Monday, 6 June 2011

Thermal bridges and nekkyo

There seems to be some confusion over thermal bridges. The concept in English, or in fact German where it arose, is connected with multidimensional heatflow.  Insulation is not really rocket science. All materials conduct heat, the better at conducting, and the thinner the layers, the more heat is conducted. The worse at conducting and the thicker, the less heat is conducted. Everyone knows this, so we'll put on a thick down jacket in the winter rather than a thin cotton shirt, and put a pile of newspapers rather than some aluminium foil on the table underneath a hot baking tray. It's very easy to calculate heat flow in one dimension. If there is an infinitely large wall, a uniform thickness of a uniform material with one temperature on one side, and another temperature on the other, then heat is going to flow in a straight line, in proportion to the conductivity of the material and the temperature difference, and in inverse proportion to its thickness.

Unfortunately, buildings are not made up of infinitely large walls of uniform materials. Or we may consider this to be fortunate if we appreciate having our own buildings, and the luxury of doors, floors and roofs. When the wall hits a window, or hits a corner, of if the wall is made up of different materials and contains pillars and beams, then the calculations change. The walls we are using in our house are basically made from a wooden frame of 120x120 mm pillars and beams, with the gaps between filled with insulation, then one layer of 50 mm insulation on the inside, and 100 mm of insulation on the outside, held in place by plastic brackets. If we forget about the internal and external layer for a moment, and just consider the 120x120 mm beams and pillars, filled with insulation, then we can do a simple calculation of the insulation performance by assuming that around 10% of the wall is wood, which has a conductivity of about 0.17 W/mK, and 90% glass wool, with a conductivity of about 0.04 W/mK.  Average them out and you get about 0.05 W/mK.  

This is not really a thermal bridge in English, but is often referred to as Nekkyo (literally heat bridge) in Japanese. In one sense the wood is acting as a bridge for heat to get through the insulation. The same effect could happen with nails or various other heat conductors that find their way into buildings, innocently going about their business of stopping the buildings from falling down. The thermal bridge effect is usually considered a secondary effect to this. Once you start introducing different materials of non-uniform shapes and varying conductivities, heat does not simply travel in one direction. It will take the easiest path, and more heat is going to rush through the higher conductors, so the wall will lose more than the simple average suggests. As the amount of insulation goes up, the significance of this error becomes greater, and care must be taken within wall and roof structures and at the boundaries between walls, and between walls and floors, ceilings, doors or windows. Windows and doors should be designed and constructed so that the thermal bridge effects are as small as possible between glass and frame, between moving frame and stationary frame, and between the stationary frame and the wall. 

There's a great piece of software called Therm that can calculate all this. http://windows.lbl.gov/software/therm/6/index.html