Tesla or GM: Who Has The Best Battery Thermal Management?

1 year ago by George Bower 133

Tesla Battery

Thermal Management Systems Play A Larger Roll As Charging Levels Increase

Coming In 2018, The e-tron Quattro Will Be Taking 150 kW Charges

Coming In 2018, The Audi Q6 e-tron Quattro Will Be Taking 150 kW Charges

It could be argued that a battery’s Thermal Management System (TMS) is equally as important as the battery cell chemistry. The battery cell’s life depends on the TMS.

As we go to higher and higher charging power levels (shorter recharge times) the TMS plays an even bigger role in rejecting even larger amounts of heat. Tesla’s super chargers are already at 120 kw with 150 kw planned for the future. Audi has announced a production vehicle to compete with Tesla that has 150 kw charging. Audi has also announced a US charging network based on 150 kw charging.

So the stage is set for higher and higher charging levels. With that in mind, let’s look at Tesla’s and GM’s thermal management systems.

When evaluating a battery TMS we need to keep in mind some parameters used to compare the systems:

-Simplicity
-Cost
-Ability to remove heat (BTU/sec)

Tesla

Tesla’s thermal management (as well as GM’s) uses liquid Glycol as a coolant.  Both Gm’s and Tesla’s systems transfer this heat to a refrigeration cycle and use electric resistance heating in cold weather. Glycol coolant is distributed throughout the pack to cool the cells. Considering that Tesla has 7000 cells to cool this is a challenge.

Turns out Tesla has a patent app for their system. It is based on a ribbon shaped metallic cooling tube that snakes thru the pack. A sandshot from the patent app is presented below.

Tesla gm cooling gb slide 1

 

This ribbon shaped cooling tube interfaces with the cells as shown below.

Tesla gm cooling gb slide 2

Now let’s look at one module in a Tesla Model S pack (16 modules total). The cooling ribbon snakes thru the cells as shown below.

Tesla gm cooling gb slide 3

Simple: Yes.
Cost effective: Yes.
Ability to remove heat:  Pretty good.

General Motors

GM uses prismatic (rectangular) shaped cells. Each cell is roughly the size of a children’s book. Sandwiched between the cells is an aluminum cooling plate.

Slide: Courtesy GM

Slide: Courtesy GM

There are 5 individual coolant paths passing thru the plate in parallel not in series as the Tesla system does. Each battery pouch (cell) is housed in a plastic “frame” The frames with coolant plates are then stacked longitudinally to make the entire pack.

Illustration: Karl Reque via CompositesWorld

Illustration: Karl Reque via CompositesWorld

Simple: Yes.
Cost effective: Yes.
Ability to remove high heat loads: GM’s system is slightly better than Tesla’s system in our opinion. However, in the grand scheme of things both Tesla and GM’s system have one big drawback. The heat transfer coefficient of Glycol leaves room for improvement.

Tags: , , ,

133 responses to "Tesla or GM: Who Has The Best Battery Thermal Management?"

  1. Anthony Castro says:

    This was great! I’d love to know about other manufacturers too. Do the automakers release BTU/sec extraction rates?

    1. Mikael says:

      Most likely not since it’s an obsolete unit. 😛
      They might release the number in watts though.

  2. ffbj says:

    Interesting and well done comparison. So what is the alternative glycol?

    1. Elroy says:

      Look at the BMW i3 system.

    2. Priusmaniac says:

      There is an alternative, carbon dioxide as a coolant. It is Benign and you could compress CO2 cool it in the radiator and then release it at low pressure in the battery compartment. It would turn back to gas instantly and flow around the cells effectively cooling them. That would make the ducts unnecessary while improving the contact surface in the same time. When fast charging, you could have a buffer tank on the high pressure side that would allow the huge cooling needed for a 10 minute charge exactly for that same amount of time. Of course a leak would be no problem and sealing the outside of the battery is much easier than a ribbon turning around all the cells. Beside the glycol can’t give the mentioned 10 minute boost like CO2 can.

      1. Bill Howland says:

        Forget it. CO2’s critical temperature is 86 degrees.

        I’ll leave it for PUPU to explain the ramifications of this.

        1. miggy says:

          86 is hot our average is only about 18c

      2. Bill Howland says:

        Now, in case you are thinking of using SuperCritical CO2, most popular commercial systems, such as DENSO, run at discharge pressures of around 1,400 PSI (that’s 10 MPA), and 212 deg F (100C).

        While the efficiencies are reasonable, the materials-science issues in a battery are somewhat daunting. Plastic piping ain’t gonna cut it.

        1. Priusmaniac says:

          Of course super critical and the piping must indeed be in more resistant materials.

          1. Bill Howland says:

            Fine, the next time I need to heat my battery really hot – this may be the way to go.

  3. Mister G says:

    Very informative…how would you know if there is a leak?

    1. gsned57 says:

      That was my concern as well. I’m assuming that ribbon has a very thin wall and would be concerned of pin holes forming and the whole thing springing a leak. I’ve never heard of that happening in a Tesla so maybe it is robust enough not to be a problem. One alternative is Nissan’s air cooled battery that turned to crap in extreme temps. I think I’d take the risk of a leak and maintain most of my AER vs. getting stranded in the winter or hot summer.

      1. Josh says:

        Nissan’s packs are not “air cooled”, as that would imply they are actively blowing air over the cells. This isn’t true.

        The pack has no active management system, only passively cooled through the air. Nissan was very reliant on the cell chemistry to be less sensitive to temperature than what GM and Tesla uses (which is the other factor not addressed here). The “Lizard” battery was supposed to fix this deficiency in the battery.

        1. Tom says:

          Correct, and in my last post that disappeared (deleted?), I’d like to know what Nissan has planned for future cars. A 150 mile LEAF (200 mile?) would, in my opinion, experience battery cooling issues…

          1. ClarksonCote says:

            Not necessarily, the vehicle takes so many watts of power to keep it moving at a given speed. The larger the battery pack, the slower the pack is being discharged as a percentage of total capacity.

            This generally means the pack won’t get as hot as a smaller pack would.

            As an example, trying to pull 100W from a 1Wh battery will cause the battery to melt down. Pulling 100W from a 100Wh battery is fine, and a typically acceptable draw and temperature rise. Pulling 100W from a 10,000Wh battery will have effectively no rise in temperature.

            1. Andrew says:

              Agreed. The 30 kWh pack will be running at a 1.7C rate on a 50 kW CHAdeMO charger whereas the old 24 kWh pack would be at 2.08C.

              Full 80 kW acceleration in the 30 kWh car will draw 2.7C whereas it would have been 3.33C in the old car.

              The larger the pack, the less thermal management it generally needs as a result of charging/driving activity.

              Mitigating external temps is where Leafs fall flat on their faces as there’s no way for them to extract a heat soak from ambient surroundings. If it’s 110 degrees F average for 48 hours (Arizona) then the pack will be *at least* 110 degrees and rise from there. A Volt or Tesla will actively bring that temp down to ~87 degrees and keep it there for the duration of the drive. Then the thermal inertia of the pack will prevent excessive heating on subsequent drives, hence the jarring contrast in degradation between TMS/Non-TMS cars.

              In any event, the larger the pack, generally the less load on TMS due to lower C rates in charging and driving.

              1. ClarksonCote says:

                Yep, they definitely will still have a problem in hot ambient temperatures without active cooling.

                1. kdawg says:

                  Isn’t Nissan now also using LG cells? So why do they not need to have a TMS for the LG cells but other do?

              2. zzzzzzzzzz says:

                Nissan battery uses different chemistry and at least new one isn’t so sensitive to high temperature. It is big advantage in my opinion as long as it is true. Much simpler system, less points of failure (think about diagnosing and fixing coolant leaks when a car gets 7 years old and they will inevitably appear just like with gas cars), no A/C kicking in at hot night and creating vampire load to cool battery. Typical house in the US with attached garage doesn’t have cooling in garage. On sunny day it may be much hotter in garage than outside, and it continues after sunset for some time. Running car A/C from time to time to preserve battery may work, but it isn’t perfect solution.

                1. Spider-Dan says:

                  Well, sure, any system would be cheaper if you could simply delete features that are not needed. But the battery degradation on the Leaf so far has clearly shown that exclusing a TMS is not the way to go.

                  For every one Volt that has future TMS maintenance issues, there will be dozens of Leafs with severely degraded batteries.

              3. Stephen Hodges says:

                My ’11 leaf has shown this to be true… rather fast degrading (32% in 4 years) caused by charging when I get home on top of a long hill (1500 ft elevation) when the battery is in the 90’s rather than the 70’s or 80’s ambient. A lizard replacement is called for.

      2. Roy LeMeur says:

        There is likely some type of reservoir with a level sensor similar to a modern brake master cylinder or engine cooling system.

  4. Omar Sultan says:

    Interesting article. I think without measuring the ability of each system to remove heat or alternatively the ability of each system to hold a pack within desired range, its hard to gauge which is “better”.

    The other variable is the sophistication of the software logic controlling all this.

    1. ClarksonCote says:

      Parallel cooling will result in a more even cooling of the whole pack, whereas serial will cause some portions to be warmer than others.

      1. Pushmi-Pullyu says:

        In theory, yes. But as they say: “In theory, there’s no difference between theory and practice. In practice, there is.”

        We can’t judge how well these systems work just by reading a patent. We’d need tests of the actual systems, a third-party comparison test by qualified technicians or scientists, to be able to definitively say which cooling system is better at maintaining a stable temperature.

        1. ClarksonCote says:

          I think my general statement is true regardless. However whether or not the max delta in a series configuration is 0.1 degrees max or 10 degrees max can be validated with actual data.

      2. ModernMarvelFan says:

        That is only true if the flow rate of the system is limited.

        In the Volt system, I believe each cells are in parallel, but modules are in series…

        Also if the system can’t handle the heat, then the entire system gets hotter for everyone… where the Tesla system will be more uneven.

        1. Bill Howland says:

          Yeah, GM’s system has the proof of the pudding.

          As far as Vex’s statement that the edges of the Volt aren’t cooled, it appears botth h sides R totaally//lly / ////////

  5. Vexar says:

    The pouches GM uses are a larger volume per cell than the 18650 and therefore have more heat at the center of the pouch than at the edges. I don’t think this analysis is sufficient; it does not take into consideration where the heat build-up is in a pouch cell. Is it at the anode or cathode, or in the lithium of the cell?
    Fundamentally, since Tesla’s Ludicrous mode requires (in part) replacing the more heat-tolerant contactors with iconel instead of steel, Tesla certainly is working harder on finding out where the heat build-up is. 10 years from now, this will be an interesting historical discussion, because battery chemistry and design will likely have had a generational improvement.

    1. Tom says:

      Plus one

      1. Stimpacker says:

        Thanks for sharing.
        Vexar makes a good point. Another consideration should be what happens when there’s a heat event? A physically compromised cell may experience a thermal reaction and generate heat at a focused spot.

        1. Pushmi-Pullyu says:

          One advantage Tesla gets by using cylindrical cells is that it naturally creates space between the cells; space which Tesla fills with a (non-circulating) “goop” which is both a fire retardant and a temperature-stabilizing heat sink.

          There are many differences between the battery packs not mentioned in this article, let alone taken into consideration.

    2. John Hansen says:

      Look at the picture, the GM engineers seem to have accounted for that. The coolant loop that goes through the center of the pouch is much shorter than the loop that goes around the exterior, meaning that the center of the pack receives greater cooling. In other words, a greater volume of coolant moves through the center of the pack, thus removing more heat.

      1. ClarksonCote says:

        Plus One

    3. Pushmi-Pullyu says:

      Vexar said:

      “I don’t think this analysis is sufficient; it does not take into consideration…”

      It doesn’t take a lot of things into consideration. Perhaps the most important is just how many BTUs per second each system can transfer from the battery pack to the condenser… assuming the Volt uses a condenser for cooling of the battery coolant. But even that info wouldn’t be sufficient to judge the potential for keeping each system’s battery pack at a stable temperature, since the Tesla system needs to be able to handle a lot more BTUs, as the Model S is able to accelerate much faster and has a higher top speed, plus the Model S is a heavier car needing more power to achieve the same acceleration.

      This article has a lot of details about the cooling systems which I find interesting, but certainly there is not enough info there to reach any conclusion about which system is “better”.

      1. ClarksonCote says:

        More power needed doesn’t mean more heat needing to be dissipated. That need depends on cell size relative to lower draw out of the cell, and kind of by extension, volume of the pack.

        If the Tesla battery needs to provide 3 times the power of the Volt, but the Tesla battery is 5 times as large, then it should not need as much heat dissipated. Of course, there’s many variables to consider, and neither of us should condense it into one dimension.

        1. Pushmi-Pullyu says:

          ClarksonCote said:

          “More power needed doesn’t mean more heat needing to be dissipated.”

          I think you mean: doesn’t mean more heat needing to be dissipated per cell. Unless you’re claiming that GM’s cells are inherently more efficient at delivering X amount of power with less wasted heat, then absolutely the entire pack is going to produce more waste heat if it has to deliver more power… that is, transfer energy to the inverter at a faster rate. And of course, the battery pack’s thermal management system has to be built to handle waste heat from the entire pack, not just any one cell.

          Some of that may be ameliorated by Tesla having a more open cell arrangement, with more space between cells, and by using smaller cells with a better surface to volume ratio.

          But there are unquestionably practical limits to the Model S in dealing with waste heat. It’s been well documented that you can’t run the Model S at top speed for long, until the car automatically goes into reduced power mode because something in the powertrain is overheating… probably the battery pack or the inverter, rather than the motor.

          Now, that’s not to say this is evidence that the Volt’s thermal management is “better”. It may merely be that the Volt’s lower top speed causes any waste heat to remain well within the ability of what the Volt’s thermal management system can handle.

    4. ModernMarvelFan says:

      Vexar wrote:

      “The pouches GM uses are a larger volume per cell than the 18650 and therefore have more heat at the center of the pouch than at the edges. I don’t think this analysis is sufficient; it does not take into consideration where the heat build-up is in a pouch cell.”

      That is why it is sandwiched with alumnium plate which is an excellent heat conductor to even out the heat regardless how the cell surface temperature varies. Remember that pouch is relatively thin so there is very little heat gradient within the cell where the 18650 cell can have a higher gradient from center to the surface.

      Now, Volt battery does have a higher discharging C rate, thus it will need a slightly better cooling system on per cell basis.

      1. ClarksonCote says:

        Well stated.

  6. George Bower “The heat transfer coefficient of Glycol leaves room for improvement”. So, you left that thought dangling, and without either an Opinion or a Fact to state your case on or from!

    So, what do you recommend over Glycol? What high performance cooling liquid or material do you recommend for an upgrade?

  7. Evan says:

    Great article! it makes sense that the two cars have totally different cooling not methods since they have different battery structures, but it wasn’t something I had thought about before.

    It would be fun to take a look at the BMW i3 and Nissan Leaf too. I would suspect they are similar to the Chevy volt because I know they aren’t using the 1850 cells that Tesla uses, but I don’t know! Maybe they are using another different battery format.

    1. pk says:

      I believe the i3 has a grid underneath the batteries that heats/cools using freon.

      http://darrenortiz.com/website_pdfs/BMWi3PG.pdf
      page 18 shows an exploded view of the battery assembly.

    2. Walt says:

      LEAF cells are air cooled passively

    3. Art Isbell says:

      The i3’s heat pump can cool its cabin, its battery pack, or both simultaneously. When its battery pack needs cooling, refrigerant is directed through it thus eliminating a heat exchanger that would be necessary if the battery pack was cooled with coolant (e.g., glycol). On i3’s equipped with seat heaters, electric resistance elements heat the battery pack during cold weather preconditioning. So this is different from Tesla’s cars and the Volt.

      The Leaf hasn’t had active battery pack temperature management (i.e., with a coolant or refrigerant for cooling and resistance elements for heating). I don’t think this has been added to its new 30 kWh battery pack. The VW e-Golf doesn’t have active battery pack temperature management, either.

      The i3’s battery cells are large format prismatic cells similar to those of the Volt and Leaf.

      1. Djoni says:

        Nissan Leaf have a 300 watts electric heating resistor to keep the battery at no colder than -15c° for protecting any damage.
        I3 has a inert gas coolant system, not sure that is Freon, and can be heat up to optimal prior to drive with an 8 hour time provision.

        Heating the battery externally take some time.

    4. Pushmi-Pullyu says:

      “…the BMW i3 and Nissan Leaf… I would suspect… they aren’t using the 1850 cells that Tesla uses, but I don’t know! Maybe they are using another different battery format.”

      Tesla is the only auto maker making a battery pack using cylindrical 18650 (not 1850) cells. Most or all of the other plug-in EV makers are using “prismatic”, or block-shaped, cells. And there’s no standardization of the prismatic cells, either. Pretty much each auto maker is going his own way, here. That’s good for competition; let the best system win!

  8. taser54 says:

    It would appear that serial cooling of a pack would provided more cooling at the front of the pack and less cooling at the end of the pack.

    1. Nix says:

      That was my observation too. The first battery the coolant travels past will have the highest heat differential, and dissipate the most heat. The coolant will we warmer by the time it reaches the last battery, and potentially not cool it as much.

      But it is impossible to know the full impact of this without knowing the temp of the coolant at each end of the tube, and what the safe temperature ranges are for the batteries.

      Because all that really matters for cooling is if the batteries are kept in their safe range. It wouldn’t really matter if the batteries towards the end of the cooling tube’s route were 5 degrees warmer than the first cells. As long as they were still at a safe operating temp, they are fine.

      So even if the Tesla system might not result in as even cooling as the GM system, it might be a case of “good enough” being completely sufficient. So trying to go for “perfect” might be a waste of resources.

      Simply routing the coolant from the top and rear of the battery, running it towards the bottom and front of the pack may be sufficient enough to balance out other factors, such as airflow going from the front to back of the pack, and the nature of heat rising upwards in the pack.

      In the end, if both systems keep their respective batteries inside the cell’s safe operation temp. range, they are both equally effective at keeping the cells inside their safe operating range.

      1. Pushmi-Pullyu says:

        Nix said:

        “The first battery the coolant travels past will have the highest heat differential, and dissipate the most heat. The coolant will we warmer by the time it reaches the last battery, and potentially not cool it as much.”

        But this will be affected by how much coolant there is in the system — the thermal mass of the coolant itself — and how fast it’s circulated.

        “Because all that really matters for cooling is if the batteries are kept in their safe range. It wouldn’t really matter if the batteries towards the end of the cooling tube’s route were 5 degrees warmer than the first cells.”

        That’s right. The fact that the Volt cells are larger and therefore have more heat differential between the core and the surface, may impact battery operation and life more than the differential in heat between the Tesla’s coolant at the entry point vs. the exit point.

        It’s at least problematic, if not futile, to pull out one single factor in a system this complex, and try to figure out from an armchair analysis just how much (or how little) impact it will have.

        1. ModernMarvelFan says:

          “That’s right. The fact that the Volt cells are larger and therefore have more heat differential between the core and the surface, ”

          Why would it have more differential? Due to the power dissipation or due to the large format?

          The surface area of the pouch cells are much larger with respect to its volume. That is the why the aluminium plate will evenly distribute the surface temperature across the entire cell. In the Tesla case, there will be a temperature gradient between the center and edge of the cell.

          Lastly, the GM cells are sandwiched on both sides by cooling plates where Tesla cells are only contacted on 1 side where the coverage factor is less than 1/3.

          1. Pushmi-Pullyu says:

            ModernMarvelFan said:

            “The surface area of the pouch cells are much larger with respect to its volume.”

            Hmmm, no, I don’t think so.

            The surface area to volume ratio isn’t merely depending on shape, but also on size. The cube-square law rules here; larger objects have a lower surface-to-volume ratio. In other words, the larger the pouch cells are, the more heat will build up inside them.

            One advantage Tesla gets in battery thermal management, from using a thousands of small cells, is the improvement in surface-to-volume ratio, as compared to other auto makers which use far fewer, larger cells. That includes GM’s pouch cells.

            As I recall, the Leaf’s individual cells are about the size of a paperback book, much larger than the 18650 cells which Tesla uses. Are the cells GM is using much smaller than the Leaf’s cells? I doubt it.

            1. ClarksonCote says:

              “The surface area to volume ratio isn’t merely depending on shape, but also on size. The cube-square law rules here; larger objects have a lower surface-to-volume ratio. In other words, the larger the pouch cells are, the more heat will build up inside them.”

              Perhaps if objects are larger in three dimensions, but an object that is long and very narrow in depth, does not have a low surface to volume ratio. For example, a Manila envelope.

              1. Pushmi-Pullyu says:

                If the Volt’s battery cells were as flat as an envelope, more two-dimensional than three-dimensional, then you’d have a point. But they’re not, and you don’t.

                See photo of a typical pouch cell below:

                http://batteryuniversity.com/_img/content/pack6%281%29.jpg

                Furthermore, the Volt’s configuration of stacking cells back-to-back, makes the pack more akin to a solid block than an envelope. Contrariwise, the Model S’s more open arrangement, with the cylindrical cells, round in cross-section, automatically leaves open space around each cell.

                Now, if you can find the actual thickness of a Volt pouch cell, so we can compare that to the Model S’s 18650 cells, which have a maximum thickness of 18 mm, then we might have a more profitable discussion of the point. Personally, my Google-fu failed to find that figure.

            2. ModernMarvelFan says:

              When an object is as thin as the pouch with larger surface area, then the ratio favor the pouch. The thickness matter a lot.

              Let us do some math? =)

              18650 cells are 18.6 mm (diameter) x 65.2mm (height)

              So, the math for cylinder is:

              Volume = Pix(D/2)^2 x H where D is diameter, H is the height.
              Surface area = Pi x D x H + 2xPix(D/2)^2

              So surface to volume ratio is: (PiDH + 2Pi(D/2)^2)/(Pi(D/2)^2 X H)

              If we ignore to top and bottom surface, the the ratio is approximately 4/D or in this case about 0.21 [units are in mm]. If we include them, the ratio is 0.2457.

              Now for pouch cells:

              Volt pouch cells is about 127mm x 177mm x 6.35mm (AxBxT)

              Surface = 2AB+ 2AT+ 2BT
              Volume = ABT

              Surface to Volume ratio: 2(AB+AT+BT)/ABT = 0.34

              0.34 is a lot better than 0.24.

              If we only use the contact surface to calculate the ratio to volume, then the pouch surface to volume ratio is basically 2/T where T is the thickness. In the 18650 case, it was about 4/D and D is the diameter.

              The cross over point is when 2/T = 4/D where D is 2T.

              BTW, that is also assuming that the entire surface (beside the top sides) are cooled. In the 18650 case, they aren’t. In the pouch case, most of the large surface is.

              1. Bill Howland says:

                Yeah, MMF is correct here. Clarkson’s comment (sorry) doesn’t apply with pouches. A given pouch thickness that is 10 times as long and 10 times as wide will have 100 times the volume (and therefore, capacity) , and also 100 times the surface area. I tried to say this before,but disques was not letting me type – that the cooling percentage area appears far better in the Volt than in the S.

                The S only has ‘coolant’ on a small fraction of the cell, whereas it is on the vast majority of the pouch.

                As far as the volt ‘batteries’ being in series, as far as I know the delta-t on each battery is low, so 3 or 4 delta-t is a triviality. The fact the pouch coolers are in parallel allows a significant pressure drop around each pouch, thereby intrinsically forcing even flow through all of them because the delta ‘p’ on each pouch can therefore be HIGH since there are only 3 or 4 such drops.

                Ex: Lets say the pressure drop on the pouch is 1 psi. (that’s around 7 kpa for you metric lovers). Since the pressure is so high on each cooler, they can be designed for this drop at a certain flow rate, therefore all will get it and they don’t have to worry about surface tension effects since the pressure will overcome it.

                But there’s no real reason to ‘armchair quarterback their design. Chevy said in 2010 that the battery may be charged at 11.5 kw no problem. It also seems to do 60 kw for 10 seconds just fine also.

                As far as PUPU saying this is a poor design, has he ever used heat sinks of anykind, or had to calculate delta t’s for anything whatsoever? I just love it when people grandstand without the slightest rationale or evidence.

                1. Pushmi-Pullyu says:

                  “The S only has ‘coolant’ on a small fraction of the cell, whereas it is on the vast majority of the pouch.”

                  Looks to me like the picture above shows the Volt’s coolant running in only five fairly small lines past each cell. And contrary to what someone posted in this discussion, the Volt’s cells are not each sandwiched between two cooling plates. Rather, each cell is part of a set of two with a layer of thermal insulation between, which means the cooling plates are only between every other cell, not between each. In other words, each cell is up against only one cooling plate, not two. Of course, the aluminum plate itself will help absorb heat, but where is that heat gonna go? The cooling plate is sandwiched inside the pack with only the edges exposed. So most of that heat is gonna build up inside the pack unless the circulating coolant carries it away.

                  If I understand the diagram in the Tesla patent correctly, the Model S’s TMS has a flat, ribbon-like tube/hose carrying coolant, a tube/hose which runs the full height of the cell. The potential for rapid heat transfer would seem to be much greater for Tesla’s arrangement, unless the flow rate of battery coolant in the Volt is exceedingly high.

                  Again, we need to look at the entire picture, not just pull out one factor and ignore everything else.

                  1. ModernMarvelFan says:

                    ” Of course, the aluminum plate itself will help absorb heat, but where is that heat gonna go? The cooling plate is sandwiched inside the pack with only the edges exposed. So most of that heat is gonna build up inside the pack unless the circulating coolant carries it away.”

                    LOL. That is funny. Isn’t that the entire point of flowing coolant?

                    The aluminium plate is designed to absorb heat and even distribute it and the coolant takes it away.

                    The key here is thermal junction resistance. The biggest thermal junction resistance is within the cell and at the cell contact surface. The pouch cells are very thin thus the lower gradient across its thickness. So all the temperature gradient will be taken care by the AL cooling plate. The contact surface is also very good as it is very close to 50%, thus a much lower thermal junction resistance.

                    You assume that coolants can’t carry away sufficient heat from the cells. What evidence do you have to make that assumption?

                    “If I understand the diagram in the Tesla patent correctly, the Model S’s TMS has a flat, ribbon-like tube/hose carrying coolant, a tube/hose which runs the full height of the cell. The potential for rapid heat transfer would seem to be much greater for Tesla’s arrangement, unless the flow rate of battery coolant in the Volt is exceedingly high.”

                    1. Volt cooling system is pressured thus it can be high.
                    2. Regardless how high the coolant is flowing in the Tesla system, its largest thermal junction resistance are within the 18650 cells due to its thickness and its VERY LOW contact surface where less than 30% of it are in full contact with the cooling surface.

                    “Again, we need to look at the entire picture, not just pull out one factor and ignore everything else.”

                    Absolutely. But the key about “Heat transfer” is about thermal junction resistance. The thermal junction resistance are within the cells, cell to cooling plate and coolant capacity and coolant to exterior surface.

                    We don’t know how much coolant capacity and coolant to exterior surface heat exchange capacity are, so we can’t make assumptions (which you did based on the “tiny size” of the Volt tubes without any actual information).

                    What we do know is the fact that cell internal thermal junction is lower if it is thinner [or better surface to volume ratio] (and the cooling plate will take care of the gradient across the cell) and the higher contact surface to cooling plate is, the lower thermal junction resistance is. In both of those parameters that we know for sure, Volt cells have better surface to volume ratio and higher % contact surface to the cooling plate.

                  2. Bill Howland says:

                    Aluminum (Aluminium in the UK) is a decent conductor of heat. Thankfully GM has some engineers who have calculated the ramifications of heat removal.

                    It would be refreshing if you would stop being so insulting to those of us who have a familiarity with this stuff, – most people
                    would exercise a bit of humility after being
                    proven wrong so many times, but you just keep barreling ahead with your insults…

                    Usually, it is clear you don’t fully understand the issues in question, not withstanding you have no personal familiarity with this stuff in the first place.

              2. Pushmi-Pullyu says:

                ModernMarvelFan:

                Okay, thanks for finding the actual dimensions of the Volt cell; I couldn’t.

                Yes, okay, math wins; you are correct re surface to volume ratio on the cell level.

                But on the pack level, the geometry of Tesla’s more open cell packing still favors Tesla for heat dissipation. According to one description I found, the Volt’s cells do have a layer of insulation between every other cell in the pack (no telling how thick or thin the insulating layer is), but it’s still a much more closed, compact design, which means more heat buildup in the center.

                1. ModernMarvelFan says:

                  “but it’s still a much more closed, compact design, which means more heat buildup in the center.”

                  Wait, how can you come with that conclusion? Isn’t the point of the Aluminum plate is to get rid of any temperature gradient across the cell?

                  With a very thin plate (about 1/3 of a diameter of 18650 cell), how can you make a conclusion that center of the pouch cell would be hotter if the large surface is covered with AL cooling plate? That is the entire point of heat sink, to dissipate and even out the temperature across a given surface.

                  Tesla’s design barely has 1/3 of the cell in contact with the cooling plate, combined with much larger diameter, which format would have a “greater” temperature gradient across the cell?

                  1. SparkEV says:

                    MMF is correct, but there is another factor to consider. Heat transfer across thin walls of 18650 on farther side of coolant carrying ribbon will be poor. On GM, while the plate is thicker, it has more even distribution.

                    But you don’t have to take my word for it. You can google the parameters (ie, Al thermal resistance vs thickness) and do some Physics to figure out the thermal gradient. Even a simple model will get you pretty close to the actual value. It’s tedious, but not that hard.

  9. Get Real says:

    Both systems are good and work fine. Theoretically, the more and smaller cells in the Tesla packs make it easier to cool then the fewer and larger cells in the LG/GM packs.

    1. ClarksonCote says:

      You’ve looked at 1 dimension of a 10 dimensional problem. Your conclusion has not been substantiated at all.

      1. Pushmi-Pullyu says:

        True, that’s forming a conclusion based on one single factor and ignoring everything else. But then, so is this assertion in the article: “Ability to remove high heat loads: GM’s system is slightly better than Tesla’s system in our opinion.”

        1. ClarksonCote says:

          They stated it was their opinion. They were not being overly conclusive by any means so I don’t think George is out of line here.

          And I suspect his humbly stated opinion is based on many other factors as well, considering many more dimensions of data than presented, as well as empirical data on battery degradation. Teslas have had more than Volts, which have had none.

          1. Pushmi-Pullyu says:

            There seems to be a consensus, and I won’t claim it’s “fact” since GM hasn’t confirmed it’s true, but there seems to be a consensus that GM has reserved more of the Volt pack’s capacity than is normal for EVs. That is, GM has chosen to use a lower Depth of Discharge (DoD) for the Volt pack; I think 65% was claimed in one post in the current discussion. Normal DoD for li-ion batteries is 80%, so if that’s true, GM has reserved 15% to allow for capacity loss over time without allowing that loss to impact the car’s range. That is, the assumption is that over time, the Volt releases more of its capacity, increasing DoD, reducing the “slack in the system” to make up for loss of capacity.

            If all of that is true, then the lack of range loss by the Volt isn’t due to a superior battery TMS (Thermal Management System); it’s due to the fact that GM accepts a lower range for a new Volt than they would have to, if they followed industry standard.

            One can certainly applaud GM for being willing to sacrifice electric range for a new Volt in favor of reliability in range over time, at least over the first few years of ownership. (The oldest Volts are now five years old.)

            But GM apparently reserving Volt battery capacity for future use has absolutely nothing to do with the subject of this article.

            1. SparkEV says:

              I doubt they’re reserving only 65% for 2015 SparkEV that has LG batteries and cooling topology similar to Volt (2014 SparkEV was different). It won’t go 82 miles on 12 kWh. We will see in few years how degraded SparkEV will be. That will be closer to apples-to-apples in terms of discharge state.

    2. ModernMarvelFan says:

      Exact the opposite. The larger cells have better surface to volume ratio thus easier to cool.

      The problem is the packaging and cell energy density which favors the smaller package.

  10. jerryd says:

    The Volt is clearly superior cooling but packaging means it is a dead end as can’t fit in a floor battery.
    Plus it is heavier and not flexible.
    That said I just bought a Volt pack for my next EV as I’ll just design around it.

    1. John Hansen says:

      The T-shaped form factor is not relevant to the cooling characteristics of the battery. My understanding is that GM uses the same cooling technique in the Spark, which is not a T-shaped battery. There’s no reason why the same cooling technique couldn’t also be used in a lower and flatter battery like the Tesla’s battery.

      1. Josh says:

        I have heard from some battery experts, that the vertical orientation of the cells makes a big difference in letting the heat travel up in the space between the cells. Both the Model S and Volt have this orientation.

        Many of the LEAF cells do not, on top of not having active thermal management.

        Prismatic cells stuffed under the floor of the vehicle pretty much have to be laid horizontally, putting them at a cooling disadvantage.

        1. Foob says:

          That’s interesting, since Nissan’s prototype 60kwh pack from the IDS has all the cells arranged vertically.

        2. Art Isbell says:

          The photos that I’ve seen suggest that the i3’s prismatic cells are stacked vertically in each battery pack module.

          1. Pushmi-Pullyu says:

            Yeah. If the individual cells are small enough, even pouch cells can be laid in on their side edges in a flat battery pack, so they have a vertical orientation. I don’t know if large format pouch cells can be practically made by making them a lot narrower side-to-side than they are tall, so they could be laid on their side and fit in a battery pack less than 3 inches high. But at worst, they could just use a larger number of smaller cells, closer to Tesla’s engineering.

            1. ModernMarvelFan says:

              Large format cells can be “customized” into any shapes you want, thus the “pouch” design.

              There is no reason why it can’t be made into a 3×7 size. But eventually, you have to make a decision on energy density as well.

              We will see how the Bolt battery is packaged to fit into the floor.

      2. ModernMarvelFan says:

        Spark EV battery aren’t cooled the same way as the Volt.

        The cooling plates are only running on the top and bottom of the pack. It is not sandwiched per cell as in the Volt.

        1. SparkEV says:

          Top/bottom cooling plates were on 2014 SparkEV with A123. 2015 with LG supposedly has same type of cooling as Volt with plates between cells.

    2. pjwood1 says:

      I’d rather my weight be where the batteries were, than sitting on top of them. I don’t think the T-battery is doomed.

      There are a number of Tesla owners well aware of how far down the seats don’t go.

      You can’t do an electric coupe, with a skateboard battery, IMO. Compare the seating position of the i3, against other BMWs, for example.

      1. Art Isbell says:

        There are certainly pros and cons to each battery pack packaging system. I prefer the cabin openness of the skateboard battery pack to the T-battery that makes the Volt’s cabin feel cramped. But for the lowest center of gravity and roofline, a T-battery works better. However, if the Volt’s battery pack were as large as those of BEV’s, would there be enough room using its T-battery packaging? Will the Bolt’s much larger battery pack be a skateboard or T-battery?

        1. Pushmi-Pullyu says:

          The GM Bolt, like the BMW i3, looks like a tall car compared to how long it is. I’m pretty sure that the Bolt, like the i3 and the Model S, has the battery pack in a flat layer under the floor.

      2. Pushmi-Pullyu says:

        Lowering the seats below the level of the battery pack is exactly what causes the hump in the rear of the Volt. That’s hardly a benefit, now is it?

        Moving forward, I think we can be fairly confident that all, or nearly all, plug-in EVs which are designed from the ground up will use the “skateboard” design.

        And speaking of battery cooling, putting the batteries in a flat layer, as in the Model S, prevents a dangerous heat buildup in the center of the pack. There are a lot of reasons why the Volt battery pack’s “T” design isn’t the best approach. If the Volt hadn’t been built with the idea of sharing as many parts with the Cruze as possible, perhaps GM wouldn’t have chosen that awkward design.

      3. ModernMarvelFan says:

        The way GM cools its battery doesn’t have to be a T shape.

        It can be I shape, H shape or even A shape….

        It requires a minimum height that is at least the height of cells plus the packaging.

        But there is no reason why GM can’t can lay them flat or at a 30 degree angle to reduce height.

  11. Stephane says:

    We all know that the LEAF got the best system 😛

    1. Tom says:

      Affirmative…?

    2. SparkEV says:

      Leaf and VW eGolf have same system. They work great as you can see from Leaf’s DCFC speed.

  12. Chris C. says:

    As mentioned in several comments above, this article really would be well served to also tackle the matter of water-based (glycol) vs refrigerant-based (R134/Freon) cooling systems. BMW’s position is that a refrigerant system, besides being more efficient, is SAFER because you’re not taking the risk of exposing the batteries to water (e.g. in crash scenarios). I’ve had an BMW engineer tell me outright that their system was better than Tesla’s for this reason. And another recent article here mentioned that Tesla already does (may?) have a refrigerant-based loop in the Model S, just not for battery cooling per se.

    1. Nix says:

      Yes, the Tesla has an A/C system for cooling the interior. It is an R134a (or similar) compressed coolant system. In theory it could be used for both tasks.

      However, it might be overkill. Let me fabricate/makeup/pull-from-my-backside some numbers as an example. If you only need to remove 500 BTU’s worth of heat from a battery pack, it is actually very hard for a compressor based system rated at 10,000 BTU’s to efficiently supply that small of an amount of cooling.

      The compressor will end up drawing a whole lot of electricity to remove a small number of BTU’s. It is the same idea as sizing your AC system in your house to be the correct size, and not installing a system that is way over-rated.

      1. Priusmaniac says:

        But more cooling capacity is paramouth to enable faster charging at the megawatt level, so BMW is right. The onlymistake is that the need to use the benign carbon dioxide as a coolent not the usuall coolant gases. CO2 is cheap and less problematic for leaks. It is also very interesting for heat recovery from the outside.

        1. Nix says:

          The other way to cut down on heat when charging, is to simply have a bigger battery.

          Then you can add more miles worth of charge per minute to a larger pack, compared with a smaller pack, without eating each cell as much.

    2. Pushmi-Pullyu says:

      Chris C. said:

      “…this article really would be well served to also tackle the matter of water-based (glycol) vs refrigerant-based (R134/Freon) cooling systems. BMW’s position is that a refrigerant system, besides being more efficient, is SAFER… I’ve had an BMW engineer tell me outright that their system was better than Tesla’s…”

      I imagine a Tesla engineer would point out that circulating a glycol-based coolant requires far less energy; all you need to do is run what amounts to a water pump. Using a refrigerant-based system means you need to power a compressor, which takes far more energy.

      It’s easy to say one system is “better” than another, but in a system this complex, that often boils down to individual opinion regarding which characteristics are most important. That’s a subjective opinion, not an objective one.

    3. ModernMarvelFan says:

      I believe refrigerant system has more heat transfer capacity. Thus lower weight for the overall system for a given heat capacity.

      However, the pressurized system will be required for refrigerant system thus cost more and it is also more complex.

      1. Nix says:

        You can’t make assumptions about the comparative weight of a high pressure refrigerant system vs. a much lower pressure coolant based system on volume of coolant/refrigerant alone.

        Hoses and connectors that handle high pressure, high temperature differential refrigerants have to be much stronger than a low pressure coolant pipe.

        1. ModernMarvelFan says:

          That is a good point.

          The high pressure tube and hoses can be heavier overall

  13. jim stack says:

    From all reports it appears the GM system is the best. They have no battery loss in high mile Volt. The GM SPARK EV is great too with no loses of capacity. Also don’t miss the FORD Focus EV cooling and Energi and hybrids from FORD. All we have measured has no loss at all after 3 years in the HOT Phoenix Sun.

    Now to be complete Telsa does have much higher power and heat generating Super Charging that others don’t have.

    So they are all great and are very sustainable. Last longer than most will ever need and still can reuse the batteries no lower stress solar and wind back up.

  14. Brian says:

    Exactly why I love my volt. Battery is well protected. Have met so many LEAF owners that have battery issues. What was Nissan thinking? Did they believe fast charging using chademo and not having any way of drawing away the heat caused was a good idea?

  15. John Hansen says:

    GM’s conservative approach will eventually win. Many early adopters were willing to accept some battery loss, but that won’t be accepted by the general public. Cars should run until 100k miles without any appreciable loss of power or range, and then only minimal loss up to 150k. After 150k most cars are trashed and have lost all their resale value, so it doesn’t matter as much past that. GM has accomplished that with the Volt. Tesla almost accomplished that with the Model S (probably close enough). The Leaf has absolutely not accomplished that, which is reflected in give-away used prices. Ultimately, everyone will move in the direction of GM’s very conservative (but unexciting) approach.

    1. Josh says:

      The biggest advantage in the Volt is the 65% DoD. There is a ton of margin in the battery. Who knows over time as degradation sets in, they may open the DoD to keep the available KwHs up.

      1. Art Isbell says:

        The Volt’s 65% DoD is also a disadvantage because of the reduced usable capacity that wouldn’t work well in a BEV and that adds to the Volt’s porky weight.

        1. Pushmi-Pullyu says:

          Actually, I think John Hansen will prove to be right in the long run. Sure, the Volt takes a hit in electric range when it’s new by using only 65% DoD (assuming that’s correct), instead of the industry standard 80% for li-ion batteries. (Presumably the Volt is engineered to let the DoD creep up to 80-85% as the car ages, to counter loss of battery capacity.) But that will let GM brag “Our plug-in electric cars don’t lose range after a few years, as others do.” That’s a pretty strong selling point, and should strongly affect the car’s resale value.

          The fact that this requires a heavier battery pack is a rather minor issue. The Tesla Model S did quite well even in 2012, when its battery pack weighed approx. 1250 pounds. As energy density continues to advance year after year, this will become less and less an issue.

    2. SparkEV says:

      It’s hard to say if everyone will move to GM design. There’s something extremely compelling about Leaf/eGolf: no active cooling, lower cost. Given that Leaf is second highest selling EV, I doubt consumers will care about battery issue.

      What they will care about is how slowly the non cooled battery will charge using DCFC. But if they don’t use DCFC, they won’t know or care.

  16. Great insights into the inner-workings of using fluids to cool batteries. However, in the end this article is just an “opinion” article lacking scientific measurements of actual cooling/heating testing. Of note, the rate of flow of fluid in the system and size/type of radiator used are also significant factors, as is the mass/density of individual cell components.

    Re: Just an opinion.
    “GM’s system is slightly better than Tesla’s system in our opinion.”

    1. M Hovis says:

      Though all of your points are accurate Brian, I would rate this article a “9” on a scale from 1-to-10. What George has successfully accomplished is to raise the overall understanding in the EV community on the fundamentals involved in thermal management. All the suggested scientific data, while being even more informative, would still likely not provide a conclusive winner, which I am reasonably sure was not George Bower’s goal. People take for granted that everything about Tesla is wonderful. What George has done, is to up the awareness of GM’s accomplishments, while simultaneously educating. Well done George Bower! Maybe your best work yet.

      1. ModernMarvelFan says:

        I actually agree with George Bower’s opinion.

        GM design is better (regardless other issues which Tesla is superior in other area) in term of cooling because of the following reasons.

        1. GM has far more cell surface contact than Tesla regardless of surface to volume ratio. GM design has aluminium (great heat conductor) in direct contact in most of its cell surfaces which would reduce “thermal junction resistance” the most. Tesla only has about 1/3 the surface in contact with cooling surface (copper which is even better than aluminium) but the rest of the cells aren’t in contact which is subject to uneven cooling. Yes, gels help, but it is not nearly as good as Al or Cu.

        2. Surface to Volume ratio is better on the pouch cells used in the Volt.

        3. Volt sort has to have a more powerful or more “evenly” distributed system. Its cells requires more cooling since it discharges at higher rate. So, the parallel design is required as it can’t afford a greater variance in cell temperature due to larger format.
        In a way, Tesla is superior here since it has far more cells so the “cheaper” serial design is sufficient since each cell is smaller.

        Overall, I think the GM design is more robust in cooling. But Tesla system is very smart for what it needs.

        From an overall system cooling for the purpose of the cell goes, they are about the same. For a specific cell cooling design, GM is more robust, thus better.

        From an overall battery and car design and packaging point of view, Tesla is better.

        So, engineering judgement is often complex and it depends on the need of overall system. Tesla system fits that need very well, better than GM in overall packaging. But GM’s individual cell thermal design is superior in my engineering opinion.

  17. It would be interesting to compare the fluid cooling system of the Tesla and Volt with the forced cooler-air systems used in Kia Soul EV and Mitchfishi iMieV.

    There are clearly some advantages/disadvantages like cost and weight. Does battery pack size favor one thermal management system over the other?

    1. Pushmi-Pullyu says:

      Brian_Henderson asked:

      “Does battery pack size favor one thermal management system over the other?”

      Simple physics, and the cube-square law, mandates that all other things being equal, larger battery packs will have more heat buildup at the center, and thus will need a more robust thermal management system.

      But other than the obvious fact that passive cooling (as with the Leaf) becomes less effective with larger battery packs, I don’t think we can come to any conclusions about which cooling system is “best”. Clearly circulating a coolant will provide superior cooling. Again, that’s basic physics. Heat transfer works best (fastest) by conduction, worst (slowest) by radiation… and passive cooling depends entirely on radiation.

      Both GM’s system and Tesla’s system circulate fluid for temperature stabilization. Aside from that, I think the most important factors would be the flow rate of the coolant, and just how much coolant is present… the thermal mass of the coolant itself.

      The geometric configuration is also important. Obviously the size and shape of Tesla’s small cylindrical cells — the cylindrical shape naturally forcing space between the cells — have an advantage over GM’s larger flat cells, which are apparently sandwiched back-to-back with only a thin cooling plate inserted between.

      So Tesla’s system has a geometry more favorable to cooling, with the tradeoff of the entire pack being larger per kWh. But GM could at least partially compensate for its more compact battery pack shape by circulating the coolant at a faster rate, altho that would take more energy.

      1. ClarksonCote says:

        “The geometric configuration is also important. Obviously the size and shape of Tesla’s small cylindrical cells — the cylindrical shape naturally forcing space between the cells — have an advantage over GM’s larger flat cells, which are apparently sandwiched back-to-back with only a thin cooling plate inserted between.”

        I still am very skeptical of these assertions. A Manila envelope has higher surface to volume ratio than a cylinder. And it would appear Tessa’s cooling method has much less uniform contact with the cells than GM’s cooling method.

        Because of these reasons, I’m very suspect of your conclusions quoted here, still seems like far too many details being assumed.

      2. ModernMarvelFan says:

        “Simple physics, and the cube-square law, mandates that all other things being equal, larger battery packs will have more heat buildup at the center, and thus will need a more robust thermal management system.”

        Why are you making those statements without doing some math?

        The ratio favors the large and thin pouch cells. Also, GM uses aluminum plates to distribute the temperature so there is very little gradient across the cell.

        The 18650 cells are much thicker, so there is actually a higher gradient between the center and the surface.

        Of course, all of these all depends on the type of chemistry and discharging rate which impact the amount of heat generation.

        when everything are equal, the GM system has far lower temperature gradient than the Tesla system.

        But that is probably because Volt requires it since its discharging rate is higher in C.

        Lastly, don’t forget that GM is cooling the entire large surface where Tesla is only cooling less than 1/3 of the surface.

        1. Pushmi-Pullyu says:

          MMF: Please read what you’re responding to more closely. I very clearly wrote “larger battery packs”, yet you responded as though I had written “larger battery cells”.

          I also wrote “all other things being equal”, yet you completely ignored that, too.

        2. Pushmi-Pullyu says:

          “Lastly, don’t forget that GM is cooling the entire large [cell] surface…”

          No, at best GM is cooling only one side of the pouch cell, less than 50% of the surface. Each pair of cells is fixed in a frame with an insulating layer between; obviously there’s not a cooling plate between the two. So the configuration is |cell|cell|plate|cell|cell|plate|, not |cell|plate|cell|plate|

          And again, you’re ignoring the volume of flow of the coolant. GM’s cooling plates appear to have rather tiny tubes in them; the Telsa cooling system appears that it might allow a much greater flow of coolant.

          Rather ironic that you say “when everything are equal”, since nearly everything is rather different when you compare the two systems!

          1. ModernMarvelFan says:

            “No, at best GM is cooling only one side of the pouch cell, less than 50% of the surface.”

            Less than 50% but very close to 50% of the total surface where the Tesla design would have less than 1/3 of the surface. Which one has more coverage factor?

            “And again, you’re ignoring the volume of flow of the coolant. GM’s cooling plates appear to have rather tiny tubes in them; the Telsa cooling system appears that it might allow a much greater flow of coolant.”

            Does it? Do you know that for sure? What is the flow rate of the Tesla system vs. Volt system? The “tiny tubes” are 5 in parallel and per plate where the Tesla system is made of “thin ribbon” which has to connect in series across over 8,000 cells…

            “Rather ironic that you say “when everything are equal”, since nearly everything is rather different when you compare the two systems!”

            Well, it is harder to compare if you don’t have all the details.

            We know that Volt system requires a high rated cooling system due to its C discharging rate which is higher than Tesla, thus the more close contact of the cells to cooling plate.

  18. Bill Howland says:

    GM definitely wins in the efficiency game, since in moderate weather the system simply fan cools the anti-freeze. The refrigeration system is only used when it is absolutely needed.

    In very cold weather GM is also clearly superior with its ‘buttoned up from the elements’ battery location, needing little extra heat to keep warm while being charged. In fact, the 3.3 kw charging rate is usually enough to keep the batteries warm enough to charge, using the resistance of the batteries themselves to maintain temperature while charging.

    My roadster was quite inefficient while charging, with losses coming from running the airconditioner to keep the battery cool, as well as running the cooling fan to keep the inverter temperature at bay.

    The only time the air conditioner DID NOT run was in the very coldest weather, when charging at 6.3 kw. Of course in that weather, most of the time was spent heating the battery with the 1 kw heater, before you could even begin to THINK about charging the battery.

    From everything I’ve read about the “S”, it is even more inefficient, seeing as Broder lost 1840 watts of energy over the duration of his hotel stay in 10 degree F weather.

    1. ClarksonCote says:

      1.8kWh of loss?? Ugh, it pains me that all the inefficiency of the Teslas while parked are not reflected anywhere in the MPGe numbers.

    2. Pushmi-Pullyu says:

      Bill Howland said:

      “From everything I’ve read about the ‘S’, it is even more inefficient, seeing as Broder lost 1840 watts of energy over the duration of his hotel stay in 10 degree F weather.”

      So… your argument is that the Model S battery pack’s refrigeration system is less efficient than the Roadster’s, and you offer as evidence the fact that it loses a lot of energy running the battery heater in very cold weather.

      Riiiiight… LOL!

      Or maybe, just maybe, the Model S uses more energy to run its battery heater because it has a larger, 85 kWh battery pack vs. the Roadster’s 52.8 kWh battery pack, and a larger thermal mass needs more energy to maintain a stable temperature.

      But no, that’s too obvious. [/snark]

      1. Bill Howland says:

        Go ahead PUPU , laugh it up. But first tell me, what the quiescent loss of the Roadster in 10 degree F weather is.

        You have no idea what it is? I thought not.

        Hint: Of the 3 cars under identical conditions, the volt proportionally is the best (assuming we normalize the battery to each 10 kwh section of battery,of battery loss) – that levels the playing field.

        1. Pushmi-Pullyu says:

          The fallacy here, Bill — or rather, one of your fallacies — is your assumption that what applies to a battery pack TMS’s ability to heat applies equally to its ability to cool.

          On the contrary, with the Volt battery pack’s more compact, more densely packed design as compared to the Model S’s pack, I would assume the Volt pack holds heat better… with the tradeoff that it has to use more energy per cubic foot to cool the battery pack. But if it holds heat better, then obviously it doesn’t need to spend as much energy heating the battery.

          1. Bill Howland says:

            Your insulting response doesn’t deserve another one. There are 2 separate issues here, – Argue the point with MMF, since he is one of the few conversant with heat sinking.

            Priusmaniac’s desire for CO2 refrigeration absolutely with not work since these systems require a COLD WATER heat sink.

            CO2 *CANNOT* be condensed to a liquid at *ANY* pressure if 86 degrees fahrenheit or higher – the so called ‘critical temperature’.

            PUPU, you are getting a ‘basic physics’ lesson here, as opposed to the drivel you spout with almost every post.

            I’ve worked on CO2 refrigeration systems that were cooled with city tap water. Even with this cold heat sinking, the “Discharge” pressures were around 1,000 PSI, and the compressor “Suction” pressures were around 400 PSI. Typical efficiency was also HORRIBLE, usually around 2 1/2 HP per ton of refrigeration effect.

            Now you’ll probably tell me that CO2 can’t form a liquid. No, it cannot at 14.7 PSIA, since it sublimes. But that is an irrelevancy here.

            1. Bill Howland says:

              Now, in case you are thinking of using SuperCritical CO2, most popular commercial systems, such as DENSO, run at discharge pressures of around 1,400 PSI (that’s 10 MPA), and 212 deg F (100C).

              While the efficiencies are reasonable, the materials-science issues in a battery are somewhat daunting. Plastic piping ain’t gonna cut it.

  19. JakeY says:

    For the GM example, do they have a plate for every cell or every other cell or even more? The diagrams don’t make it clear.

    1. ModernMarvelFan says:

      Every cell, thus sandwiched on both sides..

    2. Pushmi-Pullyu says:

      Quoting from “Chevy Volt battery pack: Rugged but precise”:

      “The battery features 288 individual battery cells housed in a lightweight glass fiber-reinforced thermoplastic battery pack that incorporates an integral thermal management system. Enclosing the cells are 135 repeating frames, each approximately 250 mm wide and tall by 15 mm thick (9.8 inches square by 0.60 inch thick), and 18 similarly sized end frames (see diagram). Each repeating frame holds two lithium-ion cells, one on each side, separated by a layer of insulating foam.”

      So clearly there can’t be an aluminum cooling plate between each cell and the next. At most, there is one between each frame and the next… which is to say, between each set of two cells.

      http://www.compositesworld.com/articles/chevy-volt-battery-pack-rugged-but-precise

      1. Pushmi-Pullyu says:

        Edit/correction:

        Or maybe each frame has a cooling plate sandwiched between the pair of cells, as the article linked below states. That would actually make more sense, from an engineering perspective, and it would again indicate one cooling plate per two cells, not one per cell as ModernMarvelFan keeps claiming.

        http://www.torquenews.com/119/volt-battery-uses-dana-corp-fins-maintain-temperature

  20. heisenberght says:

    Once again I use a battery-thread to go slightly off-topic 😉

    This time it’s Rice University who use laser-induced-graphene for supercapacitors…

    IMO this laser-graphene stuff is really cool, cause it allows for homemade graphene. Don’t believe it?

    “http://spectrum.ieee.org/nanoclast/semiconductors/materials/laserinduced-graphene-looks-to-diplace-batteries-with-supercapacitors”

    All that graphene is expensive BS will soon be worthless words! accelerate progress!

    Why this is only slightly off-topic? Well I dislike hybrid and like hybrid 😉

  21. heisenberght says:

    An just another tidbit:

    It’s bout Na-Ion batt. I like that they don’t exaggerate too much 😉

    “http://spectrum.ieee.org/energywise/energy/renewables/a-first-prototype-of-a-sodiumion-rechargeable-battery”

    1. Nix says:

      For both battery ideas — Wake me up once they’ve proven themselves in smaller scale applications (like consumer electronics) in the real world.

      Until then, the sleepy boring world of 14% yearly price per kWh improvement in commodity lithium-ion cells is actually where the action is at. That is halving costs roughly every half a decade, with no theoretical wall for a couple of decades.

      It sounds boring, but in the real world it is a statistic with a decade and a half of proven record behind it, and the cells are ready and in cars today. It isn’t glamorous, but at least we know it works.

  22. Pushmi-Pullyu says:

    Hey, George Bower: Thank you for this article! Very informative, there is lots of info there I didn’t know.

    BTW — What does “sandshot” mean, as in “A sandshot from the patent app…”

    Was that perhaps supposed to be “screenshot”?

  23. Someone out there says:

    I’ve been thinking of another idea: The 18650 cells are made by rolling a couple of sheets of materials into a cylinder. The center of this cylinder doesn’t contain much material so why not just make a hole through the cell and have the cooling fluid run right through the battery?

  24. Bob says:

    I like bigger cells, they are easy to exchange.

  25. DNAinaGoodWay says:

    So, how much range is lost when TMS is used? Is it recommended that TMS cars stay plugged in when parked in extreme heat or cold to mitigate losses from TMS?

  26. tom says:

    It doesn’t matter which has higher heat removal. All that matters is if each is adequate. So lets see…150kW with ~90% fast charger efficiency is about 135kW into the pack, divided by about 400V pack V for Tesla is about 338A, or about 1.4C for the ~240Ah modules. The cells are 3.3Ah, so that’s about 4.6A/cell. Assuming a ballpark guess of 2mOhm cell resistance then gives about 0.042W per cell, and with ~6800 cells around 288W dissipation for the pack. Doesn’t take much to remove that. I suspect they are more concerned with heating the cells in winter than with cooling them. That’s certainly been more important for the pack in my converted car with LiFePO4 cells. They have no cooling whatsoever and have done fine over the last 6 years, but heating them in winter is very necessary to prolong cycle life. I’ve kept them heated to ~65F, cooling to at most ~50F when parked outside for several hours, and they have done fine in winter. This is accomplished with an array of resistive heaters with total power of 350W. They kept the pack at its set point at -5F ambient. Passenger heating uses far more energy.

    1. Bill Howland says:

      Tom, you are saying 0.002 ohms per Tesla S cell.

      Do you have the specific equivalent series resistance:

      1). When almost fully charged, and charging at 1.4C?

      2). When almost fully charged, and Discharging at 1.4C?

      3). When at 10% State of Charge, and charging at 1.4C??

      4). When at 10% SOC, and Discharging at 1.4C?

      Thanks in advance.

    2. Bill Howland says:

      I guess I can see why you haven’t answered my questions: apparently the answers aren’t being published for some reason.

      I DID find on ‘industrialpanasonic’ info on the NCR18650B, which, in bullet points say <= 45 milliohms, and then in the chart says equivalent series resistance: 0.11 (or 110 milliobms).

      So that would make your '288 watt dissipation' either 6480 watts per pack, or 15,840 watts, depending on the particular spec used.

      Since I don't own a model S, I don't know how much heat comes from the car when supercharging… Is is a little or a lot?

  27. ampzilla says:

    wow lots of input. how about the design of my 2013 coda tms. just the reason i bought it for. closed loop system that cools, heats dehumidifies my old school super durable LI FOSFATE 4 batterries. this is a 300,000 mile battery design. im going to prove it. this is ny so im stuck with level 2 charging yet it works for me.

  28. RFS says:

    Sorry if this is a bit off track , however something that really appeals to me about a standard off the shelf cylindrical battery is the ability to keep the car on the road well into the future , anyone who likes driving will always like something that goes 0-100k in 3 seconds. So even if a more efficient battery does exist it may not be easy to find such a battery in the future , a classic marketing strategy for gadgets , thus Tesla have probably created a classic while others may have created a throw away gadget.