|
|
The number of solid-state lighting (SSL) products being sold on the market continues to increase at a rapid pace. Performance barriers fall on what seems like a regular basis, with overall performances climbing steadily upwards while costs drop. But it’s easy to forget that the technology itself is still quite new and that, in many respects, SSL remains a largely unknown quantity, with a host of questions that have yet to be answered. A major reason for this is the fact that LED lighting products simply haven’t been around long enough for us to see how they perform in real-world settings over time. For example, many LED luminaires carry lifetime claims of 50,000 hours, which works out to nearly six years of continuous, 24/7 use. But the technology remains too new for any LED lighting product to have been on the market anywhere near that length of time.
As a result, glitches of one kind or another are bound to emerge as the body of field experience grows. These growing pains should be expected, because they’re a normal part of the development of any new technology. Theory can only take you so far, after which a real-life proving ground is needed to identify issues that may not have been anticipated in the confines of the laboratory. Such issues are typically addressed by the manufacturer in successive rounds of fine-tuning, until we finally have a product that’s relatively glitch-free.
Three recent Gateway demonstrations (www.ssl.energy.gov/gatewaydemos.html) provide a glimpse of this multilateral learning process and show why, at this early point in the development of SSL, it makes sense to take a measured approach to its implementation rather than to plunge in over one’s head with one big, enthusiastic leap. What we’re learning about the technology at times challenges our most basic assumptions and is instrumental in shaping how we manufacture, install and even use SSL products.
FDR Drive, New York City
A Mixed Bag in the Big Apple
A
s part of a municipal effort to reduce energy use by 30 percent, the New York City Department of Transportation (NYCDOT) is in the process of evaluating its street lighting. Noting that recent roadway and street lighting demonstrations in other cities (Minneapolis, MN; Oakland, CA; and Portland, OR) have yielded energy savings of between 12-40 percent for SSL products, NYCDOT installed a number of LED luminaires on the Franklin D. Roosevelt East River Drive (FDR Drive) in August 2009 to see how they would perform.
The FDR Drive is a freeway that runs along the East River for 9 miles in Manhattan and is three lanes in each direction for most of its length. Vibration from vehicles along its elevated sections causes a high rate of premature failure in the freeway’s incumbent high-pressure sodium (HPS) lighting. Because of this, it was hoped that the vibration resistance of solid-state lighting would result in significant maintenance savings.
Luminaires from four SSL manufacturers (BetaLED, eLumen, LED Roadway and LSI Industries) were selected for evaluation in the FDR Drive demonstration. Six LED luminaires from each manufacturer were installed back-to-back on three consecutive poles mounted to the median barrier and spaced ˜165 ft apart, for a total of 24 LED luminaires on 12 poles. A single pole with two 150-W (nominal) HPS luminaires separates the LED luminaires of each manufacturer from those of the other two.
Of the four different LED products installed, three were found to exceed the average initial illuminance levels of the HPS system (Table 1). However, one of those three raised concerns pertaining to the continued depreciation of light output expected over the course of an LED luminaire’s lifetime. Because LED luminaires typically don’t burn out but instead slowly fade over time, their lifetime is defined as lasting until light output falls below 70 percent of initial levels (expressed as “L70”). In contrast, the light output of an HPS luminaire typically depreciates just 10-15 percent over the lamp’s life. Thus, LED luminaires used in roadway applications must emit more light at the outset than correctly sized HPS luminaires in order to maintain acceptable illuminance levels over the full course of their expected lifetime.
All of the LED products installed on the FDR Drive used less energy than the existing HPS system–anywhere from 25 to 50 percent less– but in two cases this was at least partly offset by lower illuminance levels. Despite the usual superiority of SSL in terms of uniformity, in this case none of the LED products were found to be as uniform as the incumbents in terms of light output, although two came close. In all likelihood, this was due to the fact that merely substituting an alternative product (whether LED or conventional) into a lighting system that was designed for another type of product, without making corresponding changes in the system’s design, will often result in diminished performance.
Another lesson learned from this project had to do with light measurement. The two meters used in the FDR Drive demonstration yielded results that differed significantly from one another. This somewhat surprising finding raises the issue of instrument accuracy and suggests that the results obtained in studies of this kind could be largely dependent on the meter used. The inaccuracy of the meter becomes increasingly important at lower illuminance levels, which is especially relevant to SSL for two reasons: 1) LEDs are gaining traction in exterior applications, where illuminance is much lower than in interior environments; and 2) no automatic end-of-life mechanism exists for an LED luminaire, and thus the reduced illuminance will have to be verified at some point in the future to confirm that it is still meeting the needs of the application. DOE will be taking a closer look at this issue.
In the eight months that these LED products have been installed on the FDR Drive, problems have surfaced in three of the 24 luminaires, each from a different manufacturer. One product was cycling on and off and was subsequently replaced, and two failed due to water incursion after a heavy rainstorm and were also replaced. Water incursion is a familiar issue with street lighting, regardless of the type of light source, and some SSL manufacturers are still learning how to deal with it, as well as with the power supply, in designing their luminaires. What we’ve learned from demonstrations like the one on the FDR Drive, as well as from studies in the lab, is helping us get a better understanding of the causes of failure in SSL luminaires. We’re used to thinking of luminaire failure in terms of the light source, but with SSL many other factors come into play. These factors may even change over time, as manufacturers make adjustments to their products in response to feedback from the field. Question to Ponder: Can LEDs match HPS on roadways?
WALMART, LEAVENWORTH, KS
A Whole Lot of LEDs
In what could possibly be the largest LED site-lighting installation in the U.S. to date, Walmart
installed 92 pole-mounted LED luminaires made by GE in the 500,000-sq ft parking lot of a new
Supercenter store in Leavenworth, KS, in May 2009. This full-scale installation was preceded by
several smaller pilot demonstrations of the technology at other Walmart locations. These yielded
promising results, including improved facial recognition and an enhanced feeling of security, in addition
to energy and maintenance savings.
The customary lighting design for the parking lot of a Walmart Supercenter employs 1,000-W pulsestart
metal halide (PMH) luminaires. In this case, the default lighting design also included 22 175-W
wall-mounted PMH luminaires to supplement the pole-mounted lighting. The exterior lighting systems at
Walmart Supercenter stores operate from sunset to sunrise, 365 days a year, which works out to more than
4,000 hours annually. At this level of usage, Walmart relamps the PMH luminaires every two years, which
results in significant maintenance costs, thereby creating a situation where the longevity of SSL can be an
important consideration.
In addition to investigating the use of LEDs, Walmart took the extra step of designing the system to
meet the LED Site Lighting Performance Specification published in May 2009 by the DOE Commercial
Building Energy Alliance (CBEA). Strictly speaking, the standard 1,000-W PMH system exceeded the
specification’s allowable lighting power density of 0.06 watts per sq ft for the lighting zone applicable to
the site, and thus did not represent an entirely appropriate baseline with which to compare the LEDs.
As a result, designs for two PMH systems (1,000-W and 400-W) were reviewed via computer simulation.
In the end, neither PMH design was found to comply with the CBEA specification. The lighting
power density for the PMH systems are 0.11 watts per sq ft at 1,000 watts, and 0.07 watts per sq ft at 400
watts, whereas the LED system installed at the Leavenworth site draws 0.04 watts per sq ft, making it
the only system that meets the CBEA specification.
The LED system equaled or surpassed the performance of the two PMH systems in other ways. The
exceptional dark-sky performance allows for adherence to light trespass ordinances and keeps the
residential neighbors happy, something much more difficult to accomplish with high-intensity discharge
sources. Customers at the Leavenworth Walmart reported that they liked the LED lighting and
found it brighter, more reflective and more focused on the parking spaces than the lighting they were
used to.
The light output of the LED system (Table 2) was found to be more uniform and controlled than
that of the PMH systems, and it cut down considerably on energy use–44 percent less than the 400-W
PMH, and 63 percent less than the 1,000-W PMH. In addition, the LED system reduced projected maintenance
costs significantly–by 65 percent compared with the 400-W PMH system and by 45 percent
compared with the 1,000-W PMH system (the PMH maintenance numbers vary because the 400-W
system required more poles and luminaires). As a result, opting for LEDs is projected to bring dollar
savings, despite a substantially higher first cost, and has a projected payback of less than six years–
without even taking into consideration other possible incentives.
Because energy and maintenance costs in Kansas are lower than they are in most other parts of the
country, Walmart’s Leavenworth demonstration represents a kind of acid test for the applicability of
LED site lighting as well as the related CBEA specification. In an area where energy and maintenance
costs are higher, the economic advantages of the LED system would be more pronounced and the
payback period shorter. Still, there are other factors–such as parking lot shape and size, local light
ordinances and acquisition costs–that go into determining whether LED site lighting is the right solution
for any given location.
One lesson that was strongly reinforced by the Gateway demonstration at the Leavenworth Walmart
was that field measurements of lighting systems don’t always match the computer simulations, despite
the best attempts of the lighting design team. For instance, while taking field illuminance measurements
at the new installation, Walmart discovered that a pole was located 13 ft off its intended location
in the site plan. Most site owners don’t bother taking such measurements post-installation, and if
Walmart did not have a policy of requiring them, this disparity might never have been detected, even
though it greatly impacted the illumination achieved in the corresponding area. Some other CBEA
retailers have already taken notice of the Walmart experience and are implementing field measurements
as a kind of double-check procedure. Question to Ponder: How will regional differences in
energy and maintenance costs influence LED site lighting?
Field Museum of Natural History, Chicago
Getting on Track
Chicago’s Field Museum of Natural History, one of the country’s largest and oldest research museums, has been looking to increase sustainability as well as maximize the preservation of its artifacts. In October 2009, the halogen track lighting in the museum’s Booker T. Gallery was replaced with 26 LED products. The gallery is a long and narrow space with a 14-ft ceiling, and at the time of the installation it housed an exhibit of the entries in a design competition for a memorial honoring noted architect Daniel Burnham, whose design plan for Chicago has guided the city’s growth over the past century.
The gallery’s halogen track heads–which had housed 35 luminaires with PAR 36, PAR 38 and MR16 lamps–were replaced with LED-dedicated track heads. The LED luminaires were made by Lighting Services Inc. and received recognition in the 2009 Next Generation Luminaires competition. They featured Xicato light engines, which have good color quality and consistency, both of which were important considerations for the museum. Another concern for the museum was ultraviolet light, which can damage artifacts and thus must be filtered out when conventional halogen lighting is used, but which is not emitted by LEDs.
Ambient conditions can also hasten damage to artifacts, and thus museums are often cooled to 65-68 deg F to avoid the high temperatures and humidity that are the primary contributors. However, the portion of lighting power that’s converted to heat rather than light is a direct burden on the cooling load—the general rule being that 3 watts of lighting adds 1 watt to the HVAC system. Therefore, higher-efficiency light sources that also provide good color and can be dimmed offer museums a win-win scenario.
In addition, the Field Museum was interested in the long life of SSL to help cut down on maintenance. The importance to a museum of maintaining the accent lighting that highlights specific objects on display means that any lamp failures must be addressed as soon as possible–sometimes immediately. What with the logistics of maneuvering a lift or ladder into place among exhibits and visitors, and the sheer number of lamps in use (approximately one every 5 sq ft), maintaining lamp operation can be extremely costly and time consuming, and museums often have employees devoted exclusively to that task.
For the most part, the LED system met or exceeded the illuminance of the halogen lighting it replaced in this application (Table 3). The new system was found to produce more lumens, and in some cases a slightly wider beam angle, thus allowing for a reduction in the number of luminaires. What’s more, the LED system drew 63 percent less power than the halogen system (335 watts vs. 894 watts) and thus used 63 percent less energy (963 kWh vs. 2,603 kWh, figured over the course of a year).
Considering that the new technology was more efficacious (45 lumen per watt vs. 13 lumens per watt) and produced twice the lumen package (1,000 vs. 500), one would expect an even greater energy savings, but in this case those savings were mitigated by two factors: accent lighting needs and baseline consumption related to dimming. Accent lighting in a museum setting generally requires at least one luminaire per artifact, so doubling the lumen package doesn’t translate into halving the number of fixtures, as it might in other applications. As for dimming, there is a minimum level of energy use required by the LED driver and the halogen system transformer, regardless of lamp wattage. This minimum draw makes the potential savings from dimming LEDs asymptotic–rather than roughly linear, as is the case with a halogen system. At the Field Museum, dimming the LEDs to 15 percent of their maximum light output only reduced overall system power consumption by about 50 percent, which was the result of this particular driver-dimmer combination. This issue is not inherent to LED systems. Driver design, dimmer design and a suitable matching of the two are central to achieving optimal performance. The precise explanation of the dimming inefficiency in this case is unknown. DOE will continue to examine this issue.
In addition, the correlated color temperature (CCT) stayed relatively constant when the LEDs were dimmed, in contrast to halogen lamps, which become warmer in color temperature as they’re dimmed. In a survey of museum staff before and after the installation, a number of respondents said the LEDs made the gallery feel cool, although there was overall satisfaction with the LED lighting. The difference in color temperatures is more pronounced in the dimmed state used in the gallery, and at least one respondent suggested that the LEDs might be made to shift towards the red spectrum as they’re dimmed, so as to match the halogen products. This kind of user feedback can be helpful to manufacturers in their product development. Question to Ponder: Do the specific requirements of museum applications minimize some of the typical benfits of LEDs?
June 2010
|
