Charging efficiency

Charging efficiency

I apologize if there is already a thread on this. I am interested in how much electricity is lost in the charging process. If you pump 85 KWH into your battery, I am sure that your electric meter on your house will show that you used more than 85 KWH. My question is how much more. All of Tesla's analysis assumes that there is no loss - i.e. perfect transfer - from the AC current in your house into the DC battery in the Model S. Does anyone know how much that loss is and are there any factors that impact that loss - e.g. temperature, current level... And is there anyway to measure the loss. It seems to me that knowing that is critical to being able to calculate cost per mile.

riceuguy | 11 décembre 2012

I believe it's 92% but I'm sure someone can verify that!

sergiyz | 12 décembre 2012

92% maximum efficiency per the specs page.
From my experience by comparing kWh spent vs what Chargepoint reports as used during a charging session, it's about 80% or less.
Not sure how much it depends on charging current (Chargepoint is 30A) and weather conditions (if the battery is hot, tesla is running a cooling pump during charging).
It would be interesting to gather more data.

Rod and Barbara | 12 décembre 2012

@ Etzbach – Currently there is no way to know how much electric power you pull from the grid during a charge cycle unless you install a dedicated power meter in line with your charging outlet. This is unfortunate since as you correctly state, this information is needed to calculate your cost per mile to drive the Model S. The Roadster provides this info (electric power pulled from grid during charge cycle) so it is frustrating that Tesla did not include this same feature in the Model S. I have contacted the Ownership Experience Team to express how important I think it is to include this feature in a future software update. I encourage other owners to do the same. On another front, I semi-moderate (there are no official moderator in this forum) the Prioritized Software Enhancement List (a private thread for reservation holders/owners only) and this feature is number 11 of 84 suggested software enhancements in that thread.

You can back out the efficiency from the data Tesla provides in the Calculator Section of the Go Electric > Charging page. These data imply 98% efficiency for 110V at 12A, 92% efficiency for 240V at 24A, and 94% efficiency for 240V at 40A/80A. I’m not sure I would put a lot of faith in those numbers. By way of comparison, using actual data collected while charging my Roadster I observed 82% for 240V at 70A and 54% for 110V at 12A. I did try to back out the charging efficiency during one charging cycle in my Model S by observing the voltage and current reported on the instrument panel every 15 minutes during the charging cycle. The voltage and current were pretty steady at 236V and 69A. The calculated efficiency from this rough method was 98%. This seems overly optimistic to me. I know that the charger in the Model S is improved over the Roadster so I expect a charging efficiency close to 90% for 240V charging in the Model S. The charging efficiency sergiyz mentions during a Chargepoint charging session seems low to me, but I am not familiar with Chargepoint chargers.

Etzy | 12 décembre 2012

Rod and Barbara - This is very interesting. Your comment about a dedicated meter gave me an idea. What if you tripped every breaker in the house except the ones that go to the dedicated 240 volt charger. Read the meter, charge for a couple of hours, then read the meter again. Only downside is that you would have to reset the clock on the microwave (and I wouldn't want to do it when my wife was watching Modern Family). Would this work?

Volker.Berlin | 13 décembre 2012

Rod and Barbara, I wonder if it is even possible for the car to tell how much electric power it pulls "from the grid". Where would the car's meter have to be located for that? At or before the outlet, as you mention.

Even if the car would precisely measure the electric power that passes its charge port, it would still miss losses from the charging cable (which does emit substantial heat, based on other remarks in this forum).

nickjhowe | 13 décembre 2012

I have a TED 5000 monitoring my electricity usage. If no-one beats me too it I'll measure the draw once I get my car in early January.

mrspaghetti | 13 décembre 2012


I don't see why that wouldn't work.

jat | 13 décembre 2012

I have a TED5000 monitoring my garage breaker panel, which is 99.9% used to charge the LEAF. I see around 82% efficiency, if the car's measurements are accurate. Some of that could be lost in the charging dock and cables, but it is mostly just a relay and GFCI in the power circuit and not much electronics to run, so I think it is mostly in the rectifier and batteries in the car.

I would guess the Model S would be around the same, though the battery chemistry is different (the LEAF is LiMn) so it might not be.

jat | 13 décembre 2012

@Etzback - that would work, except the meter may not be accurate enough to get a good measurement after only a couple hours. Plus, you are living without power the whole time -- much easier to monitor just that circuit, it isn't very expensive.

space09 | 13 décembre 2012

I guess I should switch over to 110v if I just drove a few miles and/or for maintenance charge?

Brian H | 13 décembre 2012

110V is the least efficient.

Per O | 15 décembre 2012

I got my P85 today!!
And my monitoring is with 1% accuracy power meters

Numbers are ugly
From model S reporting
Drove 63.1 miles consuming 22.9kWh
Charged in the evening for app 2.5hr and car reported 26kWh

But my power meter reported 32.8kWh
This is only 70% efficiency!

I am an engineeer with some knowledge of designing boost DC-DC converters. app95% is realtively easy doable. Then there is additional RI^2 loss in wiring and battery, and the cooling of the battery and finally the chemestry loss. But definitely looks ugly so far

I'll report again Monday night after my first 90 mile work commute
(most of it 65MPH steady highway). Should cost me 27kWh consumed.

Timo | 17 décembre 2012

But my power meter reported 32.8kWh

Is the car only thing using that input? If not then that could explain the huge difference.

Per O | 17 décembre 2012

Yes, i am monitoring that circuit only with 1% accuracy.
But I think I know the problem. The carging was set to "Max Range" instead of "standard".
I'll have an update tonigth or tomoroow after recharging following my first 90mile/30kWh commute

Rod and Barbara | 18 décembre 2012

@ Per O – I have a Roadster and a Model S. I have lots of data on my Roadster, not so much on my Model S. With respect to your charging data reported on 12/15/12, the charging efficiency can be calculated by comparing the kWh added to the Battery (26 kWh) and the kWh drawn from the grid (32.8 kWh). This is 79% efficiency (not 70%), which is significantly lower than I expect from the Model S. With respect to the difference in kWh between what the car reported consuming and what was added to the Battery, there are two things to be aware of. First, the Battery SOC at the end of the each charge cycle (the one before the driving and the one after the driving) has to be the same. Second, the difference between the kWh consumed while driving and the kWh added to the Battery does not represent charging efficiency, it represents the power that was consumed by the car while it was turned off and sitting idle. This certainly affects the overall efficiency of the car, but it is not related to charging. It is related to the fact that you are driving a computer that you can never turn off. After four years with my Roadster my results are as follows: of the energy taken from the grid, 78% is used to propel the car (245 kW/Mile), 5% is used up by the car while it is sitting idle (17 kW/Mile), 16% is lost inside the car during charging (54 kW/Mile) and 1% is lost getting from the grid to the car (3 kW/Mile). So the charging efficiency of my Roadster is 271 kW / 328 kW = 83%.

Rod and Barbara | 18 décembre 2012

@ Volker – Concerning your post on 12/13/12 about whether it is possible for the car to tell how much electric power it pulls “from the grid.” You are right to point out that the car can’t be expected to have that data. What I should have said in my post to Etzbach is the Roadster reports how much electric power is delivered to the charging port on the car. As you can see from my prior post to Per O, the power loss from the grid to the car’s charging port on the Roadster is very small. I used to have an inline meter on my Roadster High Power Charger so I could collect all the pertinent data, but my meter stopped working about a year ago and I have not replaced it yet. Hence I really miss the fact that the Model S does not report the electric power delivered to the charging port.

Volker.Berlin | 19 décembre 2012

Rod and Barbara, thank you for the clarification.

kent | 19 décembre 2012

I charged yesterday at a ChargePoint location. The total drawn was 6.250 Kw and energy into the car showed 5.081 from their information. That gives an efficiency of 81%.

Per O | 19 décembre 2012

@ Rod
I'll update after charging tonight app 250 miles total driven
That should reduce uncertainty regarding the SOC
First charge was after Tesla delivered, so some uncertainty
I now have ver4.0 software and it should be powering everything down at night
And i won't drive the car next 15 days, so that would give a good read on idle drain by Jan 3rd!

Per O | 19 décembre 2012

And BTW, since some losses is R*I^2, we should get higher efficiency by charging more slowly.
In jan, i'll try to reduce from 10kW to 5kW.
In my case, that will still leave plenty of time to charge overnight

jodyk | 22 février 2013

Curious if anyone has an update on this post?

Hans (Amsterdam) | 23 février 2013

EPA data Tesla model S 85 kWh:

range 265 miles
combined fuel economy 38 kWh / 100 miles (from the grid)

effective battery capacity (range charge): 95% x 85 kWh = 80,75 kWh
total kWh from the grid: 265 x 38/100 = 100,7 kWh

charging efficiency: 80,75 / 100,7 = 80,2 %

Charging efficiency will be a little better since the Vampire Load will account for some losses during charging.

Timo | 23 février 2013

You get full 85kWh in range mode. Car just claims zero range when it still has something like 4kWh to go and you have already used 81kWh.

Hans (Amsterdam) | 23 février 2013


It all depends on the rated capacity of the battery. If rated capacity indeed is 85 kWh than it is not possible to use the full 85 kWh because in range mode the charging voltage is limited to 4,15V.

Tesla Motors just states 'Energy Storage 85kWh'. Is this rated of range mode capacity ?

From: A Bit About Batteries By Martin Eberhard:

There is a huge difference in cycle life between a 4.2V/cell charge (defined by the manufacturers as “fully charged”) and a 4.15V/cell charge. 4.15 volts represents a charge of about 95 percent. For this reduction of initial capacity (5 percent), the batteries last a whole lot longer. Unfortunately, further reduction of charge has a much smaller benefit on cycle life. Understanding this tradeoff, Tesla Motors has decided to limit the maximum charge of its cells to 4.15 volts, taking an initial 5 percent range hit to maximize lifetime of the pack. We also limit discharge of our battery pack to 3.0V/cell and will shut down the car when the batteries reach this level.

full article:

EVTripPlanner | 23 février 2013

Agree with all above that efficiency, claimed at max of 92%, is not constant/linear...and it is NOT average for an entire charge cycle regardless so real-world results will always be less. I also don't think 80% is the right number - unless you're always "topping off". Also remember parasitic loads that amount to 12-14 miles/day (about 150 Wh/h, which is about 1.5% efficiency hit NOT including battery conditioning that is going on). If you are plugged in for 12 hours but only charging for 4 hours, then the parasitic is more like 5% hit to overall efficiency. I have some spreadsheets with various related calculations/tables at if you want to play around with charge times, etc. I need to update for time- and temp-based parasitic loads.

jat | 23 février 2013

It's still early to account for all the variables like battery heaters coming on, vampire load, etc, but so far I am seeing about 86% efficiency (ie, 86% of the energy that comes from the wall makes it into the battery) charging at 40A/240VAC. Some of that is lost in the Mobile Connector, other in the charger in the car, and the rest in the battery itself due to internal resistance (which gets higher the faster you charge, since power dissipation goes up with the square of the current).

Alex K | 23 février 2013

@Timo | FEBRUARY 23, 2013: You get full 85kWh in range mode. Car just claims zero range when it still has something like 4kWh to go and you have already used 81kWh.

That's what I'm seeing empirically. After charging in range mode (270 available miles), the last trip I took was 219mi using 65.7kWh with remaining range of 50 miles, or about 15.7kWh remaining. So total battery usable capacity was about 81.4KwH.

nickjhowe | 23 février 2013

The REST interface that folks are starting to program to includes what appears to be a SOC number.

I just dumped this from my car:

[battery_range] => 230.49
[est_battery_range] => 219.5
[ideal_battery_range] => 265.28
[battery_level] => 87

When you do the math, 230/265 =.87

There is nothing in there about actual KWh.

Alex K | 23 février 2013

@nickjhowe | FEBRUARY 23, 2013: There is nothing in there about actual KWh.

Yea, that's too bad. I was hoping to just start logging the streaming location information and occasionally get battery state information and correlate the two. I was planning on just running this while I was out driving the car. I would think the battery level is directly related to the number of kWhs remaining in the battery, so could be used to determine the kWh used during various segments of a trip.

Timo | 24 février 2013

@Hans (Amsterdam); I believe car battery is actually closer to 90kWh, but you can use only 85kWh, and it shows you zero at 81kWh. At 85kWh point car shuts everything down to prevent battery damage, but it does not "brick" at that point yet. This is only a guess of course, but it is consistent to what I have seen reported here: car still works when range shows zero, Tesla says that car can still recover from "zero charge" quite a bit after it reached zero.

dortor | 18 avril 2013

charging efficiency is very very low - and I'm wondering if I have a problem.

I've been suspecting this for a while - but I finally got all the pieces together in one place yesterday and now I know:

I used 12.5 kWh yesterday in my daily driving - and then proceeded to dump 22.51 kwh into the car to charge it - that approaching only 50% charging efficiency and dramatically alters the cost of driving this car.

Does anyone know if this is expected?

jemartin | 18 avril 2013

50% is low. I've been using an Aeon Labs ZWave Smart Energy Monitor on my 14-50 feed for the past 50 days or 2,300 miles. My charging efficiency (kWh driven / kWh charged) is 71.3%.

I estimated that about 14% of my total kWh charged is attributable to the energy lost while the car was parked. If I subtract this 14% vampire loss from the total kWh charged, I get a 83% charging efficiency (kWh driven / (kWh charged - kWh lost while idle))

HzSoGood | 18 avril 2013

We have had our car for 18 days. We have logged 1070 miles. The car has always been charged at home on a dedicated meter.
Total kWh used per meter is 462
Total kWh used by car trip B (never been reset): 340
Charged up every night.

74% charging efficiency (including vampire loss at night)

EVTripPlanner | 18 avril 2013

I just updated my Electricity Cost calculations to account for charger efficiency AND "vampire load" (draw while parked), and once you look at all that, the overall efficiency (driving energy / wall energy) is like 70% (even if the average charger efficiency is 84%). That means that if your dashboard is telling you that you are using 345 Wh/mile, the real from-the-wall usage is more like 490 Wh/mile. If vampire load was reduced to about 10% of what it is now, these numbers would be more like 82% and 420 Wh/mi. Anyway, you can play with the spreadsheet by downloading from (or just see PDF snapshot at

See more on actual efficiency in this thread -

tesla | 19 octobre 2015

80% Efficient charging? That is leaving a lot of heat on the table. By my calculations that would be around 1900W of heat dissipation while charging (depending on what charging phase) using the 10kw connector, and 3800W of heat dissipation while charging with the 20kw charger. That is like running a space heater on high for 8 hours in your garage, or two space heaters on high for 4 hours in your garage. At least my garage will stay warmer in the winter?

sule | 19 octobre 2015

Well, not all of it is (straight) heat and not all of it is in the car. At max amps about 0.5 kW is lost in my house between the meter and the car. Some power goes to the computers - seems around 150-200W. That is at the peak charging rate. Extra inefficiency comes from the battery itself, the fact that there is consumption even when not driving and less efficient charging at some voltages and near 100% state of charge.

sonja.gallego | 21 juin 2019

Charging efficiency of about 80%, at least in practice, tried to measure - it turned out something like this. Also, the efficiency depends on the settings set - page 94, but values ​​will still be about the same +/-.

dlinkeg | 20 février 2020

Charging yesterday, Feb. 19, 2020...12.61kW listed by ChargePoint EVSE, but only 12.15kW reported by Stats app, giving me a 96% charging efficiency. This charging session, with Sentry Mode running for 2 hours, 14 minutes, was in the evening with ambient temperature around 50 deg. F. This EVSE is usually 210-220 V and 30 amps for ~6kW charging.