diff --git a/docs/sap-spec/PARITY_FINDINGS.md b/docs/sap-spec/PARITY_FINDINGS.md
index 2e24969a..6563288e 100644
--- a/docs/sap-spec/PARITY_FINDINGS.md
+++ b/docs/sap-spec/PARITY_FINDINGS.md
@@ -50,6 +50,16 @@ Worst 15 (residual = predicted − actual):
 5. **S-B-bungalow-investigation** — Bungalow residuals don't fit the flat-surfaces pattern (bungalows have full floor+roof). Hypothesis: thermal-bridging y-factor + storey-count interaction over-counts envelope. Probe specifically before deciding.
 6. **S-B-pump-fan-default** — We default to 130 kWh/yr; SAP 10.3 Table 4f says higher for systems with mechanical ventilation. Marginal but consistent.
 
+## RdSAP10 Table 32 Energy Cost Deflator drift (P5 finding, 2026-05-19)
+
+Surfaced while transcribing the SAP 10.2 worksheet into [test_bre_worked_examples.py](../../packages/domain/src/domain/sap/tests/test_bre_worked_examples.py) for P5. The calculator uses **`ENERGY_COST_DEFLATOR = 0.36`** (SAP 10.2 Table 12, item (256) on worksheet page 146 of [sap-10-2-full-specification-2025-03-14.pdf](sap-10-2-full-specification-2025-03-14.pdf)). The newer **RdSAP 10 specification Table 32** (page 95 of [rdsap-10-specification-2025-06-10.pdf](rdsap-10-specification-2025-06-10.pdf)) states the deflator is **0.42**, noting *"this table is equivalent to Table 12 in SAP10.2 specification"* — i.e., it supersedes the SAP 10.2 value for RdSAP assessments.
+
+**Why it matters.** The deflator scales the ECF numerator, which drives every SAP rating. Switching 0.36 → 0.42 changes every SAP rating numerically; in the linear regime the rating shifts by `−16.21 × ΔECF`, in the log regime by `−120.5 × Δ(log10 ECF)`. For a typical dwelling this is roughly **−2 to −4 SAP points** across the cohort.
+
+**What it doesn't explain.** The session-B residuals above are dominated by per-dwelling shape errors (mid-floor flats, bungalows) measured in tens of SAP points — a uniform 2-4 point shift won't move that needle. The deflator is a calibration call, not a model-shape fix.
+
+**Decision needed.** Whether the calculator targets SAP 10.2 ratings (keep 0.36, used for full SAP / new-build EPCs) or RdSAP 10 ratings (switch to 0.42, used for existing-dwelling EPCs derived from reduced data). ADR-0010 sets the target as SAP 10.3; the rdSAP10 publication is more recent than ADR-0010's reference points and may be the operative spec for the cohort we probe against. **Not changed in P5** — flagged for ADR-level resolution.
+
 ## How to reproduce
 
 ```bash
diff --git a/packages/domain/src/domain/sap/tests/test_bre_worked_examples.py b/packages/domain/src/domain/sap/tests/test_bre_worked_examples.py
new file mode 100644
index 00000000..a836baf9
--- /dev/null
+++ b/packages/domain/src/domain/sap/tests/test_bre_worked_examples.py
@@ -0,0 +1,339 @@
+"""SAP 10.2 worksheet trace — comprehensive `intermediate` lock for a
+synthetic baseline dwelling.
+
+**Provenance.** The SAP 10.2 worksheet template (pages 131–148 of
+docs/sap-spec/sap-10-2-full-specification-2025-03-14.pdf) is the canonical
+calculation form every SAP implementation must mirror, using the item
+reference numbers (1a), (4), (33), (39), (40), (91), (92), (93), (257),
+(272), (286) etc. The PDF's form fields are non-functional, so this test
+is **spec-formula-derived** — each expected value is computed independently
+from the worksheet formulas applied to the baseline inputs below, not from
+BRE-published worked-example tables. BRE worked-example values were not
+located in any of the three SAP-spec PDFs in docs/sap-spec/; if they
+surface later, only the expected numbers need updating, not this file's
+structure.
+
+**Scope.** Locks every key currently published on `SapResult.intermediate`
+to the matching worksheet item, plus the top-level rating/emission/PE
+fields that mirror items (257)/(272)/(286)/(287). Per-month decomposition
+((38)_m, (66)_m, (74)_m–(82)_m, etc.) lives on `result.monthly` and is
+checked elsewhere; this file is the annual/aggregate trace.
+"""
+from __future__ import annotations
+
+import pytest
+
+from domain.sap.calculator import (
+    CalculatorInputs,
+    WindowInput,
+    calculate_sap_from_inputs,
+)
+from domain.sap.worksheet.dimensions import Dimensions
+from domain.sap.worksheet.heat_transmission import HeatTransmission
+from domain.sap.worksheet.solar_gains import Orientation
+
+
+def _baseline_dwelling() -> CalculatorInputs:
+    """Synthetic 100 m² semi-detached gas-boiler dwelling, UK-average
+    climate. Mirrors the baseline used in `tests/test_calculator.py` so
+    both trace suites exercise the same canonical fixture. All values
+    chosen to land near a real RdSAP cert (HLC ~150 W/K, τ ~100 h,
+    SAP ~70)."""
+    dim = Dimensions(
+        total_floor_area_m2=100.0,
+        volume_m3=250.0,
+        storey_count=2,
+        avg_storey_height_m=2.5,
+        ground_floor_area_m2=50.0,
+        ground_floor_perimeter_m=30.0,
+        top_floor_area_m2=50.0,
+        gross_wall_area_m2=150.0,
+        party_wall_area_m2=50.0,
+    )
+    ht = HeatTransmission(
+        walls_w_per_k=60.0,
+        roof_w_per_k=20.0,
+        floor_w_per_k=20.0,
+        party_walls_w_per_k=0.0,
+        windows_w_per_k=25.0,
+        doors_w_per_k=5.0,
+        thermal_bridging_w_per_k=20.0,
+        total_w_per_k=150.0,
+    )
+    windows = (
+        WindowInput(
+            area_m2=4.0,
+            orientation=Orientation.S,
+            pitch_deg=90.0,
+            g_perpendicular=0.63,
+            frame_factor=0.7,
+            overshading_factor=0.77,
+        ),
+        WindowInput(
+            area_m2=4.0,
+            orientation=Orientation.N,
+            pitch_deg=90.0,
+            g_perpendicular=0.63,
+            frame_factor=0.7,
+            overshading_factor=0.77,
+        ),
+    )
+    return CalculatorInputs(
+        dimensions=dim,
+        heat_transmission=ht,
+        infiltration_ach=0.7,
+        region=0,
+        windows=windows,
+        control_type=2,
+        responsiveness=1.0,
+        living_area_fraction=0.30,
+        control_temperature_adjustment_c=0.0,
+        thermal_mass_parameter_kj_per_m2_k=250.0,
+        main_heating_efficiency=0.85,
+        hot_water_kwh_per_yr=2400.0,
+        pumps_fans_kwh_per_yr=100.0,
+        lighting_kwh_per_yr=600.0,
+        space_heating_fuel_cost_gbp_per_kwh=0.07,
+        hot_water_fuel_cost_gbp_per_kwh=0.07,
+        other_fuel_cost_gbp_per_kwh=0.07,
+        co2_factor_kg_per_kwh=0.21,
+    )
+
+
+def test_baseline_dwelling_worksheet_trace() -> None:
+    # Arrange — see module docstring for provenance. The baseline is a
+    # 100 m² gas-boiler dwelling matching tests/test_calculator.py.
+    inputs = _baseline_dwelling()
+    ht = inputs.heat_transmission
+    tfa = inputs.dimensions.total_floor_area_m2
+
+    # Act
+    result = calculate_sap_from_inputs(inputs)
+    inter = result.intermediate
+
+    # Assert — §1 Dimensions ----------------------------------------------
+    # (4) Total floor area = Σ(1a..1n)
+    assert inter["tfa_m2"] == pytest.approx(100.0, rel=1e-12)
+    # (5) Dwelling volume = Σ(3a..3n)
+    assert inter["volume_m3"] == pytest.approx(250.0, rel=1e-12)
+    # (9) Number of storeys n_s — exposed as a float for trace uniformity
+    assert inter["storey_count"] == pytest.approx(2.0, rel=1e-12)
+
+    # Assert — §3 Heat transmission (A×U per element) ---------------------
+    # (29a) external wall A×U
+    assert inter["walls_w_per_k"] == pytest.approx(60.0, rel=1e-12)
+    # (30) roof A×U
+    assert inter["roof_w_per_k"] == pytest.approx(20.0, rel=1e-12)
+    # (28a)/(28b) floor A×U (ground/exposed combined here)
+    assert inter["floor_w_per_k"] == pytest.approx(20.0, rel=1e-12)
+    # (32) party wall A×U
+    assert inter["party_walls_w_per_k"] == pytest.approx(0.0, rel=1e-12)
+    # (27) window A×U (effective U per §3.2)
+    assert inter["windows_w_per_k"] == pytest.approx(25.0, rel=1e-12)
+    # (26)+(26a) door A×U
+    assert inter["doors_w_per_k"] == pytest.approx(5.0, rel=1e-12)
+    # (36) linear thermal bridges Σ(L×Ψ); calculator uses the y-factor
+    # shortcut (36) = 0.20 × (31) per the spec's default-bridging clause
+    assert inter["thermal_bridging_w_per_k"] == pytest.approx(20.0, rel=1e-12)
+
+    # Assert — §2 Ventilation ---------------------------------------------
+    # (16) infiltration rate (sheltered) — input air-changes-per-hour
+    assert inter["infiltration_ach"] == pytest.approx(0.7, rel=1e-12)
+    # (38)/m equivalent in W/K: 0.33 × ach × volume — annual-equivalent
+    # value (calculator carries the same scalar across months when wind
+    # adjustment is disabled in trace mode)
+    expected_infiltration_w_per_k = 0.33 * 0.7 * 250.0
+    assert inter["infiltration_w_per_k"] == pytest.approx(
+        expected_infiltration_w_per_k, rel=1e-9
+    )
+
+    # Assert — §3 aggregates ----------------------------------------------
+    # (37) total fabric heat loss = Σ(26..32) + (36) + (36a). With the
+    # baseline's HT.total_w_per_k already summing element-level A×U,
+    # (37) equals that sum.
+    fabric_plus_bridges_w_per_k = ht.total_w_per_k  # (33) + (36) effectively
+    # (39) Heat transfer coefficient HTC = (37) + (38)/m
+    expected_htc = fabric_plus_bridges_w_per_k + expected_infiltration_w_per_k
+    assert inter["heat_transfer_coefficient_w_per_k"] == pytest.approx(
+        expected_htc, rel=1e-9
+    )
+    # (40) Heat loss parameter HLP = (39) / (4)
+    expected_hlp = expected_htc / tfa
+    assert inter["heat_loss_parameter_w_per_m2k"] == pytest.approx(
+        expected_hlp, rel=1e-9
+    )
+    # τ time constant — Table 9b: τ_hours = TMP × TFA / (3.6 × HTC).
+    # The 3.6 converts kJ → W·h (TMP is kJ/m²·K, HTC is W/K).
+    expected_tau_h = (
+        inputs.thermal_mass_parameter_kj_per_m2_k * tfa / (3.6 * expected_htc)
+    )
+    assert inter["time_constant_h"] == pytest.approx(expected_tau_h, rel=1e-9)
+
+    # Assert — §5 internal gains + §7 mean internal temp (annual averages)
+    # (73) total internal gains; (93) adjusted mean internal temp.
+    # Both are positive for a heated dwelling at UK-average climate; exact
+    # values depend on Table 5 monthly profiles, so locked by sanity bound
+    # not a closed-form expression.
+    assert inter["internal_gains_annual_avg_w"] > 0
+    assert 17.0 < inter["mean_internal_temp_annual_avg_c"] < 21.0
+
+    # Assert — §8 Space heating -------------------------------------------
+    # (98c) total space heating requirement kWh/yr — exposed as the
+    # top-level SapResult.space_heating_kwh_per_yr too. Positive for a
+    # heated dwelling; the test_calculator suite already locks the chain
+    # via direction tests (zero-HTC, doubled-efficiency, colder-region).
+    assert inter["useful_space_heating_kwh_per_yr"] == pytest.approx(
+        result.space_heating_kwh_per_yr, rel=1e-12
+    )
+    assert inter["useful_space_heating_kwh_per_yr"] > 0
+
+    # Assert — §10/12 Fuel costs per end-use ------------------------------
+    # (240e), (242e), (247), (249-split), (250) — per-end-use costs in £/yr.
+    expected_main_cost = (
+        result.main_heating_fuel_kwh_per_yr
+        * inputs.space_heating_fuel_cost_gbp_per_kwh
+    )
+    assert inter["main_heating_cost_gbp"] == pytest.approx(
+        expected_main_cost, rel=1e-9
+    )
+    # (242e) secondary-heating cost — zero for baseline (no secondary)
+    assert inter["secondary_heating_cost_gbp"] == pytest.approx(0.0, abs=1e-9)
+    # (247) water heating cost
+    expected_hw_cost = (
+        inputs.hot_water_kwh_per_yr * inputs.hot_water_fuel_cost_gbp_per_kwh
+    )
+    assert inter["hot_water_cost_gbp"] == pytest.approx(
+        expected_hw_cost, rel=1e-9
+    )
+    # (249) split — pumps/fans at "other" tariff
+    expected_pf_cost = (
+        inputs.pumps_fans_kwh_per_yr * inputs.other_fuel_cost_gbp_per_kwh
+    )
+    assert inter["pumps_fans_cost_gbp"] == pytest.approx(
+        expected_pf_cost, rel=1e-9
+    )
+    # (250) lighting cost at "other" tariff
+    expected_light_cost = (
+        inputs.lighting_kwh_per_yr * inputs.other_fuel_cost_gbp_per_kwh
+    )
+    assert inter["lighting_cost_gbp"] == pytest.approx(
+        expected_light_cost, rel=1e-9
+    )
+    # (252) PV export credit — zero for baseline (no PV)
+    assert inter["pv_export_credit_gbp"] == pytest.approx(0.0, abs=1e-9)
+
+    # Assert — §13 Energy cost rating -------------------------------------
+    # (256) energy cost deflator (Table 12 = 0.36; see PARITY_FINDINGS for
+    # RdSAP10 Table 32 update to 0.42 — separate spec drift, tracked there).
+    assert inter["deflator"] == pytest.approx(0.36, rel=1e-12)
+    # (257) ECF = [(255) × (256)] / [(4) + 45.0]
+    floor_area_offset_m2 = 45.0  # baked into (257) formula
+    expected_ecf = (
+        result.total_fuel_cost_gbp * 0.36 / (tfa + floor_area_offset_m2)
+    )
+    assert inter["ecf"] == pytest.approx(expected_ecf, rel=1e-9)
+    assert inter["ecf"] == pytest.approx(result.ecf, rel=1e-12)
+    # The 45 m² offset and ECF=3.5 log/linear regime boundary — spec
+    # constants from §13 worksheet (257) and equations (8)/(9)
+    assert inter["floor_area_offset_m2"] == pytest.approx(45.0, rel=1e-12)
+    assert inter["ecf_log_threshold"] == pytest.approx(3.5, rel=1e-12)
+
+    # Assert — §14 CO2 emissions (per end-use; aggregate is (272)) --------
+    # (261) main heating CO2 = (211) × emission factor
+    assert inter["co2_factor_kg_per_kwh"] == pytest.approx(
+        inputs.co2_factor_kg_per_kwh, rel=1e-12
+    )
+    assert inter["main_heating_co2_kg_per_yr"] == pytest.approx(
+        result.main_heating_fuel_kwh_per_yr * inputs.co2_factor_kg_per_kwh,
+        rel=1e-9,
+    )
+    # (263) secondary CO2; (264) hot water; (267) pumps/fans; (268) lighting
+    assert inter["secondary_heating_co2_kg_per_yr"] == pytest.approx(
+        result.secondary_heating_fuel_kwh_per_yr
+        * inputs.co2_factor_kg_per_kwh,
+        rel=1e-9,
+    )
+    assert inter["hot_water_co2_kg_per_yr"] == pytest.approx(
+        inputs.hot_water_kwh_per_yr * inputs.co2_factor_kg_per_kwh, rel=1e-9
+    )
+    assert inter["pumps_fans_co2_kg_per_yr"] == pytest.approx(
+        inputs.pumps_fans_kwh_per_yr * inputs.co2_factor_kg_per_kwh, rel=1e-9
+    )
+    assert inter["lighting_co2_kg_per_yr"] == pytest.approx(
+        inputs.lighting_kwh_per_yr * inputs.co2_factor_kg_per_kwh, rel=1e-9
+    )
+    # (272) total CO2 — sum of the five end-use components reconciles
+    # with the top-level co2_kg_per_yr field.
+    co2_sum = (
+        inter["main_heating_co2_kg_per_yr"]
+        + inter["secondary_heating_co2_kg_per_yr"]
+        + inter["hot_water_co2_kg_per_yr"]
+        + inter["pumps_fans_co2_kg_per_yr"]
+        + inter["lighting_co2_kg_per_yr"]
+    )
+    assert co2_sum == pytest.approx(result.co2_kg_per_yr, rel=1e-9)
+    # (238) delivered fuel kWh/yr — the input to (272)/(286) chains
+    expected_delivered = (
+        result.main_heating_fuel_kwh_per_yr
+        + result.secondary_heating_fuel_kwh_per_yr
+        + result.hot_water_kwh_per_yr
+        + result.pumps_fans_kwh_per_yr
+        + result.lighting_kwh_per_yr
+    )
+    assert inter["delivered_fuel_kwh_per_yr"] == pytest.approx(
+        expected_delivered, rel=1e-9
+    )
+
+    # Assert — §14 Primary energy per end-use, per m² ---------------------
+    # (275)+(276) space heating PE / (4); (278) hot water PE / (4);
+    # (281)+(282) other PE / (4); (283)/(4) PV PE offset.
+    expected_sh_pe = (
+        (result.main_heating_fuel_kwh_per_yr
+         + result.secondary_heating_fuel_kwh_per_yr)
+        * inputs.space_heating_primary_factor
+        / tfa
+    )
+    expected_hw_pe = (
+        inputs.hot_water_kwh_per_yr * inputs.hot_water_primary_factor / tfa
+    )
+    expected_other_pe = (
+        (inputs.pumps_fans_kwh_per_yr + inputs.lighting_kwh_per_yr)
+        * inputs.other_primary_factor
+        / tfa
+    )
+    expected_pv_pe_offset = (
+        inputs.pv_generation_kwh_per_yr * inputs.other_primary_factor / tfa
+    )
+    assert inter["space_heating_pe_kwh_per_m2"] == pytest.approx(
+        expected_sh_pe, rel=1e-9
+    )
+    assert inter["hot_water_pe_kwh_per_m2"] == pytest.approx(
+        expected_hw_pe, rel=1e-9
+    )
+    assert inter["other_pe_kwh_per_m2"] == pytest.approx(
+        expected_other_pe, rel=1e-9
+    )
+    assert inter["pv_pe_offset_kwh_per_m2"] == pytest.approx(
+        expected_pv_pe_offset, rel=1e-9
+    )
+    # (287) dwelling PE rate = (286) / (4). Reconciled via the per-m²
+    # components, floored at 0 (Appendix M PV offset cannot drive PE
+    # negative).
+    expected_total_pe_per_m2 = max(
+        0.0,
+        expected_sh_pe
+        + expected_hw_pe
+        + expected_other_pe
+        - expected_pv_pe_offset,
+    )
+    assert result.primary_energy_kwh_per_m2 == pytest.approx(
+        expected_total_pe_per_m2, rel=1e-9
+    )
+
+    # Assert — top-level rating sanity ------------------------------------
+    # (258) SAP rating — integer-clamped via §13 equations (8)/(9). Tests
+    # in test_calculator already lock the curve direction (higher
+    # efficiency, colder region, doubled HTC). Here we just sanity-check
+    # the integer/continuous pair stays consistent for the baseline.
+    assert 1 <= result.sap_score <= 100
+    assert abs(round(result.sap_score_continuous) - result.sap_score) <= 1