Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Cold Weather Impact on EV range

This site may earn commission on affiliate links.
I am interested in the impact of cold weather on EV range of the Model S and the reasons why cold weather has such an impact on EV range. The range calculator at Tesla's web site seems to be extremely optimistic--I don't think I trust it. I do not have a model S, so I am providing data I observe for my car (Ford Fusion Energi).

Below, I have plotted kWh/mile I have observed for my 8 mile EV commute to work vs. temperature. I do not use heat or air conditioning for the vast majority of the trips. During the winter, I precondition the car before I leave so it is warm when I leave. The commute is only 12 minutes so it doesn't cool down all that much when it is below 0 F outside. There is no heater for the battery in the car, so the variation in kWh/mile is solely due to temperature, weather, wind, and traffic (and not the use of heating or a/c). It is a city commute with posted speed limits of 55 mph.

As you can see, when the temperature is 80 F, the kWh/mile is approximately 0.236 kWh/mile. When it is 0 F, it is approximately 0.320 kWh/mile. It requires 1.36 times more energy for my commute at 0 F vs. 80 F due to the colder temperature. This is far more than I would expect due to greater air density alone (1.15 times greater at 0 F vs. 70 F). There must be other factors coming into play than denser air.

I suspect that there is a lot of internal friction in the car at cold temperatures, i.e. wheel bearings, transmission fluid, gears, etc. When I first start out in the winter, it takes more than three times the power initially to propel the car than it does in the summer. Since I am travelling at low speeds, most of this must be due to this much greater friction in the winter.

Has anyone done a similar analysis for their commute to work in the Tesla, i.e. kWh/mile vs. temperature. Ideally, it would be without the affects of heating and a/c or the battery heater.



Temperature vs. MIleage.png
 
Did you keep the tires the same between summer and winter?

This will likely be the largest influencer in your case - especially if you didn't reinflate them in the winter. But even if you did, the rolling resistance will be more.
 
Did you keep the tires the same between summer and winter?

This will likely be the largest influencer in your case - especially if you didn't reinflate them in the winter. But even if you did, the rolling resistance will be more.

I used the same tires all year round and maintained tire pressure around 40 psi throughout the year. The tires are Michelin Energy Saver A/S.

---

To estimate the amount of energy used by the heater (not included in the plot above), the heater consumes a max of 5 kW of power. At 0 F, it would consume that amount for the entire 12 min trip. So that is a total of 1 kWh for the 8 mile commute or an additional 0.125 kWh/m. For a/c, it will consume 5 kW of power (more than the central a/c for my house) for the first 4 minutes and then about 0.6 kW thereafter. That is an additional 0.41 kWh for 8 miles, or 0.05 kWh/m.

Accessories, including headlights, consume about 530 watts and are included in the plot. They account for 0.015 kWh/m.
 
Last edited:
IMHO, the vast majority of the increase results from the heating of the battery. The main evidence for this is that turning off the heat (for the cabin) makes a big difference in consumption. If we could turn off the heat to the battery (not a good idea to do routinely) I believe the difference would hardly be noticeable. Why do you claim "there is no heater for the battery in the car." There certainly is!

Are you able to prewarm your car for your return trip? Is that part of the trip included in your data? If you don't prewarm, my experience is that a huge amount of energy comes from the battery initially (or from the wall if you do prewarm). Even when you do prewarm, the car must still be maintained at a livable temperature of the battery, if not yourself, and the battery is out in the cold compared with the passengers, so it is significant.

The air conditioning load is minor compared with heating because the Model S uses a heat pump for that, but amazingly not for heating, so all heat is wastefully generated using resistance strips, one for the battery and one for the cabin. Tesla should make this easy fix, since knowing that your range is strongly temperature dependent, to an unknown degree, is the biggest single cause of range anxiety. Data such as you show here is a significant aid to reducing uncertainty, and hence anxiety, so thanks! However, keep in mind that you are not allowed to interfere with the management of the battery in a safe and healthy temperature range.
 
Last edited:
IMHO, the vast majority of the increase results from the heating of the battery and the cabin. The main evidence for this is that turning off the heat (for the cabin) makes a big difference in consumption. If we could turn off the heat to the battery (not a good idea to do routinely) I believe the difference would hardly be noticeable.

Are you able to prewarm your car for your return trip? Is that part of the trip included in your data? If you don't prewarm, my experience is that a huge amount of energy comes from the battery initially (or from the wall if you do prewarm). Even when you do prewarm, the car must still be maintained at a livable temperature for you and the battery, and the battery is out in the cold compared with the passengers, so it is significant.

The air conditioning load is minor compared with heating because the Model S uses a heat pump for that, but amazingly not for heating, so all heat is wastefully generated using resistance strips, one for the battery and one for the cabin. Tesla should make this easy fix, since knowing that your range is strongly temperature dependent, to an unknown degree, is the biggest single cause of range anxiety. Data such as you show here is a significant aid to reducing uncertainty, and hence anxiety, so thanks!

When the coolant temperature falls below 15 F, the ICE will turn on. I use the engine block heater at home to prevent that. But I can't do that for the return trip. So when it is below about 10 F, the ICE turns on for the return trip. I don't include trips when that happens. Otherwise, I do not preheat or use the heater for the return trip. I park in the sun. That does a good job of warming the cabin when it is above 10 F. Its not a long commute.

Note that there is no heater for my battery. I do not have a Tesla. When it is cold, the battery temperature falls. I have observed it to get down to -7 F.
 
When the coolant temperature falls below 15 F, the ICE will turn on. I use the engine block heater at home to prevent that. But I can't do that for the return trip. So when it is below about 10 F, the ICE turns on for the return trip. I don't include trips when that happens. Otherwise, I do not preheat or use the heater for the return trip. I park in the sun. That does a good job of warming the cabin when it is above 10 F. Its not a long commute.

Note that there is no heater for my battery. I do not have a Tesla. When it is cold, the battery temperature falls. I have observed it to get down to -7 F.

Ok, clearly you are talking hybrid, while I was assuming you were talking Tesla here on this forum, though you did say "EV". A 36% increase without any heating seems unaccountable to me. Once warmed up with heat off, our Tesla consumption falls to the normal range (<=200 Wh/km) even in cold weather, unless there is snow or freezing rain. I find it hard to believe that rolling or air resistance can account for that much increase. Could it be tires? Tesla 19" Michelin Primacy inflation is 45 psi.
 
I think that you will find for any car, including EVs, significantly more power is required to overcome internal frictions when you first start out when its cold. The wheel bearings are much more stiff and hard to turn. The transmission fluid is much more viscous. It will probably require three times the normal power to propel the car the first mile, over any additional energy required for a battery heater. Since my commute is only 8 miles, I suspect these components don't really get much of a chance to warm up. For a longer commute, the mileage would improve
 
I now have an explanation why the kWh/mile is so much greater in the winter vs. the summer. From the plot in the OP, kWh/mile is 0.225 at 75 F and 0.31 at 0 F, i.e. 1.38 times greater at 0 F vs. 75 F. The main reason for this increase is that cold weather significantly impacts the effectiveness of regenerative braking. About 13% of the increase is due to greater air density, increased tire rolling resistance, and increased internal friction at colder temperatures. The remaining 38% - 13% = 25% is due to the impact of cold temperatures on regenerative braking.

The following plot shows how regen is affected by temperature. The percentage regen is the amount of energy recaptured via regenerative braking divided by the total energy used to propel the car. Regenerative braking is far less effective at colder temperatures. Much of the kinetic energy that would otherwise be available for regen is lost due to increased friction from higher air density, increased tire rolling resistance, and increased internal frictions, among other factors. In addition, if the battery is cold, then it can't accept as much regen as when it is warm. Any regenerative braking above the maximum charge limit of the battery is lost and cannot be stored in the battery.

During my city commute to and from work with many stop lights and a 55 mph speed limit, regenerative braking is critical to attaining good mileage. In the winter, the effectiveness of regenerative braking is cut in half. The very feature that makes EVs so efficient in the first place is rendered significantly less effective by cold weather.


Percent Regen vs. Temperature.png
 
How are you gathering this data from your vehicle? (Ford Fusion Energi per OP)

The car sends all trip data to the MyFord Mobile web site which provides a summary of data for each trip, including:

date and time of the trip,
plug-in energy consumed (kWh) from the battery,
MPGe (which is converted to kWh/mi by 33.705/MPGe),
trip distance,
EV miles for the trip,
brake score (indicates if you used the friction brakes vs. regenerative braking),
driving score (indicates energy efficiency of your driving),
regen miles (regen miles is the number of miles driven using energy recaptured from regenerative braking)

Tesla should provide a web site where you can access information about each of your trips. You could probably capture this data for the Model S from the REST API.

For more detailed information for each trip, I use an OBD II scanner and Torque Pro running on a Android Tablet PC. This allows me to record torques, rpms, power, and temperature, and hundreds of other quantities during the trip.
 
After more analysis, it appears that approximately 25% out of the 36% total increase in kWh/mile consumed at 0 F vs. 70 F is due to increased internal friction. The transmission fluid is more viscous, the gears and bearings are harder to turn, etc. When the transmission fluid is at 0 F, it takes an additional 2 kW or more of power to propel the car vs. when it is over 100 F. A pure EV would experience a similar phenomena, but it is unclear just how much of impact it would be on mileage. Since an EV is so efficient, it takes a very long time to warm up the transmission and other drive train components.
 
After more analysis, it appears that approximately 25% out of the 36% total increase in kWh/mile consumed at 0 F vs. 70 F is due to increased internal friction. The transmission fluid is more viscous, the gears and bearings are harder to turn, etc. When the transmission fluid is at 0 F, it takes an additional 2 kW or more of power to propel the car vs. when it is over 100 F. A pure EV would experience a similar phenomena, but it is unclear just how much of impact it would be on mileage. Since an EV is so efficient, it takes a very long time to warm up the transmission and other drive train components.

I don't see why it would take any longer to warm up the transmission/reduction gear/differential, since exactly the same amount of power has to go through it to keep the car moving at the same speed. If anything, it would warm up quicker, because the car is heavier.
 
I don't see why it would take any longer to warm up the transmission/reduction gear/differential, since exactly the same amount of power has to go through it to keep the car moving at the same speed. If anything, it would warm up quicker, because the car is heavier.

As an example, for a commute in January when the outside temperature is -10 F, the transmission temperature starts out at 25 F (the temperature in the garage). After the 8 mile commute, the transmission warms up to 50 F. At 25 F, the additional power required to overcome cold internal drivetrain friction is more than 2 kW. At 50 F, it still requires about 1.5 kW of additional power. For the 15 minute commute, that is about an additional 2 kW * 15 min / 60 min/hr = 0.5 kWh of energy (approximately) to overcome the additional internal friction of a cold transmission.

For a commute in June when the outside temperature is 65 F, the transmission temperature starts out at 65 F and rises to 90 F. At 65 F, the additional power required to overcome cold drivetrain friction is about 0.8 kW of power. At 90F, it is about 0.2 kW. For the 15 minute commute, that is about an additional 0.4 kW * 15 min / 60 min/hr = 0.1 kWh (approximately).

The entire commute takes about 1.7 kWh of power in the summer (1.6 kWh without cold drivetrain losses). In the winter, 0.5 / 1.6 = 31% of the energy is used to overcome cold drivetrain friction. In the summer, 0.1 / 1.6 = 6% of the energy is used to overcome cold drivetrain friction. The transmission has to warm up to well over 100 F to minimize cold drivetrain power losses. In a PHEV, you can start the ICE to quickly warm up the transmission if you have a long commute that cannot be driven entirely in EV mode. But my commute is short and can be done entirely in EV mode, so I don’t use the ICE to warm up the transmission.
 
On Saturday morning, I measured the electric power consumed by the electric motor vs. transmission fluid temperature while driving over the same section of level road several times in EV mode with cruise control set to 30 mph. The outside temperature was 30 F and there was no wind. I did not use climate control. The power measurements exclude power used for accessories, so it is exclusively power consumed by the motor. To warm up the transmission faster, I used the ICE. Otherwise it would take many miles to warm up the transmission fluid temperature over 100 F. The following plot shows the results. When cold, it required about 6 kW of power to go 30 mph. When the transmission was warmed up it took 4 kW of power. A cold transmission has significantly more internal friction than a warm transmission and will consume significantly more energy. It would be interesting to see similar results for a pure EV.

30 mph power vs tft.png
 
I have been tracking the Energy Capacity and Temperature of the battery each morning for a year. The following plot shows how the energy capacity of the battery varies with temperature. The capacity of the high voltage battery (HVB) is 7.2 kWh. Capacity falls by about 8% as the HVB temperature falls to 10 F. Below 10 F, capacity falls rapidly. Since I park in a garage, the HVB temperature rarely falls below 30 F. So a cold battery has much less impact on EV range than a cold transmission for my car.

I'm not sure how accurately the Battery Management System (BMS) computes energy capacity of the battery. The upper markers in the plot are from this fall when temperatures were falling. The lower markers are from this spring when temperatures are rising. I suspect it takes a while for the BMS to adapt to changes in the weather.

HVB Energy vs Temp.png
 
The chart from the link below shows the viscosity of transmission fluid vs. temperature. At 0 C (32 F), the viscosity is many times greater than what it is at 50 C (120 F). I would expect the transmission fluid used in EVs behaves similarly. This explains the significantly reduced mileage that I experience in sub-zero temperatures.


http://www.viscopedi...sion-fluid-atf/
 

Attachments

  • Transmission Fluid Viscosity.png
    Transmission Fluid Viscosity.png
    77.4 KB · Views: 216
Last edited:
The following chart shows the viscosity of transmission fluid vs. temperature. At 0 C (32 F), the viscosity is many times greater than what it is at 50 C (120 F). I would expect the transmission fluid used in EVs behaves similarly. This explains the significantly reduced mileage that I experience in sub-zero temperatures.
Some Prius owners preheat the power split device to reduce the fluid friction when it's cold. It would be a significant boon if you could preheat the reduction gears in the Model S.