Traditional stone walls in Scotland - validation of RdSAP U-value calculation methodology: research

4. Conclusions and recommendations

4.1 Conclusions

U-values are inferred in RdSAP calculations as there is generally insufficient information available on wall constructions for existing buildings to accurately predict U-values. The approach taken for stone walls in RdSAP is based on:

• measured wall thickness;
• assumptions on the conductivity of stone (for a very limited number of stone types) and an assumed ratio of stone to mortar in the wall.

This approach has the effect of presenting an ‘average’ performance around which the actual performance of walls is expected to vary. In addition, the thermal performance of a wall will also often vary to some extent due to seasonal variation of moisture content, which can influence the thermal conductivity of the stone (or other components in the wall) (See Appendix B).

Most of the measured U-value results from this study are lower than the respective RdSAP 10 calculated values. This suggests that the updated RdSAP 10 equations (which generally result in lower stone wall U-values than the RdSAP 2012 calculations) are somewhat more appropriate than those in RdSAP 2012, since they more closely align with the measured values from this study, despite not matching them exactly.

It was not possible to draw any other trends for the sandstone or granite and whinstone walls respectively, as the spread of measured results was too varied.

Impact on overall property energy assessment

The EPC energy assessment outputs from the property scenarios based on the measured U-values are closer to the outputs based on the RdSAP 10 U-values than to those that assume the existing RdSAP 2012 default approach. This again suggests that the updated RdSAP 10 stone wall equations are generally more appropriate than those in RdSAP 2012.

In general, the changes to the energy performance outputs due to the different U-value calculation approaches were modest. However, a few exceptions showed more notable variation in the space heating demand in excess of 10%, and EPC score changes of -4 to +6.

4.2 Recommendations

As with other studies of this nature, an obvious recommendation to assist in improving the confidence in stone wall U-value assumptions would be to obtain more measured U-value data from a sufficiently large sample of properties deemed representative of Scotland’s building stock. However, this would ideally need to be supported with more intrusive surveys of the test walls (e.g. to confirm the exact construction make up including any linings, materials present, thermal conductivity, density and moisture content of the stone, etc). This would provide a better understanding of any variability that may be witnessed, so these aspects could be factored into any new approaches for estimating the U-value of stone walls. This may however present challenges on a sufficiently large scale given the potentially wide variability of stone wall constructions across Scotland, e.g. ensuring a robust sampling method, project cost, engaging households (as property owners may not be willing for their walls/properties to undergo such physical testing), etc.