Post by David B. Benson on Mar 31, 2019 13:49:20 GMT 9.5
The simplified grid has no interconnections. We assume that the omnipresent load, the so-called baseload, is in fact a constant, B MW. For simplicity, the load is divided into 5 minute units, 288 per full 24 hour day. The total load is
Lt = B + It
where t is the time parameter and It is the varying, mostly daytime intermediate load. For simplicity, the occasional peak loads are ignored;assumed nonexistent.
It is highly efficient to use constantly running generators, traditionally coal and also nuclear powered to match the baseload. Assuming that the output of these constant units sums to C, we set
C = B
and only have to dispatch more flexible generators to match It as it varies over time t.
Along comes wind turbines, even lower cost to run when the wind blows. These offer generation Wt, time dependent upon the wind. But now there is no baseload for constant generators. All that is left is
B - Wt
so the constant generators can no longer supply C = B; they lose money and go out of business, being obsolete in this Brave New Climate of grid electricity generation.
Post by David B. Benson on Mar 31, 2019 14:22:53 GMT 9.5
Continuing, let the traditional variable generators sum to Dt with, by varying the generation each 5 minutes, Dt = It so that the total load is met by the total generation
C + Dt = B + It.
Now with wind generation, Wt: assume that the maximum wind generation is M and of course wind power is always first in the dispatch merit order. This moves, in effect, M MW of the baseload to variable, so-called intermediate, load. We now have
B'= B - M I't = It + M
and so the new constant generation is less and the new variable generation is more.
Post by Roger Clifton on Apr 1, 2019 15:29:50 GMT 9.5
"wind power is always first in the dispatch merit order"
Do you mean the "dispatch merit order" in the sense of the ranked prices bidded by the would-be generators of five-minute blocks of power? Wind can only be involved in this system if it can guarantee to supply across the full five minutes. That guarantee will increase the price of the bid by the wind generator, because they would have to make backup contracts with single stage gas turbines elsewhere on the grid. That is to say, wind would often be too expensive to get the priority, when they would have to feather their blades instead. It is a further inconvenient (costly) fact that their friendly single stage gas turbine would be ramping up a bill as it idled, waiting for a few minutes of duty every now and then.
While the gas component of the cost of levelised wind is more than the fuel component of its steam competitors, steam would always take priority. That is to say, baseload would always be assigned to steam.
All this is separate from that subset of wind/solar generators that have been able to establish a "must-take" relationship with the grid, outside the dispatch system, whereby their generation is injected into the grid as a raw variable. Levelling out the noise is then left as a problem for the grid operator, who must find single stage gas turbines (or hydro) willing to provide that much reserve within the dispatch system. That "priority" for wind/solar is entirely political, at the whimsy of the government of the day.
Post by David B. Benson on Apr 1, 2019 15:51:22 GMT 9.5
Roger Clifton, thank you for contributing. What I have done is simplify. Think of each wind farm as equipped with a short duration, nominally one hour, flow battery. In that case the wind farm can meet the contract.
Here in the Pacific Northwest the wind farms have must-take contracts but there is ample hydropower to compensate for the wind farm variations --- and much more. In the ERCOT Texas grid there are an increasing number of natural gas generators as more wind farms have been installed and the older coal fired generators turned off.