DOE workshop discusses LED success stories, labeling programs and quality issues (MAGAZINE)

Sept. 13, 2011
Presenters at the DOE SSL Market Introduction Workshop looked at the SSL success stories, analyzed different cost-reduction strategies and ways of reporting luminaire reliability, and examined the implications of the FTC Lighting Facts label. LAURA PETERS reports from Seattle.

This article was published in the September 2011 issue of LEDs Magazine.

View the Table of Contents and download the PDF file of the complete September 2011 issue.


The annual US Department of Energy (DOE) SSL Market Introduction Workshop took place in Seattle, Washington on July 12-14 and was attended by 275 industry professionals including lighting designers, lighting manufacturers, LED and driver manufacturers, utility companies, and energy-efficiency companies. Highlights of the workshop included several real-world success stories, a discussion of the new FTC Lighting Facts label and how it differs from the DOE Lighting Facts label, a new report on reliability, and ways of bringing down SSL costs throughout the supply chain.

Jim Brodrick, who runs the DOE’s SSL program, stated in one of his weekly Postings emails that the Workshop indicated just how far the SSL industry has come over the past year. He said that “while SSL is not yet suitable for every application, it’s become a definite force to be reckoned with.” The industry is clearly getting some of the help it needs – from government incentives to utility rebates to technical developments – to succeed in the replacement of incumbent, less-energy-efficient lighting options.

The maturing of the LED lighting industry can be seen in the increasing number of large-scale installations across the country. At the Workshop, the principal installers discussed the lessons learned from installing LEDs in two art galleries, the city streets of Seattle and US Navy parking lots.

Streets of Seattle

The home of the DOE workshop this year also happens to be a showplace for LED street lighting. At the workshop, Bruce Harrell, Council Member of the Seattle City Council, discussed the successful installation of over 6000 street lights by the publicly-owned utility Seattle City Light (see Fig. 1). The project is part of the plan to replace 41,000 residential street lights in Seattle by the end of 2014, a program that is already saving the city $300,000 per year ( Once completed, the city council estimates a $2.4 million reduction in operating costs will be achieved.

FIG. 1. Edward Smalley, council member of Seattle City Light, said the decision to install LED-based street lighting was the result of LED’s demonstration of the illumination performance, controllability and operational efficiency (48% energy savings) needed to satisfy the city’s lighting needs.

Council members were also swayed by the tremendous savings in maintenance costs. “Every two years, we would pay workers overtime to quickly replace the high-pressure sodium lamps before the winter came,” Smalley said. “Now that cost has been essentially eliminated.”

Smalley said that feedback from local law enforcement indicates the LED street lights make criminal activity easier to identify. He said this is due both to the better quality of the LED light and the elimination of “hot spots,” which are prevalent with HPS street lights.

Smalley is the director of the DOE Municipal Solid-State Street Lighting Consortium (, which has more than 270 members. Consortium membership is free and open to cities, utilities and investors involved in street- and area-lighting projects. The consortium builds a repository of shared field data to help accelerate the learning curve for LED street lighting.

Lighting in art galleries

Lighting designer Scott Rosenfeld had no complaints about the light quality in the Smithsonian American Art Museum and the Renwick Gallery, both in Washington, DC, but he wanted a more energy-efficient alternative. Rosenfeld, chair of the IES Museum Lighting Committee, explained the unique challenges associated with illuminating art. “Any light causes damage, so we use the smallest quantity of light necessary to experience what is relevant about the object,” he said.

Table 1. Rosenfeld turned to LED lighting after evaluating fluorescent and metal-halide lighting technologies and determining that they could not match the performance of the existing halogen products in the galleries. Fluorescent, in particular, would cause additional damage to artwork due to the ultraviolet light component.

Effective illumination of artwork involves precise control the lighting intensity, angle and distribution. Rosenfeld chose a mixture of LED wall-washing and spotlighting lamps, including replacements for PAR30, PAR38 and MR16 lamps, with beam angles from 4 to 54 degrees. Some lights had to remain incandescent, because there was no suitable LED alternative, for instance, for 250W incandescent spot and flood lamps used with the high ceilings (34 ft) of the Renwick Gallery.

Rosenfeld commented that the early part of the LED installation was dominated with flicker issues, especially with MR16 lamps. The museum overcame these issues by using magnetic ballasts.

The switch to LEDs in the two galleries is expected to reduce lighting power density in the Renwick Gallery from 3.9 W/ft2 to 1 W/ft2, and in the Smithsonian American Art Museum from 2.9 W/ft2, to 0.5 W/ft2,. Overall energy consumption in the two galleries is expected to decrease by 75%.

As part of this Gateway study, the DOE surveyed members of the viewing public to determine their preference for the quality of light among three LED flood-lamp options and a halogen lamp (see Table). One LED lamp scored particularly highly (lamp 2), though the halogen lamp also scored highly. This survey confirms what Rosenfeld has observed with the LED replacements: viewers discern little difference, if any, between a high-quality LED replacement and halogen lighting.

US Navy

Paul Kistler of the US Navy (Seattle, WA) shared his experience with changing out parking-lot lights from HPS lights to LED luminaires on naval bases in Ventura, CA and the naval station at Pearl Harbor, HI. In Ventura, the Navy replaced 14 400W HPS luminaires with 207W LED luminaires, achieving a 21% improvement in minimum illumination (lux) and a $1412 per year energy saving.

At Pearl Harbor, the Navy changed out 34 150W HPS luminaires for 104W LED luminaires, improving minimum illumination by 225% while still reducing the energy cost by $1280 per year. “We are mandated by Congress to reduce our energy use by five percent per year, and at the same time these lighting projects have become showpieces for the Navy,” explained Kistler.

Kistler added that the installations provide better lighting quality and security, and the Navy expects to improve the energy savings further through increased use of sensors and intelligent controls. With such projects, the US Navy looks for utility incentives, 5-year warrantees on the products they purchase, and luminaires with corrosion resistance, efficient thermal management and good ventilation.

He also promoted the value of checking LM-79 and LM-80 data for any products that are under serious consideration, checking DOE Caliper data for the class of products considered, and assessing whether lighting products can dissipate heat well.

Meeting aggressive cost targets

FIG 2. The industry has set very aggressive cost targets, including a 50% reduction in luminaire cost from 2010 to 2012 and greater reduction between 2012 and 2015 (see Fig. 2). As Fred Welsh, President of Radcliffe Advisors (Chestertown, MD) explained, luminaire manufacturers are moving toward a more integrated model, using fewer components to reduce assembly and testing costs.

A breakdown of packaging and material costs (see Fig. 3) indicates that cost reduction in several areas is needed. For high-volume luminaire production, new equipment is needed to maximize yield and consistency of manufacturing results. Welsh said that increased equipment throughput is needed with increased automation for MOCVD, improved testing and inspection techniques, and improved upstream process control. From the LED packaging side, simplified designs are required and the industry should move to wafer-scale packaging technologies.

Mike Watson, Senior Director of Marketing at Cree (Durham, NC), called for more radical change throughout the manufacturing supply chain. He suggested building a new ecosystem and new business models for the SSL market. In this ecosystem, financial-service companies should monetize the energy savings for the consumer. This can be provided through lighting and energy-control companies, lighting-systems integrators and leasing experts.

FIG. 3. Mark Hand, Director of New Products and Technology at Acuity Brands (Atlanta, GA), talked about the need for more realistic specifications to reduce cost. For instance, specifications of 50,000-60,000 hours may be reasonable for applications that run 24 hours a day, 7 days a week, but he said that 10,000-20,000 hours may be reasonable for a majority (80%) of applications. This could allow the LEDs to be run hotter, either yielding a higher light output or reducing the cost of the thermal-management components.

Hand also advocated delivering only the CRI that is necessary for the application, and suggested that customers specify only the necessary level of restrictions on binning to reduce cost.

Quality of light

A recurring theme in several workshop presentations was the need to assess the quality of light from SSL. Maria Thompson, Principle Applications Scientist at Osram Sylvania (Danvers, MA), said that the industry is making rapid progress in performance, having already broken the 200 lm/W efficacy barrier that was set as a target for 2020. However, the potential for delivering improved color quality is essentially untapped.

Thompson described the limitations of CRI. “CRI, which is a color fidelity metric, reflects the ability to reproduce colors as close as possible to those perceived under sunlight, or blackbody radiation,” she said. “However, CRI is based on eight unsaturated colors, and it not only doesn’t account for color saturation, but it actually penalizes for color saturation.”

Thompson explained that color distortions with increased chromatic contrast (increased color saturation) are typically preferred by viewers. The Color Quality Scale (CQS) developed by the National Institute for Standards and Technology (NIST, Gaithersburg, MD), does take into account color saturation, though it also is an average value, like CRI. Given the limitations of both CRI and CQS, Thompson suggested further investigation into improved metrics for general lighting applications for SSL white light.

Chad Stalker, Regional Marketing Manager of Philips Lumileds (San Jose, CA) took a slightly different approach to defining the quality of light, and discussed the different color-rendering needs in terms of application. He compared food displays, which require excellent color rendering, versus street lighting, which requires lower CRI. He explained that LEDs require color consistency in five areas:

  • Color spectrum and rendering: Measured by CCT, CRI, R9, CQS
  • Color consistency between sources: <3 SDCM (standard deviation of color matching)
  • Color in application: Measured at 85 degrees junction temperature
  • Color consistency in beam: <10 points on u'v' scale
  • Color consistency over lifetime: <5 DSCM
  • Stalker advocated testing and binning LEDs at their operating conditions to reduce costs, simplify the design process and raise purchasing confidence.

    Practical color variations

    Daniel Salinas, a lighting-systems designer with Nelson Electric (Seattle, WA) provided feedback from the perspective of a designer and installer of lighting fixtures. He said that color variance for an overall project can be minimized by purchases products from a small number of luminaire manufacturers. To simplify installation and maintenance, he also suggested minimizing the number of control technologies and drivers used on a project. “I recommend that the designer and the customer establish an acceptable level of color variance or color shift, make that a part of the binning requirements, and include it in the contract agreement upfront,” he said.

    Salinas also suggested that lighting samples be made available during the design process and that these be tested together with the materials of the room (wood, carpet, artwork, etc.) to ensure customer satisfaction.

    Salinas defined project life as typically 5-7 years, and said that during that time, the light fixtures will need to be cleaned and maintenance will typically be needed on drivers, so accessibility to both is important. He added that adequate ventilation for LEDs is critical. Finally, he said that luminaires should be checked for UL listing before the project begins.

    Update on Lighting Facts labeling

    FIG. 4.Effective January 1, 2012, the US Federal Trade Commission (FTC) will require that manufacturers of medium screw-base bulbs (table-lamp-size bulbs), whether incandescent, fluorescent or LED-based, include the new FTC Lighting Facts labels on the front and back of the bulb’s packaging. Also, the bulb itself must contain the light output (in lumens) and the fact that it contains mercury (if it does).

    The US FTC mandates that these labels (see Figs. 4 and 5) appear on products sold on and after January 1, 2012. The front label contains light output (lm) and energy cost ($/yr), while the back FTC Lighting Facts label contains light output, energy cost, lifetime estimate (years), appearance (CCT), required power (W) and presence of mercury.

    Until very recently, the lighting industry has used the voluntary DOE Lighting Facts label. The DOE has stated that its label is not in conflict with the new FTC Lighting Facts label, and will still be used by lighting professionals, utilities and retail buyers. The FTC label strictly targets consumers. However, the DOE will no longer use its label on medium screw-base products after January 1.

    FIG. 5. A notable difference, however, is that the DOE Lighting Facts label program has attempted to control the label’s content by encouraging third-party testing and verification at certified laboratories. Though this action was voluntary, Lighting Facts partners had a greater tendency to publish data that was independently verified with LM-79 test data than non-partners.

    Now, the FTC labeling is required. Unlike the DOE, the FTC in principle has the power to ensure that manufacturers only publish accurate metrics on the label. Only time will tell whether the FTC will take action against manufacturers that publish false performance metrics.

    Luminaire reliability

    The SSL Quality Advocates, a working group formed jointly by the US DOE and the Next Generation Lighting Industry Alliance (NGLIA), announced a new report at the workshop entitled “LED Luminaire Lifetime: Recommendations for Testing and Reporting.”

    Members of the working group included Fred Welsh, Terry Clarke, CEO of Finelite, and Steve Paolini of Lunera Lighting Inc. Welsh pointed out that many people in the industry still ascribe lumen depreciation to luminaire lifetime. In a simple example, he explained that lumen depreciation of an LED might lead to lifetime estimates of 60,000 hours, but when driver lifetime (55,000 hours) is taken into account (failure probabilities are multiplied), the lifetime of the system will be more like 52,000 hours.

    The working group recommends defining the standard luminaire lifetime as the time when half the product population falls below 70% of initial light output for any reason. Additional specifications might be provided for color shift, such as an amount of color shift after a measured period of time.

    Clarke emphasized that color shift in LEDs is significant, complex and not well understood. He added that not all color shift comes from the LED itself: different luminaire designs impact color; different optics and designs can age differently; different environmental operating conditions may cause color shift; and the luminaire may or may not use active color management. Though the working group could not come to an agreement on ways to characterize color shift as a metric of reliability, they consider it a significant factor in luminaire performance.

    Paolini talked about the value of temperature cycling and highly-accelerated life testing – designed to identify the weakest component in the luminaire – to enable design improvements. “These tests can help you improve the design, but only long-term reliability testing at normal use conditions will give you an indication of how long the luminaire will last.” For more on the recommendations, see