Non-REZ IDs N0/V0 not templated in network_geography

[EllieKallmier]

The NSW and VIC non-REZ IDs N0 and V0 don't appear in the IASR table renewable_energy_zones,
so they don't get templated into the network_geography table. At the moment this isn't
causing anything to break, and in templating new entrants doesn't cause problems because
we can just use a 'not-subregion' filter that leaves rows with N0/V0 geo_id grouped
with other REZs explicitly defined in the geography table. However, as we add in the
wild-card expansion system and in case other templating/translation steps end up relying
on the network_geography table, this might end up causing silent data losses.

[dylanjmcconnell]

Hey - not sure if strictly relevant? But just flagging this from the ADR in the trace-parser:

AEMO's non-REZ zones (N0, V0) parent at a real subregion like every other zone (AEMO attributes "NSW Non-REZ" to CNSW and "Victoria Non-REZ" to VIC in its new-entrants modelling inputs). This keeps every zone on the same zone -> subregion -> region chain with no special case in the resolver.

[EllieKallmier]

Good to note! I think at the moment because they're present in the summary tables we're not losing them entirely, but because we collapse down to just one geo_id we do lose the subregion/region mapping (which is only contained in that network geography table).

Also though just noticed that in v7.5 and 7.8 workbooks the V0 subregion for new entrants (summary mapping table) is set to 'SEV' - this might be an inconsistency across sources?