A brief article I wrote for tED Magazine. Reprinted with permission.
While LED promises long service life as one of its primary advantages, accurately projecting lifetime remains a challenge. The Next Generation Industry Alliance’s (NGIL) LED System Reliability Consortium (LSRC) recently updated a key report, LED Luminaire Lifetime: Recommendations for Testing and Reporting, to shed some light on the issues involved.
The lighting industry is predominantly using lumen depreciation to express expected life of LED products. A published life of 50,000 hours at L70, for example, means average light output for a population of products is expected to decline by 30 percent after 50,000 hours of operation. The lumen depreciation may be for any reason—general lumen depreciation among functional LEDs, depreciation within the luminaire, or a percentage of LEDs failing outright.
The industry uses the LM-80 standard to test lumen depreciation for LED packages, modules and arrays up to 10,000 hours. TM-21 provides a procedure for extrapolating this information to L70 or some other target value. Because luminaire design affects light output, LM-80 testing data provides a baseline for estimating lumen depreciation for LED products under certain conditions. As such, luminaire design may affect lumen depreciation, as could stress factors such as heat and humidity experienced in the field. Lifetime claims for LED luminaires and lamps are made by manufacturers based on their own methods and data, as there is currently no industry standard for doing this.
While lumen depreciation in the light source is popularly used as an indicator of product lifetime, LED products are really systems featuring other components that may fail first. This challenges earlier assumptions centered around the light source being the leading point of failure. In fact, research conducted by RTI International—which resulted in the 2013 report, Hammer Testing Findings for Solid-State Lighting Luminaires—suggests that other luminaire components are far more likely to fail before the LEDs do. This has several big ramifications.
First and foremost is that LM-80 and TM-21 get us to a future lumen depreciation value, not necessarily service life. As such, it’s one indicator of life.
In its updated recommendations, LSRC also added color shift as a consideration in defining service life for applications where color quality is important. This is an issue of color stability over time in a given product and color consistency between products installed in the same space over time. Because predicting color stability is difficult and there is no industry standard for it, for color-critical applications, designers should ask manufacturers about how they ensure color stability and negotiate an appropriate warranty, if needed.
Different projects have different needs, and the definition of service life for an LED product may change based on the project. For a given project where color quality is critical, for example, service life might be defined by as when average light output has declined by X percent (due to lumen depreciation or LEDs failing) or when color has shifted beyond a specified limit of Y, with X and Y being defined by the designer and/or the owner based on the application. If the LEDs are visible with occupants having a direct view, the point at which a certain percentage of LEDs fail (go dark) could be another failure mode, one based on aesthetic considerations.
Because of the close dependency of such parameters and demand for products, LSRC recommends the industry delineate LED products into three reliability categories—lamp-replacement grade, standard grade and specification grade. Each category would be defined within a range of parameters, from light output to color quality. This would help manufacturers minimize costs by designing products for specific markets rather than a one-size-fits-all approach.
LSRC advises LED product buyers to focus on qualifying suppliers. Determine how manufacturers support their reliability and service life claims, and ask them to back up these claims by revealing their methods and data. As industry-consensus robustness tests develop, buyers should require their suppliers to follow them.
Next, understand the product warranty, what it covers and what it doesn’t, and whether the warranty period covers a reasonable fraction of claimed life. Ask if the manufacturer will maintain a stock of replacement products and components and for how long.
LED is still young, and it’s still evolving not only in terms of technology, but how we evaluate and rate performance. As the LED revolution continues to develop, buyers should protect themselves by qualifying suppliers and asking the right questions, and qualifying products by evaluating the application need and associated parameters for defining useful life.
