Fail Of The Week: Taking Apart A Tesla Battery

It takes a lot of energy to push a car-sized object a few hundred miles. Either a few gallons of gasoline or several thousand lithium batteries will get the job done. That’s certainly a lot of batteries, and a lot more potential to be unlocked for their use than hurling chunks of metal around on wheels. If you have an idea for how to better use those batteries for something else, that’s certainly an option, although it’s not always quite as easy as it seems.

In this video, [Kerry] at [EVEngineering] has acquired a Tesla Model 3 battery pack and begins to take it apart. Unlike other Tesla batteries, and even more unlike Leaf or Prius packs, the Model 3 battery is extremely difficult to work with. As a manufacturing cost savings measure, it seems that Tesla found out that gluing the individual cells together would be less expensive compared to other methods where the cells are more modular and serviceable. That means that to remove the individual cells without damaging them, several layers of glue and plastic have to be removed before you can start hammering the cells out with a PEX wedge and a hammer. This method tends to be extremely time consuming.

If you just happen to have a Model 3 battery lying around, [Kerry] notes that it is possible to reuse the cells if you have the time, but doesn’t recommend it unless you really need the energy density found in these 21700 cells. Apparently they are not easy to find outside of Model 3 packs, and either way, it seems as though using a battery from a Nissan Leaf might be a whole lot easier anyway.

86 thoughts on “Fail Of The Week: Taking Apart A Tesla Battery

      1. Would think they’d drain those batteries first. Not just to the low shutoff level but completely dead drained. Otherwise crushing or chopping the batteries could start massive fire.

        Maybe a custom press similar to tomato dicer that can be used to instantly cut through glue and pop apart individual cells for a safe deep discharge then recycling.

          1. water and lithium dont mix very will.

            there are youtube videos of lithium in water.

            even the water of the sweat hitting lithium can cause problems even a meth cook was burned.


            video may be disturbing so watch carefully

            even electronics recycling centers are leary about taking devices with batteries in them even devices as small as apple’s air pods witch contains a tiny lithium battery of only a couple watt hours

            however once the lithium has been safely deactivated then water may be used to keep the dust down

        1. I work in a nickel refinery, and some of iur source material is now Li-Ion batteries. I’m sure there are several methods, but for our part the batteries just go in the smelter, a large electrode oven. Cobalt, nickel and copper come out in the smelt, plastics and carbon come out in the slag.

          For this method discharging doesn’t really matter.

        2. One recycling process freezes the cells with liquid nitrogen for two reasons: it makes the metals and the liquid electrolyte and plastics inside the battery brittle so they crush instead of mashing into a paste, and the extreme cold and pure nitrogen atmosphere prevents runaway chemical reactions inside the crusher.

          The resulting powder is self-igniting once it warms up, but at this point that no longer matters because the mixture is sent to an incinerator which burns off the volatiles anyways. The remaining chemical energy simply burns off.

          1. The main achilles heel of lithium battery recycling is the high energy consumption of the process. It takes about 5-10x the energy to reclaim the metals compared to using virgin materials, which pushes the ESOEI (Energy stored per energy invested) of recycled lithium batteries down somewhere on par with lead acid batteries.

            In the worst case, you spend just as much energy to rebuild the lithium battery as the amount of energy the battery can store in its lifetime. That means the recycled battery has a total energy efficiency of just 50%

    1. It all looks extremely difficult to separate those batteries from the glue. I have an idea. Source a piece of thin-wall stainless steel tubing, that is a ‘sliding fit’ over the battery diameter. Sharpen the ‘business end’ of the stainless tube. Attach the sharpened piece of stainless – steel tube to a heat source (such as an Industrial Soldering Iron). Allow the stainless tube to heat – up, and simply place this tube directly over the battery that you would like to remove. When firmly pressing down, the heated (180C) stainless tube should easily cut through all of the layers of bonding, simply because you are supplying an incredible amount of pressure at the sharpened-end of the stainless tube, that will cut through almost all hot-melt adhesives, and even epoxy resins at the thin bond-layer between each battery. I have never tried this method, but I would not be beaten by something that looks ridiculously simple to achieve. Hot wire will not cut it, and you are only attempting to break the bond down one side of the adhered batteries. The heated tube method would cut through all bonds, around the outer circumference of each battery, freeing-up the battery in 5 seconds.

      1. Not really. The salt flats are close to pure lithium carbonate/hydroxide whereas the batteries are almost anything but. There’s only a couple kilograms in a 500 kg battery, and it’s dissolved or compounded with other stuff, so it’s not trivially separated. You have to crush the battery, then run it through various chemical baths to separate all the different compounds, then separate the lithium separately from each of them.

        It’s a very energy intensive process that, if done for the lithium alone, takes 10 times as much energy as the mined lithium.

  1. I bet there is some solvent that will strip that glue, but I’d not be surprised if it was nasty enough stuff that it would only be practical to use in an industrial setting with robots handling the task of dunking whole sleds of cells at once so no human has to endure the vicious fumes / skin melting / transdermal absorbtion toxicity / whatever other flavor of nasty that particular solvent embodies…

    Like many manufacturing cost reductions the very same thing that makes it cheaper and faster for a robot to do it likely makes repair, modification, and manual disassembly a time consuming pain in the ass. Like sub-millimeter 256+ ball BGAs, self-aligning snap-fit enclosures, and of course those phones whose metal shells are glued onto either side of the board with thermal potting epoxy, this is one more manufacturing optimization which costs us tinkerers all the time it saves the manufacturer. (On the other hand, they build (and eventually recycle) orders of magnitude more units than hackers repurpose.

    It may also be the case that the glue is actually an important heat conducting component in this assembly serving a thermal as well as mechanical role.

    1. YES!!!
      I imagine the same scenario of specialized factories recycling cells by the million. Fast and efficiently.

      I am sure there is a plan.
      And we are not going to know about it up until all trademarks, patents and gigantic plants are in operation.

      Then it will be OK for hobbyist.

      Actually it does make sense.
      Whilst Europeans rejoiced using shiny screws, washers, bolts and symmetrical nuts, good old “America” was pumping ahead using riveting everywhere.
      History repeating itself ??????

      Didn’t Apple started all that glue’ing madness?

      1. Ugh. Europeans and their screws! Sorry Europe but I will never forgive you for your EU mandates that now mean I need to go get a screwdriver just to replace a gd battery in pretty much any commercially produced device!

      2. “I am sure there is a plan.”

        The willingness we (collectively) have to assume that companies operate a certain way because somebody smarter knows something we don’t is baffling.

        Even if the problems are obvious, it’s often too much to assume somebody considered them.

        Even if somebody considered the problems, it’s often too much to assume a solution or mitigation fit their motivation.

    2. There’s a far simpler solution to the problem, which all the other EV manufacturers are using: prismatic (pouch) cells instead of cylindricals.

      Tesla started by leveraging Panasonic’s existing manufacturing technology to produce cells which are aimed at laptops and power tools. They went for “available now” instead of “optimized for EVs” and this is the price they’re paying.

        1. Tesla? Expensive assembly and disassembly, larger weight for all those steel canisters, higher chance of faults due to thousands of tab welds, loss of usable capacity due to the varying capacity of the individual cells, cost of testing and binning the individual cells according to capacity…

          There’s a whole bunch of problems in using thousands of laptop cells for a very large battery, but Tesla went that way because it’s what they had available in 2009 – that’s what they could already buy without spending money and time in developing a new battery type and a production process for it.

          That’s what gave them the head-start on the market, along with other compromises. Now they have all their money invested in production lines optimized for this type of battery, and if they were to change they’d have to take a huge leap backwards in cost and production capacity. They’d have to play catch-up with all the other EV manufacturers who now have prismatic cells of their own, who are optimizing production and cutting costs down with expanding production volumes, and Tesla doesn’t.

          Tesla is stuck with the worse batteries because Musk took the businessman approach to engineering. He even calls it his “first principles method”, which means that if you have A and B then you should be able to do A plus B and it just works. In this case, “batteries” plus “cars” equals “electric cars”. Of course the practical reality is different, but you can make long ways of superficial progress that way by pretending that all the real engineering issues don’t exist.

          1. Oh, more disadvantages:

            Non-optimal packing of the cells in a module, because cylinders leave so much empty space between them. Even contact with cooling channels is made more difficult, leaving cells warm/cold on one side more than the other. Uneven heat distribution within the cells because the battery is a long rolled sheet where one end is left in the middle; causes current/voltage differences within the cell when the temperature changes.

            About the only advantage of cylindrical cells is that they’re by nature compartmentalized into small cells, which retards the spread of catastrophic failures. However, this is only an advantage if you are using a battery chemistry type which is prone to catastrophic failure. You shouldn’t.

      1. Uh, cylindrical currently is easier to manufacture at bulk/high speed. That’s the reason. Tesla’s battery cost is WELL under basically everyone.

        If ever you want to think Tesla’s done something short sighted on the manufacturing side maybe you need to step back and figure out what you’re missing, because that’s a lot more likely.

        As pouch manufacturing processes improve that’ll change (Tesla’s made a battery advancement announcement implying that) but for now cylindrical is better.

        1. Prismatics are fundamentally simpler in construction and simpler to make. Other EV manufacturers already have it pretty much down to cost, but there are other considerations and compromises that Tesla is making, that is keeping their costs below the competition. For example, Tesla is using a cell chemistry that gets them higher energy density and higher power for a lower price, at the expense of worse fire safety.

          Tesla tried to shoehorn the same NMC chemistry that Nissan and others are using, but since their batteries are twice in capacity than the competition, the higher cost turned out prohibitive, so the Model 3 is using the same NCA cells as the Model S – with the same problems in crash and fire safety. (Needs more armoring, cooling…)

          “Tesla’s done something short sighted on the manufacturing side”

          They have, but you have to mind that they started doing this 10 years ago when large format prismatic cells were simply not available. Nobody was making them. Tesla went in and invested billions in cylindricals because they needed the cells, and now they have these huge sunk costs in a technology that is falling behind the curve. It was either that, or no Tesla.

          Tesla cast their own cement boots by rushing to the market without doing any fundamental product development: they just bought everything off the market, which is patently a bad idea if you think about it: If everything is already available, then why isn’t anyone else doing it?

          Rule #1 of business: just because someone’s spending money doesn’t mean they know what they’re doing. People warned about these problems back in 2009 when Tesla announced the Roadster, but all the fanboys just went “YAAAY Electric sports car!”. Cue in meme of Philip J. Fry.

          1. Though you have to remember just how they sold the idea to investors back then. Musk basically said, Okay, we’ll start with this toy car here made out of parts cobbled together from wherever, just to show you that it can be done, and then we’ll get enough experience and expertise to build a proper EV that is cheaper and actually useful. There was a three step plan where each stage funds the next stage, ending up with the electric people’s car.

            The Model S was originally supposed to cost $50,000 for a 300 mile EV. Didn’t happen. The goalposts started slipping right off the bat, and Tesla didn’t actually show any signs of solving any of the fundamental problems uncovered by the Roadster, or later the Model S, but by that time there was enough hype and enough sunk cost that people just let it slide.

  2. “Apparently they are not easy to find outside of Model 3 packs”
    A quick google gave me a full page of different local shops that sell them. All of them vape-shops for some reason.

  3. My desire to own a Tesla just dropped a notch. Save gas and the environment by going electric but suffer environmental pollution from all the post-processing required to recycle the packs, not to mention all the non-biodegradable plastic waste.

        1. At the end of the day, it’s all about making money at the excuse of saving the environment.

          But you have to remember that cost = energy. All the economic activity needed to pay for a technology means consumption of energy. Energy is the true source of all the real value in the economy that all the money represents. If it costs more than what you had before, you either have to reduce energy use elsewhere (lower living standards), or you end up spending more than you gained trying to do both things.

          That’s why things like solar subsidies are really not about saving the environment. They’re about taking money from pocket A (the public) and putting it in pocket B (the industry) while the economy suffers the added cost. If it was about saving the environment, the subsidies would go into R&D to push costs down to parity with other means of power generation, instead of paying people to install more solar panels bought from China.

  4. I am a chemist, and methylene chloride(MC) in the pure liquid state should work to penetrate the blue plastic and make it very fragile and easy to remove. Only apply this MC to the lower portion – keep it away from the positive electrode as it will also penetrate and expand the plastic seal. You can buy a gallon from a building supply place.Avoid the fumes and avoid skin contact. It is not flammable or corrosive but it will extract all the lipids (oils) in your skin and make it crack. See if you can buy MC proof gloves. I could not tell if the cell are all side by side with hte + to one side and the – to the other or if they are alternate + and – on each side. If all the metal cans are down, you could place it in bath of MC and let it break down the plastic

    1. If you look at the pack, preventing the MC from getting on top of the cells would be quite difficult…also shouldn’t MC exposure be completely avoided, as it breaks down into carbon monoxide inside the body?

      Personally I’d first and foremost discharge the pack as much as it can be without damaging the cells and then slowly heat it up in something like a paint drying box to prevent hotspots, 50-60°C should still be safe. It would also slightly soften the thermoplastics.
      Or if that turns out to make the plastics gooey and actually more difficult to remove, freeze the thing. -20°C should not do anything bad to the cells and the plastic will become brittle.

      1. Yes, the seal is teflon, but MC does enter it to a small degree and thus swells the polymer with possible entry of the MC? MC is not that toxic with care, but care mu be taken to avoid ingestion. MC is the main in gredient in some paint strippers. This page tells about it.
        Your cooling suggestion might work if the plastic gets brittle, if it does not embrittle….. It looks as if they have selected the potting materials to make access hard – that is what the highly evolved potting industry is all about.

  5. My take – what people think of as “glue” is actually an important constituent for heat removal through a phase-change material. This is what people refer in the industry as a “heat capacitor”.

    Some of the methods of (temporary) heat storage with the highest heat storage density are traditionally implemented using a wax-like solid which turns into liquid and, in the process, is able to absorb a tremendous amount of heat.

    Heat capacitors are traditionally used in aerospace missions where you need to briefly store (and, subsequenty, release) a large amount of heat using a small quantity of material

      1. It’s function is also to serve as a flame retardant that would prevent a chain reaction throughout the module should a single cell catastrophically fail. That’s a large reason for the design choice of this material.

        1. Lithium batteries don’t burn like wood burns – one cell heats another cell to the point where there’s an internal chemical breakdown that produces more heat. You can’t really douse it with a flame retardant chemical – unless that chemical can somehow absorb large amounts of heat without rising in temperature.

          1. ” intumescent foam-packed modules”

            That says it all. The foam expands on the application of heat, insulating the intact cells and pushing the failing cells out of the fault area.

    1. ok, at what level do we consider something serviceable?

      you can easily remove and replace the pack, but fuck no that’s not good enough we need to replace EACH INDIVIDUAL CELL in a pack of THOUSANDS or its “not serviceable”

      1. The smallest replicable unit in the 3’s battery is a module which makes up one quarter of the entire battery.
        The Bolt comes apart into eight modules and individual cells can be replaced the Prius pack comes apart into small low cost modules as does the Nissan Leaf.
        Those companies did a better job on their pack design both from a serviceability stand point and a recycling standpoint.
        Repairiability and a long service life is very important for an EV or hybrid to actually be green as it can take up to 12 years for it’s fuel savings to amortize enough to offset the higher emissions cost of it’s manufacture.

      2. If the item needing replaced costs about the same as the depreciated cost of the product…like when your five year old phone battery costs more to replace than a new phone with similar spec…or when it’s cheaper to buy a new printer than buy replacement cartridges for the old one…. then maybe it’s not ‘serviceable’.
        Reference used Tesla with bad battery depreciation.

  6. These 20700 cells are actually fairly easy to come by, SONY and SANYO make great 20700 batteries (check out any vape store) it’s just that the Tesla 20700 cells are a whole bunch better than your average 20700 which is why they are desirable.

    1. TIL Sanyo began as an offshoot of Panasonic in the 40’s. Originally led by the brother in law of Mr. Matsushita the founder of Panasonic. In the 2010’s Panasonic bought Sanyo for the battery tech.

  7. Nobody seems to be least bit concerned about a possible thermal runaway event… dinky little fire extinguishers will do jack shit against an angry pack. You need a literal firehose to contain it or a way to throw the whole thing outside the buiding.

    1. Sometimes I wonder if even an ICE hybrid such as a Prius greener.
      Sure the Prius is kinda lame but it is reliable and when they do break it is not that difficult or expensive to repair them.
      Toyota definitely doesn’t glue the batteries together.

      1. A plug-in hybrid with 10-20 kWh pack and a range extender would deal about 80% of common driving on electricity alone, and the other 20% can be done on biofuels.

        If you must. The main issue is that these plug-in hybrids consume more fuel than necessary because they have to run the engine occasionally to keep it operational. Especially in colder climates, they usually start the engine just for the heat.

  8. Another idea worth trying is to pull really hard on the aluminum bus bar ( @ 8:24 )
    This probably needs a small winch, but the forse is put on the Alu Bus bar instead of individual cells.
    With a bit of luck this pulls of the bus bar with a row of cells, and then by bending the bus bar the other way, more of the glue gets damaged and the cells may be easy to pull off.

  9. I suppose thats one way to stop the competitors getting at your tech.

    Incidentally its also possible that these 20700s are not standard ones as others have mentioned: even a “EOL” one might be better than something you can buy as a consumer.
    If so then the casing might be different, or even be something new like a Li-Al alloy (Cough NASA /Cough) or some other composite. I did once encounter plastic NiMH 9V batteries before.

  10. is there anyone breaking the modules down into smaller modules for ebike use?

    i have determined that one long module is 25 bricks of 42 cells in parallel so it is 25 series of 1s42p bricks.

    a 48 volt ebike would need 14 bricks witch is 36 inches long

    while 36 inches is too long and would stick out the back of an ebike it is better to make in 2 blocks of 7 bricks seriesed

    2 blocks would be only 18 inches long witch would be better manageable.

    there is a 2 part video on the batteries

    The Truth About Tesla Model 3 Batteries Part 1 and The Truth About Tesla Model 3 Batteries Part 2

    the bond wire that jumps from each cell to the bus plate acts like a fuse so if a cell goes bad it isolates it’s self from the rest of the pack preventing a fire though the range would be reduced.

  11. Smeltingis a heat based process whereby the source material is placed in a large crucible. Lithium packs will be reduced by grinding after the freezing someone else mentioned. Processes vary, but usually a salt is added as a flux, read on from here in the wiki.
    Since lithium when hot explodes with air, they might use an inert atmosphere – nitrogen is fine. The whole thing melts and the metals sink to the bottom. There will be several layers of metals. One copper rich, one aluminium/zinc rich. They often do this below the iron/steel melting temperature and use a large steel mesh box shape which they dip into the funace full of the melt in process. With air, the carbon based stuff burns. With inert it makes an ash which is in the top layer with the molten fluz, often salt or borax based. This is a contonuous process and each layer has a pipe to allow that level to be removed but not the others. The copper layer is the bottom and gold, silver, tin, lead etc end up with the copper, The zinc and aluminum make a layer as well, with varied other metals in there. There is a bit of everything in all layers. The copper layer will have a little aluminium and zinc in it, and the aluminium will have copper etc. The lithium will burn with the air heated one and the lithium oxide will end up in the floating flux layer. In inert gas smelting the lithium will pobablly end up in the copper zone – I am not sure.
    From time to time they lift the mesh container which will have all the unmelted stuff, steel, ceramic, titanium, platiunum etc and send to a higher temperature iron circuit – each ingot they make from each level tells them where to place that ingot from further processing. Most will be copper and aluminum based ingots and they will have a standard process for that. Occasionally another alloy layer will build up with unusual metals making a layer that does not mix well with copper and aluminum. Instruments on the crucible detect another interface layer by ultrasound that tells them they have an unusual layer to deal with, so they find the top/bottom and arrange how to pipe it out.
    Litthium is new in the recycling chain and better processes are being researched. Bulk processes, like smelting are a lot cheaper than mechanical battery disassembly – which can cost 5-10 times as much as newly mined Lithium. They are hard at work on wet chemistry processes where batteries are dissolved in exotic acids and the ions in solution reacted with other ions and precipitated out by filters etc, but these are not yet mature. I expect about 5 years before a mature water based chemical based lithium refining process is perfected.

    1. It could be the batteries have been potted to keep the batteries from vibrating and eventually losing connectivity.

      I work in electronics repair and subwoofer circuit boards are the worst.

  12. If you cut the battery open with a band saw could you easily remove the contents? In order to ensure the is no arc would you need to insulate the wheel the blade rides on and the base plate. Or could you insulate your saws all blade connection and cut between the gaps of the battery then place the battery on a lathe to trim off all the excess material. Just wondering if a machine can be designed to slice between the cells then you would have better leverage and accede on the long side vs the end.

  13. From experience with these modules, I’d recommend blowing the interstitial filling out with an air gun, then running along the top and snipping all of the wires. At that point you should be able to spray it down with isopropanol (let it sit for a while, then spray it down again) and rip the whole IC clamshell and current collector off of the top with channel locks (Gently, if you rip a cell cap off it can lead to a thermal event). Cut the plasic couplers between cooling tubes, cut an inch wide strip of the NIC clamshell off on both sides using a dremel with a bamboo skewer, and you should be able to blast out the rest of the potting with an air gun, then pull the bandoliers (two strips of cells with the cooling tube in the middle) out one by one and pull the cells off from there. Getting the grey adhesive between the sidemounts and cells off should just take a couple passes with isopropanol and gentle persuasion. Also a few words of warning: Take the busbar off first, then the VSH strip (the conductive strip under the tape on the NIC side). Don’t touch both ends at the same time and don’t let more than one person touch it at a time. I’ve seen some people get jammed by these modules and it’s not pretty (both guys are okay now) but it’s not worth the risk. (IC = Interconnect, NIC = Non-interconnect) As always, wear your HV gloves.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.