Tesla Model S Recycles Waste Heat to Warm the Battery

1 year ago by George Bower 40

Inside Tesla Model S Battery Management

Inside Tesla Model S Battery Management

The Waste heat comes from the drive motor and power electronics.

Range loss in cold weather is a huge problem for pure battery electric vehicles and manufacturers use various approaches to deal with the problem. The simplest approach is resistive heating in the cabin and/or battery. This is the low cost approach but it also results in the largest penalty on range.

Nissan LEAF Utilizes A Heat Pump

Nissan LEAF Utilizes A Heat Pump

Other manufacturers like Nissan use a heat pump. The heat pump improves range loss but adds expense. A heat pump is an air conditioning system run in reverse that pulls heat from ambient air.

Have you ever wondered why the Tesla Model S doesn’t have a heat pump? Certainly cost is no object in a luxury car costing as much as $100,000.

Perhaps the reason Tesla does not use a heat pump is that they have a better way to do it.

Tesla has a patent app outlining a practical approach to cabin heating where waste heat from the drive motor and power electronics is used to warm the cabin. As it turns out, the production model S uses a similar but subtly different approach. Waste heat from the drive motor and power electronics is still used to warm the vehicle. However, Tesla chose to heat the battery instead of the cabin with the waste heat.

What are the finer points on how the production vehicle and the patent differ?

The patent shows waste heat being used in the cabin as shown in the following simplified sketch. A detailed technical sketch is provided in the original article.

Simplified Tesla patent sketch shows waste heat being used to heat the cabin

Simplified Tesla patent sketch shows waste heat being used to heat the cabin

In the patent the cabin liquid glycol heating loop can be connected to the drive motor and power electronics glycol cooling loop with some valves.

Note that the patent shows 4 loops: 3 glycol loops and one refrigerant loop.

Now let’s look at a simplified sketch of the production model S

Production Model S uses waste heat to heat the battery instead of the cabin and has eliminated 1 Glycol loop

Production Model S uses waste heat to heat the battery instead of the cabin and has eliminated 1 Glycol loop

There are two big changes between the patent and the production vehicle: the waste heat is used in the battery instead of the cabin and one loop has been eliminated. The production vehicle now has only three loops, not four. Tesla eliminated the Glycol cabin loop.

Do we know this for sure? Is it 100% proof positive that this is the case in the production vehicle?

We have no direct information from Tesla itself. However here is a quote from an article in the MIT technology review:

Tesla takes a different approach. Once you start driving, heat generated by the motor is used to heat up the battery. This approach is more efficient, since it uses waste heat rather than electricity. But it takes a while to work because the motor doesn’t produce much heat. As a result, it might take several minutes before the battery is warm enough to provide full acceleration.

In addition we have some screen shots of the Model S in “diagnostic mode”. This diagnostic mode schematic was posted on the Tesla Motors Club forum in regards to our article Tesla outlines practical approach to cabin heating with the following comment by member “Ingineer” (an electrical engineer).

While there is no way to use drivetrain waste heat to warm the cabin, they can use it to warm the pack when it’s cold.

Here is the diagnostic mode system schematic.

Diagnostic mode showing battery loop in parallel -not connected- with the drive motor and power electronics loop. Courtesy: Tesla Motors Club member “Ingineer”.

Diagnostic mode showing battery loop in parallel -not connected- with the drive motor and power electronics loop. Courtesy: Tesla Motors Club member “Ingineer”.

In another sketch we see the drive motor and power electronics loop hooked in series –connected-with the battery cooling loop.

Battery loop in series–connected-with drive motor and power electronics loop. Front radiator is bypassed

Battery loop in series–connected-with drive motor and power electronics loop. Front radiator is bypassed

So while we have no direct confirmation from Tesla that this cooling scheme is used in production, all evidence indicates that Tesla Model S warms the battery with waste heat from the drive unit and power electronics.

Implications for Chevy Bolt 200 mile EV

2017 Chevrolet Bolt Looking Good At NAIAS: (InsideEVs / Tom Moloughney)

2017 Chevrolet Bolt Looking Good At NAIAS: (InsideEVs / Tom Moloughney)

We do not have direct word from GM as to whether the Bolt EV will have a heat pump. However, some GM-Volt forum members who attended the NAIAS claim to have heard Pam Fletcher, executive chief engineer of electrified vehicles, say that the Bolt EV will not have a heat pump but that the Bolt EV will have a similar heating scheme as the Volt EREV only “improved”.

What could this improvement be?

Perhaps it is waste heat recovery from the power electronics and/or drive unit. The scheme is easily implemented, low cost, improves cold weather range and is in tune with the “keep it simple” approach GM has used in the Chevy Bolt 200 mile EV. In another interesting twist, the Bolt EV uses a glycol loop for cabin heating –Tesla doesn’t- so GM could use waste heat in the battery OR the cabin.

Stay tuned.

About the authors: George Bower is a retired mechanical engineer with over 20 years experience in gas turbine power systems.

Co-Author of the piece, Keith Ritter is a mechanical engineer, and licensed professional engineer with over 35 years of experience in heating ventilation and air conditioning systems.


40 responses to "Tesla Model S Recycles Waste Heat to Warm the Battery"

  1. mhpr262 says:

    The EV manufactureres should just offer an additional little bio-ethanol fueled burner for their cars to assist with heating in cold climates. Much less range penalty in cold weather that way.

    1. Priusmaniac says:

      If they include an ethanol burner, then let it at least be a Thermophotovoltaic generator equipped burner, so it will quietly be producing some electricity as well. It is not much more complicate than a normal photovoltaic generator, the only difference is that it is made to work with the infrared light emitted by the burner instead of the visible light emitted by the sun. If mass produced the specific panels can be produced at a much lower cost and better efficiency than what is presently the case.

      1. Bill Howland says:

        Wake me up when you find a manufacturer to do that. The only place I have a thermopyle in my house is to run the solenoid gas valve on my hot tub heater.

        I’ll be sleeping a long time.

        1. Priusmaniac says:

          Perhaps there could be a company quiet well known to Elon Musk, that happen to be in the related photovoltaic business. You know, one called Solar city.

          1. Bill Howland says:

            I own 38 non-Solar-City panels. Every one of them is Flameless.

            Or, maybe you are saying you can run your solar panels off of an ethanol flame.

            Now that’s the question: which of your Inventive Ideas is the better one?

            1). Running solar panels off of an Ethanol Burner?

            2). Quick charging your car with a 50,000 volt charger?

            1. Priusmaniac says:

              Apart from a little sense of humor you seem to have, I would say that essentially the Physics knowledge related to solar photovoltaics is close to the one for thermophotovoltaic because it is fundamentally the same goal of transforming the electromagnetic energy of photons into electricity. You do need different substrates to transform different wave lengths so it is another technology but the fundamental physics is the same.
              Here is a link to photovoltaics: http://jxcrystals.com/drupal/thermopv

              As for the voltage of 50000 volt that might be right to recharge a short range defense electric submarine or an electric version of a 747 but for cars you don’t need to go that high. But again fundamentally 50000 v or 5 v is the same thing only the electron energy is different so it requires a different technology. Rest assure if 50000 v somehow scare you, 1000 or 2000 v will do for cars at least if we want to access megacharging at 1000 KW. The reason is that since maximum copper section of the charging cable is limited the only way you can increase the power further is by increasing the voltage. At least 400 v is not going to make it, you can actively cool the line but that’s not going to be enough, you just need more voltage. Porsche is apparently testing 800 v which is double but for true high power that will need another doubling to 1600 v. Now if they go even higher to use even thinner wires why not. It will be up to balancing power capacity, technological feasibility and cost. But fundamentally, there is no limit. High voltage kind of inspire fear to some but that shouldn’t be. Putting yourself limits at the basic concept level is the worst you can do to find new things out of the box. Latter on there will always be more than enough naysayers and critics to make you fine tune the concept to better match what is feasible, but at first just don’t care about it and ask why not.

    2. zzzzzzzzzz says:

      Maybe they should provide whole ICE with catalytic converter. Oh wait, they already done it!

  2. krona2k says:

    I really do hope the Bolt has an improvement on the Volt HVAC.

    The Volt heater is quite poor, worse than the no heat-pump version of the LEAF. I know this having owned both cars through winter.

  3. SJC says:

    I figured this out 20 years ago.

  4. Darth says:

    Seems silly to not use a heat pump for the cabin if you already have AC anyway. But using waste heat for battery warming is brilliant, I wish the Leaf had it.

    1. Priusmaniac says:

      That sound strange indeed, if you have the airco compressor you can as well add some extra valves to make it work in reverse. It sound simple but it actually is indeed that simple. Not doing it is a really very strange.

    2. Nathanael says:

      Tesla does in fact use a heat pump for cabin heating, *and* for battery heating.

  5. jerryd says:

    There isn’t much heat from the electronics or motors so not much use for cabin
    Only 1kw or so at 60mph and they would be lucky to actually collect 30% of that.
    Most likely they have a resistance heater for the battery too. Leaf, Volt do.
    But these are only needed at really low temps, under 30F in most cases.
    And lithium doesn’t lose power anywhere near as much when cold as lead batteries.
    Here in Fla the Volt pack I’m using I’m not even putting any heating on the battery, just circulating the coolant to keep the cells all the same temp, the most important thing.
    Though easy to put a block heater in the hose line if needed.
    If I lived up north I’d just insulate, letting the battery mass keep them warm.

    1. james says:

      Don’t forget that the battery itself will generate heat while being used.

    2. lewl says:

      300W for one hour is an extra 2km.
      All for the cost of a valve and controller to merge the coolant loops.

      Seems worth it to me.

    3. no comment says:

      it looks like tesla made the right choice in the production model. the system described in the telsa patent (the application has issued into a patent) doesn’t look like it would be very effective at heating the cabin (and ultimately, that should be the objective).

      electricity is a very poor energy source for heating purposes. a heat pump might work in mild climates, but you don’t want to have to hang you hat on a heat pump (which is prone to shutting down) if you live in colder climates. one of the nice things about the ERDTT mode in the Volt is that it is really good at quickly heating the cabin and maintaining a comfortable cabin temperature.

      nobody is going to be putting triple glazing in automobiles and since single lite glass is a very good thermal conductor, the cabin in an automobile is a poor thermal envelope. it is better to use the waste heat to warm the battery because you can insulate the battery to a greater degree than you can a cabin.

      1. Frank says:

        ERDTT in very cold conditions is actually bad.

        It keeps revving the engine when engine oil is at it’s coldest (bad). This is also when the engine must run richer in order to keep a good combustion while cold(bad). The catalytic converter does not reach it’S optimal temperature to properly do its job (bad to very bad). Overall gas consumption for 10-20 miles is crap when ERDTT is on. I got 30 mpg at times… While still using up all of my battery on 18 miles of range(again, very bad).

        All the while, a simple little bio-ethanol fueled burner would do a better job at 0.5L per hour. I would actually burn less fuel than using the full gas engine, while maintaining my normal electric range.

        I would definately offer this as an OPTION for colder climates. This is the best, cheapest, less poluting solution right now. We can review this when 100+kwh batteries are the norm.

        1. Doug says:

          Actually that isn’t correct. The Volt’s coolant loop warms the engine oil prior to starting the generator

          1. Bill Howland says:

            I don’t see how pre-warming the oil is possible. In cold weather both my cars start the engine immediately, over which I have no control.

            The 2011 volt seems better than my 2014 elr, since it totally allows the electric heater to function at low fan. This helps keep the engine off for much longer since you can keep the coolant temperature warm enough to prevent restarting of the engine.

            The 2014 ELR will constantly cool the coolant under any conditions except if the recirculate is on, and the cabin fan is off. But then thats only good for a minute or so until the entire windshield fogs up.

            Of course, it would be simpler if in both cars the driver/owner actually could start the engine when he wanted to, but then that’s just another example of GM’s arrogance that must be tolerated if I want their electric cars.

            I wonder that gotchas are going to be in the BOLT.

            As far as the “S” is concerned, it then means a ‘conservative pedal’ won’t heat the battery much at all, so I disagree that this is a ‘better way’. If the car is driven hard, the battery should self-heat, and there will be so much waste heat from the controller/motor combo that it in that case won’t be necessary and it will have to be wasted.

            The commenter who stated a low-temp heat pump (like the ones made by Mitsubishi) would provide coefficients of performance far greater than 1 for all but the lowest of temperatures.

            Its too bad with all the discussion of the “S” there have only been two examples of very cold “S” range loss – first the Broder test – , and then a few months later, the Minnesota Road Trip Lady.

            1. ClarksonCote says:

              “As far as the “S” is concerned, it then means a ‘conservative pedal’ won’t heat the battery much at all, so I disagree that this is a ‘better way’”

              I agree, and even the MIT review says the motor doesn’t produce much waste heat, so it takes a while. I just don’t see much advantage to all the additional complexity. If motors had a ton of waste heat, they wouldn’t be so superior to a gas engine.

              As to GM forcing owners to start their engine in cars like the ELR and Volt, I’m hoping that they that nobody will have an unexpected ERDTT event in the Bolt… But at this point even an engine starting on the Bolt wouldn’t surprise me, with the way they’ve handled the Volt and ELR 😉

          2. lewl says:

            Actually, this isn’t correct.
            The only coolant that can be warmed while plugged in is battery coolant. Which does not mix with engine coolant. There is no engine warming via electricity in any form

  6. Stretch2727 says:

    I think the best idea is to redirect the regenerative braking to a resistance heater to heat the battery or cabin. Part of the low efficiency during cold is the loss of regenerative braking until the battery is warm. You do not want to push a lot of current to the battery until it is warm. If the motor could push electric current to a electric heater this would speed the process of warming the battery and save some brake pads as well. Consistent regenerative braking as well. I am always a little surprised in the cold weather when I first lift off of the accelerator and there is little or no regen.

    1. lewl says:

      If your battery heater is on, it already does this.
      Heater takes xW. If you regent more than that it is dumped in battery, otherwise the power just flows to whatever is using electricity on the HV circuit.

      1. Djoni says:

        Nope, not close.
        When the battery is cold it does not take any regen and the only heat that could be feed is thru the existing resistive heater integrated in the battery.
        Beside producing internal heat when cycling charge and discharge, this is all you can get.
        And all excess energy is simply shed in the classic friction brake and wasted to any good use..

  7. ModernMarvelFan says:

    All “wasted” heat should be recovered…

    As far as Heat pump goes,

    Heat pump is almost useless below freezing…

    Its best use is when temperature is around 40 degrees which is way more efficient than resistive heating…

    1. Bill Howland says:

      Used to be true MMF but no longer. Mitsubishi heat pumps are effective down to about 0 deg F, and are quite efficient heat movers at only normal cold like 20 deg F.

      1. ModernMarvelFan says:

        Ok, maybe my information is old…

        What kind of “refrigerant” is used?

        What is the system max pressure?

    2. DavidCary says:

      This statement is repeated so often it is infuriating. Almost useless implies a COP of 1.2. You don’t have to get to Mitsus excellent performance to be much better than 1.2 down to 10 degrees. Mitsu has COP at 2 at -3 degrees.

      I haven’t seen testing down to -10 but it looks like the Mitsu might get almost useless at -10. So the wording needs to have a slight change – 50 degrees.

      1. Bill Howland says:

        Just checked the mitsubishi website.

        Difficult to distill the information there, but Mitsubishi claims 2 facts for their “Hyper-Heat” split system heat pumps. I used the largest 48,000 BTU/Hour heating model.

        1). FULL capacity at 5 deg F (that is amazing)

        2). Coefficient of performance (equal to 8.2 EER) of 2.4 at 17 degrees F – equally fantastic. Meaning each 1000 watts of electrical usage gives 2400 watts worth of heating effect.

        I know this is not a car system, but the principle applies. Too bad they don’t make an upscale BEV that would include this system.

        1. Bill Howland says:

          The above remark should be tapered since there will be quite a bit of defrosting at those temperatures but even so, when running the effectiveness is remarkable.

          Now THAT is something worth being patented.

          I don’t get Tesla trying to patent something that is obvious and has been done by everyone for at least the past 100 years. People all the time and certainly manufacturers have used the waste heat from things heat other things.

          Since the electric usage is so high, I bet this is one of Tesla’s free patents that no one will use – that is if its patentable in the first place.

          Example: I just bought a cheapie Frigidaire (aka Electrolux) 21 cubic foot freezer. The defrost water falls into a cup in contact with the ‘hot gas’ area of the compressor discharge. The ‘waste heat’ is lowered by the water (and therefore the compressor energy input is lowered since the condensing temperature is also lowered from what it would have been), and it evaporates the water in the cup at the same time. Now that’s a 2 for one. Its a better efficiency improvement than the one mentioned here about ‘saving heat’, and even so, I doubt Frigidaire even applied for a patent since it is so obviously an easy thing to do.

          1. ModernMarvelFan says:

            I just looked up some of the residential heat pump specs by Mitsubishi.

            It is impressive. (with COP around 2.2-2.6 at 5 deg F) However, their claims of 100% efficiency down to 5 deg F doesn’t specific the output temperature or the calculation doesn’t include the heating of the outside unit for defrosting.

            Yes, it is still efficient but what is the output temperature?

            The only thing they state is 100% capacity which still doesn’t specify temperature.

            Looks like they are still use

          2. ModernMarvelFan says:

            What is the temperature differential between condenser and evaporator?

            Yes, that is the original reason why heat pumps won’t work well near freezing. When the evaporator gets around freezing, it will start to freeze over the fins which limits the amount of the air exchange required in order for the system to properly work. If additional defrosting is required, then the energy has to come from somewhere which will cuts into efficiency.

        2. Sting777 says:

          This explains why Pennsylvania houses are converting out of oil into heat pumps. Winter nights don’t typically go below 20 degrees F.

  8. wavelet says:

    Sorry, but I don’t see how having only a waste-heat heating system is sufficient:

    1) What happens when the car isn’t being driven? There’s no waste heat at all, but the battery still has to be kept warm if it’s to be used soon.
    In fact, there are a lot of reports (including several on this site) about the Model S’s “vampire” electricity draw, which is very noticeable, esp. in subfreezing weather; miniscule draws like an alarm system wouldn’t account for it… Has to be battery warming.

    2) Ergo, there has to be another battery heating method in use. A heat pump doesn’t work in very cold weather (e.g., home airconditioners can heat a room if the outside temps are a max of 7-9°C below the desired indoor temp); therefore, so while there may still be one (efficiency tends to be double that of a resistive heater, and there is already an A/C system as people note), it’s a safe bet there’s a resistive heater as well.

    Bottom line, the article pretty much has to be wrong — maybe waste heat is sufficient while the car is being driven, but the design can’t have eliminated all other methods.

    Has InsideEVs tried simply emailing Tesla and asking? I hardly think the basic existence of such systems would be a secret…
    If not, why not? Simply reporting on one source without adding value is half the job.

    1. lewl says:

      It’s not either/or.
      It uses this and/or standard resistive heat.
      Kia soul also has this tech.

      1. wavelet says:

        Of course, but the article wrongly says there’s isn’t a heat pump:
        “Other manufacturers like Nissan use a heat pump. The heat pump improves range loss but adds expense. A heat pump is an air conditioning system run in reverse that pulls heat from ambient air.

        Have you ever wondered why the Tesla Model S doesn’t have a heat pump?”

  9. Djoni says:

    It all depend of which cold we need to overcome.
    Up north, all the heat you can get, recover, or efficiently produce is a must.
    Motor, inverter, braking, pre heat and accumulative heat of high thermal mass should be done.
    Insulating the heating loop and cabin should be done thoroughly as much as possible.
    So, recover, recycle, pre heat, resistive regeneration, heat pump and insulation has to be improve to work fine in the coldest climate.
    Having AC and not using it as a heat pump is simply stupid, because that’s what an air cooling is anyway.
    Not everybody live in “cold” Cal or Fl, some live in “polar” Minnesota or Canada.

  10. Phr3d says:

    northers need a ‘cold weather package’ and confidence that the ‘packager’ did their homework for efficiency.
    Heat pumps are the first thing that come to mind for a Norther’s experience – whether forced to electric heat for lack of NG, or other reasons, the Heat-Pump was born and died in the eighties – Many people found that they needed to ‘bundle up’ in their homes as the device simply couldn’t heat well below 10 degrees and bought space heaters.
    Glad to hear that the efficiencies have improved per Bill’s info above, cuz I can think of a few thousand Northers that get a bit Testy when the subject of Heat Pumps is even brought up in polite conversation, lol

  11. Proton says:

    Clearly the best idea is to live in south Florida and all problems are avoided: no freezing, no hills or mountains, and grid power at less than $.12/kW.

  12. Speculawyer says:

    Tesla did some real engineering.