RogerBW's Blog

2020 documentary, dir. Kate Stirr: IMDb.

The Makani project tried to solve a simple problem: wind turbines are already big and expensive, but there's more wind at higher altitudes. So why not put a wind turbine on a kite rather than on a pole?

That in itself is a decent idea (turbine-generators are heavy, but it looks basically solvable); but every time there was a decision to be made, the engineers took the most complicated option. Let's not just loft a turbine; let's have it circling in the sky, sweeping through a larger area of air. Yes, there will indeed be more power. But this means the thing has no stable gliding state: it will constantly be looping on its tether.

The progress of the project seems to be a catalogue of repeating bad ideas discarded from historical aircraft designs. The initial version of the kite was a flying wing; then later it changed into a tail-sitter, which would transition from a rotor-borne takeoff into a forward gliding mode and back again; and it was intended to land and take off (vertically) from a ground station which would also handle the tether. Oh, and all of this is supposed to happen in all weather conditions without human intervention.

Flying wings are hard; only one (the B-2) has gone into production, and the extent to which it works is distinctly arguable. Converting tail-sitters are hard; nobody has ever got one working without severely compromising its performance in conventional flight, and I don't think one has ever gone into production. VTOL is hard; the aircraft which do it are dedicated to it to the point that that capability becomes their primary virtue. All of these things add weight and complexity. There were plenty of smart engineers involved… but couldn't they have found one with a hobby of aviation history who could tell them why this particular idea didn't work on all previous occasions it had been tried?

To me this seems like a cautionary tale about working to the maximum extent of your capability: if you design something so complicated that you can just barely understand it, then when things fail you don't have enough margin of understanding that you can work out what's going wrong and try to fix it. This documentary makes it feel as though each person was doing that: they are taken entirely by surprise by wind state excursions that didn't happen in their simulations. They design their tether for a constant load plus safety margin and apparently don't account for jerk. (Or long term fatigue from repetitive stress cycles; they didn't ever fly for long enough to gather data on that.) Their kite has protruding fins and they're surprised when the tether gets caught on them. They change the design at full scale without getting a working small model first. They are committed people doing many things well, but there doesn't seem to have been anyone here with an overall picture of the project, just lots of specialists having fun working on their own bits of it.

Of course there's a problem in that the team had to work to artificially short deadlines in order to get the investors to maintain their interest and continue to supply money (e.g. "if we don't manage crosswind flight by the end of the year they'll pull the plug"). The engineers definitely made decisions to fly before they were ready, and that kind of need to produce a flow of short-term results that don't aid the overall project is an intrinsic problem with relying on investors. But the investors kept paying as the system got more complicated, going from an initial fabric kite to a huge eight-motored 600kW-generating carbon-fibre beast merely as the single unit proof of concept – presumably they had nobody working for them who understood the thing either. They kept paying as conventional ground-based wind turbines got better and cheaper, and the only potentially profitable niche the project could find was deep water offshore, adding a new set of environmental challenges. Eventually they stopped paying, but it would be fair to argue that they should have done so rather earlier (obviously that's an easy decision to make in hindsight). There's basically nothing here about relations between the project team and the investors, and it would be fascinating to know what control the latter tried to exert other than by setting deadlines.

One traditional way in which capitalism externalises development costs is by letting the entity that paid them go bankrupt and allowing a new entity to buy its assets and leave the debts – that's how both Iridium and Globalstar got their starts. But nobody seems to have been lining up to buy any of the Makani hardware or patents.

Thanks to the Yorkshire Ranter for pointing me at this documentary. It's freely available on YouTube.