Test reVX pipeline for ReEDS technologies

[mmowers]

Why this feature is necessary:
To make sure reVX is running and producing outputs for all ReEDS technology types: onshore wind, offshore wind, urban PV, and rural PV.

I am providing onshore and offshore wind config files now. I will attach urban and rural PV config files later this week.
wind.zip

Charge code
WEPG.10348.08.01.02

Urgency / Timeframe
Onshore wind: Monday, 2020-03-02
Offshore wind: Monday, 2020-03-02, dependent on NREL/reV#124
Urban PV: TBD
Rural PV: TBD

[grantbuster]

@MRossol and @mmowers, FYI the new land-based wind farm profile method will require a few tweaks to onshore_wind_config.json.

Here's a pytest example of how the land based wind farm rep profiles are now run:

reVX/tests/test_reeds.py

Lines 301 to 309 in 2c9e7cf

	gid_col = 'sc_gid'
	profiles_dset = 'rep_profiles_0'
	weight = 'capacity'
	profiles, meta, _ = ReedsProfiles.run(cf_profiles, rev_table,
	gid_col=gid_col,
	profiles_dset=profiles_dset,
	weight=weight,
	max_workers=1)

And here's what I think the config should look like. This effectively groups the "reg_cols" then calculates the representative profile based on the capacity weighted collection of supply-curve-point-wind-farms. So in cf_profiles file, each SC point has its own aggregate wind farm profile.

  "profiles": {
    "cf_profiles": "placeholder for new land based wind farm cf profile file",
    "n_profiles": 1,
    "profiles_dset": "rep_profiles_0",
    "rep_method": "meanoid",
    "err_method": "rmse",
    "weight": "capacity",
    "gid_col": "sc_gid",
    "reg_cols": ["region", "class"]
  },

@MRossol, reV is finishing up the land based wind aggregate profiles now. Should be done in 30 min.

[grantbuster]

Alright @MRossol you should be good to go for the reV output wind files here. There are both land based and offshore data: /lustre/eaglefs/shared-projects/rev/projects/reeds_jan2020/wind

[mmowers]

Hi @MRossol and @grantbuster, I noticed a spurious space in the "mean_res_max" column header for onshore wind. Here is the corrected file:
onshore_wind_resource_classes.zip

[grantbuster]

@MRossol, sorry for the bad data. All onshore points weren't getting passed through properly. Fixed in a commit to reV/master. It's honestly kind of a testament to how flexible our pipeline is that the rest of the pipeline just kept on going with the flexible input! at least thats the brightside im choosing to see :)

The data is updated and located at the dir below. I plotted up the SC points and everything looks as expected. You've got "rep_profiles" which are really just area weighted mean profiles for each SC point including each offshore SC point. You've also got sub_type as a column in the SC table.

Let me know if you need anything else.

/lustre/eaglefs/shared-projects/rev/projects/reeds_jan2020/wind

[mmowers]

Hi @grantbuster and @MRossol, here are the PV files:
pv.zip
Note that urban and rural pv share the same resource class definitions. Let me know if you have any questions. Thanks.

[MRossol]

@mmowers
onshore and offshore wind runs executed without any apparent bugs. Can you please check the outputs ASAP to make sure they look good:
/shared-projects/rev/projects/reeds_jan2020/wind/onshore_wind/output/
/shared-projects/rev/projects/reeds_jan2020/wind/offshore_wind/output/

[MRossol]

@mmowers
rural pv runs complete here:
/shared-projects/rev/projects/reeds_jan2020/rural_pv/output/

@nickwg03 do you have an ETA on the reeds_jan2020 urban_pv gen and sc runs?

[nickwg03]

Should have them ready by Monday. I'll double check.

[mmowers]

Hi @MRossol, i can't access the onshore and offshore wind folders, could you change the permissions?

[MRossol]

POS Eagle...
@mmowers try again

[mmowers]

Thanks @MRossol! checking through some of the results now with Trieu etc

[MRossol]

@mmowers Urban PV runs are done:
/shared-projects/rev/projects/reeds_jan2020/urban_pv/output/

[MRossol]

Working on master