Revealing Occupancy Diversity Factors in Buildings Using Sensor Data

Pierrick Bouffaron (1, 2) , Therese Peffer (2)

1 CMA - Centre de Mathématiques Appliquées
2 CIEE - California Institute for Energy and Environment

The definition of the number of people that occupy a particular space and for what duration is difficult to characterize because human behavior is considered stochastic in nature. Occupants’ locations within a building vary throughout the day and this distribution can be valuable information when evaluating demand control strategies. Occupancy diversity factors have not been studied as extensively as for example lighting and plug loads diversity factors. Some reasons for fewer studies of occupancy is limitations accessing existing occupancy datasets and challenges interpreting the data. In a research building at UC Berkeley, we were able to add sets of passive infrared (PIR) or motion sensing for occupancy and carbon dioxide sensors in 67 private offices and 2 conference rooms, as well as in multiple open offices. In this work we study deterministic and stochastic building occupancy models based on data from the deployed sets of sensors. Data is analyzed to show major variations of occupancy diversity factors in private offices and conference rooms for time of day, day of the week, holidays, and month of the year. The impact on the building electrical load is highlighted: people usually operate electric lights, computers, and other common office devices when in a space, and this equipment is often turned off or in sleep mode when the space is not occupied. The diversity factors presented in this study can differ as much as 40% from those published in the literature or in the latest ASHRAE energy cost method guidelines, a document commonly used by energy modelers for building simulations. This may result in misleading simulation results and may introduce inefficiencies in the systems design and control. Therefore we argue that building occupancy is a basic and key factor in energy simulations. Occupancy sensors can certainly help in calculating better diversity factors, but what is the optimum number and distribution of sensors to improve performance and justify the cost?

Mots clés: diversity factor; demand control strategy; building energy consumption; building occupancy

Domaines: Planification énergétique; Prospective long-terme

Pierrick Bouffaron, Therese Peffer. Revealing Occupancy Diversity Factors in Buildings Using Sensor Data. Behavior, Energy and Climate Change Conference, Precourt Energy Efficiency Center (PEEC) at Stanford University, American Council for an Energy Efficient Economy (ACEEE), and California Institute for Energy and Environment (CIEE) at the University of California, Berkeley., Dec 2014, Washington, DC, United States. ⟨hal-01108017⟩