Sky Windpower’s high-altitude wind generation system will likely fail to launch

Last update: January 8, 2014

A recent assessment of Google’s Makani flying wind generator led to a suggestion to look at other suggested approaches. Sky Windpower was one of them and has sufficient public material for an initial analysis to look for challenges and opportunities.

Artist's rendition of Sky Windpower device

Artist’s rendition of Sky Windpower device

Sky Windpower’s proposed high-altitude wind generation solution raises major concerns:

  • For serious generation it requires building by far the world’s largest, most powerful and fully autonomous helicopter / quadcopter, and many of them.
  • It needs to get the price per unit for these devices down to a fraction of the price of the world’s largest helicopters today.
  • It likely needs to consume more land without secondary uses than comparable conventional wind generation.
  • It requires that significant additional airspace be declared to be fully restricted.
  • It is likely limited to non-winter operation, and may be challenged with icing at the altitudes it will fly at.
  • Tether weight alone will likely be greater than the largest load lifting helicopter capability today.
  • They must at least invert normal maintenance to flight times for helicopters from the common 3.5-4.5 hours of maintenance to a single hour of flight.
  • They are very optimistic regarding capacity factors due to weather- and maintenance-induced landing requirements, as well as energy costs to fly the devices to their operating altitude and back.
  • It is unlikely that they have fully analyzed and accounted for failure conditions in their projections; testing required to satisfy aviation authorities and insurers will likely take more than a decade if they are fully funded.

What is Sky Windpower’s solution?

At heart, it’s simple. Fly a quadcopter up to 15,000 to 30,000 feet where the air velocity is higher, have the blades spin in the wind like a kite, generate electricity from reversing the electric motors similar to regenerative braking on an electric car and transmit the electricity to the ground down the conductive tether.  They have a small (1:13 or 1:17 scale) prototype that they have tested as of December 2011 flying with additional safety tethers through a limited range of required manoeuvers. More testing may have been done and not publicized on their website. Also note that comments have been made that Sky Windpower has abandoned the high-altitude goal and is targeting a 2000′ foot ceiling, but this was not reflected in their publicly available material as of November 2013; it is worth assessing the high-altitude implications regardless.

Sky Windpower is foregoing most of the cubing of wind velocity Makani takes advantage of by cross wind flight while still suffering from a large loss of swept area. To get significant generation with this, they point out they would need to be flying in the 15,000-30,000 ft (4500-9000m) range. To be conservative this implies a minimum 9 kilometer electrified tether which is at any given time some where in a half sphere with a 9 km radius with a volume of about 1500 cubic kilometres.

What are the implications for scaling?

Sky Windpower projects one quarter the swept area over four rotor sets for equivalent generation faceplate capacity to a conventional wind turbine. Let’s consider a five MW conventional wind turbine as a comparison. It has about a 128 m rotor diameter and a swept area of just under 13,000 square meters.

A small amount of math using Sky Windpower’s factor tells us that the Sky Windpower device would require four 32 m (112 ft) diameter rotors with a total craft size of about 70 meters (245 ft) including blades per side. For context, the MI-26 heavy lift helicopter has a rotor diameter of 32 m with its eight blades; check the rotor droop and imagine four MI 26’s bolted together with less fuselage. Considering a two MW conventional wind turbine, the Sky Windpower device would have 23 m rotor (80.5 ft) diameter and the device would be roughly 51 m (179 feet on each side including blades. This is still far from a trivial scale.

MI 26 heavy lift helicopter

What about the tether weight and length?

Tether weight is a glossed over issue on their site; Makani’s one km, carbon fibre and aluminum tether is projected to weigh 3,660 kg; let’s assume that is accurate and representative. The minimum length 9 km conductive tether would weigh about 33,000 kg assuming adequate strength. A 9 km ceiling would require a roughly 18 km tether weighing 66,000 kg. The MI 26, one of heaviest lifting helicopters in the world, has a maximum lift of about 20,000 kg above its own 36,000 kg weight and a hover ceiling of 1700 m.

Sky Windpower are projecting making devices that are much better than anything created by the aerospace industry which strikes me as unrealistic, not to mention nontrivially expensive on development and a per unit production basis. It would be unsurprising if a 9 km carbon fibre and aluminum conductive tether cost more than the mast of an equivalently generating conventional wind turbine, and for tether plus device to cost significantly more than a conventional wind turbine of equivalent capacity. For comparison, an MI 26 by itself costs $15-18 million, and a 5 MW Sky Windpower device would be much bigger and fully autonomous.  Also for comparison, the MI 26 weighs about 36,000 kg including fuel; take off most of the fuselage, replace the turbines with electric motors, remove the fuel tanks and then multiply by four. It’s likely that a 5 MW equivalent Sky Windpower device would weigh 20,000-30,000 kg by itself. The device has to lift 53,000 – 63,000 kg to 4.5 km to get into its operational zone. No weight or dimension calculations for the Sky Windpower device are included in their public documents, so these are necessarily estimates.

What about weather conditions for safe flying?

Sky Windpower’s current plan per public material is to land the devices when lightning is forecast. They make no mention of how long it will take their device to ascend to its 4.5 km operating minimum or descend from that altitude safely, but they project future developments which will make this necessary less often. They are silent on other weather conditions which would require grounding of the device; there are easily half a dozen which could or will require this. They are also silent on energy required to lift the potentially 20,000 kg device and 33,000 kg of cable to operating height and what the net energy output impact will be. For context, one conventional wind farm’s draw from the grid for blade start and other peripheral items was 1:320th of the annual generated electricity.

Like Makani, Sky Windpower does not address temperature variation as a concern in their public material. However, while Makani can safely operate in tropical and subtropical climates, or in summers in temperate climates, the Sky Windpower device is projected to fly at altitudes where the temperatures are often below zero regardless of surface temperatures. The rule of thumb is 3.5 degrees F (2 degrees C with rounding) per 1000′ of altitude gain. At minimum suggested operational altitude of 15,000′, the Sky Windpower would be operating in temperatures on average of about 53 degrees F or 30 degrees C below surface temperatures. At 30,000′, operating temperatures will be 105 degrees F (60 degrees C with rounding) lower on average. In their publicly available material, there is no mention of de-icing technology requirements on the ground, or operational factors which will prevent ice build up aloft on the device or along the tether. It is possible that when operating, airframe vibrations will prevent ice build up, as will tether oscillations, but these assumptions are unstated. Regardless, de-icing on the ground is a significant unstated operational requirement for any climate which experiences winter conditions.

What about maintenance?

Like Makani, maintenance requirements will be much higher for this device than for a conventional wind turbine. The ratio of maintenance to flight hours for helicopters is often in the 3.5-4.5 to 1 range, meaning that a helicopter flies one hour for every 3.5 – 4.5 hours of maintenance. Sky Windpower is silent on why their quadcopter technology is expected to achieve at minimum the inverse of that ratio, and why this is going to be better than the guaranteed 95% availability modern conventional wind turbines have while under warranty. Maintenance is unlikely to require shutdown of significant portions of the generating farm however. With icing and weather related downtime as well as significant energy loss to get their device to altitude it’s unclear that they will achieve the significantly greater capacity factors to overcome other challenges. Remember that new conventional wind turbines in the US operate at 35% capacity factor in lower wind category zones, and up to 47% capacity factor in the best wind category zones per LBNL. Achieving significant multiples of those numbers is non-trivial.

What about spacing and safety?

Spacing requirements will be large.  Ground equipment and tethers are not worked out in any detail on their public site.  Given multiple devices at multiple altitudes with multiple tether angles ascending and descending, turbulence at different altitudes, tether slack potential and tether oscillation, a single 70 m (245 ft) per side device will need a large amount of operating leeway on all sides. If a device manages to get the cable of a nearby device caught in its blades, both are likely to be in serious trouble and this would very likely cause a domino effect in other devices which would have to be taken account of in spacing.

For an idea of the best possible scenario of a high-strength cable caught in large rotor blades, watch this video of a helicopter catching a cable.

A worst case scenario is a higher-altitude Sky Windpower device having a tether broken a kilometre or two below the device due to an airplane strike or a malfunction on another Sky Windpower device. The Sky Windpower device would be blown downwind rapidly, with the 3700-7400 kg of severed tether penduluming beneath it. Once it stabilizes, it will not have power so it probably won’t be able to make headway and will likely still be going downwind if more slowly, and with limited onboard control ability due to lack of significant power.  All devices downwind will have the tether dragged through their operating airspace and potentially fouling their rotors. This implies at minimum very large gaps between devices as well as a very large amount of synchronous flying automation given the oscillating tether. While quadcopter automated synchronous flying is advancing, at present it is only done with tiny quadcopters in safely enclosed indoor volumes. Sky Windpower asserts that they could bring a device down safely in the event of a severed tether, but it’s unclear that they have thought through the scenarios. In any event, the implication for spacing is that safety requirements for the foreseeable future will required significant spacing, likely more than conventional wind turbines just as with Makani.

What about land usage?

The area under a Sky Windpower farm will be the equivalent of an airport. Due to very large devices landing and taking off, it is likely that no other economic uses would be allowable due to workers safety and insurance reasons. Unlike Makani, maintenance would likely not require nearby generating devices to be shutdown however. This once again is a comparative disadvantage to conventional wind farms which are mostly interspersed with agricultural and other land uses.

Sky Windpower claims 1:400th of US airspace would be required to provide all of the USA’s electrical demand, so they’ve done some density math, but once again do not show that math. Coarse assumptions could be five MW devices, a very generous 60% capacity factor and 1000 m spacing between devices based on the above assessment. At roughly 3500 TWh per year annual consumption, the average would require roughly 133,000 of these devices covering about 2% of the total land mass of the contiguous United States with technology that makes any other uses unsafe. It’s unclear what Sky Windpower’s calculation of 1:400th — an order of magnitude less than this calculation — is based upon, but it is highly suspect as a power density calculation without engineering or safety considerations.

What about flight permissions?

Sky Windpower indicates that they have had initial discussions with the FAA on operating conditions and requirements, and have some ideas on how to achieve acceptance of their device. Their comparison to blimps is disingenuous just as Makani’s comparison to antenna tower guy wires is disingenuous; blimps have passive lift and are much safer than an auto rotating quadcopter with four rapidly rotating blades of the scale of the Sky Windpower projected device. Imagine four MI 26 heavy lift helicopters bolted together auto rotating down in a housing complex in a high wind. Imagine a blimp coming down in the same place. The FAA and insurers will imagine this.

Like Makani, Sky Windpower is more optimistic about the FAA accepting their arguments than I am, and if this moves forward it will likely after be a decade or more of dedicated and restricted testing. One comparison to blimps may be apt, however; currently blimps carrying radar equipment at 15,000 feet do not require tether markings in restricted airspace according to a point made on the airborne wind energy systems forum. It’s unclear if aviation authorities will be convinced that the more passively safe blimps are equivalent to massive quadcopters however.

What does this all add up to?

Makani has a much more complete solution publicly available than Sky Windpower’s, and one that didn’t depend on what appears to be unlikely leaps in technology. Makani’s solution has significant safety limitations and is unlikely to be economic onshore in many places, but the challenges pertaining to the Sky Windpower solution cover a much, much larger potential impacted area. Makani, as with other vendors in this space such as Skysails, realizes that the constraints imply a much greater likelihood of offshore use for high-altitude wind generation, but Sky Windpower has no reflection of this in their public material.

Sky Windpower is projecting building the biggest multi-rotor helicopter ever, the largest autonomous helicopter ever, the heaviest lifting helicopter ever, the highest-flying autonomous helicopter ever and for a required price point much lower than a single MI 26. That isn’t a recipe for success in any near term time frame. Sky Windpower does not have appeared to have worked out even a small fraction of the engineering challenges. It’s unclear that it’s even possible given air density with increasing altitude to lift the 33,000 kg tether although they assert that helicopter limitations don’t matter due to using rotor kiting; they don’t show their math so its impossible to determine their assumptions.

Sky Windpower’s solution is unlikely to be allowed, it requires solving a number of non-trivial engineering problems and if they have thought through all of the above implications of their solution, they are not articulating it in their public documents. It would likely require enormous amounts of land be set aside for its use.  The firm does not have on its website an economic analysis, but merely stated that there is an economic case with some already significant assumptions.

Sky Windpower doesn’t pass the sniff test. It’s even less likely than Makani’s approach to generate any significant amount of energy.


  1. “It likely needs to consume more land without secondary uses than comparable conventional wind generation.” Not sure since jet-stream is expected to be unidirectional,the tether keeping its position during operation which can set several months.

    1. The biggest flaw is the expectation of massively different ratios of maintenance to flight hours for rotorcraft than has ever been achieved before. At 3.5-4.5 to one ratios the generators will be landing a lot and seeing 20% capacity factors from this issue alone. This ignores weather groundings. These things would be up and down like yo you but mostly down. As such it’s a busy airport not a field with some cables pointing into the sky.

  2. ” The minimum length 9 km conductive tether would weigh about 33,000 kg assuming adequate strength. A 9 km ceiling would require a roughly 18 km tether weighing 66,000 kg.” More due to additional losses!

    1. I was being a bit generous there I agree.

      1. …If the weight of 9 km conductive tether is about 33,000 kg…

  3. “It is likely limited to non-winter operation, and may be challenged with icing at the altitudes it will fly at. “No, since the system being stationary there is no apparent wind (like for crosswind kite-tether) cooling the conductive _ the conduction producing some heat_ tether. Concerning the quadcopter itself,the problem of icing is the same than for airliners flying in low temperatures.

    1. It is true wind speeds being higher are more able to cool anything.Making comparison with electrical cables working il low temperatures.

  4. On one place:”Sky Windpower was one of them and has sufficient public material for an initial analysis to look for challenges and opportunities.”.On another place:
    “…they don’t show their math so its impossible to determine their assumptions.” What is true?

    1. Sufficient to do an adequate comparison. More information would be better and would push some assessments from likely to either true or lower possibility. There’s certainly enough information to realize it’s very unlikely to be economic or even technically possible.

  5. An intreasting analysis – while I won’t fault the rest, are you not making the assumption the blades operate must be like helcopter blades ? If the blades are fixed with fewer moving parts like more typical quadcopters than surely the cost/maintenance/scaling issues are more surmountable.

    I would say there are considerable uncertainties but we know the cost profiles of the certain systems arn’t good enough, there is enough to past a sniff test – if it was on Kickstarter I would chip in, I would say there is enough here to worth funding a bigger prototype.

    1. An interesting question. My take is that pitch control will be necessary for survivability. And regardless, the quote about beating the air into submission still applies.

      If I’m remembering correctly the first working prototype of this approach was 1986. Not a good sign.

      1. On their website Sky Windpower explains by using four rotors there is no need of “cyclic pitch”. Only “collective pitch” is used for changes of direction when it is needed. So without “cyclic pitch” the cost of maintenance should be far lesser.

      2. Except that I don’t believe that will be the case as the device scales up. Quadcopters are all small today and as such have great lower to weight ratios that allow brute force to replace elegance. As they scale up, I believe it’s likely that they will find the simplicity of their current device is lacking capability and safety. And they’ll still be beating the air into submission. A long time rotorcraft expert agrees with my take on Sky Windpower so I’m comfortable with my assessment that they wouldn’t be able to achieve an extraordinary reversal of current rotorcraft maintenance cycles.

      3. “Except that I don’t believe that will be the case as the device scales up.” With numerous rotors (what is the limit?),by keeping the size of blades.
        See the helicopter with multirotor as possible AWES on .

      4. Except of course that the rotors are above the airframe pushing air down toward it in this instance. In a generation device, the more rotors, the more airframe and the more the rotors are flying in the lee of the airframe or in turbulent air increasing potential failures and reducing generation.

        Ground-based wind turbines moved away from simpler behind the mast rotors in the mid-80s due to flow challenges. Cleaner air rules for generation and reliability.

  6. “Sky Windpower projects one quarter the swept area…” Another question is the out put of projected swept area as wind turbine by taking account of aerodynamic features, a blade going upwind while the other blade goind downwind.Are there measures as wind turbine?Is it really possible in a good way to link both needed lift as autogyro-helicopter and needed torque as wind turbine?

    1. My assumption is that blades will be suboptimal for both lift and generation, a compromise. Makani and Sky Windpower will both have lesser output than possible as a result. This is fine on medium sized prototypes but as the devices scale up to the mega scale shown as needed, it’s probable that they will have to be optimized more for flying and sub optimized further for generation.

  7. Mike,
    The design of Sky Windpower is elegant, being the natural meeting point of both wind turbine and airplane as rotorcraft. However its quotation is low also because of an ambitious target. My idea is a lighter version implementable offshore for lower winds can be a possibly interesting target, eliminating problems of crosswind tether making the device safer and better locatable. Another advantage can be the possibility to join the modules of 4 rotors to go progressively towards utility-scale, making a single large device rather several small devices with the inherent risk of mess due to long tethers.

  8. […] or the company website if a detailed assessment hasn’t been published yet: Makani, Sky Windpower, Windlift, Altaeros, University of Delft, Kitegen’s Stem, Ampyx, Magenn (now […]

  9. […] assessments, or the company website if a detailed assessment hasn’t been published yet: Makani, Sky Windpower, Windlift, Altaeros, University of Delft, Kitegen’s Stem, Ampyx, Magenn (now […]

Be nice, be respectful, be relevant.

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 1,229 other followers

%d bloggers like this: