LBNL Report Number
We present a detailed consideration of the energy flows entering a building space and the effect of random measurement errors on determining fenestration performance. Estimates of error magnitudes are made for a passive test cell; we show that a more accurate test facility is needed for reliable measurements on fenestration systems with thermal resistance 2-10 times that of single glazing or with shading coefficients less than 0.7. A test facility of this type, built at Lawrence Berkeley Laboratory, is described. The effect of random errors in this facility is discussed and computer calculations of its performance are presented. The discussion shows that, for any measurement facility, random errors are most serious in nighttime measurements, and systematic errors are most important in daytime measurements. It is concluded that, for this facility, errors from both sources should be small.