Scope Princeton Net Zero

[TrentonBush]

• [x] what are outputs?
• [x] AFTER reviewing airtable questions, what looks like a good fit?
• [x] what is data size/shape?
• [x] how hard to integrate?

Come up with a quick and dirty scope and a more robust scope and present to Matt

[TrentonBush]

Princeton Net Zero America (NZA) Study

Matt and Eric hope to use this dataset as a medium-to-long term forecast of the geographic location of renewable development. This will help them prioritize their local political action. The Princeton Net Zero America simulation study explores five future scenarios of US decarbonization by 2050. Each simulation is a least-cost optimization that uses simplified models of transmission, generation, evolving demand, siting constraints, etc. The outputs of these models include shapefiles of projected wind and solar locations and transmission expansion locations.

Suitability

I don't think this study is directly suitable for the kind of forecast they want. By directly I mean taking the output shapefiles literally and simply aggregating them to the county level. This is because the NZA locations are suspiciously concentrated and don't look like the diffuse development we have seen in real RE projects over the past 20 years. Maybe there is good reason for this (constraints due to transmisison, land use, etc) but my intuition says this is a modeling artifact rather than a real effect.

The intended purpose of the NZA study was to demonstrate that national decarbonization is plausible and affordable given the resources available in the United States. I think their primary concerns were cost curves, not geography. I can imagine that two areas with nearly equal resource quality could be in very different locations, making a small amount of noise in cost produce a large change in location.

Paths forward

Post-processing

We could 'fuzz' the NZA shapes (like a gaussian blur or KDE) to try to account for uncertainty and reduce false spatial precision. This is a bit of a research project though. (How far to blur? etc)

Alternative Data Sources

I imagine nobody knows siting issues better than developers, and nobody has as complete a view of developers like investors and the independent consultants that review every project. Perhaps we can reach out to DNV GL or BlackRock for help with this?

Dataset Notes

• Each of the 5 scenarios has a set of shapefiles for RE siting and another for transmission expansion. They are on the order of 10s of MBs each.
• Attributes include: projected annual energy production, capacity, year built
• The metadata states explicitly that no updates are planned for these datasets.