"In the LED industry, historically, we have done a 10X improvement in lumens per dollar per decade and a 20X improvement in lumens per package," says Willem Sillevis-Smitt, director, strategic marketing and business management at Philips Lumileds (San Jose, CA). "This pace is picking up right now. Our ambition is to 10X improvement in five years. That's what we think is possible."
"There are lots of things that have come together," says Vrinda Bhandarkar, analyst, LED Practice, Strategies Unlimited (Mountain View, CA). "The utilities want energy efficiency because they don't want to invest in new power plants. LEDs have the right technology."
There are multiple approaches to producing white light with LEDs. One method involves mixing the output of red, blue, and green LEDs. It can produce a good quality white light but involves more components, increasing both cost and points of failure. A bigger issue is the fact that the efficiencies of the different colors of LEDs vary, as do their responses to temperature and aging. For best performance, they must be electrically balanced.
An easier approach uses phosphors to down convert the output of an LED, typically a blue gallium nitride (GaN) device. Yellow phosphors such as yttrium aluminum garnet (YAG) down convert part of the output from the LED and mix it with the primary emission to produce a white light (see Figure 1).
Figure 1. In a typical solid-state lighting bulb, yellow phosphor down converts a portion of the blue light from a gallium nitride LED, mixing it with the primary a mission to produce white light. (Courtesy of Cree Inc.)
Of course, there are many different colors of white. The incandescent bulb does a very good job of mimicking the solar spectrum and providing accurate color representation. Like fluorescent lights before them, first-generation white-light LED designs produced a very cool white light. The typical color rendering index (CRI) was 70 to 75, corresponding to a correlated color temperature (CCT) of around 4000K (see sidebar). That tone doesn't provide accurate color rendering, nor does it supply a comfortable tone. Innovations like adding a second, more greenish-yellow phosphor to the mix can improve the CRI, but not to the warm white tones that match what people expect from indoor lighting. To achieve warm white tones, manufacturers need to broaden the spectrum by adding red light to the mix. This can be performed two ways. Perhaps the easiest is to add a second phosphor that downconverts blue light to red. The approach offers the advantage of simplicity and manufacturability, although the Stokes losses inherent to the wavelength conversion process reduce overall device efficiency.
An alternative is to add another LED to the device to generate red emission directly. The approach avoids additional Stokes losses, but the red LEDs have inefficiencies and power demands of their own that partially offset the benefits. These designs also present some of the same challenges as the RGB approach-greater complexity and the need to electronically balance the outputs of the different LEDs. Depending on the approach, adding red to the mix can raise CRIs as high as 85. As in all engineering, the choice of which approach to use is a matter of trade-offs based on the application. A cost-sensitive application with forgiving performance demands may be best served by the phosphor approach whereas if emission is critical, adding a red LED to the mix may be the right choice.
Color variation presents another issue solid-state lighting must overcome. The color performance of LEDs differs not only from batch to batch but from device to device. Classical LED users are aware of this variation, which manufacturers typically solve by testing and binning devices. Users accustomed to getting identical emission from any 60-W incandescent bulb they pick up may be less forgiving of these inconsistencies. Indeed, tackling the color performance issue is one of the primary goals identified in the U.S. Department of Energy's updated Solid-State Lighting Research and Development: Manufacturing Roadmap. This need for consistency, reliability, and desirable quality of light underpins the recent efforts behind the development of standards for the performance, testing, and lifetime of LED luminaires.
Crossing Market Barriers
High efficiency and long lifetime have been the twin calling cards of LED lighting. Compared to the luminous efficiency of incandescent bulbs (around 15 lm/With) and compact fluorescents (around 50 lm/With), white light LEDs boast impressive numbers - manufacturers have reported chip-level efficiencies as high as 208 lm/W in the lab. Die performance is far from the whole story, however. Unlike incandescent, fluorescent, and halogen designs, LEDs cannot simply be put into a fixture. "That's the most difficult part for LEDs," says Bhandarkar. "You start with a 100 lm/W LED and you add inefficiencies because of the driver, the phosphor, the optics. The output of the luminaire drops to about 60 to 65 lm/W." Heat degrades device performance and limits lifetime. Unlike incandescent bulbs, which radiate heat out over the front face of the glass envelope, LEDs generate most of their heat out the back end of the device. As a result, thermal management becomes crucial. LED bulbs are more likely to show up in purpose-built fixtures than as bulbs that can be threaded into your existing lamps the way compact fluorescents can. This approach allows lighting designers to engineer the entire structure for heat management. In the case of retrofit bulbs to replace incandescents, manufacturers are tackling the issue with passive heat sinks (see Figure 2).
Figure 2. To dissipate the heat generated on the backside of the chips, LED bulbs feature various types of heatsinks. (Courtesy of Philips Lumileds)
Perhaps the biggest issue that solid-state lighting is up against it is that of cost. High-end LED bulbs that provide quality CRI can cost as much as $70 or $80, compared to $0.25 for a typical incandescent bulb. "We've been used to a commodity price of for incandescent lights," says Bhandarkar. "Trying to compete with that is going to be difficult. There's a whole paradigm shift that has to happen. People have to learn that not needing to change light bulbs is an added utility. Lighting has to go from a commodity to a system, which requires investment" Cost is relative, of course. A factory running a 24/7 manufacturing line can lose hundreds of thousands of dollars for a few hours of stoppage while overhead lights are switched out. Commercial businesses like hotels with lights on around the clock will happily pay more to ensure customer safety and usability while spending less on maintenance. "Or if you run a fast food restaurant and your margins are low and you don't want to pay for the electricity," says John Edmond, director of Advanced Optoelectronics and co-founder of Cree Inc. (Durham, NC). "It's not just the longevity, it's the electricity use. [Solid-state lighting] is 85% more efficient than an incandescent, it's 50% more efficient than a CFL, so you get that and you get the lifetime." In the case of Wal-Mart stores, for example, ceramic metal halide lamps used in the produce department caused fruits and vegetables to degrade more rapidly. Even though solid-state lighting cost more upfront, it ultimately paid for itself by eliminating waste.
Surprisingly, the single biggest growth area right now for solid-state lighting is outdoor roadway illumination. The technology that was once only bright enough for flashlights has grown up. The city of Los Angeles, for example, last year announced plans to retrofit approximately 140,000 street lights with LED fixtures. Elsewhere, the devices illuminate parking lots and garages. Although the cost per bulb is higher, the reduced cost of ownership speeds return on investment. "The payback is great, the energy savings is great, and the color is tremendously better," says Edmond. "They're used to the CRI of a sodium vapor bulb, which is 26 or 27. You just put in a cool-white LED with a CRI of 65 or 70 and its beautiful light."
Ultimately, of course, the consumer market beckons. One of the big concerns in the industry is buyers will get turned off by purchasing bad product that fails prematurely. Inexpensive LED bulbs, which still retail for 10 or 20 times the price of an incandescent, are probably poor quality and destined for premature failure-assuming they work at all. "There have been bulbs you can buy at warehouse clubs for a few dollars and they're just garbage," says Edmond. "I bought five of them once about a year ago and none of them lit up, none. That was the intro to LED bulbs and it was just a joke. Now, they're starting to get some nice bulbs out there."
Indeed, this year may well be a tipping point, as evidenced by the recent release of a dimmable 9-W A19-style bulb by Lighting Science Group (Satellite Beach, FL), retailing for under $20. At 400 lm, it provides a viable replacement for a typical 40-W incandescent. Manufacturers such as Philips and GE have also fielded entries (see Figure 3). "There are going to be light bulbs with very high efficacies that are going to be introduced to the market this year," says Bhandarkar. "Right now, it is not a question of whether LEDs will be part of the general lighting market; they are going to be part of the general lighting market. The question is how long the conversion will take."
Figure 3. Designed as a replacement for conventional incandescent polls, this LED bulb features a standard A19 form factor, as well as cooling fins around the base and up on the glass envelope. (Courtesy of Philips Lumileds)
"You're already starting to see the first signs of good bulbs at a reasonable price levels for consumers," says Sillevis-Smitt. "Last year at Light Fair, people were introducing their 40-W replacement bulbs. This year people were announcing their 60-W bulbs which is in the range of 800 lm. I think sometime next year it will really start to pick up and that's when you can expect consumer adoption."
Tidak ada komentar:
Posting Komentar