"Liquid-cooled" battery

"Liquid-cooled" battery

I've seen multiple references to the Liquid-cooled battery pack of the Model S, but then every mention I can find pretty much blows straight through to the next topic, as if it's too obvious even to mention.

Dave-LasVegas | 5 juli 2012

I'm an electrical engineer, so maybe I'm the only one curious about this. However, I also happen to live in a climate where a car's "cold" temperature can start out annoyingly hot, perhaps over 50°C/122°F on the very worst afternoons in summer; over 38°C/100°F in the coolest (!) part of the night. It's not a coincidence that roads around here have names like "Sahara."

Can anyone here add a few words on the subject? I assume since the key word is liquid that we're not talking about water. In fact, I'm pretty sure you want to keep water away from Li-ion batteries (corrosion issues). What is the mysterious coolant? How much of it is present? Is there a hazmat issue? Is there a pump and radiator involved? Does it degrade? How does its performance over time compare to the overall expected lifespan of the battery?

Thanks in advance to anybody who can shed a little light here. I'm looking forward to my Model S.

-- Dave (P1736)

ggr | 5 juli 2012

I can't (yet) comment on the Model S battery, but the Roadster battery is also liquid cooled. I believe the cells are in small sealed compartments, with standard anti-freeze coolant circulating around them. There is a small reservoir and pump. The main purpose is to keep the cells at an EVEN temperature, to prevent bad cells and/or thermal runaway. If the temperature goes too high, the air conditioning unit switches on too, and the coolant goes through a heat exchanger to actively cool the cells. Mostly I've heard this happen during charging, only a few times when driving.

VincentA | 5 juli 2012

There are Roadsters that "live" in very hot climates (Arizona, for example), and have worked fine. This does put somewhat more stress on the car, and will somewhat reduce efficiency and range, but Tesla has gone out of their way to ensure that the Li-ion cells are carefully temperature-controlled in even extreme weather and climates (arctic cold also!).

jerry3 | 5 juli 2012

- Is there a pump and radiator involved?

Yes, the model S has a heat exchanger. This video is old, but it shows the three heat exchangers.

- Does it degrade?

It's just automotive anti-freeze so presumably you have to change it occasionally as the anti-corrosion additives get used up.

Sudre_ | 5 juli 2012

With any luck they will bring a car (or 10) to a hot climate for test driving. If they make it thru a full day of supercharging and continuous driving I think your worries should be laid to rest. You will never torture your car more than what I've seen everyone doing on the videos. I wonder if Elon cringes when he see the videos of people driving his baby?

jerry3 | 5 juli 2012

There doesn't appear to be any hot weather problems with the Roadster and the Model S' system is more sophisticated, so I wouldn't expect any problems with it either.

bfranks273 | 5 juli 2012

And isn't the motor also liquid cooled in the Model S? Wonder how that works. One fluid or two separate ones to manage two different temps?

Brian H | 5 juli 2012

As I posted in General, Nissan seems to have been bitten in the asp by lack of Leaf battery temp control:
Siry chimed in, referring back to his Jan. '10 prediction of same:

Nissan seems to be in denial mode at the moment. Danger!

Theresa | 6 juli 2012

I own a Roadster 1.5 and I live in the Midwest. It not only gets hot here it gets humid. So far I have experienced a few days of power limiting if I drive hard enough (which doesn't take much to do in the heat) to push the battery and/or PMC to the yellow range on the temperature gauge. The last few days here have been in the 110 heat index range (99 actual temp) and when I get in the car the fluid is circulating and sometimes the fans are also running. When the humidity is high is when I have trouble getting the temps to fall on the PMC. The battery doesn't seem to have nearly as much trouble cooling as the PMC does. When the humidity is low the car never power limits.

The fluid is standard antifreeze.

Timo | 6 juli 2012

@ Theresa, do you mean PEM with PMC? Power Electronics Module?

That's also liquid-cooled in Model S. Pretty much entire drivetrain is.

Theresa | 6 juli 2012

Yes I meant PEM. That's the problem with being an engineer. Too many acronyms!

I am excited that the entire drive train is liquid cooled as you can tell from my post that it is a problem here to only be air cooled.

Larry Chanin | 8 juli 2012

And isn't the motor also liquid cooled in the Model S? Wonder how that works. One fluid or two separate ones to manage two different temps?

I'm not positive, but I believe that there may be three cooling loops. There are three small heat exchangers. One for the batteries, one for the passenger cabin and one to cool the motor/inverter combination.


Teoatawki | 8 juli 2012

Yes, Larry. You are correct. 3 systems to condition cabin, battery pack and motor/PEM combo. Each of the three have very different requirements. The cabin conditioning system is attempting to maintain a very narrow range specified by the occupants. The motor/PEM system is only concerned with preventing overheating. The battery pack system must make sure the battery pack is both warm enough to function efficiently and cool enough to prevent damage.

Larry Chanin | 9 juli 2012

Thanks. Let me apologize in advance for going slightly off-topic.

I was told by a Tesla rep at a private factory tour that technically the Model S doesn't have a PEM like a Roadster. There's the part of the powertrain consisting of the motor, gear box and inverter that is liquid cooled. Then the chargers are no longer part of this assemby, they are located separately under the rear passenger seating. So what was called the PEM on a Roadster is just called the inverter on the Model S.

I'm going to try to attach a photo I took while on the tour.

The cooling inlet and outlets are located on the right. One of them can be seen as an orange cap.


Brian H | 9 juli 2012

A bit bigger than the 130-lb watermelon described for the Roadster! Don't think I'd like to lift that baby up -- or even just the motor half of it.

Brian H | 9 juli 2012

Just flashed a Stupid Human Trick with Letterman attached to the rotor/axle and spinning so fast he blurred ...

EdG | 9 juli 2012

The photo makes it look quite a bit bigger than it appears in person. If you have a local store with a "skateboard" on display, you can check it out yourself. Here's a view more in line with how I remember it:

Timo | 9 juli 2012

EdG +1, I think that's perspective illusion caused by camera zoom, it just looks like it is further away from the camera than it really is. I have been victim of the camera zoom distortion once. Because I'm rather big individual I took a place in a corner of a group picture and tried to look as small as I could. Big mistake. Camera distorted the edges of the picture so that I look absolutely gigantic, like a giant in a fantasy film. It looks like I could hold the two smallest woman in the middle of the picture on a palm of my hand at the same time.

That said it is quite a bit bigger than Roadster motor. It is the whole package, PEM, "gearbox", motor all cased for liquid coolant. the axle is actually not going in the middle of that package but behind it and the person leaning over is not looking at the foreground cylindrical objects but the background one that is nearly invisible from this angle.

I wish we had a decent 3D screens by now. It would be so much easier to look around Model S designs with one of those.

Larry Chanin | 9 juli 2012

Hi Guys,

Interesting theories. ;-)

Here's another view from the side.

Sure it looks smaller in the context of the entire "skateboard" when the photo is taken at a distance, but up close it really is rather large.


Timo | 9 juli 2012

Second picture gives better context of comparison. I think I know why it looks that big in that first picture: the system is rather high on a pedestal, but in first picture it looks like it is standing onto ground (in conscious mind you can see it in first picture too, but like in most visual illusions, even when you know the trick your eyes still claim otherwise).

Larry Chanin | 9 juli 2012

However, I do admit that my first photo did make it look larger.


EdG | 9 juli 2012

That's about right. It's big, but not big enough to fit an extra child-sized passenger if it were emptied; that's what the first photo seems to imply.

If memory serves (it usually doesn't when it comes to this) I'd guess it to be about 18" diameter by about 42" wide, or about 45 cm by 100 cm for those who don't speak English system.

I'm sure others have access to actual measurements.

Vuckoko | 28 augustus 2012

Hello Everyone!

I don't think having read a comment about that, so let me share my thoughts: how is it possible to maintain the battery temperature of the roadster (for instance) below 35°C, when it's hotter outside? I imagine that the roadster can still work in Florida in summer?

For those who know a little bit about physics, It would be contrary to thermodynamic laws. From my point of view, the only way to achieve a cold source temperature cooler than the atmosphere would be to use some sort of refrigeration cycle.

So how does the cooling circuit look like? A water-glycol liquid pumped system inside the battery connected through a liquid-to-liquid heat exchanger to a vapor compression cycle?

If we take an example, the batteries being at 35°C, the coolant should be at less than 20-25°C (I suppose) taking into account the thermal resistance of the battery packs.
- If we suppose that there is no secondary circuit, the outside temperature should be at least 5-10°C colder to permit an efficient air forced convection at the front of the car. It is hard to believe that the cooling system would not be able to maintain a 35°C battery pack temperature, once the outside temperature is above 15-20°C?
- So is there any other cooling circuit in series? Or only a vapor-compression cycle directly implemented in the battery.

I hope this question will interest you!

jkirkebo | 28 augustus 2012

The Roadster uses the AC unit to cool the coolant. My fridge can maintain -18C in +25C environments, I don't see why the Roadster AC would be any different.

Vuckoko | 29 augustus 2012


Their patents provide additional information on what they might be using:
- The liquid system inside the battery :,
- The global cooling system with three cooling loops:


Timo | 30 augustus 2012

My understanding about A/C is that it is basically a heat pump. Quite simple and quite energy efficient, it just gets hotter onto outside than air to pump that heat out and vice versa when it needs heat the battery (or interior of the car). Isn't very efficient at heating in very cold environment (around -20) so additional heater element is required there.

julian.thrussell | 29 december 2015

The first use of water cooled batteries was in a Royal Navy submarine in the 1920's. It was used to allow for fast charge / discharge cycles.

It seems that even new ideas are often old ones.

Bob.Calvo | 29 december 2015

Ditto for threads. ;-)

Cat4Change | 26 augustus 2016

Do you think the P100D has a liquid cooled battery?

murphyS90D | 26 augustus 2016

All S and X cars have a liquid cooled battery.

Cat4Change | 26 augustus 2016

Yes, it goes through tubes, but could the cells be submerged in the liquid and avoid the complex tubing? And thereby be able to transfer more heat in less space?

Cat4Change | 26 augustus 2016

And do you think the packs were built in the Gigafactory?

murphyS90D | 26 augustus 2016

With cells submerged in liquid there is no way to guarantee a uniform flow rate, especially in the corners. It only takes one overheated cell to cause problems.

I have no way of knowing what is happening in the gigafactory.

Cat4Change | 26 augustus 2016

Thanks, makes sense. Must be a more straight forward solution than all those tubes.