Hacker News Reader

Comment by joshe | Hacker News Reader

The article isn't very clear.

The car still has a battery, this goes in the stationary charging system. It's used to increase the current when charging electric batteries. Most electric batteries can accept much more current than commercial electricity drops can provide. This system stores the current in stationary flywheels and then discharges it quickly when a car pulls up and plugs in.

So this provides something like a Tesla super charger, without having to call up the power company to rewire your gas station. They call it a "Kinetic Battery". In the video below, the CEO makes the claim that flywheels are much better for this because you can get many more discharge/charge cycles out of flywheels than with chemical batteries.

You can imagine that rewiring electric infrastructure all over the country would be quite a bit more expensive than just plugging in their system. You could also imagine a solar powered system (off the grid even) slowly being charged and then recharging a car in 10 minutes.

Cool idea!

Interview with CEO https://www.youtube.com/watch?v=YxRfPtYmTDE

View on HN

> This system stores the current in stationary flywheels

Well, rotating flywheels.

It amazes me just how much energy you can story in a flywheel. They're really neat devices.

I believe Verizon has a datacenter that has a number of 15 ft concrete flywheels that provides emergency power.

This is a pretty common way to bridge the gap between when utility power drops and when emergency generators are started and ready to handle the load. I believe flywheel systems can be more compact and don't have the same maintenance demands, but may be more limited as to energy capacity and duration of service.

Didn't Twitter have one of these on its roof in SoMa? I vaguely remember reading that somewhere.

What is amazing is the rotational speeds involved. Say the flywheel weighed as much as a tesla. Think of the energy. Say you want to store enough energy so that a tesla can accelerate itself from zero to 100mph ten times. That amount of energy in a flywheel weighing the same as the tesla would mean the flywheel spinning at +/- 1000mph. Either this flywheel weighs hundreds of tons, or it is spinning at supersonic speeds. Either way, you better have some good bearings.

The magical word here is „angular momentum“. The energy that a chunk of mass can store in a flywheel is squared with the distance of that mass to the center of its rotation axis. If you distribute the Tesla sized mass around the edges of a 4m sized Flywheel now (spare a few % for the wheel and structural support), it can store quite a few dozen times the energy of the same sized Tesla going the same speed.

I'm pretty sure it doesn't matter whether the eneergy is stored in rotating body, or one moving in a straight line. Kinetic energy of rotation is 0.5 * I * w^2, or for that extreme case of "concrete ring" - 0.5 * m * r^2 * w^2, which is exactly equal to 0.5 * m * v^2, or kinetic energy as calculated for body moving in straight line.

Are you sure that you aren't confusing force with energy? Angular or not, it's just accelerated mass.

Modern flywheels spin extremely fast in a vacuum, so supersonic is moot.

So let's do a little high school "perfectly spherical cow of uniform density" physics here.

Imagine we take that 1 Tesla mass and turn it into a infinitely thin ring of infinite density material. Now we can calculate how fast it's need to be spinning to get all that mass doing "1000mph".

Lets make our infinitely dense 1 Tesla mass into an infinitely thin ring 1m in radius. Each point on that ring moves around it (2pir) circumference every revolution so ~6.3m per revolution. 1000mph =~ 440m/s SO we only need to spin this flywheel ay ~70 revolutions per second, just over 4000RPM.

4000RPM doesn't seem unreasonable...

(Note:I don't think your 1000mph stacks up though. The kinetic energy of a moving mass is 1/2mv^2. So the speed of a 1 Tesla mass object which has 10 times the energy of a 1 Tesla object doing 100mph is only sqrt(10)*100 closer to 320mph than 1000. My high school physics flywheel might only need to spin at 1500RPM...)

When I was in middle school? maybe high schoo? 1990s somewhere, there was a long article about flywheel batteries for cars, they were claiming some very impressive range and power numbers.

I remember it was one of a million things I read about in sciencey magazines that never went anywhere, but maybe it was the one I was most sad about never happening (at least so far)

The one product I do remember happening that I first read about back then was e-paper! That happened! Cool.

They do (or did) exist.

https://en.wikipedia.org/wiki/Gyrobus

Was it Omni magazine? I think I read the article as well. If memory serves, they discussed a flywheel for each wheel of the car?

I always wondered how they dealt with gyroscopic forces in flywheel vehicles. Maybe I'm being naive and the effects can be compensates for with gimbals, but I'm stuck with the mental image of cars that don't turn corners

-I've seen a small-ish one disintegrate under test.

(Some 480kg of mass in a ~600mm diameter configuration, at some 8-9000rpm maximum if memory serves)

It amazes me how hard it is to contain that kind of energy if it wants to get out. They're really terrifying devices.

Particularly when left in the hands of engineers who do not fully appreciate that any sufficiently rapid release of stored energy is indistinguishable from an explosion.

The ones in Formula 1 cars had 140.000 rpm, but not much mass. It's terrifying indeed, so much that Kubica refused to sit in such a thing.

Having seen a flywheel + clutch assembly disintegrate on a drag car in the lane next to me (10,000rpm - bits of flywheel came through the bodywork of his car), I can understand why you'd be hesitant to sit on top of something spinning at 140,000rpm.

Isn't that fairly typical turbocharger turbine speed though?

I saw a Mazda RX3 Sport Sedan (a circuit racing class here in Australia ~25-30 years back) spectacularly blow a flywheel/clutch at the end of the main straight (at Oran Park) a long time back - it pretty much perforated all the bodywork in the plane of the flywheel - bonnet and both guards - as though it was a tearoff line on a set of stamps. We could hear bits landing 10+ seconds later hundreds of meters behind us in the parking lot. Depending on what port job he had, that could have been up around 12k rpm (if my vague 30 year old car modifying memory serves...)

> Isn't that fairly typical turbocharger turbine speed though?

I believe so, but compressor/turbine wheels don't have a lot of mass to them, so fragments don't usually escape very far if they blow up. I've seen a few compressor housings split, and the main damage is done to the engine itself really. I guess if your turbo was sitting high up and exposed, it might be a danger. Flywheels however just seem to fail more regularly from my experience.

(Also, yeah anything rotary powered deals with scary rpm's, haha!)

For anyone interested, here's some in-car footage of what we're describing here: https://www.youtube.com/watch?v=qOHUwL0zeXc and an image of what the result can look like: http://images.thetruthaboutcars.com/2011/04/LNJ11-BlowedUpPa...

He wasn't sitting on top of it, it was rather behind his ass. He either felt the vibrations or the hizzing sound, and that made him very uncomfortable.

Figure of speech, apologies.

I was thinking about this the other day, why not use a liquid as the mass? Say water? If the spinning disk is treated like the tanks in a liquid tractor trailer, with slots to divide the mass up evenly.

When the disk starts spinning all the liquid turns into essentially a solid, if there is a massive failure, simple venting could easily disperse the liquid 360 degrees or away from people, etc... (ie, a giant massive mist explosion)

I've worked with power washers that are dangerous when focused, but when the water is dispersed, it's practically harmless.

Is there some obvious thing I am missing here?

Density, cost, complexity.

That's too bad. But that sort of describes every technology at one point in time...

Fluid flywheels are used for hydraulic couplings in automobile transmissions.

They're not good for long-term energy storage, since the fluid loses energy to friction and turbulence much faster than a solid-- think about how long a swirling vortex lasts in a cup versus how long spinning top can continue.

Ya, I was thinking that the water wouldn't be sloshing around, only being inside a sectioned donut wheel. The water could be hermetically sealed (with some kind of preasure relief for temp changes) in chambers, no sloshing or friction issues. Externally it would look exactly like any other fly wheel.

The thought being that since it's a safe medium to explode out, you could spin it faster than other mediums. The steel would just be a shell to hold the water, and therefore the containment vessel wouldn't have to be as robust, as the amount of steel that would fail would not be as much, and as soon as the water exploded out , the remaining steel would dramatically loose kinetic energy and possibly not even flyout.

Another comment said that 20k rpm would hold 4x the energy as 10k rpms. That implies that if you truly aren't afraid of the spinning medium hurting anyone, it would seem that water/steel hybrid would be better than solid steel, because you could spin it faster more safely.

That is not a flywheel.

Power-density and energy-density.

The "secret" to dense flywheel energy storage is in the equation: Energy = 1/2 * I * w^2

Where "w" is rotational speed in radians/sec, and I is rotational-inertia / rotational-mass. So you store the most energy by rotating it as fast as possible. 20,000 RPM stores 4x the energy of 10,000 RPM.

The issue with water, or really... anything outside of steel and other such incredibly strong materials... is that water doesn't hold itself together very well. As such, the RPM limit would be smaller than a steel-flywheel.

Indeed, steel is still not used in advanced flywheels, because some forms of plastic IIRC are superior at the kinds of stresses we're talking about. Thus, these plastic / composite material flywheels, while much lighter, can spin much faster.

Leading to more efficient energy storage.

-----------------

In effect, you need every part of the flywheel designed to hold itself together to reach maximum efficiency. Water is the exact opposite. Its a lot of mass but basically no ability to handle stresses. Solid Steel would be better and would spin faster.

That is really interesting, thanks for explaining it.

I want to ask then, based on other comments, that size can greatly improve the amount stored. So wouldn't a larger safer flywheel work in many places?

Also, my thought on the water as a solid, consider a steel wheel where instead of steel all the way through, put a liquid inside of a cavity. Since water doesn't compress, it would essentially be part of the same wheel, only safer in a collision? (arm-chair engineer)

I enjoy reading comments like yours so I can move on/let go of my from my silly pet ideas. :)

> Also, my thought on the water as a solid, consider a steel wheel where instead of steel all the way through, put a liquid inside of a cavity

Well, consider that solid steel is considered too weak for a modern flywheel. So that questions why you'd be removing steel instead of adding more of it.

Modern flywheels spin really, really, really fast. Again, fast enough that modern solid steel is too weak. Newer composite materials are superior, but still need to be designed for maximum structural integrity.

Once a flywheel spins fast enough, it literally rips itself apart due to centrifugal forces. Any "cavity" weakens the structure and will break sooner.

Carbon fiber composite (not exactly plastic) is strong and light, which is what you want for a flywheel.

There a few more relevant expressions for flywheels to express that you want them to be on the edge of breaking at maximum energy storage. http://large.stanford.edu/courses/2010/ph240/wheeler1/

You don't want light in a flywheel. You want mass.

Well, in the context of storing energy, you want mass in the flywheel. Having a lightened flywheel is beneficial in, say, racing applications however... But that's not really important in this discussion.

No - as the immediate prior post correctly stated, stored energy increases linearly with m (flywheel mass) and as the square of rotation rate (omega^2) So a flywheel having 0.1m rotor mass spinning at 10omega stores 10 times as much energy as a flywheel of 1.0m spinning at 1omega. 10% of the mass, 1000% of the stored energy of the heavier, slower rotor.

Of course, the rotor has to be able to handle increased circumferential and radial stresses resulting from increased omega. Carbon fiber (like T1000 grade) has extremely high tensile strength (lets you spin a carbon fiber rotor very fast) but low density. T1000 rotors store more energy at lower mass than steel (or any other metal) because of the proportionality to omega^2 and their high tensile strength.

But for commercial storage, it's really about the $/kWhr/mass in terms of the overall rotor economics. By that metric, it's hard to beat a composite rotor with E-glass as the major fiber. It's cheaper than carbon fiber by easily 100X and has about 30% to 50% of CF tensile strength. It's a bit denser than CF, but its cost metric is why most modern large energy storage flywheels use glass/CF hybrid composites and magnetic bearings (for high speed and zero wear).

In the old days, they made flywheels out of solid steel, then switched to piano wire (higher tensile strength). Then fiberglass, aramids and carbon fiber happened.

It depends - the "v^2" term quickly passes the linear "m" term for some scenarios. If I've got a choice between making it ten times as heavy or capable of spinning 10 times as fast, I don't want more mass there.

I'm surprised that Aluminum wins over Steel in this use case. I figured Steel's higher strength would be more important, but I guess the increased mass makes it have issues.

Thanks for the additional math reference.

You are missing something (but don't worry, it is good to brainstorm like this).

At maximum speed (maximum energy storage) you want every part of the flywheel to be about to break. By introducing a non-structural material (liquid water), you have added mass but not the strength to get to higher speeds.

Imagine filling a bucket with water and spinning it around your body. All of the force to keep the bucket from flying away has to be held by the handle of the bucket. A better bucket flywheel would be all handle... Hopefully someone else can give you a better analogy.

>A better bucket flywheel would be all handle...

Ok, that makes some sense... still not feeling it, but I can see the logic. Need more math in my education perhaps.

Aha, a whole new meaning for RUD.

When I used to work on turbo generators, the kind you find at power plants I saw carnage of what happens when those blow up. The retaining rings on the ends, typically a few thousand pounds were prone to breaking on older generators because of the alloy used. When that would happen it would send the projectiles through steel and concrete walls.

The balance bunker for load testing and balancing these at the shop I worked in was a pit 40 feet underground with 50 ton caps placed over top.

Velkess (bill gray) solved that problem before they went bankrupt.

The answer is to make the flywheel from a flexible kinetic lasso such that when it blows up, the forces tangle the parts in a manner that it is safely destroyed.

> Most electric batteries can accept much more current than commercial electricity drops can provide.

Just how much headroom is there? I'm under the impression that avoiding excessive heat is the key to battery longevity, so the question is whether or not this extra current increases heat levels to degrees that will have a measurable negative impact on the long-term capacity of the battery.

Depends on C values, thermal limits, etc.

The single data point I have of 75k miles, once a week charging at 120kW from ~20% -> 60% shows very little degradation to the pack on our Model S.

Also, when it's warmer out I've seen warnings that in-cabin AC is lowered to divert to help cool the pack so I'm sure it's a question of thermals + number of parallel modules.

Accordingly to Musk the limit charge power for the Tesla in the future superchargers is around 200kw if I remember correctly.

But Tesla limits supercharging after too many cycles:

https://electrek.co/2017/05/07/tesla-limits-supercharging-sp...

That's still a lot of DCFC(~20k miles based on 0.33kwH/mi) and even then it's limited down to 90kW from 120kW.

FWIW you only see 120kW on the first ~35% of the pack and quickly tapers down to sub 100kW after that approaching home charging speeds for the last 80-100%.

I believe that turned out to only be for one particular revision of the battery pack.

TLDR; Faster chargers are useful now and more of them would be great!

The cells charge in parallel (there are on the order of 15,000 cells in car batteries), so at least right now the charger power is the constraint and not the battery chemistry.

The most power you'd ever really need to deliver is maybe 3-400 miles of range in 5 minutes. So at a point, that we are pretty close to, you don't really need to charge any faster.

Using Tesla as an example of our current charging constraints:

The Level 1 chargers that plugin into a normal outlet are at 2 kW.

In-home charging (Level 2) maxes out at about 17 kW, (close to the most the average house circuit can handle). An average gas station or retail parking lot is probably close to this. More likely it is down at 11 kW.

At Super chargers, the Model 3 can currently accept a max of 120 kW, and the chargers can put out a max of 200 kW.

Tesla has teased doing more, including a Megacharger for their semi's, that seems like it would work at more than 1000 kW. See https://www.teslarati.com/tesla-semi-megacharger-charging-po.... And also "Super charger V3" at something more than 350 kW.

Very roughly, kW is proportional to how much range you get for the amount of charge time. So 120 kW charging is about 10 times faster than 11 kW Level 2 charging.

Also very roughly, a Tesla Model 3 could charge it's 300 mile (75kWh) battery in about 5 minutes at a 1000 kW charging rate.

It is worth noting is that this particular need, dumping 75kWh of power in 5 minutes, will exist even as battery technology improves. Fast charging is not gating on electric cars because home and work charging works quite well. But there will be demand for fast charging as a convenience to electric car owners.

Good links and details here: https://en.wikipedia.org/wiki/Tesla_Supercharger

According to WolframAlpha that 75 kWh over 5 minutes equals 900 kW or about the power developed by the most powerful road going car - Bugatti Veyron Super Sport.

You _seriously_ needed Wolfram Alpha to divide 60 by 5 then multiply 75 by 12???

Unit conversions and arithmetic are computers' work.

Tesla already does this in some locations, using stationary li-ion batteries to even out demand with their Superchargers.

-We're about to try out the same idea locally, on a larger scale - ferries are about to go electric, the grid is not stiff enough to handle the peak load (4.7MW for 5 minutes every 30 minutes) and hence bunkers (for lack of a better term) containing tons and tons of li-ion batteries are needed to make for a sustainable draw from the grid.

The battery storage facilities won't win any architectural awards as they are currently envisioned, but I can't wait to see whether this will actually work reliably; our first few attempts at electric ferries had some teething problems, mostly with the charge process taking too long to establish. The ferries run on a tight schedule; hence charge time dropped significantly as establishment was slow, and ferry batteries slowly depleted over the course of the day.

Anyway - 4.7MW translates into some quite massive shore power connectors, and they need to mate quickly and reliably - and when the weather gets rough, everything in the surrounding environment will be soaked in salt water. Interesting times.

That’s really cool. A seaworthy quick connector capable of handling 4.7MW sounds all kinds of exciting.

Sounds like a potential use case for flow batteries.

> Most electric batteries can accept much more current than commercial electricity drops can provide.

Current or voltage? I thought Telsa Superchargers use higher voltage to charge quicker. More "pressure" (voltage) versus more flow (current).

It would most likely be current in this case. 480V is available commercially, based on a quick search. Tesla batteries seem to provide around 400V.

To fully charge an 85 kWh battery in one hour, using a 480V supply, you would need to draw over 177A, assuming perfect efficiency. That is an absolutely ridiculous amount of current, and that would only charge one vehicle.

My house can load 105A. I sometimes use 120 kWh on a really cold day, so it's not peaking anywhere near that current, at least not for long.

I'm not driving 400km every day so the grid will definitely handle it. But there is a lot of potential to reduce network costs if the charger negotiated with the grid.

Your house can load 105A, but that's most likely at 120V/220V, and as you said, it doesn't maintain that for long.

Based on 120 kWh over 24 hours, your house, on a cold day, draws around 500 W. To fully charge a single 100 kWh battery, like you'd find in a new Model S, in an hour, you would need to draw 100 kW. That would be 200 times the average load on your house, drawn continuously, and that's only to charge one vehicle.

Something like a flywheel battery does have a lot of potential to save money, or even earn money, if it communicated with the grid. A large rotating mass could potentially store a lot of energy, and would be increasingly useful with the rise of renewable energy.

120 kWh / 24h = 5000 W, not 500

Many houses in cold climates using electrical heaters will actually use at least 2-3x this energy.

So yeah, 100 kW is quite streched, but we're not _that_ far from the power available in a home.

You're correct, my mistake.

Still, that would be 20 times the average load on those homes, to charge a single vehicle. To move that power 100 feet, like from a busbar (I think, I'm not an electrician) to the charger would require a 3/0 AWG copper cable (which has a diameter of around 2 1/2", or 5.8mm). That cable costs around $2.59 a foot, or $259 to get power to that charger. That would be undersized, since I calculated it out at exactly capacity, and that price wouldn't include conduit (since the cable isn't direct burial) or installation.

Multiply that by three chargers, which I'd consider to be on the very low end of what a commercial installment would need, and you have quite a high demand for power. Five hundred amps at 480V is no joke. On the plus side, all this could spur adoption of renewable energy; with enough demand for power, investors will shovel money into renewable energy faster than it can even be used.

It doesn’t really matter. The charger will convert whatever voltage it’s supplied with to the voltage needed by the battery at any given moment. You could theoretically run a fast charger off 120V, you’d just need massive conductors to handle the massive current.

Is it conceivable that the same type of battery could be used in cars as well, or would the capacity/falloff/external forces be too limiting?

Gyroscopic forces are a big problem as is trying to get enough mass.

Angular momentum increases linearly with mass but with the square of the radius. So another key solution to capacity is to increase the size of the flywheel. Obviously there are real limits to this in a car.

I could see it being used as a "quick-charge" portion of the capacity. i.e. you have a full-sized battery, but also a small flywheel so that if you're really in a pinch you can rapidly charge enough to get somewhere without waiting for an hour. It'd be like having a small SSD paired with a large HDD.

You'd get better results using the same weight for more batteries.

https://en.wikipedia.org/wiki/Flywheel_energy_storage#Transp...

There have been experimental race cars from Chrysler, Porsche, and others, using flywheels in hybrid systems.

I think it could be useful in congested cities with lots of stop and go driving. If there were magnetic flywheels in the street that mated with magnetic flywheels in the vehicle, energy could be transferred at stop lights. Provided the field didn't move the whole car, and if gyroscopic forces were mitigated, as another poster pointed out

There are a ton of batteries used for this application already -- regenerative brakes combined with a small battery is a "mild hybrid".

No.

Flywheels are useful when you have a fixed installation with no weight constraints because you get certain benefits over LiPo cells, namely longevity. As a matter of energy density though, they are awful compared to LiPo and could never efficiently be used in a vehicle. Also, the gyroscopic forces would be such that it would be impossible to steer the vehicle.

> the gyroscopic forces would be such that it would be impossible to steer the vehicle.

That's a solvable problem, and has in fact been solved by F1 teams that have used it - Mclaren won two titles with their KERS equipped cars. The cars are demonstrably far from 'impossible' to steer...

All you have to is mount two of them rotating in opposite directions to cancel it out. The axles need to be able to handle the load but the rest of the car sees nothing.

But a KERS system stores only a fraction of a kWh.

I think the first year F1 had kers some of the teams used flywheels, but I think since then everyone seems to have moved to standard battery pack type systems, so I assume there are downsides to them.

At least if you drive a bit above the speed limit when commuting :)

Thank you for explaining clearly what article failed to mention. I assumed some of this, but not the whole.

So now they have a cache invalidation problem, right? When do they move power to the flywheel (friction expires the electrons) and when do they allow for cache misses to hit their connection with the power company?

Reply on news.ycombinator.com