Carbon Dioxide Gasification of Kraft Black Liquor Char in a Laboratory Char Bed Reactor

In the current study a series of pyrolysis and CO2 gasification experiments were conducted on a bed of kraft black char under a perpendicular impinging jet. The experiments demonstrated that the char bed carbon gasification rates are externally mass transfer limited. The mass transfer is further decreased by a net molar out-flow caused by the conversion of one mole of reactant CO2 to two moles of product CO via gasification. Predictions of the mass transfer coefficient using the COMSOL MultiPhysics finite element program were comparable with those obtained from the experimental data. Introduction The char bed is perhaps the least understood portion of a kraft black liquor recovery boiler. The comprehensive studies of A.D. Little reported by Merriam & Richardson in 1978 and 1979 still provide most of the present day understanding about the char bed. A one-dimensional model of the mass and energy flows inside the char bed was also developed. The predictions of the model were in general agreement with the experimental results and earlier bed and gas temperature measurements reported by Borg et al. (1974). Data collected suggests that CO2 gasification plays a key role in consumption of organic carbon in the active layer of the char bed but no comprehensive rate expression was developed. The char bed combustion processes have been reviewed by Grace and Frederick (1997) and Grace (2001). Based on the experimental studies referred to above, there is a general consensus that the burning rate of the char bed is limited by mass transfer of oxygen and gasification gases (CO2 and H2O) and by radiation heat transfer to the bed surface needed for the endothermic carbon consuming reactions (sulfate reduction, gasification and sodium carbonate reduction). These reactions are: C(s) + O2 (g) → CO2 (g) (1) C(s) + CO2 (g) → 2 CO (2) C(s) + H2O(g) → CO(g) + H2 (g) (3) 2 C(s) + Na2CO3(s,l) → 2Na(g) + 3CO(g) (4) 4 C(s) + Na2SO4 (s,l) → Na2S(s,l) + 4 CO(g) (5) The formation of CO2 is not included in the overall reactions (4) and (5) because reaction (2) quickly leads to CO formation at the temperatures in the bed. It was realized by Merriam & Richardson (1978) that there is a net outflow of gas from the bed due to formation of two moles of CO per mole of CO2 consumed in reaction (2). As a result nitrogen would be swept out of bed, and its concentration would be lower than that in a combustion gas. Merriam & Richardson (1978) assumed in their model that reaction (2) would be at equilibrium with carbon present in excess. Additional CO produced by sulfate reaction (5) was also considered, but reactions (3) and (4) were not. Based on a recent study of gaseous samples collected from a char bed (Connolly & Van Heiningen 2004) it was shown that the change in gas composition inside the bed is consistent with a net outflow of gas from the bed. In earlier laboratory char bed experiments (Brown et al, 1989 & Kochesfahani et al, 1998) air jets parallel to the bed surface (with or without CO2 and H2O added) were used to simulate the primary air jets sweeping over the char bed. Since the reactivity of oxygen with organic carbon is very high, all oxygen reacts at the interface rather than penetrate into the char layer as might occur for CO2 and steam (Sutinen, Karvinen, & Frederick, 2002). A unique aspect of the two experimental char bed studies is the continual replenishment of organic carbon at the char surface by using a moving bed arrangement. Lee and Nichols (1997) studied gasification of black liquor char by feeding CO2 through a packed bed of the char. They found that the gasification rate was limited by chemical kinetics below 700C, but that diffusional limitations became important at higher temperature. In the present study the interaction between carbon gasification by CO2 and mass transfer processes was investigated for a black liquor char bed at temperatures of 675 -815C with a gas jet impinging on its top surface. The measured carbon gasification rate is interpreted in terms of external and internal mass transfer resistances, and the influence of net outflow of gasification gas from the char bed is quantified. Experimental Setup The impinging jet reactor set-up is depicted in Figure 1. The feed gas flow system includes mass flow controllers and a coiled heat exchanging tube inside the reactor to deliver an accurate flow and composition of gas preheated to the bed temperature. An Agilent 3000 Micro GC equipped with a TCD cell and two separation columns was used to analyze the product gas for permanent gases, N2, O2, CO2, H2, CH4 & CO every three minutes. CO and CO2 gas concentrations were also continually monitored by an online Siemens CO/CO2 analyzer. The aluminum oxide cup under the impinging jet contained the bed of pyrolyzed kraft black liquor char. The reactor element was heated by a pottery kiln. Product gases were cooled by a shell and tube heat exchanger followed by water trap filter prior to gas analysis. The gas temperature just below the nozzle and just above the char bed was measured by thermocouples. The inner diameter of the impingement nozzle is 7 mm, the diameter of the char bed is 105 mm, and the distance from the nozzle to the char bed surface is 149.3 mm. To better understand the CO2 gasification behavior of the kraft black liquor char bed, five variables were investigated: bed thickness, impinging jet flow rate, gas diffusivity (helium versus nitrogen as carrier gas for reactant CO2), bed temperature, and CO2 concentration in the jet. The experimental reactor was brought up to the desired steady state temperature under an impinging flow (1 L/min) of nitrogen. When the desired temperature was reached, CO2 was added at a flow rate to obtain the desired gas composition and total flow rate. The char bed behavior during the heat up period will be reported as char pyrolysis. The steady state CO2 gasification conditions are summarized in Table 1. Figure 1. CO2 Gasification Reactor Setup Table 1. CO2 Gasification Experimental Conditions Experiment Gasification Gas Impinging Jet Flow Rate [L/min] Bed Mass [grams] Bed Thickness [mm] Bed Temperature [C] 1 N2 , 10% CO2 3.3 100 19 815 2 N2 , 44% CO2 3.3 100 19 815 3 N2 , 89% CO2 3.5 100 19 815 4 N2 , 83% CO2 2.9 50 10 815 5 N2 , 84% CO2 2.9 200 38 815 6 N2 , 15% CO2 4.5 100 19 815 7 He, 18% CO2 4.1 100 19 815 8 N2 , 85% CO2 2.9 100 19 675 9 N2 , 88% CO2 2.8 100 19 750 10 N2 , 85% CO2 1.0 50 10 815 11 N2 , 85% CO2 6.0 100 19 815 Data Analysis The solid residues remaining after gasification were dissolved in water. The dry weight of the solids remaining after dissolution was determined. The following analyses were performed on the solution: cyanate (OCN), sulfite (SO3), thiosulfate (S2O3) and chloride (Cl) using ion chromatography, carbonate (CO3) as CO2 using headspace gas chromatographic analysis, sodium and potassium using ICP, and hydroxide, sulfide and carbonate using ABC titration. The focus of this paper, however, is on the gas phase results. Reaction of gasification gas with the reactor wall Close examination of the effluent gas composition showed that the gasification gases produced by the char bed reacted with the outer reactor vessel made from Inconnel. This made it impossible to determine the organic carbon consumption from the CO production rate using the CO2 gasification reaction (equation 2). A series of experiments were performed with different CO2 and CO concentrations fed to the empty reactor (Table 2). It was found that excess CO was converted into CO2 on a one to one molar basis and vice versa. Table 2. Empty Reactor Experiments with CO and CO2 Feed Total Volumetric Feed rate [liters/min] CO Feed rate [mmole/min] CO2 Feed rate [mmole/min] CO Conversion Rate [mmole/min] CO2 Conversion Rate [mmole/min] Molar ratio CO/CO2 1.395 8.18 12.39 -1.74 1.71 -0.98 1.498 12.27 12.31 -2.18 2.43 -1.12 1.167 4.09 11.94 -0.22 0.22 -1.02 3.617 10.22 24.50 -4.96 4.59 -0.93 4.005 28.63 118.16 -5.88 7.08 -1.20 3.464 6.13 24.29 -2.01 2.38 -1.18 3.568 10.22 24.29 -3.27 3.27 -1.00 3.799 0 31.08 1.17 -1.24 -1.07 Based on literature information (Da Silva, 1960) it seems likely CO reacted with magnetite (Fe3O4) of the Inconnel producing CO2 production by the equilibrium reaction (6): Fe3O4 (s) + CO (v) ↔ 3FeO (s) + CO2 (v) (6) At 815C the equilibrium of reaction (6) is shifted towards the right, i.e. it favors CO2 formation from CO (Table 3). Furthermore this reaction is thermodynamically unfavorable below 375C, which confirms our finding that 5%CO in the feed to the reactor was stable below 375C (see Figure 2). This figure also shows that CO in the feed is converted into CO2 on a one to one molar basis during the heat up of the reactor from about 350 to 600C. Above 600C both CO and CO2 formation occurs due to further pyrolysis and gasification of the black liquor char. Table 3. Equilibrium Constants for Reduction of Magnetite (Fe3O4) by CO (equation 6) Temperature (C) 300 400 500 600 700 800 900 Equilibrium Constant (keq) 0.576 1.108 1.684 2.181 2.525 2.692 2.697