Areva, the French nuclear group, is strongly pro-wind.

In the long run, we're all dead. John Maynard Keynes

One is that at present levels of renewables in medium-sized countries (e.g. Germany, the UK, France, to some extent importing-from-France Spain, but not mismanaged Italy), already existing excess capacities are enough. The main reason is that excess capacities were built to be capable to jump in when a large power plant, say a four-block nuclear power plant with 3.2 GW, breaks down - and this excess capacity is even ensured on a regional basis. Intermittance fluctuations are of the same magnitude.

Second, as there are economies of scale, renewables intermittance (the part that is not day/night or seasonal), especially short-term fluctuations, is also reduced in a larger system (at the price of transmission losses) - e.g. when whole weather systems are covered, the wind always blows somewhere. This is of a bit limited worth for Europe, with most wind being along the Atlantic Coast that may have low winds on the entire length, but for example for the USA, most of the intermittance could be balanced thus. (BTW, you may check in real-time the intermittance of Spanish wind power here - where I note Spanish wind is strongly concentrated in one small region, into Galicia province which is the part above Portugal. The average level to compare to is around 25% maximum capacity.)

Third, note that different renewables could balance each other's intermittance. Most directly hydro and wind (as already practised to some extent in Scandinavia): hydro would be run at varied rather than constant power to balance wind's intermittance (something that also revitalises floodplains and the river ecosystem downriver), and the water spared during strong winter winds means higher reservoir levels during summer droughts. For a more distant future,  the balancing of solar and wind is another issue: intriguingly, the day/night and weather patterns of these can be combine to give not a constant power, but one roughly following the daily human usage curve!

Fourth, but this again is an issue for later, there are non-ittermittent renewables that could provide excess in the system: tidal and geothermal. I'm a bit sceptical about the former (low total potential, possible danger to marine life); as for geothermal, it is still rather expensive if we disregard non-renewables external costs, but the potential is there (I wrote more on geothermal, again with a focus on Germany, here).

_{*Lunatic*, n.
One whose delusions are out of fashion.}

20% of kWh means a lot more in terms of capacity (MW) as windfarms typically produce one third or less of the kWh of the baseload plants per unit of capacity due to their intermittent nature (i.e. a 200 MW wind farm produces one third of the actual electricity of a 200 MW gas powered plant).

So the cost of intermittence on the system is NOT a valid argument today and will not be for a long time to limit the construction of windfarms.

In the long run, we're all dead. John Maynard Keynes