Change CCS policy lever to be based on % CO2 captured

[jrissman]

Currently, the CCS policy lever scales in an amount of CO2 to be captured based on an abatement potential (specified in mass units in ccs/CSA). This potential is then divided up between the electricity sector and the industry sector, and between different power plant types and different industries, using hard-coded percentages in ccs/CPbE. We have found this approach to be unsatisfactory, because the percentage allocations in CPbE don't change to reflect the BAU case and policy cases. For example, in 2.1.0-US, the model assigns about a third of the electricity sector CCS potential to biomass plants, but there are not enough biomass plants in the BAU case to take advantage of this potential, so the potential CCS amount by the electricity sector actually drops in the latest years of the model run. This is incorrect.

A modified approach will use a set of input data to specify the percentage of the CO2 emissions from each power plant type and each industry that can be captured. (Industry percentages are inclusive of CO2 process emissions.) Then the master policy lever will set what percent of this potential is achieved. Since the potential will be based on a percentage of the CO2 emissions of what actually exists in the model, it will no longer be possible for there to be a mismatch between the potential assignments (e.g. in CPbE) and what CO2-emitting things exist in a given model run.

This will likely involve removing ccs/CSA and changing ccs/CPbE to use a new format and units. We probably should subscript the main CCS policy lever by sector (electricity sector and industry sector) at the same time.

[jrissman]

Note that this builds upon work done for issue #47.

[jrissman]

Done in commit bdd2ace. As is usual with these sorts of things, it ended up taking longer than I expected. But it is working very cleanly now. It has quite a few benefits:

• It's now easier to provide input data for the CCS sector
• CO2 reductions are allocated adaptively based on which power plant types or which industrial facilities actually exist in a given model run
• New input variables allow a model adapter to specify maximum CO2 capture limits by industry and by plant type (or exclude certain industries or plant types from using CCS)
• It allows energy-related emissions and process emissions to be captured, including emissions from the fuel burned to power the CCS process itself
• The policy now scales in based on a schedule set in FoPITY, like most other policies. This policy previously used an outdated approach that required the input data itself to contain potential that scaled up with time. It's better to have the phase-in be controlled by FoPITY, as this gives the web interface user control over the timing of CSS scale-up.

The max lever bound in the web app is set to 30%, which should be a very generous and optimistic level for most scenarios. In the guidance text for the CCS policy, I mention an edge case where higher values might be reasonable, and I point out the user can download the EPS if he/she wants to build a scenario like that.

[jrissman]

While the upfront CCS-related calculations had been overhauled in bdd2ace, I hand't realized that we needed some re-wiring of the back end to allocate the CO2 emissions reductions correctly in the industry sector. 5d1d8ee completes the work by allocating industrial CO2 sequestration between energy-related and process emissions, and within the energy-related category, it allocates the sequestered CO2 by fuel. This fixes inaccuracies in various Industry-related output graphs.