ABSTRACT

The study aims to develop a method for glare analysis, based on recorded luminance variations in a space which is useful for post occupancy building energy performance analysis in terms of change in energy use patterns caused due to glare and visual discomfort and the resulting occupant behavior and interaction with lighting controls. Such a method would be useful in predicting occupant interaction as well.

1.0 INTRODUCTION

Occupant interaction with lighting and lighting control systems has been found to significantly affect the energy use patterns of spaces. The extent of occupant interaction with lighting controls depends on the degree of visual discomfort experienced by the occupant in the space. Visual discomfort in a space is primarily due to high contrasts, and luminance variations in the field of view of the occupant. It is clear that occupant sensing and photo sensing lighting controls in buildings, succeed in reducing their electrical as well as air conditioning requirements, but the effect of glare or visual discomfort on the occupant and the resulting interaction with the lighting control systems has not yet been included in determining energy use patterns. It is also not clear how glare and visual discomfort effects occupant attitude and productivity in a space and how the subsequent modifications made by the occupant governs the performance of the lighting systems within the space. Glare and occupant response and attitude to it causes significant changes in the predicted energy requirements of the space and hence needs to be carefully evaluated.

2.0 BACKGROUND

2.1 Existing Theories

At first glare or discomfort was understood to be caused due to high levels of illumination in a space. But the simple understanding that the higher light levels outside a building did not cause discomfort, radically changed this thinking. Glare was then thought to be caused because of the high contrast ratios within the room. By experimentation it was found that a ratio of 10:1 was problematic and would cause discomfort within the space, and that achieving maximum contrast ratios of 3:1 within a space would make the space comfortable. However this theory can be disproved easily. A bright sheet of paper with black ink printing on it has a contrast ratio of more than 10:1.

The first glare theory was developed by Hopkinson in his Glare Index Method (1)(2)(3). Based on extensive experimentation and subjective testing he proposed that glare was dependent on the background level within a space. The field of view and glare source were defined in terms of steradians. This method seemed to address the complex phenomena of glare in terms of contrast between high intensity glare source and background luminances.

Visual Comfort Probability (4)(5)(6) is the current state of art method to determine comfort within an interior space. Comfort is estimated on the basis of how many people out of 100 would feel comfortable within the space. This method involves extensive calculations of subtended angles, average luminances and position indices of each luminaire in the field of view of the occupant. The field luminance or the background luminance is calculated as the sum of the luminances of each luminaire and its contribution to the

background by interaction with the reflective surfaces of the room.

Relative Visual Performance (7)(8) is another existing method in which the comfort within a space is defined on the basis of the speed and accuracy of performing a task. Visual performance is studied as contrast changes between paper and ink, or task contrast and the effect of any glare source in the field of view of the occupant or change in the adaptation level in the space is accounted for as a change in the time taken to perform the task or in the increase in the number of errors. This method, although suitable for determining performance at a single task, is not capable of evaluating the luminous quality of the entire space. The underlining concern therefore in glare quantification is the need to be able to numerically compare all the luminances within a field of view simultaneously to the background level of the space.

2.2 Significance of Luminance Measurements

While evaluating glare and visual comfort it is necessary to know not only the individual luminances of small areas of an interior, but also the average luminance of the larger area or the background luminance. Typical luminance measurements do not describe the visual phenomenon. Rather luminance meters(9)(10) either average the visual field at some angle of view to provide a single number value or measure the detail within the visual field to describe some specifically interesting point. These measurements are in no way proportional to the complex visual environment.

Video Photometry (11)(12) has been found to offer the capability to map and characterize the visual environment. The primary benefit of using video cameras is being able to record different solar positions, shadows etc. over time and to be able to correlate action with each variable. Video data can be digitized and used for later analysis in different algorithms to evaluate the luminous environment and then correlated with the recorded occupant behavior. However there are extensive calibrations required to offset camera gains, settings and spectral responsivity etc. The new Luminance Distribution Method (13) offers a solution to the prevalent drawbacks in video photometry like recording f-stop, range etc. by introducing a known luminance box into the image which acts like a self calibration scale.

3.0 ENERGY IMPLICATIONS

To evaluate the extent of change in the predicted energy requirements of a space because of glare and occupant response, an office space was modeled using DOE2.1E.

There is a 65% reduction in the percentage of light energy reduced by utilizing daylighting when glare controls are exerted over a space (See Fig. 1). The peak cooling load increases by 35% and the heating load by 2% (See Fig. 2). The electrical energy (KWH) for this space increases by 60% compared to that over which no glare control is exercised.

The space modeled is a single room with 2 occupants and a lighting schedule incorporating a stepped control system and a 70% probability of manual interaction with the lighting controls.

The maximum glare index was defined as 22, and the view azimuth was 0° , that is the occupant is looking straight up at the window. DOE does not model other view azimuths, the results are the same as when the space is modeled without any glare control. However the extent of change in the predicted energy requirements of the space caused due to glare and resulting occupant response can be judged from the results of this study and the importance of addressing glare control issues during the initial energy analysis stage is substantiated.

4.0 DETAILS OF GLARE ANALYSIS METHOD

The initial study was conducted at the Collins Center, Cal Poly, Pomona and deals with the recording and evaluation of the luminances over time within an office space. Glare is analyzed in terms of contrasts and luminance variations across the space.

4.1 Description of method

The known luminance box uses long life fluorescent lamps and opal glass diffusers to provide a known luminance of 250 fL. Each box has an absorptive surface of approximately 0 fL providing 2 surfaces within the image, one at 0 fL and the other at 250 fL to provide the range of

the pickup by the camera. Since the image is now certain in absolute value, other portions of the image could be determined in relation to or calibrated from this known absolute value. The boxes were located against the window wall below the desk level, so that there is a minimum effect of the other light sources on them. The camera was mounted facing the wall, and perpendicular to the window outside. The luminance of the window, position of blinds, solar position, sunspots and shadows within the room were recorded. Occupant behavior was recorded in terms of whether the venetian blinds were let down or not and whether the occupant changed his task location to escape the feeling of discomfort.

4.2 Analysis of Histograms:

1. Shape and Distribution of the bell curve: There is a distinct bell curve observed in the histograms of images of almost all days tested. This bell curve can be assumed to be representative of the background level. The shape of the bell curve is sensitive to the luminance distributions within

the space. A wider bell curve implies a more uniform distribution of light intensities across the space. A narrow bell curve implies lesser distribution of intensities across the space.

2. Field of view: From the histograms it is possible to evaluate the number of pixels within that field of view. Within any fixed field of view there is a linear relationship between the number of pixels in any image and the actual steradians in the field of view.

3. Relative Range of Intensities: The histograms also provide a relative range of intensities which are present in the space. These are then correlated with the color-coded images of the space to get the absolute values of the intensities. However just the relative range of intensities provide us with useful information on the contrast ratios present within the space. For example, there are ratios of

1:200 to 1:250 present within the space, but not all of them produce discomfort. The ratio of intensities of the glare source and the background can be established from the histograms and used to predict glare situations.

4. The spike: The spike indicates the intensity reached by the maximum luminance values of pixels in the space. The position of the spike on the histogram and its relation to the bell curve determines the visual comfort within the space. A spike at the low intensity indicates that there is little difference between background intensities and peak areas. A spike at the high intensity could be a potential glare source, and can be determined from the images. The relationship between the spike and the bell curve is what determines glare or no glare situations.

5. Numerical Analysis: The histograms were numerically analyzed in terms of the median pixel intensity, number of background pixels, maximum intensity and ratio between the maximum intensity and background level (See Fig. 3).

4.3 Analysis of Recorded Images

The relative values of luminances within the room are available from the images. The images are color coded, each color corresponding to a specific range of intensities.

The image is analyzed for high intensity surfaces, which could be either the window or a patch of light on the wall or floor. By identifying the surface which is at the highest intensity we can then evaluate its importance in the field of view of the occupant. Occupant interaction with visual environment is seen through the position of blinds, change in task location etc. Examples of a few of the cases are discussed below.

April 5, 5:00 PM Room 1

The known luminance box is in the 160-192 relative range on the histogram. The window sill and table along the far opposite corner of the room are at the highest intensity and correspond to the spike at the high end of the histogram (See Fig. 4). The bell curve is tall and narrow signifying low background luminance levels and greater contrast, which is confirmed by the image (See Fig. 5).

This causes extreme discomfort within the space. The blinds were let down over the top portion of the window, to cut out the direct sunlight, as a means to reduce the discomfort. There is internal glare within the room, caused by the excessive brightness of the wall compared to the desk surface and monitor or the contrast between the papers on the desk on the far side of the room compared to the rest of the background. The window sill would cause discomfort if it was within the field of view of the occupant. As such there is plenty of internal glare caused due to high contrast between background and areas of excessive brightness.

April 5, 5:00 PM, Room 2

The known luminance box is in the 224-256 relative range on the histogram. The bell curve is low and wide. The portion of the window seen beneath the blinds and table tops are at the highest intensity of 250 fL (See Fig. 6). The spike on the left side of the histogram corresponds to a large number of pixels at a low background level (See Fig. 7). This spike, which is to the left of the background curve is not a problem. Glare was caused as a result of direct light into the room and is corrected by letting down the blinds.

Another observation made here was on the significance of the absolute range of luminances. The absolute range of luminances in this room is lower than that in Room 1. The background levels are higher than those in Room 1. The resultant contrast between background and source is high, but not as high as in Room 1, and source area is limited.

October 2, 12:00 noon, Room 1

The known luminance box is in the 224-256 relative range on the histogram. The window and the known luminance box are at the highest intensity, corresponding to the spike at the right end of the histogram (See Fig. 8). The bell curve is tall and narrow, and the likelihood of causing glare is high. The predominant background level is low. The wall behind the monitors is graded into the high luminance of the window. The glare is not internal, in that there is no high contrast between surfaces within the room. The highest intensity is that of the window (See Fig. 9).

The glare would be caused because of the contrast between the window and the background. Hence an actual feeling of discomfort will depend on where the occupant performs his task and whether the window forms a part of the field of view.

October 2, 12:00 noon, Room 2

The known luminance box is in the 224-256 relative range on the histogram. The window, known luminance box and table surface are the highest intensity of 240 fL. The histogram shows pixels at almost all intensities, with a smooth bell curve and a spike at the right high intensity region (See Fig. 10). The general background level is higher and hence there is less possibility of the high intensity causing a glare. The wall behind the monitors also is graded in its intensity and this is confirmed by the presence of pixels in the range between the bell curve and the spike.

4.4 Conclusions

1. The method seems to be applicable from a numerical standpoint, by analyzing the histograms. The histogram is capable of establishing the background level or the adaptation level within the space, the percentage of field of view that the glare source and the background occupy, as well as the absolute values of intensities within the space and contrast of highest luminance with that of the background level.

2. The image makes it possible to define which surface has the highest intensity, and establish the range of absolute intensities within the space. It also gives a record of the window blind positions and any interaction of the occupant with the visual environment.

3. However there is no way to find out if the blinds were let down as a result of visual discomfort at that particular instance or were in that position since some other situation previously experienced. These observations need to be tested against a large number of individuals occupying a similar visual environment to determine their reaction to that situation.

4. The position of the camera is another critical factor in the method. These observations have been made with the entire space in the field of view. However there are other locations which are more interesting. The camera can be placed to mimic an occupant at his/her task location, either reading on a desk, looking at a computer screen etc. This information, although valuable, might prove to be difficult to obtain since in a real life situation the space is occupied by the occupant. This brings up the suggestion of setting up a test cell.

5. It is necessary to obtain occupant observations and evaluations in the form of surveys or questionnaires to back up the numerical analysis and conclusions. Since glare is such a subjective phenomena it is essential to reinforce numerical data with subjective impressions of which ratios and percentages of view cause glare, and which are acceptable.

6. The critical factor determined from the histograms is the ratio of the extreme intensity to the median of the background intensity. There are actual intensity levels exceeding 1:250 within the space, but the ratio of highest intensity to that of background intensity is more crucial in determining glare conditions. From the histograms it is found out that a ratio of 2:1 or greater between the peak and the median begins to feel uncomfortable. Any ratio of 3:1 or greater positively produces a sensation of discomfort and should be avoided.

7. This method is very useful for post occupancy building analysis. Any significant area can be examined and monitored along with the subjective impressions of the occupant to provide useful information in understanding the glare response of occupants.

8. It would also be useful in determining glare behavior in building energy simulations such as DOE than current algorithms.

9. To further analyze the glare response of occupants within the space, a test cell would have to be set up, where exact luminances corresponding to specific ratios of background to extreme can be produced and tested with occupants being surveyed while being exposed to those controlled visual conditions. These conditions can be videotaped and digitized for later numerical analysis. The occupants would be surveyed simultaneously as the test progresses. Specific ratios of background to extreme and field of view can be tested to find conditions of visual comfort and glare response.
 

5.0 ACKNOWLEDGMENTS

We would like to thank the following for their help in conducting this study.

Prof. Hofu Wu, Cal Poly Pomona, for arranging access and coordinating illuminance measurements at the Collins Center, Kelly Andereck and Gregg Ander at Southern California Edison for sponsoring work and video equipment on the Collins Center project, and Prof. Murray Milne, University of California, Los Angeles, for loaning the luminance meters.

6.0 REFERENCES

(1) Hopkinson, R. G., Evaluation of Glare, Illuminating Engineering, Vol. LII, June 1957, Pg. 305

(2) Hopkinson, R. G., Architectural Physics: Lighting, Her Majesty’s Stationery Office, 1963

(3) Hopkinson, R. G., Petherbridge, P., Longmore, J., Daylighting, Heinemann, London 1966

(4) DiLaura, David L., On the Computation of Visual Comfort Probability, Journal of the Illuminating Engineering Society, Vol. 5, July 1976, Pg. 207

(5) Subcom. on Direct Glare(1972), Preamble by Calculation Procedures Com.(1991), Computing Visual Comfort Ratings for Interior Lighting, RQQ Report No. 2 with the 1991 Preamble Outline of a Standard Procedure for Computing Visual Comfort Ratings for Interior Lighting, IES LM42, 1991

(6) Guth, S K, Computing Visual Comfort Ratings for a Specific Interior Lighting Installation, Illuminating Engineering, Vol. LXI, Pg. 634.

(7) Rea, Mark S., Toward a Model of Visual Performance: Foundations and Data, Journal of the Illuminating Engineering Society, Vol. 15, Summer 1986, Pg. 41

(8) Rea, Mark S., Toward a Model of Visual Performance: A Review of Methodologies, Journal of the Illuminating Engineering Society, Winter 1987

(9) Subcommittee on Guide for Measurement of Photometric Brightness of the Committee on Testing Procedures of the Illuminating Engineering Society, IES Guide for Measurement of Photometric Brightness (Luminance), Illuminating Engineering, Vol. LVI, July 1961, Pg. 457

(10) Walsh, J. W., Photometry, Constable, Lewin and Baker, London, 1958, 3rd Edition.

(11) Orfield, Steven J., Photometry and Luminance Distribution: Conventional Photometry Versus CapCalc, Lighting Design + Application, January 1990

(12) Rea, Mark S., Jeffrey, I. G., A New Luminance and Image Analysis System for Lighting and Vision, Journal of the Illuminating Engineering Society, Vol. 19, Winter 1990, Pg. 64

13. Schiler, Marc E., Report of Luminance Study Method, July 1994

ASES 1995