|
|
Just as an archer is judged by how close the arrow comes to the target’s center, a technology is judged by how well it meets the need that spawned it. In the case of solid-state lighting, the “target” is a composite one that encompasses a host of needs in addition to illumination—not just energy efficiency, but things such as durability and controllability. And because LED technology is still at an early stage, with many of its advantages yet to be fully developed and a fair amount of bugs still to be worked out, the bull’s-eye is attached to a moving target.
Another thing that has become clear is that solid-state lighting (SSL) is a long way from being the right solution for every application. A major reason is cost. With prices for LED products still generally much higher than those of conventional lighting, cost-effectiveness remains the toughest challenge and almost always requires factoring in benefits other than energy efficiency, such as reduced maintenance, for the numbers to work out favorably. These are the kinds of issues that lighting decision makers across the country are grappling with, as they consider switching to LED products.
Three recent Gateway demonstration experiences illustrate the current SSL situation nicely. In the first situation, the LED product fell short even when it came to saving energy and matching illumination levels of the incumbent technology. The second scenario yielded mixed results, in large part due to the unique needs of the application. The third situation is a “direct hit” on the target—the SSL solution saved energy, improved illumination and yielded an acceptable payback.
Missing The Mark
Dept. of labor HQ
At the headquarters of the U.S. Department of Labor (DOL) in Washington, D.C., many energysaving improvements have already been made to this Energy Star-rated, 2 million-sq ft facility. Government agencies are under an Executive Order to continue to save energy. In this building, after HVAC, lighting is believed to be one of the few areas still offering energy-saving potential, so DOL considered replacing linear fluorescent lamps with LED luminaires. The building’s primary luminaire is Lithonia’s recessed 1-ft by 4-ft parabolic louver troffer containing two GE T8 fluorescent lamps.
 To help determine whether wholesale lighting changes were advisable, the LED product under consideration was installed in a single DOL conference room. DOE was asked to assess the initial installation for its suitability as a potential Gateway demonstration. Lighting levels in that room were found to average 96 footcandles on the work plane at 61 watts with two fluorescent lamps, and 50 fc at 40 watts LED Testing & Application October 2009 with one fluorescent lamp (Table 1)—although the installation of a one-lamp ballast would drop the 40 watts to 30 watts. By contrast, the room averaged 63 fc at 29 watts with two LED replacement lamps per fixture, and 34 fc at 14 watts with one LED replacement lamp.
Subjective evaluation of the LED lighting was favorable: the color temperature of both the fluorescent and LEDs was the same (4,100 K); the color rendering of the LED appeared very good; and the distribution was the same (the luminaires using LED lamps created the same scallop effect on the wall that the fluorescent fixtures created).
However, the LED lamp could not be used in-line with the existing ballast, which therefore would have to be disconnected, adding labor cost. In addition, it was discovered that the workspace was over-lighted with two lamps—a very common condition—and disconnecting the ballast to switch to LED lamps would entail roughly the same labor costs as replacing the ballast with one designed for a single fluorescent lamp, which would save just as much energy.
That’s why it was concluded that switching to LEDs does not make good sense at DOL headquarters, and a full-scale Gateway demonstration will not be undertaken. LED replacement lamps cost between $40 and $130 each, compared with just $1.50 to $5.00 for a fluorescent lamp, and the lifetimes are comparable (24,000-46,000 hours for fluorescent, vs. a projected 50,000 hours for LED). The LED system missed the bull’s-eye because the lamp did not provide the same amount of illuminance and failed to save energy in comparison to the fluorescent lamps. In addition, the LED lamp costs a lot more, so it’s missing the mark on all counts.
Tom Pruitt, director of facilities management for the DOL headquarters building, recognizes that the cost of SSL is dropping as rapidly as the technology is improving, and says that just because SSL didn’t work out in this particular situation doesn’t mean DOL won’t consider it in the future.
A Partial Hit
Ohio vehicular tunnel
SSL fared somewhat better in the preliminary analysis for another potential Gateway demonstration, this one involving a traffic tunnel under construction in the Waterville, OH, area. Interested in green technology as well as the potential to save money and energy by using LED lighting instead of the proposed high-pressure sodium (HPS), the Ohio Department of Transportation (ODOT) asked DOE to make an assessment of its tunnel plans.
When it opens in 2012, the tunnel will have separate lighting systems for daytime and nighttime. Illuminance requirements during daylight hours are high in any tunnel, so that entering motorists can see well immediately without having to adjust to a lower level of light. High light levels are most critical at a tunnel’s entrances and can decrease toward its interior without adversely affecting motorist vision, but because of the relatively short length of this particular tunnel (335 ft), high daytime light levels must be maintained throughout its expanse.
In general, Gateway analysis for multiple installations has shown that there are few hard and fast rules of thumb—every potential application needs to be reviewed based on its own characteristics and merits to determine if LEDs will be a good fit. The ODOT tunnel installation proved no different. Although LED luminaires have proven competitive against 250-W HPS in other applications, they have a much harder time competing against 250-W HPS in this tunnel.
The very high illuminance levels needed for this environment can be achieved either by using high wattage fixtures, by increasing the number of fixtures, or through a combination of both. The design is governed by the desire to minimize the overall cost of the installation. HPS technology works well in this kind of application because it increases in efficiency as wattage goes up, while capital cost remains relatively the same, thereby eliminating one variable in the equation. In contrast, the cost of LED products increases proportionally with required lumen output; more LEDs typically require more heat-sink material and a larger housing, for example. In short, the higher the required output, the less cost-competitive LEDs become with respect to conventional products.
As a result, neither of the two daytime LED systems evaluated were found to be cost-effective for tunnel use, compared to 250-W HPS. Installing an LED daytime system instead of HPS would consume an additional 1,122 watts overall during the day and cost substantially more; therefore, it was not recommended by DOE at present—although it was acknowledged that LED technology is improving so rapidly that products available a year from now may be substantially more efficacious and cost-effective than those investigated for this study. ODOT has decided on an HPS system for daytime use in the tunnel.
Nighttime use is a different story, however. Tunnel lighting levels can be significantly lower at night than during the day, because there’s no competition from the sun, and motorists enter the tunnel from a dark environment. One obvious way to reduce tunnel light levels—using lower-wattage lamps—makes maintenance more complex, time-consuming and costly because it increases the variety of lamps that have to be kept in inventory. Reducing the number of fixtures would also bring light levels down but, in a tunnel environment, would risk causing “spatial flicker”—a strobing effect on passing motorists that can cause headaches, nausea and even seizures. A natural example of this effect occurs when trees along a road are planted apart on the same spacing. Bursts of sunlight between the trees can sometimes cause a similar spatial flicker for drivers. Consequently ODOT was leaning toward using the same high-wattage HPS lamps at night as would be used during the day, even though it would result in the tunnel being overlighted at night to levels three times those recommended by IES.
This overlighting was a key factor in DOE concluding that an LED nighttime system is a viable option for the Waterville tunnel, because many LEDs allow for easy control and adjustment of the lighting level. A second factor in favor of LEDs was maintenance; according to Aaron Behrman of ODOT, the state’s relamping schedule for HPS is three-to-five years, which is about half of the expected lifetime of LED products. And because LEDs fixtures look so different from HPS lamps, there’s no danger of the two types being confused, thus potentially adding to maintenance time and costs.
 DOE found that using LEDs at night instead of HPS would save 53 kWh per day and yield a simple payback of approximately eight years—assuming $0.11/kWh electricity cost, $45 per luminaire for relamping and hand cleaning, and no available incentive money (Table 2). Although this may be somewhat of an “apples-to-oranges” comparison because of the higher light levels produced by the HPS system, those higher light levels are unnecessary, so the savings are real. That’s why, according to Behrman, ODOT is still evaluating the nighttime options for lighting the Waterville tunnel, and LEDs are still in the running.
A Bull’s-Eye
Supermarket parking lot
Despite the cost factor, there are already some applications where LED products are not only in the running, but make good economic sense. A Gateway demonstration involving the parking lot of a Raley’s Supermarket in Sacramento, CA, provides a good example. The parking lot is lighted with pole-mounted, 320-W metal halide dropped-lens cobra heads. For this demonstration, headed by Pacific Gas & Electric’s Emerging Technologies program, roughly half of the existing fixtures were replaced with bi-level (dim- and high-state operation) 100-LED luminaires that were controlled by motion detectors so that light output was reduced when there was no activity in the vicinity for a period of five minutes.
In high-output operation, the LED luminaires matched the metal halide baseline for illuminance and provided a more uniform lighting distribution. In low-output operation, the LEDs delivered improved uniformity but, as expected, lower average illuminance than the metal halide baseline (roughly half)—though because this corresponded to times when no movement in the vicinity was detected, it exceeded the IES recommendations in RP-20. These quantified improvements were qualitatively supported by the results of a survey of 17 store employees, who rated the new lights significantly better than the old ones with regard to parking-lot appearance and, in more than one case, independently cited an increased sense of security with the LED lights.
With regard to energy use, the LED luminaires drew significantly less power than the metal halide luminaires—an average of 149 watts on high power and 52 watts on low power, compared with 346 watts for metal halide. On average, the LED luminaires were on high power 55 percent of the time, and on low power 45 percent of the time, resulting in a time-averaged demand of 105 watts and an estimated annual energy savings of 70 percent over the metal halide.
 When maintenance and replacement costs for metal halide luminaires were combined with energy costs, it was calculated that the bi-level operation LED luminaires, which were assumed to have zero regular maintenance cost over the course of their useful life, would each yield annual savings of approximately $278. The total incremental cost of an LED luminaire was calculated at $925.83 in a new-construction setting and $1,300 in a retrofit scenario, resulting in simple payback periods of 3.3 years and 4.7 years, respectively; the 15-year life cycle savings were $2,661 and $2,287 (Table 3).
According to Randy Walthers, Raley’s energy manager, the company considered the demonstration such a success that it has made a commitment to use LED lighting in all new stores and major remodelings where it has control of the parking lots. It should be noted, however, that the economic advantage and relatively short payback period for the LEDs in this application was mainly due to cutting down on spot replacement, which is the strategy used by Raley’s for replacing burnouts and significantly drives up maintenance costs. Raley’s uses a private lighting contractor for maintenance of the parking lot lighting and reports annual maintenance costs of $200 per luminaire per year, so the assumption that the LED luminaires will have zero maintenance is a key factor in their payback analysis. In other settings, where maintenance costs are lower, the payback would not be as favorable, and LEDs might not make as much economic sense.
That underscores an important point deserving of special emphasis. As stated before, it’s usually the benefits in addition to energy efficiency that make SSL economically viable at this stage. But as we’ve seen from the examples presented here, the value of these other benefits is very much site-specific and can’t be extrapolated from case studies that may involve completely different parameters and requirements. What this means is that there’s no substitute for due diligence and that someone on the evaluating team needs to be knowledgeable about SSL technology so that the right questions get asked, and the significance of the answers is fully understood.
October 09
|
