(599c) Kinetics of Sulfur Deactivation of Steam Reforming Catalyst | AIChE

(599c) Kinetics of Sulfur Deactivation of Steam Reforming Catalyst

Authors 

Czernik, S. - Presenter, National Renewable Energy Laboratory
Jablonski, W. - Presenter, National Renewable Energy Laboratory


We are developing a gasification process to produce clean syngas from biomass feedstocks. The syngas can be further processed to produce fuels and chemicals such as gasoline, diesel, alcohols, olefins, synthetic natural gas, and high purity hydrogen. The raw biomass-derived syngas contains unwanted components such as methane, other gaseous hydrocarbon and polycyclic aromatic compouds (collectively called tars), which must be removed prior to further processing. Steam reforming using nickel-based catalysts is the preferred method to remove these impurities and convert them to additional amounts of carbon oxides and hydrogen. The main challenge in the tar removal process is the presence of hydrogen sulfide, a known catalyst poison, in biomass-derived syngas. Depending on the biomass feedstock, the syngas produced in our pilot plant contains 20-600 ppm H2S and steam reforming catalysts loose activity after a relatively short time on stream (30-50 minutes). In support of our catalyst development effort, we conducted a catalyst deactivation and regeneration study using a fixed-bed micro reactor system (0.5-2.0 g catalyst) with model tar compounds (ethylene and benzene). This work reports on the results of the catalyst activity during steam reforming of ethylene containing different amounts of H2S. The product gas composition was measured at the reactor outlet every three minutes during the test, which allowed us to determine conversion of ethylene as a function of time on stream. When the H2S concentration in ethylene was 100 ppm or higher, the catalyst performance measured as ethylene conversion decreased very rapidly and the sampling frequency was too low to provide reliable time-conversion data for kinetics calculation. At lower H2S concentration the loss of the catalyst performance was slow enough and allowed an attempt to determine the deactivation kinetics calculation. The experimental data are well described by a catalyst deactivation model with residual activity, which will be presented in this paper.