Here at Energy House Labs, we’ve been studying heat loss from a typical 1930s semi-detached home to measure the real impact of retrofit insulation, and our findings suggest that current estimates of energy performance may be underestimating how much heat is lost.
Working with colleagues at Leeds Beckett University, we’ve published our findings in the SCI journal Energy Science and Engineering, in a paper titled Whole‐House Thermal Bridge Modelling for Accurate Retrofit Energy Prediction. Our research shows that seemingly small weak spots in a home’s insulation create a route for heat to escape. These crucial gaps, known as thermal bridges, occur around walls, floors, roofs and windows, and they can make a significant difference to how much heat is lost, even after a home has been insulated.
We found that a carefully designed whole-house retrofit reduced the home’s overall heat transfer coefficient by as much as 63% using external wall insulation and 61% using internal wall insulation, compared with the original building.
However, we also found that the default figures commonly used to estimate heat loss are often not representative, particularly after homes have undergone piecemeal insulation retrofits. This matters because energy models are used to predict the effects of improvements to Britain’s housing stock, and they inform frameworks including Energy Performance Certificates and retrofit assessments.
Dr. Christopher Tsang, who led the study here at Energy House Labs, explained:
“Retrofitting insulation into homes has become increasingly important to help meet climate targets. However, as buildings become better insulated, thermal bridges can become a more significant proportion of the heat loss.”
“Rather than simply asking how much insulation should be installed, more attention needs to be paid to how insulation joins together across a whole building to fill in the thermal bridges.”
“As well as heat loss, poorly addressed thermal bridges can also produce colder internal surfaces, which in turn can increase the risk of condensation and mould. We found that whole-house approaches could potentially help mitigate surface-condensation risks as well as reducing overall heat loss.”
Our call to government and industry
We believe better data is needed on where heat escapes from different types of homes. Using more realistic figures could improve predictions of how well insulation and other retrofit work will actually perform. We’d also like to see assessments take better account of the benefits of improving the whole house together, rather than treating measures such as wall, roof and floor insulation separately.
Our study draws particular attention to the simplified assumptions used when assessing existing houses. Under the Reduced Data Standard Assessment Procedure (RdSAP), bespoke calculations of thermal bridge heat loss are not normally used. Instead, a single default value is applied across a property, irrespective of factors such as its age, type and insulation level.
We recognise that this simplification has practical advantages. Individually modelling every building junction would be costly and time-consuming, and the precise construction of an older home is frequently unknown. But we argue that relying on generic assumptions becomes a bigger problem when modelling the effects of fabric retrofits.
We’re not suggesting that every home in Britain needs bespoke computer modelling. Instead, we’re calling for a more comprehensive and representative database of thermal bridge values, covering common building types and PAS 2035-compliant retrofit scenarios, which could make energy-performance predictions far more reliable.
The research team
The project was carried out by Dr. Christopher Tsang from our team at Energy House Labs, School of Science, Engineering and Environment, University of Salford, alongside Felix Thomas, David Glew and Matthew Brooke-Peat of the Leeds Sustainability Institute, School of Built Environment, Engineering and Computing, Leeds Beckett University.
The work was funded by the UK Department for Energy Security and Net Zero.