Despite some localized emission increases due to ramping and cycling, they estimate total aggregate avoided NOx and SO2 emissions damages in the region from wind and solar to be between 19 and 468 million dollars between 2010 and 2020.

This is resulting in a higher reliance on the local dispatchable plants that are available, and ultimately more extreme emissions outcomes. The paper documents characteristics of the communities in which these more extreme outcomes occur.

These outliers are located in areas with fewer local dispatchable plants and higher shares of renewables, and strikingly, nearly all the outlier plants are in areas with limited access to electricity markets—i.e., limited ability to trade when there is an unexpected change in wind or solar output.

For example, this figure shows key differences in heat rate and emissions rate responses to solar, versus overall emissions and generation responses:

Balancing renewables might increase emissions due to ramping or cycling—if a plant must frequently adjust output, its heat rate and/or emissions rate may increase. They find that while most plants reduce output, some increase fuel intensity and emissions rates, consistent with ramping or cycling.

They use a high-resolution empirical design to obtain operational responses for each plant in the western U.S. Most plants modestly reduce NOx and SO2 emissions in response to wind or solar, but emissions change dramatically at 15 to 20% of outlier plants. Emissions at certain plants even increase.