Model/backend/documents_parser
Khalim Conn-Kowlessar 58a9547210 Slice S0380.168: Bio-liquid mapper extensions + Table 32 FAME price flip
Mapper extensions (`_ELMHURST_MAIN_HEATING_EES_TO_FUEL_CODE`):

  "BFD": 71,  # HVO        — corpus variant oil 2 (SAP 127)
  "BXE": 73,  # FAME       — corpus variant oil 3 (SAP 128)
  "BXF": 73,  # FAME alt   — corpus variant oil 4 (SAP 129)
  "BZC": 76,  # Bioethanol — corpus variant oil 5 (SAP 126)
  "B3C": 75,  # B30K       — corpus variant oil 6 (SAP 126)

`_ELMHURST_MAIN_FUEL_TO_SAP10` water-side labels:

  "Bio-liquid HVO from used cooking oil": 71,
  "Bio-liquid FAME from animal/vegetable oils": 73,
  "Bioethanol": 76,
  "B30K": 75,

Values are direct Table 32 codes (the bio-liquid codes 71/73/75/76
don't collide with any API enum value so they pass through
`unit_price_p_per_kwh` etc. unchanged). Spec: SAP 10.2 Table 12
(PDF p.189) notes (d)/(e)/(f).

Pre-slice all 5 oil 2-6 variants raised `MissingMainFuelType` per
S0380.132. Post-mapper-extension cascade results:

  oil 2 (HVO):       SAP / cost / CO2 / PE all EXACT first try ✓
  oil 5 (Bioethanol): SAP / cost / CO2 / PE all EXACT first try ✓
  oil 3 (FAME):      SAP +17.34, cost −£398
  oil 4 (FAME alt):  SAP +16.06, cost −£367
  oil 6 (B30K):      SAP +3.05,  cost −£70

Slice S0380.131 had left a deferred TODO in `table_32.py` for FAME
code 73 ("worksheet 7.64 vs spec 5.44 — flipping has no measurable
cascade effect today, deferred until a cert that exercises it
surfaces"). Now exercised — flipping `73: 5.44 → 7.64` closes 85 %
of the oil 3/4 cost gap:

  oil 3 (FAME):      SAP +17.34 → +2.59,  cost −£398 → −£62
  oil 4 (FAME alt):  SAP +16.06 → +2.56,  cost −£367 → −£57

The Elmhurst-engine canonical 7.64 ↔ spec PDF 5.44 divergence is the
same pattern S0380.131 applied to heating oil (code 4: 7.64 → 5.44)
per [[feedback-software-no-special-handling]].

Remaining residuals on oil 3 / oil 4 / oil 6 are cascade-side
(HW kWh under by ~250-900, SH demand small diff, CO2/PE blend
artifacts) — pinned at observed values as forcing functions for
follow-up slices. Open fronts:
  - HW kWh discrepancy on FAME (cascade applies different efficiency
    path than Elmhurst for SAP codes 128/129)
  - B30K (oil 6) Δcost −£70 with prices matching: SH/HW kWh gap

Closures `oil 2` / `oil 5`: ±0.0000 on all 4 metrics. Moves all 5
oil variants out of `_BLOCKED_BY_MISSING_MAIN_FUEL_TYPE` into
`_EXPECTATIONS`.

Blocked tier now: 6 variants (community heating × 5, no system).
Cascade-OK tier: 32 variants (up from 30), 30 EXACT + 3 (oil 3/4/6)
pinned with non-zero residuals + 1 (pcdb 1 SH residual closed in
S0380.165).

Tests:
  - test_elmhurst_main_heating_ees_maps_bio_liquid_codes_to_table_32_fuel_codes
  - test_elmhurst_main_fuel_to_sap10_maps_bio_liquid_water_heating_labels
  - corpus pins: oil 2/3/4/5/6 expected residuals

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-06-02 10:14:10 +00:00
..
handler address JTK review comments 2026-04-20 15:11:17 +00:00
tests Slice S0380.168: Bio-liquid mapper extensions + Table 32 FAME price flip 2026-06-02 10:14:10 +00:00
__init__.py Map to RdSapSiteNotes from site notes JSON 🟥 2026-04-16 13:54:03 +00:00
db_writer.py include updating epc_property_data to pashub to ara workflow 2026-04-29 09:55:14 +00:00
elmhurst_extractor.py Slice S0380.140: §4 cylinder storage loss — extractor picks up §16 thermostat lodging + Table 2b note b restricts ×0.9 to boiler/warm-air/HP systems 2026-05-31 19:03:58 +00:00
extractor.py Handle wall thickness "Unmeasurable" 🟩 2026-04-30 16:41:16 +00:00
local_runner.py update local runner to work for elmhurst 2026-04-24 14:01:36 +00:00
parser.py load ecmk site notes to db 2026-04-29 11:20:47 +00:00
pdf.py update local runner to work for elmhurst 2026-04-24 14:01:36 +00:00