Understanding the impact of thermal bridging



Heat loss through building fabric is made up of heat loss through plane elements (e.g. components like windows, walls, floors etc) and thermal bridging which may be thought of as the extra-over heat loss that occurs at junctions of those components.  Thermal Transmittance, or U-value (in W/m2K), is the measure of plane element heat loss. Linear Thermal Transmittance, or Psi-value (in W/mK), is the measure of thermal bridging.  Geometrical thermal bridges occur where the same component turns a corner and construction thermal bridges occur where, for example, a floor meets a window, wall meets roof, or even two portions of the same wall, each with different thermal characteristics, meet.


The Y-Factor (in W/m2K) is a way of representing the sum of all heat loss due to localised linear thermal bridging in a dwelling expressed in terms of the whole building envelope.  It is referred to in Appendix D of the Irish Technical Guidance Documnets (TGD)- L(2007) and is an input in DEAP building energy rating software.  Because it has the same units as U-value it may therefore also be directly compared.  When the averaged U-value for a dwelling is ascertained the Y-Factor may be directly added to get a sense of the full extent of building fabric heat loss.


This study shows how plane element heat loss and thermal bridging are closely related.  Perversely insulating the plane elements more can lead to a significant increase in heat loss at junctions and even colder temperatures at those surfaces if not carefully dealt with.  This is the ‘Achilles heal’ of so many energy-focused retrofits.  Figure 1 shows this change very clearly: in the isotherm diagrams warm colours indicate warmer temperatures, while in the thermal flux diagrams warm colours indicate faster levels of heat transfer.  The two graphs on the far right show how plane element heat loss (in yellow) can hugely reduce due to a particular retrofit without necessarily reducing the overall heat loss at that junction (yellow and orange).  This is poorly understood in the Construction Industry.  Even if the greater level of thermal bridging can be accepted at junctions, the increased risk of surface condensation (measured by the Temperature Factor, fRsi) may not be acceptable, due to the resulting potential for mould.


To further complicate matters TGD L(2007) & the DEAP manual  state that the ‘worst case’ Y-Factor for any works other than a new building designed to meet the Acceptable Construction Details is 0.15 W/m2K.  As we can prove that a poor retrofit can have levels of thermal bridging that are multiples of this figure, a good BER (e.g. C1) could seriously underestimate the actual building fabric heat loss.  The only answers for now are for the Government to remove this in accurate and misleading ‘worst case’ value and for professionals to calculate those thermal bridges and the resulting Y-Factors so as to see the real value of different design strategies.  In time, when the lessons been learnt through calculation and the level of understanding has grown, this may not be as necessary and codes based on long experience may make assessing this energy loss easier.



Wednesday, September 22, 2010