Summary

A long-term study of a new academic building with advanced control hardware at Carleton University revealed that the overhead occupancy-based lights in private offices were on 1.6 times as long as the occupants were present. This motivated us to investigate: why the lights were on so long, what the corresponding energy implications were, whether the occupants were satisfied with the lighting controls and whether the controls could be modified to improve occupant satisfaction and energy performance simultaneously.

A five-month monitoring study of the existing lighting control scheme showed that many occupants were dissatisfied with the lighting control scheme because it was set to turn on lights automatically despite the good daylight spaces. Many of the occupants took their own interventions without contacting facilities management. A quick modification to the control logic (approximately five minutes per office), whereby DIP switches on each light switch device were changed, yielded energy savings of over 60% while greatly improving occupant satisfaction (based on measurements of one additional year). Aside from the specific findings, this study also revealed the importance of 1) centralizing all building controls functionality (including lighting) into a single building automation system; (2) subtle controls decisions’ impact on overall building energy performance and occupant satisfaction; and (3) continuous monitoring of building performance to identify issues and opportunities.

Background

Richcraft Hall (formerly River Building) at Carleton University in Ottawa, Canada is a 17,000 m2 academic building with a variety of functions including lecture halls, conference spaces, and private offices. It features dozens of highly daylit private offices around the east, south, and west-facing facades (Figure 1).

Study period

In 2015, we noticed that occupancy-based lights in the private offices were on for much of the day, regardless of occupancy. To investigate the cause of this energy-adverse observation, we performed a monitoring study. The building is equipped with state-of-the-art controls equipment (including a DNS-L24 (Figure 2) in every office). However, the lighting controls in Richcraft Hall are decoupled from the central building automation system (BAS). To monitor lighting use, we affixed light state data loggers to the lamps in 25 offices (Figure 3). This approach required us to manually install the devices, obtain consent from all office occupants, and collect and post-process frequent data collection. We collected data for 15 months starting in March 2016.

Accompanied by the light use monitoring, the lead researcher collected numerous anecdotes of unsatisfied occupants, as indicated by the photographic evidence in Figure 4. In contrast to the frequent-cited complaint that lights automatically turn off as a result of automation (e.g., daylight or vacancy-based control), Richcraft Hall occupants tended to be annoyed that the lights were automatically turning on despite often having ample daylight. The original controls were set to turn on the lights upon detected occupancy and turn them off 15 minutes after the last detected motion, during the five-month study period (and since building occupancy five years earlier).

The results of the five-month monitoring study indicated that for the 25 offices, the median ratio of lights on to occupied period was 1.6. In other words, the lights were on at least 60% longer than necessary (if the lights were on for all occupied periods). And in fact, most occupants did not want the lights on even during occupied periods. The reason for this ratio being above 1.0 is that these occupants routinely leave their office for meetings and to teach the class. Thus, each departure event was associated with as much as 15 minutes when the occupant was not present.

Intervention

Following the study period, the lead researcher switched the lighting controls to manual-on/vacancy-off after 30 minutes. In other words, the occupant must turn on the lights manually, and the lights turn off after 30 minutes of no detected occupancy. The 30 minutes is used to prevent false negative presence states where the occupancy sensor fails to detect a near-motionless occupant. The lighting control product in use required us to change in position of several DIP switches which we accessed under the cover of the light switch in each office. However, for the sake of reducing disruptiveness and manual labour, it would have been ideal if we could make the change centrally.

Post-intervention

Following the modification to the controls, the lighting use in the 25 offices was monitored for an additional 10 months ending in June 2017. The data show that the simple change of control schemes reduced energy use by an average of 62% (Figure 5). Moreover, some occupants did not use lights at all for certain months.

When the lead researcher met with the occupants, most occupants revealed that they are more satisfied with the post-intervention lighting controls. Only two of the occupants expressed a preference for the initial controls, as they liked to be able to walk into their offices with their hands full and not need to turn on the lights manually.

Key lessons learned

1. Our study greatly reinforces the change in ASHRAE/ANSI/IES Standard 90.1-2016, upon which most building energy codes are based, that new buildings require lighting controls to be manual-on and vacancy-off after 20 minutes.
2. Centralizing all building controls functionality (including lighting) into a single building automation system is critical for optimizing controls for occupant comfort, satisfaction, and energy. In the current case, centrally controlled lights that integrate with the zone-level HVAC controls would have allowed much more efficient deployment of the new scheme while also allowing lights to be controlled based on the DNS-L24 occupancy sensor. 
   
3. Subtle no- or low-cost controls decisions can have a tremendous impact on overall building energy performance and occupant satisfaction. Despite the apparent relative impact of subtle control scheme choices, they receive a disproportionately low amount of attention during design meetings. Further research and post-occupancy evaluation studies are needed to help inform best practices. In this case, the post-intervention control scheme improved occupant satisfaction and energy use; but the upgrade was only made five years after occupancy.
   
4. Continuous monitoring of building performance is essential for identifying controls-related issues and opportunities. Thanks to the data archiving capability, the researchers were able to identify a major cost-savings opportunity and present ample evidence to facilities management.