«SOLID-STATE LIGHTING PRODUCT QUALITY INITIATIVE THIRD EDITION SEPTEMBER 2014 Next Generation Lighting Industry Alliance LED Systems Reliability ...»
LED LUMINAIRE LIFETIME: RECOMMENDATIONS
FOR TESTING AND REPORTING
PRODUCT QUALITY INITIATIVE
Next Generation Lighting Industry Alliance
LED Systems Reliability Consortium
TABLE OF CONTENTSAcknowledgements
1 Introduction and summary
1.1 Background and objectives
1.2 What makes LED-based luminaires “different”?
1.3 Key conclusions and recommendations
2 Defining lifetime
2.1 Combining abrupt and gradual failures
2.2 Product reliability classes
3 Studies of failure modes
3.1 Accelerated life tests
3.2 Power supply considerations
3.3 LED package failures
3.4 Field data
4 Warranties and shared risk
5 Reliability testing and standards
5.1 Accelerated testing
6 Reliability modeling
7 Additional references
The Next Generation Lighting Industry Alliance wishes to acknowledge the valuable contributions of Dr. James R. Brodrick and the U.S. Department of Energy, and DOE support for Dr. Lynn Davis of RTI International, whose reliability testing work contributed significantly to the knowledge base for this document. We are grateful to the current and former members of the LED Systems Reliability Consortium, who offered their considerable time and expertise over the past three years to the development of this document.
Please send any comments regarding this document to Amy Oriss, firstname.lastname@example.org.
1.1 BACKGROUND AND OBJECTIVES As the market for solid-state lighting (SSL) begins to take hold and grow, our understanding of the technology has advanced significantly. Early on, there was a good deal of confusion about expected product life, and many relied on the gradual lumen depreciation of the LED (light-emitting diode) source as the best indicator— resulting, on occasion, in unrealistic claims for product life. To address this issue, the Solid-State Lighting Program of the U. S. Department of Energy (DOE) together with the Next Generation Lighting Industry Alliance (NGLIA) formed SSL Quality Advocates and a Reliability and Lifetime Working Group, which published two editions of the document LED Luminaire Lifetime: Recommendations for Testing and Reporting.1 The intended audience included manufacturers; specifiers, including architects, interior designers, lighting designers, and engineers; utilities and energy efficiency organizations; standards organizations; and regulators and other government agencies.
Since our earlier reports, the number and variety of LED packages and LED arrays has increased. They range from low-power LEDs that operate at less than
0.05 watts per package to COB (chip-on-board) modules that can consume over 50 watts per module.
This has allowed a corresponding increase in the number of LED replacement lamps and LED luminaires that are available. Existing data from numerous manufacturers show that today LED sources, when operated conservatively, are highly reliable and long lived. For most LED products, manufacturers strive to take advantage of the long source life to realize a longlived end product. There is now widespread understanding of the degree to which LED drive current and operating temperature (LED and driver) affect system reliability, with better designs as a result.
However, a wide range of design choices to meet Figure 1. Total system or luminaire reliability is the product specific application and market needs is available for
of all of the individual reliability considerations:
LED luminaires, hence a potentially wide range of RLuminaire = RLEDs * ROptical * RPCB * RFinish * RMechanical * RThermal * product life. The numerous other subsystems and RHousing * RGaskets/Sealants * RConnections * RDriver * RManufacturing components in a luminaire introduce other potential failure modes which will affect and may actually Source: Philips Hadco dominate the determination of system lifetime.
Figure 1, from the first Recommendations report, illustrates this concept.2 This complexity can make the prospect of measuring and characterizing LED system reliability and lifetime very daunting, and indeed it is.
1 U.S. Department of Energy, LED Luminaire Lifetime: Recommendations for Testing and Reporting, http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/led_luminaire-lifetime-guide_june2011.pdf.
2 In this figure “reliability,” or Rcomp, is the probability that the component will be functional at a given time, sometimes referred to as the “availability” of the component, or “probability of success.” The probability of failure by that time is (1-R).
1 daunting, and indeed it is. However, with good design, the overall luminaire reliability will be dominated by a small number of “principal components,” thus reducing the challenge.
Since the first and second editions of the Recommendations were published, SSL Quality Advocates was expanded to include more stakeholders and continued its work as the LED Systems Reliability Consortium (LSRC) under NGLIA. The group has contributed to a number of studies, reviewed international efforts for developing tools to project LED-based luminaire lifetime, and researched real-world experience with numerous types of LED-based luminaires. Our objectives have been to better understand what types of failures are most likely to affect system lifetime, to examine possible ways the failures of the principal components might be accelerated to shorten testing times, and to consider the attributes of system models that might assist in the reliable design of LED products. The intent is to provide guidance to both manufacturers and users as to what factors should be considered in developing and testing reliable products, and not to provide specific test protocols or data on acceleration factors for specific components nor any type of universal model to predict system reliability. Some manufacturers have devised suitable proprietary methods to estimate the reliability of their specific products, and subsystem manufacturers have improved the information available to designers on the reliability of those products. We encourage those efforts as well as the continued sharing of best practices among producers in the interest of continuous improvement.
Accordingly, it is appropriate to summarize what we have learned, to suggest directions for further work, and to update the LSRC’s recommendations for describing the life and reliability of LED-based luminaires. To that end, here is the third edition of LED Luminaire Lifetime: Recommendations for Testing and Reporting.
1.2 WHAT MAKES LED-BASED LUMINAIRES “DIFFERENT”?
LEDs convert electricity to light directly within the device. They do not radiate heat but rather retain it so dissipation must be by convection or conduction. This generally results in higher local temperatures within the luminaire, the extent depending on design, which will accelerate changes within the LED and in nearby electronic components. LEDs do not generate white light directly, necessitating either conversion via phosphor or other means, mixing of multiple monochromatic sources, or both. LED sources emit light potentially for a very long time, and integrated lamps appear to be capable of at least 25,000 hours of operation. Such fundamental differences from traditional lighting technologies have implications for product reliability, a few of which are
For LED luminaires, the end of life may not be evident. For practically all traditional lighting technologies, the end of life is evident: there is no light. And while it is true that traditional fixtures can corrode or color can shift beyond acceptable limits, requiring replacement, and many technologies show some level of light depreciation, for the most part lights-out failure is what people have come to expect. In contrast, LED sources—the core of an LED lighting system—emit light for a long time. Over that time, depending on design, the light output may continuously fade or the color may slowly shift, possibly to the point where low light output or an unacceptably large color change constitutes practical failure. Indeed, during the early stages of the development of LED technology, lumen depreciation appeared to be the principal mechanism that would define the end of life.
However, that means that the “end of life” may no longer be clearly evident, a fundamental departure from our traditional understanding.
The order of failure may change. The traditional lamp + fixture paradigm is that the lamp or light source will fail long before the fixture as a whole, so the system is designed to easily replace a relatively inexpensive lamp (or, for some types, ballast) while the fixture lives on. That same model can still be implemented as a “serviceable” LED system. However, because the LED source may well be the part that “lives on,” and because the technology is rapidly evolving, which complicates maintaining a replacement inventory, the traditional lamp + fixture paradigm may not always be the most economical solution. A better way may be to think about an LED lighting system as an appliance—more expensive than a traditional lamp but with a much longer life expectancy, and that eventually will need maintenance or replacement.
Useful life is difficult to verify. For LED systems designed to have a very long lifetime, there are new challenges in testing. Straightforward operational aging tests of LED luminaire products can be prohibitively expensive and time-consuming, and manufacturers are increasingly concerned about mandated testing intervals that delay new product introduction and add cost. Many manufacturers have developed proprietary means to estimate product life for their own designs using data on principal components such as the LED package, driver, and optical components, which allows an estimate of the overall luminaire performance. While such practices exist for specific product lines and applications, there is no industry-consensus protocol at this time.
1.3 KEY CONCLUSIONS AND RECOMMENDATIONS
Recommendations from the Second Edition:
The following is a summary of the recommendations from the second edition of LED Luminaire Lifetime:
Recommendations for Testing and Reporting. Remarks in italics indicate whether these recommendations were unchanged, modified, or dropped in this edition.
Lumen depreciation is not a proxy for luminaire lifetime.
Unchanged: This remains an important point. The LED package may not be dominant in determining product lifetime.
Consider only light output in defining lifetime.
Modified: This recommendation now includes color shift for those applications where it is relevant.
Use overstress testing to identify design flaws and manufacturing defects.
Unchanged: The concept of “robustness” testing, while not a full reliability test, can provide useful screening of products.
Indicate if a product is serviceable or not.
Unchanged: Additional discussion of what serviceability means in terms of reliability is in Section 2.2.1.
Develop standard ways to characterize drivers for SSL use.
Unchanged: While driver characterization is better than it was, there is still a need for testing standards.
End of life as defined in this document excludes color shift.
Modified: For products designed for applications requiring color stability, this is not the case.
To deal with color shift, designate products in one of three categories (lamp replacement, standard grade, specification grade).
Modified: Designate as color-stable, if relevant, and redefine categories.
Develop standard qualitative descriptions of the degree of color shift.
Unchanged: Discussed in this document.
Define standard luminaire lifetime.
Dropped: There is no consensus around this approach at this time.
Reported lifetime should have at least a 50% confidence level.
Unchanged: This recommendation has not been widely observed but should be.
Use LM-795 for full luminaire characterization.
Develop and document a change control process.
Unchanged: This is good business practice for manufacturers, but not intended to suggest any industry standard.
Develop a capability for statistical system design for reliability.
Dropped: Because of the variety of designs and applications, this is not realistic as a general industry objective, although proprietary methods for specific product groups can be helpful to manufacturers.
Add standard LED luminaire lifetime to the LED Lighting Facts® label.6 Dropped: While this may have been useful when there were a relatively few types of products available, with the proliferation of products and applications, there is now too much variation in the lifetime for a general recommendation.
Recommendations for Manufacturers:
Provide clear product warranties, report compliance with industry standard measurements, and communicate your use of and commitment to best practices for designing and developing long-life, reliable products.
3 IES LM-80-08, Approved Method: Measuring Lumen Depreciation of LED Light Sources, http://www.ies.org/store/product/approvedmethod-measuring-lumen-maintenance-of-led-light-sources-1096.cfm.
4 IES TM-21-11, Projecting Long Term Lumen Maintenance of LED Light Sources, http://www.ies.org/store/product/projecting-long-termlumen-maintenance-of-led-light-sources-1253.cfm.
5 IES LM-79-08, Approved Method: Electrical and Photometric Measurements of Solid-State Lighting Products, http://www.ies.org/store/product/approved-method-electrical-and-photometric-measurements-of-solidstate-lighting-productscfm.
6 LED Lighting Facts, http://www.lightingfacts.com/.