Why a fabric target exists
Before April 2014, new homes in England & Wales were assessed in Part L against a carbon target only (DER/TER). That skewed designs in odd ways — you could build with inefficient fabric and rescue the CO₂ number with renewables or low-carbon heating. Sensible engineering (and the energy hierarchy) says reduce heat losses first, then add renewables.
So Fabric Energy Efficiency (FEE) was added to Part L in the 2013 edition (effective 6 April 2014). In SAP it appears as DFEE/TFEE — Dwelling vs Target Fabric Energy Efficiency, measured in kWh/m² per year. The dwelling’s DFEE must be equal to or lower than the TFEE of the notional dwelling.
What drives it
Fabric energy efficiency is exactly what it sounds like: the heat-loss performance of the building fabric itself —
- u-values of walls, floors, roofs and openings
- air tightness
- thermal bridging at junctions
Renewables can’t help you here — that’s the point. If your DFEE is marginal, the fixes are fabric fixes, and calculated junction Ψ-values (try ΨMonkey) are often the cheapest one available, since they improve the number without changing anything on site.
How the TFEE target is actually set
Your target isn’t a fixed national number — it’s built from your dwelling. SAP takes a notional dwelling of identical size, shape and orientation, gives it a standard reference fabric specification, and calculates its fabric energy demand. Then, following the 2013 consultation, it adds a 15% flexibility allowance so designers aren’t pinned to the exact notional detail:
TFEE = notional dwelling fabric energy demand + 15%.
Because the notional building is the same shape as yours, the target automatically reflects how compact your design is. A blocky mid-terrace has far less exposed surface per m² of floor than a sprawling detached bungalow, so its TFEE lands lower — the more exposed fabric a form has, the more heat it can lose, and the target moves with it. This is why two dwellings on the same site can have very different TFEE figures, and why “just copy the plot next door’s spec” doesn’t always work.
A worked example
Say the notional version of your dwelling comes out at 40 kWh/m²/yr of fabric demand. Add the 15% allowance and your TFEE is about 46 kWh/m²/yr. Your job is to design fabric whose DFEE lands at or below that 46. Get your walls, roof, floor, glazing, air tightness and junctions modelled and your DFEE comes in at, say, 44 — you pass the fabric test with 2 kWh/m²/yr to spare. Come in at 49 and you’ve failed it, no matter how much PV or how efficient a heat pump you bolt on afterwards.
(The 40 and 46 here are illustrative — real figures depend on dwelling type, size and compactness, so always work from your own SAP model rather than a rule of thumb.)
How FEE fits with the other SAP targets
Fabric energy efficiency is one of several tests a new-build SAP has to pass at once, not the whole story. Under SAP 10 / Part L 2021 a new dwelling must satisfy:
- Primary energy — DPER must beat TPER (the headline metric in Part L 2021).
- Carbon — DER must beat TER.
- Fabric energy efficiency — DFEE must beat TFEE (this page).
- Limiting standards — individual u-values, air tightness and fixed building services must all sit within their backstops.
Miss any one of them and the dwelling fails. FEE is the one that can’t be bought back with kit — it’s a deliberate floor under the fabric, so a home stays efficient even as the grid and its heating system change around it.
DFEE looking marginal? Where to claw it back
Because FEE is fabric-only, the levers are all fabric levers — and some are far cheaper than others:
- Calculated junction Ψ-values. Usually the cheapest win. Poorly-detailed junctions can be 20–30% of fabric heat loss, and swapping punitive SAP defaults for calculated Ψ-values improves the number without changing a thing on site.
- Air tightness. Tightening the target the build can actually achieve (and prove on test) directly cuts fabric demand.
- U-values. Better walls, roof and — often overlooked — glazing and its frames. Marginal upgrades to insulation thickness or a better window spec move the DFEE.
- Reduce needless glazing / sort orientation. Oversized north-facing glazing loses heat with little solar gain in return; the fabric metric notices.
An assessor reads the DFEE/TFEE gap to show you exactly which of these is the cheapest route back to a pass — often it’s the junctions, not another 25mm of insulation everywhere.
What the Future Homes Standard changes
Worth knowing where this is heading. Under Approved Document L 2026 — the Future Homes Standard, published 24 March 2026 and in force 24 March 2027 — the separate DFEE/TFEE pass/fail looks set to be superseded by a whole-building approach built around carbon and primary energy (targeting roughly a 75% cut in emissions versus 2013), backed by tighter limiting fabric U-values and a new renewables requirement.
That doesn’t mean fabric stops mattering — the opposite. As walls, roofs and floors get tighter, thermal bridging becomes a bigger slice of what’s left, and fabric performance re-emerges as one of the four headline metrics on the reformed EPC. And until 24 March 2027, DFEE/TFEE still applies to your Part L 2021 submissions, so it’s very much a live test today.
The bottom line
Fabric energy efficiency is the test you can’t cheat with renewables — get the u-values right, tighten the air test, and nail the junctions. If a DFEE is coming up marginal, calculated Ψ-values are frequently the cheapest single fix: model yours free with ΨMonkey, then hand the numbers to your SAP assessor — no subscription, no sign-up.