Potassium Capture by Kaolin and Coal Fly Ash

Abstract

Combustion is an important way for bio-energy production. Suspension-firing boilers, also called pulverized fuel combustion boilers (PF-boilers) are increasingly used for production of power and heat from biomass. Combustion of biomass in suspension-fired boilers can produce renewable, CO2-neutral electricity with a higher electrical efficiency compared with that of grate-fired boilers. However, during the combustion of biomass, significant amount of K-species are released to gas phase in the boiler chamber, as KOH, KCl and K2SO4, consequently leading to deposit formation, corrosion as well as de-activation of SCR (Selective Catalytic Reduction) catalysts. One option to tackle these ash-related problems is to use additives to capture gaseous alkali species in flue gas. Kaolin and coal fly ash are two effective Al-Si based additives to capture alkali species. The mechanisms and kinetics of alkali capture by kaolin and coal fly ash have previously been investigated in several studies published in the open literature. However, most of these studies were conducted in fixed bed reactors where kaolin or coal fly ash pellets or flakes were utilized. In suspension firedboilers, kaolin and coal fly ash particles are well dispersed in flue gas, with a total  residence time of a few seconds and the controlling mechanisms could be considerably different from that in fixed bed reactors. In this Ph.D. project, the reaction between gaseous potassium species (KOH, K2CO3, KCl and K2SO4) and different Al-Si based additives (kaolin, mulliteand coal fly ash) under well-controlled suspension-fired conditions was investigated by performing experiments in the DTU entrained flow reactor (EFR). The K-capture level of additives CK (g K/(g additive)) and K-conversion XK (%) were quantified by analyzing the solid products (CK is the mass of potassium captured by 1 g of additive; XK is the percentageof fed potassium captured by additive). The impact of different parameters, such as K concentrationin flue gas (50-1000 ppmv), molar ratio of K/(Al+Si) in reactants (0.048-0.961), reaction temperature (800-1450 °C), gas residence time (0.6-1.9 s), additive particlesize as well as the type of coal fly ashes on the K-capture reaction was studied. Corresponding equilibrium calculations were carried out using the equilibrium module of FactSage 7.0, to shed light upon how far the EFR reaction system is from the equilibrium and provide information for understanding the EFR experimental results. The results of the K-capture experiments using kaolin at 1100 °C or 1300 °C showed that for all the four K-species, KOH, K2CO3, KCl and K2SO4, the K-capture level (CK) increased considerably when the K-concentration in the flue gas changed from 50 ppmv to 500 ppmv(molar ratio of K/(Al+Si) in reactants varied from 0.048 to 0.481). However, no obvious increase of CK was observed when the K-concentration increased further to 750 ppmv and1000 ppmv (molar ratio of K/(Al+Si) in reactants was 0.721 and 0.961). This is probably because all kaolin Si has been consumed forming K-aluminosilicates at 500 ppmv. Results of K-capture experiments using kaolin at different temperatures show that, for KOH, KCl and K2CO3, the K-capture level (CK) and K-conversion (XK) by kaolin generally followed the equilibrium predictions at 1100 °C and above, when using a kaolin particle size of D50 =5.47 μm and a gas residence time of 1.2 s. This reveals that a nearly full conversion of kaolin to K-aluminosilicates was achieved without kinetic or diffusion limitations under the applied conditions. At 800 °C and 900 °C, the measured conversions were lower than the equilibrium predictions, indicating that the reactions were either kinetically or diffusion controlled. For K2SO4, the measured CK was obviously lower than the equilibrium predictions. Kaliophilite (KAlSiO4) was predicted by the equilibrium calculations; however,the XRD analysis results revealed that leucite (KAlSi2O6) was actually formed. Results of KOH capture experiments by kaolin of different particle sizes showed that, fine kaolin powder (D50 = 3.51 μm) and normal kaolin powder (D50 = 5.47 μm) behaved similarly, while coarse kaolin (D50 = 13.48 μm) showed a smaller K-capture level (CK) at 1100 and1300 °C. This is probably because KOH diffusion into the kaolin particles became a limiting factor for the coarse kaolin at 1100 °C and above. At 900 °C the difference was smaller, probably because the reaction is more kinetically controlled and the additive particle size did not influence the reaction significantly at 900 °C. Results of KOH capture by kaolin at different residence times showed that the K-capture reaction reached equilibrium at 1300 and 1450 °C, with a gas residence time of 1.2 s and akaolin particle size of D50 = 5.47 μm. However, at 800 °C, CK is obviously far away from the equilibrium even with a longer residence time of 1.9 s, showing that the reaction is more kinetically or diffusion controlled. Similar results were observed for KCl capture by kaolin. Results of K-capture experiments by kaolin using different K-species show that, for KOHand K2CO3, leucite (KAlSi2O6) was formed at low K-concentration of 250 ppmv, while athigher K-concentration (500-1000 ppmv), kaliophilite (KAlSiO4) was detected in theproducts. But in the experiments with KCl and K2SO4, only leucite (KAlSi2O6) was detectedby XRD analysis. Another difference was that the CK of K2CO3 was comparable to that of KOH, while CK of KCl and K2SO4 by kaolin were both relatively lower.The results of K-capture experiments using coal fly ash showed that the behaviors of thestudied four K-species were similar to what was observed when using kaolin. CK and XKincreased when K-concentration increased from 50 ppmv to 500 ppmv (molar ratio of K/(Al+Si) in reactants varied from 0.048 to 0.481), and they did not increase further at Kconcentrationof 750 ppmv and 1000 ppmv (molar ratio of K/(Al+Si) in reactants was 0.721and 0.961). One difference observed was that at 250 ppmv K and above, the measured CK and XK of coal fly ash was lower than the equilibrium predictions. In addition, compared with kaolin, although the types of formed K-aluminosilicates agreed with that of kaolin, coalfly ash captured the K-species less effectively at a K-concentration higher than 250 ppmv(molar ratio of K/(Al+Si) in reactants changed from 0.240 to 0.961).Results of K-capture experiments by coal fly ash at different temperatures showed that, at 800 °C, the KOH-capture reaction was kinetically controlled. At 900-1300 °C, the reaction was both diffusion and kinetically influenced. At 1450 °C the reaction was equilibrium and diffusion influence. Results of 500 ppmv KOH capture by coal fly ash (with a molar ratioK/(Al+Si) = 0.481 in the reactants) showed that, CK of coal fly ash generally increased whenthe temperature increased from 800 °C to 1450 °C, but the CK of coal fly ash was lower than that of kaolin through the whole temperature range studied. At 50 ppmv KOH (K/(Al+Si) =0.048) which is comparable to the K-concentration under practical full-scale wood combustion conditions, CK of the coal fly ash was comparable both to the equilibrium data and to the CK of kaolin. Results on CK of KOH-capture by coal fly ash of different particle sizes showed that decreasing the coal fly ash particle size could increase the K-capture level. Comparison of KOH-capture by kaolin, mullite and coal fly ashes at 500 ppmv KOH showed that kaolin is the most effective additive for alkali capture followed by mullite and coal flyashes. However, no obvious difference was observed between the three additives at 50 ppmv KOH. In addition, when the reaction temperature was varied from 800 to 1450 °C, CK of kaolin firstly increased then decreased, reaching peak at 1300 °C. CK of mullite and coal fly generally increased with increasing temperatures from 800 to 1450 °C.