However, renewables, especially with increasing grid penetration and geographical extent, can be predicted, predicted with sufficient precision to be considered in the Mittellast planning -- the same way the temporary decommissioning of a large coal or nuclear power plant can be.
While no weather predictions are foolproof, the ability to predict likely wind generation has been evolving strongly over the past decade. Accuracy increases as penetration increases over wider areas, as well.
I think it's important to set a frame on how to review power generation. The frame is that we are somewhere past the beginning of a major transition period, from the old world of centralized generation to the distributed world of intermittent generation coupled with a load-following smart grid, including rate structures which allow certain industrial load to be switchable.
They are truly two different systems, and right now we're "Lost in Transition," as we re-calibrate the system. The experience in Denmark and northern Germany has provided much insight. Most importantly, the utilities involved were quite negative about windpower penetration a decade ago, but operating experience at higher and higher penetration levels turned the opinion around.
In 2008, windpower produced more than 40% of net energy consumption in three Federal states, Lower Saxony, Mecklenberg West Pomerania, and Schleswig Holstein. These are areas with some of the weakest grid infrastructure in Germany!
The German Energy Agency Dena produced a huge grid integration study in 2005, including the wind industry and the major transmission companies E.ON Netz, RWE Netz and Vattenfall Transmission. Summary here. Briefing Paper Here.
As far as I know, the study itself is only in German, though I hope I'm wrong. Ongoing work is certainly updating the results from 2005. "Life shrinks or expands in proportion to one's courage." - Anaïs Nin
Wind energy does not require construction of additional `balancing' power stations The amount of control and reserve power to balance the increasing amount of wind power has been calculated. The amounts of control energy are strongly depending on the accuracy of the short term wind power forecast, and the deviation between forecast and actual feed in. The quality of the forecasting tools keeps on improving. It is found that the additional required balancing power (positive and negative regulation power, from secondary and hourly reserves) can be provided by the remaining conventional plants. There is no need for new investments in additional power stations for this purpose to meet the balancing needs of the German system with 36 GW of wind power. An overview of the required balancing power capacities is given in table 2. In 2015 on average 3.2 GW positive regulation power representing 9% of the installed wind energy capacity and 2.8 GW negative regulation power representing 8% of the installed wind energy capacity is required.
