Battery Charge Loss: Overnight

Battery Charge Loss: Overnight

Does anyone else notice that their Model S loses charge over night? I went to Tesla Menlo Park yesterday and told them this, but hey said that the charge loss would replenish itself once I started driving. They explained that this was due to the cold temperature at night, which caused a misreading in the morning. My takeaway from the conversation was that the battery under cold temperatures acts similar to that of a car driving up or down a hill: the "gas tank" misreads the level of fuel in the tank.

So I took my car home last, checked the mileage, went to breakfast this morning, and not only did the car register 30% less charge but it did not replenish. It also lost 2 miles of charge for a 1 mile trip. I drove normally. I checked my phone and it had been logged in to when I checked the mileage last night before I went to sleep: 61 miles
It changed to 38. It was at 40 before I drove to breakfast.

I believe it lost 30% the previous night. 102 went to 70 something.

I just picked up my car from the factory on Thursday 2-14-13, so I'm a little new to this. Does anyone else have this problem?


jat | 17 février 2013

Without sleep mode (which was added in 4.0 but removed in 4.3), the computers in the car use up about 8mi/day regardless. Also, if it is cold, the car will use some energy to keep the battery warm enough to avoid damage. This energy used doesn't come back (though if you keep the car plugged in, it will periodically charge to replace the power used).

Aside from that, there is the issue you discussed, which is that the computer calculates the range based on how much energy it can extract from the battery at the current temperature, not taking into account that the battery will warm up with use. That is the only "charge" that comes back, and it simply reflects the calculation reflecting the warming battery temperature.

One other thing is that if you make short trips, you will use much more energy than typical, because you will be running the heater full blast for the entire trip to warm the cabin, and then by the time you get back in the car it will be cold again. As an example, I typically average around 370Wh/mi the way I drive, but when I was out of town and making lots of short trips in below-freezing weather, I was regularly using over 900Wh/mi on those short trips.

You really should keep the car plugged in when you are at home.

David Trushin | 17 février 2013

I have never seen my rated miles actually go up. On the other hand, I have never seen it decrease by 30% when out in the cold and not charging. Usally it is about 1 mile per hour, maybe 1.5. I have seen the projected miles higher than the rated miles. Also, there appears to be a characteristic cold spike in the energy graph that occurs when you drive the car cold. This goes away after a period determined by how cold it is, how cold the car is, and how fast you are driving. jat is right that the worst driving is short trips followed by long periods in the cold. I drive about 35 miles to work and I find that the Wh/m starts over 400, but by the end of the trip is in the 320-340 range. I second that you should really keep the car plugged in and I have tried a few times to warm the car before unplugging. the best result I got was when the charging kicked in and warmed up the battery prior to leaving. You can actually force this by shifting to range mode and then back after it starts charging. I don't think that that hurts the battery.

noel.smyth | 17 février 2013

I received my car about a month ago and its been pretty cold here in the northeast. I would say most nights in the 20s or low 30s and several in the teens. I see about a 1 M loss per hour in the cold and another tax of about 10 M to warm the battery when initial driving. You also lose all or some of the regen capability when warming the battery. What I have done to combat this is I set the charging amperage such that it completes charging about when I am leaving in the morning, i get the full range available (not that I really need it) and more importantly I dont have to warm the battery and have regen from the start. this has worked very well for me as a stop gap until they deliver a software update to allow scheduling of the charging.

CC | 17 février 2013


What you are experiencing is about the same as the guys that just did a trip to Groton CT to replicate NYT's debated test drive. After super charging (MaxRange charge) at Mildford CT and arrived at Groton (55 miles). They left the cars outside of a hotel at 22 degree F overnight WITHOUT plugging.

Car 1:On arrival 196 mi rated remaining -> after over night 169 mi
Car 2:On arrival 198 mi rated remaining -> after over night 171 mi

Loss of 27 mi each car, which is about the same as yours. Then they drove highway around 65 to 70 mph with heater at 70 to 72 degrees, which used more energy compare to traveling ideal 55 mph with no heat on. So from Groton CT to Tesla super charger station:

Car 1 drove 81 mi with 73 mi remaining = 154 mi
Car 2 drove 67 mi with 74 mi remaining = 141 mi

So the rated miles in the morning before they left is pretty close to the final tally, depending on speed, heater use.

Another example, Consumer reports reported 140 mi rated remaining, after overnight, 65 mi (probably even colder temp over night) rated remaining, then he drove locally for 30 miles, and the car states 55 mi remaining, which means the car can go total of 85 miles, that is 20 mi more than the rated 65 mi in the morning.

Anyhow, the car will use power to keep the computer running and keeping optimal condition for the battery. I rather the car use the power to extend the battery life than not. The low outside temp also plays a role, and after you start driving, the battery heats up plus the heat from the electric motor is diverted to heating the battery, the range will improve a little, which in consumer reports' case, it replenished about 20 mi.

Bottom line, plug in the car whenever you can, especially in the winter. Topping off the battery everyday will also extend it's life.

July10Models | 17 février 2013

I have a 40mi round trip commute into NYC. In subzero degrees Celsius temperatures, my drive to work consume between 400 and 500 Watts/mi. In the evening albeit when the temps have climb above the freezing mark, my car consumes electrons at the same rate but completes the 20mi home leg with no loss in range. All battery chemistries shrink in the cold including Pb acid batteries.

nickjhowe | 17 février 2013

Small correction to @Jat's post to avoid any confusion: sleep was removed in 4.2; 4.3 isn't out yet.

jat | 17 février 2013

@nickjhowe - what I intended to say was removed in 4.2, and should be back in 4.3, thanks for correcting.

nickjhowe | 17 février 2013

Guessed it was! Didn't want the noobs to jump in with all the "I don't have 4.3 yet...." comments. :-)

And no offense meant to the noobs. You're most welcome here.

EVTripPlanner | 19 février 2013

Here is some more data from a few days where my Model S was parked (temp ranged from 40-70F). As you can see, loss averages around 0.5 mph (12 miles per day), but estimated range does go up when temperature warms.

Date/Time EPA Range Avg Loss Rate (mph) Incr Loss Rate (mph)
2/15/2013 15:45 121
2/15/2013 21:59 118 0.48 0.48
2/16/2013 12:41 100 1.00 1.22
2/16/2013 16:26 106 0.61 -1.60 (WARMING)
2/17/2013 21:28 97 0.45 0.31
2/18/2013 6:45 81 0.63 1.72 (OVERNIGHT COOLING)
2/18/2013 8:42 77 0.68 2.05
2/18/2013 14:47 87 0.48 -1.64 (WARMING)

I'm an engineer and am able to estimate range/conditions/etc (and enjoy it!), but to be really mainstream, I think Tesla can/should do more to help drivers manage their driving - they have not been really clear/open about parking- and temperature-related losses and the on-board tools and the app don't really help you estimate where you really stand. For example, they could integrate Nav, Weather and Energy apps and really give you a trip planner that takes a lot more into account. Right now, for example, I have to consider speed, route (climbs) and real-world consumption to plan a trip. For example, while you might have enough total energy to get home based on average consumption, if there is a big climb shortly before you get there you can still run out before the top! As more non-tech and non-early-adopter types buy, it will be too easy for them to mis-estimate and get stuck with the current tools.

You can download the reference sheets I use from

EVTripPlanner | 19 février 2013

(formatting on table above got screwed is also in the spreadsheet)

jat | 19 février 2013

As we get more data from the streaming API, you can see exactly how the charge goes down over time, and coupled with ambient temperature readings we could more accurately predict actual loss rates.

Brian H | 19 février 2013

Here it is using the "pre" tag, just about the only way to tabulate in this forum (H2 added for clarity; 'pre' shrinks the font somewhat):

Date/Time EPA Range Avg Loss Incr Loss
Rate (mph) Rate (mph)
2/15/2013 15:45 121
2/15/2013 21:59 118 0.48 0.48
2/16/2013 12:41 100 1.00 1.22
2/16/2013 16:26 106 0.61 -1.60 (WARMING)
2/17/2013 21:28 97 0.45 0.31
2/18/2013 6:45 81 0.63 1.72 (OVERNIGHT COOLING)
2/18/2013 8:42 77 0.68 2.05
2/18/2013 14:47 87 0.48 -1.64 (WARMING)

Brian H | 19 février 2013

Somebody typo'd the "Steaming App". Sounds more dynamic!! ;) But rude ...

c.bussert67 | 19 février 2013

Ran my own little test when I got my car. Garage temps were between 54-59 degrees. Charged up to rated at 241 and unplugged it.
Checked SoC at 6pm daily to monitor losses.
Started at 241,

Monitored battery temps as best I could by measuring bottom of car.
At the point I unplugged, it was at 71 deg. After that it dropped to around 64 degree +- 1 and maintained that temp regardless of garage temp changes.

riceuguy | 19 février 2013

So 12-14 mi/day of loss seems like the vampire load regardless of temp when sleep is not available. Any 60kwh owners measure whether the loss is comparable, and if so is it a comparable % loss or a comparable loss of miles?

jat | 19 février 2013

@riceuguy - it should be more like 8 rated miles/day, not 12-14. | 19 février 2013

I lose about 8-10 miles for 12 hours not charging (overnight).

nickjhowe | 19 février 2013

My car is sat at Miami airport - and will be until Saturday morning. Here's the story so far:

Brian H | 20 février 2013

By advanced mathemagics, I deducts that it deducts 10 mi. day. Like Macbeth, it wishes it could sleep.

jat | 20 février 2013

@BrianH - but would it dream of electric vehicles jumping over a fence?

bsimoes | 20 février 2013

@jat: very funny!
@nickjhowe: thank you for the straight-forward presentation. All the charts and graphs were quite confusing. Your idea was clear and concise.

Sudre_ | 20 février 2013

I have a 60kWh battery. I unplugged this morning around 178 rated miles and drove 6 miles to work on a 19F degree morning. I think I had 171 miles when I got here. My app now shows 161 miles but that's over 3 hours of sitting.

Brian H | 20 février 2013

I dunno, I'll ask Phil when I see him.

jat | 20 février 2013

@Sudre_ - note that is conflating actual charge loss from heating the battery and running the electronics, with a reduced estimate because of the temperature of the battery -- the estimate only considers the battery's current temperature, and not that it will warm up with use.

kristian.skogstroem | 20 février 2013

It should be possible to adjust the min temp for the battery heater. My Leaf's battery only starts battery heating below -22C..

kristian.skogstroem | 20 février 2013

Or, even better: When unplugged do not start battery heater unless temp goes below -22C. And if battery heater starts in this case send me a warning via SMS.

Brian H | 20 février 2013

the general rule for capitalizing unit abbreviations is whether they are names of people. So kilo Watt hours is kWh. mega Hertz is mH. Kelvin and Celcius and Farenheit are K, C, F. And so on. Numbers and time periods are lower case.

HansJ | 20 février 2013

@Brian H - I hate to get off topic but but megahertz is MHz and millihertz is mHz. H is for Henry (unit of inductance) not Hertz (unit of frequency). mH would be millihenry.

Metric prefixes are lower case for 1 000 (kilo) and smaller, and upper case for 1 000 000 (mega) and larger. See

Brian H | 21 février 2013

right you are. Not an SI native, so ...

jat | 21 février 2013

@kristian - I don't know that I would want to adjust the temperature at which battery heating occurs -- I assume Tesla has set it as a balance between preserving the life of the battery, having it usable when you get in the car, and not using too much power to do so.

The only time it should matter is if you are leaving the car unplugged in the cold for a long time, and is there really any reason you would do that rather than hook up a 120V outlet? Even if you park a regular car for long-term storage, they will come out and charge the battery every so often, and they could do the same for a Model S (perhaps a bit more frequently).

nickjhowe | 21 février 2013

Here's the latest...(click for a bigger version)...another day and a half to go. Good thing is I'll only use about 60 miles to get home.

c.bussert67 | 22 février 2013

Here's my update. The post I had earlier was on the old 4.0 software. So just to make it current and fair I left it unplugged again to see my daily losses with the new 4.2 update.
Garage temps are at 57 Deg...
Pulled in with exactly 100 miles to go.
Next night at same time had 87,
Following night, same time was 75.
It seems to me, my losses are the same regardless of software updates so far. I also thought that it might be less if battery levels were lower as it might try to 'survive' at some point and draw less power. I also used the app to check levels this time, where as before I had to open the door to check it and the car would wake up.

nickjhowe | 23 février 2013

Picked the car up this morning. Here's the full week's results:

Miami this week was 70's during the day, and 50's over night. After initially losing 0.6 m/hr in the first 12 hours, it settled down to a (fairly) steady 0.4 m/hr for the rest of the week (horizontal axis is time)

eq1 | 23 février 2013

This seems really, well, bad... I'm no expert, but if my calculations are about right, they mean there's about a 140 watt draw while your car is parked (in whatever state it's in). Here's what I did:

-you lost 189-131=58 miles of range.
-at an average of say 350 watt-hours per mile, that 58 mile loss is equivalent to 58 X 350 = 20,300 watt-hours.
-you lost this amount over about 6 days, which equals 6 x 24=144 hours.
-20,300 watt-hours divided by 144 hours = 141 watts.

Could this be right? - could the car be consuming 141 watts while parked?? If so, that's way too much...

Pungoteague_Dave | 23 février 2013

Wow. TM needs to get the sleep function back in service ASAP.

nickjhowe | 23 février 2013

Yup. TM knows it. They are keeping everything running at full power all the time. As soon as they get sleep mode reimplemented in the firmware the drop should be closer to 25-30 w/hr.

Hans (Amsterdam) | 23 février 2013

It's unfortunate that all the controls have to be on the display, otherwise they could add a physical switch that disconnects the 12V battery. Like any ICE car.

Getting Amped Again | 23 février 2013

@nickjhowe - great information and presentation!

nickjhowe | 23 février 2013

Thanks Got Amped.

@Hans - except the switch doesn't disconnect completely. The battery on an ICE car still powers the alarm, the remote key sensors, the lights,... All the key does is shut down the switched 12V circuits.

The S uses contactors to isolate the battery (and by the sound of it different elements within the battery) and has switched 12V circuits like an ICE car - try using the cigarette lighter when the car isn't on.

To get the best user experience they decided to keep all the computers powered on all the time - I guess they didn't figure that many people would leave the car not plugged in for a long time (bad assumption) or that it would draw quite as much as it does.

Clearly they know how to fix this (it was fixed in 4.0), but didn't quite get enough bugs out.

Hans (Amsterdam) | 24 février 2013


Thanks for explaining.
Indeed I will keep my Model S unplugged for most of the time, since I only can use public chargers. I have had no problems with my Nissan Leaf doing this so I am a little worried with the Model S. Hope they will fix this.

eadolphe | 24 février 2013

Note that getting the sleep mode back functioning may very well disable the Mobile App.

dqb | 24 février 2013

They should implement sleep with a periodic heartbeat, which would allow communication with TM and also the mobile app. The app communicates through a TM system, and there's hardly any info exchange needed. This could be a very low for car->TM, and fully wake the computer only if the mobile app needs to dynamically connect and perform an to turn on climate control. Else stats could be proxies by TM server.

Pungoteague_Dave | 24 février 2013

The mobile app also goes offline whenever the car is out of cell phone range. This is to be expected, but is an issue at a lot of EV charging stations. For example, I left my S at the BWI daily garage this week for a couple days. The EV charging stations at BWI and at most garages are on the lowest level, in this case in the basement where there is no cell service. So I wasn't able to check on my car with the app at all. I suspect this is the case at many garage locations, especially under buildings.

drp | 24 février 2013


Do you recall, once the car got warmed up and was driving, did the picture look similarly or did add some miles? It would make sense to me that, while it is sitting there, it thinks it can't go as far as it really can, again just like a gasoline car driving downhill or up in incline and engage tips one way or the other. Let me know if you noticed anything at all or maybe it's slipped your mind by now.


drp | 24 février 2013


I like how you took the snapshots. This could be really useful to all of us and maybe we should start a new thread with just this type of information. Any thoughts from anyone else?

nickjhowe | 24 février 2013

@drp - when I got to the car in MIA, the ambient was already c.65° so not much range to be gained by warming, it at all.

astrotoy | 25 février 2013

I had my parking sensors installed this past week. Initially we thought it would take two days, so I brought it in on Thursday morning, to pick up Friday afternoon. So I didn't bring my charging cord. However, it took longer and they had to keep the car over the weekend, when they don't work. So they finished this morning - total of 4 24 hour days without a charge. The car was kept in their garage, with typical night time temps in the low 60's. I went from 198 to 135 in the four days, without the car being driven. They did use some power to test the sensors, but I think it was nominal. Anyway, it looks like a loss of 16 miles per 24 hour day. I checked remotely every night and the loss looked pretty linear. I have a standard MS85 pretty well loaded.

Pungoteague_Dave | 26 février 2013

I noticed this morning via the app that my car dropped from 235 to 165 overnight. The car is at the Rockville service center, and I can see that the doors, hatch, and frunk are all open, as is the charge port. It is very cool to know that the car is actively being worked on. They must have driven a bunch of test miles to kill off that many miles since last night.

New function for the app: odometer reading

Brian H | 26 février 2013

And/or current current use! The "teenager surveillance function".

EVTripPlanner | 26 février 2013

The quick-and-dirty estimate I've been using: 0.5 mph loss while parked. Goes up and down with temperature, but I think that evens out when battery warms up after starting.