Battery Balls, fast way to fill your electric vechicle?

[Bert]

 

A traffic spike to our website from the eevblog? I better get in on the action, before we're piled up with the Solar Roadways, Perpetuum mobile, and other BS stories.

We're for real. Let's do some debunking.

I smell BS! Can't be done! Marketing hype! Boo!

Skeptics are cool. I'm one of them.

But I also happen to have designed and built the HW, so there is another perspective, of which I'll share some here.

I might have to do some posts in between because this will take some typing.

[helius]

Welcome to the forum.

Your tone is a little upsetting. I haven't seen anyone here say that it can't be done or that it is simply BS. But they are questioning why it is better than other simpler ideas, that do not require an array of new technologies to be perfected at the same time. Engineering is not, like politics, the art of the possible: it is about finding solutions that are better and cheaper, not "if you build it, they will come" faff. To give you an example, wireless power in the home, such as the WiTricity concept, is obviously possible. Is it economic to install at this time, or in the foreseeable future?

[taylor12]

Engineering is not, like politics, the art of the possible: it is about finding solutions that are better and cheaper, not "if you build it, they will come" faff.

  wait what? i must be studding the wrong thing 

[Bert]

another solar road way type technology.
http://tanktwo.com/

https://youtu.be/8BAk4l98GtY
I think its a Gimick, low energy density (casing and air gaps and electronics) and low current path  due to point contacts and random connections will mean random voltage ranges.

 

1) the best way to convince you would be to make you read the patents. It is heavy duty legalese mumbo jumbo, but the facts are there in all the glorious detail.

2) Let's do some numbers. The casing shape is an ellipsoid, with a very specific semi axis ratio (approximately 1.2:1:0. which is the shape with the highest known random packing density of any shape. It beats cubes, spheres and everything else.

Here is an interesting reference for you to dig into if you don't buy that argument to begin with: http://mathworld.wolfram.com/EllipsoidPacking.html

Wolfram alpha is cool, but after that read also the M&M packing density paper (fascinating, and you get to eat the experiment afterwards). By the way, the internal code name for the String Cell is "candy". We're not big on poetry, it's ok if you think that's lame.

So: "our" ellipsoids have a random packing density of approximately 76%, which we round down to 73% due to "wall effects" in small "tanks" and the process which is not perfect.

Next, the shell is already less than 10% of the volume at the current size (and there is lots of room for improvement). We also use a couple of percent of the volume for the PCB and the ASIC. Leaving, conservatively, 60% for electrodes and other real battery stuff.

Compared to a regular battery pack, that's pretty damn good to begin with, because the 20-something percent which is "lost" is used for cooling, something which other packs need to do too.

And a cylindrical cell, or a pouch cell for that matter, needs a casing of some sort too, which is arguably less in volume than our current plastic cell, but it is really not much to begin with - we're just a few percent behind already now. And we have our ways of filling that space entirely, batteries are not always rectangular or cylindrical - think wrist bands for wearables)

But, only then the things get interesting. We don't need an external BMS (battery management system), because it is integrated and distributed.

1) We don't need bus bars or thumb-thick cell wiring (a Tesla Model S P85D, to take a rather cool example, draws a bit of current when it is putting that Lambo Aventador to shame in an acceleration test). [Note: I'm not saying that the P85D has bus bars in their pack, but it is a nicer visual image than the Nissan Leaf. (Hi Nissan, I promise that if you want to buy our batteries I'll never say that again in public)].

2) We use light plastic tanks with no rigidity requirements, which are a major issue if you have a half ton plus pack. You can't have 7k cells sloshing around in a standard pack. This saves space, but more importantly a LOT of weight. Weight is harder to save on when you design a BEV.

The list is much longer beyond this, but from this point onward we are already on parity with the legacy systems.

Not all  cells will be used as only those that meet the algorithm.

Well yes, obviously. But the exact utilization rate is what matters. And of course one of the things that the company is making money of is by developing algorithms that make better use of packs.

About the absolute numbers: even with simple routing algorithms (Think Dijkstra algorithm, which we all have learned in school, you can get already 90%+ utilization rate. Easily. Of course this depends on a boatload of other parameters, such as number of contacts on the string cell, placement of contacts, size of the "air gap", size, shape, number of channels (the contacts placed to the walls of the string battery, which is the enclosure or the "tank" if you will), and a ton of other variables. Getting close to 100% is easy, although admittedly not intuitive. Simulations and hardware prove it though.

This is basically a box of dry joints, any vibration would cause intermittent open circuits and arching.

Nope, wrong. Two reasons:

1) the dense ellipsoid packing naturally leaves little empty space, which also means there is little room for rattle. If you just fill a container, and shake it violently, you can barely hear any rattle to begin with. This is also not intuitive, but it's basic physics. In geometrical terms it is simply a matter of degrees of freedom. Think shaking a bag of sand vs. a bag of marbles.

2) the string battery container has a silicone bladder, which is pumped up with compressed air. Even with modest pressure applied, the string cells remain rock solid in place. The G forces found in a normal passenger vehicle don't come even close to what is needed to even marginally upset the arrangement.

3) contact resistance is a well understood section of engineering, as the connector wasn't invented yesterday. You need surprisingly little mechanical force to achieve an acceptable contact resistance value, if you use the appropriate alloys. In the early prototypes, we used gold. That works well, but is a little on the expensive side. Tinned copper, for example, has worked very well - a little bit to our surprise admittedly. Apparently the oxidation effects don't harm operation at all, and contact resistance is in the milliohm range. We probably won't use tin much going forward for certain reasons, but it is an illustration.

4) the string cells are smart, meaning they measure and control stuff, and also measure the current going through their contacts. The connections go through N-channel power MOSFETs, and when contact pressure is accidentally reduced and contact resistance goes up, the power FETs have disconnected the load long before the mechanical connection is severed. Hot plugging and unplugging requirements are nasty on connectors, but we remove that requirement by anticipating "hot unplugging" events it and removing current before it can cause a spark. It's surprisingly simple.

with most energy density you want the most chemical in the smallest volume, thats why you use liquids or powders or gels/pastes

Sure. That's why you won't see the Boeing or Airbus Leaf soon.

hell of a lot easier to just transfer electrons at high current and voltage then transfer battery balls.

Of course. That's why a car with our batteries can still be charged in the same way than any other electric car. And even when you have the capability to swap them, you won't do so most of the time.

Is there any spec on the cell? Ah capacity? dimensions? volume?

Of course. We don't quote capacity without NDA because there are variations on the battery chemistry options with impacts on price, longevity, peak current, and other parameters. But there is nothing special going on there, the battery chemistry are variants of the "standard" Lithium-ion cells and all options are available to our customers too.

I smell BullSh!T

Try getting new firmware for your nose.

Thanks for the feedback anyway. This will only make it better!

Bert

[Bert]

Slick site with very little actual data. IMHO It's an interesting but impractical idea, I'll go as far as saying it's a bad idea. 

I could name a handful of issues of the top of my head but instead I'll speculate the project will never go into production.

You're wrong with your speculation, but still would like to hear your issues. It is only by eliminating the issues one by one ourselves (the ones we could come up with) that we got it to be a product.

nope. the thing exists. I know the guy who invented it.

As a prototype? All the "photos" on the site appear to be CGI renders.

Okay. Let's post a few pictures of a proof of concept proto run from a year and a half ago. This is version three, still built with discrete components, and therefore not a lot of room for battery. But it is fully functional and was used to develop software on. Picture is in attachment.

[Bert]

Seems TankTwo is at the get investors stage so slick presentations and press releases are top priority. But I couldn't find any actual data or independent test info at all.

We opened for business two weeks ago. Perhaps you are the kind of person with 100 or 200 million spare cash, but unfortunately we are not one of those.

So we had to start and grow like most other businesses grow: build something, sell it, build something better, sell something better. If you have a better idea, please share it!

We designed the best we could, and we built prototypes as many and as well as we could that are being evaluated by potential customers.

I simply don't see any practical use for it... I'm amazed this got off the drawing board.

That's fine, there are people that do see it. It's not surprising you are amazed because it is really quite cool - and indeed rather complex to execute.

[electr_peter]

Hi, Bert,

Welcome to the forum and thanks for sharing your comments.

Could you comment a bit about tank size requirements and how it is supposed to fit in a car? What kind of capacity and voltage can you expect from single module (5-10 Wh at 3.7V)? Is the outer shell size determined at this point and will be fixed in the future?

Okay. Let's post a few pictures of a proof of concept proto run from a year and a half ago. This is version three, still built with discrete components, and therefore not a lot of room for battery. But it is fully functional and was used to develop software on. Picture is in attachment.

Can I have few of those assembled PCB for Easter? Those modules look like very funky eggs

[Bert]

A well-maintained Tanktwo battery-equipped EV can have more capacity at resale than when it was new, allowing for unseen resale values.

....Small battery modules may have advantages (gradual replacement/upgrade of battery cells, quick refill) only if they are treated as such. If they are treated as a single battery pack, many advantages vanish (this would be the case if little or no infrastructure is built) in comparison to single battery pack unit.

Good you realize some of the ecosystem-level advantages, but your assessment of the benefit vanishing without infrastructure being present is fortunately not correct. While it is certainly true that it decreases, the net benefit is still significantly positive.

Many details are still in unreleased patents, but I'll share three elements that bring us to break even.

From the Li-ion battery wear data you can see various complex phenomena occurring at the same time, compounding and influencing one another in various ways, but certain basic things hold true.

Therefore (although I'm simplifying for the sake of the argument):

- the fuller you charge and the deeper you discharge, the more wear is induced
- the shape of the "SoC vs. wear curve" is a standard bathtub
- fuller charging and deeper discharging effects at a dynamic rate over the lifecycle, ie. the edges of the tub become blurred (think Monte-Carlo analysis curves)
- limiting temperature excursions is very important for longevity
- Commercial BEV cells have bricking-reserves and lifetime optimizing reserves roughly between 15 and as much as 40%

Therefore:
1) if you optimize SoC limits for each individual cell as it ages and the variance grows, you can extract a significantly higher useful charge out of the pack as a whole
2) if you throttle cells individually to optimize for limited delta t, (caused by wear induced ESR changes, less optimal cooling conditions in the corner of a pack etc.), the longevity can be maintained while extracting more capacity (or more longevity @ same capacity).
3) if you choose to widen your SoC limits, and accept a higher FFR down the road, you can simply replace those cells at a maintenance center after, say, 5 years. You can build the cost of that replacement rate into the sticker price, and push the pack as a whole significantly higher. This is purely a business model exercise and requires no service stations - only a cell segregator at the maintenance center which is in essence a glorified shop vac.

[electr_peter]

I have not thought about dynamically enabling/disabling  separate cells to improve overall wear, charge levels and other parameters. In comparison, big sealed single battery pack is not that flexible - some cells wear out unevenly and cannot be replaced in the future.

With this in mind, even single EV with such technology and little external  infrastructure could benefit from improved battery management.

[Bert]

If you wanted to recharge an electric-drive car's battery pack in 1h, you would need an electrical circuit rated for 400A/240V and a battery able to absorb that much energy that fast. Easier said than done and some people may have trouble handling cables that thick.

Tesla is doing fine with their Supercharger concept, but it has some limitations, grid load wise and for a few other reasons too. It seems though that cramming 130kW through a rather modest cable works fine.

The dry joint issue can be mostly solved by simply adding an air bladder at the top to compress the balls after they have been loaded so cells won't shift or break contact so easily.

That's how we do it! Other methods have proven to work, too.

I agree that energy density, both by volume and mass, would be questionable at best - stacked lithium cells are already at the limits of being practical without the extra weight and volume of fully enclosing individual cells, along with the routing MOSFETs' losses.

See my comments elsewhere. Your skepticism here is warranted but the numbers work out great. The RDSon values make the losses insignificant at the currents involved.

As for some cells not getting used, that is not necessarily a significant issue as long as you only get billed for the difference between the charge in the cells you dumped and the cells you picked up.

That, too is correct. Even though we don't talk much about that yet.

It would be much simpler to simply come up with standard EV battery pack sizes, capacities and locations so they can be quickly swapped out when fast-charging is either not possible or still not fast enough. If swappable batteries become part of the energy delivery infrastructure rather than individual property, battery packs would get progressively upgraded over time to deliver more billable capacity at lower labor and material costs.

Better Place spent $1.1B pushing that model like there was no tomorrow, but it failed. The list of reasons is a mile long, but the three most important ones seem to be that

1) one size fits all does not apply to transportation. Granularity is required.
2) swapping stations for packs are expensive and fragile
3) You don't want to receive a $20k battery of unknown history - the reverse logistics needs to be trustworthy before you want to participate.

[Bert]

Seems very much like the strategy of swapping battery packs for range extension. Except for the differences:
1. Since there are many small batteries, they can be derated/replaced at a finer granularity
2. Dynamic mesh power reduces impact of single failed cells, and can keep using marginal cells

Correct, good analysis. Not complete though - but I'll comment on your your perceived disadvantages below.

those seem to be the only real advantages. Disadvantages compared to single swappable packs:
1. Less energy and power density because of spherical packing arrangement

This is not correct. I've explained elsewhere a few reasons why this is not the case.

2. Higher cost per Wh capacity because of the construction of the balls and contacts

That is not correct either. It is true that while at the moment we cannot yet build you a battery at a competitive price, but already at very modest volumes we can be cost competitive.

Another, not earlier mentioned argument, is that smaller cells cost less per Wh than large ones, the most important reason being the same why large dies suffer from lower yield, but also because guaranteed FFR rates can be lower with our batteries vs. the competition.

3. Cooling the cells efficiently is impossible

How so? That's what the space between the cells is for. Air cooling is even sufficient, keep in mind that each cell does duty cycle control to begin with. The cells can also be liquid cooled but so far no application has required it.

4. Contradictory requirements for materials: to make contact reliably, the patches on the outside of each ball must be compliant. But that means they have to be thin, reducing their power handling ability and making wear and tear a problem

That is partially correct, in the sense that it is an engineering trade-off.

However, the current that each cell sinks our sources is so low that the power handling is not even close to being a limiting factor. The 18650 cell fuse links in a Tesla Model S battery are high double digit AWG wires too, nothing special going on there.

Wear resistance is also fairly easy to achieve, but keep in mind that most people will only occasionally do a battery swap. They'll charge their car in their garage or on their driveway most of the time, as they do today.

5. Requirement for a large storage tank inside the vehicle, which you can calculate must be larger than either a liquid fuel tank or an EV battery pack. The tank needs to be large in each dimension so it is nearly cubic

The volumetric requirements for the container are roughly comparable to a pack of comparable capacity. There are few constraints on the shape, which can be extremely irregular.

6. Multiple unproven technologies required: mesh power, randomly oriented contacts, air delivery (easily the wildest proposal if you understand how pneumatic tubes work), and on and on.

That's true, the system implements various new methodologies from several disciplines.

Progress would not be possible without someone proving unproven things.

[helius]

I appreciate that you are responding point by point to comments. That is really everything that should be expected for PR in an engineering forum, so good job there.

I still don't understand how you can say that the packing is space efficient. Surely if it was, then every large storage battery would contain rounded instead of flat elements? I suppose you're including the space taken by the bus bars, cooling channels, module BMS, and module rails in an EV pack in that comparison.

The comment about efficient cooling is based on teardowns of Tesla battery packs, which are liquid cooled. I could be wrong, but liquid cooling over the cells only seems feasible if connections that lie inside the liquid volume are permanently welded. You might be able to use gas-tight connectors but they add cost (expensive connectors but even more in assembly time cost). Liquid cooling also seems incompatible with an air delivery system. So the alternative is air cooling, but that requires a higher W/K cell package to achieve the same system efficiency.

[firepower]

Welcome to the forum Bert, I originally read about TankTwo on Electric Motorcycle Forum. Glad to have helped in highlighting your product. It is a original and unique idea,  just does not seem practical / economical unless you planning on a monopoly like the fuel industry has.

http://electricmotorcycleforum.com/boards/index.php?topic=4496.0
By the way the Zero motorcycles have the highest capacity/kg of any battery pack 184Wh/kg air cooled, compared to Tesla Roadster 117Wh/kg and Model S 140Wh/kg liquid cooled.

We needed specs for the most important part of your smart battery, the cell capacity. What is your expected Wh/kg?.
Then we will be able to make fair jugdement.
Every week there is a new upstart or lab promising break through in new battery technology.

I like to apologize for being harsh on your product, making statements without facts, thank you for taking the time to answer our questions. I support any tech that promotes EV and gets us better efficeny than burning dinosaurs.

[Bert]

Don't forget a delivery system that can exchange hundreds or thousands of those balls through forced air tubes without jamming in less than three minutes. That I'd like to see.

It's not done the way you think it would be done, but it is simpler than you think.

Moreover, pneumatic conveying is a very mature technology which you can buy off the shelf from many industrial suppliers already today.

I think it is a great idea, with a lot of problems, some of them are breaking it totally. Namely the contact resistance of these batteries. If you put a lot/all of them in series, you basically have a system, where you have a lot of batteries connected to each other by only a small surface, and small forces. The real EV batteries (not these tesla notebook ones) all have screw terminals and all use busbars and thick metal plates to make the connections. Replace it with two rounded surface touching each other randomly and you are asking for trouble.

Obviously we are not cramming half a kA through two metal thingies that gently nudge one another. The system is distributed to the extreme and the contact resistances are not an issue in any way.

Probably all the cells have to be in series to reach a voltage level to decrease the current, such the contact resistance doesnt matter anymore, but at around 500-600V you start having trouble with the electronics, MOSFETs stop behaving and IGBTs are pain.

You're observation about the semiconductor limitations is correct - but you are going down the wrong path for the solution. The FETs see a Vds of a couple dozen volts max.

So far I would have prefered the Renault-Nissan battery swap, but that went down the drain.

Sometimes businesses fail, and then someone else tries to do it better.

Does anyone care to offer odds on the transfer system jamming badly enough that one or more charged spheres suffer sufficient shock damage to cause them to catch fire?

You can't shoot them from a cannon, but their exposure to G forces encountered during all foreseeable handling scenarios won't do them any harm.

Additionally, there are safety mechanisms in place to deal with the theoretical worst-case scenario.

How many catastrophically failing spheres can the charging system hopper contain without the whole lot going up?

That's a good question, but they aren't spheres, for a reason. See my answer above.

If the spheres have enough impact absorbing and fire protection material round their batteries to make these concerns dismissible, the energy density will be even lower . . . .

A fair assumption, but that's not how we do it. Think how airplanes decelerate to get an idea.

This technology seems ahead of its time. Like maybe about 2 days ahead.

It's the technology of tomorrow, so you're about a day off.

Seems like a bunch of mobile phone guys got together with a bunch of marketing guys.
http://tanktwo.com/team/

Not everyone is on there. But they are saying hi anyway. We're fans of EEVblog too

Here's the patents.
http://worldwide.espacenet.com/searchResults?DB=EPODOC&submitted=true&locale=en_EP&ST=singleline&compact=false&DB=EPODOC&query=tanktwo
https://tanktwo.com/tanktwo-blog-post/
They're tiny!

Let me know if anyone has chewed through the patents. Apparently it shows it wasn't Stephen King who wrote them (sorry).

Bumpy roads could be a problem: <bump>*rerouting*........<bump>*rerouting....  *rerouting*

Good thinking, but no. Explained elsewhere.

Even in some magical universe where spherically packing a bajillion separately packaged batteries was somehow not a spectacular waste of time, space, and weight; how do they expect this to gain critical mass among car manufacturers and charging stations?

You should leave cyberspace and visit our universe some day. The solution is called "trying to run a business."

Success is not guaranteed and luck is needed, but unless you're a fan of North Korea, you'll have to agree that this is how it's done.

PS: they are not spheres.

I'm pretty sure that no engineer would ever want to be associated with such an awful project

Unless your name is Einstein, Hewlett, Packard, Horowitz, Shockley, Bardeen, Noyce, Grove, Woz, Williams or Pease, I think that's not how you are supposed to behave on the playground.

If you're Heidi Klum I'm ok with it though.

I suppose the algorithms for routing paths through the spheres could have some niche uses, but nothing like what they are advertising.

Maybe you don't understand it all well enough yet? Please read the patents.

PS: They are not spheres.

[max_torque]

but from this point onward we are already on parity with the legacy systems.

No, no you're not.

  If you have a OEM approved productionised battery system that is in fully type approved and in volume production with the same energy density as conventional EV batteries then yes, you have parity. 

If you have a bunch of designs, prototypes and some sums / simulations, but nothing else you are so far away it's not even funny.  I'm not sure if you have worked directly in the Tier1 / OEM automotive industry, but until you have that experience i can tell you that bringing a theoretical system to volume production is no small task.

For example, on a low volume(375 units) ultra high performance hybrid project that i worked on, we had the prototype battery system working in under 3 months.  It took another 2.5 years and approx £10M to get that to production.......

This is the sort of thing you will be up against:
http://www.cenex-lcv.co.uk/2013/presentations2013/day2/dome2/D2D2S2-Allan-Paterson.pdf


i.e. the largest, most established, Teir1 suppliers in the business are investing millions already in this field.
You seem to be incredibly nonchalant about the thermal performance of your system, and yet, this is CRITICAL to the systems real world performance, more so that things like cell Ri, or absolute maximum energy density etc.

Sorry to say, whilst i wish you the best of luch, there's no way i'd be investing in your company, based on the facts at hand.

[max_torque]

By the way the Zero motorcycles have the highest capacity/kg of any battery pack 184Wh/kg air cooled, compared to Tesla Roadster 117Wh/kg and Model S 140Wh/kg liquid cooled.

That ^^^ should set some alarm bells ringing i'd hope, as it tells you something very important about passenger car batterys!


(namely, that even a company as forward thinking/innovative/cash rich as Tesla, has had to compromise on Energy density in order to meet practical and legislative limitations(and use a liquid cooled pack, like EVERYONE else........)

[peter.mcnair]

I suppose the algorithms for routing paths through the spheres could have some niche uses

Cool - a 3D travelling purchaser problem (TPP) - and NP-hard 

[eneuro]

2) Let's do some numbers. The casing shape is an ellipsoid, with a very specific semi axis ratio (approximately 1.2:1:0. which is the shape with the highest known random packing density of any shape. It beats cubes, spheres and everything else.

Yep, they have a big chance to make epic win, but in this kind of research http://www.improbable.com/ig/ 


I like math too... Did you calculated this battery energy density per m^3 (cubic meters)?
Probably you should from this perfect eliptic shape substract volume of ... those blooddy PCBs needed to show frustrated investors that this thing can powerup something and they need invest more to run EV vehicle 

Yep, I was thinking that my flywheel battery is  but looking around and findidng projects and ideas like this battery balls, probably I have much higher energy density even in made at home KERS and no need to fill battery space with... PCBs, but something that can store energy 

Anyway this
http://en.wikipedia.org/wiki/Vanadium_redox_battery is for years in industry and... charged fluid can be simply replaced with used one, so no need to mess with bloody  balls battery 

[DanielS]

Better Place spent $1.1B pushing that model like there was no tomorrow, but it failed. The list of reasons is a mile long, but the three most important ones seem to be that

1) one size fits all does not apply to transportation. Granularity is required.
2) swapping stations for packs are expensive and fragile
3) You don't want to receive a $20k battery of unknown history - the reverse logistics needs to be trustworthy before you want to participate.

Granularity can be achieved with standardized battery packs too: decide on a baseline unit, say 200V 50AH, then equip vehicles with a multiple of those in combination series-parallel to meet performance and range requirements.

As for having a battery with unknown history, the same goes with TankTwo: you have hundreds of unknowns histories in the tank. By buying into any batteries-as-infrastructure scheme, you are trusting the battery ecosystem to maintain the battery/cell park and refurbish/retire bad packs/cells as necessary. The packs/cells individual history should not matter since the ecosystem should be ensuring the packs/cells it circulates are still within specs. An energy ecosystem where people cannot trust the infrastructure is not going to last long. If you are going to tell me TankTwo cells have built-in history, monitoring and logging circuitry could be added to hypothetical standardized-format packs too.

[Bert]

200A current thru those contacts? Yeah, sure... 

How about 200A through 15k contacts?

I read their patents, and they are bullshit.

Either you haven't read them, or you don't understand them.

It's more like a many batteries auto connecting system, a "high power network" router?

Yes.

There is no storage innovations. I think they are making small battery cells, and make them auto connecting to each other, and replace a batch of cells when they're depleted.

If you are unable or don't feel like plugging it in, that is an option, yes.

Might be a good idea for fast "refueling", but I can not see why it is better, compared to swappable bulk li-ion batteries preconfigured as a module.

Maybe because you haven't read the patents?

At least we know who's going to do their marketing...

Update your link - our marketing guy has a beard.

The very front page asks a question: why not make the batteries smartphone smart: Anyone who's seen the latest EEVblog mailbag where Dave's phone craps itself randomly will know why not.

And you are reading the forum from a rotary phone, a typewriter or a clay tablet?

And the front page looks like it has marketing smeared all over it. 

If you want to sell your stuff, you need to market it. Think about it as having to accept that cow dung needs to exist if you want to have ice cream.

Can't believe they the size of ping pong or squash balls, bugger all capacity. By the time you add up all the tech it meant to have and the construction of the casing for strength and shock. You be filling up more often than current EV.

Maybe you need to look at the facts before you make broad claims about something. It's bad for your street cred.

Tesla are designed for fast charging, fastest possible as well as battery swap. EV are topped up every night at home, always full tank every morning.

That's no different with our solution. You  just have an extra option.

just think about the sound of exchanging them i would sound like you won a slot machine  ding ding ding

Unfortunately not. Should we add it to the feature list?

And the grand prize is a fire, and a burnt car.

Finally something to put on YouTube!

The ideal solution would be replacing a complete battery pack, but the trick is designing a car where such a large, heavy component can be quickly and easily removed without unduly compromising safety or performance.  And of course, coming up with a standardized pack that can be fit into a wide range of vehicles.  Probably the best way to do it is to have the pack slung under the car, but then you need to have access to the bottom of the car at every battery swap station--probably means a lift or a pit unless it's an SUV.

Pack swap is a well tried and tested but failed technology, mainly commercially. Better Place tried it. Tesla tried it. Maybe Tesla succeeds, but they don't really need it because they have the Superchargers.

I wonder if a better solution than battery balls would be to use a standardized cylindrical cell.  You could still have a large 'tank' of arbitrary size/shape that only needed a small outlet and inlet opening for getting spent and charged cells out and in, but cylinders will rest and pack more densely and predictably than balls, and can have larger contact surfaces on their ends.  You could have the two contact-bearing sides of the walls press in to provide high contact force once the cells are in place.  Contacts can be arranged to give whatever series/parallel arrangement is required.

Good thinking, it has been tried. I don't have a link at hand but patents describing exactly that have been filed. It's mechanically a lot harder than what we are doing. We seriously considered it as an option.

The big challenge is feeding them in and out reliably; without a proper set of guides the cylinders will be even more prone to jamming then balls.  Maybe having them snap together into a belt would be beneficial, with internal sprockets to help feed them into place.

Also the size of the system is problematic, so is the swapping speed.

Mechanisms for feeding cylinders in and out are already highly evolved in automatic firearms. Look for a design called the HK73.

That's a relevant reference actually. It isn't however implemented as easily in practice for a battery.

Aluminum, Gallium and water to power Hydrogen fuel cells, the byproduct is aluminum oxide (alumina) and the Gallium is reused. At the "pump" you can recover the alumina that can be converted back to aluminium using solar or wind energy so that new aluminum pellets are available for the next customer.

http://www.purdue.edu/uns/x/2007a/070515WoodallHydrogen.html

Not sure where this research is at, there are several companies trying to bring it to market with their own versions of the reactor to produce hydrogen on demand, to be used directly or via fuel cells to produce electricity.

I wish them the best. We pick the winning technology, at the moment it is Li-ion.

The real EV batteries (not these tesla notebook ones) all have screw terminals and all use busbars and thick metal plates to make the connections. Replace it with two rounded surface touching each other randomly and you are asking for trouble.

Not such a big deal-those surfaces will be.... spot welded in rush current and they will have useless EV wehicle full of soldered metal balls and sparks will destroy NASA mission to Mars while jamming their radio equipment 

I think Tesla won't agree that their batteries aren't real. And Nasa isn't buying, now I know why!

I wonder if a better solution than battery balls would be to use a standardized cylindrical cell.

...like an 18650?

 

Welcome to the forum.

Thank you! As I said elsewhere, we're big fans of the EEVblog. Quiet ones though. Except maybe today?

Your tone is a little upsetting.

Apologies, that was not intentional. In my defense, I'm sure you understand that people calling us stupid (and our baby ugly!) doesn't bring out the best in us. We have a much nicer side though, which will surface over time.

I haven't seen anyone here say that it can't be done or that it is simply BS.

I have, but that's ok. The vast majority of people though are open for a reasonable discussion so it's for them that I spent the last few hours typing here. It's not the best investment of my time for our young company, so see it as an engineer's disease that can't be cured.

But they are questioning why it is better than other simpler ideas, that do not require an array of new technologies to be perfected at the same time.

Fair enough! One of the things that is cool about our approach is that it works with manageable technology and complexity for today, which makes the technical and business risk acceptable. It uses methods that don't "belong" in the battery business, but that's the new insight which helps us ahead.

Engineering is not, like politics, the art of the possible: it is about finding solutions that are better and cheaper, not "if you build it, they will come" faff.

I could not agree more. We looked at Better Place, analyzed their flawed business model, and built a new perfected business model based on technology we have within reach.

To give you an example, wireless power in the home, such as the WiTricity concept, is obviously possible. Is it economic to install at this time, or in the foreseeable future?

Total agreement here. When I have urges to build something because I can, I design a receiver or whatever that has no commercial viability whatsoever. That does not apply to what we do  here.

[Ian.M]

Compare with commercial distribution of compressed gasses:

There are two basic options: rental cylinder (owned by the distributor) or customer owned cylinder.  In most cases the cylinders are identical apart from a paint-job and the distributor's stamp on the neck of a rental cylinder  Both must be hydro-tested at regular (but fairly long) intervals, the owner being responsible for testing costs.   Rental cylinders are usually simply exchanged for a full one, though dealers may offer an exchange on a customer owned cylinder if they stock the same size/type.

Presumably the customer would rent the battery modules and the distributor would be responsible for depreciation and maintenance costs as long as the built-in dataloggers didn't report abnormal external conditions during storage or use.   Probably a guaranteed minimum trade in value backed by a 3rd party bond would be required so customers could be assured of the percentage of their deposit they'd get back at end of vehicle life.

[DanielS]

Anyway this
http://en.wikipedia.org/wiki/Vanadium_redox_battery is for years in industry and... charged fluid can be simply replaced with used one, so no need to mess with bloody  balls battery 

Except VBR has energy density roughly on par with lead-acid. Also, I am not convinced it would be safe to have people draining and refilling tanks containing a relatively strong Vanadium acid and base. I bet people cross-contaminating their positive electrolyte with negative and vice-versa would be an issue too: even if it does not harm the cell stacks themselves, it will degrade energy density and performance until the contamination gets flushed out.

Look at the size of those 175kW cell stacks in their videos. Even half that size to provide about 100HP worth of electrical power would still use quite a bit of space, add quite a bit of weight and that's without counting the  50L tanks (or more) for each electrolyte, the pumps, nitrogen extractors to flush other gases out of the system to prevent corrosion and whatever other support infrastructure the system needs.

Sounds great for stationary application where size and weight do not matter much and the system becomes a closed loop after commissioning aside from occasional maintenance but not so much for individual mobile use as a frequently opened loop.

[Bert]

Could you comment a bit about tank size requirements and how it is supposed to fit in a car? What kind of capacity and voltage can you expect from single module (5-10 Wh at 3.7V)? Is the outer shell size determined at this point and will be fixed in the future?

The standard design uses 3.7V Li-ion chemistry. Capacity scales almost linearly with size except in the smallest models. Our customers could order (or make) them in any size roughly between a hazelnut and a lemon. So far however, no one has asked for a different size than the standard 42mm one. Capacity varies as stated before, but low double digit Wh is the right ballpark.

The string battery (the enclosure) is sized to provide the required capacity. I cannot provide detailed numbers, but the energy density per volume unit is in the 300Wh/l range. So 30l will give you 10kWh, a bit less for cheaper batteries, a bit more for the dearer ones.

Can I have few of those assembled PCB for Easter? Those modules look like very funky eggs

That depends on which customer you work for

I have not thought about dynamically enabling/disabling  separate cells to improve overall wear, charge levels and other parameters. In comparison, big sealed single battery pack is not that flexible - some cells wear out unevenly and cannot be replaced in the future.

With this in mind, even single EV with such technology and little external  infrastructure could benefit from improved battery management.

Correct analysis.

I appreciate that you are responding point by point to comments. That is really everything that should be expected for PR in an engineering forum, so good job there.

I won't be doing this for a long time but for now all questions are fair game! Excuse my brevity due to the large volume I am going through.

I still don't understand how you can say that the packing is space efficient. Surely if it was, then every large storage battery would contain rounded instead of flat elements? I suppose you're including the space taken by the bus bars, cooling channels, module BMS, and module rails in an EV pack in that comparison.

Correct. It is a pack-for-pack volume and mass comparison.

The comment about efficient cooling is based on teardowns of Tesla battery packs, which are liquid cooled. I could be wrong, but liquid cooling over the cells only seems feasible if connections that lie inside the liquid volume are permanently welded. You might be able to use gas-tight connectors but they add cost (expensive connectors but even more in assembly time cost). Liquid cooling also seems incompatible with an air delivery system. So the alternative is air cooling, but that requires a higher W/K cell package to achieve the same system efficiency.

Good thinking, but keep in mind that ours have a large metal outer surface with excellent thermal conductivity.

Welcome to the forum Bert, I originally read about TankTwo on Electric Motorcycle Forum.

Thanks, and I didn't know we were covered there! Thanks for the coverage.

Glad to have helped in highlighting your product. It is a original and unique idea,  just does not seem practical / economical unless you planning on a monopoly like the fuel industry has.

Not really, because you can swap any EV battery with our battery, forget about the swappability and the idea still works without that one feature. In the same way as many Tesla Model Ses support pack swapping but won't ever use it.

http://electricmotorcycleforum.com/boards/index.php?topic=4496.0
By the way the Zero motorcycles have the highest capacity/kg of any battery pack 184Wh/kg air cooled, compared to Tesla Roadster 117Wh/kg and Model S 140Wh/kg liquid cooled.

And that depends how you measure it (do you include the coolant circulation system, etc.)

Same story here with the specs in that they vary, but we do a little better than the numbers above what the specific density is concerned. Air cooling does limit certain sustained high peak power cases, such as the P85D requires at full acceleration, which is not feasible for a long period of time without liquid cooling..

We needed specs for the most important part of your smart battery, the cell capacity. What is your expected Wh/kg?.
Then we will be able to make fair jugdement. Every week there is a new upstart or lab promising break through in new battery technology.

I said above already way more than I should.

Yes, we've been promised the sky (and then some) many times over. What we're doing different is that we don't require any nobel prize level progress in chemistry or any other magic to happen, it is just a different way of handling things.

GTG, more later.

[firepower]

By the way the Zero motorcycles have the highest capacity/kg of any battery pack 184Wh/kg air cooled, compared to Tesla Roadster 117Wh/kg and Model S 140Wh/kg liquid cooled.

That ^^^ should set some alarm bells ringing i'd hope, as it tells you something very important about passenger car batterys!


(namely, that even a company as forward thinking/innovative/cash rich as Tesla, has had to compromise on Energy density in order to meet practical and legislative limitations(and use a liquid cooled pack, like EVERYONE else........)

motor bikes don't have the luxury of size and volume of cars, hence they using different battery technology and packaging, Tesla deliberately choose 18650 cells and use 7,140 of them because they are the cheapest cell available in volume, used in power tools etc.

Zero cells are custom design.

Zero gives a 5 year 100,000 miles warranty on battery pack system.

Here is an interview of Zero design ideals.
http://www.cycleworld.com/2015/03/25/on-the-record-abe-askenazi-cto-zero-electric-motorcycles-cycle-world-interview/

[BlueBill]

Bert.
How many Wh is each cell? (300Wh per cubic meter?)
How many kWh for a typical tank? (the Tesla has 60 - 85kWh under the seats)
How much does a typical cubic meter of cells weigh?
How long to recharge in the battery tower refill station?
How do you insure the batteries are topped off when you refill?
How does the tank put pressure on the cells to maintain good contact?
Does the tank work if let's say 50% full?
How do you locate bad balls? I assume they're sorted during the tank empty process.

IMHO.
They only features I see are 3min fillup time and and dead cell replacement. You'll need distribution stations and you're going to have a near impossible task in the US (look at the crap that Tesla has to put up with in many states). You'll also need auto manufactures on board or you have zero chance go getting off the ground (do you have any confirmed vehicle manufactures on board?). It's also going to need real world testing, that'll take significant time and money.

As battery technology improves and it will, a typical EV will likely average 200 or more km per charge, for most commuters this is plenty. Longer range consumers will opt for hybrids as they do now and fill up on fossil fuel once a week. 2016 is likely to see Hydrogen vehicles getting hyped by manufacturers over EV.

A far better solution is a replaceable battery pack, it's been tried but because it adds cost & weight it's a hard sell to money conscious consumers it's down the ladder of must have auto accessories.

My crystal ball sees no future for battery balls. It does see self driving cars that don't need handholding to keep their cells topped up. It also sees inductive coupling at intersections so EV can get a quick top up if needed.