> This is the first time a project of this kind will be used anywhere in the world and ESO believes it could be a “huge step forward” in running a zero-carbon electricity grid.
I'm trying to find some stats on this, because flywheel UPS aren't a new idea:
What is different in this case is it is not really an energy storage system (at least that is not the main intended purpose), but only to provide stability to the grid frequency. A huge flywheel in a vacuum to provide a reference to millions of distributed PLLs. I don't think the idea is to always have it spin at a fixed rate with minimal losses.
In the case where a flywheel is used as a UPS, the rate at which it is spinning will vary in accordance with the stored energy.
The point here is that the flywheel is so massive that the amount of energy required to bring it below (or above) the allowable frequency range is practically very large.
Large steam/gas turbines running on the grid have a similar effect. Even without hot gas running through them, they can provide rotational inertia for grid stabilization purposes. This flywheel is basically an extremely heavy version of the same idea, and powered exclusively by the grid itself.
You wont be able to ride out a scenario where there is a long-term lack of generation capacity, but this flywheel could buy the precious seconds/minutes required to spin up peaker plants and other contingencies.
Rotational energy goes up worth both the square or the radius and the speed, so it is much better bang for your buck to have a the mass centered on an outer rim connected to a hub by spokes and spinning really fast rather than just something uniformly heavy. The mass close to the centre doesn’t do anything for you.
The heavier it is also requires bigger more expensive bearings and accompanying system to get the thing spinning in the first place.
I looked at a flywheel that spun at 12,000 rpm. It was going to be located in an underground vault in case it ever got off its pedestal bearings it wouldn’t mow people down.
Like anything that stores energy but doesn’t generate the economics aren’t great.
You can get constant spin rate but variable energy storage / release by adjusting the moment of inertia, e.g. if you have movable weights within the rotor you can move these inboard or outboard, moving it inboard uses energy while increasing the spin rate, and moving it outboard releases energy while slowing the rate. Using this in combination with "conventional" direct axle energy input / release means you can store or release energy from the system as a whole while also keeping a constant rate (within the bounds of the total energy storage capacity of the system).
Is this used in practice? It sounds like a rather substantial engineering challenging considering how much energy the wheel is storing. I remember MIT had one of these for sparking the old tokomak, they had to plan for it to fall off, destroy several buildings without killing anyone, and then land in the river... and that was a solid wheel that did nothing but spin.
But isn't the point that it provides inertia towards stabilizing the frequency? I don't think it matters whether you can vary the energy stored if it's just acting as a, well, wheel... if the goal is to give you a bit more time to bring more generation online or take it offline then it seems like a simple system would be fine. It's not storing power, so it can only ever slow the drift to give more flexibility.
As you say, I suspect it probably isn't used much in grid scale applications. The device image in the article would imply that it is just a simple damper, i.e. the angular momentum acts to reduce any frequency change, up or down. It doesn't look like it does any storage, but then the article is somewhat content free.
You've seen what your washing machine does when you get a blanket off-balance in it.
Imagine if one of the motors on the flywheel stopped responding to inputs.
Windmills had governors which were horizontal wheels with two or more pendulums attached to the outer edge. The angular velocity is trying to lift the pendulum and gravity is pushing it back down, so velocity increases much slower than momentum.
The only problem you need to worry about there is seized bearings or foreign bodies unbalancing the load.
There are other devices that replace gravity with springs. You might be able to do the same with permanent magnets. But I believe each of these relies on the device being able to put up with larger strains than the pendulum.
It's kind of off topic, but your description reminded me of the executioner's sword "Terminus Est" from the Book of the New Sun series. It had a liquid metal (mercury I think) core that would collect at the base when held upright, making it easy to hold. When it was swung, the liquid would flow to the tip, putting more weight at the point of contact.
I've seen a system meant for hospital ride-through that packed an amazing amount of power into a relatively small enclosure using carbon fiber wound flywheels. For short durations you can get away with this because you're only going to operate one of those. Larger installations would need multiple of those and then you need to get very serious about containment so that if one of the flywheels ever fails it doesn't set off a chain reaction.
The nice thing about the fiber would units is that when they break they self destruct completely within the enclosure. The fiber acts as energy sink by breaking up, as opposed to a steel wheel that will spit out sizeable chunks across fairly larger distances (many meters).
I remember back in the early 2000’s my father was investing in a company building a flywheel. They were working on installing one in New York, not really sure what happened but I believe they installed one or two before the company went bust.
Yeah, I remember going to work with my dad (~20 years ago) at a datacenter for a large national bank. Power to the building was supplied by three large flywheels, which were in turn (usually) kept spinning by electric motors driven by street power. If they lost outside power, they had three diesel engines that they could spin up and clutch in before the flywheels spun down too far.
Agreed. I did a napkin-back calculation on a giant (50m diameter) grid-scale flywheel back in grad school in '95 (when there was talk of powering cars with them) and failures would have been catastrophic since the flywheel would dissipate all of the energy nearly immediately (in comparison to gasoline). A car powered by a flywheel (0.5m diameter) would have been a guided missile... Fortunately, grid-scale flywheels can be placed in out-of-the-way locations.
The spinning flywheel doesn't just spin on its own. It gets spun up and kept spinning by an electric motor/generator I have to imagine. So you're limited as to how much power can go in or out by the size of that motor/generator.
If it's a 100kw peak electrical machine then it can provide at most 100kw of instantaneous power to the grid while at the same time slowing down by however much is required to deliver that power.
If I understand correctly the idea is a lot like capacitors for DC voltage. The DC voltage has to go down in order for the capacitor to supply current, but the capacitor ensures that the voltage can't go down as quickly as it would otherwise.
Thats not the theoretical limit though. This is a physical device with moving parts. What if the bearing fails? What if the weight disintegrates due to a manufacturing fault?
Suddenly all that stored energy wants to go somewhere.
The flywheel will have breakers to prevent excessive inflow or outflow of power.
In a conventional plant, the spinning machinery is monitored by vibration sensors. If the vibration exceeds some threshold, a "Turbine Trip" occurs and the offending generator is disconnected from the grid. It then slowly spins down, dissipating energy via it's own friction.
It's a synchronous motor which normally idles at the Mains frequency, but if the Mains frequency attempts to change, the rotor inputs massive amount of current (leading or lagging) into the mains which prevents the change.
There is no need for a drive motor.
In the industry they are known as a "Synchronous Condenser" or Syncom. They have been in use for a long time.
So are you saying that the entire mass of the whole flywheel is all the synchronous motor?
I guess I had assumed -- especially based on the picture and description -- that a portion of the mass of the device was the synchronous motor and that a portion of the mass of the device is just "dead weight" whose purpose is just to act as energy storage via rotational inertia.
The parallels to the steam turbine would seem apt; the turbine end of a generator has a lot of mass relative to the electrical end. Isn't that a part of the point? Or did I miss something?
I think you're misunderstanding how this flywheel operates. It doesn't provide 50hz AC in the same way the grid operates, it just is attached to a DC motor that provides electricity which can then be converted into AC to power the grid.
That means it can spin at any frequency and it doesn't really matter what state the grid is in.
It's a synchronous rotor which is spun by the mains. When the mains frequency tries to change, the momentum of the flywheel supplies massive leading or lagging current to prevent the change.
It behaves exactly as an un-powered generator idling on the mains.
There is no DC motor.
In the industry they are known as a "Synchronous Condenser" or Syncom. They have been in use for a long time.
Of course it provides electricity to the grid. It doesn't provide any NET energy to the grid, since it's not a generator. It will pull and push energy to the grid in small amounts as needed to regulate the frequency, which is directly related to the voltage of the grid.
Does anyone have any real information about this project? The Guardian article is essentially the same as the press releases on both GE's and Statkraft's web sites and says nothing about the scale of the project.
They say it's a 25M pound budget, and show a model about 2x the height of a person, so it looks like it may just be that single unit. The moment of inertia looks comparable to that of a large steam turbine.
Thanks for the link. It is heartening to see utility-scale storage for a reasonable cost being developed.
20 years ago, there were a couple companies working on flywheel energy storage for mobile applications. They were using carbon-fiber flywheels, which could spin up to 100K RPM, but were expensive to manufacture. I would have predicted (back then) that carbon-fiber would have won out eventually for infrastructure applications as well, but I guess they were not able to get the costs down sufficiently.
Does it make sense for a flywheel to have a horizontal axis of rotation, as shown in the concept image? Seems like that would be the hard way, from the standpoint of engineering.
Interesting. Why does every other flywheel on the market have a vertical axis (that is, normal to the surface of the Earth, regardless of the place it is installed)?
I was thinking more about the fatigue of the shaft.
This is a consequence of the UK having little capacity for pumped-storage, i.e. hydro-electric. We don't really have the height, and we have "areas of outstanding natural beauty" that cannot be built in easily, so we need alternatives to store the energy.
The rest of the world just builds a dam, creates a new lake, and moves on.
That's not quite how the grid works. Consider a case where you wire two AC motors together. You spin one, and the other one spins. That is kind of like how the grid works; many loads are AC motors that spin at the line frequency, and generators try to also spin at that frequency. If the load on the motors increases, the load on the generators increases. Eventually there won't be enough energy going into the generator to spin at the nominal line frequency, and the motors attached to the mains will spin more slowly (and correspondingly use less power, and output less work).
This all becomes more complex when the loads are switch-mode power supplies and the generators are solar panels. The flywheel project adds some smoothing to the grid, as spinning generators are replaced with DC devices. But the principles are the same; without any extra devices, electricity has to be generated at the exact instant that it is consumed, and the consumers have a large physical effect on the generators.
(Frequency is not proportional to voltage in general, only in this spinning-generator connected to a spinning load case. You can obviously switch DC on and off at whatever frequency you desire, and a perfect on/off cycle consists of infinitely many frequencies at various amplitudes.)
I'm not sure where you are getting that frequently is proportional to voltage. Voltage in a synchronous generator is controlled by varying rotor excitation current. Frequently is controlled by varying the power to the prime mover.
But your main point is correct. This has to store energy to control frequency.
That's a good point. But I'm not sure it's quite as simple as you make it sound to construct a dam with 2.5 million cubic meters of concrete, divert all of the water for a major drainage basin into the newly-created reservoir, and flood 500 square kilometers of land to a depth of 100 meters, and then just move on.
Hydro storage and flywheel storage have very different use cases, typically.
Flywheels contain tiny amount of energy, but can respond nearly instantaneously To provide frequency regulation on the grid.
Hydro has massive amounts of energy storage (days+ worth of output), but takes a looot longer to respond than flywheels.
I think we will see very little flywheel or dam construction in the coming decades; battery storage is becoming cheap enough that it will hit the non-linear inflection point in deployment in the next few years. First, lots of lithium ion, and likely after that lots of flow batteries.
Oh interesting! I had heard of some hydro in the US planning to instal batteries to help with faster response. But I guess there is great variety in design.
I'm just cynically generalising - the rest of world compared to the UK has typically less stringent planning requirements, and more mountainous terrain. To store energy in other countries (except perhaps Holland..!) you build a hydro dam, and pump water into those dammed lakes using excess electricity to store it up for later release.
We have mountains (Wales, Lake District, Scotland), but the likelihood of creating lakes for hydro use is close to zero. And you would need many, many, flywheels to substitute for one lake.
Also, pumped-storage plants are massively big and expensive projects. E.g. the pumped-storage plant near my home town [1] (which I think was the biggest in Europe for a while) took two decades from planning to running [2].
There are two pumped storage stations in Wales that I know of. Dinorwig (where the power station is inside the mountain, as a student I got to go in there) and Ffestiniog. Wales has a long history of having dams built to provide water to English cities, see the Elan Valley, Llyn Vyrnwy, Llyn Clewedog and many more. The displacement of people and 'theft' of water even spurned a terrorist organisation to stop it called the Free Wales Army!
In case anyone's wondering what this is about, in the 1960s the Tryweryn Valley in Wales contained a village (Capel Celyn) which was evacuated and the valley flooded to make a new reservoir to provide water for Liverpool. There's a portion of a nearby wall painted with these words.
Or rather, the orthography is - spoken Welsh is ancient, but the way of writing it with Latin letters is relatively new, so it was worked out fairly sensibly.
Because saltwater is going to escape the artificial lakes even if they have impermeable beds, and we'll, there's a reason “salt the earth” is something that historically people have done, or at least talked about doing, to mortal enemies’ lands and not their own.
Have you seen how small the UK is vs. our population density? I highly doubt you could find a site big enough and at the correct elevation to make a noticeable difference.
We're the size of Michigan with 67 million people.
This sounds like a fancy syncon. They have been a thing in power network engineering for a long time.
I can believe it has "new features" but the underlying mechanics of a Syncon are not new. This may just be something good, being over sold by a company with an interest in boosting.
Life gets difficult for this model because the heat induced due to current 'turbulence' and maybe skin effect gets 'stuck' inside the chamber. Fun engineering problems.
Even more Sci-Fi (I think; I couldn’t find definitive figures as to the size Teraloop aims for): https://www.ecn.nl/news/item/floating-train-at-2000-kmh-set-.... I don’t know whether that idea died on the drawing table since 2015, but I guess it could have.
Why do we care about keeping frequency stable when more and more products simply put the power through a rectifier to turn it to DC? The lamps all have LED "bulbs". The computers have power bricks.
Do we really care if the compressors in the air conditioners and the refrigerators turn at a slightly different rate? Okay, maybe a rapid shift in frequency could be damaging, but a slight drift sounds okay mechanically.
This flywheel is exactly about compensating rapid changes in production/consumption so that they do not cause rapid shifts in frequency which can cause disruptions to the whole grid.
The household electric devices or the long term frequency drift are not relevant there, it's about keeping the huge turbines everywhere else running at a stable rate. The other way to achieve rapid compensation for frequency drop is to disconnect consumers, i.e. automatically triggered blackouts.
A change in frequency generally changes the efficiency of voltage regulators / voltage converters; over a very large network this can exhibit behaviour that looks a lot like negative resistance, this can induce runaway voltage conditions.
In the electrical grid, frequency is proportional to voltage. When the frequency drops, it means that the supplied voltage is also falling. Likewise, if you go over-frequency, then the voltage will rise.
Traditionally, the inertia of thousands of tons of spinning generator turbines across the system provided this inertia, but with more and more nonlinear sources (and also nonlinear loads), that balance is disappearing.
Batteries are also becoming popular for the same reason. The more batteries you have the easier it is to deal with fluctuations in supply and demand as you can switch them on and off in milliseconds and they can serve to both supply GW to the grid or absorb it from the grid.
Flywheels are useful because they store a lot of energy and don't require a lot of energy to keep them spinning (i.e. topped up with energy). Simply connecting them to a generator can be done (relatively) quickly and allows them to supply power for a relatively long period of time. It's basically a mechanical battery.
Both have the advantage that they are cheaper to operate than a typical peaker plant, which is increasingly the role of remaining coal plants that are otherwise too expensive by orders of magnitudes to operate continuously. Switching those on is a last resort for energy companies. The more battery they have, the less need they have for those. And the less they get utilized, the more expensive they are to keep around. Gas plants are better but they take a long time to turn off and on again and doing that is also not cheap.
Prices have actually turned negative a couple of times in e.g. the UK in cases where the power companies were literally paying people to use their excess power just so they could avoid having to turn off plants that are expensive to turn back on. Basically, grid storage capacity allows electricity companies to smooth out peaks in demand and supply and respond extremely rapidly by either soaking up or supplying many GW.
We haven't even burned coal for the purposes of providing power[0] since the 11th of May. That's the longest period without coal being burned since 1882.
[0] Sort of. A couple of coal plants underwent maintenance and had to be fired up and provide power to fully test them. We had a total of around 21 hours of coal plants supplying energy and we're now at day 18 of a coal-free run, and prior to the test we had almost 68 days!
The frequency of the grid is a strong indicator of the balance between supply and demand. If it has drifted off by even a small amount the system is close to a runaway collapse (modern switch-mode power supplies are actually quite bad from this point of view: when the voltage drops their effective power draw actually increases, in comparison to a lightbulb or traditional motor). And when the system collapses in an uncontrolled way it can quite easily cause severe damage to generating equipment which will take it offline for a long time. So there's a lot of stuff which goes into cutting out load and/or generators before bad stuff happens, and it tends to err on the side of caution, because a 1 hour blackout is far better than a significant amount of generation capacity being offline for months.
Inputting energy to the grid must occur at grid frequency. You want that stable. And there are still applications which themselves rely on grid frequency for timming and synch. Change or vary freq and you'll violate lots of assumptions.
I don't really know a lot about energy. Is this similar to something like Energy Vault[0] or is stabilizing the frequency different to storing power for later use.
This energy vault device will not work for a couple of reasons.
Like the astonishingly low energy density of such a structure or the difficulty in running and maintaining the cranes to pull the weights. It is basically a mechanical version of pumped hydroelectric energy storage; just clunky and inefficient.
Flywheels store orders of magnitude more power per ton and have very low reaction times (in the order of milliseconds).
This makes them great for frequency control which just means really short term energy storage.
The power grids frequency is directly co-related with the power throughput. If there is too much demand the frequency will drop.
A flywheel can quite easily buffer very short term demands and flatten the somewhat erratic output of wind turbines or solar.
The rotational energy of the planet is absolutely massive. I don't remember the numbers but you'd need millions of atomic bombs fired at once just to make a dent - and it would still not affect the orbit of this now radioactive fireball at all.
They aren't particularly, but there was a fairly large one last year, and grid stability is become harder as more and more of its energy comes from renewables (as the article mentions, the grid is favoring fossil fuel power sometimes in order to keep things stable).
"The power plant was notable for being one of the few that could restart generating electricity without an external power to restart its generators in case of a wide-scale power-outage. This was partly due to cold-war concerns that an attack could take out London power supplies.
To be able to restart the plant, it made use of versions of jet engines based on those used on Concorde to jump start the main generators — and they were used to keep the lights on following the October 1987 storms."
Flywheels (no matter their size) can be operated in near vacuum with magnetic bearings[0]. Friction can be really low. A funny aside is that they have to be aligned with earth's rotation so they don't resist it.
My thought is that the weight of the flywheel matters. To be able to absorb / release a tremendous amount of power, I imagine this flywheel would need to be very heavy. If so, wouldn't it non-linearly differ in efficiency from the near vacuum & magnetic bearing designs we've built before?
Perhaps the proposal is to build a large array of smaller flywheels that don't have to deal with the problem I imagine?
I'm not familiar with any part of engineering here -- just a curious soul -- sincerely asking!
Not that I know much about flywheel engineering. But machines tend to become more efficient the bigger you build them.
With flywheels, I don't see the bearings as difficult. I mean it's sure a challenge to keep a few tons afloat, but nothing unsolved. I'd assume (but don't know) that you can scale magnetic bearings a few orders of magnitude with their properties staying the same.
On the other hand, the faster you spin flywheels, the more energy they store. And here comes the limitation: The material they're made of has to sustain all tearing force. So at some point you'll add mass instead of spinning faster.
Haven't these been around forever? I went on a tour of a university tokomak in the 80's and they had some giant flywheel for short-burst electrical energy.
Framing a flywheel as a "simulation of a turbine" is really bizarre, unless there's something quite weird about this particular flywheel that the article fails to make clear.
It is to mimic the spinning turbine generators, which themselves act as flywheels, but in this case without the turbine part. The rotational inertia gives stability in that the frequency cannot suddenly change. I thought that was quite clear from the article?
The article isn't unclear, it's just weird. Smoothing out the input power is doubtlessly the most common purpose of a flywheel, second to power storage, and both of those is what this flywheel is for too. Describing a flywheel in terms of a turbine due to turbines acting as flywheels is the part that's weird.
As somebody without the domain knowledge, I thought the article did a great job of making it clear what the flywheel is doing (though of course it could be misleading in some way and I wouldn't know). I'm honestly a little confused as to the nature of your objection about which object is described in the context of the other.
I'm not 'objecting' to anything and I don't think it's misleading. I just think it's weird to describe a simple machine in terms of a more complex machine by pointing out that in one aspect the more complex machine is doing what the simpler machine does.
It's like describing a light-switch in terms of relays. Not wrong, not even misleading, but normally you'd expect the comparison to be made the other way around; describing relays in terms of light-switches or the rotational inertia of turbines in terms of flywheels.
It's from the Guardian Open Platform[1], which has a clause in the TOS[2] requiring that that the "Powered By The Guardian" logo be displayed where you display Guardian content
> Include a "Powered by The Guardian" logo (or such other Guardian logo as we may require from time to time) on the same webpage as any republished OP Content, or any tool or function that is based on OP Content. Such logo must be a reproduction of the "Powered By" file found at http://www.theguardian.com/open-platform/logos, and comply with any special terms set out by us. The "Powered by The Guardian" logo must not be used in conjunction with any content other than the OP Content.
I'm trying to find some stats on this, because flywheel UPS aren't a new idea:
https://www.finning.com/en_IE/products/new/power-systems/ele...
https://en.wikipedia.org/wiki/Flywheel_storage_power_system#...