Rework CCS implementation

Issue

The current CCS implementation is a bit odd - each CCS technology is 'hard coded'. Thus, implementing a new CCS technology would currently require adding code.

I think it would be useful to rework CCS such that the inputs are in dictionary ( or list / object) form.

Viability/Strategy

I expect this to take in the vicinity of 10-20 hours (just a guess), but I think learning how the variables and constraints work would be useful.

From quickly looking over the optimization code, it seems most constraints involving CCS are linear combinations of CCS parameters. Sums over list (dict) comprehensions seem easy enough to implement.

Currently, each ccs (pyomo) parameter is written as an attribute of m (i.e., m.ccs2_capacity_co2[node], m.can_ccs2[node]). Resolving this issue requires looking into an alternate way of handling that. I will need to look over the Pyomo documentation for the various options.

Because of the way things are currently handled, it would be rather easy to implement one or two more CCS technologies with some boiler-plate style code additions. If that is the case, I would say resolving this issue is not worth the effort. If more than two or three different CCS technologies are to be used by the model, than I think this issue should be considered.

When reworking the CCS, several sets of inputs/outputs from the prior version should be used to cross-validate the new version.

Expected Issues

Things may get hairy with IO for csvs in this form, which would require some clever additions to those codes. Or, we could limit IO files to only JSON. Although I don't imagine reworking the outputs would be too difficult. I'm not sure I see much of a reason to continue using input csvs?

webber-energy-group / HOwDI