I've tossed dozens of short operating life CFLs. I've never tried but, does anyone expect any kind of product support from a no name Chinese manufacturer?
Each of the three lighting technologies discussed has their advantages and disadvantages. And each has its uses. LEDs are stupid expensive, will they cost more than incandescents over their life? Probably. If lighting were possible that used zero electricity, it still wouldn't make sense to use it if it costs $10 million per bulb. CFLs last no time at my house but if you fixed that they still wouldn't work in my oven or outdoors on cold nights.
In the end, the consumer should be informed in a truthful manner and allowed to pick appropriate lighting for the situation. He should not be forced into new technology by a bureaucracy.
"Please . . . let's let the market decide which products succeed and which fail without resorting to the temptation to ban things that people use and need every day."
Unfortunately, that train left the station a long time ago. Our Besserwissers can't allow us to have choices since we might make the wrong ones.
Because of their mercury content, the premature failure of CFLs is an issue in regions where there is no recycling of bulbs. My local Home Depot has recycle bins; otherwise, I wouldn't use CFLs since I find they typically fail too soon (after several thousand hours) to prevent an emission of mercury from coal plants greater than their own mercury content.
In amplification of what Bert22306 has said, a "standard" 100W GE 48690 bulb is rated for about 1260 lumens output, or 13 lumens per watt, and lower wattage lamps tend to be less efficient with a 40W GE 81869 lamp at 370 lumens output, or 9.3 lumens per watt. A Cree XMLAWT-00-0000-000LT50E4 is rated for 133 lumens per watt output at 2.03 watts. Assuming an 85% efficient drive circuit (a fairly good one) and an operating LED junction temperature of 120C (20% flux output derate), that will result in 90 lumens per watt. This is closer to reality, providing about an 7X to 10X electrical savings, mostly in environmental heat rejection.
Even more of interest is to look at the plot of white LED efficiency with time noting that incandescents plateaued decades ago, the LEDs are still rapidly improving.
As far as I'm concerned, the inefficiency of incandescents is THE issue. So if it's not dicussed enough, that is indeed a problem, and perhaps one that the public at large doesn't fully appreciate. CFLs use about 1/4 of the power of incandescents, for a given amount of light. LEDs need about 1/5 of the power of an incandescent. These are big savings, when you consider that the load of electric lighting is a hefty percentage of the total, especially in gas-heated homes.
As to your point, I'd agree that the heat created by inefficient light bulbs might take some of the load off the furnace, but light bulbs even do that job very inefficiently too. No one would defend the efficiency of resistive electric heating, as a means to heat a home. It's better to save that electricity, and either apply it to heat pump heating, which uses the heat content of the ambient outside air to heat the home, or save that electricity entirely and let an efficient gas furnace do the heating job. There's nothing 100 percent efficient about any form of resistive electric heating.
And then, of course, there are the other seasons of the year.
I'm surprised that this topic usually generates little discussion about the purported inefficiencies of incandescent lighting. While it may be true that much of the energy used by these devices is given off as heat, this doesn't mean that that energy is wasted. This time of year, it just means that my furnace doesn't have to work quite so hard to heat my house. What this means to me is that for about 7 months of the year, my incandescent bulbs are 100% efficient. That makes their lifetime efficiency average much better than the 5% - 7% normally quoted. Not bad for an inexpensive, non-toxic, and reliable product.
Here in Australia, you can't buy incandescents except in special sizes (eg for sewing machines, "golf ball" bulbs etc - and a lot of those are the halogen type.
I've never seen a disposal bin for CFLs here though, and the shops are flooded with cheap chinese crappy ones which I've learned to avoid.
At one point the govt gave energy credits for CFLs so you could get them from some utility companies for free - usually awkward long sizes though. I have a whole bunch which I am slowly getting through.
It may be true that CFLs do more damage to the environment (although here in the US, you can dispose of old ones in special bins at certain stores). But I think the main point is, the government is not mandating use of CFLs.
The ideal situation would be if lighting companies would get serious and develop LEDs or other types.
To me, this is very much like the fuel injection evolution in cars. The government did not tell auto companies HOW to achieve the low emissions. The feasibility of doing via catalytic converters had been demonstrated, so the bureaucrats knew that this could be done. But that's as far as they went. It was the auto companies that realized that using 3-way catalytic converters, along with precisely metered fuel and air, was the way to go. And everyone came out ahead.
In this case, CFLs are certainly not the answer. But ONLY if the archaic incandescents are finally banned will the marketplace develop the better answers. And we all know they can, because we're all engineers, right?
Many incandescent in higher-wattage applications(show lighting spotlights, for example) actually run a small amount of current through the bulb when it is nominally off—to increase the life by minimizing thermal shock to the filament. You don't want these bulbs blowing out when you turn up the lights during a show. And even if they don’t blow out during the show, they are expensive bulbs and are often awkward to replace. So a little bit of leakage current and "wasted power" pays big returns.
Blog Doing Math in FPGAs Tom Burke 23 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...