Models

Small interactive models, each built to make one mechanism visible.

These are teaching toys, not research. They run on fixed coefficients, nothing in them is optimized, and no number they produce should be cited. What they are good for is the shape of an argument — the thing that is hard to see in a table and obvious once you can move it.

Where the numbers come from

Every coefficient in these models says which kind it is — a published figure from the IEA, EIA or USGS, or an order-of-magnitude estimate made for the page. Roughly a third are estimates. That is fine for arguments that turn on ratios of two hundred to one rather than on ten percent accuracy, but it is not fine to hide, so the label travels with the number and every model carries its figures as a table as well as a picture.

Where a model cannot compute something, it says so and does not draw it. The day-shape diagram on What a Grid Is Made Of is a schematic that deliberately does not respond to the controls, because the model underneath has no time resolution and a diagram that reacted to the sliders would be claiming to have calculated something it has not.

The real versions of these questions — a mix that is chosen rather than given, under capacity limits, lead times and recovery — are the actual research, and do not fit on a web page. That work lives in the lab and in the papers.