Can LED lamps beat the products they replace?
Panelists at the US DOE SSL Workshop discussed a Caliper study on LED replacement lamps purchased at retail, which found the LED lamps generally did not meet the performance of the benchmark lamps.
The report, based on a Caliper (Commercially Available LED Product Evaluation and Reporting) study, is available for download on the DOE SSL website.
Standard LED replacement lamp testing and analysis revealed “great differences between the best and the worst performing lamps,” said Kelly Gordon, of the Pacific Northwest National Laboratory (PNNL, Portland, OR). The DOE tested 33 products that were anonymously purchased from eight commercial retailers and ten different manufacturers, finding that most of the replacement lamps failed to meet basic performance parameters for the incandescent or halogen lamps they were benchmarked against. Gordon and others on the panel noted that LED performance is continually improving, and newer models are much more likely to meet such targets.
PNNL’s Gordon was joined by Jason West of D&R International (Silver Spring, MD), an environmental consulting firm, and Marc Maldoff of the home improvement retailer Lowe’s (Mooresville, NC). West provided an update on Lighting Facts activities, while Maldoff relayed his experience with consumers and solid-state lighting (SSL) products.
Retail lamp results
The DOE’s Caliper program, which compares actual product performance to the manufacturer’s claims and incumbent products, has evaluated over 400 products to date. However, many of these lamps are not available on the retail market. This specific study that Gordon described was undertaken to evaluate LED replacement lamps that are available to the public through retail stores.
Samples of 33 LED replacement lamps, three units each, were first evaluated for light output, efficacy, power factor, chromaticity and color rendering. The lamps were then operated for 1000 hours and tested again. They were tested using the IES LM-79-08 standard (Electrical and Photometric Measurements of Solid-State Lighting Products). Measured product performance was compared with the manufacturer’s claims on the product packaging. The initial and 1000-hour test data sets were compared to provide an indication of the likelihood that products will meet their lifetime claims.
A range of types of SSL lamps were purchased that mimic conventional technologies in shape and size, including five A19, four B10 (candelabra), two C7 (night light), eleven MR16/PAR16, four PAR20, and seven PAR30. The authors of the report warn that this study represents only a sampling of the available replacement lamps and should not be used as a judgment of specific products, retailers or manufacturers.
Only a few of the products met or came close to meeting the average light output of the benchmark products. Two of the five A19 lamps came close to meeting the output levels of a 40W incandescent A19 lamp; four of the eleven SSL MR16/PAR16 lamps came close to meeting the light output levels of a 20W halogen MR16; none of the four PAR20 SSL lamps achieved the light output of a 35W halogen PAR20; and three of the seven SSL PAR30 lamps came close to meeting the light output levels of a 50W halogen PAR30 lamp.
|CRI vs. CCT|
For instance, most products are marketed as replacements for incandescent products, implying a CCT of 2700-3000K. However, ten products had CCTs above 4000K. Thirteen products had CCTs above 3000K, perhaps making them a colder white than might be acceptable, while 16 were in the acceptable 3000K region. When all aspects of color quality were considered (CCT, CRI, and/or Duv), over half of the products had one or more unacceptable metric.
Fourteen of the 33 products had a power factor above 0.80. The other products had a range of power factors from 0.29 to 0.70.
The manufacturers’ claimed product lifetimes were in the range of 12,000 to 50,000 hours. After 1000 hours, three products operated at less than 70 percent of initial light output, while 12 other products exhibited less than 97 percent of initial light output. After 1000 hours, the remaining eighteen products exhibited between 97 and 100 percent of initial light output, with the greatest chance for meeting lifetime expectations. Notably, four of five Lighting Facts-labeled products performed at or near the original light-output levels after 1000 hours.
When lamp performance was grouped by retailer two retailers featured lamps with better performance overall, indicating that certain retailers may be implementing stricter screening procedures on the lamps they sell.
Packaging claims for most of the products included light output (lm), CCT and lifetime. The rated claims for light output and CCT were close for most of the products. However, when equivalency claims were provided (such as “5W = 40W light output”), they were inaccurate. The products did not provide the equivalent output.
The DOE suggests that consumers use data from the Lighting Facts label or LM-79 testing rather than manufacturer’s equivalency claims. They recommend improved education regarding color quality, light output, and the Lighting Facts and Energy Star programs to allow more informed purchasing decisions.
Lighting Facts next steps
|Details of the Lighting Facts label.|
Jason West spoke briefly about a new Lighting Facts Quality Assurance program that is being developed to help confirm the values listed on the DOE Lighting Facts label by the manufacturer. Currently, the manufacturer supplies LM-79-verified performance data to the DOE when it applies for a Lighting Facts label. The data supplied and shown on the label includes the light output in lumens, power required to light the product (W), efficacy (lm/W), CRI, CCT (metrics which are tested with the industry standard testing procedure (IESNA LM-79-2008) and verified by the DOE before the label is granted), model number, lighting fixture type, and brand,. These values are only monitored randomly using the Caliper testing program and the new Lighting Facts Quality Assurance program.
The new QA program will require compliance of four out of five samples of the lamp or luminaire within tolerance levels of the values listed on the LF label (the rating). According to West, if any rating proves to be exaggerated and not within standard tolerance levels, manufacturers will be asked to update their label.
The DOE finalized the details of the program earlier this year, and is currently undergoing the first round of procurement and testing. Lighting Facts selects the laboratories that perform the testing based on technical credentials, availability and cost to the program. More details can be found in the Lighting Facts Partner Participation Manual, page 20.
Lowe’s, a Lighting Facts partner since 2009, requires that all its LED suppliers be Lighting Facts partners, provide LM-79 testing on all their products, and use Lighting Facts labels. Marc Maldoff discussed the vast improvement in LED performance from 2010 to 2011, stating that Lowe’s highest performing A19 bulb went from a CRI of 82 to 90 and light output levels of 350 lm to 810 lm.
“As lumens doubled on several LED replacement-lamp types, the retail prices stayed about the same, something that customers may not be fully aware of,” Maldoff said. From a retail standpoint, he emphasized the need for education among consumers, and said, “beyond shifting the discussion from watts to lumens, which is very important, customers also need to understand the true cost of light, which includes the electric bill, replacement bulb costs and the hassle of replacement.”
During the question and answer period that followed the panel’s presentations, panelists were asked if higher price tended to correspond with higher-performance lighting products. Maldoff said that it definitely did.
Another audience member asked Kelly Gordon whether the DOE was considering including the new color quality scale (CQS) developed by the NIST for inclusion on the Lighting Facts label. “We would love to include it when the industry is ready and has decided on the best implementation. At this point, CQS has not been finalized or standardized,” she said.