Energy accounting

scaesare · Jul 1, 2013

I've come across several threads that discuss the various energy consumption display... some of which were as much conjecture as anything... which is understandable given that we are all guessing at exactly what our Model Esses are doing under the covers, and the fact that he official guides are relativity light in discussing the subject.

I am trying to reconcile my personal experience based on my understanding of what information the car is presenting to me in the several areas regarding charge and range. In other words: "Where's my energy going?"

BACKGROUND: In my month+ of ownership, it's been my experience that after a charge, there is some energy usage at rest (aka the "vampire loss"). I've also just hit a lifetime average of the magical 308 watts/mile (after 3000 miles on the odometer).

I also just took two 500-mile road trips, on one of which I had opportunity to drive the S from a full-range charge down to 0 miles (actually about a mile past).

All numbers used in this post are with the displays set to "Rated" miles.

OBSERVATIONS: Here is what I've noted regarding the displays-

Dash "Speedo Range": After a "standard" (90%) I typically have on the order of about 242 miles of range. A "max" (100%) charge nets about 270. For both my normal driving habits, as well as the road trips, my miles driven have always been less than what the speedo range estimate has been. On one leg of my road trip, I drove a mile past 0 in to "CHARGE NOW", despite having started with 270 miles estimated, and only 245 miles driven, and my reported trip average being below 308 W/mi.

Trip Meters: After a standard charge, which ends a couple of hours before I leave in the morning, my trip-meter w/mi "Avg since last charge" numbers are often very-high when I first start driving. Sometimes as high as 500-800+ W/mi. As soon as I start driving, the averages fall dramatically, and I typically net less than 308 W/mi per charge. My "miles driven since charge" trip meters are always less than the speedo range estimate.

Console Energy App: Using the 30 mile / Average settings, I typically see my energy usage in the 280-310 W/mi range after driving for a bit. The "estimated range" indicator (pointer) typically indicates a few miles MORE range than the speedo range estimate, which was the case for the majority of each leg of my road trip.

CONCLUSIONS/QUESTIONS: I have a few conclusions, but each seems incomplete and raises additional questions. Particularly I'm trying to figure out where energy used for non-locomotive draws (A/C, headlamps, heat, pack thermal management, etc...) is accounted for:

1. Given that the trip meters start with very high energy usage averages after charging, but before driving, I assume they are accounting for some energy usage elsewhere? Yet if this is the case, why do I get less than range-estimated miles if this average is under 308 for the duration of the time since last charge? Does it account for some energy usage, and not other? Or is the artificially high immediate energy usage an artifact of the system "dividing by zero" before the first tenth of a mile is registered on the trip meter?

2. The speedo range estimate always decreases faster than actual miles driven (according to the odometer). Is this accounting for ALL energy used in the system? Was my being able to drive slightly past "0 miles range" because I was below 308 W/mi and therefore the estimate was wrong, or did I dip in to a "battery reserve" Tesla built in to the system?

3. The console energy app would seem to think that my averages would give me more range than my dash does. Which is right? Does the console ignore non-locomotive energy draw?

All in all, it feels like the speedo-range estimate is "safest"... primarily because it seems like the most conservative. However it does feel like there's a disconnect in that even when using 10-15% less energy driven than the rated estimate of 308 W/mi, it still counts down faster then miles driven...

Any thoughts appreciated.

sublimaze1 · Jul 1, 2013

scaesare said:
CONCLUSIONS/QUESTIONS:
1. Given that the trip meters start with very high energy usage averages after charging, but before driving, I assume they are accounting for some energy usage elsewhere? Yet if this is the case, why do I get less than range-estimated miles if this average is under 308 for the duration of the time since last charge? Does it account for some energy usage, and not other? Or is the artificially high immediate energy usage an artifact of the system "dividing by zero" before the first tenth of a mile is registered on the trip meter?

This may be an answer. I have never thought of it in this terms. I considered it the movement of a very large piece of metal a short distance from a standstill divided over a short interval of time - which gives you that initial 1513wH/mi tick on the screen before it settles down to 342 after a few miles.

2. The speedo range estimate always decreases faster than actual miles driven (according to the odometer). Is this accounting for ALL energy used in the system? Was my being able to drive slightly past "0 miles range" because I was below 308 W/mi and therefore the estimate was wrong, or did I dip in to a "battery reserve" Tesla built in to the system?

I do not think it is a reserve phenomeon. Try this: Let your charge go down to 45 or 50 (say on Friday) and look at the number before you go to bed. Get up Say am (you will be somewhere from 37-48) and drive around the neighborhood at 25 mph. You will see that range go up, simply due to the fact that the computer now sees you are at the top of the range bell curve, because you are Driving Miss Daisy. It is more pronounced in the Rated Range than in the Ideal Range setting.

3. The console energy app would seem to think that my averages would give me more range than my dash does. Which is right? Does the console ignore non-locomotive energy draw?

I'll pass on this.

All great concepts and great questions. !!!

scaesare · Jul 1, 2013

sublimaze1 said:
This may be an answer. I have never thought of it in this terms. I considered it the movement of a very large piece of metal a short distance from a standstill divided over a short interval of time - which gives you that initial 1513wH/mi tick on the screen before it settles down to 342 after a few miles.

Yeah, this may just be a math quirk, and not accounting for non-locomotive draw at all...

sublimaze1 said:
I do not think it is a reserve phenomeon. Try this: Let your charge go down to 45 or 50 (say on Friday) and look at the number before you go to bed. Get up Say am (you will be somewhere from 37-48) and drive around the neighborhood at 25 mph. You will see that range go up, simply due to the fact that the computer now sees you are at the top of the range bell curve, because you are Driving Miss Daisy. It is more pronounced in the Rated Range than in the Ideal Range setting.

Hmmm, I've not ever seen my range go up on the speedo estimate before... but I'll try that. That begs the question then: In my scenario while averaging <308W/mi why did my range estimate not revise itself, and instead let me get in to "zero territory"? If i was using less than expected energy, shouldn't the range estimate have started decreasing MORE SLOWLY than my odometer was coiunting off miles driven?

Thanks for the reponse... hopefully as a community we can figure this out...

sublimaze1 · Jul 1, 2013

scaesare said:
Hmmm, I've not ever seen my range go up on the speedo estimate before... but I'll try that. That begs the question then: In my scenario while averaging <308W/mi why did my range estimate not revise itself, and instead let me get in to "zero territory"? If i was using less than expected energy, shouldn't the range estimate have started decreasing MORE SLOWLY than my odometer was coiunting off miles driven?

Thanks for the reponse... hopefully as a community we can figure this out...

Maybe I did not read carefully. Sorry. But, I know you have to be somewhat heavy on the usage for the computer to give you that credit.

That is:

If you are spending 335wh/mi and then go to 308wh/mi to save, the computer won't help you out. It doesn't see that as much a difference. Now, hoon the beans out of your car for the charge and use 375, then cut to 25mph for 10 miles or so, and you should (I most certainly did) see that calculated mileage go upwards. I was on my way back from work in December after having the car outside in the 35 degree temperature for the whole day. Sucked the life out of it (I was on firmware - like - 0.0.1). I had 47 miles rated to get 33 miles actual. About 10 miles up the Tollway, I was on 17 rated and one could have sharpened a pencil in my ...

... anyhow, that was my first case of range anxiety, and having been obsessed with this car for a long time, I knew that 25 was the magic number. Got on the access road, and putted around for five minutes at 25 mph, and my range (rated) went from 17 to 28. Putted all the way home. So I know for a fact (at least in the older firmware) that this does, indeed, happen.

scaesare · Jul 1, 2013

I'm sure you are right, and it does... I'm just trying to figure out why it wouldn't also do it (albeit more slowly), if your average energy consumption is gradually dropping over a more extended period of time... after all numbers are numbers.

scaesare · Jul 10, 2013

Revisiting this....

I specifically looked at my "Since Last Charge" counter this morning when I unplugged and left for work. Energy usage was 0 until I registered the first tenth of a mile on the mileage counter, and then it jumped up to 1143 wh/mi. It then ramps down, so as we earlier surmised, it's accounting for SOME non-locomotive energy.

I also thought about the speedo "Rated Miles" estimate... it will decrease when parked (due to vampire losses from onboard computers, etc...), so it ALSO must be accounting for non-locomotive energy.

The question then remains: if both of these counters are clearly including non-locomotive energy usage, why does the rated miles estimate fall below the actual miles traveled, even wen the avg. energy usage is 308Wh/mi or less?

In other words, if the rated miles is based on the entire car using 308W for every mile traveled, then shouldn't the rated miles = distance traveled @ 308wh/mile avg usage over the lifetime of a charge?

The only things I can think of are:

1) The energy meters account for SOME additional energy usage (i.e. computers), but not others (HVAC & headlights).

or

2) The speedo Rated Mile estimate is conservative in that it will calculated "downwards" based in the initial large avg energy usage, but doesn't as aggressively "return" miles to the estimate as the avg continues to drop over a trip.

I just did another ~400 mile road trip, and in all cases my range estimate dropped by 15 miles or so as compared to actual distance traveled in the first 1/3rd of the trip, and then settled to dropping slightly faster than actual distance for the remainder of the trip.

I had ~140 miles in the pack, and 120 miles to my destination. After 40 miles, I had 85 miles estimated range left, and 80 miles to go, with an "since charge" average of ~320Wh/mi at that point. After another the next third of the trip, I had 42 miles range left, and 40 miles to go, even though I was averaging only 305Wh/mi since charge by that time. At the end of the trip, I arrived home with exactly 0 miles range left, but had dropped the since charge average down to ~290Wh/mi.

So I lost 20 miles of range, despite averaging "better" than 308Wh/mi...

I still don't quite get it.

bluetinc · Jul 10, 2013

SC,

I've seen the same thing in my trips of late, and I've documented it with pictures

. I'd also like to know what's going on.

Peter

scaesare · Jul 10, 2013

Ah... pics are good.

I might drive at a steady 55 for a while... and then turn on every accessory I can think of (heat, headlights, wipers, dome lights, defrosters. etc...) and see if the dash/console meters register a change...

boilerbots · Jul 24, 2013

Let me throw in another piece of information that I just discovered.

I put a power meter on my cars outlet to try and determine the cars charging efficiency. So, I drive every day during the week and let the car sit in the garage over the weekend. Since I am trying to focus on charging I try to unplug the car soon after charging is complete. During the week things look reasonable, trip meter shows power coming out, my power meter logs the power going in. The loss is pretty consistent until the weekend. I wake up Saturday morning and unplug the car and it sits all weekend. I get in to drive Monday morning and the estimated millage is already down 13 miles.

I pull up the trip log before I move and it still says 0.0 kW used, funny. I drive to work and home then charge. The next day the power meter shows that it took an additional 4kW/hr of power even though the trip meter shows that my entire day used the same total power as the previous week day trips.

Interesting to note that 13 miles of lost range times my average of about 305w-hr/mile equals 4kW/hr or the missing energy from the trip log.

What might be going on here?

For starters the energy was lossed over about 50 hours of time. That is about 80w per hour. Since the dashboard power meter seems to be able to measure at least 120kW on my car and a mere 80w is about 0.06% of 120kW this just might be to low to measure. Perhaps the trip meter decays based on idle time and perhaps can not accurately measure small power consumption.

All I know is that if you want to know how much power you are actually using per mile you need to measure it with an external power meter.

mknox · Jul 24, 2013

scaesare said:
BACKGROUND: In my month+ of ownership, it's been my experience that after a charge, there is some energy usage at rest (aka the "vampire loss"). I've also just hit a lifetime average of the magical 308 watts/mile (after 3000 miles on the odometer).

For the first time since owning the car, I left it sitting while I was away on vacation. I had it charged to 60% (with charging scheduled for 10:00 PM) and left it plugged in. Temperatures were typical to above typical for July in Southern Ontario, Canada.

I noticed that every other day, the car re-charged itself using exactly 5 kWh of electricity according to my home energy monitor (dedicated set of CTs on the EV outlet). On the mobile app, I could see that every other day, the car would show "Scheduled for Charge", then commence the 5 kWh charge at 10 o'clock, which would top it back up to the 60% level I had set.

It looks like the car will sit and discharge until some pre-determined percentage below the set point at which time it will schedule a charge if the charge time is used. I presume it would just start to charge immediately when this threshold is reached if a charge time is not pre-defined.

So in conclusion, it appears my car on v4.5 of the software will lose 2.5 kWh per day (76 kWh per month) to the "vampire".

scaesare · Jul 24, 2013

Interesting additions to the discussion.

So it appears that we can quantify "externally" what the power loss is, it's still difficult to reconcile what the displays are reporting.

It seems that the trip meters indicate non-locomotive power usage associated with some things, as the above mentioned "immediate power usage" spike when starting on a drive after a charge indicates (i.e. immeidately showing a 1100Wh/mi when you first start out).

It seems that when I first sit in the driver's seat and the car wakes up, it immediately spins up HVAC (and possible battery conditioning) compressors, among other things. It feels like the harder these systems are running before I actually put the car in Drive (i.e. a hot garage and the HVAC ramps up to full), the larger the initial reported usage spike.

So I'm going to venture that perhaps the trip meter energy usage reports Power Consumed while the car is Awake. I suspect vampire losses while the car is asleep is NOT reflected.

What this still doesn't explain is: If the trip meters account for energy usage while the car is awake, why does rated miles still drop faster than driven miles if my average is below 308W/mi AND I start drive immediately after charging and thus no vampire losses are accrued?

mknox · Jul 24, 2013

scaesare said:
So I'm going to venture that perhaps the trip meter energy usage reports Power Consumed while the car is Awake. I suspect vampire losses while the car is asleep is NOT reflected.

I think you're right. I believe the car's meter only records power consumed (from all sources) when the car is actually on. It therefore doesn't account for the "vampire" losses nor the charging inefficiencies.

scaesare · Jul 24, 2013

mknox said:
I think you're right. I believe the car's meter only records power consumed (from all sources) when the car is actually on. It therefore doesn't account for the "vampire" losses nor the charging inefficiencies.

So then... if it's "from all sources", I can't figure out why the car projects I have 242 rated miles, yet if I average under 308Wh/mi for 200 miles immediately after a charge, it reports only ~20 miles left?

Where did that extra 22 miles go? I don't think it's that the car is using some value other than 308 for its "rated" projections, as the loss is non-linear.... I lose 15 miles of rated range in the first 1/3rd of the trip.... after that rated range drops only slightly faster than my actual miles driven, but it nonetheless does...

mknox · Jul 24, 2013

I thought Rated was 300 Wh/mi. I believe that is what my car's display shows.

265 miles * 0.3 = 79.5 kWh which is close to the pack's 85 kWh capacity since the EPA rating is supposed to include charging losses.

brianman · Jul 24, 2013

mknox said:
the EPA rating is supposed to include charging losses.

My understanding... This is accounted for in the "how much battery gets" vs. "how much you consume from your power supplier". The capacity of the battery is a different matter entirely.

- - - Updated - - -

mknox said:
I thought Rated was 300 Wh/mi. I believe that is what my car's display shows.
265 miles * 0.3 = 79.5 kWh

There are many ways to come across these rates (rated and ideal). One is the in-car display. One is consumption statistics (people writing notes / journal entries). From the charging "rated miles per hour" shown on Tesla's charging page, etc. Another option is the REST API telemetry.

Let's look at this last one a bit. What do you think the "Ideal" consumption rate is set at?

Some data for you to mull over...
Model S REST API - Page 40

battery_range=231.48
est_battery_range=194.64
ideal_battery_range=266.42
battery_level=90

timestamp,speed,odometer,soc,elevation,est_heading,est_lat,est_lng,power,shift_state,range,est_range
1373590398452,8,11277.6,90,135,263,36.253632,-120.238269,12,D,231,193

Using the first quote numbers and your 300 Wh/mi number...
(231.48 * 300) / 266.42 = 260.65 Wh/mi = "Ideal" consumption

For comparison, if you just use raw numbers with no padding you get...
Rated: 85 kWh / 265 mi = 320.75 Wh/mi
Ideal: 85 kWh / 300 mi = 283.3 Wh/mi

mknox · Jul 24, 2013

For me, I just use my Energy Monitor's number against the odometer in the car and come up with my "real" Wh/mi. I compile this once per month. The only "flaw" is that I don't account for "free" power I get when charging at free public stations, so this makes my number artificially low. Interestingly though, my calculated number is still usually higher than what the car reports (I re-set the trip odometer monthly) even with the "free" power.

schueppert · Jul 24, 2013

Interesting discussion. One possible explanation for the disconnect between the range display and the trip meter is that they are based on different measurements, not just different accounting. I suspect the trip meter calculates power consumption by a continuous measurement of the current and voltage delivered by the battery. In contrast, I suspect the range display is based on an instantaneous measurement of the battery voltage only, which declines along a predicable curve as state of charge decreases.

boilerbots · Jul 24, 2013

I have never seen this "spike" in power when I first start to drive. I looked on the trip summary and the power graph. I even checked again today because my firmware updated last night. The trip summary often starts with an estimated "-0.0" kWh used but just slowly increases as I drive away.

We have someone with an 85 reporting about 2.5kWh lost per day, my number is lower just under 2kWh per day over a weekend for my 40 (software limited 60). The vampire is really sucking a lot of juice if you add up all the Teslas on the road.

Now that I have something to look for I will dig a little deaper, perhaps the rest API can reveal more information.

scaesare · Jul 24, 2013

boilerbots said:
I have never seen this "spike" in power when I first start to drive. I looked on the trip summary and the power graph. I even checked again today because my firmware updated last night. The trip summary often starts with an estimated "-0.0" kWh used but just slowly increases as I drive away.

I'll see if I can capture a quick video of it next time I drive after a charge...

jerry33 · Jul 24, 2013

boilerbots said:
I have never seen this "spike" in power when I first start to drive.

Sometimes I do and sometimes I don't. It seems to depend on variables like "did the car just finish charging", climate control on, ambient temperature, etc.

Energy accounting

Well-Known Member

8Dec2012 / Leeroy Jenkins

Well-Known Member

8Dec2012 / Leeroy Jenkins

Well-Known Member

Well-Known Member

Member

Well-Known Member

Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Burrito Founder

Well-Known Member

Member

Member

Well-Known Member

(S85-3/2/13 traded in) X LR: F2611##-3/27/20

Similar threads