Window U-Values Explained: What the Numbers Really Mean

One quote says 1.2. The next says 1.6 for what sounds like the same thing. A surveyor has told you your existing sashes are "about 4.8", and the council's conservation officer has mentioned 1.2 centre pane as though everybody knows what that means. None of these numbers are being used dishonestly. They are just different numbers measuring different things, and nobody has told you which is which.

Once you know which number is which, two quotes that looked contradictory turn out to be comparable after all. It also becomes clear why an original sash, however carefully it is upgraded, will never reach the figure printed on the front of a glazing brochure.

What a U-value is measuring

A U-value is a rate of heat flow: watts lost through one square metre of the element, for every degree of temperature difference between inside and outside. The units are W/m²K, sometimes written W/(m²·K). Lower is better. Double the U-value and you double the conducted heat loss through that square metre.

Two things follow from the definition, and they are the two things people get wrong.

First, it is a rate per square metre, not a total. A large window with a good U-value can lose more heat than a small window with a bad one. Second, it measures conduction, convection and radiation across the element under controlled conditions. It does not measure air leaking around it. A window can have an excellent U-value and still make the room feel cold, because the U-value takes no view at all on the gap between your meeting rails.

The three U-values, and which one is in your quote

SymbolWhat it coversTypical use
UgThe glass alone, measured in the middle of the pane, ignoring the edge spacer and the frameThe number in glass-unit datasheets and most advertising
UfThe frame section aloneRarely quoted to homeowners; used in calculations
UwThe whole window: glass, frame, and the edge effects where they meetThe number that Building Regulations care about

Ug is what the industry calls the centre-pane U-value, and the Approved Document defines it exactly that way: the U-value determined in the central area of the glazing unit, making no allowance for edge spacers or the window frame.

That exclusion is the whole story. The edge of a sealed unit is where the two panes are held apart by a spacer bar, and a spacer bar is a thermal bridge. Aluminium spacers conduct badly; warm-edge spacers in composite or stainless are better but still worse than the cavity they sit next to. Around that you have a timber frame, and around that the interface with the wall. Every one of those is worse than the middle of the glass.

So Uw is always a worse number than Ug, and by more than most people expect. A slim unit advertised at Ug 1.2 will typically land somewhere around 1.6 to 2.0 as a whole window once it is sitting in a period sash. That is not a failure of the unit. It is arithmetic. If a quote gives you one number and does not say Ug or Uw, assume Ug and ask.

The test window is not your window

This is the piece almost nobody explains, and it matters more for period sashes than for anything else.

Approved Document L allows a window's U-value to be calculated on the actual size and configuration, or on a standard window: 1.23 metres wide by 1.48 metres high. For a vertical sliding sash, the standard configuration is defined as a central horizontal divider with two opening lights. In other words, a plain one-over-one, at a fixed size, with a single meeting rail.

Now picture what is actually in a London terrace. A Georgian six-over-six with a dozen slender glazing bars. A Victorian two-over-two in a narrow bay. A margin-light Edwardian with a border of small panes around a large centre pane.

Every one of those has far more frame and far more sealed-unit perimeter per square metre of window than the test configuration does. Frame and edge are the weak parts. So the same glass, in a heavily barred sash, gives a genuinely worse whole-window U-value than the certificate on the glass suggests, and the difference is not trivial. A six-over-six can be 0.3 to 0.5 worse than a plain sash glazed with the identical unit. If your windows have glazing bars, the published figure is optimistic for you specifically.

Ask for the Uw calculated on your actual window sizes and configuration, in writing. Any supplier who genuinely does this work can produce it. One who cannot is quoting you a glass datasheet.

Window Energy Ratings, and what they reward

The other number you will be handed is a letter. Window Energy Ratings run from A++ down to E and combine three things: heat loss through the window, solar gain through it, and air leakage around the opening parts. Building control can accept a Window Energy Rating of Band B or better in place of a U-value figure for a replacement window.

A rating is useful because it includes air leakage, which a U-value ignores. It is misleading because it also includes solar gain, and it takes no view of which way your window faces. A window that scores well largely on solar gain earns that score on a south elevation on a sunny February afternoon, and not on a north-facing bedroom at three in the morning in January. For a period house where you are trying to make a cold room comfortable, the U-value is the more honest of the two figures.

What Part L actually requires

The current guidance in England is Approved Document L, Volume 1, 2021 edition incorporating the 2023 amendments. A 2026 edition has been published but does not take effect until March 2027, and it keeps the window standard unchanged.

For an existing dwelling, the limiting standards for new and replacement fabric elements are:

ElementStandard
Window1.4 W/m²K, or Window Energy Rating Band B minimum
Rooflight2.2 W/m²K
Door with more than 60% of its internal face glazed1.4 W/m²K, or Doorset Energy Rating Band C
Other doors1.4 W/m²K, or Doorset Energy Rating Band B

Two exceptions matter enormously to anyone with original sashes.

Repair is outside all of this. The Approved Document says a controlled fitting means the entire unit of a window including its frame, and states plainly that replacing the glazing, or replacing a window within its existing frame, is not providing a controlled fitting and does not need to meet the energy efficiency requirements. Splicing a rotten rail, re-cording, re-puttying, draught-proofing, even fitting slim double-glazed units into your existing sashes: none of that is caught by the 1.4 figure. That is a matter of what the guidance says, not a loophole. The wider picture of what does and does not trigger a notification is in our article on building regulations for windows.

Character gets an alternative route. Where a window cannot meet the standard because of the need to maintain the character of the building, the Approved Document gives two named alternatives: a centre-pane U-value no worse than 1.2 W/m²K, or single glazing supplemented with low-emissivity secondary glazing. That second option is the one conservation officers reach for, and it is why secondary glazing in listed buildings keeps being approved when replacement windows are refused.

Separately, work to a listed building, or a building in a conservation area, does not have to comply fully with the energy efficiency requirements where compliance would unacceptably alter the character or appearance of the building, though it must comply to the extent that is reasonably practicable. Your borough's conservation officer and building control body are the people who decide where that line sits on your house. Ask them before you commit to anything.

Real numbers for real period windows

Build-upIndicative whole-window U-value (W/m²K)
Plain 4mm float glass, centre pane onlyaround 5.4
Original single-glazed timber sash, whole windowaround 4.8
Single glazing plus plain secondary glazingaround 2.5
Single glazing plus low-emissivity secondary glazingunder 2.0
Slim double-glazed unit retrofitted into an existing sashroughly 1.6 to 2.1
Modern 24mm argon-filled unit in a new timber sashroughly 1.2 to 1.6
Vacuum glazing in an original sashUg as low as 0.7 claimed; whole window well above that

Treat all of these as indicative rather than as promises. Historic England's testing puts a typical single-glazed timber window at about 4.8, with the glass itself nearer 5.4 and the timber frame doing the work of dragging the average down. Their figures for secondary glazing are about 2.5 for a plain installation and better than 2.0 with low-emissivity glass, largely because the secondary unit decouples the primary frame and sashes from the room as well as adding a second pane. The constraints on retrofitting slim units into original rebates are covered in slimline double glazing for sash windows.

Why the number understates your problem, and overstates the fix

Two honest caveats before you spend money on the basis of a U-value.

Air leakage is not in it. The heat losses from a traditional window are predominantly through the gaps around it, not through the panes. A U-value is measured on a sealed specimen in a test rig with no wind. Your window has a shrunken parting bead, a 3mm gap at the meeting rails, and two open pulley mortices venting the room into a hollow weight box. Nothing in a U-value describes that, which is why sealing a rattly sash usually changes how a room feels more than any glazing upgrade of comparable cost. That is what draught-proofing is for, and it is the cheapest genuine improvement available on a period window.

Comfort is about surface temperature. What you feel standing near a window is a combination of the draught crossing your ankles and the radiant loss from your body to a cold pane. Halve a window's U-value and the inside face of the glass runs warmer; that is why better glazing reduces condensation on the pane and stops the near-window seat feeling like a fridge door. If you have water running down the inside of your glass every winter morning, the mechanism and the fix are in our article on condensation on the inside of windows.

Six questions that make a quote comparable

  1. Is that Ug or Uw? Write the answer on the quote.
  2. Was the Uw calculated on my actual window sizes and configuration, or on the standard 1.23m by 1.48m test window?
  3. What is the spacer bar, and is it warm-edge?
  4. What is the cavity width and what gas fills it?
  5. Does the price include making good the rebate, and if the rebate has to be deepened, is that in the number?
  6. Is this a replacement window, which has to meet 1.4 or Band B, or glazing into my existing sashes, which does not?

Any supplier who works on period buildings will answer all six without hesitating. If a question gets deflected, that is information too.

Where to start if the room is cold

Get a still, cold evening and a lit taper or a stick of incense. Hold it a hand's width from the meeting rails, the staff bead and the bottom rail in turn, and watch the smoke. If it streams, you have an air-leakage problem, and no U-value on any quote will fix it. Seal that first and reassess in a month.

If the smoke hangs still and the glass is simply cold to the back of your hand, then you have a conduction problem, and the glazing conversation is worth having. Even then, work out the cost per degree before you commit: secondary glazing usually delivers more improvement per pound on a period sash than replacing the glass does, and it leaves the original joinery untouched.

Our sash window repair service covers every London borough, and we will tell you honestly when a window needs sealing rather than glazing. If you want the whole range of options set out in order of cost, from a strip of brush pile to vacuum glass, start with our guide to insulating sash windows.

Need this doing? We handle draught-proofing across London — see our sash window repair service or request a free estimate.

Related reading