I'm looking for references to studies of the actual backup requirements for widespread wind power systems where their contributions are greater than 30 - 50%. Specifically I'd like to know more about how much non-wind backup is required to maintain 100% capacity at all times and how widespread the grid interconnects would have to be. The problem obviously is that there are times of the year/day when calm winds prevail over large areas and those times aren't correlated with low power demand. The few serious studies I've seen seem to imply that even when there is wide geographic distribution, e.g. the entire upper Midwest, there would need to be on the order of 10 times the nameplate generating capacity at least 20% of the time to maintain full nameplate generation. Maintaining it for for greater than 90% of the time seems pretty impractical without almost 100% backup. It surprises me that such studies aren't widely available, they don't seem that difficult to do.