Reduce columns of Generators_variability.csv file

[xuqingyu]

Hi Colleagues,
It is likely that some generators share the same time-series. For example, most thermal units use all 1s and many wind/solar farms that only differ in CAPEX.
I would like to suggest creating a column called VARIABILITY in generators_data.csv that store the column names of Generators_variability.csv file. In load_generators_variability.jl, we reorder the columns in gen_var according to the entries in VARIABILITY.

So that the size of Generators_variability.csv can be greatly reduced, and new generators that share the same variaiblity can be easily added without modifying Generators_variability.csv file.

I have tested it successfully.
Best,
Qingyu

[cfe316]

Hi Qingyu,

• Note that thermal units that are all 1's can already be removed, as any missing columns are automatically added as a vector of 1s.

I think this is generally a good idea, and I've been thinking of a variant of this, but I like your concept as well.

Right now there's an ugly limitation that (internally) pP_Max is a matrix which is indexed by resource number. In general, as you mention, not all resources need a timeseries --- but some might want multiple timeseries. I think another good solution would be

• Make the matrix pP_Max[t,y] into a more general df_timeseries[t, :column] dataframe.
• Then :column could be a resource name (FRCC_solar_1) or a more general zone + resource-type (FRCC_solar), as you describe
• or even could be something more specific like ISONE_hydro_1_inflow with another ISONE_hydro_1_minoutflow for inflow and outflow profiles, respectively.

Different types of resources could handle a VARIABILITY column differently (perhaps with intelligent defaults, like searching for a column with the same name as the resource).