An Interview with Tesla Battery Hacker [wk057]

We covered [wk057] and his Tesla Model S battery teardown back in September. Since then we had some time to catch up with him, and ask a few questions.

You’ve mentioned that you have a (non hacked) Tesla Model S. What do you think of the car?

It’s the best car I’ve ever driven or owned, period. Not to get too into it, but, I love it. I’ve put almost 20,000 miles on it already in under a year and I have no real complaints. Software feature requests… but no complaints. After almost a year, multiple 1700-miles-in-a-weekend trips, and an overall great experience… I can never go back to a gas vehicle after this. It would be like going back to horses and buggies.

A salvage Tesla Lithium battery had to be expensive compared to a Lead Acid setup. What made you go with the Tesla?

Actually, if you consider that the Model S battery is already pre-setup as a high-capacity pack, contains the wiring to do so, and the modules are much more energy and power dense than any lead acid battery bank, it’s actually almost cheaper than a comparable lead acid bank and all the trimmings.

I haven’t officially weighed them, but the modules from the Model S battery are roughly 80 lbs. 80 lbs for a 5.3 kWh battery is around 15 lbs per kWh, which is impressive. For comparison, a decent lead acid battery will have a little over 1 kWh (of low-rate discharge capacity) and weigh almost the same.

Also, the Tesla pack is much more powerful than a lead acid bank of the same capacity.
Generally a lead acid battery bank would have a capacity that would only be realized with slow discharges, so, 1/20C. Much over that and you sacrifice capacity for power. 1/20C for an 85kWh pack is only 4.25kW, barely enough for a central air unit and some lights without losing capacity.

Now the Tesla pack can be discharged (based on how it does so in the vehicle) at up to 3.75C for short periods, and at 1/2C continuously without really affecting the overall capacity of the pack. That means I can run 10x more power than lead acid without a loss in overall charge capacity. Leads to a much more flexible battery solution since the loads will, in reality, always be so low that this will not even come into play with the Tesla pack, but would almost always be a factor with lead acid.

Charging is also somewhat better with the Tesla battery. Charge a lead acid battery at a 1/2C and it will boil. Charge the Tesla pack at 1/2C (42kW) and it might warm up a few degrees. Oh, and the charging losses at high rates are much less than lead acid also.
Overall, without continuing to yack about the technical aspects, it’s just a much better battery, takes up less space, weighs less, and has more power available.

There are likely decent arguments for other solutions, but the rest aside, this one won out because it was definitely more interesting.

Click past the break to read the rest of our interview with [wk057]!

tesla-4

Was it hard to find a salvage battery? How much did it cost?

I actually stumbled upon a listing for someone selling one from a salvage on TMC when I was considering options for my project.
It was pretty expensive, around $20k. However the cost per kWh was significantly lower than other comparable options, overall, especially after considering things I noted above. So, it was a no brainer.

What about buying bare 18650 cells and building up a system from scratch?

I had considered this, and actually purchased some and assembled a small module of cells. Many hours, pretty sore hands, and one small soldering iron burn later I had a 0.5kWh pack that in total, not counting time, cost 2.5x as much per kWh as the Tesla pack I bought. Add in labor and it was probably over 5x more expensive. Definitely not worth it.

Tell us a bit about the rest of your solar setup. How many panels does it have, and which inverters?

I’m setting up the system with just under 30kW (DC) worth of large commercial SunPower 20% efficiency 435W panels (69 of them) that I was able to get a hold of at an awesome price (< $0.80 per watt).

For inverters, I already have a couple of now are the Outback Radian Series GS8048A, 8kW off-grid. They’re programmable and work perfectly with the way I’ve reconfigured the Tesla modules (44.4V nominal, ~1900Ah). They’re stackable, so I plan to add more to basically have the same amperage AC service available from the setup that I do from the grid.

With a charged battery and no solar input I expect to be able to run the home, not counting Model S charging, for several days without issue. With solar input including Model S charging it should work indefinitely as a buffer for the power. I plan to generate around 35,000 kWh per year and run the home with the standard Tesla style “no compromises”.

Tesla Model S

Are you going completely off-grid, or are you going to sell back to the grid when your pack is fully charged?

Depends on how you look at it. I plan to power everything completely off of my solar setup/battery pack, drawing zero from the grid. However, I’m not cutting my grid service.
The inverters I’ve chosen are grid interactive. In the event that for some reason I need more power than the battery and solar can provide, they can recharge the battery from the grid.
So, 99.9% of the time I will be completely off-grid, as I’m sizing my setup accordingly. The grid will basically be my backup generator (along with my actual backup generator).

I do not plan to sell back to the grid at all. I decided to go off-grid initially because the concept of net-metering is just flawed, in my opinion. The grid is not a battery, but that is what many people are using it as with net metering. Eventually policies will change to reflect this and the cost benefit of doing so will likely be lost, and soon. Going off-grid assures that I’m in control of my power production and usage constantly. The price per kWh could jump to $5 next year (unlikely) and it wouldn’t phase me.

In the event I generate excess power, which may happen a bit in the summer, I plan to have several diversion setups in place to dump the power. The first being my Model S if available and at < 90% charge. The next being some climate tweaks (make it 1 degree cooler or 1 degree warmer to “store” that energy in the house itself). Auto modify the pool-pump timer schedule for that day. Heat the hot water a couple more degrees, etc. I don’t plan to waste the excess, but “store” it in other forms. Will take a little bit of custom hardware to make this happen, but that’s a project for another day. :)

Have you talked to any Tesla employees about your project? What did they say?

I had spoken with several people at Tesla regarding the project. Their standard responses seemed to be that they could not support my efforts in any way and wouldn’t provide information about any of the components in the battery pack to assist. I found this unfortunate.

What was the hardtesla-2est part about tearing down the pack?

Few things kind of rank together…
First, safety. Since I was not sure of the exact configuration and layout of any wiring and components, and could not find any definitive information about this, the added time and effort taken to disassemble it safely while essentially blind was probably the hardest part, but certainly important and worth it. A wrong move here could mean instant death.
Second, the pack was put together with strong adhesives almost everywhere. This physically made the tear down difficult.
Third, moving the thing. It was *heavy*, nearly a full ton. Took 4 people to move it around even with the wheels I put on it.

tesla-batt

At one point you had casters on the full pack, did you consider putting a motor on it? It almost looked like a go-cart.

The casters barely held the weight of the pack, and several were bent by the time I removed them when I scrapped the frame.
My brother had mentioned the go-cart thing, actually, and I admit… the thought had crossed my mind and would have taken the project in a totally different direction. A go-cart with hundreds of miles of range… hmm…

10. Do you think you could have added the battery to your Tesla Model S to increase the range?

Funny you mention this. Directly, no. I don’t think there would be any way to directly increase range by adding the additional battery. Weight considerations aside, the pack would need to end up in parallel with the existing pack to make it work. So, assuming I could somehow fit the thing on or in the car, this would almost certain confuse the heck out of the electronics in the car and probably would end badly.

Indirectly, however, I plan to potentially try this. I’m going to have to move my test setup (where I currently have a couple of inverters, the pack, and a couple of the solar panels setup) to its permanent location soon. If I have the time, I may wire up half of the pack in the back of the Model S along with one of the inverters. Then I’d drive, stop along the way somewhere, and charge the Model S from the pack in the trunk.
Probably pretty pointless, but, a fun “Yeah, I did that” project for sure.

11. Do you think Elon Musk would be happy about you pushing the limits of solar with your Tesla battery pack?

I believe Elon Musk is already involved in similar more official projects if I understand correctly. As for whether or not he would be “happy” about my project with a salvage pack, I couldn’t say. I think he may at least appreciate the recycling aspect of it, in any case.

12. If you could ask Elon Musk anything – be it about Tesla, Solar City, or SpaceX, what would it be?

I would actually be curious as to what he thought about my particular project, if he had anticipated such projects, and what, if any, impact he feels similar projects have or will have on Tesla in general, now or in the future.

We’d like to thank [wk057] for taking the time to answer our questions, and wish him luck with his Tesla Supercharged home solar power system!

tesla6

63 thoughts on “An Interview with Tesla Battery Hacker [wk057]

      1. By my calculations (based only on this posting), he’s spent around $44,000… that is a lot of money for a hobby budget… >_<
        Plus he apparently owns a Tesla S… (another $60-100k)…

        Note to self… get a job in whatever field this guy is in, because good googa, he must be raking in the dough… =p

        (This calculation was based on the stated $20k for the battery, and then he said "435W panels (69 of them) that I was able to get a hold of at an awesome price (_>)

        1. bleh… formatting screwed up my post…

          “435W panels (69 of them) that I was able to get a hold of at an awesome price (< $0.80 per watt)."

          $.80/W * 435W * 69 Pannels = $24k, plus the additional $20 he spent on the battery…

        2. It may be a hobby, it may be a lot of money, but it is clearly a capital improvement to his home as well.

          Try that if you are trying to justify some of your projects to yourself (or your spouse). Just make sure you finish them, and they work.

          1. Now if he wanted to sell his home together with the setup. Would he be able to retrieve the invest? Would potential buyers see this (very non-standard) system as an asset or a liability?

          2. It is a lot to budget for, but still in the same general cost range as a lot of home renovations, plus you can take a lot of it with you when you move. (And that’s a good solution around the issue of selling the house with all the nonstandard stuff in it…)

  1. “So, assuming I could somehow fit the thing on or in the car, this would almost certain confuse the heck out of the electronics in the car and probably would end badly.”

    What about with a firmware update?

    Tesla Model SEx, extended range model S.

        1. @Old’un
          Stop turning AA off. If text looks like crap with AA, you have some combination of a crappy monitor, a crappy font, or poorly configured AA settings. I realize that you may not be in a position to fix some or all of these issues, but you need to understand that hackaday can’t reasonably accommodate every oddball usecase.

      1. This is my first time comment on a website layout, but holy crap what happened? I’m all for updated features, but this totally pulled the “linear form” out of this site. There is now “older posts” button all of a sudden. This might force me to RSS, which I really don’t want to do.

      2. Disable infinite scroll. Not everyone runs gaming rigs!

        The rest of the layout, eh whatever everything is a little too big not a huge deal. Kind of looks like something someone whipped up using wordpress templates in less than an hour.

      3. Does the “update” include editing our own posts? Adding skulls, plusses, or minusses to comments or posts? If we can’t do those, is it really and “update” or a lateral move?

        1. There are a few people who would say upvoting and downvoting posts and comments would be a downgrade; i.e. stupid floats to the top.

          This isn’t a lateral move; it’s combining the blog, hackaday.io, featured posts, stuff from the store, and everything else in the hackaday ecosystem. It only makes sense that we would put all that stuff in one place, and this is the solution to that.

          Since this is going to be the ‘official’ hackaday off-the-cuff comment, THERE ARE A TON OF YOU THAT WILL COMPLAIN ABOUT THE NEW LAYOUT. JUST LIKE WHEN WE LAST UPDATED THE LAYOUT. Everyone complains for a day or two, and a month from now when people look up the old design on archive.org, everyone comments how dated the old layout looks. Trust us on this one. It’s better. Stop being afraid of something different.

          1. I agree with the sentiment of complaining because it’s different, but I’d be more of a fan without the slideshow on top. Another site I read, Consumerist, has a similar thing on top of their page getting in the way of accessing the posts I came to read. It won’t keep me from perusing either site, just a design choice I’m not a fan of.

          2. Hey, I don’t think its that bad, weird maybe, odd, maybe, but as far as I can tell no structural or system flaws, just what I’m used to. The main home page I do say though is weird, really weird. The blog is more of a side note, with the home page being almost a landing for each thing individually, I’m not sure that’s really intuitive, As most are wanting to read, so the blog should somehow be more accessible, more forefront, not treated equally. I do say too the all black background I think fitted the hackeresc chic look.

        1. Its so odd to me that Tesla wouldn’t be that innovative. From what I’ve read about Elon, and his personality, the idea of innovation and old fashioned entrepreneurial would be dubbly supporting this and you.

    1. can confirm. i’m still pissed off by the fact that the read more links are all BEFORE the content. by the title.

      so my flow is now: read the title, see the pic, read the first part of the article while slightly scrolling (720p and all that)… then i want to read the rest, and … no link to be found!

  2. My main concerns with this whole “project” would be longevity of the power storage system. At ~$20k per pack, it had better last a lifetime.

    Having lived completely off-grid for the last 5 years or so we’ve picked up a bit of experience in this matter. I think despite the slower ( much ? ) discharge rates of Nickle/Iron, I’d rather go this route and have a storage system that will last several lifetimes. Some of us just do not care to spend $20k every 5 to 10 years . . .

    1. I’m not sure why you’d think it would only last 5 to 10 years.

      Over time the total power available will certainly degrade, but I wouldn’t expect to ever have to actually completely replace the pack in my lifetime, especially with such an easy load on it like a residence. These cells are happy with huge charge and discharge rates, and a residential off-grid solar setup will never even come close to those limits. (See what I wrote about this in the interview.)

      Kept in a climate controlled environment (ie, indoors) and never fully discharged (~5-10% buffer) and rarely charged to 100% (90 or 95% max) I expect the pack should last for many decades.

      1. I have a really old cell phone battery (lithium ion) that is about 15 years old and it still works. Probably only a quarter or so of its original capacity but it does still work. In contrast, lead acid becomes completely unusable in 5-8 years. NiMH (the kind I work with at least – the automotive grade stuff lasts much longer) seems to last 5 years or longer, with a few over 10 years old still working but also many that failed in about 7 years.

        1. The reason that Toyota guarantees the Prius M-H battery is that they declare 40% chg as “discharged” and 60% as “full charge”. I suspect that the 5-10% low and 90-95% high numbers are a tad optimistic for long life. I’ve seen it claimed that terminating LiPoly at 4.15v (verses 4.20 spec) decreases capacity 7% but nearly doubles cell life. Nonetheless, I’m still getting great service out of 18650s from a laptop running Windows 98!

      2. Might be worthwhile doing a custom controller for the hot water system. Or buying a high impedance heater that can be run constantly at lower wattage.

        An electric hot water system calls for 4kW of heating and then turns off suddenly on the thermostat. So you have rapid switching between high discharge rates and then no load at all.

        Either that or switch to solar hot water.

          1. Actually, the load of a normal electric hot water heater (~4-5kW?) is really nothing to this battery pack. 5kW is ~0.06C draw. Literally about 200mA of current per cell in the pack for a 5kW load. The 18650 cells will happily put out 200mA without skipping a beat.

        1. My hot water approach is …
          Dedicate 4 to 5 p.v. panels and run the DC into the bottom element in a dual element 240v heater. It’ll run at 1/4 to 1/3 heat capacity as long as the sun is out. Let the tank get up to 160 degrees F at peak and use a mixing valve to mix the up-to-160 hot with well water to achieve a 110 to 120 degree tap temperature.

  3. So, that works out to what, about $3/cell. Not bad. New old stock laptop packs with similar capacity cells seem to go for $3-5 a cell on ebay, though I’ve found a few for ~$2.25/cell. And then, of course, there is all the labor avoided by reusing intact subassemblies.

  4. So, that works out to what, about $3/cell. Not bad. New old stock laptop packs with similar capacity cells seem to go for $3-5 a cell on ebay, though I’ve found a few for ~$2.25/cell. And then, of course, there is all the labor avoided by reusing intact subassemblies.

  5. Thanks for this brilliant interview ! For sure it is a budget but it shows how it is better spending his money in useful project ! Making you think about essential needs like energy (more and more concern) rather than spoiling money in the Titan Zeus TV for more than a million dollars.
    WK057 you rock and it is motivating seeing this kind of project and much more important sharing it !!
    Thanks to you and to Hackaday team !

  6. Great interview.

    To recover some of the battery costs, alternative uses for electric car battery systems are being considered by the electric automobile vendors once their used packs fall below 80% capacity and get changed out. Since range is everything for a car, putting used battery systems into grid energy storage schemes is one of the many things being considered. There may be legal and liability hurdles to overcome before we see used batteries on the open market.

  7. Very cool project. I can see how a properly sized solar PV system could, on an annual average, be self sufficient with little use of the grid on a NET basis. I can see how the 85 kWh battery and other adjustments can load level over a span of days.

    What I don’t understand is how seasonal changes can be managed to make the system independent of the grid 99.9% of the time, without generating excess to the grid during some months of the year and/or drawing a lot during other months of the year.

    My 7.7 kW (DC) solar PV system makes 4 times more energy during the month of June than it does in January.

    To not be a grid consumer 99.9% of the time, I’d need a ~30-35 kW (DC) solar PV system to cover my needs during the winter months. But that system would end up generating an enormous excess during the late spring and summer, with about 3/4 of my generated solar energy being sent to the grid during a period spanning several months, totaling tens of thousands of kWh.

    Alternatively I could size my solar PV to just cover 100% of my needs in the summer. But then I’d be drawing a large majority of my electricity during the winter months from the grid.

    An 85 kWh battery can level load on a time scale of days. But it’s not nearly big enough to bank an excess of solar energy generated during summer months to be used 6 months later in the winter. Even with other means of buffering such as charging a Model S, and adjusting the home thermostat or pool temperature, I’m not understanding how one can transfer thousands of kWh of solar energy from summer to winter.

    wk057, can you explain in kWh the expected solar production, storage, and consumption for a couple of month-long scenarios — January and June?

    1. The battery isn’t meant to store excess from the summer. The system should generate more than enough power in all months, using the battery as a buffer. The system is oversized (mine is around 30kW). I expect to have some excess in the cooler summer months, which I would try to dump into various diversion loads. I’m also quite possibly adding a second 85kWh pack to my setup.

      Estimates for my setup show that in the lowest month, December, the system should generate 2000 kWh (64kWh/day). In the best month, June, it should generate 3700kWh (123kWh/day). Annual production around 35MWh. This should be more than enough for my needs.

      This is all calculated after derating for the DC->AC process.

      In June, consumption should be highest. A/C running to cool the house, pool pumps, lawn sprinklers, etc.

      In December, consumption should be lower, since heat is supplemented by natural gas. Eventually I may add solar capacity to not have to use gas, though.

  8. Hi, during my research on Li+ cells I have found that “irreversible” degradation on LiFePO4 and also possibly LiNiMnO2 can be almost fully reversed with near 100% capacity when before they were measuring less than 20%.
    Turns out that the literature on these is actually wrong, overcharging is actually the killer.

    Is anyone interested?
    Also managed to unmuck totally discharged ie 0.7V LiFePO4 and it looks like the NiMnO2 can be repaired as well as have two e-bike packs with “bad” cells.
    Has to do with thermal management, it turns out that a lot of problems with solar lights using these can be similarly fixed.

  9. Have some recovered cells here still in use (years later) so it obviously works.
    Many of the problems with capacity loss are actually issues with the cathode and NOT copper shunt, this only generally occurs if a cell goes below 0.5V with this chemistry whereas Li-Ion (ie CoO2) it is more like 1.8V
    Some recent work conclusively proves that a memory effect exists with these and this further supports what I have done ie using the correct treatment during the 3 step recovery charge brings the cell back to life.
    The problem is that with the overdischarge the active cathode transitions to an inactive state and my method reverses this and could be applied to any battery with a ferromagnetic cathode.

  10. i am having 2amp 3.7vol cells ………. i want to make my battery capacity as 36amp and 48vol with using those small batteries.plzzz help mee .. how to connect those batteries to get my out put

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s