We are about a year away from the date on which mandated energy efficiency standards will increasingly make it impossible to sell incandescent bulbs in the U.S. For my part, I hope LED-based lighting is a cheap alternative to ordinary household incandescent by then. Compact fluorescent lights, despite a lot of hype, just don't seem to have lived up to the energy efficiency promises of their promoters.
I found this out the hard way when I installed my first CFL. The promotional material on the package claimed lifetimes in the 10,000-hr range. My bulb didn't last quite that long. It was dead within six months after occasional use in my garage, eventually logging something less than 100 hours of ON time.
Apparently my experience wasn't extra ordinary. In May of last year, the Calif. Public Utilities Commission released a study about a CFL subsidy program in that state. It concluded that distributing low-cost CFLs there saved about 73% less energy than local utilities had projected. Part of the disappointing performance was chalked up to unwarranted optimism about CFL lifetimes. The useful life of the bulbs had initially been put at 9.4 years. It turns out a more realistic figure is a bit over six years.
In my own case, it looks as though relatively short ON times may have done in my CFL bulb. That was one of the factors brought up by a fluorescent systems engineer at GE with whom I spoke about how CFL makers come up with lifetime ratings for bulbs. It turns out engineers define CFL lamp life the same way they do for incandescent bulbs: It is the median of the distribution curve for all lamp lifetimes (in other words, where 50% of the lamps die).
Fortunately that distribution curve doesn't have a sharp peak at the median. GE says the spread of CFL lifetimes is a Weibull distribution, though with relatively few bulbs on the infant mortality tail. Still, only 50% of CFLs will give the lives listed on their packaging. A key point, though, is in how these lifetime data get measured. An ANSI standard procedure spells out that sample bulbs be powered on for three hours and off for 20 minutes at room temperature. This test in no way resembles the situation in my garage. There I was more likely to flip on the light for just a few minutes several times a day, and temperatures could vary from below freezing to 100°F.
High temperatures have been known to fry CFLs or, more specifically, the CFL ballast. They are a particular problem in recessed light fixtures where heat can build up. And short duty cycles? My GE contact says the company doesn't have a lot of data on this question, but admits some studies suggest CFL life drops with frequent cycling.
I've had no luck finding hard numbers on what frequent cycling does to CFLs. Consumer Reports, though, says its own tests indicate CFLs last longer if left on for 15 minutes or more. Which is why I look forward to day when I can put an inexpensive LED in my garage light fixture. I can't see leaving that light on for 15 minutes with nobody in the garage, just to get a reasonable life out of the bulb.