C116 Outward-Open Window K Value: What 1.7 Measures, and What Your Opening Will Score

The C116 outward-open window carries a published K value of 1.7 W/(m²·K). That sentence is on the spec sheet, it is true, and it is not the number your building will be assessed against. It is a number measured on one specimen, built one way, glazed one way, sized one way, under one country's test conditions. Change any of those four things and you have a different number — sometimes by more than the difference between two competing series. This article is about the gap between the printed figure and the figure you can actually submit.

We have written elsewhere about how the thermal test itself is run and how to read a thermal test report. This piece starts where that one ends. It is not about the apparatus; it is about the object that was put into it.

The first thing to get out of the way: the opening direction does not change the requirement

A title that names "outward-open" invites the assumption that the direction is what makes the number good or bad. It is not. A UK installer's buyer guide says this plainly: replacement windows in England must reach a whole-window U value of 1.4 W/(m²·K) or a Window Energy Rating of band B under Approved Document L, whichever way they open. Direction is a decision about furniture, cleaning access, swing clearance and egress. It is not a thermal decision.

What direction does change is the shape of the specimen. A side-hung outward sash and a top-hung outward sash are different objects with different maximum sizes and different sealing geometry. One European manufacturer's published brochure — a Passivhaus-oriented window line — lists separate typical maximum sizes for outward opening: fixed light 2500 x 2500 mm, side hung 950 x 1520 mm, top hung 1300 x 1650 mm. Same system, same profile family, three different objects. The rating belongs to one of them.

So the useful question is never "is outward-open thermally better?" It is: which specimen produced this 1.7, and how far is my order from it?

K, U and U-factor: one quantity, three sets of test conditions

K value and U value are the same physical quantity — heat flow per unit area per unit temperature difference — reported under different rulebooks. The Chinese building energy system uses K, tested to GB/T 8484 (the standard for determining thermal insulating performance of external windows and doors). The European and international system uses U, with EN ISO 10077 for calculation and EN ISO 12567 for hot-box measurement. North America uses U-factor under NFRC procedures. The unit is identical: W/(m²·K). The number is not.

Two practical consequences follow. First, a number without its rulebook named next to it is not a number you can submit — it is a number you can discuss. Second, the unit conversion between the North American and metric worlds is the easy part and the part people get wrong in the wrong direction: an NFRC U-factor in Btu/(h·ft²·°F) is converted by multiplying by about 5.678. As one technical commentary puts it, the unit conversion is right and the boundary conditions are not — so a converted figure is still not the same measurement.

For a project outside China, the honest position is this: 1.7 is a series-level figure from the Chinese system. It is not an NFRC label value and it is not a certified EN whole-window value. We do not hold either, and we would rather say so than let a number travel somewhere it was not measured.

The reference window: the number belongs to a size, not to a series

Every rating system needs a fixed specimen, because a whole-window value is an area-weighted average and the weighting changes with size. The European reference size for a single-sash window is 1.23 m x 1.48 m. The Passivhaus component certification scheme anchors to the same window. The UK Passivhaus-conference brochure quoted above prints the caveat directly under its U values: "U values calculated for standard single window 1230 x 1480".

Here is the detail almost nobody tells a buyer. A UK glazing-cost guide explains that the standard figure must be calculated for that reference window in the standard configuration — and that for a casement, the standard configuration is a central vertical divider with one opening light and one fixed light. In other words, the conventional quoted casement number is measured on a two-panel window with a mullion through the middle of it. If you are ordering single-sash windows, your real number has no mullion and should be slightly better. If you are ordering a three-panel elevation with two mullions, it is worse. Either way, the quoted figure is not your figure.

This is no longer a technical nicety. Under the direction of travel described in UK guidance for Part L 2026 and the Future Homes Standard, the calculation has to be run on the actual window dimensions rather than a declared reference value, which the Glass and Glazing Federation is reported as describing as "almost a doubling of the work" for suppliers. One published summary states the consequence bluntly: a window that achieves 1.2 W/(m²·K) at 1200 mm x 1400 mm may fail at 800 mm x 900 mm. Under that approach, a single certificate no longer covers a product range — each opening size carries its own number.

Figures and regulatory thresholds quoted here come from public sources and move over time. Some UK sources cite a 1.4 W/(m²·K) replacement floor, others a 1.6 backstop, and devolved nations differ. Do not take a threshold from an article — including this one. Take it in writing from your building control body or energy consultant, and then ask us for a number computed to that threshold's method.

Small openings score worse, and that is the surprise in the arithmetic

Glass insulates better than the frame around it. So the more glass a window has as a share of its own area, the better its whole-window number. That single sentence produces a result most buyers find backwards: the same profile, the same glass and the same hardware give a worse number on a small window than on a large one. Finepoint Glass states it directly — as windows get larger there is proportionately more glass than frame, and big windows end up with improved overall U values; as size decreases, the higher proportion of framing lowers the whole-window figure.

Let us put numbers on it. The standard area-weighted formula is:

Uw = (Ug × Ag + Uf × Af + Ψg × lg) / (Ag + Af)

Ag is the glazing area, Af the frame area, lg the perimeter of the glazing (the spacer line), Ug the glazing value, Uf the frame value and Ψg the linear thermal transmittance of the spacer. Now take three openings made from the same C116 profile. To do the split we need one number the spec sheet does not give: the frame's projected face width. We assume 60 mm of visible frame band all round for this illustration — that is our assumption, not a published figure, and it matters, as the next section explains.

The illustrative column uses inputs taken from public ranges, not from our catalogue: Ug 1.1 for a generic low-E double unit, Uf 2.5 (public sources commonly place thermally broken aluminium frames in a 2.5 to 4.0 band), and Ψg 0.05 at the upper end of the warm-edge range. Do not read the absolute values as a product claim — read the spread. The same system moves about 0.28 W/(m²·K) between the largest and smallest opening purely because of size. That spread is bigger than most of the differences buyers argue about between competing series.

The design consequence is uncomfortable but useful: on a facade with one large window and three small ones, the small ones are your thermal problem, and they are the ones nobody models. Ask for a per-opening calculation, not a series figure.

None of the dimensions on the sheet is the one this arithmetic needs

Look at what C116 actually publishes: frame wall width 116 mm, glass sash thickness 81 mm, mullion 58/22 mm, profile 6060-T66 at 1.8 to 2.0 mm, insulation strips PA66GF25 across a 34 to 44 mm range, oblique press line giving a slim 28 mm frame, flush glass sash and screen at a 14 mm ultra-slim interior profile. Every one of those is a real and useful number. Not one of them is the frame's projected face width, which is the number that decides the Ag/Af split in the table above.

• 116 mm is how deep the frame sits in the wall — a wall-intersection dimension, not a face dimension.
• 81 mm is the depth of the glass sash stack — how much glass build-up the sash can carry, which is not the same measurement as the glazing rebate.
• 58/22 mm is the mullion: the first figure is its depth, the second its visible face. Sightline and section are different properties, and only one of them keeps the window stiff.
• 28 mm is the press line — the thing you see, and therefore the thing that gets quoted, and therefore not the thing that decides the heat flow.

This is why the first document to ask for is not a U-value and not a certificate: it is the section drawing with dimensions. From a section you can compute the area split yourself, or hand it to your energy consultant. From a single number you can do neither.

The mullion tax, and the second glass edge

A mullion is frame material running through the middle of your view. It costs you twice: it replaces glass with a worse-insulating material, and it creates a second perimeter of glass edge. Both terms appear in the formula — Af goes up, lg goes up — and neither appears in the headline number unless the specimen had a mullion in it.

The UK guidance on the new calculation methodology makes the same point with a starker example: in a Glass and Glazing Federation worked example, adding aluminium glazing bars is reported to push a window from 1.27 to 1.60 W/(m²·K). Decorative divisions are not decorative in an energy model. They are frame.

C116's mullion is published as 58/22 — a slim 22 mm visible face, which is a view feature, sitting on a 58 mm deep section, which is a stiffness feature. The slim face is what gets sold. The depth is what carries wind load and lets a wide elevation stay straight. Anyone selling you the 22 and staying quiet about the 58 is selling you a sightline. For thermal purposes, the one that matters is the projected width of the whole mullion in the plane of the window, and that comes from the section, not the brochure.

What sits inside the 81 mm: the glass, the spacer, and the number nobody prints

The glass sash on C116 takes an 81 mm build-up, which is generous by casement standards and is one of the reasons this series can be specified with a heavy insulating glass unit. But the sash depth is not the glazing rebate. The rebate is the pocket the unit actually sits in; it is a different measurement on a different part of the section. If you want to know the thickest unit the sash will take, ask for the rebate dimension on the section drawing, not for an inference from 81 mm.

Then there is the spacer — the strip that holds the panes apart at the edge. Its linear thermal transmittance is the Ψg term, and it is routinely ignored in sales conversations. A worked EN ISO 10077 example published by a composites manufacturer runs the reference window with Ug 0.70, Uf 1.0 and Ψg 0.04 and lands on Uw 0.88; it notes that published warm-edge values run roughly 0.03 to 0.05 W/(m·K) against roughly 0.08 for a conventional aluminium box spacer, and that spacer choice alone can move a whole-window value by about 0.1 — in that author's words, the difference between passing and failing a 0.80 specification.

One more reading from that same worked example is worth carrying: in it the frame occupies barely a quarter of the area, yet because its U value is higher it contributes almost as much heat loss as the much larger glass area. Swap that frame for a thermally broken aluminium one and the same glass loses its rating tier without anything about the glass changing. The glass flatters; the frame decides. This is the argument for asking about the frame section and the spacer before you negotiate over glass coatings.

The screen is almost certainly not in the number

C116 is specified with a flush glass sash and screen — a 14 mm ultra-slim interior profile — and the screen options include 80# black or grey stainless mesh and black high-definition anti-PM2.5 mesh. On an outward-opening casement the screen sits on the room side, which is also the side you operate it from.

Now the question that decides whether your number is real: was the declared value measured with the screen installed? Ratings are generally produced on the glazed sash assembly; screens are commonly treated as an accessory outside the rated construction. That is a reasonable convention — the screen is not a sealed unit and does not behave like one — but it means a closed screen creates an unsealed air layer that was not in the calculation. It also changes how the window behaves in daily use: a closed screen reduces air movement across the inner face of the glass, and in a cold climate that is exactly the condition under which internal surface condensation appears.

We are not going to claim a condensation performance figure we have not measured. What we will say is the ordering: condensation risk is driven by interior surface temperature, indoor humidity and airflow at that surface — heavy curtains and closed blinds lower the surface temperature by cutting the room's air from the glass, regardless of how good the K value is. If condensation is the concern on your project, the question to ask is for the temperature factor (fRsi) at your conditions, not for a lower K.

K is a state, not a property

An outward-opening casement achieves its thermal result the same way it achieves its weather result: the sash closes onto a rebate and the locking system compresses a continuous seal around the whole perimeter. That is why outward-opening windows can be specified to tight thermal targets at all, and why the same factory's sliding products cannot be quoted the same way — a sliding sash must keep a running clearance in order to roll, so it can only be contact-sealed, not compression-sealed.

It also means the number describes a condition the window has to be put into. A closed but unlocked casement does not have the K value. A sash whose eccentric locking cams have never been adjusted does not have the K value. A sash with a flattened or cracked gasket does not have it either. The published figure assumes the sealed state, and the sealed state is created by hardware, maintained by adjustment, and lost by wear.

C116 is built for that state: 34 to 44 mm multi-chamber PA66GF25 insulation strips, a unified isothermal line, triple sealing strips with seamless welded corners, and full glue-injection corner assembly. The C116 hardware is German Wehag as an upgrade option, with 304 stainless high-load hinges in a built-in hardware groove. Those are the components that decide whether year-one performance is still there in year five.

The honest gap: C116 publishes no air-leakage class, no water-tightness class and no wind-load class. A thermal number without a measured air-leakage number is a conduction number. Air moving through a joint carries energy that no conduction calculation captures. If your compliance route needs a leakage figure, you need to say so at enquiry stage, because it is a different test on a different specimen.

Why the casement carries a K and the sliding series do not

Our published K values sit on the casement and tilt-and-turn families: 1.7 for C116, 1.6 for C120, 1.3 for AW90. The sliding series publish none. There is a reason, and it is not that the sliding products are worse.

A sliding sash has to remain movable. That requires a designed clearance, and the clearance is a thermal and air path that a compression seal does not have. The frame section can be excellent and the whole-window number still governed by the interface. Publishing a single number for such a product would either flatten that reality or overstate it.

The side-press window in our range exists precisely to attack this problem with a mechanism rather than with more rubber: it moves the sash out of the sliding plane before it seals, which is how a sliding-format product gets near casement-level sealing. It still does not turn a sliding product into a casement, and we will not print a number that implies it does.

The buying rule: an unpublished number is not a worse number — it is an absent one, and absent means you must ask for a calculation on your own configuration instead of comparing two decimals. Comparing 1.7 against 1.6 across two series is already shaky, because the two figures were produced on two different specimens. Comparing anything against a blank is impossible.

Where the lowest number is the wrong target

A K value describes heat flow caused by a temperature difference. It says nothing about heat arriving as sunlight. On a cooling-dominated project, or on a heavily glazed elevation, the solar term can dominate the conductive one, and chasing the lowest K can make the building worse.

• Glazed enclosures. In a british sun room or any largely glazed room, the conductive term is often not what makes the space unusable in summer. Solar gain is. A specification built entirely around the lowest K can produce a room that is warm in winter and unbearable in July.
• Large openings. A big folding door assembly usually scores better per square metre than a small window, because the frame share is lower. It also loses more heat in total, because there is far more of it. Per-square-metre values rank products; they do not size a heating plant.
• Mixed facades. A steel glass entry door and a casement window do not share a reference size, a frame-to-leaf ratio or a rating method. Do not average them into one number for a submission. Average the areas and keep the products separate, or the whole calculation is unsupported.

The general rule: K is ranked per unit area, heat is lost in total watts, and comfort is decided by surface temperatures and air movement. Three different questions, three different numbers.

What we publish, and what you have to ask for

Ten lines to put in the enquiry so the answer is usable

The single most useful thing you can do is send a data block rather than a question. Here is the block, with the kind of answer that sounds fine and tells you nothing.

Frequently asked questions

Is 1.7 a good number?

It is a competitive figure for a thermally broken aluminium casement, and it is meaningless without a threshold and a method next to it. Public buying guides for temperate markets commonly quote replacement-window floors around 1.4 W/(m²·K) and new-build targets nearer 1.2, while cold-climate and ultra-low-energy routes go lower. Those are other rulebooks' numbers, not conversions of ours. Compare your project's written requirement against a whole-window value computed to that rulebook's method.

Can you give me the value for my exact window sizes?

Yes, if you send the sizes, the panel layout and the glass build-up. What we will not do is take 1.7 and present it as a whole-window certified rating for an opening nobody has calculated. The series figure is a starting point; the calculation is the deliverable.

Why does the same window have a different number in Europe and in China?

Because the test conditions are different. Chinese industry sources describe outdoor conditions around minus 20 degrees Celsius for the K test, against around 0 degrees for the EN/ISO calculation, and note that the warmer cavity under EN/ISO conditions often produces a value about 0.1 W/(m²·K) higher. Same product, different rulebook, different answer.

Do outward-opening windows insulate better than inward-opening ones?

Not because of the direction. Published buyer guidance in the UK states the thermal requirement applies whichever way the sash opens. What matters is that a casement — in either direction — closes onto a rebate and is compressed by its locking points, which is what lets it reach a tight number at all.

Will a lower K stop condensation?

It reduces the risk and does not eliminate it. Condensation forms when a surface is colder than the dew point of the air touching it, which depends on indoor humidity, room airflow and anything — curtains, blinds, a closed screen — that cuts room air off the glass. If condensation is the design driver, ask for the temperature factor at your conditions, not for a lower K.

Can you supply an NFRC label or CE documentation?

No. We can supply profiles, dimensioned sections, material and test evidence, and physical samples for you or your representative to have tested locally. Certification and importer registration belong to the party placing the product on your market, and we do not hold them for you.

The short version

• The 1.7 on the C116 sheet is a series-level figure from the Chinese system, measured or calculated on one specimen — not a whole-window certified rating for your opening.
• Opening direction does not change the thermal requirement. It changes which specimen the rating came from.
• Small windows score worse than large ones from the same profile, because the frame share rises. Our illustration put the spread at roughly 0.28 W/(m²·K) across three ordinary sizes.
• A mullion costs you twice: it adds frame area and a second glass perimeter. Decorative divisions are not decorative in an energy model.
• The spacer is the cheapest decimal in the window and the one nobody specifies. Warm edge against an aluminium box moves the whole-window value by about 0.1 in public worked examples.
• The screen is almost certainly not in the declared number, and a closed screen changes how the glass behaves in cold weather.
• K is a state, not a property. It belongs to a closed, locked and correctly adjusted window with intact gaskets.
• Send opening sizes in mm, the panel layout, the glass build-up and your written threshold. Ask for a value, not a decimal.

If you send us the opening schedule in millimetres, the panel layout and the glass build-up you are considering, we will tell you what we can calculate, what we have to test, and what we simply do not have. That answer is slower than a number, and it is the only one that survives a submission.