1 Department of Chemical and Biochemical Engineering, Technical University of Denmark
This thesis describes experimental and theoretical investigations of wet flue gas desulphurisa-tion (FGD). A review of the current knowledge of the various rate determining steps in wet FGD plants is presented. The experimental work covers laboratory studies as well as pilot- and full-scale experiments. In the theoretical part of the work, the laboratory and pilot plant observations are investigated using mathematical modelling.The mechanism underlying the rate of dissolution of finely grinded limestone particles was examined in a laboratory batch apparatus using acid titration. Three Danish limestones of dif-ferent origin were tested. A transient, mass transport controlled, mathematical model was de-veloped to describe the dissolution process. Model predictions were found to be qualitatively in good agreement with experimental data. Deviations between measurements and simulations were attributed primarily to the particle size distribution (PSD) measurements of the limestone particles, which were used as model inputs. The measured PSD was probably not representa-tive of a given limestone sample because of agglomeration phenomena taking place in the dis-perser, preventing a stable and accurate measurement. Other factors, such as convective mass transfer, porosity and porosity changes, and perhaps surface reaction, may also influence the rate of dissolution. However, those effects could not, due to the uncertainty of the PSD, be confirmed. Empirical correlations for the dimensionless mass transfer coefficients in a pilot plant (falling- film column) were determined. The correlations are valid at gas phase Reynolds numbers from 7500 to 18,300 and liquid phase Reynolds numbers from 4000 to 12,000, conditions of industrial relevance. The presence of inert particles in the liquid phase was found to decrease the rate of gas phase mass transport with up to 15 %, though the effect could not be correlated.A detailed model for a wet FGD pilot plant, based on the falling film principle, was devel-oped. All important rate determining steps, absorption of SO2, oxidation of HSO3-, dissolution of limestone, and crystallisation of gypsum were included. Model predictions were compared to experimental data such as gas phase concentration profiles of SO2, slurry pH-profiles, sol-ids contents of the slurry, liquid phase concentrations, and residual limestone in the gypsum. Simulations were found to match experimental data for the two Danish limestone types (Faxe Bryozo and a chalk, Mikrovit) investigated. Gas phase mass transport was found to be the dominating rate determining step, though the liquid phase mass transport resistance could not be neglected. Simulations and experimental data both showed the same degree of desulphuri-sation and absorber pH profile for the two limestone types using a holding tank pH of 5.5, but the residual limestone in the gypsum was significantly lower for the chalk. Furthermore, simulations showed that between 10 and 30 % of the limestone dissolves in the absorber de-pending on the process conditions. A typical holding tank pH of 5-5.5 (also used in full-scale wet FGD packed towers) was found to be a reasonable compromise between residual lime-stone in the gypsum and the degree of desulphurisation. Simulations were only slightly sensi-tive to the temperature in the interval 313 - 333 K, pertinent for full-scale wet FGD packed towers. The possibility of co-firing straw and coal was investigated in a full-scale power plant. No ef-fects on the overall performance of the wet FGD plant were observed, though laboratory ex-periments with fine dust and fly ash from the full-scale experiments showed a decrease in limestone reactivity. However, the test period was only about one week, probably not allowing the FGD plant to reach steady state operation. Pilot-scale experiments were initiated to investigate the possibility of oxidising spray dry scrubber by-products (TASP) to gypsum in wet FGD plants. The investigations showed that it was possible to oxidise the TASP at concentrations of up to 300 g/litre (feed tank basis). The residual Ca(OH)2 in the TASP could be exploited for desulphurisation resulting in a reduced limestone consumption and an increased removal rate of SO2. The simultaneous presence of Al3+ and F- did not, as reported in the literature, inhibit the rate of dissolution.The results of this work demonstrate that mathematical modelling can be a powerful tool in the process of obtaining a detailed understanding of the complex interactions between chemi-cal reactions and mass transport phenomena in wet FGD plants.