Steam reforming of bio-oil: Effect of bio-oil composition and stability

Hydrogen can be obtained from biomass pyrolysis liquids by catalytic steam reforming. Catalyst deactivation by coking and the formation of carbon deposits are the major known limitations although the specific causes are unidentified. It is proposed that these limitations could be reduced by selectively reforming specific fractions of the bio-oil. The hydrophobic fraction mainly composed of heavy oligomers can be separated by water extraction as a high-value product. In the current work, the particular interest is on efficient hydrogen production from the residual aqueous fractions. There are certain challenges introduced due to the complexity of these mixtures. An improved understanding of the underlying steam reforming characteristics of pyrolysis liquids compounds will provide the opportunity to predict hydrogen yields and to potentially identify undesired compounds. Different fractions were studied to determine which ones can be used more efficiently to produce hydrogen with existing steam reforming technologies. A fractionating condenser system was used to collect different fractions of fast pyrolysis liquids. The hydrophilic components from the different fractions were analyzed and steam reformed using a nickel-based commercial catalyst loaded fixed bed reactor. First it was found that bio-oil injection into this kind of reactors is an issue since there has to be quick and effective contact of the bio-oil with the catalyst and avoiding preheating of the oils. When reforming at 500°C, an overall positive effect in hydrogen yields was observed for the fractions with higher concentrations of lower molecular weight oxygenated compounds like acetic acid and acetol while the effect heavier compounds may be detrimental. Higher selectivity towards hydrogen suggests less selectivity towards carbon deposits. Mostly the samples that contained the heavier compounds also contained less of the light compounds so that their individual effects need to be deconvoluted. The coking phenomena needs to be further studied to find the species and reaction conditions that promote its deposition. It was found that the changes caused by aging effect of aqueous bio-oil had a detrimental effect in hydrogen production but the causes are not well understood. Also seemingly insignificant traces of potential catalyst poisons found in bio-oils can irreversibly deactivate the nickel reforming catalyst difficulting their study.