Energy-Tech Magazine

The EPA Utility MACT regulations will require existing coal-fired plants to reduce mercury emissions by 90 percent on average. The final Mercury and Air Toxics Standard (MATS) was issued on Feb. 16, 2012, and in accordance with the Clean Air Act, the three-year compliance period begins April 16, 2012. A number of individual states already have imposed mercury reduction legislation, in some cases more stringent than the EPA’s guidelines. The leading candidate by far with regard to mercury reduction is sorbent injection into the flue gas, with subsequent collection of the reacted material in either an electrostatic precipitator (ESP) or fabric filter device, aka baghouse.

While sorbent collection of mercury can be effective, misunderstanding of both sorbent and flue gas chemistry can lead to less-than-desired results. This article examines many of the most important issues with regard to this process.

Mercury speciation
During combustion in the furnace, mercury is primarily released in its elemental form, Hg0. As the combustion gases cool

References Presentation by ADA Carbon Services, December 2011.

in the backpass, chlorine that has also been released will oxidize some of the mercury to the +2 valence state. Mercury oxidation is an extremely important aspect regarding its removal by sorbents and, for that matter, in wet or dry flue gas scrubbers. Eastern and Midwestern bituminous coals (in boilers with selective catalytic reduction (SCR) systems) often contain enough chlorine to oxidize most of the mercury. However, the chlorine content is typically low in sub-bituminous coals, and most notably those from the Powder River Basin (PRB). Thus, the bulk of mercury produced by PRB combustion will exist in the elemental state, without some form of artificial treatment.

Sorbent details
The sorbent that has proven most effective to date is powdered activated carbon (PAC). The material is prepared so that is has an enormous surface area, in which each particle has thousands of microscopic pores.

In fact, 1 lb. of activated carbon has a surface area equivalent to 50 football fields.¹

Activated carbon is utilized for many adsorption purposes in both gaseous and liquid streams.  The pore size within activated carbon particles is very important, where the ideal range depends on the application and the impurity or impurities to be removed. For flue gas mercury removal, pore diameters within a general range of 10-150 angstroms are reportedly most effective.¹ This then leads to the question, “What is the best starting material to produce activated carbon that has the most effective properties for mercury capture?” The 4 most common starting materials in general are coconut shells, coal, lignite and wood. These are all heated in an oxygen-free environment to produce activated carbon. Many suppliers use lignite as the starting material for flue gas mercury PAC, while others use bituminous coal.

Treating sorbent to oxidize mercury
Even though PAC has a large surface area, if mercury exists in the elemental form much of it will not adsorb to the carbon. But with PRB or lignite, the chlorine content is just too low to provide much oxidation. Enter brominated activated carbon, where the material is impregnated with calcium bromide to provide the necessary oxidant. Calcium bromide might also be added to the fuel or injected in the furnace to oxidize mercury.

Sorbent poisons
By far the most problematic PAC deactivator is sulfur trioxide (SO₃). This compound blocks the pores in carbon particles, preventing mercury capture. Small quantities of SO₃ are generated by minor oxidation of sulfur dioxide (SO₂) in the boiler backpass. However, additional SO₃ can come from two other sources. In units equipped with SCR, perhaps a percent or so of SO₂ might be oxidized to SO₃. Potentially much more troublesome is the SO₃ that is commonly injected to improve electrostatic precipitator performance (ESP) for units firing PRB coal.