An instantaneous point-of-use electric water heater (WHC 907/909, no
cylinder) on an off-peak tariff was billed 100% at the off-peak LOW rate
by the generic electric-off-peak else-branch. But SAP 10.2 §12 (PDF line
2680) computes the off-peak on-peak HW proportion via Table 13, "dependent
on the total floor area and the CYLINDER size" — it presumes a stored-water
cylinder charged overnight. An instantaneous heater has NO cylinder, heats
on demand, and cannot shift to the off-peak window, so 100% of its
consumption bills at the HIGH rate (Table 12a WH high-rate fraction 1.0).
Both the SAP-cost rate (`_hot_water_fuel_cost_gbp_per_kwh`) and the
ADR-0014 / CO2-PE fraction (`_hot_water_high_rate_fraction`) fixed
consistently (low-rate scalar -> 7-hour high rate; fraction 0.0 -> 1.0).
Localised by deep-diving corpus cert 74061136 (HHR-storage mid-floor flat,
WHC 909): PE matched lodged (+1.6, roof+floor zero-loss) while SAP over-rated
+7.72 — the cost-only signature. Its DHW was ours £59 vs lodged £344 (5.8x);
the tariff was the whole gap. Fix: +7.72 -> -1.25 (residual is separate small
fabric). 7 corpus certs carry electric-instantaneous DHW on off-peak; the 3
outside 0.5 all move sharply inward (MAE the win, not within-0.5 crossings).
RdSAP 10 §12 tariff routing confirmed spec-correct (Unknown meter + storage
409 -> off-peak 7-hour, Rule 2) — the bug was the DHW rate, not the tariff.
Unit-pinned in test_cert_to_inputs; RealCertExpectation 74061136 = 72.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The gov-API `secondary_fuel_type` enums 5 (anthracite) / 9 (dual fuel) /
33 (coal) collide in VALUE with same-valued RdSAP-10 Table-32 codes for
OTHER fuels (5=bulk LPG, 9=LPG SC11F, 33=elec 10h-low). The main-fuel
boundary and the CO2/display `_secondary_fuel_cost_gbp_per_kwh` already
canonicalise these via `canonical_fuel_code`, but the SAP-DRIVING
`_fuel_cost` cascade had a SEPARATE inline secondary rate that passed the
raw enum straight to the price lookup — so the two paths diverged and the
cost path silently mis-priced.
Anthracite (enum 5) was billed at the bulk-LPG code-5 rate 12.19 p/kWh
instead of 3.64 — a 3x over-cost on the secondary that under-rated every
solid-fuel-secondary dwelling. Worst: corpus cert 100050355518 (semi,
2 extensions, anthracite room-heater secondary) lodged 36 / engine 20.8
-> 35.5. Its CO2/PE already matched lodged (the demand cascade was right)
— a pure cost-side gap, which is what localised it.
Fix: canonicalise `secondary_fuel` before the Table-32 lookup in
`_fuel_cost`, mirroring the main-fuel boundary. 55 corpus certs carry a
colliding secondary.
Gauge: within-0.5 75.5% -> 77.0%, SAP MAE 0.708 -> 0.668 (floors
ratcheted). The prior enum-9 test only guarded the display helper; new
test_fuel_cost_secondary_colliding_fuel_priced_at_canonical_rate pins the
_fuel_cost cost path directly. RealCertExpectation pinned for
100050355518 (36 = lodged).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
RdSAP 10 §5.11.4: a pitched roof lodging thickness "NI" with description
"...insulated (assumed)" was billed at the §5.11.4 observed-retrofit
50 mm row (U 0.68) while the SAME description lodged with "ND" fell
through to the Table 18 age-band default (0.40 for bands A-G). The
ND-vs-NI sentinel is lodging-software noise; "(assumed)" means the
insulation PRESENCE is an age-band assumption, not an observed
retrofit, so Table 18's "thickness cannot be determined" clause
governs both forms. The bare "insulated" description (observed
retrofit, no qualifier) keeps the 50 mm row.
Evidence: the 27-cert loc4+NI corpus cohort was systematically
under-rated (~-1 SAP; zero positive movers among "(assumed)" certs),
vs the 102-cert loc4+ND cohort already on Table 18 at ~0 bias.
Gauge: within-0.5 74.2% -> 75.5%, SAP MAE 0.721 -> 0.708, CO2 MAE
0.09 -> 0.074, PE MAE 3.5 -> 3.2 (floors ratcheted). Unit-pinned in
test_rdsap_uvalues (assumed -> Table 18; bare-insulated -> 0.68
regression guard); integration pin in test_heat_transmission updated;
RealCertExpectation pinned for 10094975827 (66 = lodged, was -1.87).
Ledger: C002 103001004 deferred pending Elmhurst arbitration
(single-immersion off-peak pricing + declared cylinder loss vs
Table-2); C006 quadruplet marked ⚠ (LIG-21.0 software honours lodged
0 extract fans literally vs the Elmhurst-validated 0=unknown->default
convention, worksheet case 46).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Setup for the RdSAP-21.0.1 corpus gauge campaign (74.2% within-0.5,
SAP MAE 0.721 at HEAD, 1000/1000 computed):
- scripts/corpus_1000/build_worklist.py — runs all 1000 corpus certs
through the gauge's own path (from_api_response -> Sap10Calculator),
ranks |engine - lodged| descending, clusters by dwelling signature,
and (re)writes scripts/corpus_1000/worklist.md preserving per-cert
statuses and notes. 258 certs outside 0.5 in 249 clusters.
- C001 (worst cert, Δ +23.9): uprn 4510053280, ground-floor flat 47 m²
on ASHP PCDB 100053 (Mitsubishi Ecodan 5 kW, PSR table 0.2-2.0).
Dwelling PSR 2.031 -> SAP 10.2 N2 footnotes 44/45 heat-pump extension
(reciprocal interpolation toward 100% at 2x largest PSR) = 305%
space efficiency, which accredited Elmhurst also applies (golden
case 56, record 100061). The LODGED software instead treated the
out-of-range record as invalid and billed 100% direct electric +
standard schedule (reproduces lodged CO2 to 1%). Lodged-software
methodology gap, worklist ⚠, engine untouched; observed engine 75
pinned in test_real_cert_sap_accuracy.
Gauge floors unchanged (no engine change).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Portfolio 814 / scenario 1271 audit (2026-07-03): all 26 properties lodging
has_hot_water_cylinder with insulation type 0 (none) or NULL over-rate their
effective baseline by up to +18 SAP (avg +12.7) because
_cylinder_storage_loss_override returns None for any insulation label it
can't resolve and the worksheet keeps the zero-storage-loss combi default.
Same-lodged neighbours split by 19 effective points (WD5 0DP pids
742215/742216, both lodged 54, effective 50 vs 69) and the inflated members
distorted neighbour-divergence cohort medians (E17 6EB pid 742355 Δ+18).
New HIGH check cylinder-storage-loss-dropped fires on the deterministic
input pattern (cylinder present, insulation type ∉ {factory, loose jacket}
or thickness missing) — 28/338 on the motivating portfolio (the 26 above
plus 2 same-pattern rows with no lodged score). The calculator fix
(uninsulated Table 2 loss, or RdSAP Table 29 age-band default) is separate
follow-up work.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
39 Hyde-796 property_overrides rows ('100% secondary glazing (sap 9.94)')
plus the classifier-cache row; the cache phase self-gates on the
'Secondary glazing' pgEnum label landing (assessment-model#345).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Addresses PR #1402 review: the guard only claimed dominant-single, leaving
the symmetric bug open — the LLM can flatten a dominant-double/triple split
onto its minority single (e.g. "96% Double glazing 2002 or later, 4% Single").
A dominant double/triple whose era is explicit ("pre-2002" / "2002 or later")
is just as fully determined as era-free single, so the guard now claims it via
_DOMINANT_MEMBER. Only a genuinely ambiguous era ("unknown age", unstated) still
defers to the LLM — the "96% double -> None" contract now holds solely for the
era-unknown case, not the era-stated one.
Backfill script reuses the same guard, so it now corrects any dominant split;
renamed reclassify_dominant_single_glazing.py -> reclassify_dominant_glazing.py
to match. Tests cover double/triple x pre-2002/2002-or-later and the still-
deferred unknown-age case; 14 green.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The landlord property-type description is a "<dwelling type>: <built form>: <floor>"
split whose leading token IS the dwelling type; the built-form tail is not. The
LLM occasionally over-read the tail and flipped the type — a handful of
"Bungalow: EndTerrace" / "Bungalow: MidTerrace" dwellings were stored as House.
Adds property_type_guard (claims the recognised leading token: House / Bungalow /
Flat / Maisonette / Park home; defers unrecognised phrasings to the LLM) and wires
property_type through a GuardedColumnClassifier, so the built-form tail can never
flip the type and the live path is deterministic.
Applied the scoped backfill to portfolio 796 (Hyde): 3 rows corrected from House
back to Bungalow. No enum migration needed — the targets are original members.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
`plan-stops-short-of-goal` fires on anything below the goal band — 322/338
on portfolio 814, mostly flats a single band short of B, which drowns the
real hits. Add a sharper, band-keyed check that isolates the genuinely-
stuck tail: a plan that leaves the dwelling still rated E/F/G with the band
UNCHANGED from baseline on an unlimited-budget scenario — the "E->E, no big
movement" worklist. On portfolio 814 / scenario 1271 it flags just 4 (pids
742121, 742210, 742265, 742347) vs 322. No SAP-points threshold; the
strictly-worse case is left to plan-below-baseline-band so the two
partition cleanly.
Also refresh the find-weird-recommendations skill's Phase 4 catalogue:
register the new check in Phase 1, correct the stale 742121 note (it is an
electric maisonette whose HHRSH IS offered but scores -6.3 SAP, so the
Optimiser correctly drops it — not the old mains-gas story), and add the
742265 community-flat findings (HHRSH/ASHP gated out by heat-network
topology; vaulted-ceiling roof with ND insulation).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
MainFuelType had no individual wood-logs member — only "biomass (community)" —
so the LLM classifier funnelled "Solid Fuel: Wood Logs" into the community fuel,
inventing a community heat network the dwelling isn't on (and mislabelling the
connection). main_fuel had no deterministic guard at all, so nothing caught it.
Verified against domain/sap10_calculator/docs/specs: RdSAP 10 Specification
Table 32 lists "wood logs" as a solid fuel (code 20, 0.028 kgCO2e/kWh); the
calculator's input scheme (the gov EPC API fuel enum) codes it 6 -> Table 32 20
(sap_efficiencies._API_TO_TABLE32), and water_heating_overlay already pins the
same fuel to 6. So _FUEL_CODES["wood logs"] = 6 is confirmed, not guessed.
Adds MainFuelType.WOOD_LOGS + the _FUEL_CODES entry, a main_fuel_guard mirroring
water_heating_guard (claims the "wood log" token; dual fuel keeps its own member
since it has no "wood log" substring), and wires main_fuel through a
GuardedColumnClassifier so the live path is deterministic.
Applied the scoped backfill to portfolio 796 (Hyde): 21 rows off
"biomass (community)" -> "wood logs". property_overrides (TEXT) only; the
classifier-cache pgEnum member is deferred to the FE Drizzle migration.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A landlord glazing split like "4% Double glazing 2002 or later, 96% Single
Glazing" is a near-uniform *single*-glazing dwelling, but the LLM classifier
latched onto the 4% minority and wrote a "Double glazing…" override — over-
crediting the dwelling.
glazing_mix_guard rescued genuine mixes (no type >= 90%) to MIXED but returned
None for near-uniform splits, trusting the LLM to have applied the dominant type.
It hadn't. Single glazing is era-free, so a dominant-single split is fully
determined: the guard now returns SINGLE for it (dominant double/triple still
carry era ambiguity, so they stay the LLM's job — the existing "96% double" case
still defers).
Adds a portfolio-scoped backfill (reuses the same guard, so it can't drift from
the live path) and applied it to portfolio 796 (Hyde): 115 glazing overrides
corrected from a spurious Double back to Single.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The neighbour SAP-divergence cohort was (postcode, property_type, built_form),
which mixed electric- and gas-heated dwellings. Electricity scores materially
lower in SAP than mains gas for identical fabric, so a mixed-fuel postcode
produced cross-fuel false outliers — an electric dwelling flagged only for being
electric among gas neighbours.
Add the main-heating fuel class to the cohort key (electricity variants — 29/30
+ off-peak 31-40 — collapse to one class; every other code is its own bucket).
Now every flagged divergence is a same-fuel comparison worth investigating. On
portfolio 814 this refined 20 raw divergences to 17 same-fuel ones while keeping
the genuine within-fuel outliers (e.g. the all-electric WC2B 4AW flats at SAP
26/38 vs a cohort median of 67).
Reaches the fuel via epc_main_heating_detail through the indexed property_id —
still never touches the recommendation table. Surfaced by the portfolio-814
recommendation audit (Work item B, #1388).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
One-time script (dry-run default, --apply in a transaction, idempotent) reusing the
live glazing_mix_guard so the backfill and classifier cannot drift. Maps every
genuine percentage-mix override (neither type >= 90%) off a flattened single type
onto MIXED (no overlay -> the cert's per-window glazing is kept). property_overrides
TEXT updated now; the glazing pgEnum cache write for 'Mixed glazing' is deferred
until the FE adds the member. Dry-run against the audited DB reports 647 rows.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A focused sibling to audit-ara-portfolio: that skill audits baselines/plans/SAP;
this one audits the *recommendations themselves* — why a measure was or wasn't
offered. Motivated by the portfolio-814 review (Khalim's HHRSH-on-community-
heating, missing-HHRSH, missing-secondary-heating-removal, and a neighbour split).
Adds:
- .claude/skills/find-weird-recommendations/SKILL.md — scan -> neighbour scan ->
live re-model deep-dive -> root-cause -> codify, with a seeded known-bug
catalogue and the query-safety rules inherited from audit-ara-portfolio.
- scripts/audit/anomalies.py: new `plan-stops-short-of-goal` HIGH check — the
default plan ends below the goal band on an unlimited-budget scenario (the
deterministic worklist for "why didn't this get recommended X"). Adds
scenario_budget to the bundle/query so budget-capped scenarios are excluded.
- scripts/audit/neighbour_divergence.py: groups a portfolio by (postcode,
property_type, built_form) and flags effective-SAP outliers vs the cohort
median. Never touches the 26m-row recommendation table, so it is safe
portfolio-wide.
- Tests for both (12 passing).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
One-time script (dry-run default, --apply in a transaction, idempotent) mapping a
flat roof's stated depth (flat: Nmm) onto the nearest Table 16 rung member so the
overlay scores it by thickness, not the age-band default. Updates property_overrides
(TEXT) immediately; the roof_type pgEnum cache writes are deferred until the FE adds
the FLAT_*MM members. Dry-run against the audited DB reports 102 rows, deferring
Flat 50/100/150 mm.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
One-time script (dry-run default, --apply in a transaction, idempotent) that maps
every party-ceiling override (another/same dwelling or premises above) currently
stored as an external roof value onto the party-ceiling member, reusing the live
roof_party_ceiling_guard so the backfill and the classifier cannot drift. Updates
property_overrides (TEXT, the modelling read path) and the roof classifier cache;
dry-run against the audited DB reports 106 rows. No FE migration — the party-
ceiling values already exist in the roof pgEnum.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
One-time script (dry-run default, --apply in a transaction, idempotent) that
NULLs the four lodged_* columns on every predicted-source baseline row — a
predicted EPC exists and no lodged one does — leaving Effective, the bill block,
and rebaseline_reason intact. Dry-run against the audited DB reports 12,236 rows.
Must run after the FE-owned Drizzle ALTER ... DROP NOT NULL lands.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Deterministic catalogue remap (no LLM) that re-resolves the stored archetype for
the canonical RdSAP heating descriptions the under-populated taxonomy
mis-classified. Dry-run by default; --apply writes inside a transaction;
idempotent. Updates property_overrides.override_value (TEXT — what the modelling
reads, the actual band-G fix: 3,028 rows incl. 2,210 community boilers + 770
panel/convector room heaters). Cache (the Drizzle-owned main_heating_system
pgEnum) updates are deferred per-value until the FE-repo enum migration adds the
new archetypes — surfaced explicitly by the script.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replaces sfr/principal_pitch/2_export_data.py, which read the retired
plan_recommendations m2m and recommendation_materials table. The new model
links a recommendation to its plan directly (recommendation.plan_id), keeps
materials inline on the recommendation (material_id), marks the chosen plan
per (scenario, property) with is_default, and stores post-works SAP/EPC and
savings on the plan row (the new SAP calculator's output).
Takes a portfolio id, resolves every modelled scenario (those with plans),
and writes one workbook with a properties sheet per scenario. EPC descriptive
fields are sourced live from the EPC service (property_details_epc is dead);
property_type falls back override -> cert.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The `recommendation` table (~26m rows) has no index on `plan_id`, so any query
reaching it via `plan_id` — including the audit's own rollup — seq-scans the
whole table and saturates the shared DB (it blocked the DB during the
portfolio-796 audit). Make the audit safe-by-default:
- statement_timeout (120s) on the audit connection — a hard ceiling so a bad
plan aborts instead of hammering the DB.
- The recommendation rollup (the two solar checks) is now opt-in via
--with-recommendations, and EXPLAIN-gated: it refuses to run (raising
RecommendationScanError) when the plan contains a Seq Scan on recommendation,
which it does on any large portfolio until idx_recommendation_plan_id exists.
- SKILL.md documents the plan_id-no-index trap, the reach-via-property_id /
EXPLAIN-first / confirm-with-user rules, and the index as the real fix.
Verified on 796/1268: default run is bounded and completes (2,952 anomalies over
31,919 properties); --with-recommendations aborts pre-scan portfolio-wide but is
allowed for a single property.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
