Add the FK relationship to the new EIA860 storage direct support plant/generators

[cmgosnell]

There are new sets of plant_id's and generator id's in the core_eia860__scd_generators_energy_storage table which are all about whether a storage generator supports another generator. Goal of this issue is to add these new id's into special FK relationships. It is possible that we will need to harvest these new ID's in order to actually have all of the ID's in the harvested generators table. Another possible barrier is that there definitely nulls in these columns. Is it okay to have nulls in a column with a foreign key relationship to a table that doesn't have any nulls?

These are the columns at question:

"plant_id_eia_direct_support_1",
"generator_id_direct_support_1",
"plant_id_eia_direct_support_2",
"generator_id_direct_support_2",
"plant_id_eia_direct_support_3",
"generator_id_direct_support_3",

Running pudl_check_fks without the new alembic schema produces the following errors:

pudl.etl.check_foreign_keys.ForeignKeyErrors: Foreign key error for table: core_eia860__scd_generators_energy_storage -- core_eia__entity_generators (plant_id_eia, generator_id, plant_id_eia, generator_id, plant_id_eia, generator_id) -- on rows [1523 1548 1576 1589 1596 1727 1728 1795 1796 1827 1864 1865 1879 1892
 1895 1909 1920 1922 1934 1943 1946 1977 2028 2046 2059 2069 2088 2101
 2102 2105 2108 2109 2110 2111 2112 2116 2125 2126 2133 2143 2224 2236
 2315 2316 2317 2408 2448]

Which have the following values:

	plant_id_eia_direct_support_1	generator_id_direct_support_1	plant_id_eia_direct_support_2	generator_id_direct_support_2	plant_id_eia_direct_support_3	generator_id_direct_support_3
1523	NaN	None	NaN	None	NaN	None
1548	NaN	None	NaN	None	NaN	None
1576	NaN	None	NaN	None	NaN	None
1589	NaN	None	NaN	None	NaN	None
1596	NaN	None	NaN	None	NaN	None
1727	62975.0	SYNLB	NaN	None	NaN	None
1728	NaN	None	NaN	None	NaN	None
1795	63490.0	WIL	NaN	None	NaN	None
1796	NaN	None	NaN	None	NaN	None
1827	64185.0	DRAPV	NaN	None	NaN	None
1864	64436.0	WLB	NaN	None	NaN	None
1865	NaN	None	NaN	None	NaN	None
1879	64520.0	SOL1	NaN	None	NaN	None
1892	NaN	None	NaN	None	NaN	None
1895	NaN	None	NaN	None	NaN	None
1909	NaN	None	NaN	None	NaN	None
1920	64852.0	DAGGB	64852.0	DAGGP	NaN	None
1922	NaN	None	NaN	None	NaN	None
1934	NaN	None	NaN	None	NaN	None
1943	65055.0	77FPV	NaN	None	NaN	None
1946	65058.0	WAIBA	65058.0	WAIPV	NaN	None
1977	NaN	None	NaN	None	NaN	None
2028	66076.0	20409	NaN	None	NaN	None
2046	NaN	None	NaN	None	NaN	None
2059	NaN	None	NaN	None	NaN	None
2069	NaN	None	NaN	None	NaN	None
2088	67457.0	SHPV1	NaN	None	NaN	None
2101	61169.0	60798	NaN	None	NaN	None
2102	NaN	None	NaN	None	NaN	None
2105	NaN	None	NaN	None	NaN	None
2108	61720.0	30	NaN	None	NaN	None
2109	61722.0	32	NaN	None	NaN	None
2110	61752.0	49	NaN	None	NaN	None
2111	61807.0	66	NaN	None	NaN	None
2112	62494.0	SCSOL	NaN	None	NaN	None
2116	NaN	None	NaN	None	NaN	None
2125	63541.0	63257	NaN	None	NaN	None
2126	NaN	None	NaN	None	NaN	None
2133	NaN	None	NaN	None	NaN	None
2143	NaN	None	NaN	None	NaN	None
2224	NaN	None	NaN	None	NaN	None
2236	NaN	None	NaN	None	NaN	None
2315	65550.0	64921	NaN	None	NaN	None
2316	65550.0	65789	NaN	None	NaN	None
2317	65550.0	None	NaN	None	NaN	None
2408	NaN	None	NaN	None	NaN	None
2448	67282.0	5654	NaN	None	NaN	None

### Tasks
- [x] Add new `foreign_key_rules` into `core_eia__entity_generators` for each of  these direct support column names for the plant/gen ids. (examples of other tables that do this are: `core_eia__entity_utilities`, `core_eia860__scd_utilities`)
- [x] Regenerate the alembic schema
- [x] test the relationships w/ `pudl_check_fks`
- [x] fix plant ID dtypes
- [x] Rule out incorrectly written string nulls in SQLite DB by directly querying in DB Browser
- [ ] debug NaNs / Nones failing the FK check

[e-belfer]

When running just on the direct_support_1 and with the field type updated for the plant IDs:

[1523 1548 1576 1589 1596 1727 1728 1795 1796 1827 1864 1865 1879 1892 1895 1909 1920 1922 1934 1943 1946 1977 2028 2046 2059 2069 2088 2101 2102 2105 2108 2109 2110 2111 2112 2116 2125 2126 2133 2143 2224 2236 2315 2316 2317 2408 2448]

[e-belfer]

Adding these IDs to the harvesting process seems to have fixed the issue, and pudl_check_fks is now passing on #3699.