«Solid-State Lighting Product Quality Initiative SECOND EDITION JUNE 2011 Next Generation Lighting Industry Alliance with the U. S. Department of ...»
The working group recognizes that these measurements are expensive and time consuming, and not all manufacturers may have the ability to comply when the product is first introduced. If the recommended minimum confidence level is not achieved, manufacturers may prefer to warrant performance for a specific period of time instead of making a specific lifetime claim. In that instance, the working group recommends stating the number as a “warranted lifetime,” as distinguished from a substantiated lifetime estimate.
When specifying a product’s L70 lumen maintenance claim of B50 lifetime, or any other value, the manufacturer should validate the product life not only upon first release, but also in the event of any changes to the product during its lifetime. Such changes may relate to design (which includes product variants), component, subsystem material, supplier, or manufacturing process—in short, anything that may affect product life.
The rest of this section addresses some of the necessary steps to demonstrate a product’s lumen maintenance life claim by separately considering a new platform, and then how to deal with variations in that platform. While some examples are shown, they are not intended to specify a standard procedure. Responsibility for specific standards lies with standards organizations and is beyond the scope of this guide.
NEW PLATFORM LUMEN MAINTENANCE
The preferred tool for measuring performance of LED systems is IES LM-79-08 Approved Method:
Electrical and Photometric Measurements of Solid-State Lighting Products. LM-79 prescribes uniform test methods under controlled conditions for photometric and colorimetric performance as well as electrical power measurements for LED fixtures as they would be manufactured for production. A key systemic element of LM-79 is that the LED fixture must be tested using absolute photometry which measures LED performance in situ.
For a new platform it is the manufacturer’s responsibility to demonstrate life performance compliance by testing luminous flux, in accordance with LM-79, in a sufficient sample of product for a sufficient amount of time to have confidence in the lifetime figures.
Summary table showing in lumen maintenance (percent) change in input power (percent), and change in color after 6,000 hours of testing • LM-79 reports at T = 0 and T = 6,000 hours Description and details of the product under test and test setup •
Although extrapolated LM-80 data for packaged LEDs has been used as a proxy for luminaire lifetime (and is an ENERGY STAR requirement), it only deals with lumen depreciation of a luminaire indirectly, and then only depreciation due to the LEDs themselves. There is also an issue regarding how to accurately extrapolate data of limited duration. The TM-21 committee of the Illuminating Engineering Society of North America (IES) has been exploring ways to address the latter. As of this writing, the results of the committee’s work have not been published, but when available, they may be used to extrapolate to the L70 value of the LED packages, but not for the luminaire. The working group therefore recommends LM-79 testing of the complete luminaire to determine lumen output over time.
PRODUCT VARIATION OF NEW PLATFORM
Recognizing possible platform variations to extend the product line for other applications (product groups) or material or design changes, additional measurements may be needed to ensure the platform is still qualified. Consideration may be given to minimize the number of test hours to demonstrate long-term life performance, as described above. In this regard, it is reasonable to consider the different types of change (or model variations) and their likely impacts on lifetime.
Ultimately, this is the customer’s choice, but it is recommended that manufacturers develop and document specific rules for change control to maintain the integrity of their products.
For example, changes in the following areas may be deemed to require significant retesting:
Housing/chassis Thermal management/heat sink • Change of assembly method or materials • Light source (includes operating current, Vf, and LED supplier) •
Other changes, such as in finish or out of the optical path, may require less requalification.
Analytical data may often be used in part to demonstrate that the change has not influenced the luminaire’s lumen maintenance performance. But typically, a small number of luminaires may need to be retested for some, perhaps shorter, period of time. If the manufacturer cannot demonstrate via analytical data or limited testing that life performance is not diminished, then the luminaire should be treated as a new platform and subject to full qualification requirements.
LED Luminaire Lifetime Recommendations, June 2011 Page 25 The working group recommends that manufacturers develop and document their own change control process, and that they be responsible for providing sufficient justification to their customers so that any change will be accepted as having no material, deleterious effect on product lifetime.
The discussions above outline current understanding of the issues surrounding LED luminaire reliability and lifetime. A basic recommendation for a standard, default designation of product life is provided, along with some suggestions for additional characterization in certain applications or markets. During the course of discussion to update this guide, the working group considered a
number of additional aspects of product lifetime, namely:
Design for system reliability Product labeling •
Warranties • •
DESIGN FOR SYSTEM RELIABILITYA thread running through many working group conversations was that the methods available today to demonstrate full luminaire performance over time (using LM-79) may be too costly or timeconsuming to be practical in many circumstances, especially given the rapid evolution of SSL products. The result is often a fall-back to the use of LED package lumen depreciation as a proxy for product lifetime, and the fear is that this will lead to customer dissatisfaction with this new and promising technology.
LED operation is interdependent upon drive electronics, the thermal management system, and the optical system, as well as upon proper and controlled materials and manufacturing processes.
Because of the resulting complexity, it may be more efficient and effective to take a system reliability approach in the design of the LED lighting fixture. By understanding how each system component contributes to failure, one can estimate overall reliability and optimize the design for best performance. Figure 4 (page 11) simply illustrates how information on various individual parts of the system can be combined to arrive at an overall estimate of lifetime. If statistical information on failures were available even for the major subsystems of a luminaire, it would be a big step toward more realistic estimates of product lifetime. While this would not obviate the need to verify lifetime anytime soon, it could eventually reduce the overall testing requirements.
As noted earlier, however, some failure modes may be as a result of interaction between parts of the system, which requires still more sophisticated analysis. For example, LED manufacturers typically use pulse-mode operation with a very short pulse—typically 10 or 20 milliseconds— which will not heat up the LED. On the other hand, there may be numerous LEDs in situ, often in close proximity to one another, operated continuously. These conditions elevate TJ above 25°C, thereby affecting photometric and colorimetric performance of the LEDs. This effect is offset by LED Luminaire Lifetime Recommendations, June 2011 Page 26 various thermal management approaches incorporated into the luminaire design, which must be accounted in the system reliability design.
The DOE SSL Manufacturing Roadmap recommends that LED driver manufacturers provide uniform data for LED fixture manufacturers which include a number of specific performance results to assist in luminaire design. A similar approach might also be applied to other components and materials.
Once such information begins to become available, software to analyze and combine the data into an overall system reliability estimate will be needed. The working group believes that the industry working together to arrive at standard methods to report and process statistical reliability information could be highly beneficial in terms of the effort and cost needed to achieve reliable product design, as well as the accuracy of the resulting lifetime predictions.
lifetime, L70/B50, as defined in this guide. Recognizing that some manufacturers may not have the The DOE Lighting Facts® label should be augmented to include the standard LED product ability or time to demonstrate lifetime in accordance with these recommendations, the label should not require that lifetime be stated, but if it is, then it should be reliably established with standard tests. Otherwise, a specific lifetime should not be claimed at all. If lifetime is not stated, then it is recommended to include at least one of these other options: lumen depreciation, a warranty, or an estimate based on accelerated testing of components.
The label should also be augmented to indicate whether the product is “serviceable” or “nonserviceable” as defined in this guide.
The DOE Lighting Facts label should also be modified to include, optionally, a warranty period to cover either lumen lifetime, color shift, or both. The warranty period may or may not coincide with lifetime (if claimed) at the manufacturer’s discretion. This warranty may be particularly useful to manufacturers who, while not able to demonstrate lifetime in accordance with the recommendations, are nonetheless prepared to guarantee performance over some period of time.
This may be especially useful for new products, also.
Absent lifetime data as prescribed, the manufacturer may alternatively or in addition to a warranty state the LED lumen depreciation time to L70 based on LM-80 data and a projection based eventually on TM-21. Because present practice is often to use this figure as a proxy for lifetime,
however, the label should be explicit:
The time shown should be described as “LED lumen depreciation to L70,” not as lifetime. Moreover, this depreciation time value should only be shown if the manufacturer “Lifetime,” and should avoid any implication that it represents tested product can assure that the LEDs are operating at or below the temperature of the LED LM-80 testing used to project the designated value, and that the drive current in the luminaire does not exceed that used for LM-80 testing during the stated time period. Additional information that should be readily available to the purchaser, although not on the label itself, should include maximum ambient temperature of operation to achieve this LED Luminaire Lifetime Recommendations, June 2011 Page 27 depreciation performance, the number of hours actually tested on which the projection is based, and the type of projection if not in accordance with TM-21.
If a manufacturer has developed proprietary techniques for accelerated component testing and system modeling, as suggested in the previous section, then it may choose to designate an “estimated system life” on the label. In no case should this figure exceed the lumen depreciation period estimate, however.
The Lighting Facts program may require substantiation of any lifetime claim put forth for use on the label. In addition, the burden of proof lies with the manufacturer to show that any platform changes have not materially affected lifetime.
The label should further be modified to include, optionally, certain color shift information
depending on the market segment:
A general qualitative description of color shift, or A maximum color shift warranted for some period of time, or •
For the short term, color shift, if quantitatively specified, could be described in terms of change in CCT. For some applications, and for professional use, actual limits on the change in color coordinates should be specified, but it is not recommended to include them on the Lighting Facts label.
Since the original publication of this guide, the Federal Trade Commission has developed requirements for a label similar to Lighting Facts for certain replacement lighting products. While this label has the same name, it is formulated somewhat differently, which may lead to some confusion. The working group hopes that the FTC adopt the same definition of lifetime as recommended in this guide, if possible, to preserve a degree of uniformity among products and labels.
THE ROLE OF WARRANTIESGiven the early stage of LED lighting technology, it is difficult to accurately predict product lifetime.
Warranties provide a means for the potential buyer to reduce risk. While this guide provides some guidance on what might be included in a warranty, it is ultimately the manufacturer’s decision on how best to provide this protection and how much risk to take in doing so. The warranty may logically be shorter than the standard lifetime as defined in this guide. As there will be a distribution of failure times for any product, the degree of risk on the part of the manufacturer depends on the level of confidence in the projected lifetime and the amount of variation in that distribution.
LED Luminaire Lifetime Recommendations, June 2011 Page 28
RECOMMENDED FURTHER READINGBenatti, John. “MTBF and power supply reliability.” Electronic Products, August 1, 2009.
IPC-9592A, “Requirements for Power Conversion Devices for the Computer and Telecommunications Industries.” Mao, George and Marshall Miles. “LED driver lifetime and reliability hold the key to success in LED lighting products.” LEDs Magazine, September/October 2010, 33–37.
Weiss, Bill. “Lighting for Life.” Digi-Key TechZone, TZL101.US (November 11, 2010), 8–14.
LED Luminaire Lifetime Recommendations, June 2011 Page 29
ACKNOWLEDGEMENTSThe Department of Energy and the Next Generation Lighting Industry Alliance wish to acknowledge the valuable contributions of the members of the DOE Reliability and Lifetime Working Group who offered their considerable time and expertise to the development of this document. Present members of that group and their organizational affiliations are listed below.
Please direct any comments regarding this guide to the DOE Reliability and Lifetime Working Group at email@example.com.
* NGLIA member †On behalf of the U.S. Department of Energy
LED Luminaire Lifetime Recommendations, June 2011 Page 30