Bench-scale Kinetics Study of Mercury Reactions in FGD Liquors

Abstract

This document is the final report for Cooperative Agreement DE-FC26-04NT42314, 'Kinetics Study of Mercury Reactions in FGD Liquors'. The project was co-funded by the U.S. DOE National Energy Technology Laboratory and EPRI. The objective of the project has been to determine the mechanisms and kinetics of the aqueous reactions of mercury absorbed by wet flue gas desulfurization (FGD) systems, and develop a kinetics model to predict mercury reactions in wet FGD systems. The model may be used to determine optimum wet FGD design and operating conditions to maximize mercury capture in wet FGD systems. Initially, a series of bench-top, liquid-phase reactor tests were conducted and mercury species concentrations were measured by UV/visible light spectroscopy to determine reactant and byproduct concentrations over time. Other measurement methods, such as atomic absorption, were used to measure concentrations of vapor-phase elemental mercury, that cannot be measured by UV/visible light spectroscopy. Next, a series of bench-scale wet FGD simulation tests were conducted. Because of the significant effects of sulfite concentration on mercury re-emission rates, new methods were developed for operating and controlling the bench-scale FGD experiments. Approximately 140 bench-scale wet FGD tests were conducted and several unusual and pertinent effects of process chemistry on mercury re-emissions were identified and characterized. These data have been used to develop an empirically adjusted, theoretically based kinetics model to predict mercury species reactions in wet FGD systems. The model has been verified in tests conducted with the bench-scale wet FGD system, where both gas-phase and liquid-phase mercury concentrations were measured to determine if the model accurately predicts the tendency for mercury re-emissions. This report presents and discusses results from the initial laboratory kinetics measurements, the bench-scale wet FGD tests, and the kinetics modeling efforts.