Single vs Double Glazing: Real Numbers for a Period Home

Nearly every double glazing quote for a period house arrives with two numbers on it. Something like 4.8 for the window you already own, 1.2 for the one being sold, and an implied heating bill in the gap. They are not measured the same way, and neither describes the route most of the heat in an old London house takes when it leaves the room.

The one sash window that has actually been measured

Almost everything said about single glazing comes out of a calculation. One exception was tested instead. In 2008 Historic Scotland's Technical Conservation Group commissioned Dr Paul Baker, at Glasgow Caledonian University's Centre for Research on Indoor Climate and Health, to put a traditional sash and case window, the Scottish cousin of the London box sash, through the ordinary upgrades one at a time.

The whole window went into the guarded hot box at the National Physical Laboratory first. It measured 4.5 W/m²K as received and 4.2 after draught-proofing, a gap narrower than the ±5.5 per cent measurement uncertainty, so the report settles on 4.4 for the assembly. One figure there is worth carrying: the glass was 55 per cent of the window's area and roughly 72 per cent of its heat loss.

The upgrades were then tested in an environmental chamber, with heat flux sensors stuck to the pane. Note what they are: centre-of-glazing values, as the report states plainly, not whole-window U-values, so they cannot be set against the 1.4 a replacement has to meet. What the table does honestly is rank the options against each other on one window, the comparison nobody selling glass will offer. Uncertainty on the U-values is 0.3 W/m²K.

Read the slim sealed unit against the four rows below it. The expensive intervention came out behind sealed secondary glazing, and behind a blind, shutters and curtains all closed on a cold night. None of that makes double glazing a bad buy. It does make the case that the cheap measures are no consolation prize, and that a house with its original panelled shutters still folded into the reveal already owns most of the performance available to it.

Why the 1.2 on the quote never turns up in your wall

Two U-values circulate and they get used interchangeably. Ug is the centre-pane value, the glass alone, measured in the middle of the unit with no allowance for the edge or the frame. Uw is the window as an assembly. Uw is always the worse figure and it is the one Building Control works to.

They diverge because of the spacer. In a sealed unit the perimeter bar bridges the cavity and conducts heat straight round the gap the cavity exists to create, so performance falls away in a band an inch or two wide all round every pane. In a picture window that band is a small fraction of the area. In a six-over-six Georgian sash, almost everything is edge, and there is a glazing bar between every light.

Historic Scotland's 2010 work on slim-profile units in traditional windows shows what that penalty is worth. Count the frame and the bars, and the whole-window figures came out at 1.9 to 3.4 W/m²K on Georgian-pattern windows and 1.4 to 3.0 on Victorian ones, where the same windows single-glazed sat at 5.2 and 5.1. In-situ centre-pane readings for those units ran between 1.0 and 2.8, so the glass alone was doing better than the window. Secondary glazing on the equivalent windows landed at 2.0 and 2.1, mid-range against the slim units and without a chisel near the sashes.

So the pattern of your sashes changes the answer before anything else does. A supplier's 1.2 is almost always Ug for a standard test unit rather than Uw for your opening, so ask which the quote means and get the answer in writing. Anchor the conversation to Uw, and read what a window U-value actually describes before the salesman arrives.

The 86 per cent that never appears on a quote

Go back to the hot box. Draught-proofing that window did nothing measurable to its U-value. It cut the air leakage by 86 per cent, and left the sash tighter than the 4,000mm² trickle vent that Scottish building standards recommend for a kitchen or bathroom.

Both are true at once, and together they explain why a U-value is a poor guide to how an old house feels. It describes conduction through the assembly in still air. It says nothing about warm air leaving through the gap between the meeting rails, past a shrunken parting bead, out through the pulley mortices and down the hollow weight boxes into the wall. On an unrestored sash that path is usually the larger loss, and it is not in the number at all.

Historic England puts about one fifth of a home's heating out through the windows, most of it through air gaps rather than the glass. Testing at the Building Research Establishment on draught-proofed traditional windows found air leakage cut by more than half and energy performance improved by more than a quarter, for a fraction of what glazing work costs. If one thing gets done before next winter, make it proper draught-proofing of the sashes, and it is the cheapest work we do.

Cold glass, and the two things it does to a room

Surface temperature is the number nobody quotes and the one you can genuinely feel. The bare pane in that chamber sat at 12°C on the room side; every option in the table lifted it to between 18 and 21.

Glass at 12°C radiates your body heat away, and you feel that as a draught even in still air. Sit within a metre of a large single-glazed sash on a January evening and the room feels colder than the thermostat claims. That is radiant asymmetry, and it is why people replace windows performing exactly as designed.

The same cold surface causes the other complaint. When the inner face of the glass drops below the dew point of the room air, moisture condenses on it, which in a bedroom with two people breathing and the door shut happens most winter nights. The water runs down onto the bottom rail, gathers in the corners of the glazing rebate, and starts the putty and the timber decaying from behind, where nobody looks until the paint blisters. The same run-off feeds the black spotting along the reveals, traced out in why condensation forms on the inside of windows.

Lifting that surface temperature stops it, by whichever route. One detail decides whether secondary glazing helps or hurts: the secondary unit should be the airtight one, with the original sash left slightly leaky, so moisture reaching the cavity dries outwards. Build it the other way round and the condensation moves to a surface you cannot wipe.

What the rules control, and what they leave alone

A good deal gets sold on this point. Approved Document L treats the entire unit including the frame as a controlled fitting, and its scope stops there.

The middle row surprises most people: slim sealed units retrofitted into your existing sashes have no U-value target at all. A firm telling you your repair needs FENSA registration is quoting you for a replacement.

Then there is the relaxation, which has a much bigger reputation than its scope. Take the two routes it opens first: where a window cannot reach 1.4 without compromising the character of the building, the approved document accepts glass with a centre-pane value no worse than 1.2 W/(m²K), or the original single glazing kept and backed with low-emissivity secondary glazing. Both are written to leave old windows in the wall, which is why conservation officers reach for the second one.

What sits in front of those routes is narrow, and no building qualifies simply by being old. Statutory listing, a designated conservation area or a scheduled monument gets you to the door, as does traditional vapour-permeable fabric that has to breathe. Opening it is not the owner's call: Building Control decides, on the conservation officer's advice, one building at a time. Building regulations for windows walks through the rest of the regime.

None of that is planning consent, which runs on its own track and runs first. If the property is listed, or sits under an Article 4 direction in a conservation area, settle that before you collect a single quote.

What the saving is worth against London prices

Energy Saving Trust figures for putting A-rated double glazing across an entirely single-glazed, gas-heated semi have generally landed between one and two hundred pounds a year, and they rise and fall with the price of gas. That is every window in the house, replaced.

Now set London prices beside it. Replacement timber sashes with heritage-detailed double glazing generally run from about £1,200 to £2,500 a window in softwood, more in hardwood or Accoya, before scaffolding, plaster and decoration. A four-bedroom terrace with ten or twelve openings is a five-figure job. Divide that by a saving in the low hundreds and the payback runs to several decades, longer than the sealed units are guaranteed for.

The other route is priced differently.

None of it is a quote until someone has stood in the room. Two identical windows on one street can differ by half if one needs a tower, or a borough licence to stand scaffolding on the pavement, and condition moves them again once the beads come off.

Where a sealed unit is the right purchase

We fit retrofit double glazing, and there are houses where it is plainly the sensible buy.

The clearest case is joinery that has finished. Rails and stiles decayed through their section with the tenons gone means new sashes either way, and if sashes are being made, have the rebates cut deep enough to carry a proper unit. The second is a reveal that will not take secondary glazing: measure from the face of the plaster to the front of the sash, because a deep window board or shutters that must stay operable can leave a secondary frame nowhere to sit. The third is habits, since shutters and curtains are the best value in the table above and return nothing in a house that never closes them. The fourth is timing: if the putty has failed all round and the cords are coming out anyway, glazing slim units in at the same visit costs far less than commissioning that work later.

Two things get missed at quoting stage. The glazing rebate in a Victorian sash is usually only 10 to 12mm deep and slim units run thicker, so it normally has to be machined out, taking timber off a section that was never generous. Ease off one staff bead and measure yours before accepting a price that assumes otherwise. The weights then need rebalancing, because a slim unit weighs roughly twice what 3mm float glass does, and a sash that will not hold its position is a worse window than you started with.

If traffic noise rather than heat is the real complaint, price secondary glazing first. Acoustic performance comes from the width of the air gap, not the U-value, and a 12mm cavity has very little of it. We weigh the two routes against each other, cost included, in secondary glazing against sealed units.

The order to spend it in

Air-tightness first, on the evidence above, and it is the cheapest line in either table. Then surface temperature, by whichever route the building allows: shutters and blinds if you have them, secondary glazing if the reveal takes it, slim units if the rebate does and the sashes are sound. Replacement last, decided on the condition of the joinery rather than a U-value, the only ground on which it reliably wins.

If the honest answer for your house is draughts, slack cords and cold glass, that is one job rather than three, and our repair work across the London boroughs is built around that sequence. If the sashes are genuinely past saving, buy the replacement on a whole-window figure and a set of section widths in millimetres, not on the number printed largest.