Windows · Daylight

Glazing ratio and daylight

A window is a hole you cut in the insulation to let light in, and both halves of that sentence have a number. The light depends on the glass that is actually left after the frame has taken its share — usually only two thirds of the opening — on how clean it is, on how much sky it can see and on how pale the room is. The cost is the heat that leaves through it every night for the rest of the building's life. This page puts the two on the same screen, and tells you what one extra percentage point of daylight is going to cost you in watts.

The room

Length m

Width m

Height m

What the room is for

The targets are the long-standing BRE / BS 8206-2 minimum average daylight factors. They are a floor, not a goal: a room at exactly its target still needs the light on for a good part of a working day.

The opening

Width mm

Height mm

Window type

Vertical bars

Horizontal bars

Bar width mm

How many of these windows in the room

Glass and sky

Visible transmittance τ_v

Maintenance factor

Visible sky angle θ °

Surface reflectance R

θ is the angle of sky visible from the middle of the window, measured in section from the horizontal. Painting a room white instead of mid-grey is worth more daylight factor than most people expect, and it is the cheapest move on this page.

What it costs in heat

U_w window

U wall

ΔT K

The window replaces wall, so what it costs is the difference between the two, not the whole of it. Feed the result into the whole-house heat loss calculator to see it in context, or work out U_w properly with the window U-value tool.

—

Glazed area

— m²

Average daylight factor

— %

Glass to floor

— %

Heat cost

— W

The two sums

what is actually glass A_g = ( W − 2f − n_v·b ) · ( H − 2f − n_h·b ) // per window, then × how many glazing ratio = A_g / (W · H) average daylight factor — the BRE formula ADF = ( T · A_g · θ ) / ( A · ( 1 − R² ) ) T = τ_v · maintenance factor θ = angle of visible sky, degrees A = every internal surface: floor, ceiling, all four walls R = area-weighted average reflectance of those surfaces what the light costs Q = ( U_w − U_wall ) · A_window · ΔT // watts, design condition

Three things in that formula are worth staring at.

The frame is not a rounding error. An opening casement in a 1200 × 1400 hole gives 1.18 m² of glass out of a 1.68 m² opening — the frame has eaten 30 % of the daylight before anything else happens. Split the same window into six panes with glazing bars and you lose another chunk. This is why a fixed light next to a casement, same size, is visibly brighter.

The room is in the denominator, not the window. Daylight factor is not a property of the window; it is a property of the window and the room it opens into. Put the same window in a room twice as deep and the factor halves. That is also why painting the walls white is a real move and not a decorating one: going from R = 0.3 to R = 0.6 lifts the answer by about 40 %, for the price of paint.

The sky angle punishes cities. θ is 90° for an unobstructed window and maybe 45° with a building across the street — so the same window in the same room delivers half the daylight, and no amount of better glass gets it back.

What the number means

ADF	What the room feels like
< 1 %	Gloomy. Electric light on all day, whatever the weather outside.
1 – 2 %	Daylight is noticeable but not sufficient. Supplementary light most of the time.
2 – 5 %	Predominantly daylit. The lights stay off on a bright day.
> 5 %	Very well daylit — and now glare, summer overheating and winter heat loss are the problem instead.

Worked example you can check by hand

The living room this page opens with: 4.0 × 3.5 × 2.5 m, one opening casement in a 1200 × 1400 mm hole, double low-E glass.

glass (1200 − 210) × (1400 − 210) = 990 × 1190 = 1.178 m² opening 1.68 m² → glazing ratio 70.1 % room A = 2(4·3.5) + 2(4·2.5) + 2(3.5·2.5) = 28 + 20 + 17.5 = 65.5 m² 1 − R² = 1 − 0.25 = 0.75 ADF T = 0.70 × 0.90 = 0.63 (0.63 × 1.178 × 90) / (65.5 × 0.75) = 66.80 / 49.13 = 1.36 % heat (1.4 − 0.30) × 1.68 × 23 = 42.5 W 42.5 / 1.36 = 31 W per 1 % of daylight factor

So the room misses the 1.5 % living-room target — by not very much, but it misses it. To reach it the glass has to go from 1.178 to 1.300 m², which on the same 1200 mm width means a window 1523 mm tall instead of 1400. That extra 123 mm of height buys 0.14 % of daylight factor and costs about 4 W. Making the same window a fixed light instead of an opening one would get there for nothing at all — the frame is where the light went.

What this page does not do

Average, not distribution. One number for the whole room. It cannot tell you that the far corner is dark, and in a deep room that is the thing that actually matters. As a rule of thumb useful daylight reaches about 2 to 2.5 times the head height of the window into the room, and beyond that the average is a lie.
No orientation, no sun. Daylight factor is defined against an overcast sky on purpose, so it is deliberately blind to which way the window faces. Real rooms get direct sun, and a south window behaves nothing like a north one of the same size.
The heat cost is night-time and worst-case. It ignores the solar gain that pays part of it back through a heating season — which is real, and on a south facade can be most of it — and it ignores summer overheating, which makes large glazing much more expensive than this number suggests.
ADF is the old method. BS 8206-2 has been withdrawn and EN 17037 is the current standard: it asks for a target illuminance over a fraction of the space for a fraction of the daylight hours, assessed by climate-based modelling against real local sky data. The BRE average is still the right first sketch, and still what most people mean when they say "daylight factor", but it is not a compliance answer.
No blinds, curtains, deep reveals or balconies overhead. All of them cut the figure, sometimes by half.

Read this. Use this to size a window early, to compare options and to argue about a frame. Anything that has to be demonstrated — a planning condition, an EN 17037 claim, a rights-to-light matter — needs climate-based daylight modelling of the actual geometry.