Mississippi Power’s Victor J. Daniel Plant Unit 1 is a 540-MW_G pulverized-coal unit which typically combusts a 60 percent bituminous and 40 percent Powder River Basin (PRB) subbituminous blend. The unit’s flue gases are cooled by two parallel preheaters, each of which discharges to a pair of cold-side ESPs. The tests were conducted on one ESP, which is ¼ of the total flow. The flue gases are conditioned with gaseous SO₃ prior to treatment in the ESPs. The FGC system is in service continuously during typical operation.

In April and May of 2005, ADA-ES conducted the sorbent injection demonstration. It was designed to assess the effectiveness of two carbon sorbents on a unit that can fire varying blends of bituminous and subbituminous coals. The initial premise of the demonstration was to conduct all testing with the flue gas conditioning system in service, according to the plant’s normal operating practices.

The initial tests with standard PAC yielded low efficiencies with mercury removal of only 40 percent at an injection rate of 10 lb/MMACF. Following these tests, an interruption caused the FGC system to be taken out of service and suddenly the mercury removal levels nominally doubled. This dramatic performance increase caused the test team to reevaluate the test plan.

Subsequent to this discovery at Plant Daniel, similar results have been observed at other test sites. For many utilities, the issue of flue gas SO₃, whether due to native fuel content or SO₃ flue gas conditioning for enhanced ESP performance, inhibiting sorbent effectiveness is a key technical hurdle. Finding a suite of solutions to this problem is a primary concern to utilities that would use PAC injection to meet the new mercury emission limits.

At this point, the physical and chemical mechanisms by which flue gas SO₃ interferes with mercury adsorption are not fully understood. The Mercury Information Clearinghouse discusses a series of sorbent breakthrough curve studies, conducted by EERC, that investigate the role of sulfuric acid on sorbent capacity. EERC concluded that sulfuric acid and mercury compete for the same active binding sites on the sorbent surface and that the sulfuric acid is preferentially bound. Consequently, the concentration of sulfuric acid determines the number of remaining sites available for mercury capture. It is implicit from the proposed mechanism that flue gas SO₃ would have the same effect. Another important finding showed that the measured breakthrough curve often peaked above the inlet mercury concentration. This indicates that “captured” mercury is being desorbed, suggesting displacement by preferential flue gas species.

The demonstration at Plant Daniel starkly defined an installation where the success of sorbent injection as a viable mercury control technology hinges on the influences of the flue gas conditioning system. Where the FGC system is in normal operation at full SO₃ injection without any sorbent injection, native mercury capture was approximately 29 percent. A high ratio of sorbent injection in conjunction with the FGC improved mercury capture to about 60 percent. When SO₃ injection is fully stopped, mercury capture peaks at 88 percent. The inlet mercury concentration was nominally steady throughout these tests, the unit was held at a constant full load and the fuel was unchanged. This stepwise examination clearly delineates the positive effects of sorbent injection for mercury removal and negative effects of SO₃ injection. To put this another way, the use of FGC may result in more than a doubling of the unit cost of removing mercury. (Note that costs are specific to this example.) However, FGC may be essential in meeting particulate emission limits.

With federal and state mercury regulations looming and sorbent injection currently the most commercially-developed control technology, the industry is urgently trying to identify a means to avoid the substantial SO₃ inhibition effect. For many units, such a solution will be the difference between achieving economical compliance or expending a large capital investment.