In recent years, with heightened environmental awareness and the development of new lighting control technology, various energy-saving measures for luminous environment have been adopted in offices. Among them, partial-off or dimming lighting on the ceiling, which was performed in the aftermath of the Great East Japan Earthquake, is certainly one of the effective means for energy saving, however it may cause uncomfortable non-uniformity in luminous environment. Uniformity ratio has been generally used as an index for evaluating the uniformity of lighting environment on task areas, whereas it would not always be a robust index for the uniformity of ambient lighting environment. This study aims to examine and verify several indexes for evaluating the non-uniformity of ambient lighting by comparing the occupants’ evaluation in some real work areas with the potential indexes which are calculated from luminance and illuminance distribution. The workspaces in this study are limited to open-plan offices under evenly distributed ambient lighting.
In the experiment the occupants evaluated the lighting environment after spending at least one week under each lighting condition. The experiments were conducted at three different actual work areas. The first experiment was conducted in a room of the university laboratory. The size of the space was 4,350 mm wide, 9,650 mm deep and 2,700mm high. The graduate students in their early 20s participated in the first experiment. The second and the third experiments were conducted in the actual office. Each size of the space was 8,550mm wide, 14,350 mm deep and 2,600mm high, and 7,645mm wide, 14,350mm deep and 2,600mm high, respectively. The occupants in their 20s to 50s evaluated the luminous environment in these experiments.
As a result of the analysis, a single regression analysis between the coefficient of variation (explanatory variable) calculated from the true values of the ceiling luminance (including light sources) and the mean values of the occupants’ assessment on non-uniformity of the spatial brightness (objective variable) showed a relatively high coefficient of determination of 0.54. Path analyses suggested that the spatial brightness had an effect on the evaluation of the non-uniformity, therefore, a regression analyses was also performed by adding the arithmetic or the geometric averages calculated from the true values of the illuminance and the luminance. The highest coefficient of determination 0.73 was obtained when both the coefficient of variation and the arithmetic mean calculated from the true values of the ceiling luminance (including light sources) were used as the explanatory variables. However, there were still some gaps between the estimated values and the evaluated ones, suggesting that there may be differences in evaluation tendencies among the different spaces. Therefore, trivariate multiple regression analyses were also performed by adding dummy variables 1on the spaces. The result showed that the explanatory power of the multiple regression analysis using the illuminance and the luminance increased from the bivariate cases to 0.87 and 0.83, respectively.
It was found that the luminance distribution on the ceiling (including light sources) can roughly estimate the the evaluation of the non-uniformity of lighting environment in workplaces under evenly distributed ambient lighting, however, there is a possibility that the other characteristics of space that could not be analyzed in this paper would also affect the non-uniformity evaluation, and further study would be necessary on that.
