In order to secure the energy supply to an increasing population and at the same time limit the damage to Earth, i.e. avoiding a fatal climate change as a result of anthropogenic emissions of greenhouse gases (primarily CO2), immediate action is necessary. This includes reducing the energy consumption, increasing the energy conversion efficiency, and using renewable energies. The transport sector is the one most dependent on fossil energy and it stands for a significant part of the energy consumption in the world. For instance, in EU-25 transportation stands for 30 % of the total final energy consumption and relies to 98 % on oil. Being the only renewable energy possible to convert into liquid fuels biomass, as a means for reducing the CO2 emissions from the transport sector, will play an important role in the near future.
The conversion of biomass into transportation fuels is preferentially done via gasification followed by the fuel synthesis. The whole production chain from biomass to final fuel is very dependent on R&D, in order to become competitive with the fossil fuels. Fischer-Tropsch (FT) diesel made from biomass is a viable option for reducing the CO2 emissions from transportation since it may be blended with conventional diesel in any concentration. Furthermore, since its composition is almost the same as of petroleum-based diesel (although cleaner) the same distribution system and engines may be used, which facilitates its introduction on the market. This thesis presents the results of the laboratory work performed in 2003 – 2007 at the Department of Chemical Technology, KTH, and at the Department of Chemical Engineering, NTNU (the Norwegian University of Science and Technology) in Trondheim. Part of the work has been performed in close cooperation with the Department of Chemical and Biological Engineering at Chalmers University of Technology.
All FT experiments were performed in a fixed-bed reactor at 210 ºC and 20 bar. Pure mixtures of H2, CO and N2 were used as feed to the reactor. Steam was also occasionally introduced. Selectivity to C5+ was used as a measure of the catalysts’ ability to grow long-chain hydrocarbons, which is desirable when diesel is the product aimed for.
The first part of the thesis deals with the direct conversion of a H2-poor syngas, which is obtained upon gasification of biomass, into FT hydrocarbons. “H2-poor” means that the H2/CO ratio is lower than what is required by the stoichiometry (~ 2.1) of the FT synthesis (reaction 1). In order to increase the H2/CO ratio to the required one, internal water-gas-shift (WGS) is needed (reaction 2).
FT: CO + 2H2 “-CH2-“ + H2O (1)
WGS: CO + H2O CO2 + H2 (2)
The H2/CO usage ratio has been used as a measure for the internal WGS activity, it is defined as follows:
where S is the selectivity (of total C-containing products), and the factor F indicates the number of H2 moles required for one CO mole to form the product (e.g. for producing high molecular weight n-paraffins, 2 moles of H2 per mole of CO are required). For the fraction C2 – C4, F will have different values depending on the selectivity to C2, C3 and C4, and it also depends on the olefin/paraffin ratios for those hydrocarbons. In order to reach the highest once-through conversion of the syngas, the H2/CO usage ratio should be equal to the inlet H2/CO ratio. The lower the usage ratio, the higher the relative WGS activity.
The combined FT and WGS reactions with H2-poor syngas have been tested for 12 wt% Co and 12 wt% Co – 0.5 wt% Re catalysts supported on γ-Al2O3. It was found that with lower H2/CO ratios in the feed, the syngas conversion and the CH4 selectivity decreased, while the C5+ selectivity and olefin/paraffin ratio for C2-C4 increased slightly. The WGS activity was low for all catalysts, implying a H2/CO usage ratio close to the stoichiometric one (2.1), even for inlet H2/CO ratios of 1.5 and 1.0.
By incorporating significant amounts of Fe (20 % of total metal) in the Co catalyst referred to above (by co-impregnation) in order to achieve a 12 wt% bimetal loading on γ-Al2O3, a slightly lower usage ratio was obtainable (1.92) for an inlet ratio of 1.0 for dry conditions. Different Fe:Co ratios ranging from 100 % Fe to 100 % Co (12 wt% bimetal) were tested for an inlet H2/CO ratio of 1.0. The characterisation results indicated that Fe was enriched at the surface, hence covering the more FT-active Co sites, even at low percentage Fe. Not even upon significant replacement of Co by Fe (≥ 20 %) were the usage ratios for dry conditions lowered to any significant extent.
The WGS reaction at the low temperature used could be boosted by addition of external water. However, since the catalysts with the highest WGS activity had surface enrichment of Fe, high water partial pressures negatively affect the FT rate and also lead to rapid deactivation by re-oxidation of the FT-active iron phases (iron carbides) or by sintering. Surprisingly, also the WGS activity rapidly deactivated at high partial pressures of water, although Fe3O4 is believed to be the WGS-active phase. Possibly, surface oxidation of Fe3O4 into γ-Fe2O3 may take place during these water-rich conditions.
The WGS reaction (reaction 2) is thermodynamically favoured at low temperatures. However, the WGS kinetics over the tested catalysts is far too slow at the FT reaction temperature used in typical low-temperature FT (LTFT) applications, with diesel as the desired product.
The second part of this thesis deals with the hydrocarbon selectivity over Co-based catalysts in the FT synthesis with stoichiometric H2/CO feed. An alternative catalyst preparation technique, the microemulsion (ME) technique, was used as a complement to the conventional incipient wetness (IW) impregnation. A great deal of effort was put into the development of the synthesis of Co particles of monodisperse size in a ME and the subsequent deposition of these onto a porous support material. By using the ME technique it was possible to prepare relatively small Co particles in a large-pore support such as TiO2, which is not an easy task using the conventional impregnation technique.
With the prepared ME catalyst on TiO2 it was possible to study the effect of Co particle size on FT selectivity irrespective of the pore size of the support. It was found that, at least for Co particles above 10 nm, it is not the Co particle size that is the principal parameter determining the selectivity of a catalyst, but rather the physical and/or chemical properties of the support.
Due to the smaller Co particle size (~ 12 nm) of the ME-TiO2 catalyst as compared to a corresponding IW catalyst (~ 26 nm) supported on TiO2, the FT activity of the former was 100 % higher as fresh catalyst. After 120 h on stream the ME catalyst still gave 60 % higher rate to C5+ (expressed as gC5+/gcat,h) compared to both an IW-TiO2 and an IW-γ-Al2O3 catalyst, all catalysts having the same composition (12 wt% Co, 0.5 wt% Re).
Stockholm: KTH , 2007. , xiii, 90 p.
2007-05-25, Kemisk teknologis bibliotek, KTH, Teknikringen 42, 1tr, Stockholm, 10:00