Scope integration of FERC-714 XBRL load forecast data

[zaneselvans]

• @aesharpe and @zaneselvans briefly inspected the load forecast table in the FERC-714 XBRL data to estimate how much work it will be to bring it in.
• Getting the hourly time series data in will probably be straightforward (with the exception of needing to munge timezones to get it into UTC -- but that can be based on the existing process).
• We will also need to set up a little bit of infrastructure in the FERC-714 Extract module to pull data from the XBRL derived SQLite DB, which we haven't yet had to do.
• The main source of uncertainty / risk here is the need to do some ID mapping / record-linkage that connects the recent FERC-714 XBRL Respondent IDs to:
  • PUDL Utility IDs or EIA BA IDs.
  • Old FERC-714 respondent IDs
  • EIA Utility and/or Balancing Authority IDs.
• This ID mapping can be done manually if need be -- there are only 91 respondent IDs that show up in the FERC-714 XBRL data.
• The non-FERC respondent IDs that the respondents report used to be all EIA IDs (BA or Utility -- this was somewhat ambiguous) but in the new form, there's no indication as to what kind of ID it should be so there's a wider variety of IDs being reported. Sometimes EIA, sometimes it's the same as the FERC-714 XBRL Respondent, and in a few cases it's something else entirely like an SEC ID. Also it's not clear if these are being reported consistently across the different years of data.

Particularly because of the need to deal with the new universe of IDs in the context of integrating the 3 new years of theFERC-714 Respondents table, and the need to get a bit of new infrastructure outlined for pulling tables from the XBRL derived SQLite DB, we gave this a fairly generous estimate of 50 hours. It could be less than that if the ID mapping turns out not to be too bad.