Biomass Conversions to Liquid Fuels via BTL Processes

Biomass to Liquids (BTL) processes are being designed based on the thermo-chemical platform for converting biomass to biofuels. In an indirect liquefaction process, synthesis gas (CO+H₂) is first produced via gasification of solid biomass or liquid bio-oil produced by the fast pyrolysis of biomass. Synthesis gas (syn-gas) can be converted to synthetic gasoline, jet-fuel or diesel using Fischer-Tropsch synthesis of hydrocarbons on cobalt-based catalysts or synthetic alcohols, such as ethanol using rhodium-based catalysts. These gas to liquids (GTL) technologies, previously utilizing coal or low-cost, remote natural gas as feedstock for liquid fuels production can be adapted for the conversion of bio-mass to liquids (BTL).

Typically, larger plant capacity is required to make the BTL process more economically competitive. This means that a large amount of biomass feedstock needs to be transported over a long distance, which can cause logistic issues. In order to reduce the cost of biomass transportation, the BTL process can be decoupled into two steps. In the first step, the biomass is converted to a liquid form via fast pyrolysis at distributed facilities close to the source of biomass. The liquefied biomass, commonly called as bio-oil, is then transported to a much larger central facility where it is gasified/reformed into synthesis gas at high pressure. The syn-gas is subsequently cleaned/upgraded and then converted into liquid fuels (see schematic in Figure 1). The advantages of this two step process are lower cost of transporting biomass feedstock and significant reduction of power requirement for producing high-pressure synthesis gas. In addition, compared to its biomass source, bio-oil contains significantly lower ash, sulfur and nitrogen compounds, which are poisonous for the GTL reaction catalysts. The synthesis gas cleaning step at the centralized facility can thus be made significantly less strenuous.

Several ongoing research activities on the BTL processes have provided CSET facilities with equipment and instrumentation necessary for demonstrating the production of liquid fuels from biomass through thermo-chemical approaches. The new bench-scale high pressure GTL reaction system, as shown in Figure 2, is capable of converting syngas to liquid fuels using fixed-bed reaction mode or slurry-type reaction mode. Analysis of the synthesis gas composition is carried out using a Varian 4900 microGC and the liquid products composition using a Varian 3800 GC equipped with a Saturn 2200 MS. CSET’s wet chemistry facilities are also well equipped for conducting research that requires catalyst synthesis.

Figure 2. High-pressure gas-to-liquid (GTL) reactor system at CSET

: (a) Integrated fixed bed and slurry reactor system

: (b) data acquisition system

: (c) fixed bed reactor

Aaaaaaaaaa

aaaaaaaaaaaa

: (d) slurry reactor

Initial testing of the Fischer-Tropsch catalyst was conducted in the 500 cc slurry reactor. 20g of the ground, sieved Co-Zr/Al₂O₃ catalyst (particle size < 90 micron) was loaded along with 250 cc of paraffin. Prior to reaction, the catalyst was reduced at 623 K and 25 bar for 4 h with 50% H₂ in N₂ flowing at 500 sccm. The reaction syn-gas composition was 20% CO, 40% H₂ and 40% N₂ flowing at 1slpm. The reaction was conducted at 25 bar and three different temperatures 498, 523 and 548 K. Figure 3 shows the plot of CO conversion v/s reaction temperature. Figure 4 shows a comparison of theoretical and experimental liquid product distribution from the reaction run carried out at 498 K and 25 bar. Upon successful testing of the F-T reaction runs with the 500 cc reactor, similar reaction conditions were employed for a newly built 2 L F-T slurry reactor. The reaction was conducted at 25 bar and 523 K with the same syn-gas composition (20% CO, 40% H₂ and 40% N₂) at 4 slpm. Figure 5 shows the products collected from the first condenser of the 2 L F-T slurry reactor after this run. We intend to tune the process parameters to maximize yields in the desired hydrocarbon product range as predicted by the Anderson-Schulz-Flory distributions shown in Figure 6.