Do you know which software could do this? Also learning curve? I expect you are thinking about «grashopper»?
There have been animations published of optimised AWES flight loops
(I think from Freiburg and others)
highlighting the changing lift & drag force regime as parameters like wind speed and tether length change.
Yeah I reckon it wouldn’t be too hard to do it in grasshopper too. Hmmmmm maybe I’ll get time one day.
Offshore comparison of AWES vs HAWT should take also into account innovations for both, not only AWES.
As an example the PivotBuoy system from X1 Wind would allow to implement a far lighter rig for HAWT.
Some explains are on:
Would such an innovation be a brake or a springboard for AWES offshore?
HAWT is innovating fast, 14MW (rated power) in the pipeline, lighter blades and nacelle, alternative materials for the tower (wood laminates), floating Offshore, higher capacity factors and lower LCOE.
AWE first generation systems (the Loyd generation) are either FlyGen or YoYo. After Makani, AWE is essientially just YoYo. TU Delft has a long history with Laddermill (morphing into YoYo) and many graduates going on to work in AWE Companies with a YoYo based design.
Everyone has a ‘pet’ design, which they think is the next big thing. It’s very unlikely that their viewpoint can be changed, even when confronted by design problems. There are almost certainly Makani Engineers who still believe that the M600 can be a success.
Simply, Makani M600 has a ‘rated power’ output <= 250kW @ ~11m/s wind speed. It is less efficient (% power output at the generator) than the M30, it didn’t Scale-Up, it is too big.
Offshore (HAWT) -> ‘bigger is better’, because of CAPEX, OPEX and infrastructure. A few big machines are easier (lower cost) to install/maintain than lots of little ones.
AWE advantage (wind) is Onshore. AWE disadvantage (size) is Offshore.
AWE Companies still think it’s possible to build a 1 MW (rated power) single unit YoYo system. Makani failed @ <= 250kW.
Hi Derek: As the guy who invented the basic configuration later called “laddermill”, way back in the 1970’s as a teenager, I was first excited that someone would finally be building one, then disappointed as, to this day, as far as I know, nobody has ever bothered to build even the most rudimentary “laddermill”. The “long history” as far as I can see, was nothing but empty talk, with no followup. To say laddermill “morphed” into “yo-yo” is, I believe, inaccurate, since every kid reeling a kite in and out could imagine a motorized reel, then allowing it to run as a generator as it reels out, which must be among the very simplest AWE configurations imaginable. Laddermill, on the other hand, is a more complex design with more components, which, if successfully built and run, would ostensibly operate in a steady-state, rather than requiring an intermittent cycle. So I’d say the original concept of laddermill devolved or degenerated into yo-yo kite-reeling, due to reeling a single kite being closer to what kite-flying people were used to doing, with fewer unknowns and less experimentation required to achieve basic functionality. I don’t see any “morphing” other than laddermill was given up on, and the enthusiasm found a new home in a possibly less-promising, yet easier-to-achieve-immediate-results “plan-B” configuration, that was so simple as to be pretty much obvious, and which I would say invented itself, rather than requiring any cleverness to come up with.
Also, to say “After Makani, AWE is essentially just YoYo” is mostly in response to the highly-publicized efforts seeking lots of funding, and promoting group-selfies, rather than looking at the more broad array of possibilities out there, both in concept only, and also including reasonably promising prototypes having been built and run.
1 Like
Its time to open up the definition of AWE. This involves taking kite arrays/multiple kite systems into consideration.
To me it seems a possibility (based on speculation but also physics) that Makanis next scaling step would need more than one kite. (Proof: conductive tethers are heavy thus limits length and wingspan. A single kite cant get close to approacing the Betz limit which is necessary at some level to compete with huge HAWTs)
Once you have more than one kite - the list of possible, sensible constructions explode. Its still lift and drag mode, or like @PierreB explained me in the thread about the worst AWE, something both lift and drag based. I think the next step will be shifting away from the AWE stereotypes and accepting that at this point, we dont know what successful AWE might look like.
Here are some signs that a design could work, that we should be looking for:
- effective use of materials
- crosswind flight
- solves the practical aspects of launch and high wind safety
- production of electric energy or propulsion rather than some intermediary form of energy
I think rotary rigs have been undercommunicated in AWE articles long enough.
Until we get a lighter weight composite material, advances in soft kites or carbon tube tethers, I think we should also realize that it will be hard for AWE to compete with HAWT in pure scale. At 14 MW, which is probably 30 MW before AWE enters any largish market, the scale is already such that tether mass alone is a crippling factor for AWE. If you want kilometers of tether, there us just no two ways about it, that the mass scales in a cubic fashion. Then realize the other limitations of huge flying vehicles that cant be solved by engineering.
2 Likes
Y configuration is often seen as an ultimate. Several wings sweep their common annulus, while the main tether is fixed. Launching could be achieved by using a ground area equivalent to said annulus, the station being at the center of said ground area, and by using the generator(s) as motor(s), VTOL being achieved by considering the set as a rotor.
Y configuration could allow to maximize the space used by approaching the Betz limit, and mitigate tether problems (drag and weight). However Y configuration could be too dependent on a computerized control, and the presence of a single main tether would constitute a serious risk. We dare not imagine the consequences as the tether brokes. It is a reason of the possible suitability of a design like Dr. Beaujean's airborne wind turbine comprising several fixed tethers, and where the blades are the alone moving elements.
A theoretical advantage of AWES over HAWT is a not limited space. And AWE aims high altitude wind energy. So, IMHO, AWES will only be able to compete with units that are much larger than the HAWT which will have moreover made great progress. The current prototypes can be seen as first means to explore AWE potential, not as achieved systems, and that for and until a while. It is also why we don’t see significant intermediate markets.
Makani Launch and land Perch was smaller than the wing.
Surely a multi wing ground station doesn’t need to use the same launch area as the flown ring.
Doug : TU Delft has the ‘Loyd’ mindset -> a single AWE unit (single kite) can be bigger (MW) than HAWT. This was the basis of the Loyd paper (1980). Ampyx still think that they can make a 5MW system with one kite. Laddermill is a better design, with multi-kites and continuous power output, you were ahead of your time with that idea. However, the idea was not taken forward by the masses so I didn’t include you in the ‘first generation’ (by investment) tag. Things might have looked a little different, if the Sky-Serpent or other Laddermill variants had been pursued by Delft. Today, AWE is essentially just YoYo (by investment). Of course, there may soon be more interest in Laddermill designs, and other Utility Scale designs, as the ‘Loyd generation’ comes to an end.
Tallak : We can wait for new developments in materials, but HAWT can do the same. We could just build an AWE system now that matches (and surpasses) HAWT. Solar seems to do OK, it’s made up of lots of little panels, and when they are put together into multi-MW farms they produce cheap electricity. ‘Scale-Out’ and not ‘Scale-Up’.
Pierre : Do you think that the ‘Y’ configuration (dancing kites) is a good idea? Do you think ‘bigger is better’ (in the HAWT sense), i.e. bigger blades are better, so bigger kites (wings) are better?
Rod : Launch & Landing systems are part of the initial design. A good design will allow the groundstation size and wing size (pattern size) to be independent. The Makani perch is a good example, so a vertical multi-perch would use the same ground footprint (that’s not a recommendation).
Makani : Good show, and good decision to open up the archive.
Let’s concentrate on Utility Scale systems, say >~100kW ‘rated power’. The M600 is the first AWE system at this scale.
Theory :
One of the Laws of Physics, the ‘law of conservation of energy’ concerns the interchange of Potential Energy (PE) and Kinetic Energy (KE). This is described in detail earlier in this topic (July 2019), just before the OktoberKite (MX2) 3 month Design Review.
-
The Law says that an object (kite) moving around a loop (or figure 8) has a continuously changing velocity. It speeds up on descent, and slows down when climbing back to the top. It’s affected by gravity.
-
Miles Loyd came up with a theory to make lots of power using a big kite. The kite is assumed to fly at a constant (optimum) velocity around a loop (for a given wind speed). A constant speed loop. Any deviation from this ‘sweet spot’ will reduce the power output over the cycle.
There is a mismatch between reality (1) and Loyd’s Theory (2).
Design :
The kite gets a ‘push’ (PE -> KE) on the descent, but must pay it back on the climb (‘There’s no free lunch’). Simultaneously, the kite gets an Aerodynamic Energy (AE) ‘push’ for the entire loop (from the wind). It’s like a ‘double push’ on the descent, some for generating power, and some (PE) to save for payback later. If the (PE) is not saved, then the payback comes out of the AE, and the average power output decreases.
The M600 was intended to generate power on the descent and the climb, a constant speed loop (Loyd). It didn’t. It generated power on the descent, wasted the PE (by depower -> ‘brake on’), and then had to use the AE on the climb to replace the lost PE, so couldn’t generate any power (PE > AE, see below). Also, the minimum airspeed has to be maintained, so more energy is consumed.
Much later, in the MX2 Design Review, Makani came up with a ‘Loyd patch’ (page 131, ref mx1). Store the PE to the grid on the descent, then take it back out for the climb. It’s a sort of ‘pass the parcel’, only with a giant parcel, lots of wrapping paper (PE) and a small gift (AE), where the parcel is passed back and forth via a giant tube (tether). The round trip efficiency ‘to grid’ and ‘from grid’ is low … some of the wrapping paper always falls off.
Nature tends to do things better, a ‘rollercoaster’ loop, saves all of the PE in speed, so that all of the AE is available for power generation (cf. HAWT). The downside is the need for a stronger tether (thicker -> more drag, more mass), giving a lower zeta (aero performance metric) and reducing AE. Who wouldn’t want to see the M600 in ‘rollercoaster’ mode …
Saving the PE (mgh) efficiently, is challenging, so it’s a good idea to minimise it i.e. low mass (m) and small loops (for a vertical loop, h = 2 * loop radius). ‘Mini loops’ are better than ‘Maxi loops’. A lightweight aerobatic kite is ideal.
Results (M600) :
The M600 ‘power curve’ (page 232, ref mx1), shows a ‘rated power’ output of ~100kW (electrical) @ 11m/s ‘rated wind speed’ (simulation predicted 600kW). This gives ~10% efficiency at the onboard generators (~1 MW).
Makani thought that one of the Laws of Physics was a bug -> Bug 34221767 (page 14 … 18 ref mx3).
“The flight test tether tension is systematically a factor of 2 or greater in the flight test than the
sim in the downstroke (near the 5 o’clock position)” (page 14).
Care needs to be taken when looking at the data. The flight data is real, but the Sim data could be wrong, and certainly is here, since the Sim is scaled for a 600kW ‘rated power’ system.
There’s an extra ‘push’ coming from somewhere, and it’s bigger than the aero ‘push’ … (PE > AE)
Performance :
Defining AEr as the AE at ‘rated wind speed’, then
AEr = ‘rated power’ * loop descent time (sec).
‘Performance Metric’ PM = AEr / PE. Higher is better.
Makani M30 : PE ~57kJ AEr ~64kJ PM = 1.1
Note PE : 76kg (inc tether) h = 77m (r = 50m, 40°) AEr : ~16kW ~4sec (Fig 28.7 page 486 ref b,e)
Makani M600 : PE ~5.1MJ AEr ~0.9MJ PM = 0.2
Note PE : 2080kg (inc tether) h = 251m (r = 145m, 30°) AEr : ~100kW ~9sec (Table 1 page 96 ref mx1)
For a given loop, the PE is fixed, but the AE can change, depending on the wind speed. By definition, the best performance (PM) occurs at the ‘rated wind speed’. At lower wind speeds, PE will have a greater effect, resulting in a later cut-in and a degraded ‘power curve’.
At higher wind speeds, from ‘rated wind speed’ to cut-out (11m/s to 25m/s), the AE can be much bigger. This excess energy is thrown away after the ‘rated’ limit, due to the infrequency (same as HAWT). Infact, both Makani wings could not fly in ‘high’ winds (page 232 ref mx1), due to the inadequate depower strategies.
‘Scale Up’ (wrt kite Area), means kite mass, and tether mass (kg) increase exponentially, along with ‘wing loading’ (kg/m²) and turn radius. PE will increase exponentially, while AE just scales with the kite area. ‘Scale Up’ lowers the relative performance (PM), such that the M600 performs significantly worse than the M30, largely due to the higher ‘wing loading’ and lower turning agility.
Test :
Simulate ‘Scale Up’.
a) fly your kite as normal
b) increase the loop size (e.g. ~50%), fly again
(Warning : tether tension will increase and the tether could break)
c) if required, repeat a) & b) pairs at different wind speeds from 'cut-in' to 'rated'
(to modify AE).
d) publish the results
The average power output over the full cycle will reduce as the loop gets bigger, and the cycle time will be stretched out.
Something Else :
Of course, ‘Loyd without gravity’, by flying the kite in horizontal circles or figure 8’s may seem like a good idea, but historically it never really caught on … VAWT.
References :
mx1) The Energy Kite Report (Part 1) Makani Team 2020
mx3) The Energy Kite Report (Part 3) Makani Team 2020
2 Likes
I also evoked this concern on
I mentioned on the same comment:
Dancing kites (two kites in opposition) could perhaps mitigate the irregularity issue (adding significant control issue) but losses by gravity and resulting higher kinetic energy during down phase would occur in the same way, increasing while scaling up.
To get back to the heart of the problem let us quote again, on page 13, Part I:
● The potential energy swings resulting from large loops forced us into large speed
variations. At low winds, we’re often flying much faster than optimum at the bottom of
the path, and at high winds, we must fly much slower than optimum at the top of the
path to prevent overspeeding at the bottom.
I see a double issue. Even if there was no need to prevent overspeeding at the bottom, the system would undergo loss. Kinetic energy is higher during descent than during climb, the difference increasing by the square of the kite speed, and this energy is more or less directed to the ground, preventing climb. HAWT has an axis supporting the rotor, a crosswind AWES has not.
1 Like
Not in absolute terms because of the multiplication of risks that only a flawless computerized control can avoid. This being so, the Y configuration makes it possible to better maximize the space.
Yes. The tether is already 1 km long, and delimits a huge space. So from a certain power a small or large AWES will take the same volume for a given tether length. So to compete with a HAWT an AWES should be far more powerful and large, in an equivalent proportion of the tether length compared to the tower height. That said an AWES can be composed of several smaller elements to mitigate weight penalty by scaling, even a network of connected elements.
1 Like