Open Collections

Reaction of lime and coal mineral matter with hydrogen

University of British Columbia

Removal of H₂S from gasification product gases is a necessary step in integrated gasification combined cycle power generation. H₂S capture by lime and coal ashes at 850°C from N₂ and N₂-steam mixtures with 200-2300 ppmv inlet H₂S concentration was studied in a packed bed reactor. Experiments were carried out on Texada lime (d[sub p] = 770 μrn) and ash from three kinds of coal: Highvale, Costello, Minto, with particle diameters between 144 and 303 um. The breakthrough curves of H₂S* under various conditions were obtained from N₂ containing up to 40 vol% steam content. In the absence of steam, Texada lime was superior to all coal ashes for H₂S capture, yielding spent sorbents with greater than 35 wt% S compared to the possible 44.5 wt% S for pure CaS. Coal ash from Highvale and Costello coals reacted 4 and 10 wt% S, respectively. Coal ash appeared to promote some H₂S cracking, rather than solely sulphide formation. For Mint ash, sulphide level of 18.8 wt% was achieved. Iron oxide appears to be the primary reacting species in this ash. The results indicate that the sulphur capture by lime is limited by the equilibrium of the H₂S-CaO reaction. In the presence of steam, sulphur removal by Texada lime was reduced. As the steam content increased from 0 vol% to 40 vol%, the sulphur content in spent Texada lime (0.3 g) decreased from 35 wt% to 24 wt%. For spent Costello ash containing the equivalent amount of CaO, S content decreased from 10 wt% to 8 wt%. For Highvale ash, no H₂S was captured in the presence of steam, on the contrary, some residual sulphur in ash was released to the gas. With Highvale and Costello ash, SO₂ was detected; the SO₂ production was enhanced by increasing amount of sorbent, inlet H₂S concentration, and steam content. The forward reaction of dry H₂S and CaO in the fixed bed was analyzed by available reaction models. The disappearance of H₂S could be modeled as a first order process with rate constant kʹ=1.275 (min⁻¹) in the early stage of reaction. To obtain appropriate rate expression for the late stages of reaction, the grain model was applied to the fixed bed system to describe the single particle behavior. The chemical reaction constant on the grain surface is evaluated as k=1.85x10⁻⁴(cm/s) by assuming the initial reaction rate is controlled by chemical reaction only. The influence of effective intrapellet diffusion and diffusion in the product layer of each grain were also considered in the overall reaction rate expression. The approach of Krishnan and Sotirchos where effective diffusivities decrease as increasing solid conversion, X, was applied to this case. By assuming reaction being in the extreme cases, the resulting intrapellet diffusivity, De[sub eff] and intragrain diffusivity, De , were 0.3e⁻¹⁰[sup x] (cm²/s)and l 1.1x10⁻⁵e⁻¹³ [sup x]. (cm²/s) respectively, and they both provided good prediction on the solid conversion in the present system.

Thesis/Dissertation

application/pdf

2009-01-10

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/3534

Chemical and Biological Engineering

University of British Columbia

UBCV

DSpace