diff --git a/packages/domain/src/domain/sap/worksheet/tests/test_fuel_cost.py b/packages/domain/src/domain/sap/worksheet/tests/test_fuel_cost.py
index e8b4c9a1..5c237a40 100644
--- a/packages/domain/src/domain/sap/worksheet/tests/test_fuel_cost.py
+++ b/packages/domain/src/domain/sap/worksheet/tests/test_fuel_cost.py
@@ -10,7 +10,6 @@ import pytest
 
 from domain.sap.rdsap.cert_to_inputs import cert_to_inputs
 from domain.sap.worksheet.fuel_cost import FuelCostResult, fuel_cost
-from domain.sap.worksheet.tests import _elmhurst_worksheet_000474 as _w000474
 
 
 def test_single_rate_main_only_bills_kwh_at_high_rate_price() -> None:
@@ -347,41 +346,8 @@ def test_total_cost_clamps_to_zero_when_pv_credit_exceeds_consumption() -> None:
     assert result.pv_credit_gbp < 0.0
 
 
-def test_000474_cert_to_inputs_fuel_cost_within_existing_e2e_tolerance() -> None:
-    """Cert-round-trip conformance: 000474 mid-terrace combi-gas (PDF
-    total fuel cost £655.69). Post-§10a actual lands at ~£726 (+10.7%
-    over PDF) because §4 HW kWh overestimates by +14% (2622 vs 2292) +
-    Appendix L lighting overestimates by ~3x (528 vs ~169 back-derived
-    from PDF). The pre-§10a £651.85 close-match was a coincidence —
-    wrong-prices-but-cancels-kWh; post-§10a is right-prices-but-
-    exposes-kWh-overshoot. See `project_section_4_hw_next_ticket`
-    memory — §4 HW worksheet tightening is the next ticket. Tolerance
-    mirrors the existing e2e 15% ceiling (test_e2e_elmhurst_sap_score)
-    until upstream §4/Appendix L slices land."""
-    # Arrange
-    epc = _w000474.build_epc()
-
-    # Act
-    inputs = cert_to_inputs(epc)
-
-    # Assert
-    assert inputs.fuel_cost.total_cost_gbp == pytest.approx(655.6949, rel=0.15)
-
-
-def test_000490_cert_to_inputs_fuel_cost_closes_to_within_5pct() -> None:
-    """Cert-round-trip conformance: 000490 mid-terrace combi-gas with PV
-    (PDF total fuel cost £807.54). Pre-§10a was £706.23 (-12.5%) —
-    handover blamed pre-amendment spec-version drift but the real cause
-    was wrong-table (Table 12 vs Table 32) + missing (251) standing
-    charges. Post-§10a actual lands at ~£776 (-3.9%); tightens further
-    when §4 HW closes. Marquee zero-error closure for this fixture."""
-    # Arrange
-    from domain.sap.worksheet.tests import _elmhurst_worksheet_000490 as _w000490
-
-    epc = _w000490.build_epc()
-
-    # Act
-    inputs = cert_to_inputs(epc)
-
-    # Assert
-    assert inputs.fuel_cost.total_cost_gbp == pytest.approx(807.5421, rel=0.05)
+# 000474 / 000490 fuel-cost conformance tests removed — superseded by
+# the strict `test_sap_result_pin[<fixture>-total_fuel_cost_gbp]` cases
+# in test_e2e_elmhurst_sap_score.py at abs=1e-4. The previous rel=0.15
+# (000474) and rel=0.05 (000490) tolerances permitted ~£70/£40 drift
+# from PDF — a fictional pass gate for a deterministic test vector.