A FUNDAMENTAL STUDY ON CURRENT INDUSTRIAL LIGHTING SYSTEMS AND ANALYSES OF RETROFIT OPTIONS by MATTHEW BLAIR CLINTON A THESIS Submitted in partial fulfillment of the requirements for the degree of Master of Science in the Department of Mechanical Engineering in the Graduate School of The University of Alabama TUSCALOOSA, ALABAMA 2008 Copyright Matthew Blair Clinton 2008 ALL RIGHTS RESERVED CONTENTS ACKNOWLEDGEMENTS ............................................................................................... iii LIST OF TABLES ............................................................................................................ vii LIST OF FIGURES ......................................................................................................... viii ABSTRACT ........................................................................................................................ x CHAPTER 1: INTRODUCTION ...................................................................................... 1 CHAPTER 2: HIGH-INTENSITY DISCHARGE LAMPS .............................................. 5 2.1 HID Lamp Construction ....................................................................................... 6 2.2 HID Ballasts ........................................................................................................ 11 2.3 HID Evaluating Criteria ...................................................................................... 12 2.4 HID Technological Advancements ..................................................................... 16 CHAPTER 3: LINEAR FLUORESCENT LAMPS ........................................................ 29 3.1 Linear Fluorescent Lamp Construction .............................................................. 29 3.2 Linear Fluorescent Lamps Types ........................................................................ 33 3.3 Fluorescent Ballasts ............................................................................................ 38 CHAPTER 4: FUTURE ALTERNATIVES .................................................................... 43 4.1 Induction Lighting .............................................................................................. 43 4.2 Light-Emitting Diodes ........................................................................................ 45 CHAPTER 5: LIGHTING CONTROLS ......................................................................... 49 5.1 Occupancy Sensors ............................................................................................. 49 5.2 Photocells ............................................................................................................ 53 v CHAPTER 2 HIGH-INTENSITY DISCHARGE LAMPS High intensity discharge lamps are commonly used in both high- and low-bay applications to provide, powerful, efficient, high-quality light. Suffice it to say, HID lamps have long been the standard for industrial facilities, and continue to be the lighting workhorse for the illumination of manufacturing plants around the country. Their compact design, attractive color characteristics, and high lumen efficacy make HID lamps an appealing option for industry applications. Three primary lamp types comprise the family of industrial HID lamps. These lamp types include mercury vapor, high-pressure sodium, and metal halide. Each lamp utilizes a unique set of materials, namely metals, to produce light, but all three adhere to a common form of operation. Table 1 compares each lamp according to several features [2]. Mercury vapor lamps will not be discussed because they have long become obsolete and, for the most part, are not used in manufacturing facilities. 5 6 Table 1. HID Lamp Comparison HID Lamp Comparison Features Mercury Probe Metal Halide High Pressure Sodium Color White White Yellow Efficacy (L/W) 50 60 – 85 100 – 125 Lumen Maintenance 50% 65% 90% Lamp Wattages 50 – 1000 175 – 1000 35 – 1000 Lamp Life (kHrs) 24+ 6 – 20 24+ Source: Taken from Ref. 2. 2.1 HID Lamp Construction There are two main parts to a HID lamp: an outer bulb, and an arc tube. Light is formed in all HID lamps through the discharge of an electric arc inside the arc tube. Contained inside the arc tube are two tungsten electrodes, a starting gas, and arc metal. The starting gas, either argon xenon, or a mixture of neon, argon and xenon, will facilitate starting the lamp at low pressure and ambient temperatures [1]. The arc metal consists of metals and halide components of metals. An electric field is passed between the starting electrode and the main electrode which allows the starting gas to volatize the arc metal. The subsequent decrease in resistance between the main electrodes caused by the volatized arc metal creates an arc in the tube [1]. As the arc metal evaporates into the arc discharge, different metals will release distinct lines of radiant energy [1]. Because of this, each type of HID lamp produces light characteristic of the metals contained in the arc. 7 The outer bulb, made of either a soft or hard glass, encompasses the arc tube, providing it protection from the environment. The outer tube also functions to filter out unnatural light and short wavelength ultraviolet energy, absorbing a majority of the unwanted radiation while allowing the visible light to pass through [1]. The inside of this outer bulb can also be coated with a diffusing material to reduce brightness or a layer of phosphors which absorbs the UV radiation to improve color-rendering properties. Between the arc tube and the outer bulb, a low-pressure gas or vacuum is present. Fig. 1 depicts the basic structure of an HID lamp [21]. Fig. 1. A basic HID lamp [Taken from Ref. 21]. 8 2.1.1 Metal Halide Metal halide lamps were initially created in the 1960’s for use in industrial settings to provide a compact, efficient, and intense source of light [1]. They have now replaced the much outdated mercury vapor lamp, and are available in a wide array of sizes and configurations to meet the varying demands present in commercial, residential, and industrial applications. They are commonly referred to as a “point” source of light, operating under high temperature and pressure. Because of the blue light they emit, metal halide lamps are often the most ideal option for providing a primary source of light in situations where no natural light is available. Most manufacturers offer metal halide lamps in low, mid-range, and high wattages. Metal halides are most commonly seen in wattages of: 175-W, 200-W, 225-W, 300-W, 320-W, 350-W, 360-W, and 400-W [1]. A metal halide lamp is constructed in a similar manner to its predecessor, the mercury vapor lamp, but boasts some significant improvements. Where light is produced through the vaporization of mercury in a mercury vapor lamp, a process that continues until the amount of mercury contained in the arc tube is completely vaporized and depleted, a metal halide lamp utilizes a regenerative cycle to maintain light output [1]. In a metal halide lamp, a compact arc tube contains a mixture of gases, commonly a highpressure mixture of argon, mercury and a variety of metal halides. As voltage is applied to the electrodes and an arc is formed, the argon gas is easily ionized, sustaining the arc and producing intense heat to vaporize the mercury and metal halides contained within the tube. As the temperature and pressure increase inside the tube, and the halide vapors converge on the intense central core of the discharge, the halides separate from the metals allowing the metals to produce a bluish-white light. After the metal atoms have radiated 9 their spectrum, they return to the cooler arc tube wall to recombine with halogen atoms, repeating the cycle [1]. Fig. 2, depicting normal metal halide lamp construction, is displayed below [1]. Fig. 2. Metal halide lamp construction [Taken from Ref. 1]. 2.1.2 High-pressure Sodium High-pressure sodium lamps are one of the most efficacious light sources available, with the ability to attain efficacies of nearly 150 lumens per watt. They are commonly used outdoors because of their unmatched efficacy, long life, and poor color rendition. In industrial applications they are most often seen in warehouses or low task areas where minimal color contrast is sufficient. 10 The production of light in a sodium lamp is achieved when an electric current is passed through sodium vapor [1]. In a typical high-pressure sodium lamp there exists an arc tube, inside which an amalgam of metallic sodium and mercury resides to produce light through vaporization, and a small amount of xenon functions as a starting gas. The arc tube is made of a refractory material, namely aluminum oxide (alumina), to resist the corrosive effects of the high-temperature sodium [1]. The outer envelope is typically evacuated to isolate internal electrical components from environmental effects and maintain a steady arc temperature. While low-pressure sodium lamps typically radiate energy in the principal double D lines of sodium at 589nm, the high-pressure sodium bulb will radiate energy across the visible spectrum because of the pressure broadening associated with higher sodium pressures in the lamp [1]. The resultant light output is no longer monochromatic, making colors distinguishable and thus more suitable for indoor applications. Another important feature inherent to the sodium lamp is the absence of a starting electrode and the presence of an ignitor, which provides a high voltage, high frequency pulse to start the lamp. The current application of this feature in pulse-start metal halide bulbs has seen groundbreaking results. Fig. 3 displays the typical construction of a high-pressure sodium lamp [1]. 11 Fig. 3. High-pressure sodium lamp construction [Taken from Ref. 1]. 2.2 HID Ballasts A HID lamp must also have a ballast in order to operate effectively. Because the light source does not regulate its own current consumption well enough, it must utilize the services of a ballast to perform this operation. All HID and fluorescent lamps present a negative resistance to the power supply when connected in an electrical circuit. If the ballast were not used to provide a positive resistance in the circuit, the lamp would draw an increasing amount of current until it failed or destroyed the power supply. Ballasts have the reputation of being fairly intricate devices even though their operation is quite simple. Ballast types are numerous, but there are two categories for establishing their function: magnetic or electronic. Magnetic ballasts are similar to transformers in that they use an inductive core and coil to slow down changes in current. An inherent annoyance with magnetic ballasts is their tendency to produce a flicker because of their 12 low frequency modulation. They also vibrate which causes a noticeable hum. Electronic ballasts use solid state electronic circuitry to produce starting and operating currents for the lamp. They are favored over magnetic ballasts because of their ability to take the frequency of power flowing from the mains (50 or 60 Hz) and transform it to 20,000 Hz or more [1]. This eliminates the stroboscopic effect (flicker) and constant humming prevalent in the operation of magnetic ballasts. In addition, electronic ballasts offer higher system efficacy because of efficiency gains in the ballast itself, such as the use of a capacitor for lower line losses over an inductor [1]. 2.3 HID Evaluating Criteria There are several basic criteria upon which the two lamp types can be distinguished for their performance in those areas. Some of those criteria will be used to discuss the performance of both lamps. 2.3.1 Lamp Warm-up and Re-strike Times An important aspect of any light source is its ability to yield light quickly. An attractive light source will reach its full output of light and maintain its normal operating color in the shortest amount of time, whether that be from first start or hot start (restrike). 2.3.1.1 Metal Halide Although metal halide lamps below 150-W use ignitors, which provide enough voltage across the main electrodes to start the lamp, most metal halide lamps above 150- 13 W in operation today use a standard starting probe which is part of the lamp circuit. It takes time for the metal halides in the arc tube to warm up and evaporate into the arc; the time it takes a lamp to reach 90% of full light output (warm-up) can vary between 2 and 15 minutes. Re-strike times are generally a lot longer because it takes a while for the traditional pinched-body arc tube to cool down. In order to reignite, the temperature and vapor pressure inside the arc tube must reach a reasonable level; otherwise, the high required starting voltage cannot be applied without destroying the lamp. Re-strike times for metal halide lamps typically vary between 5 and 20 minutes. Pulse start systems, which utilize an ignitor have superior warm- up and re-strike times over standard probestart systems [22]. Fig. 4, taken from the National Lighting Product Information Program (NLPIP), illustrates this point [22]. Fig. 4. Warm-up and re-strike comparison for probe-start and pulse-start lamps [Taken from Ref. 22]. 14 2.3.1.2 High-pressure Sodium To start a high pressure sodium lamp, an ignitor has always been used for creating a high voltage, high frequency pulse. Generally, the vapor pressure is lower and the starting gas is ionized more easily with this configuration than it is with convention probe start metal halide lamps. The arc is, therefore, created faster, allowing for shorter starting times, and consequently, warm-up times. Re-strike times are also shorter because of a lower vapor pressure and the presence of an ignitor. 2.3.2 Lamp Life and Lumen Maintenance The benefits from long lamp life are glaringly obvious. A lamp that lasts longer will save money in replacement and maintenance costs. The average rated life, which is provided by most manufacturers, is determined when 50% of a group of lamps initially installed during a test are still operating. “The procedure prescribes operating cycles for HID lamps of 11 h on, 1 h off” [1]. However, lamp life listed by manufacturers is generally based on run times of both 3 and 12 hours. Factors affecting lamp life include but are not limited to: burn position (vertical or horizontal), number of starts, high or low operating voltage, ineffective control devices (ballasts, capacitors, etc.), and extremely high operating temperatures. Lumen output for lamps is normally tested by the manufacturer and measured after 100 hours of operation (initial lumens) and at 40% of rated lamp life (mean lumens) [1]. A lamp that maintains its initial lumen output throughout the course of its life would certainly be ideal. However, this is only ideal, for lamps will lose lumen output gradually over the course of their life. Fortunately, most causes of lumen depreciation can be 15 corrected and have been as technology has improved. Lumen maintenance curves for three 400-W HID sources are displayed below in Fig. 5 [1]. Fig. 5. Lumen maintenance curves for 400-W HID light sources [Taken from Ref. 1]. 2.3.2.1 Metal Halide Of all HID lamps, metal halide lamps tend to experience a greater level of lumen depreciation. Also because the electrodes used in metal halide lamps evaporate more readily, they experience shorter life cycles. Toward the end of their life (the last 40%), metal halide lamps can experience significant color shift. In color critical situations where the color shift might be objectionable, the lamp will be replaced. Standard metal halide lamps generally last from 10,000 to 20,000 hours [1]. 16 2.3.2.2 High-pressure Sodium High-pressure sodium lamps utilize smaller diameter arc tubes and an electrode construction similar to that of mercury lamps. The electrodes contain metal oxides embedded within the tungsten coil which allow for a slower evaporation process. This slow evaporation process means less “sputtering” over the course of the lamp’s life yielding higher lumen maintenance values than those present in typical metal halide lamps [1]. “The life of a high pressure sodium lamp is limited by a slow rise in operating voltage that occurs over the life of the lamp” [1]. This slow rise in voltage is caused by a chain of events that cause termination of the lamp when the ballast can no longer supply a high enough voltage to reignite the arc. The typical lifetime of a high pressure sodium lamp is 16,000 to 24,000 hours [1]. 2.4 HID Technological Advancements Ever since both metal halide and high pressure sodium lamps were developed in the 1960s to improve the operational shortcomings of mercury vapor lamps, there has always existed a tradeoff between both lamps which inherently limited their versatility. For decades high-pressure sodium lamps have been used as the energy saving alternative to metal halide lamps. More specifically, they were considered the most viable choice “when color quality is secondary to low operating and maintenance costs” [2]. In manufacturing facilities, the characteristic yellow light produced by high-pressure lamps is normally reserved for outdoor applications and color-indifferent areas where low contrast light (poor color rendition) is tolerable, as is common in most warehouses and 17 storage areas. However, when higher quality white light is necessary, as is the case in providing light for the surfaces in most industrial facilities, metal halide lamps were chosen because they distribute crisp, white light with higher color rendition. For specifiers and plant managers alike, it was always a choice between a source with good color quality and a source with good operational efficiency; no single lamp presented both merits. 2.4.1 Metal Halide This dilemma in HID lighting existed for decades. While standard metal halide lamps are still keeping most manufacturing facility floors lit, and high-pressure sodium lamps maintain a relegated role in providing light to secluded areas which see little occupation, revolutionary research and development is devoted to changing this scenario. Within the last 10 years considerable research has led to significant innovations in the development of metal halide lamps allowing them greater versatility. The metal halide lamp was chosen because “of all the lamps in the high-intensity discharge (HID) family, the metal halide family of lamps has the brightest future” [3]. Despite their limitations, metal halides already offer much better light quality than mercury or sodium lamps and have a potential for modification which far outweighs other lamps in the HID family. Because they could already produce the coveted white light demanded by the industrial sector, they were an obvious choice over mercury and high-pressure sodium lamps for future development. A company on the forefront of integrated luminaire technology, focused solely on providing new metal halide lamp and ballast systems, Venture Lighting, strives “to be the 18 best metal halide lamp company in the world and to dominate in the development of metal halide lamps for general lighting” [23]. Heralded as “the fastest growing metal halide company in the world,” Venture Lighting is the only corporation concentrated on this technology [23]. Presently the owner of many design patents on pulse-start lamps and ballasts, they are providing the lighting industry with many innovative options, rectifying the problems plaguing the old standard probe-start lamps. Another company dedicated to solving this paralyzing dilemma in HID lighting, Advance, often considered the industry leader in the manufacturing of new ballasts, is regarded as the top source for innovations and designs in fluorescent, HID, and LED ballasts in North America. Leading an aggressive pursuit of this desirable integrated systems approach to solving key operational limitations of HID lighting, Advance coordinated a revolutionary cooperative effort between itself and the top four U.S. lamp manufacturers to coalesce issues facing the formerly distinct lamp and ballast industries. Lamps and ballasts had always been seen prior as separate entities in the design process, which Advance recognized as an impediment to the process needed to solve this dilemma. Advance has performed substantial research in conjunction with these lamp manufacturers to overcome many design flaws in probe-start metal halide lamps. Both of these sources will be documented as references for most of the changes in metal halide lamp design and operation. 2.4.1.1 Pulse-start Technology The quest of Advance, and Venture Lighting alike, was a complete systems approach to solving the major dilemma in HID lighting. Metal halide lamps must be 19 considered a complete system, more than any other lamp [3]. All key components (ballast, ignitor, controls, lamp and fixture) must work together to get the best performance [3]. Because almost every lamp is designed individually and is different chemically, containing different doses of chemicals, careful attention must be paid in designing the whole system. Thus, it was essential that the problem be tackled in this manner. The problem with probe-start technology surfaced when lamp manufacturers began to focus on improving performance features in metal halide lamps, specifically lumen efficacy and lamp life. In order to reach this end, manufacturers desired greater fill pressures and different chemistries inside the arc tube; however, the benefits of these improvements inside the arc tube could not be realized unless a modification was made to allow for higher starting voltages. Probe start systems, which only allowed for 600-volt peak starting voltages, could not accommodate the arc tube changes. In other words, to see an increase in lumen efficacy and lamp life, a higher starting voltage would need to be applied to the lamp, which the standard probe-start system was incapable of doing. That is when researchers seized the opportunity to apply present technology to a different lamp type. By using an ignitor, the integral device already used in high-pressure sodium lamp circuits to apply starting voltages, manufacturers were able to revolutionize the way metal halide lamps operate. Pulse start technology has existed in metal halide lamps in the 150-W range and smaller for quite some time, yet within the last five years manufacturers have started applying the technology to a broader range of lamp sizes. At first glance, the difference in a standard probe start lamp and the modified pulse start lamp seems minor, yet the 20 consequent gains in overall lamp operation are vast. A conventional metal halide system consists of a lamp and probe start ballast. Inside the arc tube are three electrodes, “a starting probe electrode and two operating electrodes” [22]. The ballast applies a high starting voltage to the probe, normally greater than the voltage applied to the main electrodes during operation, creating a discharge in the small gap between the probe and one of the main electrodes. “Electrons then jump across the arc tube to the other operating electrode to help start the lamp” [22]. Once the main arc is established and the pressure and temperature begin to rise in the arc tube, the probe electrode is removed from the circuit by a bi-metal switch. A pulse-start system does not use a starter electrode. Instead, an ignitor sends a series of high voltage pulses (usually 3 to 5 kilovolts) across the two main electrodes [22]. It is important to recognize that the ignitor is an integrated part of the ballast. Therefore, pulse-start lamps will not operate on traditional probe-start ballasts. By eliminating the probe to start the lamp, a level of performance from metal halide lamps that never existed before is made possible. Because the probe is no longer present inside the arc tube, “the amount of pinch (or seal) area at the end of the arc tube is reduced, which results in a reduced heat loss” [22]. This distinction opened up a wealth of opportunities, allowing manufacturers the flexibility to optimally reshape arc tubes and increase fill pressures and temperatures inside, producing greater light output (arc tube modifications are discussed below). Additionally, two of the key elements prone to cause premature lamp failure, the bi-metal switch and probe electrode, are removed from the circuit in a pulse start lamp, boosting lamp life to levels only attained previously by highpressure sodium lamps. Furthermore, the series of pulses heat up the lamp faster during 21 starting, creating better warm-up times. Finally, probably the most distinguishable feature in the pulse-start lamp is a reduction of tungsten sputtering, an undesirable characteristic in probe-start lamps which caused considerable lumen depreciation as tungsten blackened the arc tube wall as a lamp aged [22]. Fig. 6 displays the subtle difference in probe-start and pulse-start lamps which produces so many gains in performance [22]. The probe electrode and bi-metal switch are easily visible in the figure. Fig. 6. Construction of probe-start vs. pulse-start lamps [Taken from Ref. 22]. Manufacturers claim many achievements for pulse-start systems over their predecessor. The following list of improvements obtained from Advance and NLPIP yields many of these performance achievements [2, 22]: • Depending on lamp wattage and burning cycle, longer lamp life of up to 50% over traditional probe-start metal halide lamps (20,000 – 30,000 hrs) • Better lumen maintenance – increases by up to 33% in most lamps 22 • Greater efficacy – lumen output per watt increases by 25% to 50% • Superior color rendition – a 30% improvement of up to 85 CRI • Reduced color shift and improved lamp to lamp color consistency • Faster warm up and re-strike times – warm up time is often cut in half to 2 minutes and re-strike hovers around 3 to 4 minutes • Better cold starting capability – start at temperatures as low as -40°C (-40°F) Table 2 yields a simple comparison of probe-start and pulse-start metal halide lamps to high-pressure sodium lamps [2]. The gap in performance between probe-start metal halide lamps and high-pressure sodium lamps is evident from these criteria. Table 2. Standard HID Comparison with Pulse-start Metal Halide Comparison with Pulse Start Metal Halide Features Probe Metal Halide Pulse Metal Halide High Pressure Sodium Color White Whiter Yellow Efficacy (L/W) 60 – 85 90 – 110 100 – 125 Lumen Maintenance 65% 70% - 75% 90% Lamp Wattages 175 – 1000 35 – 1000 35 – 1000 Lamp Life (kHrs) 6 – 20 10 – 30 24+ Source: Taken from Ref. 2. 2.4.1.2 Arc Tube Improvements A direct consequence of the elimination of both the probe electrode and bi-metal switch in pulse-start metal halide systems allowed for some major innovations in arc tube 23 design. Venture Lighting has patented and crafted an arc tube design which reshapes the arc chamber to provide a contoured fit of the arc itself. This unique quartz sculpting process, resulting in Venture Lighting’s patented Uni-Form body, eliminates the pinchedseal design which prevailed for nearly 30 years in the industry [23]. Fig. 7 displays this contoured Uni-Form body arc tube in comparison to the pinched-seal design [23]. Not only is the arc tube smaller and lighter, allowing it to heat and cool at speeds faster than the previous pinched seal design would allow (better warm up and re-strike times), but it creates a uniform geometry which produces uniform heating of the metal halides in the arc tube. This allows the materials in the arc tube to reach a higher average temperature. Higher temperatures equate to a higher light output from the arc tube. Venture Lighting claims many performance improvements in their pulse-start systems because of the formed-body arc tube, namely: higher efficacy, better color uniformity, faster warm up and hot re-strike, and longer life [23]. Fig. 8 displays formed-body arc tubes inside two different shaped metal halide lamps [27]. Fig. 7. Common metal halide arc tubes [Taken from Ref. 23]. 24 Fig. 8. Formed-body arc tube inside two differently shaped metal halide lamps [Taken from Ref. 27]. An additional finding, which was the result of a tinkering similar to that of the installation of an ignitor in the metal halide ballast, came “when Phillips engineers borrowed a ceramic arc tube from a sodium lamp and stuck it into a metal halide lamp” [3]. Manufacturers soon discovered a superior improvement in color performance. This improvement in color offers CRI values in the 80 to 90 range, approaching that of halogen lamps. Possibly the most important consequence of this discovery is the almost guarantee of complete color consistency (meaning no color shift during a lamp’s lifetime). Lumen maintenance also improves in ceramic metal halide lamps. Though ceramic metal halide lamps are essentially a perfect match for retail and residential accent 25 and downlighting applications because of their ability to replace halogen and incandescent lamps, they are great for color-critical spaces often prevalent in industrial facilities. 2.4.1.3 Electronic Ballasts Traditionally, metal halide lamps have operated on a magnetic ballast, consisting of a core and coil which are relatively heavy, hot (when operating), and bulky. Many ballast manufacturers now offer electronic ballasts that are smaller, lighter, and provide better performance [3]. Electronic ballasts have been successfully used with fluorescent lamps for years mainly because of reduced energy losses and dimming capabilities, but only recently have they taken flight in high-wattage HID applications. Electronic ballasts use solid state circuitry to control voltages and currents for starting and operating the lamp in a manner more efficient than magnetic ballasts. In other words, they generally accomplish the task of current regulation with a lower input wattage to the system than do their predecessors. They also have the ability to operate at higher frequencies, often reaching 100 kHz and above. Higher frequency operation of ballasts reduces tungsten electrode sputtering, and therefore consequent deposition on the arc tube wall. This improves many facets of a lamp’s performance including: lumen maintenance, lamp life, warm up and restrike, and color stability. All electronic ballasts use pulse starting and therefore gain additional operational efficiencies from the use of an ignitor (as mentioned previously pulse starting also reduces tungsten sputtering). Another appealing advantage of electronic ballasts is their ability to allow lower dimming levels at higher efficiencies than magnetic ballasts could attain. Conventional magnetic ballasts can only attain 26 dimming levels of 40% to 50% of full output, with significant reductions in system efficiency [3]. Electronic ballasts can drop lighting levels lower while still maintaining relatively good system efficiencies. The combination of these features of electronic ballasts and the potential to apply ceramic arc tubes to metal halide lamps has produced a viable alternative source for many retail applications. The future for electronic ballasts in the industrial sector is somewhat uncertain because owners are hesitant to use them at such a high initial cost. Presently, it is difficult to consider a major capital project for replacing old magnetic ballasts on metal halide lamps because of high initial costs, but recent trends show that the price of electronic ballasts is decreasing. Fig. 9 is a photograph of Advance’s Dynavision electronic HID ballast. These ballasts operate with 320-W and 400-W HID lamps. An innovative microprocessorbased design offers significant increases in lumen maintenance with continuous dimming of the lamp. Fig. 9. Advance’s Dynavision electronic HID ballast [Taken from Ref. 28]. 27 The following is a list of advantages provided by electronic HID ballasts [13]: • Improved light color and output • Thirty percent longer lamp life and lower lumen depreciation • Continuous dimming capabilities • Reduced energy costs • Smaller size and lighter weight • Silent operation • Flicker-free • High-power factor (0.99) and low harmonic distortion 2.4.2 High-pressure Sodium Because metal halide lamps were essentially the “chosen” HID lamp for future development, not much has been done to improve the performance of high-pressure sodium lamps. Several manufacturers like Phillips, GE, and Venture Lighting offer highpressure sodium lamps with better color rendition, but these retrofit options still provide warmer color temperatures (2,200K) and a shorter life (15,000hrs). These options are attractive because they yield better quality light with a simple lamp retrofit, but they will never offer the benefits afforded metal halide lamps due to lack of new technology (see aforementioned section on metal halide advancements). Fig. 10 displays a high-pressure sodium retrofit option with improved color rendition manufactured by Phillips. 28 Fig. 10. Phillips Ceramalux® high-pressure sodium lamp [Taken from Ref. 29]. The characteristic yellow light provided by high-pressure sodium lamps is more than just a limiting factor for development; it seems it has become responsible for their demise. As the production of metal halide lamps with increasingly better efficacies and lumen maintenance values continues, high-pressure sodium lamps will most likely be utilized for fewer applications and eventually bumped from the market. Applications where the unwanted yellow light source is chosen for reasons of efficacy and lumen maintenance are diminishing. Metal halide lamp manufacturers are creating new lamps with all the color benefits of previous sources and performance qualities approaching those of high-pressure sodium lamps. Additionally, people are beginning to prefer higher quality white light for outdoor and other niche applications controlled by high pressure sodium lighting because of these advances in metal halide lamps. The T-12 fluorescent lamp has been dealt similar circumstances. It presently dominates the fluorescent market much like high-pressure sodium lamps still dominate their niche applications, but both are fast approaching the end of their useful lives because of more efficient and better quality alternatives. CHAPTER 3 LINEAR FLUORESCENT LAMPS Fluorescent lamps are manufactured in a variety of shapes and sizes. Linear fluorescent lamps represent a distinct family of straight-tube light sources that generate light through the excitation of mercury atoms and subsequent fluorescence of a phosphor coating. Three particular types of linear fluorescent lamps will be described in anticipation of the analyses that follow. 3.1 Linear Fluorescent Lamp Construction In 1857, a French physicist named Alexandre E. Becquerel began experimenting with the luminescence of substances when placed in a Geissler tube. These discoveries would lead to theorizations about the construction of fluorescent tubes similar to those seen today. Heinrich Geissler, a German glassblower, had earlier created a mercury vacuum pump capable of inducing a vacuum in a glass tube to a level not previously attainable. When current was passed through this tube, a green glow could be seen at the cathode end of the tube, creating a device known commonly as the Geissler tube. Becquerel built on this discovery by experimenting with different coatings on the inside of the tube. In 1901, the American, Peter Cooper Hewitt, patented the first mercury vapor lamp. This is considered “the very first prototype of today’s modern fluorescent lights” [11]. 29 30 A fluorescent lamp, commonly categorized as an arc or vapor lamp, operates in a completely different way than does a standard incandescent lamp. In an incandescent lamp, light is simply a byproduct of electricity running through a highly resistive filament. As electricity flows in the circuit of an incandescent lamp, high temperatures in the filament cause it to glow and emit light. In a fluorescent lamp, light is produced when electricity is applied to different gas mixtures contained in a vacuum tube. So the direct result of operating a fluorescent lamp is visible light, while an incandescent dissipates 90% of the input electrical power as heat. Fluorescent lamps are low pressure, gas-discharge devices that utilize electricity to excite mercury vapor suspended in a gas (normally argon or xenon). An arc is created when electricity runs through a cathode (coiled tungsten electrode typically coated with barium, strontium and calcium oxides – electron-emitting substances) at each end of a tubular lamp, sending current through mercury vapors in the tube. As electricity passes through the lamp circuit, the cathode heats up and emits electrons that collide with mercury atoms, displacing electrons in these atoms from their natural orbit and forcing them into a higher, unstable energy state [12]. As the displaced mercury atom’s electron returns to a stable energy level, it gives off some visible, but mostly invisible (ultraviolet) radiation (generally along principle lines of 254, 313, 365, 405, 436, 546, and 578 nm). This ultraviolet radiation is transformed into visible radiation by a special phosphor lining on the inner side of the lamp tube which lengthens the “UV wavelengths to a visible portion of the spectrum” [12]. In other words, ultraviolet photons are absorbed by electrons in the atoms of the phosphor coating which causes a similar energy level jump, and drop, emitting another photon of wavelength perceptible by the human eye. “The 31 phosphors are generally selected and blended to respond most efficiently to 254 nm, the primary wavelength generated in a mercury low-pressure discharge” [1]. Fig. 11 illustrates this phenomenon in fluorescent lamps [12]. Fig. 11. Illustration of light production in general fluorescent lamp [Taken from Ref. 12]. Fluorescent lamps are commonly designated by their shape and diameter which follow a particular nomenclature. A letter designates the shape of the lamp while a number (indicated in eights of an inch) designates the diameter. For instance, T-8 refers to a tubular bulb 8/8 in. (1 in. in diameter). Lamps can either be tubular or U-shaped. Most lamps seen in the industrial sector are tubular. Lamps vary in diameter from 2 mm. (0.25 in. T-2) to 54 mm. (2.125 in. T-17). Their lengths vary from 100 to 2440 mm. (4 to 96 in.) [1]. Similar to HID lamps, fluorescent lamps use two electrodes which are hermetically sealed inside the tube, one located at each end. These cathodes are designed for either cold or hot operation, but because hot operation (arc mode operation) is inherently more efficient, most fluorescent lamps are designed in this way. In arc mode operation, the hot cathode is typically constructed from either a single strand of tungsten 32 wire or a coiled piece of tungsten wire around which another finer piece of wire is uniformly wrapped. Thus, the electrode is called either a coiled-coil electrode or a triplecoil electrode, respectively. Both electrode types are coated with an electron emissionenhancing substance, which is commonly a mixture of alkaline earth oxides. As current passes through the tungsten coil, temperatures of 1100 °C (2012 °F) are reached, causing the coating to emit a large quantity of electrons [1]. The lamp tube is commonly filled with mercury vapor, typically maintained at 1.07 Pa (the vapor pressure of liquid mercury at 40 °C – the optimum bulb wall temperature of operation for which most lamps are designed), and a rare gas or combination of gases at low pressure to facilitate ignition of the discharge [1]. Common rare gases used in lamps include: argon, krypton, neon, and xenon. An integral part of operation is the ability of the lamp to develop and maintain a discharge between the two electrodes. Thus, the attainment of optimum temperatures and pressures inside the tube is critical for the ionization of mercury gas, which ultimately determines the fate of a consistent discharge. The phosphors which coat the walls of each fluorescent tube will establish the color of light produced by each lamp. Lamps with markedly different spectral power distributions are available for purchase, and thus, it is often a chore selecting a lamp with the most appropriate color characteristics. “Popular fluorescent lamps use three highly efficient narrow-band, rare-earth activated phosphors with emission peaks in the short-, middle-, and long-wavelength regions of the visible spectrum. These triphosphor lamps can be obtained with high color rendering, improved lumen maintenance, and good efficacy with correlated color temperatures between 2500 and 6000 K relative to 33 halophosphate lamps” [1]. However, rare earth phosphors are rather expensive and thus longer lamps (T-5 – T-12) typically employ a two-coat system. The less expensive halophosphate is located on the inside, while the more expensive rare earth blend resides on the outside, closest to the mercury discharge. A lamp utilizing this configuration of phosphors will typically have a spectral power distribution with a closer resemblance to that of the rare earth phosphors. Fig. 12 was taken from the Illuminating Engineering Society of North America [1]. The three distinct illustrations within the figure elucidate some key areas of function in a fluorescent lamp. 3.2 Linear Fluorescent Lamps Types The following discussion highlights some aspects of linear fluorescent lamps in the most common available sizes. Descriptions are pithy and not meant to provide a comprehensive review of each lamp type. 34 Fig. 12. Cutaway view of some common fluorescent lamps: (a) typical rapid start fluorescent lamp and the production of light (b) lamp electrode construction (c) detail of the electrode [Taken from Ref. 1]. 3.2.1 T-12 Lamps Despite the influx of newer, more efficient fluorescent lamps and the passing of the National Energy Policy Act of 1992 (EPACT) which banned their production after 1995, T-12 fluorescent lamps are still the most commonly used source of fluorescent light. T-12 lamps, possessing a tube diameter of 1.5 in., are commonly seen in lengths of 4-ft. and 8-ft [1]. More specifically, the 40-W, 4-ft, rapid-start T-12 lamp had 35 considerable control of the market for fluorescent lighting until more efficient and efficacious lamps exposed its inefficiencies. Today many of the nation’s industries still use the inefficient T-12 lamp, mainly in a reduced 34-W configuration, but most are making the switch to more efficient T-8 models when replacements are needed. Despite a growing awareness of more viable replacements to the T-12 lamp, many facilities are hesitant to invest in expensive capital projects to replace these lights. T-12 lamps have clearly seen their days of dominance, however; and today, T-8 and T-5 lamps are slowly uprooting the industrial sector’s workhouse due to numerous energy cost saving benefits. 3.2.2 T-8 Lamps T-8 lamps represent a viable, energy-efficient replacement to the obsolescent T12 lamp. They are a family of 1 in. diameter straight tube fluorescent lamps which are manufactured in many of the same lengths as T-12 lamps. The 4-ft version is the most common, consuming approximately 32 W. Other common lengths are 2-,3-,5-, and 8-ft. Though T-8 lamps are interchangeable with T-12 fixtures of similar length, a different ballast must be used in replacement. Most T-8 lamps utilize higher efficiency electronic ballasts which facilitate even better performance. Additionally, these thinner lamps “enable better photo-optic control of the light produced by the fixture, increasing efficiency and providing uniform distribution of its light output” [5]. The smaller diameter also allows for a more efficient use of the expensive rare earth phosphors. T-8 lamps are also offered in popular energy savings wattages of 25 W, 28 W, and 30 W where a combination of ballasts with varying ballast factors can be employed to achieve a desired quality and quantity of light for numerous applications [1]. 36 T-8 lamps installed in multi-lamp fixtures have recently gained momentum in the market for fluorescent high-bay fixtures because of their proven ability to provide good optical control and steady, uniform light. For high fixture mounting heights (those typically seen in high-bay applications), high-lumen, 32-W T-8 (Super T-8) lamps are often specified. A drawback of using T-8 lamps at this height is that more lamps may be required to produce the same wattage as a 400-W metal halide in a retrofit. Currently multi-lamp fixtures of both T-5 and T-8 lamps are being employed for both high and low bay applications (although T-5 fixtures are considered a better option at heights above 20ft.). Common high-bay applications utilize 6- high-lumen, T-8 lamps (typically 3,100 initial lumens per lamp) with high ballast factor ballasts (1.14-1.20) and very efficient reflectors. 3.2.3 T-5 Lamps T-5 lamps also represent a suitable, energy-efficient replacement to the T-12 lamp. These lamps pose an even greater threat because they are smaller in diameter than the T-8, offering the same benefits but on a different level. According to the National Electrical Manufacturers Association (NEMA), T-5 and T-5HO lamps comprise 2% of the fluorescent lamp market. This should, however, come as no surprise considering it has taken almost a decade for T-8 lamps to gain their 33% share of the market from the entrenched T-12 lamp. The T-5 lamps compose a family of 5/8 in. diameter straight tube lamps, 40% smaller than T-12 lamps, that utilize the appealing triphosphor technology. They are designed to operate solely on electronic ballasts, and for most multi-lamp systems the 37 newly refined programmed start ballast is considered optimal. This fact, in conjunction with the fact that they are only available in metric lengths and bipin bases, renders them a more difficult replacement for the T-12 lamp. Rather than just replacing the ballast for a T-12 to T-8 change, the entire fixture must be replaced to change from either of these lamps to a T-5 lamp. Despite making a retrofit rather difficult, these features are generally beneficial to the T-5 lamp. Mainly, they allow for it to provide a higher source of brightness and better optical control. Additionally, because they are so much smaller, they are designed to provide optimum light output at 35 °C (95 °F) rather than the common 25 °C (77 °F) [1]. For this reason it has been stated that T-5 lamps are better than T-8 lamps for high-bay applications where greater ambient temperatures are more common. Also, this higher optimum operating temperature allows for the design of more compact luminaires, an extremely attractive feature responsible for the rise of T-5 highbay fixtures. Another important advantage of T-5 lamps is its use of barrier coating technology. The smaller diameter lamp makes it more economical to use expensive coatings that reduce the absorption of mercury by the tube’s glass interior and phosphor coating. This barrier coating has dramatically reduced the amount of mercury needed to operate the lamp (from 15mg to 3mg), and consequently, by reducing mercury absorption which causes lamp light depreciation, increased maintenance of initial lumen output values. Typically fluorescents were not considered a viable option for high-bay applications because they could not provide the source brightness that HID lamps could at heights above approximately 20 feet. But the emergence of higher-output fluorescent lamps with excellent performance qualities, and cooler-operating, multi-lamp (4 to 6 38 lamps) fixtures has allowed T-5 (and in most cases T-8) luminairies to compete at these heights. Table 3. Comparison of Nominal Four-Foot Fluorescent Lamp Types Comparison of 4-foot T5HO T8 82 85 3,000/3,500/4,100/5,000/ 3,000/3,500/4,100/5,000/ 6,000K 6,000K 5,000 lumens @ 35°C 3,000 lumens @ 25°C ambient ambient 54W 32W 92.6 lm/W @ 35°C 93.8 lm/W @ 25°C ambient ambient Lumen Maintenance 95% @ 8,000 hours 95% @ 8,000 hours Maintained System 86.1 lm/W @ 8,000 89.1 lm/W @ 8,000 Efficacy hours and 35°C ambient hours and 25°C ambient Rated Life (3hrs/start) 20,000 24,000 Fluorescent Lamps CRI Color Temperature Initial Rated Light Output Nominal Lamp Wattage Initial Lamp Efficacy Source: Taken from Ref. 2. 3.3 Fluorescent Ballasts The interaction between lamps and ballasts is one of the most significant factors governing overall system performance. Various starting and operating methods for lamps are provided by a multitude of different ballasts and it is of utmost importance to ensure 39 that the ballast is chosen to match a particular application. These ballasts will often provide a balance of certain characteristics like lamp life, energy efficiency, and maintenance costs. Most fluorescent lamps used today operate on electronic ballasts because of their inherent superiority over magnetic ballasts. They are significantly cheaper for fluorescents than they are for HID lamps due to many decades of proven performance and field acceptance. As previously noted, electronic ballasts will provide energy savings over magnetic ballasts, but there are numerous reasons that their influence on the fluorescent ballast market has been substantial. Among other advantages, they are smaller, quieter, and lighter than their magnetic counterparts [25]. They operate at cooler temperatures, are more efficient (better efficacy – high lumen per watt), provide longer lamp life, and reduce the flicker and soft hum common in magnetic ballasts (because of higher frequency operation) [25]. Three main types of fluorescent ballasts, categorized by their lamp starting and operating methods, are seen consistently in application, but one particular type, the programmed-start ballast, has garnered considerable acclaim for its ability to operate with multiple controlling techniques in coveted high-bay applications, and will thus be the focus of discussion. 3.3.1 Rapid Start Ballasts A rapid start ballast applies continuous heat to the cathodes during the course of operation of a fluorescent lamp. The heat and starting voltage are applied simultaneously. This continuous heating of the cathodes reduces the necessary starting voltage and allows the cathodes to reach a sufficient operating temperature. High starting voltages are often the cause of tungsten sputtering and eventual lamp blackening; so a 40 reduction in starting voltage likely increases the life of the lamp. The only drawback is that additional power will be needed to provide continuous cathode heating. Additionally, because the cathode heating and starting voltage are applied simultaneously, a delicate balance must be reached for the optimal performance. The cathodes need to be heated to the appropriate temperature before the arc is established. But there is no guarantee that proper cathode temperatures will be reached prior to lamp ignition. If the cathode temperature is below optimum and the voltage across the lamp gets high enough to ignite the lamp, sputtering of the emissive material may ensue. However, as will be discussed later, programmed start ballasts rectify this design shortcoming (see section on programmed start ballast). Rapid start ballasts are very good for lamps connected to occupancy sensors, photocells, or any other lighting control devices where they might be switched on and off frequently. Note that additional circuitry needed for cathode heating warrants a higher ballast cost [25]. 3.3.2 Instant Start Ballasts An instant start ballast will provide a high starting voltage to a fluorescent lamp without prior cathode heating. Providing more of a “jump start” to the lamp, instant start ballasts put undeniably more stress on the electrodes. An obvious reduction in energy consumption will exist as extra power to provide heating is unneeded, but a shorter lamp life because of repeated electrode material expulsion may outweigh the benefits of energy efficiency if the lamp is employed for frequent switching applications [25]. 41 3.3.3 Programmed Start Ballasts Programmed start ballasts are a variation or improvement to rapid start ballasts. Their unique starting process markedly improves lamp life (50 percent in frequent switching applications) and reduces the pitfalls associated with rapid start ballasts to approach the energy saving potential of instant start ballasts. Programmed start ballasts utilize this precise starting sequence, which is broken into well-established steps that eliminate the deficiencies in other starting methods [26]. The ballast begins with the application of heat to the cathodes during a starting or preheat interval, and voltage across the lamp is maintained at a low level. The duration of this first step is preprogrammed in a ballast and is an integral part of successful operation. The ballast provides heat (voltage) to the cathode until it reaches an optimum temperature (at least 700°C) [26]. Voltage across the lamp remains at a low level until this optimum cathode temperature is reached, preventing the lamp from igniting and creating an unwanted glow current (lamp current that flows during the preheat interval – causing lamp blackening and degradation) [25]. In the second step of operation, the ballast applies a voltage high enough to begin ignition of the arc, but low enough to preserve emissive material on the cathodes. Both step operating times and applied voltages are programmed into the ballast for better performance, hence the name “programmed start.” Often, cathode heating is reduced after lamp starting to provide additional energy savings with decreased power consumption. Ultimately, programmed start ballasts are a better alternative for applications that utilize lighting control devices as they are purported to provide 50,000 switching cycles (15 min. on/ 5 min. off) and a lamp life improvement of 10,000 hours in 42 the these conditions [25]. They are simply gentler on the cathodes than either of the other ballast types because they employ such a well-defined starting process. These ballasts are currently being paired with multi-lamp T5HO fixtures for excellent optical control and light quality when operated with a variety of light sensors and control devices.