From Regions to Stacks: Spatial and Temporal Downscaling of Power Pollution Scenarios

Abstract

National energy models produce aggregate scenarios of generation capacity, energy output, and emissions. However, we need finer scales to study the impact of resource use and air pollution because timing and location determine impacts on sensitive ecosystems and human populations. We present a framework for disaggregating emissions projections to a scale compatible with air quality simulation models. The framework comprises three models that site new power plants consistent with historical patterns while recognizing water, transmission, fuel, and other factors that constrain siting, and then dispatches them consistent with those constraints. The resulting hourly emissions from individual plants are consistent with meteorology, in that peak demands and emissions occur during those hours when temperatures associated with such demands occur. Further, annual emissions vary in a way consistent with year-to-year changes in weather. An application of the framework disaggregates 2030 NO? emissions from a national electricity model in an eight-state region under two climate scenarios: no climate change (1990s) and accelerated change (2050s). Between-year variations in emissions patterns under a particular climate exceed differences between average patterns of the two scenarios. This is in part because NOx emissions are capped; thus, the total cannot change, only its distribution over time and space.