Robert Poore's findings were even more significant. Poore's firm, Global Energy Concepts, studied three groups of turbines to determine just how intermittent and unpredictable wind resources really are. The study finds that for 95% of the time, changes in wind speed and power output varied by only 1%. For 99.9% of the time, changes in a project's output will not exceed 20% of that project's rating. Other findings are that projects running at high capacity have less of a chance of wide variations and the probability of significant changes also declines with more turbines and the farther they are spread apart.
Lastly, a technical paper by Michael Milligan of NREL finds that wind plants in Minnesota have a much higher probability of meeting load than previously thought. At one extreme, some transmission organisations believe wind resources need to be backed up with 100% spinning reserves, but Milligan says a wind project really needs spinning reserves "somewhere around 5% capacity of the wind plant."