One of the key issues with renewable integration is understanding how variable generation (VG) resources, such as wind and solar, interact with the rest of the power system. As the output of VG varies up and down, the rest of the generation fleet (mostly hydro and thermal generation) needs to balance out these variations so the supply of electricity always matches the demand. So, the ability of conventional generators to vary output is an important consideration when integrating significant quantities of renewables.
When VG output is very high (say, in the middle of a nice, sunny day in California), the rest of the conventional fleet may need to reduce output to a very low level. But many of these power plants have operational constraints that prevent them from ramping down or shutting off, even when renewable resources are producing cheaper energy.
This “minimum generation” problem is now widely recognized as an important constraint on our ability to add renewable energy to the grid. If we cannot reduce the output of conventional generators, then we are forced to curtail (throw away) perfectly good wind and solar generation; this is why we worry about the “belly” of the duck curve.
Given the impact of this issue on renewable deployment, integration researchers — including us at the National Renewable Energy Laboratory(NREL) — have spent a lot of time gathering data sets on thermal and hydro generator characteristics. This data is mainly targeted towards the grid modeling community, but this topic has really important policy implications for understanding how much wind and solar we can cost-effectively put onto the grid.
In the paper, we do two things. First, we summarize data sources that grid modelers can use to improve how they evaluate whether we can cost-effectively integrate wind and solar into the grid. These data sets show how many power plants (at least historically) don’t really vary their output in response to electricity prices and could, therefore, contribute to the minimum generation problem.
Second, we show how incorrectly assessing plant flexibility can over- or underestimate our ability to integrate wind and solar. Our analysis shows that evaluating conventional power plant flexibility may be just as important as evaluating other sources of grid flexibility such as energy storage. It also demonstrates that as more renewables are added to the power system and concerns about “bottoming out” the system grow, we need to continually evaluate and re-evaluate the various flexibility options available to utilities and other system planners.
So what does this mean? We have a long way to go in studying the role of conventional generators in providing new sources of grid flexibility, especially as state and local policies continue to push for ever-higher levels of VG on the grid.
These findings are described in the article entitled How low can you go? The importance of quantifying minimum generation levels for renewable integration, recently published in the journal Energy Policy. This work was conducted by Paul Denholm, Greg Brinkman, and Trieu Mai from the National Renewable Energy Laboratory.
1 The duck curve is a famous/infamous chart published by the California Independent System Operator that shows how PV produces a pattern of net demand that resembles the profile of a duck. It is now widely used as an example of the challenges of integrating large amounts of PV on the system.