Comment on ‘Shading devices and daylight quality: an evaluation based on simple performance indicators’ by M-C Dubois

This article addresses a crucial issue in the performance of daylighted ofŽ ces. The necessity of controlling direct solar radiation under clear skies inevitably impacts daylight illumination levels and distribution. The careful and accurate methodology presented can be viewed as an initial model for addressing this important daylighting issue. In particular, the calibration of the Radiance model with a full-scale test room and the inclusion of the discussion of sky data variations are welcome. Both steps in the methodology enable the reader to understand the accuracy of the Ž nal results. Expanding the application of the methodology in future studies will allow a more signiŽ cant range of questions to be addressed. A typical difŽ culty in daylighting research carried out through parametric simulation is knowing how to bound the nearly inŽ nite number of variations modeled. This article presents a large number of variables in order to address a simple, single desk south-facing ofŽ ce. However, to generate information which can be useful in practice, the worst case conditions and conditions within the range of standard practice should be included. Of these, orientation of the ofŽ ce window is a crucial issue to broaden, in that eastand westfacing apertures are more difŽ cult to shade while still admitting useful daylight. Factors which modify the Ž nal luminance of surfaces in the visual Ž eld, such as interior re ectivity and glazing visible transmission are obvious factors to include in the next round of study. Performance of the ofŽ ce with these shading devices under overcast skies is an important issue, since most shading devices do not necessarily disappear when the sun does. Often shading devices reduce the overcast illumination levels, although they often increase visual comfort on bright overcast days. Finally, as ofŽ ces move inevitably to  at screen computer monitors, the conditions of glare and luminance may be shifting to the beneŽ t of visual comfort. Acknowledgement of the human factor in the methodology is another direction which could generate useful results. For example, if 70% of ofŽ ce employees sit near the window, these locations and the actual furniture arrangements could direct the choice and emphasis of additional parametric runs. Since full scale test ofŽ ces are available, they could be used for some user testing of conditions similar to those modeled. This could help to inform the interpretation of the data generated, especially as to the tolerance and latitude afforded to daylighting compared with electrical lighting conditions. Finally, the core issue of performance criteria is not fully resolved. Glare and visual comfort are difŽ cult to quantify and it is not clear that the four performance indicators chosen to evaluate the shading device performance are the most useful. We know that absolute illumination levels are subject to wildly different interpretations by users in daylit rooms, although nearly all illumination performance criteria still begin with the illumination level on a work plane. Uniformity of work plane illumination throughout the ofŽ ce has little to do with daylighting performance at one location and is the least revealing indicator. The absolute luminance of surfaces presents difŽ culties similar to that of absolute illumination levels. The relative luminance of surfaces in the visual Ž eld is the key factor and a more detailed presentation of these results using Radiance false colour renderings would be extremely useful.