Sulfur emission control in fossil fuel gasification plants implies the removal of H2S from the product gas either inside the furnace or in the gas clean-up system. In a fluidized-bed gasifier, in-bed sulfur capture can be accomplished by adding a calcium-based sorbent such as limestone or dolomite to the bed and removing the sulfur from the system with the bottom ash in the form of CaS. This work describes the H2S uptake by a set of physically and chemically different limestones and dolomites under pressurized conditions, typically for those in a pressurized fluidized-bed gasifier (2 MPa, 950 °C). The tests were done with a pressurized thermobalance at two p CO 2 levels. Thus, the sulfidation of both calcined and uncalcined sorbents could be analyzed. The effect of p H 2 S was also investigated for uncalcined limestones and half-calcined dolomites. The results are presented as conversion of CaCO3 or CaO to CaS vs time plots. The results are also compared with the sulfur capture performance of the same sorbents under pressurized combustion conditions.

The physical structure of a limestone or dolomite to be used in in-bed sulfur capture in fluidized bed gasifiers has a great impact on the efficiency of sulfur capture and sorbent use. In this study an unreacted shrinking core model with variable effective diffusivity is applied to sulfidation test data from a pressurized thermogravimetric apparatus (P-TGA) for a set of physically and chemically different limestone and dolomite samples. The particle size was 250−300 μm for all sorbents, which were characterized by chemical composition analysis, particle density measurement, mercury porosimetry, and BET internal surface measurement. Tests were done under typical conditions for a pressurized fluidized-bed gasifier, i.e., 20% CO2, 950 °C, 20 bar. At these conditions the limestone remains uncalcined, while the dolomite is half-calcined. Additional tests were done at low CO2 partial pressures, yielding calcined limestone and fully calcined dolomite. The generalized model allows for determination of values for the initial reaction rate and product layer diffusivity.

Regenerable mixed metal oxide sorbents are prime candidates for the removal of hydrogen sulfide (a major pollutant) from the hot coal gas in the simplified integrated gasification combined cycle processes. As part of this sulfur removal process development, reactor models are needed for scale-up. It is essential for this work to apply a reliable and simple modeling correlation for the conversion rate of zinc sulfide or oxygen in the regeneration reaction. The fit of such a model, assuming uniform conversion of the ZnS, was obtained from ambient pressure thermogravimetric analyzer test data with sulfided zinc titanate samples. An activation energy of about 140 kJ/mol was obtained for the regeneration reaction rate constant.

Fluidized bed combustion or gasification allows for in-bed sulfur capture with a calcium-based sorbent such as limestone or dolomite. Sorbent particle size, porosity, internal surface, and their variation during conversion have great influence on the conversion of the sorbent. The uptake of SO2 and H2S by five physically different limestones is discussed, for typical pressurized fluidized bed combustor or gasifier conditions:  850/950 °C, 15/20 bar. Tests were done in a pressurized thermogravimetric apparatus (P-TGA), the size of the limestone particles was 250−300 μm. It is stressed that the limestones remain uncalcined. A changing internal structure (CIS) model is presented in which reaction kinetics and product layer diffusion are related to the intraparticle surface of reaction, instead of the outer particle surface as in unreacted shrinking core (USC)-type models. The random pore model was used for describing the changing internal pore and reaction surfaces. Rate parameters were extracted for all five limestones using the CIS model and a USC model with variable effective diffusivity. Differences in the sulfur capture performance of the limestones were evaluated. Plots of the CaSO4 or CaS product layer thickness as a function of conversion are given, and the relative importance of limestone porosity and internal surface is discussed.

No abstract