Sunday, 8 April 2012

What infra red images tell you

Infra red image of the front of a house
I have been reminded several times recently of the power of infra red camera surveys to identify places where you are losing heat from your house. (I recently interviewed Heat Seekers for a local radio science program and even more recently I read 'do the math' on keeping your feet warm illustrated by an unusual thermal image of cold vs. toasty feet). Infra red images are great fun and excellent at spotting defects like draughts, missing insulation and damp patches. However, they do not tell you how much heat you are losing from your house -- what they actually show is the temperature of the surface they are looking at. The two are not quite the same, though they are related.

The camera works by detecting infra red light emitted from an object and from this, based on the amount of light and in particular the wavelength or 'colour' of the light, the software in the camera works out the temperature of the surface. Hotter objects emit more and at shorter wavelengths.  The camera then shows you a false colour image with different temperatures represented by different colours.

In practice however, getting a reliable temperature from the camera is a challenge. Most cameras are not as accurately calibrated as you might expect. Also most surfaces are at least a little bit shiny so the infra red light that the camera sees includes reflections from the surroundings (including the person holding the camera). Fortunately brickwork and matt paintwork are pretty good in this respect, though windows are very prone to reflections (You won't see much though from the other side of the glass because infra red light doesn't go through windows very well).  In damp conditions there can be distortion by absorption from moisture in the air as well.

You may be wondering now, how you can possibly tell anything with a thermal camera image at all! However, even if the camera is not perfectly calibrated you can reliably tell whether one area is hotter than another, and approximately how much hotter it is, as long as the surfaces are similar and not particularly shiny.

Heat Seekers take pictures of houses from the street but in practice you usually get better survey images, with bigger differences in temperature, from the inside of the house. This is because the inside surface has a higher thermal resistance than the outside. You can think of heat flow as a bit like electricity - the current is the rate of heat flow, the voltage which drives it is the difference in temperature. If the surface didn't have any resistance then the temperature of the surface would be the same as the ambient temperature(1) outside, regardless of the rate of heat loss.  However, because there is a resistance at the surface you do get a temperature difference (c.f. voltage drop) across it. You get a bigger difference in temperature on the inside, between the wall surface and the room, because the inside wall surface resistance is (usually) larger. There is yet another complication, though, because the outside surface resistance varies depending on the weather.
Temperature profile across a solid wall with and without insulation. The insulated wall
is warmer on the inside surface and colder on the outside than the uninsulated wall.

The diagram on the right shows the temperature profile across two different walls, simplified by ignoring plaster and paintwork etc so the solid wall  has just one layer and the insulated wall has only two. I have used nominal values for external and internal surfaces - these are  0.04 (2) for the external surface (low resistance) and 0.13 for the internal surface (higher but still no great shakes). A solid wall will typically have an overall resistance of about 0.5. This means that the temperature difference between the inside surface and the room will be 0.13/0.5 = 0.26 times the temperature difference across the whole wall. If it were 20C inside and 0C outside the inside surface would theoretically be 14.8C, as illustrated in the diagram.  If you insulate the wall then the inside surface temperature will be closer to the room temperature - and also the heat loss is much reduced. This is because you have increased the overall resistance, and the surface resistance is a smaller proportion of the total. The diagram shows an insulated wall with overall resistance 3.5 and the internal surface is 19.8C - very close to the room temperature. Conversely, if you have a cold bridge, such as a metal lintel over a window, which has lower heat resistance, then the internal temperature surface will be lower. Draughts will also show up where they cool the surface and damp patches show cold, both because damp areas have lower thermal resistance and because the surface is cooled by the moisture evaporating into the air.

On the outside, the same principles apply but because it is cold outside the badly insulated areas with low resistance are warmer rather than colder, and air leaking out shows warm.

Hence, even though you can't actually directly measure your heat loss with an infra red camera you can often identify where you are losing more or less heat and this is very useful. If you want a survey this year you'd better hurry while the weather is cold and the difference in temperature inside and outside is enough to get a good image.

Finally, it is worth another look at the temperature profile of the insulated wall. I have put the insulation on the inside (as I have done in my house) but it is usually better, if you can, to put it on the outside. This is because with the insulation on the inside the wall remains cold even though your room is nice and warm. If the wind is driving rain onto your wall it will get wet and if the weather is also cold your wall take a long time to dry out. If the weather then gets very cold the damp in the wall can freeze which will damage the brick and mortar. I am not worried about my house though because here in Cambridge it is very rarely both cold and wet. You can also protect your wall from rain in other ways.

(1) Actually it is the radiative temperature that matters, not the air temperature, because air is a poor conductor and most heat loss is by radiation. There can be a big difference, as you will notice on a sunny day when the air is cool. You will feel the warmth radiated by the sun on your skin when you are in the sunshine, and it seems much colder in the shade.  If the weather is windy, though, then wind chill also affects the temperatures.

(2) The units of thermal resistance are m2K/W. Its easier to think of the U-value which is 1/Resistance. The U-value (W/m2K) is heat flow in watts per unit area per unit temperature difference.

1 comment:

  1. This comment has been removed by a blog administrator.