Quoted from: https://github.com/romulus97/CAPOW_PY36
CAPOW is Python based. The model was built to simulate the operations of the major markets comprising the West Coast bulk electric power system: the Mid-Columbia (Mid-C) market, and the California Independent System Operator (CAISO).

System dynamics are driven entirely by time series of spatially distributed meteorological (temperatures, wind speeds) and hydrologic (streamflows) variables. These natural system inputs are converted to relevant power system inputs (hourly electricity demand, solar power production and wind power production, daily hydropower production). Power system inputs then drive a unit commitment/economic dispatch (UC/ED) simulation model, which minimizes the cost of meeting system-wide electricity demand, given an existing generation portfolio. The default model formulation presented here simulates the Mid-C and CAISO markets using two separate objective functions, and power flows between these two systems (i.e, between the Pacific Northwest and California) are modeled statistically. However, this default formulation can (and in some cases should) be altered to consider both systems as a single objective function, and make predictions of power flows between these systems a model output.
The primary outputs of CAPOW are hourly records of power production at the roughly 300 individual generators considered, as well as power flows along relevant transmission pathways, and wholesale electricity prices. In general, the model does not accurately reproduce historical hourly price dynamics in either the Mid-C or CAISO markets; however, the model does quite well at reproducing daily average prices in both.
The figures below illustrate the spatial extent and resolution of the CAPOW model. The CAISO system is divided into 4 zones, which cover the operations of three major utilities in the state. Exchanges of electricity (imports/exports) between the zones of the core UC/ED model and other zones in the larger Western Electricity Coordinating Council (WECC) footprint are modeled statistically.
