DOE OSTI2014
The lumen depreciation and color shift of 17 different A lamps (15 LED, 1 CFL, 1 halogen) was monitored in the automated long-term test apparatus (ALTA) for more than 7,500 hours. Ten samples of each lamp model were tested, with measurements recorded on a weekly basis. The lamps were operated continuously at an ambient temperature of 45°C (-1°C). Importantly, the steady-state test conditions were not optimized for inducing catastrophic failure for any of the lamp technologies—to which thermal cycling is a strong contributor— and are not typical of normal use patterns—which usually include off periods where the lamp cools down. Further, the test conditions differ from those used in standardized long-term test methods (i.e., IES LM-80, IES LM-84), so the results should not be directly compared. On the other hand, the test conditions are similar to those used by ENERGY STAR (when elevated temperature testing is called for). Likewise, the conditions and assumptions used by manufacturers to generated lifetime claims may vary; the CALiPER long-term data is informative, but cannot necessarily be used to discredit manufacturer claims. The test method used for this investigation should be interpreted as one more focused on the long-term effects of elevated temperature operation, at an ambient temperature that is not uncommon in luminaires. On average, the lumen maintenance of the LED lamps monitored in the ALTA was better than benchmark lamps, but there was considerable variation from lamp model to lamp model. While three lamp models had average lumen maintenance above 99% at the end of the study period, two products had average lumen maintenance below 65%, constituting a parametric failure. These two products, along with a third, also exhibited substantial color shift, another form of parametric failure. While none of the LED lamps exhibited catastrophic failure—and all of the benchmarks did—the early degradation of performance is concerning, especially with a new technology trying to build a reputation with consumers. Beyond the observed parametric failures nearly half of the products failed to meet early-life thresholds for lumen maintenance, which were borrowed from ENERGY STAR specifications. That is, the lumen maintenance was sufficiently low at 6,000 hours that seven of the products are unlikely to have lumen maintenance above 70% at their rated lifetime (which was usually 25,000 hours). Given the methods used for this investigation—most notably continuous operation—the results should not be interpreted as indicative of a lamp’s performance in a typical environment. Likewise, these results are not directly relatable to manufacturer lifetime claims. This report is best used to understand the variation in LED product performance, compare the robustness of LED lamps and benchmark conventional lamps, and understand the characteristics of lumen and chromaticity change. A key takeaway is that the long-term performance of LED lamps can vary greatly from model to model (i.e., the technology is not homogenous), although the lamp-to-lamp consistency within a given model is relatively good. Further, operation of LED lamps in an enclosed luminaire (or otherwise in high ambient temperatures), can induce parametric failure of LEDs much earlier than their rated lifetime; manufacturer warnings about such conditions should be followed if performance degradation is unacceptable.
32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗