This is the first interim report for project RP1037u1, an extension to the recently completed project RP1037 ‘Driving increased utilisation of cool roofs on large footprint buildings’. Progress so-far in the project and preliminary findings have been summarised in this report.
The research has been focused on two key aspects of roof thermal performance that had, up until the time of writing, not been taken into account in most investigations into cool roof technology:
1. The condensation and evaporation of dew on the roof surface, and the effect this has on roof temperature by way of:
- The latent heat that is absorbed and released; and
- Any change in the effective radiative-optical properties of the roof top surface due to accumulated water.
2. The effect of roof temperature on above-roof air temperatures, and the influence this can have on the performance of rooftop heating, ventilation and air-conditioning (HVAC) equipment.
Research literature related to water condensation on roofs has been reviewed. The review demonstrated that the heat fluxes caused by the latent heat release and absorption and the changes in the radiative-optimal properties are the two key effects of dew on the roof thermal performance. The latent heat released/absorbed can be deduced from the convective heat transfer coefficient. However, most previous studies that were reviewed adopted convective heat transfer coefficient models that were arguably not appropriate for roof surfaces. In terms of the changes in the apparent roof surface emissivity, evidence has been presented that dew could significantly increase the thermal emittance of low-emittance surfaces. However, previous studies that were reviewed did not consider this effect.
Further analysis of the existing RP1037 experimental dataset has revealed that conditions often occurred (on ~80% of nights) that would allow dew to form on large roofs, and that roof surface temperatures can drop as much as 8°C below the dew-point temperature.
A roof condensation model has been developed, to quantify the effect that water condensation can have on roof temperatures. The influence of the ambient air temperature and humidity on the dew formation process has been explored in a set of quasi-steady cases. In the set of conditions investigated, condensation could cause a roof surface temperature deviation of 0.83oC, compared to a case in which condensation was not considered. During the remainder of RP1037u1, the roof condensation model will be applied to realistic, dynamic cases, to quantify the effects of dew on cool roof and low-emittance ‘non-cool’ roof performance.
The above-roof temperature model, developed in RP1037, has been further analysed in this report. The range of weather conditions and buildings for which it is valid have been quantified, and compared to the cases to which it was applied in RP1037. With regard to weather, all of the relevant variables (e.g. degree of cloud cover, solar heat flux, ambient temperature, etc.) influence the normalised above-roof temperature field via their effect on either: i) the wind speed, or ii) the temperature difference between the roof surface and ‘ambient’ air. The range of combinations of these two variables included in the RP1037 dataset coincides with those that arose in simulations in RP1037, except for at high wind speeds (≳ 10 m s-1). This does indicate that the degree of uncertainty surrounding outputs of the above-roof temperature model is relatively high at high wind speeds; however, such wind speeds were very infrequent in the cases of interest, and above-roof air temperatures have a diminished effect at high wind speeds, since enhanced convective heat transfer drives the roof surface temperature closer to the ‘ambient’ air temperature. Further validation of the above-roof temperature model would be valuable, but it can be concluded that the model is valid for the cases to which it was applied in RP1037.
A new, improved version of the above-roof temperature model has also been developed. It was found that by changing the mathematical form of the model for unstable conditions (i.e. when the roof surface is hotter than the ‘ambient’ air temperature), a superior fit to experimental data could be achieved. Thus, this new model should predict above-roof temperatures with more accuracy than the previous version, and will be suitable for use in conjunction with the roof condensation model developed here. The new above-roof temperature model has been outlined in this report.
Further work in RP1037u1 will apply the new roof condensation model and improved above-roof temperature model to a set of building performance simulations, to quantify the effect of dew, and above-roof air temperature fields, on roof thermal performance. Thus, the importance of these factors in the performance of cool roofs will be identified, and tools will be developed to enable building scientists to easily take them into account in the future.