Waste to fuel: designing a cobalt based catalyst and process for once-through Fischer-Tropsch synthesis operated at high conversion

Doctoral Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
The production of fuels from waste on a small-scale decentralized level may enhance the liquid fuel security of Sub-Saharan Africa. The Fischer-Tropsch process can be used to convert waste into drop-in fuels. However, operating at small scale in remote locations requires a plant design with lower capital requirements, a greater level of simplicity and utility self-sufficiency. A plant design using an air-fed biogas reformer (without an air separation unit) and a single pass Fischer-Tropsch configuration is proposed. A fundamental requirement of this particular design is that it needs to operate at a higher FischerTropsch conversion than typically seen in industry (55 – 65%). High conversion conditions result in a high partial pressure of H2O and low partial pressures of CO and H2 within the Fischer-Tropsch reactor. These conditions have been reported to negatively affect the activity, selectivity and stability of cobalt-based catalysts. To date, no study has investigated the cause of this phenomenon, nor has a catalyst been developed specifically to operate under high conversion conditions. The objective of this study is to investigate the mechanisms behind these phenomena and provide catalyst design improvements that facilitate operation at high conversion conditions. Furthermore, a detailed design of the proposed once-through Fischer-Tropsch biogas-to-fuel plant will be evaluated using data from the catalytic experiments. An investigation into the effect of high conversion on the activity and selectivity of 0.05Pt-22Co/Al2O3 was conducted in a slurry bed reactor at T = 220°C, P = 20 bar, with a feed simulating synthesis gas generated from air-blown reforming (H2:CO:N2= 4:2:6). Space velocity was decreased to increase conversion to between XCO = 40% and XCO = 97%. The rate of CO consumption decreased with increasing conversion. Increasing the CO conversion was found to have negligible effect on CO2 selectivity (an unwanted by-product) up to a CO conversion of 75%, after which a strong increase was observed. This was attributed to the enhanced of water-gas shift activity of Co0 under hydrothermal conditions. The production of CO2 raised the H2/CO ratio within the reactor resulting in a large increase in the CH4 selectivity (an unwanted by-product), a decrease in the chain growth probability and thus a decrease in the C5+ selectivity (fuel product). In order to improve unfavourable selectivity obtained at high conversion in the Fischer-Tropsch synthesis, the effect of adding manganese (Mn) to 0.05Pt-22Co/Al2O3 was explored. The catalyst (0.05Pt22Co/Al2O3) was impregnated with increasing amounts of manganese, resulting in six Mn-Pt-Co/Al2O3 catalysts with Mn:Co mass ratios of 0, 0.04, 0.09, 0.14, 0.23 and 0.47:1. The optimal level of manganese promotion was found at a Mn:Co mass ratio of 0.14:1. At this level of manganese promotion, CO2 and CH4 selectivity was decreased by up to 6 C-% and 12 C-% (XCO = 90%) respectively whilst turn-over frequency was improved by 100%. The maximum in the C5+ yield as a function of CO-conversion was shifted from XCO = 78% to XCO = 91%, thus making operation at high conversion feasible from an activity and selectivity perspective. Operating Pt-Co/Al2O3 at conversion levels higher than XCO = 70% was shown to lead to rapid irreversible deactivation, with a total activity loss of 50% between XCO = 70% and 97%. Using a combination of spent catalyst characterisation via transmission electron microscopy, temperature programmed reduction/hydrogenation as well as an in-situ magnetometer, this irreversible deactivation was attributed to both sintering and cobalt aluminate formation. At very high conversion (XCO > 97%) enhanced reversible deactivation was exhibited due to the oxidation and re-reduction Co0 to Co(II)O. This oxidation/reduction cycle is the result of a thermodynamic conversion limit: at a mean Co0 crystallite size of 6 nm (as obtained with Pt-Co/Al2O3), the maximum achievable conversion (assuming a lognormal distribution of crystallites, σ = 0.5) is XCO = 88%. A log-normal distribution of cobalt crystallites with an average size of 8 nm (and the same variance) is required to obtain a maximum conversion of up to XCO = 98%. In order to limit deactivation due to cobalt aluminate formation at conversions higher than XCO = 70%, zinc aluminate was investigated as a novel support material for a platinum promoted cobalt catalyst. Zinc aluminate thermodynamically limits the formation of cobalt aluminate and facilitates the formation of larger sized cobalt crystallites. The catalyst, 0.04Pt-23Co/ZnAl2O4 exhibited minimal signs of irreversible deactivation at high conversion with a total rate loss of 0.08 mmol /min/g (0.62 to 0.54 mmol /min/g), whilst the rate loss over 0.05Pt-22Co/Al2O3 amounted to 0.47 mmol /min/g (from 0.74 to 0.27 mmol /min/g). The zinc aluminate supported catalyst exhibited equal selectivity towards CO2, CH4 and C5+ as PtCo/Al2O3 and an improved turnover frequency, thus making it a viable replacement support for cobalt under high conversion conditions. A once-through waste-to-fuel process (using biogas from the anaerobic digestion of waste as a raw material) was designed using the experimentally determined selectivity and activity data from the FischerTropsch synthesis. The syngas generation step of this design incorporates a tri-reformer and water-gas shift reactor. Syngas is then fed into the Fischer-Tropsch reactor, which produces largely waxy products at lower conversions (XCO = 60%) and largely naphtha/distillate products at higher conversions (XCO > 80%). The Fischer-Tropsch products were partially refined to distillate (low density diesel) by means of flash tanks and an atmospheric distillation column. At lower conversion levels, a hydrocracker must be used to improve distillate yields. All light hydrocarbons and syngas are fed to a combined cycle power plant, which produced electricity for the plant, thus satisfying the plant's utility self-sufficiency objective. The plant design was evaluated to find an optimal conversion at which to operate, and to gauge the effectiveness of the catalyst design improvements. An optimal conversion of XCO = 80% was found for Mn-Co/Al2O3 (Mn:Co = 0.14) at a production level of 329 bbl/day distillate from a feed of 16 tonnes of municipal solid waste per hour. This represents a 12% increase in production of distillate when compared to Pt-Co/Al2O3 at the same conversion. A shift from an alumina support to a zinc aluminate support will be necessary as the optimal conversion for this process lies above the XCO = 70% deactivation threshold.