With wind, it's not all or nothing.
My final argument against those who argue that 100% dedicated backup for wind generation is necessary:
3) Variations in wind are to an extent predictable, and not instantaneous.
As already established in my last piece (It’s not reinventing the wheel), the system of balancing electricity generation already has a framework established to compensate for the loss of generation or deviation from the expected supply/demand balance. Often this achieved by part loading existing power plants (in other words running them at less than their maximum output). This allows those who control the system to increase or decrease their output in order to keep the whole delicate see-saw of electricity supply and demand matching.
Most of us experienced a trip at home when a light goes and the circuit breaks automatically, but at transmission scale the loss of a power plant is of megawatt or gigawatt scale. It is important to remember that these sort of trips are low probability, high impact event. It means the loss of a significant amount of power in a fraction of a second on the system. This very kind of intermittent power loss cannot be predicted in advance, yet sufficient measurements must be in place to ensure the system readdresses the balance very quickly or the result will very quickly be a loss of grid stability and force rolling blackouts.
If we now come back and think of wind generation and how it changes, its nature is very different. As opposed to being intermittent (as it is often cited), wind generation is variable because changes in wind generation vary more gradually.
Gigawatts of distributed wind do not get shed in an instant, as there is not a single point of failure like when a power plant trips, hence we have more time to respond to a moving weather front. Allowing for sufficient reserve is actually easier than for conventional power plants in this respect.
Furthermore, wind variability is also predictable. Although forecasts are still out on timing occasionally, generally the pattern and rate of change is correct.
Compensating for a gradual loss of generation that you know is coming is considerably easier than planning for a sudden shock loss of a large amount of power. When thinking about the challenge from this respect, the argument is completely flipped – if anything we need dedicated backup for our conventional generation more than for the wind added to the system!
