Fig. 8: Workflow to identify extreme days and their application in expansion planning models.

The workflow involves identifying representative days via k-means clustering on an 11-year data set of weather and demand profiles. A design optimization is performed over the representative days to identify a design strategy. An operations simulation is run with fixed design variables over the full weather and demand data. Extreme weather and demand events are identified based on the solution of the simulation. Extreme days are added to the representative days, and the design problem is resolved.