Occupants interaction with electric lighting and shading systems in real single-occupied offices: Results from a monitoring campaign

• A monitoring campaign was performed in the town of Porto, Portugal.
• Eight single-occupied offices were monitored continuously under real conditions.
• Manual control of electric lighting and shading systems was assessed.
• Control patterns observed during monitoring were analyzed and discussed.

This study presents the results of a monitoring campaign aiming to further our understanding of occupants' behavior regarding the manual control of electric lighting in combination with shading control. It was performed on eight single-occupied offices in the city of Porto, Portugal during periods ranging from 28 to 60 days per office. A wide range of environmental variables including workplane illuminance, window and background luminance and transmitted solar radiation was measured with high frequency (20 min time steps). The study aimed to address a set of key research questions regarding typical illuminance ranges, luminance distribution and occupancy patterns found in offices and their relationship to electric lighting or shading control actions. It also enabled to compare observed behavior with predictions from benchmarking behavioral models found in the literature. The main findings were that electric lighting and shading control were influenced more by occupational dynamics (arrival and departure) than by the environmental conditions experienced over the day (daylight workplane illuminance or transmitted solar radiation), though with a large degree of variability between occupants and/or offices. It also revealed that while most of the behavioral models analyzed for comparison purposes were in qualitative agreement with field observations (e.g. more lighting switch-on actions at arrival for lower daylight illuminances), only one model (Pigg's model) predicted the frequency of observed lighting switch-off events. These findings strongly support the need for more numerous (and geographically more broadly distributed) office behavior monitoring campaigns to increase the robustness of such models.