See: https://www.sfgate.com/tech/article/california-solving-solar-power-problems-21207873.php That is a long read and followed by my long exposition. So exit now if you prefer. In summary that article described problems associated with dispatching solar and wind power to areas of demand, while balancing supplies from other generation types. It mentions the amount of time required to bring gas turbines up and down. It led to thoughts about dynamic modeling of this process. Here is my overview, and clever readers might offer improvements. The defined boxes and flows: Generation of each power type (solar, wind, hydro, gas turbine, other combustion, nuclear) has location (of source) and size. Output from solar and wind have dynamic connection to local conditions. Output from hydro might be constant until water reservoir drops to limit level, below it’s zero). Other types’ outputs might be always potentially maximum (‘nameplate values’). Each type has a spinup/spindown time; how rapidly it cam be brought online or decreased. Power distribution system in simplest sense links gnerators with users. It needs description of capacity along each distribution pathway. Power consumption by residential and industrial users are spatially specific. Their consumptions are driven differently by seasonal and daily local conditions. ‘The weather’ is known from other data sources with sufficient time and spatial resolution to ‘drive’ generators and consumers are described above. The idea is to build a model, most grandly for all western states. Or bite off a smaller piece. Where are generators, where are consumers, and how much distribution capacity is on each pathway. Goal is to describe distribution capacity required to optimize matching of supply/demand everywhere at minimum generation cost. We learn from initial reading that knowledge of spinup/spindown for generation types is important to modeling at hrs to days timescales. It would identify where power distribution systems are undersized to minimize generation costs, and by how much. Then move to separate study of costs of upgrading distribution systems. Amortize those over 50 years (or other), and see where upgrades can have most beneficial effects. Spatially explicit future climate projections are also available, so this can be forward looking as well as historical. I imagine this as a Engineering Graduate thesis. If readers imagine it as pure folly, be sure to say so
