meter, so unless the adapter is losing energy there shouldn't be any wiring losses that the meter can see - certainly possible upstream as I'm having to do this on a 110V circuit to use the meter. However, my driving experience is that the car is consuming significantly more power than is actually displayed on the energy display. I wouldn't have thought that the Fan, DRL, Stereo, etc. would have accounted for that much, but its possible that the energy meter is only reporting energy consumed by the power train and not the accessories?

Brian H | 2013年7月23日

The energy meter can only report what power is flowing while the car is charging. There may by accessory draw at the same time, but it has nothing to do with driving usage once unplugged.

Rod and Barbara | 2013年7月23日

@ nrb – There are two issues at play in the scenarios you describe. The first is the rated miles used during your daily commute. If you average about 327 Wh/M, then you should lose about 23 rated miles for every 20 miles you drive plus about 0.5 rated miles for every hour your car sits idle (vampire load) during the time span you are covering.

The second is the energy draw from the grid to replace those rated miles. The energy draw you describe may be accurate depending on the efficiency factor of the Model S mobile adaptor for a 120V circuit.

I have additional information and data I can provide on these issues if you want to contact me at tesla25@verizon.net or exchange posts via a private thread on the forum.

Finally, I don’t understand your statement, “the dashboard and main console display different rated ranges.” What displays are you referring to? The center instrument panel displays rated range. The Charging screen on the touchscreen displays rated range and I’ve never seen this value not agree with the center instrument panel. The Energy app on the touchscreen displays projected range based on the driving experience over the last 5, 15 or 30 miles. The projected range and rated range are not the same thing unless you happen to be driving at the rated range Wh/M over the last 5, 15 or 30 miles and then the two numbers agree in my experience.

nrb | 2013年7月23日

Ok, I'll take some screen shots and notes on driving the next couple of days, although I'm biking into work :)

TeslaOwnerBlog | 2013年7月24日

I have gotten quite a few survey results! Thanks everyone. More data will make the report more statistically correct, so if you haven't filled it out yet, please do so.

Thanks
Tesla Owner

http://teslaowner.wordpress.com/2013/07/21/rated-range-survey/

NKYTA | 2013年7月24日

@ TeslaOwnerBlog
Current results??

hamer | 2013年7月24日

1. Rated range is a constant set by the EPA, given its own mixture of driving conditions, patterns, and speed. It is like the sticker stuck on the side window of a new ICE car: it is what the EPA says that car will do, under it's "average" conditions.

2. The EPA Rated Range for an 85 kWh battery is 265 miles. If you divide 265 into 85,000, you will arrive at 320.75471 (that's where my calculator runs out of significant digits) Wh to drive 1 mile. So if you drive in such a way that you use average 320.75471 Wh/mile, starting out with 85kWh in the battery, you can expect to go 265 miles.

3. Of course, it would be dumb to let the battery run down to 0 miles of rated range. I don't know how much reserve the battery pack has but I would not want to depend on it.

4. Each battery pack is slightly different. Each cell is slightly different. They can't be exactly identical. Some of us will be lucky enough to have battery packs with slightly higher capacity and some with slightly lower capacity. If the difference between the worst and best packs could be as much as 1%, that would be a difference of about 2.65 miles.

5. Driving conditions differ from drive to drive. Outside air temperature, hills, etc. Depending on what else the battery pack has to be doing (heating, air conditioning), this will influence the range you get. Not much, from what I've read, but a small amount. If that were as much as another 1%, that could be another +/- 2.65 miles, on the average.

6. There is what is called "vampire drain," which is the current required to keep the electronics in whatever state they're kept. That seems to be about 1/2 mile per hour. What most people don't seem to think about, is that this is going on 24 hours per day. It takes the same current to keep the electronics going, whether the car is on, off, driving, stuck in a traffic jam, going uphill or down, etc. It may be as much as a mile an hour but you don't notice it unless you're not driving the car when your intuition tells you it should be zero.

7. On the trips odometer screen, where it shows figures such as average Wh/mile since you first drove the car, since you set the trip odometer to zero, it is not at all clear to me where this takes into account this vampire drain. It might be that if the car sits still a lot, eating Wh driving the electronics (a small number), that might increase the Wh/mile figure you see, because it is including Wh used while the car is not moving.

8. The car charges and stops charging while it is sitting hooked up and charging. It clearly would be stupid to have it start charging when the charge has dropped by 1 Wh since it stopped charging. So it waits until the charge has dropped by at least 1 kWh (I don't know how much at least). If you catch it the second it stops charging, you'll have a higher number of Rated Miles to drive than if you catch it 3 hours later.

9. Different people drive at different Wh/mile, using different amounts of heating, cooling, media, etc. This all influences things. That more than accounts for day to day variation. Yes, the battery is probably losing capacity, but so slowly you're not going to notice anything on a day to day basis unless something is broken.

10. The decision by Tesla to give distances in Ideal Miles while Rated miles are much more realistic causes great confusion. It may have been a terrible mistake. That's like my wife figuring that since once, at 5 AM on a Sunday morning, it took us 2 hours and 15 minutes to drive to her father's house, that's all the time we need to allow under all conditions.

TeslaOwnerBlog | 2013年7月24日

@NKYTA

I only have a very preliminary idea of the results. Surveymonkey lets me page through the data but I have to pay money for some automation.

I'm going to wait through the weekend before I do the detailed analysis of the data as I figure a few more cars will be reported by then.

Bob W | 2013年7月25日

@Rod and Barbara wrote:

"On the Energy app graph there is a solid gray horizontal line drawn at the rated mile Wh/M. This line is slightly wider than the horizontal grid lines. My rated mile Wh/M line is clearly displayed slightly above the 300 Wh/M grid line ... Is your rated mile Wh/M line drawn on the Energy app graph at 300 Wh/M or 308 Wh/M?"

On my Energy app, the solid line is drawn at exactly 300 Wh/mi; I see only one wide line in the middle, not a narrow grid line with a wider line above it at 308, like you see. Very interesting! My VIN is 28xx, delivered Dec. 27th, 2012.

Could it be that someone on the assembly line was personally responsible for hand entering the Wh/mi rated range number for each car, so each car got a different number (as shown on the Energy graph)? Could it possibly be geographically based? My car is in No. California.

I've posted the results of my survey responses to date in this public Google spreadsheet, which I've sorted by the last column (calculated Wh/mi used to determine the rated range). The number varies from about 300 to 321, with no apparent correlation to any of the other data collected by the survey.

Note that this is a completely different survey than the one created by the OP, @TeslaOwnerBlog.

Bob W | 2013年7月25日

@Rod and Barbara wrote:
"Based on some limited empirical testing I believe the wall power handles most of the vampire load but not all of it. For example, during a recent out of town trip I left my Model S plugged in for 8.8 days after the charge was complete. Normally the vampire load would have used about 105 rated miles during this time, but when I got home my Model S had only lost 4 rated miles (231 to 227)."

This is a bit misleading (and off topic, sorry). Obviously your car was waking up and recharging the main battery when it got too low, perhaps the day before you returned. You can verify this by looking at your energy company's web site if you have time-of-use (SmartMeter) data available.

If you car doesn't wake up to charge, the daily vampire drain will be the same, whether you leave the UMC plugged in or not.

Power is only drawn from the wall under three circumstances:

The main battery is charging
The heater or defroster is on
The A/C is on

You can verify this by watching the UMC while you are sitting in the car. Turn on the A/C or heater. After a short delay, the green LEDs on the UMC start to move, because energy is flowing. Turn off the A/C, and wait 30 seconds or so. You'll hear a large relay click near the charge port, disconnecting the charger. If you're outside, you'll also hear a smaller relay inside the UMC click. The green LEDs will stop moving. No energy is flowing. None. Turn on the seat heaters. The UMC LEDs will not light. Wall power is not used to power the seat heaters.

The motor, heater and A/C are powered from the main battery. Everything else including the seat heaters, displays, computers, computer cooling fans, parking brake, relays, headlights, interior lights, door lights, horn, 3G hardware, USB ports, and 12V socket are powered from the 12V battery.

The vampire load (mostly the computers) drain the 12V battery, not the main battery. But whenever the 12V battery's voltage gets too low, the car automatically activates the DC-to-DC converter, and that is when some energy moves from the main battery to recharge the 12V battery. This is what we see as "vampire drain," because we can see that this recharging process reduces our rated range.

It seems to vary a bit from day to day depending on when your 12V battery is getting recharged. We have no control over this timing.

If you put a 12V trickle charger on the 12V battery terminals overnight, but leave the UMC completely unplugged, I predict that you would see almost zero vampire drain, because the 12V battery voltage will never get low enough to require any energy from the main battery.

Despite what the Release notes appear to say, the UMC supplies power to the main battery only, nothing else. If you're sitting in the car running the heater or the A/C, the UMC eventually activates, and the charging display shows 1-6 kW of power being consumed by the car. That power is flowing into the main battery, while the same amount is being drawn out to power the heater or A/C, so the main battery stays at a constant charge level instead of being drained.

Turn off the heater or A/C, and NOTHING will come out of the UMC (after about 30 seconds, the LEDs stop moving and the amps used drops to 0).

The only way the 12V battery can get recharged is by taking energy from the main battery (or an external trickle charger). The UMC has no 12V output. The charger(s) inside the car produce about 400V DC (?) to recharge the main battery. I'm 99% certain they do not also supply 13.8V to the 12V battery. It's a completely separate DC-to-DC converter that recharges the 12V battery.

You can verify this by leaving the car plugged in overnight, but set a charge time to 10 am. Wake up at 9 am and you will see how much vampire drain has occurred overnight. It will not be 0.

tobi_ger | 2013年7月25日

@Bob W
Very nice writeup! I'd expect similar detailed descriptions from the manual (if not there, yet).

Bob W | 2013年7月25日

There is a fourth circumstance when wall power is used that I forgot to mention: if it's very cold and the battery heater gets activated.

Rod and Barbara | 2013年7月26日

@ Bob W – Thanks for the info about the placement of your rated miles Wh/M line on the Energy app graph. Also, thanks for the link to your rated mile Wh/M survey results. I look forward to examining the results. I’m confident that we will eventually unravel the rated mile Wh/M mystery. And, finally another big thank you for a very thorough explanation of the charging behavior and vampire draw mechanization of the Model S. Everything you describe fits charging behavior that I have observed, but not thoroughly analyzed. One thing I am wondering about regarding the charging system is the display of the kWh Added on the Charging screen on the touchscreen. I use this number, adjusted for charging efficiency, to determine the amount of electrical power I draw from the grid to charge my Model S. Based on my recent experience where I left my Model S plugged in while I was out of town for 9 days, 12V battery recharging (required because of the vampire load) is not included in this value as the value did not change over the 9 day period. If I had remotely used the A/C to cool the cabin sometime during this 9 day period after the initial charge cycle was complete, do you know if the value of the kWh Added display would have increased?

Rod and Barbara | 2013年7月27日

@ Bob W – I’ve looked over the data you collected in the Google spreadsheet. Your data collection and calculation method is the most efficient one I am aware of, but even in a single vehicle there is a fair amount of scatter in the calculated Wh/M for the rated mile. For example, 17 data points in your spreadsheet were from my vehicle, and these resulted in Wh/M values of 303 to 310 with a mean of 307. Therefore one needs many data points from any particular configuration in order to determine a statistically significant Wh/M for that configuration. Only the 85 kWh Standard vehicle configuration has enough data points to work with. In an attempt to determine if individual cars may have a different rated mile Wh/M assigned to them I wanted to compare individual 85 kWh Standard vehicles. So I can identify the data points associated with your vehicle, can you post the Projected Range for the data points in the spreadsheet for you vehicle? Are there any other individual 85 kWh Standard vehicles in your spreadsheet with multiple data points that you are aware of?

Bob W | 2013年7月28日

@Rod and Barbara: Thanks for the kind words.

As far as I can tell, when a charge is complete, the model S displays how many kWh were added by the most recent charge, and only the most recent (when Charging Units is set to Energy rather than the default, Distance).

So, if you leave your car plugged in for a week, and it starts a charge cycle three different times, I think the display will only show you the results of charging for that third cycle, not a cumulative sum. And of course the power taken from the wall will be about 15% more than what the Tesla displays.

I'll do a test tonight with my Kill-A-Watt meter and 117V charging to see what it reports, that is, the difference in kWh used vs. kWh charged as displayed by the Model S.

I agree that even in a single vehicle there does seem to be some variation, but I've only seen it range from 300 to 302 in my own car, never more, so I haven't bothered to enter it more than once.

For all the data points you entered for your car, did you remember to press Average in the Energy App. before writing down the numbers? Did you ever record the numbers while the Model S was plugged in and the A/C was running? It's best to do it when not plugged in or not using A/C, otherwise the rated range number can jump around a bit.

For the loaner P85 I have tonight (thank you Tesla Fremont) I'm seeing 301.9 Wh/mi for rated range, and I can just barely see a hint of a thin 300 Wh/mi grid line just below the thicker rated range line in the Energy App, so it does indeed show you (in obscure fashion) what number your particular Model S is currently using to calculate rated range, a great observation.

As to the cause of the variation from day to day, perhaps it uses one number when your battery is nearly full, and another when nearly empty? The discharge curve may not be a perfect straight line, so perhaps the algorithm tries to compensate for this by adjusting the Wh/mi number to different values at different points in the discharge cycle. Except I see no obvious pattern when I sort the survey results by projected or rated range, so I guess that theory is not correct either.

Bob W | 2013年7月28日

Using 117V charging at 12A (1.4 kW), which is about as inefficient as you can get, the Model S reported 5 kWh added to the battery, whereas the Kill-A-Watt meter reported 8.13 kWh consumed from the Power Company, which means only 61.5-71% of the energy made it into the battery (assuming 5.0 to 5.99 kWh were actually added; the Model S may truncate rather than round). The battery was nearly full, so maybe more energy is waste as heat compared to charging a half-empty battery.

A Tesla rep. has claimed that 40A charging at 240V (10 kW) is the most efficient in terms of minimal energy wasted, but I have not tried to confirm that. Someone with separate EV metering and an HPWC could do this quite easily (charge one night at 40A, charge the next night the same amount at 70 or 80A, compare power consumed in kWh as reported by the utility meter or the utility company's "Daily Energy Consumption" web app.)

Rod and Barbara | 2013年8月2日

@ Bob W – For all the data points I entered in your survey, I had Average selected in the Energy app, my Model S was on and running in Park in my driveway and not connected to my charger. The A/C may or may not have been running depending what the climate control settings were at that particular time. There is no reason a P85 should have a different rated mile Wh/M than a standard 85. If your standard 85 uses 300 Wh/M, your P85 loaner uses 302 Wh/M, my standard 85 uses 306 Wh/M and all the rated mile lines on the Energy app confirm these numbers (the line in my car is clearly displayed above the grid line, not just barely above it as in the loaner), it may be that the Wh/M for forecasting rated miles is unique to each vehicle. This value could be uniquely assigned to each vehicle to standardize the max rated miles displayed to 265 among vehicles with slightly different inherent battery capacity. I will contact the Ownership Experience team and see if they have any insight about this.

BTW, another method I have used to determine the rated miles Wh/M is to note both the rated miles added and the energy added following a charge cycle. This requires that you sit in the car and change the display setting from distance to energy (or vice versa) before disconnecting the charging cable. Dividing the energy added (in Wh) by the rated miles added produces the rated mile Wh/M. You must be adding significant charge to the vehicle in order to reduce the scatter in this data collection due to rounding/truncating errors. The value I get using this method is slightly lower (304.8) than the value I get from your method (307.6). I use an average of the two methods (306.2) as my best guess for the Wh/M in my vehicle.

Bob W | 2013年8月2日

Well I've got another data point for you. For no apparent reason, my car suddenly jumped to using 305 Wh/mile as the divisor when the rated range was around 188 miles. The line was clearly visible above the grid line.

But after charging, it went back to its usual 300 to 302 Wh/mile, and the rated miles line moved back down.

The calculation just doesn't behave in a consistent fashion, from car to car, and even in the same car. I think it has something to do with the build up of the "reserve buffer" that you documented, something about the fact that you have 0 reserved miles after an extended charge completes, but 5.1 kWh of reserve when you get to 0, and all the calculations are based on when you get to 0.

SCOOBY | 2013年8月2日

Bob - have you considered the possibility that Wh/mile depends on temperature?

Bob W | 2013年8月3日

The temperature here in CA has been pretty consistent. The rated range Wh/mi seems to vary, slightly, with no obvious explanation.

The Rated Range should go down (and rated range Wh/mi should go up), as soon as you turn the A/C or heater on, but it doesn't. I'm talking about the solid line on the Energy App, labeled "Rated Range". Sometimes this line appears just above the 300 Wh/mi grid line (302-307 Wh/mi), and sometimes it moves down and hides it (300 Wh/mi).

Rod and Barbara | 2013年8月4日

@ Bob W – I don’t think the rated miles or the rated miles Wh/M should change when the A/C or heater are turned on. Rated miles are designed to represent a fixed amount of energy use per mile and essentially act like a very finely granulated measure of energy left in the battery. The energy left in the battery doesn’t change when the A/C is turned on; the energy is just used up faster. Likewise, rated miles will be used up faster, compared to actual miles driven, when the A/C is on. Of course, there are times when rated miles don’t appear to be acting as designed, such as the recent movement of the rated miles Wh/M line on the Energy app graph in your vehicle.

SCOOBY | 2013年8月4日

The energy left in the battery doesn't change when the AC is turned on. But it does change when the temperature of the battery increases or decreases. It's nothing to do with energy usage; it's battery chemistry.

Bob W | 2013年8月4日

The rated range is supposed to give you a pretty good estimate of how far you can go, given the "normal driving", based on the amount of energy left in the battery. If you turn the A/C on, I think the rated range number should go down by 10% immediately, as it does on the Ford Focus EV (and maybe the Nissan Leaf as well, not sure), as many tests have shown that using the A/C or heater will reduce your actual driving range by 10% or so.

I really would like to understand how the build-up of the "reserve buffer" affects the rated range displayed. Starting with a full charge (265 rated), do they reduce the rated range display by an extra mile for every mile you drive, until 5.1 kWh have accumulated in the reserve? It does seems like the rated range drops much faster than actual miles driven for the first 10-20 miles, then it drops more slowly. Is that because they're "building up the reserve buffer"? If I do a standard 90% charge, which for me usually stops at 228 rated miles, how many kWh are in the reserve at that point?

SCOOBY | 2013年8月5日

I think you need to consider how you think the rated range meter works. And I don't just mean the algorithm, I mean what measurements do you think are taken?

I don't know how the rated range meter works, but I do know how the battery meter for a cellphone works: it uses measurements of the temperature and voltage of the battery. From this, a fairly complicated algorithm (which includes the rate of discharge) is used to project the remaining power that the battery can deliver. I suspect the rated range meter works in a somewhat similar way.

The "reserve buffer" exists because a voltage must be selected at which the meter reads zero. Often this is set at 80% of the nominal voltage. The battery can still deliver power at this point, but in practice the amount of power available once the battery is outside of its standard operating range is somewhat unpredictable.

Rod and Barbara | 2013年8月5日

@ schueppert – You bring up an interesting point in your Aug 4 post about energy left in the battery changing as the temperature of the battery increases or decreases. Tesla specifically had a problem with this last winter when cold weather resulted in abnormally low rated range values. The problem, as I understood it, was that the rated mile display did not take into account the thermal conditioning the system would provide to the battery as the car was driven. In software v4.3 Tesla introduced new software designed to improvement range predictions in cold climates and this eliminated the wild swings in rated range that I observed during the first several miles of driving in cold weather. If outside temperatures have some effect on measured battery capacity remaining, then this should be reflected in appropriate changes in rated miles, but it doesn’t make sense to me that this would also affect the rated miles Wh/M.

@ Bob W – The rated range provides an estimate of how far you can go if you drive precisely as efficient at as the vehicle was driven during the EPA 5-cycle test. The predicted ranges on the Energy app are designed to provide estimates of how far you can go based on current driving conditions during the last 0.1, 5, 15 or 30 miles.

The battery buffer that is created to keep drivers from being stranded when the rated range reaches 0 is built up slowly and linearly as rated miles are consumed. Rated miles are predicted using about 306 Wh/M, but they are consumed using about 287 Wh/M. So for every rated mile that is consumed the battery buffer is increased by 19 Wh (306-287). After driving 265 rated miles a buffer of about 5.1 kWh will have been created (19 x 265 / 1000). So if your rated miles read 228 after a 90% charge, the system has created a buffer of about 0.7 kWh and will add about 4.4 kWh to this buffer as you drive from 228 rated miles to 0 rated miles.

I determined this behavior based on data I collected during two recent trips of 175 statue miles each. I recorded the 30-mile average Wh/M and predicted range from the Energy app, and the total energy from the Trip display, which was reset at the beginning of each trip. Multiplying the 30-mile average Wh/M by the predicted range results in the remaining battery capacity until 0 rated miles. When this value is added to the total energy already expended on the Trip display this results in the total usable battery capacity until 0 rated miles. This value started at about 81 kWh and slowly decreased in a linear fashion as rated miles decreased to about 77 kWh at the end of my 175 statue mile trip. So, there was no battery “buffer” at the beginning of my trip and there was about 4 kWh of battery “buffer” created by the end of my trip.

SCOOBY | 2013年8月6日

Your analysis is very clever. But it assumes the rated range meter and the energy app have a shared estimate of the energy (kWh) remaining in the battery. This may not be correct. For example, here is one alternative explanation for your observations that does not make this assumption:

1. The battery management system uses a voltage measurement to estimate a state of charge in the range 0% to 100%. (The state of charge is not displayed anywhere but it is available from the API).

2 The rated range meter reads 265 miles (the EPA rated range) at 100% state of charge and decays linearly to 0 miles at 0% state of charge. The vehicle can still be driven because the battery is not dead at the voltage associated with 0% state of charge, but the remaining range is hard to predict because the battery is operating outside of its standard operating range.

3 The energy app estimates the remaining energy (kWh) using the state of charge and the temperature of the battery. It displays an estimated range using this energy estimate divided by the instant or average usage.

So for example, an 85 vehicle at 50% state of charge would always display 133 miles of rated range, but the energy remaining shown in the energy app would vary somewhat based on the temperature. The implied rated Wh/mi you are calculating would also then vary based on temperature.

I have no idea if this explanation is closer to the truth than your explanation. But it fits the same facts and is arguably simpler. It also makes some testable predictions: that implied rated Wh/mi should vary with temperature, and that vehicles at the same state of charge should always show the same rated range (though you would need access to the API to test that).

c.bussert67 | 2013年8月6日

I for one think they nailed the rated range, so kudos to the EPA. I usually average under 308 Wh/mi which is theoretically the rated usage, so I always travel as much or more than the dash shows. Even on all day trips, with 8 hours or so of vampire loss and remote cooling of the interior, I get home about 1:1 with rated mileage.
I just can't fret a few miles here and there. My gas cars never get the same mileage every tank, nor does my S. Its life. As long as my gauge is true to a degree I trust it.
So far as I can simply tell, rated is just whatever charge is remaining divided by 308. I deal is whatever charge is remaining divided by 268. Use the trips page on the dash for this to gauge your current average. If you want a range that changes based on driving habits, use the energy page. But if you drive in a respectably efficient manner, you should be able to hit rated numbers and have no worries. Ideal is truly a challenge though.
I'm so over worrying about the range. I so rarely break over 200 miles in a single day its not worth the stress. I know it will get me where I'm going.

Brian H | 2013年8月7日

Cn;
Get a job with the EPA or Tesla as the perfect test driver. ;)

c.bussert67 | 2013年8月8日

Being an avid hot rodder, I don't think many nice things about the EPA. But man I have thought about working for tesla on several occasions. But until I lose my current job and house, it's not gonna happen.
If I did work for tesla that car would not have come off the line with incandescent bulbs... Anywhere!

Rod and Barbara | 2013年8月10日

@ schueppert - In order to keep from hijacking this thread, I posted an answer to your August 6 post in this thread: http://www.teslamotors.com/forum/forums/important-new-information-about-....

DouglasR | 2013年8月11日

I finally got around to reading this thread. Interesting stuff. I do have a question about filling out the survey, however. What do you mean by "a full non-range mode charge"? I assume a range-mode charge is a maximum charge, but short of that there are multiple levels to which the car may be charged using the current firmware.

TeslaOwnerBlog | 2013年8月11日

@DouglasR

What I meant with the survey is that a full charge before the "trip" section. The maximum Daily number was the data I was looking for.

I think the case is pretty closed that driving habits in terms of Wh/mile have no relationship to Rated Range.

DouglasR | 2013年8月11日

@TeslaOwnerBlog

I must be dense, but I still don't get what you are asking for. My max charge is between 265-275, but my non-max charge can be anything from 50% to 90% of that, depending on where I set it.

I agree with you, BTW, that driving habits do not affect this number.

RZitrin1 | 2013年8月11日

Bob W and Rod,

You are two very smart dudes!

Richard

Rod and Barbara | 2013年8月11日

@ Bob W – Per my August 2 post, I contacted the Ownership Experience team about the different rated mile Wh/M exhibited in my vehicle, your vehicle and your recent loaner vehicle. I did not receive a very satisfying answer. Here is what they said:

“Hi Rod,

After receiving your email, I inquired to members of our powertrain engineering team for any additional insight. Although I did not receive specific information to explain, I did receive the following rationale:

Similar to a gasoline power vehicle, there will be slight variations between different Model S even when built with the same options. Even when brand new, 2 BMWs that are tested for horse power and emissions will have slight variations that will become further apart over time and use.

While I’m sure the formula you are using will be able to give a general sense of wh/mile measurements, it’s not the algorithm that Model S uses, so there are likely further variations based on minor discrepancies between these.

Thank you for the continued support and enthusiasm for our Tesla vehicles.”

So, I’ll continue to collect data and try to conjure up ways to test theories that might explain the differences we have seen.