At the moment, HAWTs are being built at >14 MW power.
Now let’s assume that a single kite can’t scale beyond eg 1 MW due to some factor (I believe tether weight is the most probable wall to be hit)
It seems clear that the LCOE for HAWT is very linked to scaling. The question is; is there hope for AWT for normal grid electricity production if we can’t physically scale the unit to more than 1/14 of HAWTs?
Don’t scale the unit, optimise the unit scale then array it in networks is my stock answer.
I’m getting boring…
Its ok to say it… but that needs to actually be viable too. There is a reason that larger HAWT are more cost effective.
If kites were larger I have no doubt we would all say thats a great thing.
Maybe go for larger groundstations and connect more kites to that? Perhaps the cost of a 100MW generator that can sit on the ground is much lower than 10x the cost of a 10MW generator that needs to sit on a tower + 10 towers etc. That is a question you can find the answer to. If yes, now the question becomes, how do you connect n kites to one generator?
I think that is my conclusion also… except even if that were the answer, connecting 14 kites to one generator seems a tall order as well…
It seems with the initial assumption about a scaling limit there are to avenues to success, bundling many kites to one ground station, or a redesign that attacks the scaling limit directly: eg if tether weight is the issue, use shorter tethers for the same power output.
I think you have to have a certain irreverence to think about ideas, and perhaps some experience. You go: I have no idea how to solve this but I’m going to spend some considerable time thinking about it anyway, if I wasted my time and if my ideas turn out to be rubbish so be it. If you have some experience with this, you have some confidence that often you’ll get somewhere.
There are several ideas to connect more kites to a single line. I haven’t seen ideas to connect more driving lines to a single drive shaft. One idea I had is to put every line on its own drum that couples to the (long) drive shaft. When the line is going out the drum is coupled to the drive shaft via one set of gearing, you couple it in a different way to the drive shaft when it is going in. You space out the pulleys going to your kites.
Let’s assume that works, now you have to synchronize either the speed of the lines going out (and in) or use (variable) gearing to synchronize the rotation supplied to the drive shaft, or both. I think the speed of the line going in and out varies constantly and significantly for single line yoyo, and that that speed is much more constant for multiple kites per line? If so, I want to say that multiple kites per line looks easier, now not only because of reduced line drag but also because of easier coupling to a generator, and you can probably put them closer together in the air and on the ground and to a central hub on the ground.
@Massimo conceived a carousel in order to connect far more than 14 kites to only one generator.
Beside it he relativizes the negative importance of the tether drag, even for a long tether. Perhaps we should review the arguments made by several users from his post below.
I think a carousel cant possibly be the solution, because laying a ring of electrical wires with a generator for each kite is undoubtly cheaper than laying a track and keep the generators moving along the track.
Not saying carousels are inherently bad, just no solution to this dilemma.
Also I am talking tether weight, not drag. We discussed drag in that other thread. Me and @Massimo are not in agreement here, thats ok.
I want to add some reasoning behind the tether mass scaling limit theory. Lets use kitepower’s kite because it has very low mass. Using info from this link: https://kitepower.nl/wp-content/uploads/2019-05-30_tether_failure_information.pdf
We have:
Lets assume avg power is at 12 m/s wind and 4 m/s reel out. Also add 30% production force to get avg 100 kW, considering we must have a reel in phase. This gives (using SK78 Prisma-Siri® S-12)
Looking at this (Weight of LEI kite sizes ? - Kiteforum.com), and multiplying the weight of a 10 sqm kite by 10 we get
Thus the mass/area is:
At 100 kW: (70+20)/60=1.5
Now lets assume we want to produce 10 MW. We need kite area 60*10/0.1=6000 sqm. The scaling is a factor of 10x in each dimension. Thus the tether length is now max 10.000 m. The pull must be such that a tether is selected for breaking strength 10.000 kN.
The mass/area at 10 MW is (2+64)*1000/6000 = 11.0
Even if the kite would weigh 30x times more relaively (think rigid kite) the tether weight still dominates.
Reducing the length seems only reasonable approach to reach 10 MW scale. The alternative is adding more and more wing area without increasing the power output relatively.
I do not think a carousel is necessarily a viable solution, having only written that several kites can be connected.
Both tether drag and weight are (perhaps unequal) issues in regard to viability. A fairly simple (but not perfect) way to estimate the tether weight issue is to add the tether weight to the wing weight in order to obtain the new wing loading.
At short tether length, we can guess that the weight of a rigid wing alone becomes a limit as it scales up. E.g. (p.3 http://www.energykitesystems.net/FAA/FAAfromMakani.pdf) Makani M600 expected specifications: 1050 kg wing mass, 440 m tether length, 250 kg tether mass [electrified, so heavier].
That said let us try to reply to the topic question. I would like to say that growing up is an obligation depending on the length of the tether and the space occupied, to avoid the risks of collisions that “networks” of small artificially connected units cannot prevent. Indeed, such units without a network connection would work just as well (or poorly). Do we see planes or wind turbines in smaller unit connected in networks? The answer is no. Just because the word “network” is fashionable doesn’t mean that it leads to a workable solution.
Let us think rather of a modular realization according to which the elements contribute to the unity of the whole. Perhaps also lowering the cantilever effect of a classic wind turbine by using airborne elements (like KiteX does) could lead to a viable solution.
And also a giant flexible power kite towing a boat seems to work well.
I think perhaps if we see it a necessity to scale huge (> 30 MW in 9 ms wind should put AWE on the map), one must focus on solutions that have inherent abilities to do so.
The 14 MW windmills have top height of <300 m. With only 1 km tether AWE should be able to reach 500 m. So one must fit enough surface area for 30 MW on a 1 km tether. Solidity is far higher than any (most?) concepts on the AWE concept market today.
Kite networks is one way.
Another way might be increasing wing area per watt produced.
A third way may be increasing the elevation angle as well (in addition to increased wing area).
Windmills have TSR of < 8. Kites with tether could better this with higher glide numbers. So I still think possible solutions may exist.
We can be blocked by the idea that an AWES must necessarily fly. Dramatically lightening the structure by reducing the cantilever effect for even more scaling can be a more accessible and useful goal. E.g.KiteX and also @Rodread have shown this by lightening the rotor, installing cables supporting the blades. Let’s tilt the rotor (it’s already done for Daisy), and make it bigger little by little. When it reaches 1 km in diameter it will also rise to about 500-700 m height.
Don’t forget the scaling promise hinted toward by Rachel Leuthold and Kiteswarms
Yo-yo via
Multiple stacked layers of well controlled individual (or paired or trippled…) kites rotating around nodes of a single line .
It’s a strong design when you have reliable control and they’re working really hard on it.
Can AWES be viable if it does not win scaling
AWES would need niche markets for that, and some systems can maybe already cut a market by being lightest solution / power out.
Why not combine all the AWE stations by means of an hydraulic system. Each groundgen will have an hydraulic pump which will pump high pressure fluid into the system. With this method, we only need one hydraulic generator to generate electricity. The efficiency of this system is greater than 90% as shown in the attached article. https://res.mdpi.com/d_attachment/applsci/applsci-08-00858/article_deploy/applsci-08-00858.pdf
The KG-Carousel track itself is the generator, the stator holds the power wiring and coils. Dimensioning the alternator track shows that the large ring circumference (7000+ m) allows reaching the electrical machine good thermal balance and power density, discrete alternators are not viable. I suggest computing the magnetics with 2 ton/MW of weight, all other requirements descend from this assumption. A GW scale Carousel needs 2000 tons of distributed magnetics and 150MW of heat to dissipate, with no room for discretization. Despite the extension, In less than one hour of operation, it could reach harmful temperatures. Natural heat dissipation, of the alternator track, must be improved with a distributed heat sink to find an acceptable balance without chillers.
That much heat would be relevant for a large town district heating scheme…
In winter & higher latitudes heat is more valuable
Is it possible to run tube / water-cooling inside the track coils?
Hi @Massimo, there is a document below with data including rotor and stator weights, page 14. For the Carousel AQUA 1000 (1 GW) the rotor weight is 100,000 tons while the stator weight is 150,000 tons (so 250 tons/MW for both stator and rotor). Is it correct? These values look to be quite huge even compared to the expected power. In comparison for the ground KG of 3 MW the generator weight is “only” 20-60 tons (so 20 tons/MW for 60 tons, and 6-7 tons/MW for 20 tons). Please can you explain or correct?
I think electric cabling is very hood, perhaps unbeatable. Its just the fact that eletricity is dirt cheap, and power cable installation is not free…
This basic idea to insert a hydraulic system between a wind energy collector and its generator has been suggested by one person after another, for as long as I’ve been involved. Often the reason is simply to place the generator at ground level. But then you have to install a hydraulic pump and possibly a gearbox uptower. You trade wiring for piping. Then you still need a generator, and also a hydraulic pump to turn it. Another gearbox at ground-level?.
The answer seems to be (my impression) that the hydraulic pump offers little or no advantages over a generator, while adding this additional layer of energy transfer entails unnecessary complication and higher cost, more equipment to maintain and fail, and probably in most cases lower overall efficiency. Still, might be the answer for some situation…