3 ms·

I'm not saying he is wrong, but he makes some bold claims without giving any proof. Some examples: > Twelve blades is a nice number. Why twelve? Why not 50,

I'm not saying he is wrong, but he makes some bold claims without giving any proof. Some examples: > Twelve blades is a nice number. Why twelve? Why not 50, it's a nice number as well. > Lipps turbines can operate in faster winds [...] turbines could be allowed to run 24/7 in any wind with no computer So even in the strongest winds turbines with 40 feet blades do not need to be stopped? > using permanent magnet generators instead of alternators, but those are more expensive > Use permanent magnet generators leading to [...] even lower cost. Which is it now? > what matters isn’t power efficiency per turbine so much as power production per acre of wind farm. Isn't it rather electricity production efficiency in terms of invested capital?

amelius 10y ago
I get the feeling that the best answer to the question is that empirical testing (perhaps by simulation) points out that three is the optimal number of blades. This leads me to the question: is it possible to determine a good rotor design by using optimization techniques such as simulated annealing or genetic programming? Or are the simulations too costly?
- Bartweiss 10y ago
  I can help with this a bit. The simulations are quite difficult - flow simulation is well-established, but it's a big step from that wind farms. You need to simulate many turbines interacting at a specific site, under a variety of conditions. That's not the problem with modeling single turbines, though. Rotor design is hard to simulate productively for the same reason that Cringely's analysis is flawed: many of the biggest factors are unrelated to everyday running. - Even blade counts create resonance patterns that stress your tower. - High solidity designs suffer more stress in storms (even while stopped), so they break more - Noise interferes with putting rotors near houses or over pastures - Storage is difficult and consistency is valuable, so expanding the operating range up to high winds is less useful than expanding it down to low winds And so on. If you model pure efficiency, the sims are pretty easy. If you model real value, you either can't do it, or you get a search space so jagged that finding maxima is hopeless.

amelius 10y ago

I get the feeling that the best answer to the question is that empirical testing (perhaps by simulation) points out that three is the optimal number of blades. This leads me to the question: is it possible to determine a good rotor design by using optimization techniques such as simulated annealing or genetic programming? Or are the simulations too costly?

Bartweiss 10y ago
I can help with this a bit. The simulations are quite difficult - flow simulation is well-established, but it's a big step from that wind farms. You need to simulate many turbines interacting at a specific site, under a variety of conditions. That's not the problem with modeling single turbines, though. Rotor design is hard to simulate productively for the same reason that Cringely's analysis is flawed: many of the biggest factors are unrelated to everyday running. - Even blade counts create resonance patterns that stress your tower. - High solidity designs suffer more stress in storms (even while stopped), so they break more - Noise interferes with putting rotors near houses or over pastures - Storage is difficult and consistency is valuable, so expanding the operating range up to high winds is less useful than expanding it down to low winds And so on. If you model pure efficiency, the sims are pretty easy. If you model real value, you either can't do it, or you get a search space so jagged that finding maxima is hopeless.