Testing glazing outdoors
In 1995 the Energy Monitoring Company was commissioned by DeMontfort University to collect data on the performance of a series of different types of glazing in TEST ROOM 3000. The work, which formed part of an EPSRC project, was aimed at validating the representation of glazing in current thermal simulation models. As well as fulfilling this objective, the work demonstrated that the facility was a credible means of testing the heat loss of glazing systems under real climatic conditions. The project was described in detail in a recent article in IBPSA news.
Six types of glazing were tested. The project was oriented towards the validation of simulation models, but we were naturally also keen to see whether TEST ROOM 3000 could reliably measure a parameter as simple as U-value under real climatic conditions. The table below shows the U-values currently assumed for the glazing systems tested, together with the values derived from the data collected in TEST ROOM 3000.
| Glazing type | Conventionally assumed U-value | U-value measured in TR3K |
| Single glazing | 5.5 W/msqK | 5.2 W/msqK |
| Double glazing (6mm gap) | 3.1 W/msqK | 3.0 W/msqK |
| Double glazing (12mm gap) | 2.7 W/msqK | 2.4 W/msqK |
| Double glazing with low-e coat | 1.9 W/msqK | 1.7 W/msqK |
| Double glazing with low-e coat and gas fill | 1.6 W/msqK | 1.6 W/msqK |
| Triple glazing with low-e coats and gas fill | 0.9 W/msqK | 0.9 W/msqK |
You might wonder what makes us so enthusiastic about these results. At first the fact that they agree pretty closely with 'conventional wisdom' does seem to lend some credibility to TEST ROOM 3000. However, most of that conventional wisdom stems from tests made in laboratory hot boxes or even calculation. So why spend huge amounts of money and suffer the aggravation of outdoor testing if all it does is return the same results as these simplistic (and cheap) methods ? The answer to this probably lies mainly with the nature of the parameter derived from these tests. The U-value is a combination of internal and external heat transfer coefficients and relatively well understood conductive and radiative transfers within the cavities of the multi-pane systems. This possibly explains why the divergences disappear for the lower heat loss systems, where the principal uncertainties have a smaller impact.
It would be interesting to determine the solar transmission characteristics of the glazing systems tested. The analysis required for this was not developed as part of this project. All of the data collected in these tests is available electronically from the DeMontfort University web site, and further analysis to derive, for example, solar transmission parameters might prove very fruitful. Whilst the Energy Monitoring Company can't undertake to provide formal support for such an exercise we would be happy to consider an informal collaboration with any interested parties. In the meantime anyone using the data should acknowledge the source, as a matter of common courtesy.
Where the facility will really come into its own is in measuring the performance of systems which can't be tested in the laboratory. Examples might be solar shading devices, and windows whose heat loss depends on incident radiation, either because cavities are ventilated or because they contain shading elements.
A proven facility now exists in the UK for carrying out these measurements. Enquiries will be welcomed !