How well do curtains reduce heat loss?

Curtains used in the heat loss tests for English Heritage

English Heritage reports that heavy curtains can reduce heat loss through a window by 40% [1]. A technical paper published by Historic Scotland, measured the reduction in heat loss at 14% [2]. That is a big difference - which figure is right (or at least closer to the truth)? Both reports were based on research at Glasgow Caledonian University, both list Dr. Paul Baker as author and judging by the pictures they were using the same curtains, though the windows were slightly different (both were wooden sash windows.) The reports do agree that secondary glazing saved 63% and they more or less agree that shutters were almost as good (58% or 51%).




English Heritage uses this data to present the case that you don't have to replace your traditional windows to get better thermal comfort - less damaging options such as secondary glazing, shutters or thermal curtains are also very effective. They did not include replacement double glazing in their tests - Historic Scotland did (using Slimlite® glazing which has a very narrow gap so as to fit into existing frames). They found the double glazing was not quite as good as the secondary glazing: 55% savings compared to 63%.

So how come the huge difference in savings from the curtains? The measurements were made in a laboratory under similar test conditions (with no wind so the draughtiness of the windows was not relevant). The Historic Scotland paper presented the heat loss reduction as measured by heat flux sensors on the glass. However, English Heritage only reports the U-values  that were calculated and the savings based on them. There is a big difference - the U-value calculations are based on some assumptions that are not necessarily accurate.

The U-value for a building element is the heat flow per unit area, per unit temperature difference. Other things being equal, if you halve the U value you halve the heat loss. You would have thought it was easy to measure the U-value for a building element but apparently not. For whatever reason, the approach taken in this case was to measure the heat flux and the surface temperatures on each side of the window system and add an adjustment for the surface thermal resistance. This was done in several places and the average taken. The same adjustment was made for all the options - though it seems unlikely to me that the thermal resistance of a curtain surface would be the same as a glass surface.

These two charts show some of the results from the experiments. The first shows that the results of the U-value calculations were quite similar for English Heritage (EH) and Historic Scotland (SH). This implies that the measurements in both cases were also broadly similar.

U-values calculated by Historic Scotland (HS) and English Heritage (EH)
(These are the U-values for the glass only, the whole window U-value is a weighted average of glass and frame. )

The second shows the heat loss savings calculated either from the heat flux measurements or using the U-values. The U-value calculation gives similar results for English Heritage and Historic Scotland but the heat flux savings (HS only) are always less - a little less for glazing and shutters and a lot less for the 'soft' cases of curtains and blinds.

Heat loss savings as measured by heat flux or calculated from the U-values.

Based on this I am very suspicious of the English Heritage figure for 40% savings for curtains. I think the Historic Scotland figures (14%) is much nearer the mark, though it is probably a little low.

From these two reports we can also conclude that Slimlite double glazing is almost as good as secondary glazing, so if your aesthetic sensibilities are different from English Heritage you might choose that rather than the secondary glazing. Also shutters can be almost as good but, as with the curtains and blinds, they only work when they are closed. That means the overall savings will be a bit less than the graph suggests.



[1] Chris Wood, Bill Bordass and Paul Baker (2009) Research into the thermal performance of traditional windows: timber sash windows (English Heritage)
[2] Paul Baker (2008, corrected 2010) Thermal performance of traditional windows (Historic Scotland).
[3] Paul Baker The thermal performance of traditional windows and practical measures to reduce heat loss and air leakage (This is a pdf of a presentation about the Historic Scotland survey. Pictures of the options tested start on slide 17 and there is a chart showing heat loss savings on slide 25).