I haven't heard that perspective. I smell a little slant there:
1) "[speaking of wind turbines]... hiding the fact that their actual output will waver between 100 per cent of capacity and zero. In Britain it averages around 25 per cent; in Germany it is lower, just 17 per cent."
I dunno how "hidden" that is.. Turbines work when there is wind. You're not getting 100 output out of solar or wind 24/7/365.
Apparently Germany has older turbines, much less efficient than the rest of Europe.
Here's what strikes me as a valid technical issue:
One is that it becomes incredibly difficult to maintain a consistent supply of power to the grid, when that wildly fluctuating renewable output has to be balanced by input from conventional power stations. The other is that, to keep that back-up constantly available can require fossil-fuel power plants to run much of the time very inefficiently and expensively (incidentally chucking out so much more “carbon” than normal that it negates any supposed CO2 savings from the wind).
So is the point of this article that we shouldn't use renewable energy because it causes a greater variation in demand within the grid? Meeting a great variation is probably fairly inefficient, but I'd wonder if it's as expensive as not having that renewable power at all. There certainly is some inefficiency in having power plants fired up and simply waiting to meet demand that isn't there yet. That's absolutely true.
Currently, especially in Texas, we have predictive demand models that largely follow temperature trends. In high temps and low temps, there is a high demand on the grid. These temperatures are somewhat predictable within a bound of days. I'd say that we can use the same sort of weather models to predict the availability of solar power. We know when it's going to be cloudy. Also remember that in Texas, the need for more power in the grid on hot days will directly correlate with the availability of solar power.
Wind, I can't speak to so much, but I do know that we have predictive wind models, just like rain models. I use them as a pilot to look at future weather conditions.
Maybe those predictive windows aren't enough.... But basically, I think it's a problem that isn't insurmountable.