I've copied and pasted some text about CFL life cycle research from a recent Home Energy article below. The articles was written by Richard Benware, a Cornell graduate student and former intern at EPA. Most of the research he cites was done by EPA.

Hope this helps!

"When CFLs are created, manufacturers dose the bulb with a small amount of mercury. This mercury, when electrically stimulated, releases UV light, which subsequently reacts with a phosphor coating to create visible light. Thus mercury is an essential part of every CFL; without it, the bulbs would not produce light. The typical dose of mercury is about the size of a pen tip, and these doses have been getting smaller and smaller. One reason for this is that the laws resulting from the Restriction of Hazardous Substances (RoHS) Directive have made it illegal for CFLs in Europe to contain more than 5 milligrams (mg) of mercury."

"In the United States, there are no such laws limiting the amount of mercury in lightbulbs as yet, but members of the National Electrical Manufacturers Association (NEMA) have voluntarily agreed to limit the amount of mercury in the CFLs that they produce to 5 mg for bulbs of up to 25 watts and 6 mg for bulbs of 25 to 40 watts. The average CFL on store shelves today contains about 4 mg of mercury, and nearly all the CFLs in production contain less than 5 mg. The mercury used in all the CFLs produced in the United States represents 0.18% of the mercury used in all U.S. products and industrial processes."

"CFLs do not release mercury as long as they are intact. In fact, they reduce net mercury emissions in the environment by conserving energy. For every kWh of electricity used by consumers, the average power plant emits over 1.5 lb of pollutants. If a 75W incandescent is replaced by an 18W CFL, the CFL will use 456 kWh less energy than the incandescent over its 8,000 hour lifetime. The Emissions and Generation Resource Integrated Database (eGRID) contains data on the emissions of the average power plant. Using eGRID’s information to calculate the average emissions per kWh, we find that this single CFL has prevented the release of 2.72 lb of sulfur dioxide, 1.05 lb of nitrogen oxide, 5.67 mg of mercury, and over 700 lb of CO2."

"It is important to note that these are the reductions from the average U.S. power plant. The eGRID data show that, on average, nonbaseload emissions tend to be dirtier. And in addition to reducing emissions, CFLs save money for the consumer. The Energy Information Administration (EIA) gives a 2006 average residential electricity cost of 10.08¢/kWh. Using the example given above, and basing our calculation on this figure, we find that a consumer would save about $46 on energy over the lifetime of the CFL."

"When these bulbs finally do reach the end of their useful life, there are several pathways they can take. In the best-case scenario, the bulbs are recycled. Recycling rates are increasing, thanks to state regulations—California and Minnesota have banned altogether throwing CFLs in the trash—and improved consumer awareness. In 1999, it was estimated that only 15% of all fluorescent lightbulbs were recycled. Currently, that number has increased to around 25%, with higher levels in commercial applications. Since an average of 98.9% of the mercury is successfully recovered in the recycling process, this pathway generates minimal emissions."

"Even the CFLs that are discarded in the trash are unlikely to release much of their mercury. Although most of them break under current trash disposal methods, some remain unbroken, and will not release any mercury. But those that do break are not likely to release much mercury. EPA estimates that only 0.2% of the remaining mercury in a spent bulb is elemental vapor. The rest of the mercury is in the glass, the phosphor coating, and the electrodes of the bulb. Mercury absorbed in these areas is not readily released. In fact, an EPA study found that only 6.8% of the total mercury in a broken bulb will be released. Since the average bulb on the market today contains only 4 mg of mercury, it will release only about 0.27 mg, even if it breaks when it is thrown in the trash."

"The only disposal option that could lead to the release of any significant amount of mercury is incineration. Today, many incinerators have advanced mercury control technologies. CFLs disposed of in such incinerators would release up to 90% of their mercury, but those emissions would then be removed by these technologies. Incinerators without these technologies are not capable of removing the mercury. But even after accounting for all of the emissions that occur via all of the routes listed above, CFLs represent a mere 0.01% of total U.S. mercury emissions annually."

"It is important to note that even if CFLs released all of their mercury, the environment would still be better off than it would be if nobody used CFLs. This is true because the average power plant releases 5.67 mg of mercury to power each 75W incandescent bulb. In short, replacing incandescents with CFLs is a great way to save energy, reduce mercury emissions, and save money."

Thanks.

People dump fluorescent tubes into dumpsters by the armload.
They shouldn't, but they do, and I'm willing to bet that they will continue to do so in overwhelming numbers.

If you could replace the mercury with plutonium, along with a
vigorous PR campaign urging responsible recycling, because it would be much more energy efficient than mercury, would you
endorse it? I seriously wonder.

Only 50% of the electricity produced uses coal- the rest is natural gas (19%), nuclear (20%), hydro (6%), oil (3%), and other, including renewables (2%). The figures cited by CFL proponents claims that the reduction in energy use by a CFL more than offsets the intrinsic mercury used in the bulbs; thus a conventional bulb would use about 10mg of mercury in the same time a CFL might use around 5.5mg. This figure (and it has been noted exhaustively) assumes the mercury contribution comes from 100% coal power. The actual figures for a nationwide comparison should be about 5mg or so for incandescents and about 5.25mg for CFL's given this energy mix and simple calculations. I accept that non-coal energy production also has mercury indirectly on its hands, and if we make the wild guess that the impact is roughly 25% of the contribution of coal, then CFL's use 5.6mg and incandescents use about 6.25mg of mercury.

The story doesn't end there, however. The mercury in the CFL's has been purified. Worldwide, production of mercury is only about 70% efficient. In other words, every milligram of mercury produced means 1.4mg of mercury is actually put into the environment. The extra .4 mg is the portion that is thought to enter the watersheds and atmosphere, not sequestered in any way. This means the mercury scorecard becomes incandescents: 6.25mg; CFL's: 7.2mg.

The story doesn't even end there, unfortunately. A Lawrence Livermore Labs study concluded that the lifetime of the average CFL on the consumer market had a median life well short of its reputation or rating. In fact, for realistic household use where a light would be used for an hour before being turned off would suffer a 50% reduction in its rated lifetime. The loss to incandescent lifetimes was about 20%. Even worse, if the use of the light is only a half hour, the life of a CFL is reduced by 85%, whereas incandescent bulb lifetimes suffer only 25%. This is a devastating blow to CFL's that is completely overlooked. This means that for a one hour usage model, the CFL will use 13mg of mercury compared to only 6.25mg for the incandescents.

CFL's use over 2x the mercury of conventional bulbs.

A light should go off in your head about the real impacts of CFL's. A lot of articles on this subject chide the naysayers of CFL's. The answer is easily arrived at however if you care to look.

As for energy use, CFL's will use less. Given the realistic life of a CFL bulb being 50% of its promise on the box, the financial payback is still pretty good. Then again, an LED bulb will be a better bet yet, with no mercury hangover.