In view of increasing air pollution due to industrial processes, flue gas cleaning of conventional power plants is an ongoing topic. The fluctuating power generation with renewable energies requires a flexible operation of fossil fuel power plants, which influences the flue gas cleaning and the performance of air pollution control devices. The strict emission limits for mercury (Hg) as a toxic and a trace element in coal, emphasizes the need for a better understanding of its behavior in the flue gas path and the influence of changes in the flue gas and operating parameters on the involved reactions. The flue gas desulfurization (FGD) plant is the main sink for Hg from the flue gas, as the oxidized mercury is absorbed in its slurry. However, under unfavorable conditions, the absorbed oxidized mercury can be reduced and re-emitted to the gas phase, which influences its removal efficiency. The complex chemistry of Hg in the slurry makes it difficult to predict its behavior and the effects of different operating parameters and components on its reactions, which is vital for finding adequate measures for the Hg removal from the flue gas.
This work was conducted at a lab-scale FGD with synthetic flue gas and slurry, which provided vast possibilities for detailed studies. At first, the influence of the operating parameters as well as different slurry compositions on the partitioning of Hg in the FGD was investigated. It was found that at high temperatures of the slurry re-emission of Hg increased exponentially. In spite of exothermic characteristic of adsorption, the Hg concentration on gypsum increased at 80 °C compared to 60 °C. Increasing of the pH led to a higher Hg re-emission and a better sulfur dioxide (SO2) removal efficiency. When the Hg inventory in the slurry was increased, higher Hg re-emission were observed and the adsorption of Hg on the gypsum followed adsorption isotherm behavior. The ambivalent influence of S(IV) concentration on reactions involving Hg was observed, which can either lead to a reduction of oxidized mercury or to the formation of complexes. Ammonia was shown to have different effects on SO2 and Hg behavior depending on its concentration. The presence of halides induced a significant decrease in Hg re-emission depending on the type of existing halides. The limited adsorption capacity of gypsum could be observed in various slurries.
As Hg re-emission and S(IV) concentration are closely linked, a continuous S(IV) measurement can help to predict Hg re-emissions. Depending on the pH of the slurry, S(IV) exists in different forms of sulfite (SO32-), hydrogen sulfite (HSO3-) and dissolved SO2. By shifting the pH of the slurry to a proper range, only one of the mentioned forms of S(IV) is present in the slurry, which provides the possibility of total measurement of S(IV) concentration. Two different measurement principles for a continuous S(IV) measurement were tested and one was proven to be selective even in presence of different halides. Changing the SO2 concentration of the flue gas simulated the sudden increase of S(IV) in the slurry and showed the relation of Hg behavior and S(IV) concentration in different slurries. The role of SO32- and HSO3- as reducing agent and ligand for Hg-complexes was more obvious in the absence of halides. The results gave an insight into the involved reactions in the slurry.
Furthermore, additives, such as precipitating agents and sorbents, were tested as an additional measure to prevent Hg re-emission from the slurry. All of the additives eliminated Hg loss from the slurry during the transition phase and almost all of them showed lower re-emission at the steady state of the system by shifting the Hg inventory mostly to the solid fraction of slurry. The presence of halides changed the partitioning of Hg in the additive containing slurry even further and influenced the involved reactions. For activated carbon based sorbents, the pore distribution was proven to have an important role, which affected the adsorption capacity of sorbents at different media.