Daylighting Collaborative


Johnson Diversey, Sturtevant, Wisconsin

contribute to this site

If you are aware of literature reviews that are not listed here, please contact us.

resources > literature reviews

  • A Post-Occupancy Monitored Evaluation of the Dimmable Lighting, Automated Shading, and Underfloor Air Distribution System in the New York Times Building
    Lee, E.S., Fernandes, L.L., Coffey, B., McNeil, A., Clear, R., Webster, T., Bauman, F., Dickerhoff, D., Heinzerling, D., and Hoyt, T. LBNL Report: LBNL-6023E, January 2013.

    In this report, the authors discuss the major renovation of the New York Times building in Manhattan, New York. The New York Times occupies 19 floors and 628,000 gross square feet of their building. They recently undertook a major renovation involving installing photocontrolled, dimmable lighting, automatic interior shading, and underfloor air distribution. After the project's completion, a comprehensive post-occupancy study was conducted that evaluated the energy usage, occupant satisfaction, and financial aspects of the renovation.

    This study found that the photocontrolled, dimmable lighting control system saved 3.94 kWh/ft2-year of lighting energy as compared to an ASHRAE 90.1-2001 baseline. In the daylit zones themselves, the annual lighting electric consumption was found to be 3.15 kWh/ft2-yr. This represented a savings of 56%. When combined, the three energy savings strategies reduced the building's EUI to 29 kBTU/ft2-year. The strategies also reduced the peak electric demand by 1.08 W/ft2 or between 21-25% during the summer.

    The occupant satisfaction survey showed that the majority of the space's occupants were satisfied with their new space, believing that the upgrades enhanced their productivity. A financial analysis further quantified the value of the renovation, showing a simple payback of 4 years and an internal rate of return of 25% for the daylighting system. It should be noted that this analysis found an initial installed cost of this system of $2.12/ft2.

    Finally, the report stresses that the energy and cost savings, as well as the high level of occupant satisfaction, would not have been possible without the highly integrated design process in combination with an educated and motivated owner. The savings and performance were ensured by a rigorous commissioning and post-occupancy measurement and verification period.

    LBNL has posted an article on the project, as well as link for downloading the full report.
  • Lighting Controls in Commercial Buildings
    Williams, A., Atkinson, B., Garbesi, K., Page, E., and Rubinstein, F. Leukos, Vol. 8, No. 3, January 2012, pg. 161-180.

    In this paper, the authors conduct a meta study of available research on the energy savings associated with lighting controls. In total, they summarize 88 papers, comprising 240 savings estimates of 4 controls strategies; daylighting, occupancy sensors, personal tuning and institutional tuning. Daylighting is defined as the automatic adjustment of light levels in response to daylight.

    Occupancy controls are defined as automatic adjustment of light levels in response to the presence of occupants. Personal tuning is defined as an individual adjusting their own light levels to their personal preference. Institutional tuning is defined as adjusting light levels to task-specific requirements by commissioning. 40 of the papers were published in peer-reviewed journals or presented at conferences, while the remaining 48 papers were self-published reports or case studies.

    The authors tracked savings by building type and calculation type when possible. Then, applied data filters to screen out data that was not lighting controls-specific and that included HVAC savings.

    On average, the lighting energy savings were summarized to be 28% for daylighting, 24% for occupancy controls, 31% for personal tuning, and 36% for institutional tuning. Combining multiple approaches resulted in average lighting energy savings of 38%. The study also showed that simulation typically over-predicts energy savings. In particular, simulation over-estimated daylighting energy savings by at least 10%. Taken together, this report indicates that these lighting control strategies result in significant energy savings.
  • Energy Use Intensity and its Influence on the Integrated Daylighting Design of a Large Net Zero Energy Building
    R. Guglielmetti, Scheib, J., Pless, S., Torcellini, P., and Petro, R. ASHRAE Winter Conference Paper NREL/CP-5500-49103, Las Vegas, 2011.

    The authors discuss the daylighting design process associated with the Research Support Facility, a 220,000 ft2 office building on the National Renewable Energy Laboratory campus. From the beginning of design, the project had a goal of low energy use, setting out to be approximately 50% below ASHRAE 90.1-2004. A more specific goal of an EUI of 33.3 kBbtu/ft2/yr was also targeted. Due to this aggressive goal, daylighting was chosen early as an integral part of the project's energy efficiency strategy. In fact, the RFP to the design-build contractors contained language requiring daylighting in all perimeter occupied zones, glare mitigation strategies, automatic, continuous dimming, and commissioning. The early EUI goal had the benefit of empowering the lighting designer to consider advanced controls strategies and motivated the entire team to work together to meet the energy target. This collaborative environment was realized through coupled daylight and energy simulation. The coupling was accomplished by simulating representative spaces in the Sensor Placement Optimization Tool (SPOT), a Radiance based software. For a given set of design parameters, SPOT was used to output an hourly lighting schedule, which was then imported into the energy model in lieu of its own, less sophisticated, illuminance algorithms. In this manner, several design variations were considered and an optimum set was settled upon.

    Architecturally, the design team extended the daylit zone into the building by designing tall window head heights. They achieved this by employing daylight glazing, or a set of windows above the vision glazing with high visible transmittance. A light louver system was added to the daylight glazing to bounce the light up the ceiling and even deeper into the building's interior. Further, light reflectances were chosen for the interior surfaces. When coupled with low partition heights, this minimized the amount of daylight that was absorbed and obstructed, thereby allowing a more even and deep penetration of daylight. The lighting design incorporated a task-ambient approach, utilizing direct/indirect, pendant-mounted fixture with 25 watt fluorescent T8 lamps and compact fluorescent downlights for the ambient illumination and LED downlights for task illumination. A target open office illuminance of 25 footcandles was set along with an additional 20 to 30 footcandle contribution from the task lighting. This design led to an average lighting power density of 0.62 watts per square foot. From a controls perspective, the design emphasized energy savings balanced by ease of commissioning. This approach was taken to ensure that the control system would not be confusing to the building occupants and facilities staff, a situation that often leads to the controls being overridden. The controls system itself took a layered approach; time clocks and vacancy sensors then daylighting controls. Closed loop daylighting sensors were used locally to continuously dim the electric lights to off, rather than just their minimum power fraction. Open loop daylighting sensors were used globally to override systems in large, open zones. In these large, open spaces, a north perimeter, a south perimeter, and a core set of fixtures were circuited and controlled separately. Finally, commissioning of the lighting controls was undertaken to check that the photosensor sensitivities and timeclock, vacancy sensor, and daylight controls interoperation are working properly. This final step, coupled with continuous monitoring, will ensure the building's intended level of high performance is realized.
  • Solar Zoning and Energy in Detached Residential Dwellings
    J. Niemasz, J. Sargent, and C. Reinhart. SimAUD, 2011

    The amount of solar energy that a building receives in a year greatly affects its energy usage. In densely developed areas, a building can shade its neighbors, thereby impacting their energy usage. The Solar Envelope is a zoning tool that specifies limits to a building's size, such that its neighbors are guaranteed a certain amount of daylight. The Solar Envelope is specific to the building's latitude and normally is sized such that any building whose extents lie within it will not cast a shadow on its neighbors for a specified amount of time on the winter solstice (i.e. from 10:00 am to 2:00 pm).

    This paper studies the impact of Solar Envelope zoning on a building's energy usage for 7 representative U.S. cities. An EnergyPlus model was developed of a typical two-story wood-framed residential detached. The model was then surrounded by buildings such that it was guaranteed a specified number of hours on the winter solstice. The number of hours was varied to show the impact of different solar zoning strategies.

    A low number of hours would correspond to a densely developed neighborhood, while a larger number of hours would correspond to a sparsely developed neighborhood with plenty of solar access. The modeling showed that across all seven cities the energy cost actually increased when 6 hours of solar access was compared to effectively 0 hours. However, in the heating-dominated climates, the energy usage did decrease.

    It should be noted that the shading was not optimized for each climate, but rather the same shading technique was applied to all climates. Finally, the authors point out that even though a building's energy usage may have decreased, using Solar Envelope zoning may still have a net increase in energy usage. This is due to more distance traveled by its occupants in cars and a possible decreased usage of public transportation.
  • Integrating Advanced Daylight Analysis into Building Energy Analysis
    John An and Sam Mason. Fourth National Conference of IBPSA - USA, 2010, pp. 310-320

    DOE-2.2's daylight control modeling methodology has been shown to be inaccurate. An and Mason developed a methodology to more accurately predict the lighting savings from implementing daylight controls by adjusting the DOE-2.2 lighting schedules with results from a DAYSIM analysis. Their methodology was compared to DOE-2.2's daylight modeling methodology for 1 east facing office and 1 west facing office/conference room in a campus administration building. Their methodology showed a 21% and 32% increase, respectively, over lighting energy savings in the spaces. A glare evaluation was also performed in Evalglare. However, the analysis showed that glare was not an issue in the spaces.
  • Multi-Objective Facade Optimization for Daylighting Design Using a Genetic Algorithm
    Jaime Gagne and Marilyne Anderson. Fourth National Conference of IBPSA - USA, 2010, pp. 110-117

    Genetric algorithms are a powerful method for determining the optimal set of inputs in order to achieve a desired result. In this paper, Gagne and Anderson develop a genetic algorithm for determining the set of facade parameters (window-to-wall ratio, glazing transmissivity, overhang depth, among others) that, for a given massing model, maximizes the space's illuminance within a specified range, while minimizing the space's glare potential. Their approach uses the Lightsolve Viewer coupled with an approximation of the Daylighting Glare Probability. The approach was applied to multiple spaces and validated against a known test case.
  • On the Use of Integrated Daylighting and Energy Simulations to Drive the Design of a Large Net-Zero Energy Office Building
    Rob Guglielmetti, Shanti Pless, and Paul Torcellini. Fourth National Conference of IBPSA-USA, 2010, pp. 301-309

    Daylight modeling and energy modeling have evolved on similar, but parallel, paths. Both are powerful tools that have yet to be married in a seamless way.

    This paper is a case study of the impact that these tools had on the design of the National Renewable Energy Laboratory's Research Support Facility, a net-zero, LEED Platinum building on NREL's campus. Specifically, daylight modeling was utilized to analyze the impact of different window head heights, widow-to-wall ratios, glazing transmittance, and interior finishes had on daylight illuminance levels.

    LightLouver's Daylight System was also analyzed, showing daylight penetration of up to 60' from the south-facing clerestories. A technique was also developed to integrate Radiance modeling in SPOT with DOE-2 to create a more robust daylight model inside of the energy model. This model showed an overall building EUI of 34 kBTU/ft2/yr. A portion of this low energy usage was attributable to the aggressive daylight harvesting.
  • Dynamic Daylight Performance Metrics for Sustainable Building Design
    C. Reinhart, J. Mardaljevic, and Z. Rogers. Leukos, v. 3, no. 1, 2006, pp. 1-25

    In this paper, Reinhart et al. investigates various dynamic daylight performance metrics and compares their advantages and disadvantages as opposed to static metrics.

    Daylight factor is one such static metric that is defined as the ratio of internal illuminance at a given location in a space to the external horizontal illuminance under a CIE overcast sky. This metric is often utilized by the design community due to its simplicity to calculate and is advantageous in that it accounts for building geometry, surrounding conditions, and properties of the interior materials and glazings.

    However, this metric does not account for facade orientations, season, and variable sky conditions. Dynamic daylight performance metrics are more difficult to calculate but include this variability. Recent developments in software have made the time-intensive calculations more tractable.

    One such dynamic metric is daylight autonomy, which is the percentage of a building's occupied hours that a target illuminance value is provided by only daylight. Useful daylight illuminance is another metric that varies daylight autonomy by including an upper threshold of illuminance. Continuous daylight autonomy is a third metric that incorporates partial credit for times that the naturally provided illuminance is below the target.

    The paper then goes on to calculate these static and dynamic metrics using Daysim for different design variations. Note that interior blind control was added to the analysis for three different types of controls: automated, active manual, and passive manual.

    The first set of design variations studied how the performance metrics changed for four different glazing geometries: a reference geometry including a view window and a daylighting window, a design that included a lightshelf, a design that utilized translucent glazing for the daylighting window, and a punched window. The study showed that the static metrics ranked the designs in reverse order as the dynamic metrics.

    A second set of design variations studied the impact of different types of shading control. In this study, an automated shading device ranked higher in terms of the dynamic metrics, while an active user did slightly better than a passive user. Note that the static metrics did not capture any change at all.

    A third variation studied two different climates that were at similar latitudes: Boulder, CO and Arcata, CA. The metrics were similar for both as the higher direct sunlight received in Boulder was offset by the higher use of interior blinds. Note that the static metrics again did not capture any change at all.

    Finally, varying target illuminance levels were studied. The paper concludes that dynamic daylight metrics are more useful for making design decisions and that the added time and expense associated with them has been greatly reduced. Absolute benchmarks for whether a given design is passable still need to be developed. Quantitative metrics, although important, are not all that needs to be considered in a daylight analysis, as daylight is as much an art as it is a science.
  • Lighting in Nursing Homes—the Unmet Need
    Eunice Noell-Waggoner, Center of Design for an Aging Society. Presented at the CIE Symposium and published in the Proceedings CIEx035: 2006 Proceedings of the 2nd CIE Expert Symposium on Lighting and Health, Ottawa, Ontario, Canada.

    Noell-Waggoner describes the need for improved lighting and more daylighting in nursing homes. Age-related changes to the eye tend to decrease the amount of light that reaches the retinue. This leads to higher instances of accidents and disrupted circadian rhythms. The ANSI/IESNA (Recommended Practice) RP-28-2001 Lighting for the Visual Environment for Senior Living outlines a series of recommendations for improving lighting in senior living facilities. However, there are no federal standards and few rigorous state standards for lighting in these building types. Recent studies have shown that light levels in nursing homes do not meet their residents' needs, with 45% of hallways, 17% of activity areas and 51% of resident rooms having inadequate light levels.

    Increased daylight provides higher light levels and reinforces the circadian rhythm of nursing home residents. It also promotes vitamin D synthesis, which very important for the older generation. Providing higher access to daylight has been shown to reduce anxiety and improve sleep patterns, so long as glare is kept to a minimum.
  • Sidelighting Photocontrols Field Study
    Heschong Mahone Group, Inc., Report #06-152, 2005

    The Heschong Mahone Group, supported by Southern California Edison, Pacific Gas and Electric and the Northwest Energy Efficiency Alliance, conducted this study of existing daylight-responsive lighting controls in sidelit buildings. The study covered 123 spaces in 49 different buildings mostly consisting of offices and classrooms. During the study, measurements were taken for two weeks. Each space included measurements of the controlled electric lighting's current, the uncontrolled electric lighting's current, and the vertical illuminance entering the window. Additional measurements were made of the illuminance level at the photosensor and on the critical task if possible. Through these measurements, the electric lighting savings for the actual, installed daylight control system was calculated by comparing the measured current to the total current during occupied periods.

    An energy model of each space was then constructed in DOE-2.2 using eQUEST as a front end. The simulation used real weather data collected over the same two week period as the measurements. In this way, the electric lighting savings for an idealized, perfectly operating daylight control system was calculated. The ratio of measured savings to predicted savings, or Realized Savings Ratio (RSR), was then calculated for each space. An RSR = 0 means that the photocontrols are saving no energy, while an RSR = 1 means they are working as designed.

    Of the spaces studied, 64, or a little over 50%, had controls that were either not functioning or achieving no savings. 35 of these systems were intentionally disabled by setting the sensor illuminance target too high, putting tape over the photosensor, or disconnecting the wire to the photosensor. 9 of the remaining systems had never worked or never been activated. However, older systems were actually found to save more energy than newer ones. Also, it was common that all of the spaces in a building with daylight controls were disabled together, instead of only in problematic spaces.

    The daylighting control systems in the remaining spaces were achieving actual savings of roughly half of the predicted energy savings. This equated to a lighting energy savings of approximately 1.1 kWh/sf-yr and a net peak demand reduction of approximately 0.6 W/sf of photosensor controlled area. Higher levels of energy savings were correlated with more uniform daylighting and higher levels of daylight illuminance. These were often accomplished via windows on multiple facades, utilizing glazing with high visible transmittance, ensuring that the interior surfaces had high reflectances, and minimizing partition heights. Dimming controls had higher rates of functionality with only slightly less overall energy savings when compared to stepped systems. Further, the highest performing systems were in spaces with controlled zone depths no greater than 2 times the window head height. Finally, the study concluded that both integrating the design of the architecture, lighting and controls design as well as educating the building occupants were instrumental in the success of designing daylighting controls systems.
  • A Simulation-Based Review of the Ubiquitous Window-Head-Height to Daylit Zone Depth Rule-Of-Thumb
    C. Reinhart. Ninth International IBPSA Conference, 2005, pp. 1011-1018

    In this paper, Reinhart explores the validity of various daylighting rule of thumb (DRT) design methodologies by checking their accuracy against Radiance-based models in Daysim. Specifically, the paper investigates the rule that a sidelit daylit area is defined by a depth equal to some multiple of the window head-height or ceiling height. A range of definitions exist, but the majority specify a ratio of depth to height between 1.5 and 2.0.

    A multitude of models were created representing 5 climate zones, four facades, office or classroom occupancy, high or low visible transmittance of glazing, and the presence or lack of blinds, among others. The daylit depth was calculated by finding the depth at which the space's daylight autonomy was equal to 50%. Daylight autonomy was defined as the percentage of occupied times that a task-specific minimum illuminance was provided by only daylight. The study found good agreement between the DRT and the simulation results. Reinhart concludes that "in a sidelit space with a standard window and venetian blinds, the depth of the daylit area usually lies between 1 and 2 times the size of the window-head-height." Further, in spaces that do not have blinds, this ratio can be as high as 2.5.
  • Lightswitch-2002: a model for manual and automated control of electric lighting and blinds
    C. Reinhart. A version of this document is published in Solar Energy, v. 77, no. 1, 2004, pp. 15-28

    The IESNA Lighting Handbook indicates that lighting control performance information is not readily available and that it depends on indoor daylight levels, size of space, work schedules, and occupant attitude and training. In this paper, Reinhardt develops a simulation algorithm for predicting electric lighting demand and blind usage for manually and automatically controlled systems in private and two-person offices. This algorithm is dynamic in that it uses 5 minute time steps to assess the occupancy and illuminance levels of the space. It is also stochastic, making on/off decisions regarding the lighting controls every time a user is faced with a control decision.

    Three levels of blind control are analyzed. Automated blind control comprises the blinds automatically being lowered when incoming solar irradiance is above 50 W/m2 and fully opened otherwise. Dynamic manual blind control comprises the blinds automatically being lowered when incoming solar irradiance is above 50 W/m2 and fully opened every morning. Static manual blind control involves the blinds being permanently lowered.

    The algorithm was used to model the electric lighting usage in a private office on a south-facing facade in Toronto, Canada. The office schedule was weekdays from 8:00 am to 6:00 pm and the lighting power density of the space was 15 W/m2 (1.4 W/ft2). The study showed that an occupancy sensor showed energy savings of 20%, but prevents users from utilizing available daylight. Controlling the electric lights with photosensors can save up to 60% of electric energy usage. However, this savings can be completely negated if it is not coupled with an occupancy sensor or time-clock control since the electric lights are often left on overnight.
  • Reviews of Technical Reports on Daylight and Productivity
    P. Boyce, Daylight Dividends Program, Lighting Research Center, 2004

    In October 2003, the California Energy Commission released three reports prepared by the Heschong-Mahone Group under the Public Interest Energy Research program. These reports describe epidemiological studies of the impact of daylight on:

    • Performance of office workers (Windows and Offices: A Study of Office Worker Performance and the Indoor Environment)
    • Merchandise sales in a retail store chain (Daylight and Retail Sales)
    • Progress of elementary school children (Windows and Classrooms: A Study of Student Performance and the Indoor Environment)

    Reviews of each of these studies are included in this report.
  • What's Wrong with Daylighting? Where It Goes Wrong and How Users Respond to Failure
    Vaidya, P., McDougall, T., Steinbock, J., Douglas, J., and Eijadi, D. Proceedings of ACEEE Summer Study, Panel 7, Page 30, August, 2004

    The authors summarize eight cases studies of various projects that employed daylighting controls. Each project encountered issues and the authors analyze the failure mechanism and how the users coped with them. A method of failure analysis is developed and four typical failure modes are identified. Finally, a template for resolving each kind of failure is developed. The case studies covered a wide range of space types and daylighting control systems. The issues and outcomes are summarized briefly below:

    • College Dining Hall: Non-dimmable ballasts were connected to dimming controls and lighting control zones were not matched correctly with daylit areas. The system was reprogrammed to be switched instead of dimming and a few lighting circuits were rewired. These issues would have been noticed much sooner if calibration had been attempted at the end of construction.
    • College Classroom Building: Smaller windows were installed than were analyzed in the daylight model. Pendent light fixtures blocked daylight from entering the space and darker than expected interior finishes were selected by the interior designer. Also, photosensors were located such that they measured the indirect component of the lighting system. These issues could have been prevented by including the interior designer in the daylight evaluation. Further, the location of the photosensors should have be checked in the shop drawings.
    • Office Building: A lower installed lighting power combined with dark furniture led to low light levels in the space. An overly aggressive calibration also led to numerous occupant complaints and a disabled lighting control system. This could have been avoided through education of the occupants, in order to get their buy-in.
    • Office Building: The installation of an overabundance of photosensors made it impossible to perform a full calibration. Therefore, no calibration was performed and the system remained inoperational. More collaboration between the inexperienced lighting designer and controls manufacturer could have avoided this problem.
    • College Classroom Building: The system responded extremely quickly to changes in exterior light levels, leading to occupant dissatisfaction. Further, dissimilar components were selected from different manufacturers, making commissioning very difficult. The system was reprogrammed to afford a slower response time.
    • Retail, General Merchandise: The system was working properly for over a year. However, a new store manager decided the system made the store too dark, which he believed would hurt sales. The system as therefore disabled. This example illustrates the importance of continued education in ensuring daylight savings.
    • Office Building - Existing Building Major Renovation: A photodiode type sensor was installed in an open loop control strategy. This particular photosensor was normally used to control parking lot lighting. It therefore controlled the lights off during the day and on during the night. Black tape was used to trick the photosensor into leaving the lights on, and the entire lighting circuit is manually switched off at night at the lighting panel.
    • Recreation Center: Baffles and HVAC ductwork obstructed the photosensors' field of view as well reduced natural light in the space. This was a result of a lack of coordination between disciplines.
    Four types of failures were identified throughout the case studies; under-dimming, over-dimming, cycling, and lights being left on overnight. These issues could have been mitigated through proper controls calibration or commissioning, better coordination between design disciplines, and more checks of the contractor shop drawings.
  • Windows and Offices: A Study of Office Worker Performance and the Indoor Environment
    Heschong Mahone Group, Inc., October 2003

    The Heschong Mahone Group, supported by the California Energy Commission, conducted two field studies to determine the impact of the daylight and windows on office worker productivity. Both studies were of office workers at the Sacramento Municipal Utility District call center.

    The first study tracked 100 workers' average handling time of calls over two timeframes; a 4 week period and a 3 week period. Linear regression was then applied to understand the impact of a wide range of environmental and work-related variables on the handling time. The study found that workers with a view of a large window tended to have decreased times than their counterparts with no view. The study also found that the absolute level of light, whether from daylight or an electric source, had no significant effect. Similarly, the effect of partition height was small.

    The second study tracked the performance of 201 workers on five different cognitive tests. The tests measured visual capabilities as well as short and long term memory and were given at the same time on the same day of 5 successive weeks in the fall. Having a view again increased performance, while those workers with a view reported less fatigue and overall better health. It was also reported that the quality of the view, such as having sky and trees, was important. Alternatively, those workers experiencing glare tended to have decreased performance. However, the impact of both of these variables was small.

    The report emphasizes that daylight can have both a positive or negative effect on worker productivity, depending on the way in which it is introduced. Daylighting designs with even illuminance distributions, quality views with no distractions, and minimal glare and heat gain had positive impacts on performance. Daylighting designs that introduce glare or thermal discomfort had a negative impact on performance.
  • Field Performance of Daylight-Linked Lighting Controls
    A. Galasiu, M. Atif and R. MacDonald. IES Conference Proceedings, Ottawa, Ontario, August 5-8, 2001, pp. 207-215

    This paper evaluates the performance of photosensor-controlled electric lighting under different configurations of office spaces. Four offices spaces in Ottawa, Canada were studied, each having a floor area of 14 m2 (150 ft2), height of 3 m (9.8 ft), and an occupied schedule of weekdays from 8:00 am to 5:00 pm. The offices each have a high visible transmittance view window and a low visible transmittance clerestory window along with 2 two lamp, 32 W, T8 fluorescent lamps. This equates to a lighting power density of 9 W/m2 (0.84 W/ft2). The electric lights were controlled to a work plane illuminance of 570 lux (57 fc). Two of the offices had electronic dimming ballasts, while the other two had on/off controls. One of the "dimming" offices and one of the "on/off" offices had motorized blinds tied to the photosensors. The results were preliminary at the publishing of this paper. However, the results did show that adding static blinds increased the electric lighting usage by between 40 and 45% for the "on/off" system and between 30 and 35% for the "dimming" system. Further, controlling the blind angles via the photosensors showed only a 10% increase in electrical usage.
  • TripleSave - The Investigation and Monitoring of a Combined Natural Daylighting and Stack Ventilation System
    G. Oakley et al, Institutes of Building Technology, University of Nottingham, c. 2000-2001.

    In this paper, Oakley et al develop a testing method for measuring the amount of natural daylight and passive ventilation a light pipe introduces into a test chamber. Two light pipes were analyzed, both having diameters of 0.215 m and lengths of 2.2 m. Concentric cylinders let light into the chamber and ventilation air out via natural convection. Photosensors measure the illuminance levels both internal and external to the chamber, while a tracer-gas method was employed to measure the ventilation flow rate. The measured ventilation flow rate was compared to a theoretical flow rate and showed good agreement.

    The results showed a ratio of internal to external illuminance of approximately 16%. The highest natural light penetration was during periods of time that the sun was overhead. However, the addition of a Laser Cut Panel to the top of the light pipe enhanced daylight penetration during periods of low angle sunlight. The natural ventilation averaged 8 air changes per hour of the 1.3 m x 1.3 m x 1.3 m test chamber during the study, with little to no effect from wind speed or direction.
  • On the Calibration and Commissioning of Lighting Controls
    F. Rubinstein, D. Avery, J. Jennings and S. Blanc. Proceedings of the Right Light 4 Conference, Copenhagen, Denmark, November 19-21, 1997

    Rubinstein et al. discuss commissioning and calibration of lighting control systems. Specifically, they address the importance of both as well as advice for effective performance. In this context, commissioning is a process for ensuring the lighting system performs as the design intended. The authors note that the majority of systems are not commissioned at all. Calibration is defined as the adjustment of a sensor in order to get the desired output from a given input. Specific activities for both calibration and commissioning include:

    • Verifying photosensor placement and orientation. Adjusting sensor and controller to obtain desired light level on the working plane.
    • Verifying occupancy sensor placement and orientation. Adjusting the sensitivity and time delay.
    • Inputting correct start and stop times into the lighting control software.
    • Setting upper and lower dimming limits.

    Three tips for calibrating a photosensor are outlined. The first is to place the sensor in the ceiling near the primary working area. The second is to calibrate the sensor at a distance if possible. The final tip is to use a photometer in conjunction with the calibration to provide feedback as to the working plane's illuminance level. The authors conclude by suggesting that an open-loop control systems may be properly calibrated in approximately 30 minutes, while close-looped systems are consistently more difficult to calibrate.
  • Field Commissioning of a Daylight-Dimming Lighting System
    D. Floyd and D. Parker. Presented at the Right Light Tree, 3rd European Conference on Energy Efficient Lighting, Newcastle upon Tyne, England, June 18-21, 1995

    The authors performed a lighting retrofit on an elementary school cafeteria in Florida. The cafeteria had 70% glazing on its east and west facing facades, as well as vertical blinds. The retrofit replaced the magnetically-ballasted, 40W T12 lamps with an electronic dimming ballast and two 32W T8 lamps. Additionally, the lights were controlled via photosensors to a minimum of 20%. The building's lighting and HVAC energy usage, desktop illuminance, and exterior solar insolation were all monitored both before and after the retrofit. The retrofit showed a 66% decrease in the cafeteria's energy usage. Approximately 76% of this decrease was attributable to the lower power required by the new luminaires. The remainder was attributable to the photosensor dimming of the lights. The daylight controls were calibrated in order to enhance their energy savings potential. Multiple issues arose during this commissioning. Those issues included the necessity of utilizing the manufacturer supplied shields to the photosensors. Further, it was necessary to obtain knowledge of the photosensor sensitivity range, which was not readily available. Also, the photosensors controlled the lights to different levels between day and night, requiring a trial and error approach to properly calibrate them. Finally, it was important to verify the illuminance in the room against measured data. The usage of the blinds also affected the dimming of the system. It was found that when the blinds were left open, the dimming saved 36% of energy usage, while this number decreased to 27% when the occupants were allowed to adjust the blinds.
  • Improving the Performance of Photo-Electrically Controlled Lighting Systems
    F. Rubinstein, G. Ward and R. Verderber. Presented at the Illuminating Engineering Society Annual Conference, Minneapolis, MN, August 7-11, 1988

    Rubinstein et al. analyze three different algorithms for dimming electric lights in response to changing daylight availability. The three basic components of a photo-electrically controlled light system are described: a photosensor for measuring the illuminance at the working plane, a controller for processing the produced signal, and a dimming unit for varying the amount of electric light provided. The analyzed algorithms include:

    • Integral Reset: adjusts dimming level such that the measured illuminance is kept at a constant level. This algorithm requires a "night-time calibration"
    • Open-Loop Proportional: photosensor detects on daylight and the control provides a linear relationship between the available daylight and the dimming level. This algorithm requires a "daytime calibration"
    • Closed-Loop Proportional: photosensor detects both daylight and electric light and the control produces a linear function of the difference between the available light and a night-time calibration level. This algorithm requires both "night-time calibration" and "daytime calibration"

    The different algorithms were then tested in scale models to determine their accuracy in controlling the illuminance level on the workplane to a set level. Different room shapes, window geometries, glass transmittance, and exterior shading devices were introduced. The closed-loop proportional control performed best. This was particularly true when it had a large field of view that did not include the window. Of the other two algorithms, the open-loop proportional control outperformed the integral reset control.