Can I just point out that the Imperial College paper summarized above makes a good case for saying that over- and under-production (from wind) are not major problems, for the years to 2020 and looking immediately beyond?

The discussion (of what to do with excess wind output) seems posited on a vague future state first (I think) evoked in this thread by Thomas.

Not that the uses excess output could be put to is not a valid debate.