The thing that's not obvious to most people is that the power contained in moving air is proportional to the cube of the airspeed. If you're local "average win

The thing that's not obvious to most people is that the power contained in moving air is proportional to the cube of the airspeed. If you're local "average wind speed" is 5m/s (~12mph, a "Gentle Breeze" and #3 on the Beaufort scale), and you size your wind turbine to generate useful power (say, 1000W, for example) at that speed, you have two consequences due to that speed-cubed term. 1) When there's only light wind - say half your design average speed, you're only gonna get 1/8th of your 1000W target, so only 125W which isn't very useful if you've got a 1000W requirement. 2) The _bigger_ problem is when the wind blows faster than average. If you get twice average speeds, your turbine is going to generate 8kW which you'll need to safely deal with. But twice isn't your problem - in the real world it's not uncommon for 3 or 4 times average windspeed (what we'd normally just call "a windy day" and 7 or 8 on the Beaufort scale), at which stage you're going to need to safely dump 64kW of power - if your turbine hasn't been braked/parks appropriately (or torn itself to pieces.) If you're still spinning in hurricane conditions, you've got another doubling of windspeed, and another eightfold increase in power, so now your "1000W generator" is unleashing half a megawatt into your wiring and control electronics (or the plasma in the smoke where your wiring used to be...) Beaufort scale with speeds and descriptions: https://en.wikipedia.org/wiki/Beaufort_scale#Modern_scale https://en.wikipedia.org/wiki/Beaufort_scale#Modern_scale Some real world windspeed measurements showing upper end outliers of 8 times average speed: http://www.wind-power-program.com/wind_statistics.htm http://www.wind-power-program.com/wind_statistics.htm