Lighting Calculations in the LED Era

Transcription

1 Lighting Calculations in the LED Era James R Benya, PE, FIES, FIALD Developed for Cree LED Lighting May 15, 2011 With Addendum May 22, 2011 Revision June 6, 2011 Final version June 30, 2011 Design Calculations with Solid State Lighting 1

2 Abstract For decades, lighting calculations have been the backbone of lighting designs for almost all lighting applications, indoors and out. Good practice in illuminating engineering dictates that the designer uses photometric data that is adjusted to meet the conditions of the intended application by accounting for temperature, dirt, and variations of components. With LED lighting systems, the photometric test method ( absolute photometry ) differs from conventional source lighting ( relative photometry ). In order to properly predict the performance of an LED lighting system and to compare results to competing conventional lighting systems, designers must be particularly careful to use appropriate factors in addition to properly using the two different photometric data formats. A discussion of each of the factors and examples of comparative calculations are provided. Introduction Calculations to predict lighting system performance are fundamental to the practice of lighting design and illuminating engineering. These calculations allow one to predict lighting performance and whether the design meets current performance recommendations. Practitioners use both simple and advanced methods, but in either case, the input must include all of the proper data and adjustments relative to the project. Until the advent of solid-state lighting (SSL or LED), lighting calculations evolved with the presumption that components could vary within the luminaire. Luminaire photometric testing is performed in a laboratory using a reference lamp and ballast, and then factors are applied to compensate for the actual lamps, ballasts and physical conditions of the design. These factors, called light loss factors (LLF), are scalar multipliers that account for differences in performance between the laboratory and field. There are two types of factors. Non-recoverable light loss factors are differences inherent to the lamp, ballast, room surfaces and thermal environment because the differences are always evident. Recoverable light loss factors are differences caused by lamp aging and effects of atmosphere, dirt, location, and other degrading environmental factors that can be recovered with new lamps and a good cleaning of the luminaire and room. Solid-state lighting is the first generation of lighting equipment in which user-replaceable or interchangeable lamps are not desirable. Not only does very long life virtually eliminate the need for relamping, solid-state light sources require very specific mounting to meet precise thermal and optical requirements. Compared to traditional luminaires, solid-state lighting luminaires are better seen as complete assemblies. Both photometric testing and application calculations must be done differently. Many emerging solid-state lighting systems have important differences relative to conventional lighting and to fairly compare competing technologies, apples-to-apples comparisons are needed. The use of photometrics and factors is different with solid-state lighting and their proper use in achieving proper calculation results is a key reason for this paper. 2 Design Calculations with Solid State Lighting

3 Calculations in Lighting Design Introduction General Concepts When designing lighting installations, one of the most significant requirements is to provide an appropriate quantity of light. The design target quantity is typically chosen from an IES publication such as the IES Lighting Handbook or one of IES Recommended Practices. The designer who may be an architect, engineer, lighting designer or specialist, or one of many other roles uses calculations to confirm that the design meets the desired quantity without being significantly above or below the target. Effective calculations can be simple, but many prefer computer calculations. In either case, the designer must choose appropriate photometric data and then must adjust it to suit project conditions. Photometric Data Photometric data for each luminaire type is determined in a test laboratory. There are independent laboratories as well as in-house laboratories in larger lighting companies. To make a test, the luminaire is mounted in a black room and light meter readings are taken at points at all angles about the luminaire. Each data point represents the candlepower of the luminaire at a specific angle. The result is a data file in a specific IES format. A complete test report includes all of the data and several summaries and derived data charts, including a Coefficient of Utilization (CU) table. Photometric data is typically obtained from the luminaire manufacturer s website. Free software i can be used to view the data and to produce reports such as CU tables. Manufacturers may also produce printed photometric reports and it is common to have the CU table for the luminaire on the catalog cut sheet. Lumen Method Calculations The lumen method is a hand calculation method for predicting the performance of a general lighting system providing reasonably uniform illumination. It can also be used by an experienced designer for non-uniform lighting, although the results could be misleading. The method entails four steps: For a given space, determine the room cavity ratio (RCR) and reflectance of the surfaces of the space. Choose a lighting system (luminaire) and obtain photometric reports including a CU table. Determine the CU of the lighting system from RCR and reflectance. Determine the applicable lumen adjustment and light loss factors (see below). Solve the equation for either average illumination (footcandle or lux) level for a given number of luminaires, or for the number of luminaires needed to meet a specific illumination level. The lumen method is convenient to use, requiring a basic calculator and a CU table, which is small and simple enough to be printed on a luminaire catalog sheet. The CU table is derived from a full photometric test report, and it is acceptably accurate for predicting the average illumination in a room. However, this method cannot predict specific illumination levels nor the variance of light levels in the space. Radiosity and Ray Tracing Calculations Computer calculations for predicting lighting system performance are significantly more accurate than the lumen method. In general, computer methods predict the illumination at specific points in the space ( point-by-point illumination) permitting a detailed evaluation of lighting performance. Design Calculations with Solid State Lighting 3

4 Most software programs offer a number of other calculation types and features, including a perspective rendering of the illuminated space. There are two primary types of calculation methods. Radiosity is a fast calculation method that assumes all room surfaces have a matte ( lambertian ) finish. This assumption permits the computation using radiative transfer functions from surface to surface. In a simple room, a lighting system performance can be computed in seconds, and a rudimentary rendering can be generated in under a minute using a typical Windows XP computer. Raytracing is a comparatively slow calculation method in which a large number of rays of light are traced from the light source, reflection by reflection, until diminished. Raytracing takes into account both the matte and specular reflections of every surface, and for each reflection to be carefully followed according to the surface from which it came. In a simple room, raytracing can produce acceptable results in a few minutes, but exceptionally nice images renderings can take hours of computational time on a typical Windows XP computer. Some software programs employ radiosity for speedy calculations of most of the lighting effects and then perform a raytracing layer in order to create more realistic images. In either case, input to the program consists of a complete description of the space in three dimensions complete with furniture, ceilings and walls. For each luminaire, its location is specified as well as its photometric aiming. Each luminaire s characteristic photometric report is part of the program input. The lighting program allows photometric adjustment factors for each luminaire type. Lighting software requires special training and experience. With modest training, calculations of rectangular box-shaped spaces can be very quick. Conversely, allowing for enough time, computer calculations can produce impressive reports and renderings. Conventional Practice Relative photometry When making a photometric test for conventional lighting systems, a reference lamp is placed in each socket and, with the exception of incandescent lamps, the lamp socket(s) are rewired to a reference ballast. The reference lamp(s) and reference ballast(s) are operated at specific temperature and voltage, such that the results of the test are relative to reference components. Because of the calibration of the lamp and ballast, detailed information about the luminaire can be derived from the data. For example, it is possible to calculate the efficiency of the luminaire by dividing the measured light output by the rated lamp lumens. The primary reason for relative testing is to permit the interchange of lamps and ballasts with different output but don t change the way light is emitted by the luminaire. In other words, all of the candlepower values can be multiplied by a single value that represents the ratio of the real lamp and ballast to the reference lamp and ballast. For instance, assume a photometric report prepared for a luminaire using 2900 lumen T-5 lamps. However, high performance 3125 lumen lamps are to be used. All of the candlepower values are multiplied by (3125/2900) or Note that the rated initial lamp lumens from the catalog are used in calculations. Using photometric reports For hand calculations, a photometric report with CU table is needed, and can either be found on the product cut sheet or determined using free photometric viewing software. For computer calculations and to use the photometric viewing software, obtain the photometric data file from the manufacturer s website. Note that for one luminaire model, the manufacturer may have a number of photometric tests involving different options that affect light distribution such as different lenses or different numbers of lamps. 4 Design Calculations with Solid State Lighting

5 Simplified calculation theory Whether hand calculated or done on a computer, lighting calculations result in quantities (footcandles, lux, nits, etc.) that are proportionate to the amount of light being emitted from the luminaires. If the quantity of emitted light changes, so does the measured lighting performance in a proportionate amount. But the pattern of light does not change unless there are multiple lighting systems and they are not all changed to the same amount. The light output of a luminaire can be higher or lower than the reference measurements in the laboratory. The light output in the test report is multiplied by light loss factors. A light loss factor (LLF) is a number that multiplies the candlepower or lumen values in the calculation. For instance, many designers use a generic LLF of 0.75, which means that the luminaire is emitting 75% of its originally tested light output. It is possible for LLF to be >1, which means that the luminaire is emitted more light than it s originally tested light output. A factor having no impact is 1.0. Most factors, however, are less than 1, indicating a loss of light compared to the originally tested light output. Light loss factors multiply each other. For instance, if a luminaire has three light loss factors of 90% (10% loss each), the total impact is 90% x 90% x 90% (0.9 x 0.9 x 0.9) =.729 or a combined loss of 27.1%. Light loss factors do not add. Light Loss Factors Light Loss Factors (LLF) are somewhat misnamed. They were once assumed to be factors that reduced the amount of light when comparing the tested luminaire with the actual luminaire in the application, but today, due to many technical advances, it is possible to have more light in the application. Nonetheless, they are still called light loss factors. They are broken down into two primary groups, non-recoverable LLF and recoverable LLF. Non-recoverable LLF are permanent losses (or gains) that can only be changed by changing the component or application, such as using a different ballast. Recoverable LLF are losses that can be recouped by using a fresh lamp or cleaning dirt off the lamp, luminaire and/or interior surfaces. In calculations, there are initial light level calculations and maintained light level calculations. In general, initial light level calculations only include non-recoverable LLF, whereas maintained light level calculations include recoverable and non-recoverable LLF. Non-Recoverable Light Loss Factors Non-recoverable LLF are factors that adjust the photometric data to compensate for the use of specific lamps and ballasts. They also compensate for other persistent factors that affect light output from the beginning of operation of the lighting system. Non-recoverable LLF include: Ballast Factor (BF) This factor accounts for the difference between the reference ballast and an actual ballast product to be used in the field. The standard T- 8 electronic ballast has a ballast factor of 87% (0.87), which means that the ballast will cause the lamp to produce 87% of its rated output. This is an important and commonly used factor. The ballast factor is generally given in the ballast catalog or cut sheet. With magnetic ballasts, a ballast factor is often used to represent manufacturing tolerance rather than designed light output. Design Calculations with Solid State Lighting 5

6 Ballast-Lamp photometric factor This factor accounts for the mismatch between ballasts and lamps relative to temperature. It is an uncommonly used factor as the other factors listed here are more often used. Thermal Application Factor (TAF) This factor accounts for the impact of temperature on light output. Temperature can be a major consideration for parking garage and outdoor lighting and for many indoor spaces where ambient temperatures are very low (like refrigerated storage or cases) or very high (like unconditioned warehouse or industrial spaces). You can estimate the thermal application factor from the lamp s temperature curve and an educated estimate of ambient air temperature. Note that TAF can be tricky. Take, for instance, a T-8 lamp troffer. The luminaire is tested with the air around it at 25 C (77 F) but in reality, the temperature around the lamp is probably about 35 C. But we don t care, because the photometric data already accounts for the actual lamp temperature in the fixture, and the TAF is 1.0. But place the luminaire in a refrigerated room with an air temperature of 57 F (15 C) and what happens? You might guess that the lamp operates in ambient air of around 25 C, so that you will get about 10% more light. A careful designer might choose a TAF of about 1.1 in this case. The situation is complicated by T5 lamp technology. Since the T5 lamp is designed to operate at peak with ambient air at 35 C, many luminaires have been designed to operate as close to this point as possible. But in the example above, the proper TAF for a cooler room would be a 10% drop in light, or TAF = Voltage to Luminaire Factor This factor is more applicable to incandescent lamps and magnetic ballasts. A voltage drop of 2.5% to a 120-volt lamp results in nearly 9% light loss. It could be further reduced if a dimmer, even operating at full output, is part of the circuit. However, with electronic ballasts, this is no longer an issue for non-incandescent sources. Heat extraction thermal factor This factor was intended to address the beneficial effect of drawing return air through a lensed fixture, dropping lamp temperature. Because of the diminished use of heat extraction troffers due to lower lighting power density and fewer lamps, this factor is generally rolled into the Application Thermal Factor, above. Equipment operating factor The power and light relationship of HID lamps is complex. This is one of two factors that accommodate the relationship between power and light. This factor is generally used to differentiate between rated lumens with horizontal or vertical arc tubes, before the fixture is tilted. The rated lumens in both positions are generally presented in the lamp catalog. Lamp position (tilt) factor This factor is to further compensate for the tilting of HID lamps. This is a harder figure to determine and must be estimated from manufacturer s data and the estimated aiming angle of the luminaire. Luminaire surface depreciation factor This factor is reserved for permanent luminaire deterioration that can t be recovered by cleaning. This includes the deterioration of paints, plastics and metals. The better materials used today make this factor of minimum interest unless the luminaire is exposed to an atmosphere that corrodes or abrades the optical surfaces. But with increased lamp life ratings of fluorescent lamps, this factor may become more relevant. For most common lighting applications today, the ballast factor (BF) and the application thermal factor (TAF) are the most prevalent. 6 Design Calculations with Solid State Lighting

7 Recoverable Light Loss Factors Recoverable light loss factors represent loss of light due to lamp age and the accumulation of dirt. Light levels can be recovered back to initial conditions when the luminaire is cleaned and relamped. Thus, recoverable light loss factors are only applied in either mean or maintained light level calculations. Mean light levels are the expected light level when lamps have reached 40% of rated life. Maintained light levels are reached when lamps have reached their replacement time. There are two types of lamp replacement practice, spot replacement and group replacement. Spot replacement is becoming more common as the reduced number of lamps per unit area makes individual tasks more susceptible to being in the dark when a lamp fails. However, group replacement is far more cost effective. Because lamp life with conventional lamps is based on mortality and not on light loss, maintenance becomes increasingly necessary as lamp life is approached. In conventional lighting systems, the ideal replacement time is determined from rated lamp life. The best time for group replacement in commercial and industrial buildings is between 60 and 80% of rated life ii as this will result in the fewest burnouts (about 5-15%) and is the considered the economically best time. But care is needed; lamp life is directly related to the ballast selection and operating application. The method of starting the lamp is determined by the specific ballast circuit of which there are three primary methods: Preheat ( program ) start Rapid start Instant Start The cathode is pre-heated for a brief period (up to 1 sec) followed by a voltage pulse to commence the main arc. The cathodes are continuously heated. Arc is started by a voltage pulse. Cathodes not heated. Arc is started by a high voltage pulse. Longest lamp life when often switched Long lamp life, dimmable Instant response; least temperature sensitive. Most energy efficient. Least expensive. Shortest lamp life when switched often; dimmable over a small range When activated, there is a slight delayed action. Adds to the cost of the ballast. Ballasts may be slightly less energy efficient. Not dimmable. When activated, there is a slight delayed action. Least efficient. Latest versions have heater cutout circuits to improve efficiency. Lamp life suffers dramatically when frequently switched. The life of fluorescent lamps is now carefully presented as a function of ballast type and average operating period per start. For example, the life of a premium lamp with an instant start ballast might be 35,000 hours at 12 hours per start but only 18,000 hours at 3 hours per start. The same lamp might only survive 7,500 hours at 45 minutes per start on this ballast. Hence the lamp life being used must first be carefully determined from lamp, ballast and operating situation data. The cleaning period of the luminaires is a separate issue and is determined by the space type and maintenance period. With modern long life lamps, it should be shorter than the relamping cycle. Design Calculations with Solid State Lighting 7

8 Following are the principal recoverable light loss factors: Lamp Lumen Depreciation (LLD) Lamp lumen depreciation is unique to each lamp type and can differ even among similar lamps. In the graph, the lumen depreciation curves for three lamps are shown. Note that T5 fluorescent lamps have a 92% lamp lumen depreciation factor (loss of only 8%) for mean life (8,000 hours), replacement life (14,000 hours) and rated life (20,000 hours). Luminaire Dirt Depreciation (LDD) This factor accounts for accumulation of dirt on the lens, lamp and reflecting surfaces. It can be determined using a relatively sophisticated procedure in IES publications. However, with the cessation of smoking indoors, the accumulation of dirt has dropped dramatically. Except for very dirty environments and applications, this factor should be around Room Surface Dirt Depreciation (RSDD) As with LDD, the change of laws involving indoor smoking has resulted in this factor being very minor except in particularly dirty spaces for other reasons. It is common to leave it out for most spaces now. Lamp burnout factor (LBF) This factor accounts for the percentage of unreplaced burnouts at a point in time, usually just before replacement of lamps. This is often seen as a small factor, and is typically ignored. But with the aggressive lamp life ratings promoted by state of the art fluorescent lamps, their mortality curves need to be carefully evaluated relative to the ballast and average operating period. A value of 95% (5 % burnout) is recommended for 70% lamp life replacement and 85% for 80% lamp life replacement, with the specific replacement time being set by the lamp and starting method. For most common applications today, the lamp lumen depreciation factor is significant, the luminaire and room surface depreciation factors less so. The burnout factor, however, is increasingly significant due to the lamp life issues that can easily occur with motion sensors and instant start ballasts. Special Note: T5 lamps The T5 lamp is designed to work best at 35 C, but reference lamp photometry is taken at 25 C. Yet, many luminaires actually contain lamp compartments that raise air temperature of the lamp. In these cases, the photometry is adjusted so that the total luminaire efficiency is increased. Therefore, for correct calculations, use the 25 C lamp lumen values from the catalog, not the rated lumens at 35 C. Calculations Involving Solid State Lighting Absolute Photometry When making a photometric test for solid-state lighting systems, the luminaire to be tested is mounted into the testing equipment. The testing equipment is calibrated to a reference source. The luminaire is operated at specific temperature and voltage and the measurements represent the absolute and actual performance of the luminaire. No effort is made to determine how much light each LED emits. In other words, the luminaire is assumed to be 100% efficient. Do not attempt to 8 Design Calculations with Solid State Lighting

9 use individual LED lumen ratings in calculations the proper value is the total lumens in the photometric report. Using photometric reports Photometric reports are used in just the same way as for conventional luminaires. These include the typical candlepower tables and CU tables. Simplified calculation theory The calculation theory is similar to that of conventional luminaires. The primary difference, as will be illustrated below, is that the light loss factors are a lot different. Light Loss Factors Solid-state lighting is measured as a fixed lamp and driver system. This eliminates the usefulness of many of the classic light loss factors related to changing the lamp or ballast. Instead, there are specific factors by which solid-state lighting is better addressed. Non-Recoverable Light Loss Factors For solid-state lighting, the non-recoverable light loss factors are: Thermal Application Factor The light emission from a solid-state luminaire is relatively constant over a wide temperature range, as light is generally proportionate to device current (milliamps). However, in good designs, device junction temperature is measured and if the device begins to get too hot, the current is reduced, reducing light output but preserving lamp life. But the reduction is not necessarily linear. It is necessary to assume the worst case, but to determine the applicable factor; the manufacturer should provide the appropriate curve. Note: LED rated lamp life is directly affected by temperature. Design lamp life can be plus or minus; it may exceed rated lamp life when thermal advantages are realized. Luminaire surface depreciation factor As with current fluorescent lamps, if the luminaire is to last as long as promised, the perseverance of the luminaire surfaces could become an issue. However, no data or studies are available to help set a value. Recoverable Light Loss Factors For solid-state lighting, the recoverable light loss factors are similar to conventional lighting. They include: Lamp Lumen Depreciation With solid-state lighting, the most significant light loss factor is lumen depreciation. It is generally agreed that LED sources depreciate, and the current rating system, based on accelerated aging tests and other factors, is that solid state lighting s rated life is the point at which the lumen depreciation is 30% (LLD=0.70). If comparing solid-state lighting with T8 or T5 fluorescent lamps at mean life, the approximate depreciation is about 12.5% (mean lumens = 87.5% of initial). If relamping is to occur earlier, say 80% of rated life, the LLD would be about 80%. Luminaire Dirt Depreciation This factor is virtually the same for any type of lighting. Room Surface Dirt Depreciation This factor is also the same for any type of lighting. Design Calculations with Solid State Lighting 9

10 Burnout Factor LED must be seen as a lamp whose actual mortality occurs long after the rated life. Moreover, the rated life is not affected by control type, such that any LED lighting system can be operated on any controls and still enjoy rated life. Thus, a burnout factor of 1.0 can be used and the lamp life will be as rated or possibly even longer, subject to changes in life caused by temperature (see above). This is a particularly interesting issue that is especially impactful when comparing system economics (below). 10 Design Calculations with Solid State Lighting

11 Comparing an LED Lighting System to Conventional Fluorescent and Compact Fluorescent Systems Project Description Solid-state lighting has rapidly advanced from curiosity to serious commercial and industrial lighting. The problem is that manufacturers make claims that test common sense. The best way to determine the accuracy of claims is photometric testing and analysis. Office, retail and school spaces are an excellent situation for comparative analysis. In today s market, solid-state lighting must compete with a variety of common fluorescent troffer and downlighting systems. These systems are ubiquitous, comprising by type the majority of non-residential lighting systems in North America. Any improvements in these luminaire types will be profound and serve as one of the biggest weapons to combat energy use. For ease of comparison, calculations are performed using the lumen method. The comparison space for troffers is a single room, 33 feet 4 inches by 30 feet (1,000 sf) with a 9-foot ceiling. This space is the most typical classroom size and is also representative of open office areas. A second comparison space for troffers is a room 12 feet 6 inches by 8 feet (100 sf), representative of private offices and small conference rooms. The room cavity ratio (RCR) of the former is about 2.0 and the latter, about 6.5. In both cases, the design level is 40 footcandles average maintained, a value I typically use because it provides over 50 fc in the center of the room thus accommodating task E in much of the space. It also ensures 30 footcandles in spaces with systems furniture and a furniture factor of 75%. The final comparison space is a hallway 200 feet long and 5 feet wide (1000 sf) with a 9 ceiling, representative of office and classroom corridors. The RCR is also about 6.5. Light level for design is 15 footcandles maintained. All spaces have 80/50/20 reflectance. The maintenance cleaning period is assumed to be two years and a 0.92 combined light loss due to dirt (RSDD x LDD). Relamping is assumed to be 70% of rated lamp life for fluorescent lighting and 80% of rated lamp life for solid state lighting. Troffers We propose three systems for comparison: A classic 2x4 lens troffer with 2-T8 lamps iii. Each fixture employs (2) generic 4100K 80 CRI T8 lamps and a high performance program start ballast. The CU is RCR=2 and RCR=6.5. The generic lamp is 2950 lumens with LLD ~.87 and the input power is 64 BF = 1.0 and Lamp burnout factor is A state of the art 2x4 troffer with 2-T5 lamps iv. Each fixture employs (2) state of the art T5 lamps at 3050 lumens and a high performance program start electronic ballast. The CU is RCR =2 and RCR = 6.5. The high performance lamps is C) lumens with LLD ~.92 and the input power is BF =.95 and BF = Lamp burnout factor is A state of the art LED troffer v rated at watts. The CU is RCR =2 and RCR = 6.5. The LLD at 70% of rated life is about 0.79 with burnout factor of 1. The results of basic lumen method calculations are presented in Table 1. Design Calculations with Solid State Lighting 1 1

12 System Lamps Lumens BF LLD LDD RSDD Burnout CU Luminaires Choose Watts Total watts W/sf FC final Classroom T8 lens T5 HP LED HP Office T8 lens T5 HP LED HP Table 1 Comparison of Fluorescent and LED Troffer Lighting Systems in 1000 sf classroom and 100 sf office. 12 Design Calculations with Solid State Lighting

13 What is particularly impressive in this comparison is that at the time of first relamping either fluorescent system (about 14,000-22,000 hours depending on the lamp) the lumen depreciation of the LED system will only be about 8-10%, which means that the light level in the classroom will still be about footcandles, and in the office, about footcandles. The relamping of the LED system will occur at about 40,000 hours of operation, which will be up to 20 years after installation. Downlights For most commercial lighting involving downlights that are switched, dimmed and/or used for emergency lighting, luminaires with compact fluorescent lamps have been used since the 1980 s. But the compact fluorescent lamp downlight has historically been problematic with many problems including lamp overheating, poor dimming quality and short lamp life. The LED now poses a rational option. Downlights are typically used in hallways, lobbies and other spaces with high room cavity ratios. The test case is a long corridor as might be found in a modern office building. The RCR is 6.5, and the design level is 15 footcandles. A popular compact fluorescent lamp downlight vi is compared to a state of the art LED downlight with a deep regress. Both have shielding between 45 and 50 making them appropriate for commercial applications. A combined dirt depreciation factor of.95 is assumed. The compact fluorescent uses a 26-watt compact fluorescent lamp rated 1800 lumens. A nondimming ballast operates at 28 watts at BF = 1.0. Assuming no unusual temperature issues, the ATF = 1.0 and the CU at RCR 6.5 = The LLD for a compact fluorescent is typically about The LED luminaire is rated 1019 lumens with a RCR of 6.5 = Its input power is 12.5 watts. Because the results so favor the solid-state luminaire, an alternative design using a high efficiency 2x2 with a single 28 watt CFT40 lamp was offered. The results are as follows for the 1000 sf corridor. System Lamps Lumens BF LLD LDD RSDD CU Luminaires Choose Watts Corridor Total watts CF LED T5 2x W/sf FC final Table 2 Comparison of LED downlights to compact fluorescent lighting systems in corridor lighting systems. The maintenance cycle of the compact fluorescent lamp can be as short as 10,000 hours, compared to 40,000 hours or more with the solid state lighting. As LED continues to increase in lumens per watt, it is foreseen that a 1350 lumen LED could easily be implemented to create equal luminaire quantities to the 26-watt compact fluorescent with equivalent or better savings than indicated above Likewise, a 2000 lumen LED downlight could be used to reduce the number of luminaires by about 50%. However, designers may also consider an LED 1 x 1 or 2 x 2 luminaire to increase source area and reduce the potential for glare. Design Calculations with Solid State Lighting 1 3

14 Related Considerations in Lighting Design Color Considerations Choosing interior architectural lighting requires the consideration of color temperature and color rendering quality. With modern light sources including LED and fluorescent, these are choices to be made because color options are possible. Correlated Color Temperature The correlated color temperature (CCT) of residential lighting is generally in the range of 2700K to 3000K, but for commercial lighting the more common choices are 3500K and 4100K. This is explained by the Kruithof Curve, a chart representing color temperature preference as a function of light level. For most commercial lighting situations, 3500K has been preferred because of the light color favors human complexion. Color temperature preference for commercial lighting lux Figure 1 Commercial Lighting Levels lux and Kruithof Curve (Wikipedia) Color Rendering Index The color rendering index (CRI) generally describes the color quality of the lighting, with a best quality at CRI = 100. Most fluorescent systems are between 80 and 85; the latest LED sources are now providing CRI = 90, an important improvement for a number of applications. Spectral Effects on Visibility Research at the Lawrence Berkeley National Laboratories (LBNL) has demonstrated improved visibility of tasks at high color temperatures vii. Berman believes that this is a result of changes to the response of the eye s optical pupil, which reacts (through the endocrine system) to short wavelengths and constricts, reducing visual noise and improving depth of field. IES does not acknowledge this theory as a general method to reduce recommended light levels viii, but the phenomenon may be successfully used for many tasks. Control Considerations Starting and Restriking LED and fluorescent lamps, as well as tungsten lamps, generally start immediately or within 1 second and will restart immediately as well. Some fluorescent lamps require a warm up period, but it is generally fairly rapid and full light is reached within a minute or two. HID lamps, however, require a warm up period of several minutes and if extinguished, require a cooling off and restarting ( restrike ) period of several minutes as well. Among state of the art lamps, LED exhibit the most ideal characteristics. Dimming All lamps can be dimmed to operate at lower light levels. HID lamps are the worst, having the poorest range and experiencing undesirable color shift. Fluorescent has a range dependent on the ballast with a bottom level of less than 1%, with the risk of some color shift particularly in compact fluorescent lamps. LED lamps can be dimmed smoothly to zero light with little or no color shift, a superior characteristic. To dim lamps, dimming ballasts or drivers (see below) must be used. 14 Design Calculations with Solid State Lighting

15 Emergency Operations Emergency lighting typically involves a lesser light level than normal with instant starting and restriking characteristics. Both fluorescent and LED sources are acceptable emergency and emergency/normal light sources permitting lower than normal power draw that could work well with a battery pack inverter unit. Energy Considerations Ballast and Driver Losses Most modern light sources have auxiliary devices that (a) convert incoming AC power and (b) regulate current flow. For a fluorescent lamp, the device is called a ballast and for an LED, A driver. Both drivers and ballasts exhibit power loss that reduces the luminous efficacy (lumens per watt) of the system (see below). Presently, some fluorescent ballasts may be slightly more efficient than comparable LED drivers, but the differences are small. System Efficacy The system efficacy of any conventional lighting system is defined as the product of the source lumens, the ballast factor, and the luminaire efficiency, all divided by the input watts. With LED (because of absolute photometry) the system efficacy is determined by measuring the total luminaire lumens by input watts. System efficacy is indicative of overall efficiency but it neglects room geometry, so it must be carefully used out of context. For instance, in the example above, the compact fluorescent downlight is (1800 x 60.9/28 = LPW) and the LED downlight is (1019/12.5 = 81.5 LPW). Pilot power draw Ballasts and drivers with direct line voltage switching have no pilot power draw. But some drivers and ballasts that employ electronic network control can draw pilot power when not on. Pilot power is drawn 8760 hours per year, so even 0.1 watt per luminaire is almost 1 kwh per year in lost energy, roughly the equivalent of 1 week per year of operating the fluorescent lamp 24/7 and two weeks per year of constant operation of the LED. Dimming power behavior Dimming behavior may be critical in some applications. Among high efficacy light sources, only LED has been demonstrated as fade to black, but only with specific drivers and sources. All sources become less efficacious as they dim because of increased ballast or driver losses relative to source energy use. Design Calculations with Solid State Lighting 1 5

16 Cost Benefits of Solid State Lighting For almost three decades, the cost analysis of competing lighting systems involved the using the energy cost savings of a more efficient lighting system to offset the first cost of a standard base case. There is an important difference with solid-state lighting; in addition to energy cost savings, sold state lighting requires less maintenance, is more immune to control system impacts on maintenance cycles, and continues to operate well beyond the rated end of life. These benefits have economic significance in ways previously ignored because of the similarity among fluorescent lamp and ballast systems. Energy Savings Solid-state lighting can save considerable energy relative to even the most efficient current lighting sources when properly applied and when using state of the art technology. This paper is based on new technology released in 2011 that is among the first, if not the first, to make a credible and cost effective argument for solid state lighting when competing against fluorescent in everyday applications. Simple Payback Analysis In simple payback analysis, the savings are used to perform a straight line amortization of the cost difference. In the case of the troffer analyzed here, an estimated cost difference of $50 per luminaire can be amortized by about 14 watts power difference per luminaire in about 8 years in a typical office or school application paying an average of 15 per kwh. Life Cycle Costing Life cycle costing has the potential for being an even stronger advocate of solid-state lighting. By eliminating relamping for the life of the luminaire, considerable costs are eliminated. Moreover, with expected dramatic increases in energy cost, the energy savings will be magnified. Summary Solid-state lighting has been long regarded as the future king of efficient lighting. This paper indicates that that foretold future is now here for one of the harder applications of solid state lighting, the general lighting of common commercial spaces. i Acuity Brands Lighting offers a program called Photometric Viewer for free download. ii IES Lighting Handbook, 9th Edition, Chapter 28 Planned Maintenance Activities suggests 70-80%; Staying on Schedule, Craig DiLouie, Electrical Contractor, June 2008 and IES RP both suggest 60-70%. iii Lithonia 2GT8 iv Lithonia 2RT5 v Cree CR24-50 vi Gotham AFV 26TRT 6AR vii Berman, Sam, New Discoveries in Vision Affect Lighting Practice, undated, c viii IES PS Design Calculations with Solid State Lighting