Another battery/ range question LR AWD Range 250!

[gdgtgeek]

The reason I am posting this is because what I am seeing appears to be well outside the norms of battery degradation.
Battery was charged to 80%. At about 31% (so 49% usage), I got 122.4 miles. Extrapolating this to 100% means (122.4/49% = 250 (rounded up).
That is a significant drop. I had tesla do a remote diagnostic a couple of days ago. This is running 2019.32.2.2 which was updated 2 nights ago. Car is parked in garage (Sentry mode off at home), otherwise runs Sentry Mode and Cabin Overheat Protection (fan only). Parked at work (open parking deck, a little toasty in Atlanta lately).

 

[David L]

I think you're saying that you drove for 122.4 miles and your battery % dropped by 49%. Your actual energy usage (i.e. battery %) when driving is a function of many things, such as speed, driving style, A/C or heater use, elevation change, etc., and is different than the 310 rated range. For an estimate of the rated range, change your display from % to miles, then redo the calculation as (displayed miles / displayed %). I'm guessing it'll be very close to 310 miles.

 

[Tam]

...The reason I am posting this is because what I am seeing appears to be well outside the norms of battery degradation...

To me, the norm for battery degradation is 30% because your warranty is 70% capacity.

You still have a long way to go to reach that warranty claim that it is outside of the norm.

 

[ewoodrick]

Don't extrapolate, charge to 100% and see what it says. Charging to 100% will help the batteries balance and the computer to re-calibrate.

If you are trying to say that you only got so many mile driving, then you are probably going too fast.

 

[gdgtgeek]

Thanks for the feedback. I did spend some time in stop and go traffic. My usage was 28 kWh at 228 Wh/mile

 

[ajdelange]

The reason I am posting this is because what I am seeing appears to be well outside the norms of battery degradation.

You are confusing battery degradation and driving consumption. Driving consumption will vary all over the map depending on driving conditions and driving habits. Driving uphill fast in heavy rain when it's cold out consumes a lot more wH per mile than driving downhill slowly in warm dry weather. Battery degradation refers to the gradually declining capability of the battery to hold charge.

Don't draw conclusions from a single drive. Subsribe to TeslaFi (or other similar program) which will keep track of your individual drives (including the Wh/mi) and charges (including the estimated full charge capacity of the battery after each). Look at the trends.

 

[AlanSubie4Life]

Battery was charged to 80%. At about 31% (so 49% usage), I got 122.4 miles. Extrapolating this to 100% means (122.4/49% = 250 (rounded up).

As others have mentioned, not the way to approach this. Unlike others I think you can extrapolate though and draw conclusions about your battery that are often fairly accurate (though good to do a couple times to ensure consistency). But there are caveats - must be done on a single drive (no vampire or other losses), no sitting in park allowed, and accuracy of the numbers is critical for extrapolation. If you did this (I suspect you may not have), your data suggests your battery has the following capacity for a full discharge:

122.4mi*228Wh/mi/0.49 = 56.9kWh.

For an LR battery this would be really bad. For an MR it wouldn’t be great. For an SR+ that would be impossibly good. (I suspect bad method for data gathering - the key issue being that the meter does not count any losses while in park.)

If you are concerned, do the extrapolation above incorporating the caveats. I recommend for the calculation you use rated miles though and not % (use whatever you want when driving though). It’s actually not necessary to drive at all of course - just look at your rated miles available, at a given charge % (swap between them in the GUI or cross reference via an app the miles and %). It’s quite accurate for assessing available battery energy relative to a new vehicle. Still an estimate, but time and time again here we have been able to predict imminent failures of batteries when these numbers get too far from a brand new battery.

Most complaints here are about a 5% loss or so - which is annoying but not the same thing at all.

For reference, for the LR battery, the full discharge number for a battery with about 4000 miles with minimal age effects (EPA article) as indicated on the trip meter should be:

230Wh/rmi * 310 rmi = 71.3kWh

Might be as high as 72.5kWh if using a 234Wh/rmi constant. But no more than that for sure.

To be clear, this does not imply the battery capacity is 72.5kWh. (It’s ~78kWh.)

There are a lot of threads here showing how to do the calculations and how to figure this stuff out - feel free to refer to them.

If you get a clean datapoint, would be happy to look at it if you post back here.

 

[ajdelange]

Problem is that rnd(30.5001) = rnd(31.49999) = 31 and rnd(79.50001) = rnd(80.49999) = 80 so that the actual percentage change in the SoC could have been as low as 48% and as high as 50%. Thus battery capacity as estimated using the displayed SoC vaulues of 80% and 31% could be anywhere between 122.4*228/.50 = 55814.4 wH and 122.4*228/.48 = 58140 wH. This is just to emphasize that you should not draw conclusions from a single observation.

It is much better to get the data you plug into a capacity calculation from the API through TeslaFi or a similar app than to get it from the screen and when these apps do the battery calculation for you they rely on charging data rather than running data. Same calculation though i.e. kWh_added/change_in_SoC.

 

[darth_vad3r]

The reason I am posting this is because what I am seeing appears to be well outside the norms of battery degradation.
Battery was charged to 80%. At about 31% (so 49% usage), I got 122.4 miles. Extrapolating this to 100% means (122.4/49% = 250 (rounded up).
That is a significant drop. I had tesla do a remote diagnostic a couple of days ago. This is running 2019.32.2.2 which was updated 2 nights ago. Car is parked in garage (Sentry mode off at home), otherwise runs Sentry Mode and Cabin Overheat Protection (fan only). Parked at work (open parking deck, a little toasty in Atlanta lately).

You're doing it wrong

Trip odometer miles (122.4) are real road miles. Advertised range is in "rated dashboard miles".
Switch from 31% display to miles display, read that number, e.g. 96, divide that by .31 and get 310 projected rated range.

 

[darth_vad3r]

As others have mentioned, not the way to approach this. Unlike others I think you can extrapolate though and draw conclusions about your battery that are often fairly accurate (though good to do a couple times to ensure consistency). But there are caveats - must be done on a single drive (no vampire or other losses), no sitting in park allowed, and accuracy of the numbers is critical for extrapolation. If you did this (I suspect you may not have), your data suggests your battery has the following capacity for a full discharge:

122.4mi*228Wh/mi/0.49 = 56.9kWh.

For an LR battery this would be really bad. For an MR it wouldn’t be great. For an SR+ that would be impossibly good. (I suspect bad method for data gathering - the key issue being that the meter does not count any losses while in park.)

Yeah his number is "since last charge" and if this was over the course of days and days, there's kWh and kWh of vampire drain he's not counting -- he said he uses Sentry and fan-only COP ... so ya, probably a ton of energy used that's not shown in the since last charge trip he used to do the (wrong method anyways) math

OP, just switch from percent to miles under display settings and read the number off. Then divide that by percent. It will be 305-310 is my guess.

 

[darth_vad3r]

Using my CSI skills ... OP, was the charge to 80% completed ~30 hours before you took that photo at 7pm last night?
So around 1pm Wed?

My guess is 38% used on driving, 11% used on car staying awake for Sentry Mode.

EDIT: Oh wait, AWD ... revising guess to 37% driving, 12% Sentry mode, and T minus 35 hours accounting for say 2 hours of driving ... 80% charge complete 8am Wed. 31% photo taken 7pm Thu.
~2 hours of driving and ~35 hours of sentry mode.

 

[KenC]

The reason I am posting this is because what I am seeing appears to be well outside the norms of battery degradation.
Battery was charged to 80%. At about 31% (so 49% usage), I got 122.4 miles. Extrapolating this to 100% means (122.4/49% = 250 (rounded up).
... Car is parked in garage (Sentry mode off at home), otherwise runs Sentry Mode and Cabin Overheat Protection (fan only). Parked at work (open parking deck, a little toasty in Atlanta lately).

Thanks for the feedback. I did spend some time in stop and go traffic. My usage was 28 kWh at 228 Wh/mile

Okay, so combining these two posts, you used 28kWh at 228Wh/m and drove 122.4 miles. So, driving usage is normal. It's the non-driving, phantom drain usage that you have to document.

But rather than 28kWh used, you showed almost 38kWh used. That's an additional ~10kWh used. Is it possible to use that much from Sentry and COHP? It shouldn't be but you say Atlanta has been hot, so maybe your COHP is running 24/7. As has been mentioned, I'd get either TeslaFi or Stats to monitor your car's usage, drain and possible deg. At the moment, you have no data to show deg. The data you've shown only indicates phantom drain.

 

[darth_vad3r]

Okay, so combining these two posts, you used 28kWh at 228Wh/m and drove 122.4 miles. So, driving usage is normal. It's the non-driving, phantom drain usage that you have to document.

But rather than 28kWh used, you showed almost 38kWh used. That's an additional ~10kWh used. Is it possible to use that much from Sentry and COHP? It shouldn't be but you say Atlanta has been hot, so maybe your COHP is running 24/7. As has been mentioned, I'd get either TeslaFi or Stats to monitor your car's usage, drain and possible deg. At the moment, you have no data to show deg. The data you've shown only indicates phantom drain.

10 kWh is only 40 hours of Sentry Mode give or take. I'm pretty sure OP charge his car overnight on Tuesday, drove 122 miles from Wed morning til Thu evening and then checked the consumption from 80% Wed morning to 31% Thu evening ... that's a whole lot of Sentry Mode while he wasn't driving.

 

[AlanSubie4Life]

I'd get either TeslaFi or Stats to monitor your car's usage, drain and possible deg. At the moment, you have n

I would not necessarily recommend that. On Stats I often log out to ensure my car is not being disturbed. Haven’t done exhaustive logging but for a while when phantom was very low over the summer, Stats made things slightly worse. This is speculative; not properly backed by firm data.

There is limited value to keeping track of the data if you are not into that sort of thing and tends to make for neurotic behavior.

Plus, most data can be logged historically other ways - just occasionally take a picture of your 90% charge, etc., if you really want some reference points over time.

Is it possible to use that much from Sentry and COHP?

Yes. But of course, we’d need a proper datapoint to assess whether this particular car has a problem or whether this is normal non-driving losses.

 

[ajdelange]

I would not necessarily recommend that.

I definitely would as it records actual power added on charge and power consumed on drive. Thus you are sure phantom drain is not included in the numbers you use for your battery capacity calculation.

On Stats I often log out to ensure my car is not being disturbed. Haven’t done exhaustive logging but for a while when phantom was very low over the summer, Stats made things slightly worse. This is speculative; not properly backed by firm data.

Stats definitely increases phantom drain as it pings the car for status and data from time to time. So does TeslaFi. But it is not included in the logged consumption and charging data.

There is limited value to keeping track of the data if you are not into that sort of thing and tends to make for neurotic behavior.

Yes, and there is limited value in participating in this forum if you aren't into EV's. But I am and am so interested in having the logged data as I imagine many people here would be. I am already neurotic. Let's let the OP decide whether he is interested or not.

Plus, most data can be logged historically other ways - just occasionally take a picture of your 90% charge, etc., if you really want some reference points over time.

If you really want historical data then use a logging program. Taking pictures of your displays will not give you an accurate picture because phantom drain, for example, is included as is, for another example, heat used while waiting for the car to warm up before driving away. This distorts one's picture of how much energy is used during driving. It should also be pretty clear that downloading a CSV and importing the data into your analysis software (or using TeslaFi's built in analysis) is much easier than organizing a bunch of photos, sorting them and then manually transcribing the data. The only reasons to avoid TeslaFi are
1)They charge a fee (well worth it)
2)You have to give them access to your car so they can get the data from it.

Believe it or not, there are those who will not go anywhere near it because of 2).

Yes. But of course, we’d need a proper datapoint to assess whether this particular car has a problem or whether this is normal non-driving losses.

You cannot determine the presence of a problem from a single data point because individual data points can be very misleading. You must look at at least recent history to detect a problem (flames issuing from a battery pack would be an obvious exception to this rule).

 

[AlanSubie4Life]

I definitely would as it records actual power added on charge and power consumed on drive. Thus you are sure phantom drain is not included in the numbers you use for your battery capacity calculation.

Sure, I use Stats myself so I am aware of the capabilities and utility. There are of course other ways to gather this information if you are careful, too. Nearly all of the data gathered by the programs is directly available from the car - the only things that are missing are the appropriate constants to use - which have been documented elsewhere here, and can easily be measured by the end user if desired.

Let's let the OP decide whether he is interested or not.

I was just giving my recommendation. Of course the OP can decide. It's certainly not necessary to get such a logging program to determine whether he has a battery issue, though.

You cannot determine the presence of a problem from a single data point because individual data points can be very misleading. You must look at at least recent history to detect a problem (flames issuing from a battery pack would be an obvious exception to this rule).

Agreed (as I said above). My point was one valid data point would be a start. It's way better than zero valid data points.


Yes, for TeslaFi I don't use it because anything with a monthly subscription fee I am opposed to. It's very handy though. Happy with Stats for $10 (more now of course).

I was just pointing out, mostly, that these programs will make phantom drain worse, generally speaking. At a minimum, we can say for sure they will not improve it. There are many benefits to them - there are some cool features in Stats that are not available from the regular Tesla app. And they have a Watch app which is nice.

Phantom and feature drain are easily quantified without any tracking program by knowing where your rated miles started, where they are now, then using the appropriate vehicle Wh/rmi constant, then looking at the since last charge meter, and subtracting the two numbers. Of course the apps will ensure you miss nothing (actually they miss things too, at least Stats does), and it is all logged, as you say. But not hard to get a basic understanding of the behavior and the associated costs without such tracking.

 

[AlanSubie4Life]

Problem is that rnd(30.5001) = rnd(31.49999) = 31 and rnd(79.50001) = rnd(80.49999) = 80 so that the actual percentage change in the SoC could have been as low as 48% and as high as 50%. Thus battery capacity as estimated using the displayed SoC vaulues of 80% and 31% could be anywhere between 122.4*228/.50 = 55814.4 wH and 122.4*228/.48 = 58140 wH.

Of course. That's why it would be better to have reported data in miles. Doesn't really matter anyway, since the result is completely thrown off by feature drain in this case.

It is much better to get the data you plug into a capacity calculation from the API through TeslaFi or a similar app than to get it from the screen and when these apps do the battery calculation for you they rely on charging data rather than running data.

It's possible to duplicate these specific capacity calculations the TeslaFi comes up with via direct in-car information if you want, AFAIK.

Same calculation though i.e. kWh_added/change_in_SoC.

Some of this information is available in the car during a charging session. The time-integrated input power (energy) is not available, however. (The energy & miles added to the battery is available in the car, which is not the same of course.) I don't know how accurate the tracking of the 3rd-party appss is on the input energy. In Stats it appears to use a fixed assumed efficiency rather than gathering datapoints for input voltage & current over the course of the charge, but I'm not 100% sure of that.

 

[gdgtgeek]

10 kWh is only 40 hours of Sentry Mode give or take. I'm pretty sure OP charge his car overnight on Tuesday, drove 122 miles from Wed morning til Thu evening and then checked the consumption from 80% Wed morning to 31% Thu evening ... that's a whole lot of Sentry Mode while he wasn't driving.

This is what I did!

 

[darth_vad3r]

This is what I did!

Called it

 

[AlanSubie4Life]

This is what I did!

Makes sense. Sentry Mode uses about 1 rated mile per hour (it's around 200-250W).

Anyway, easiest way to determine whether you have a battery issue is to look at your battery %, then switch to miles, and then do the calculation:

Miles / % EDIT: This was already provided above, of course...

This is subject to error as detailed above (% could be +/- a bit)
If it comes out within 5% or so of the rated range you have nothing to worry about. More than that it might be worth more investigation but it also could be a non-issue - it depends.