What is a g-value?
A g-value is the fraction of the sun’s energy falling on a piece of glass that ends up as heat inside the room. It is written as a decimal between 0 and 1. Glass with a g-value of 0.6 passes 60 per cent of the arriving solar energy through; glass at 0.27 passes 27 per cent. It is also called the solar factor or, in North American datasheets, the solar heat gain coefficient, and those three terms mean the same thing. On a roof it is the single most useful number on the specification sheet, because overhead glass takes far more solar energy than a window in a wall does. The physics of how that energy arrives and what it does to a room is on solar gain, overheating and shading. This page is what the number is and how to read it.
The scale, and the numbers real glass lands on
A single sheet of clear 4mm float glass sits at about 0.85. Plain double glazing with no coating is around 0.75. Add a standard low-emissivity coating for thermal performance and it drops to roughly 0.55 to 0.62, because the coating that reflects long-wave heat back into the room also intercepts part of the incoming shortwave energy.
Solar control coated units land between 0.25 and 0.40, with the well-known products clustering around 0.27 to 0.35. Triple glazing with two coatings falls naturally to 0.45 to 0.50 without any solar control intent, simply because there is more material in the way. Nothing sensible for a house goes below about 0.20; the deeply tinted glass that does is used on commercial elevations and it turns a room grey.
Where the number comes from in a laboratory
Two separate things add up to the g-value. The first is energy that passes straight through the glass as light and near-infrared radiation without being absorbed, which is the larger part.
The second is less obvious. Glass and its coatings absorb some of the energy, which warms the panes themselves. That absorbed heat then leaves the glass in both directions, partly back to the outside air and partly inward into the room. The share that comes inward is called the secondary internal heat transfer factor, and it can be a quarter of the total for a strongly absorbing product. It is why a heavily tinted pane is a poor way of controlling gain: it absorbs the energy rather than turning it away, then re-radiates a good deal of it into the room anyway.
g-value against light transmittance, and why they differ
Light transmittance, written LT or TL, is the fraction of visible light that comes through. The g-value covers the whole solar spectrum, which is roughly 44 per cent visible light, 3 per cent ultraviolet and 53 per cent near-infrared, and that infrared portion is heat you cannot see.
The two numbers are related but not locked together, and the useful glass is the glass where they separate. A plain tinted pane cuts both by similar amounts, so the room goes dark and only moderately cooler. A good solar control coating reflects the near-infrared while letting most of the visible light through, so the room stays bright and takes far less heat. That difference is the whole point of the product.
Dividing one by the other, and what a good ratio looks like
Divide light transmittance by g-value and you get a ratio that describes how well a product separates the two. A number above 1.8 is a strong performer, and the best coated products reach 2.0 or a little over. Plain glass sits at about 1.0.
| Build-up | g-value | Light transmittance | Ratio |
|---|---|---|---|
| Clear double, uncoated | 0.75 | 0.80 | 1.07 |
| Double, standard low-emissivity coating | 0.58 | 0.79 | 1.36 |
| Double, solar control coating | 0.35 | 0.68 | 1.94 |
| Double, high-performance solar control | 0.27 | 0.55 | 2.04 |
| Grey body-tinted double | 0.45 | 0.42 | 0.93 |
| Triple, two low-emissivity coatings | 0.48 | 0.66 | 1.38 |
The last row but one is the one worth looking at. A tinted unit at a g-value of 0.45 is passing more heat than a coated unit at 0.27 while transmitting far less light. That is why tinted glass overhead is generally the wrong tool, and why the question do solar control rooflights make a room dark has a better answer than people expect.
What the figure means for a real pane on a real roof
Peak solar irradiance on a well-oriented Gloucestershire roof plane in midsummer is around 700 to 850 watts per square metre. Multiply by the area and by the g-value and you get the heat arriving in the room.
Two square metres of standard low-emissivity glass at a g-value of 0.58, taking 800 watts per square metre, delivers about 930 watts. The same two square metres at 0.27 delivers about 430. The difference, 500 watts, is a fan heater’s worth of load removed from a room where somebody is trying to work in the afternoon. Whether that matters in your room is the subject of will a rooflight make my room too hot.
The g-value of the glass against the g-value of the unit
As with U-values, two figures circulate. The glass figure describes the glazing alone. The whole-unit figure, sometimes written g-window, accounts for the frame, which passes no solar energy at all, so it is an area-weighted number and it always comes out lower.
On a narrow roof window between trusses, where the frame is a large share of the total, the whole-unit g-value can be 25 per cent below the glass figure. That is worth knowing because it means a small unit is naturally less of an overheating risk than its glass specification implies, and it means the two figures should never be compared against each other across different products.
Low is not automatically better, because winter exists
Everything above concerns July. Between October and March the same glass is collecting free heat, and a low g-value throws that away as well.
On a south-facing pitch at Gloucester, where the December sun sits at 14.7 degrees and shines down a shallow slope for the middle of the day, useful winter gain through a standard unit offsets a real share of the conduction loss through the same glass. Dropping from 0.58 to 0.27 removes more than half of that. On a north-facing slope there is no direct gain to lose, so a low g-value costs nothing and a high light transmittance matters more. On south and west facing slopes the choice is a genuine trade, and it gets made against how the room is used rather than against a rule.
The shading factor that sits alongside it
A g-value describes bare glass. Anything in front of it or behind it modifies the result, and that is expressed as a shading reduction factor multiplied against the g-value.
An external awning blind gives a combined factor of roughly 0.15 to 0.30, because it stops the energy before it crosses the glass. A mid-colour internal roller blind gives around 0.60 to 0.80: the energy is already inside the room by the time the fabric intercepts it, and the blind heats up and radiates. Multiply the two figures to get the effective performance, so a 0.58 unit with an external blind lands near 0.12, better than any glass on its own. The comparison in full is on external shading against internal blinds.
Which g-value to ask for, by aspect
South-facing pitched slopes across the interwar semis at Longlevens, Hucclecote and Charlton Kings: 0.35 to 0.45 is the usual landing point, keeping some winter gain. West and south-west facing loft conversions: 0.25 to 0.30, because the four o’clock sun arrives low, square-on and into a house that has been warming all day. East: 0.40 to 0.50, since the gain is over by eleven and the room has the rest of the day to shed it.
North-facing slopes, including a great many rear slopes in the Forest of Dean and the Stroud valleys: leave the g-value alone at 0.55 or above and specify for light instead. Near-horizontal flat rooflights on Severn Vale rear additions and behind Cheltenham parapets see every aspect across the day and generally want 0.30 to 0.35. The products that reach those numbers are described on what glass stops solar gain, and the aspect-by-aspect detail on specifying glazing by which way the slope faces.
Where it appears on paperwork
The g-value belongs on the order alongside the U-value, the light transmittance and the coating position, and it belongs in the survey notes with the aspect and pitch that produced it. It also appears in overheating assessments for new dwellings and in some Building Control submissions for large glazed areas.
Severn records the compass bearing of the slope, the pitch off a digital level, the glass area against the floor area, and what shades the roof at what time of day, and the g-value is chosen from that. The energy efficient rooflights hub sets out how it is balanced against the U-value, since the two numbers pull in opposite directions on any slope that faces the sun.
Next on this subject.
- Question What U-value should a rooflight have? Ask for a whole-unit figure of 1.2 to 1.4 W/m2K on a manufactured roof window, and 1.0 to 1.3 on…
- Question Will a rooflight make my room too hot? It might, and whether it does comes down to four things: which way the slope faces, how steeply it is…
- Question What causes condensation on a rooflight? Condensation on a rooflight is warm indoor air meeting cold glass. Three conditions have to line up: moisture in the…
- Question Are rooflights bad for insulation? Per square metre, yes. A current rooflight loses somewhere between eight and ten times as much heat as the insulated…