Daylight Savings: Not an Easy Catch
Daylight Savings: Not an Easy Catch
Guest author David B Floyd writes about lessons learned from the field commissioning of a daylight-dimming fluorescent lighting system. To get all parts of the system working properly was not easy, and the manufacturer-supplied instructions were of little help.
Daylighting interior space has long been recognized for its health, aesthetic and energy-savings potential. New building designs incorporating spectrally-selective glazings, light gathering devices, and controllable fluorescent ballast technology can offer significant reductions in energy use. Older buildings with poor daylighting design, on the other hand, present more challenges. Of particular interest are the vast numbers of buildings with substantial daylight entering through windows with interior shielding devices such as blinds. Will a simple retrofit to a dimming ballast controlled by a photosensor be cost effective and perform properly under these circumstances? What are the commissioning difficulties and how will they affect the energy savings of such a system?
A CAFETERIA CASE STUDY
To answer these questions, the Florida Solar Energy Center commissioned a daylight-dimming lighting system in an elementary school cafeteria located in Florida. The site was chosen owing to the substantial daylight available through windows on the east and west walls. Prior to the commissioning, the building was lit by 25 four-lamp recessed luminaires with magnetic ballasts and T-12 lamps. Vertical blinds had been previously installed by the school in an attempt to control unwanted heat and glare.
The retrofit replaced the inefficient magnetic ballasts and 40-W T-12 lamps with a single electronic dimming ballast and two 32-W T-8 lamps in each luminaire. The cafeteria was divided into five linear north-south zones, with each zone of five luminaires controlled by a single ceiling-mounted photosensor. The photosensor regulates the light output (down to 20%) of the ballasts in each zone based on the available light measured in its field of view.
These changes brought the lighting levels down to those recommended by the American Illuminating Engineering Society for cafeterias (~300 lux), increased the lighting system efficacy, and allowed for more effective utilization of daylight. To measure the effectiveness of the lighting retrofit, the building is being monitored for air conditioning and lighting energy consumption, desktop light levels, weather conditions, interior conditions, and luminaire temperatures. Five photometers mounted at desktop height record illuminance.
COMMISSIONING DIFFICULTIES
While significant energy savings were anticipated from the new system, the difficulties associated with calibrating the control photosensors were not. When early data showed minimal dimming, an attempt was made to recalibrate the photosensors. However, proper dimming response was not achieved until the manufacture-supplied sunshields were installed (These shield the sensor from reflections and other light not originating from the closest window.) We believe that the sunshield prevented daylight and reflections originating from the other side of the room from saturating the sensor.
For example, a sensor in an eastern zone would be shielded from light from the west. This allowed the sensor to respond better to light entering the building near the controlled zone. Zones on the east side dimmed in the morning, while zones on the west side dimmed in the afternoon. Post-recalibration data showed that the dimming response had improved.
Photosensor calibration difficulties such as the ones experienced during the case study are by no means unique. In another case study and through side-by-side testing of different photosensors we found a number of problems. First, all manufacturer-supplied instructions on how to adjust the photosensor-ballast response were found to be woefully inadequate. Proper adjustment of the daylighting system response required knowledge of the photosensor sensitivity range. (For example, turning the set potentiometer too high can result in poor dimming response.)
We also found it difficult to adjust the photosensors to the desired illuminance level with any accuracy (plus or minus 50 lux), and the illuminance levels were often less during the day than at night. (For example, an illuminance of 500 lux set at night would measure 400 lux during the day.) We often resorted to a trial-and-error approach to obtain a good dimming response.
Lastly, the recommended mounting location of the photosensors and the use of sunshields (when supplied together with the photosensors) varied amongst manufacturers. Most suggested a location at two-thirds the distance into the room from the window. However, for our case study this did not apply since we used linear control zones.
Another factor affecting system performance is the control of the blinds since they are frequently drawn to reduce glare and localized overheating. Closing just one blind near the photosensor can drastically alter the energy savings. In the case study, we could achieve as much as 36% dimming savings when the blinds were left open, but overall savings attributable to dimming diminished to 27% when control of the blinds was left to the occupants. While such control is necessary, we wanted to maximize the amount of visible light entering the building while minimizing ultraviolet and infrared and reducing the glare.
The retrofit resulted in an outstanding 70% overall reduction in interior lighting energy use compared with the original inefficient system, with the dimming capacity of the system responsible for approximately 27% of the total savings. Our study shows that a daylight dimming lighting system has the potential to save substantial energy: However if not all factors influencing system operation are addressed, visual comfort and savings may be compromised. Proper commissioning of such systems is therefore vital to realizing system efficiency potential.
THE NEXT STEP
Since the glare and solar heat gain need to be controlled without obstructing the visible light, we plan to apply a spectrally selective window film during July 1995. Such films filter the near-infrared light (heat) and admit most of the visible light. This measure is expected to significantly lower air conditioning consumption while also affecting the performance of the dimming system.
Although adding the window film to the clear window glazings should theoretically reduce available daylight (the film we plan to use has a visible transmittance of ~70%), the way in which the lower shading coefficient (0.50) may affect user operation of the east and west window blinds is a potentially important factor that would influence the overall results. Since there will be less heat and glare, will the blinds remain open more, resulting in improved savings? It will be interesting to find out.
David B. Floyd
The author performs lighting research at:
Florida Solar Energy Center, 300 SR 401,
Cape Canaveral FL 32920, USA
Tel: +1 407 783 0300, ext. 295
Fax: +1 407 783 2571
| The cafeteria was divided into five linear north-south zones, with each zone of five luminaires controlled by a single ceiling-mounted photosenosr. The photosensor regulates the light output (down to 20%) of the ballasts in each zone based on the available light measured in its field of view. |
| A 36% dimming savings could be achieved when the blinds were left open (Sept.26), but overall savings attributable to dimming diminished to 27% when control of the blinds was left to the occupants. |