There are many ways of evaluating “green” design. Green architecture includes specification of renewable, local, and nontoxic materials, enhanced envelope design, and daylight strategies. Green electrical and mechanical systems are evaluated on energy performance and rewarded for energy reduction. When it comes to lighting, there is much more to going green than simply reducing energy. In an effort to shave watts and evaluate efficacy, designers may overlook an obvious — or perhaps not so obvious — incandescent lamp choice; namely, the 37W/ MR16/ IR. This powerful and efficient halogen source has commercial and retail applications that may make it the lamp of choice for green designs, including those seeking Leadership in Energy and Environmental Design (LEED) Platinum certification.

At minimum, energy code compliance and LEED certification require evaluating lighting system performance based on watts per square foot, a metric that is oblivious to light quality, system performance, embodied energy, and environmental waste. If a lighting system meets energy requirements as measured by a watt-hour-meter but is of poor light quality, introducing glare and providing inadequate light levels, is it truly green? What if it is difficult and expensive to operate and maintain, doesn't account for the actual power delivered to the lighting system, or contains toxic waste and throw-away electronics? Is it sustainable?

To help answer these questions, we'll make two comparisons. First, let's look at the system operation of 37W/MR16/IR/FL35 lamps and PL-T 42W compact fluorescent lamps (CFLs) installed in commercial downlights. Then, let's weigh the advantages of the 35MRC/IRC/NFL24 halogen relative to the PAR38/CDMI/FL25 integrated ceramic metal-halide (M-H) lamp installed in an open lamp holder or mono-point. Conclusions are based on a macro analysis of sustainability, which includes trading energy performance for operational performance and re-examining the way we gauge energy efficiency relative to delivered light.

Performance metrics

Efficacy, as defined by the Illuminating Engineering Society of North America (IESNA), New York, equals lumen per watt and is a preliminary measure used to compare the energy value of various light sources. However, lamp efficacy is not a constant. When a lamp is installed in a luminaire, the efficacy is reduced by the luminaire efficiency. As the lighting system ages, its efficacy degrades further. Discharge lamps — metal halides, in particular — suffer significant losses in lumen output as they age, yet the energy input remains constant. At 40% of rated lamp life, M-H lamps nominally provide 65% of initial lumen output, and fluorescent lamps nominally provide 85% of initial lumen output. The efficacy of a halogen lamp remains relatively constant, losing only about 5% of initial lumen output by end-of-life.

The candlepower distribution of a lamp or luminaire provides precise measures of light intensity and distribution, whereas efficacy provides little usable information regarding actual performance. To illustrate this simply, consider two 60W PAR38 halogen lamps from a single manufacturer: one 25° narrow flood (NFL25) and one 30° flood (FL30). The manufacturer's data rate both lamps at 850 lumens, and both lamps have the same efficacy. When calculated at 5° intervals, however, the NFL25 delivers 11,658 total candlepower with a center beam candlepower (CBCP) of 3,909, and the FL30 delivers 9,598 total candlepower with a CBCP of 2,643. If you consider candlepower per watt in place of lumen per watt, the NFL25 significantly outperforms the FL30 at a distribution range up to 40° vertical (Table 1). Light intensity measured in candela is the standard metric used in point-by-point lighting analyses and is used for evaluating lamp and luminaire performance in this article.

Environmental impact

In addition to lighting and energy performance, environmental emissions and toxic waste disposal are of primary importance when considering sustainable lamp choices. All discharge lamps, including fluorescent and M-H, contain mercury. Not all discharge lamps are disposed of in accordance with national and local hazardous waste disposal laws. Incandescent/halogen lamps do not contain mercury; therefore, they have no potential of contributing to the toxic waste stream.

Power plants emit CO2 and airborne mercury during the production of energy from fossil fuels, such as coal. The debate over the contained mercury of a CFL versus the airborne mercury emissions from the energy production required to operate incandescent and CFL lamps loses impact if the incandescent/halogen lamp requires less energy to operate than the CFL. It's an even more difficult argument if the halogen source consumes less energy and provides more usable light to the space, as is the case when comparing the 37W MR16IR lamp to a 42W CFL when installed in a commercial downlight.

In addition to the amount of energy required to simply produce light, each lamp contains embodied energy, which is the energy required to manufacture, package, transport, and dispose of the lamp and its base materials (i.e., glass, metals, plastics, electronics, rare earth phosphors, and mercury). Assume for a moment that it takes the same amount of energy to produce the MR16/IR as it does to produce a CFL or M-H lamp. It's appropriate to conclude that the embodied energy of the MR16/IR is twice that of the CFL or M-H lamp because its rated life of 5,000 hours is about half of that either discharge source. The net result is twice as many MR16 lamp changes over time. Incandescent system design has the advantage of strategies to increase lamp life and consequently reduce the embodied energy of the lamp due to maintenance. Current energy codes require overall building control systems and local dimming controls. If the MR16/IR is trimmed to 90% via a lighting control system, it will operate for approximately 20,000 hours, which exceeds either the CFL or M-H lamp documented in this report. Upon further observation, there is little or no embodied energy involved with halogen lamp disposal. Because they do not contain mercury, they do not require special packaging, transportation, or recycling at end-of-life.

Getting down to business

Downlights are typical to both commercial and retail design. In application, fluorescent downlights are often used for general illumination because they provide a soft-edged uniformity with height to distance ratios between 1-to-1.0 and 1-to-1.3, and the lamp plus ballast efficacy is relatively high. Unfortunately, commercial fluorescent downlights are extremely inefficient (Table 2). Every lighting professional should question whether soft-edged uniformity is a necessary design requirement. If the answer is yes, then is a fluorescent downlight the most appropriate and sustainable design solution, or should an alternative method for general illumination be considered?

Downlights designed for use with reflector lamps operate at much higher efficiencies, typically around 85%. Basically, the reflector lamp does all of the work of the internal luminaire reflector, and it is positioned within the downlight housing such that little light is lost. The difference in light distribution and spacing distances varies by lamp selection and should be carefully evaluated for each design. If the space does not require absolute soft-edged uniformity, and if a downlight is indeed the right design solution for the space, then a reflector lamp may be a better choice than a non-reflector lamp, such as a CFL.

When evaluating lighting performance, it's crucial to look at the operating conditions of the system as a whole and not focus on the performance of individual components, such as the lamp, ballast, transformer, or luminaire. In this article, MR16/IR and 42W PL-T lamps are installed in representative downlights — each with a respective integral transformer and ballast supplied by the luminaire manufacturer. Independent Test Lab (ITL) reports for each downlight along with published manufacturer's lamp performance data are used for the comparison shown in (click here to see Table 3). The 50MR16/IR/FL35/C lamp installed during test ITL58960 is not the lamp of choice. A candela ratio from the lamp manufacturer's data provides a look at how the luminaire would likely perform if the 37MR16/IR/FL35 lamp were installed in its place.

The 37W MR16/IR downlight outperforms the 42W PL-T downlight on many levels. In summary, the halogen system consumes less energy, and its initial candlepower is much higher, even though the CFL system efficacy is greater. At 40% of rated lamp life, the total candlepower of the halogen system remains somewhat constant at 9,169, whereas the CFL total candlepower is reduced to 7,485. When trimmed to 90%, the MR16/IR lamp operates at 33.6W, providing similar light to the PL-T. Lamp life is also extended to 20,000 hours. Dimming a CFL will reduce the consumed energy and light output, but it will not extend its lamp life beyond 12,000 hours. CFLs contain about 4 milligrams of mercury. Halogen lamps do not contain mercury. The halogen lamp provides better color quality than the CFL, is less expensive to dim, and is easier to control.

The best application for reflector lamps in general and for the MR16/IR lamp in particular is installed in a simple track-mounted lamp holder or mono-point used for accent lighting in retail and commercial spaces. In the following, the performance of the 25W Integrated PAR 38/CDMI/FL25 (25W CDMI) is compared to the 35MRC/IRC/NFL24 halogen lamp. Refer to the article in the October 2007 issue of EC&M, “Using State-of-the-Art Screw-Based Lamps,” for a description of retail accent lighting and a summary of the 25W self-ballasted ceramic M-H (CDMI) lamp.

In this comparison, the 37W MR16/IR used in the downlight evaluation is replaced with a 35W MR16/IRC, because it is from the same manufacturer as the 25W CDMI. It's assumed that lamp data provided from a single manufacturer will be consistent due to similarities in testing methods, test equipment, and performance documentation — and will result in a more reliable comparison than lamps from different manufacturers. The 25° narrow flood distribution was selected because it is the most common and versatile lamp used for retail and commercial accent lighting ((click here to see Table 4)).

At first glance, this might be considered a dubious comparison, because the 35W MR16 has a lower CBCP, lower total candlepower, and higher wattage than the 25W CDMI. But as in any green analysis, it's important to consider all of the life-cycle operating characteristics and power requirements before formulating a conclusion.

At 40% of rated lamp life, the CDMI delivers 69% of its initial light output, which continues to degrade prior to end of life. If, as in the downlight example, we extend the MR16/IR lamp life to 20,000 via high end trim, then at 40% of rated life (4,200 hours of burn time) the halogen lamp will operate at 31.5W — and the total candlepower is only slightly less than the candlepower of the CDMI.

Low-voltage transformers used with MR16 lamps have a power factor between 0.95 and 0.98. The 25W CDMI has a power factor of 0.57, and an input current of 0.36 amps at 120 volts. Assume that the input current remains constant, and the CDMI has a power factor of 1.0. In this situation, it would actually operate at 43.2W. The real power in watts would equal the complex power in volt-amperes, as in the case of an incandescent filament lamp that is strictly a resistive load or a typical CFL with a 1.0 pf corrected ballast. Because the actual power factor of the 25W CDMI is 0.57 and not 1.0, essentially 18.2W of delivered power is lost due to a displacement or lag between voltage and current. The loss is not accounted for in a watt-per-square-foot energy summary, because a watt-hour meter does not measure complex power.

Like the CFL, the CDMI contains mercury and an integral electronic ballast, which may contain additional environmentally toxic materials. Whether the ballast does or does not contain toxic materials remains to be documented by the manufacturer. In either case, the CDMI contains an added level of embodied energy due to disposal of the contained mercury and production of the integral electronic ballast that is thrown away with each lamp change. The MR16 contains neither mercury nor electronics, making it a better choice with regard to environmental waste streams and embodied energy.

Other lighting issues to consider when specifying the 25W CDMI are color rendering, dimming, restrike time, bulb size, and system cost. The CDMI has a rated correlated color temperature (CCT) of 3,000K (± 200K) and a rated color-rendering index (CRI) of 87. It is true that the color quality of ceramic metal halide is superior to that of standard metal halide and, in most applications, provides a brilliant color that some designers consider better than halogen. Ceramic metal lamps — and this lamp in particular — have a spectral color spike between 525 nanometers and 625 nanometers, which renders skin tones in the yellow range. This may be a detriment in clothing retail, where looking good in a mirror is critical to a purchase.

As with all M-H lamps, the CDMI has a restrike period estimated at about 5 minutes from power fluctuation or power loss to restrike and an additional 3 to 4 minutes before the lamp reaches 80% of full light output. Unlike halogen and fluorescent lamps, the CDMI cannot be dimmed or controlled via occupant sensors. In rooms or spaces that are largely unoccupied during much of the normal building operation, CDMI lamps will likely remain on unless building occupants learn to accept the warm-up and restrike times associated with switching them off.

Weighing your options

Aesthetics and construction costs are not green design issues per se; however, they are integral to the success of every project. Evaluating the aesthetic appropriateness of the PAR38 CDMI lamp versus the MR16 is a question of scale within a space. The MR16 may be lost in a large room whereas the PAR38 may take up all of the space in a small room. Finally, there's the question of cost. The price of lamp holders and track for each system will be very similar, but the lamp cost is significantly different. Transformers for the MR16 low-voltage lamp holders may cost more initially than the total cost of the CDMI, but will only be purchased one time and will typically operate up to 15 years. The cost of each CDMI lamp includes the cost of the integral ballast, which can only operate for the life of the lamp — about 2½ years operating at 12 hours per day.

Because the decisions required for sustainable or green designs are numerous and complicated, they must be carefully weighed based on environmental stewardship, overall performance, and building economics. In the end, the operational requirements of the lighting system, the architectural program and details, and the project sustainability goals will guide designers to make appropriate system choices for each design application. Although the MR16/IR lamp may not be the best design solution for every downlight or accent light application, it should be carefully considered as an option — not automatically dismissed simply because it is incandescent.

Gallina is a lighting designer based in Minneapolis. She can be reached at