This article was published in the July/August 2013 issue of LEDs Magazine.
View the Table of Contents and download the PDF file of the complete July/August 2013 issue, or view the E-zine version in your browser.
Times change quickly in the solid-state lighting (SSL) world. Just two years ago, LED and lighting manufacturers were making hyperbolic claims of 50,000–100,000-hour product lives and many poorly designed SSL products were failing very early. Now a maturing industry can reliably deliver lighting products with such long life, but should they? I've become increasingly convinced that, while exceptions exist, the industry needs to focus on 10,000–15,000-hour lifetimes for many mainstream products.
Don't misunderstand. I applaud the developments of the last few years. The LED industry has embraced LM-80 testing and the TM-21 methodology to project component life. Product developers have learned to implement thermal and driver technology to ensure that SSL systems deliver on long-life potential.
Still, in many cases, the industry will be better off with a more moderate expectation of lifetime. We can ensure a more robust ongoing lighting economy with regularly refreshed product lines and retrofitted installations that still deliver all of the energy-efficiency promises of LED technology.
Now I can't claim this revelation as original thinking. I've heard industry insiders make similar points, so credit where credit is due.
Last year at the Street and Area Lighting Conference, Mark Hand, director of new products and technology at Acuity Brands, discussed how LED products are essentially over-specified for most applications. Ironically, he presented at an event where long life may make sense. But Hand's point was that specifiers can manage the cost of products by being realistic and only demanding long life, or other premium features such as high CRI, when justified by the application.
Venture capitalist Vinod Khosla took a stronger stance in a Lightfair keynote that I also covered in my column from the previous issue. He simply said long life doesn't matter in most applications, and used 20,000–25,000 hours to characterize long life.
Khosla made his point with a simple image. He showed a photo of a phone that we may have used 25,000 hours ago and asked if anyone still wanted to use that phone today. LEDs are going to enable amazing new form factors and features in lighting products. So even if someone installs a 50,000-hour product today, they are apt to change it long before end of life comes.
If we reduce the expectation to 10,000–15,000 hours, what is the impact? I don't have all of the answers. But we could use fewer LEDs driven harder and slash upfront cost. Maybe we really don't need such complex thermal management in all products. Simpler thermals could lower costs and yield more attractive products.
Obviously, my stance doesn't apply everywhere. Industrial lighting that burns 24 hours a day and is rarely retrofitted warrants products designed for long lifetimes. There is a light fixture on a high ceiling in a basement stairwell at my mother's house that will get an expensive Philips L-Prize lamp, which will probably last 50,000 hours, installed when it next burns out. Clearly, street lights aren't retrofitted often and long life is a huge part of projected payback.
For many applications, less may be more. Even office spaces are often retrofitted more quickly than the lifetimes that are commonplace with new SSL products. So great job to the industry, but let's rethink the strategy in many cases.