Graduation Year

2013

Document Type

Dissertation

Degree

Ph.D.

Degree Granting Department

Chemical Engineering

Major Professor

Babu Joseph

Keywords

Cobalt, Fischer Tropsch, Fixed Bed Reactor, Modeling, Silica

Abstract

This dissertation explores catalyst technology for the production of renewable liquid fuels via thermo-chemical conversion of biomass derived syngas. Fischer-Tropsch synthesis is a process for converting syngas, i.e. a mixture of CO and H2, into energy rich long chain hydrocarbons and oxygenated compounds. This synthesis process involves a number of elementary reactions leading to an array of polymeric products. The economic operation of an FTS process lie in the interplay of both catalyst and reactor design. In relation to catalysis, the nature of chemisorbed species, and the fractional availability of active metal sites determines rate, conversion and yield. Similarly, reactor design decides the operational envelope and determines the economics of an FTS process.

Eggshell cobalt catalysts are used in CO hydrogenation reactions due to their ability to maximize the use of precious cobalt metal. The thickness of the shell can be utilized to control the product yield and distribution. In this study, during catalyst synthesis stage, metal-support interaction has been exploited to control the thickness and hence, the product distribution. The catalysts are prepared using precipitation of cobalt nitrate (dissolved in ethanol) on silica support. The metal deposition rate and the location are controlled through optimized non-polar solvent imbibing, followed by water addition to a Co(NO3)2-ethanol solution and hydrolysis by urea. The eggshell coating thickness (in the absence of restricting solvent) onto silica gel substrate was modeled via theoretical equations and experimentally verified during catalyst preparation through microscopic analysis of catalyst samples. Bulk precursor solution properties such as viscosity and surface tension along with substrate properties such as tortuosity are analyzed and included in the theoretical analysis for tailoring the catalyst eggshell thickness. Polar and non-polar solvent interactions with silica gel are exploited during cobalt precipitation to control the eggshell thickness. The catalyst samples were characterized using hydrogen chemisorption studies. The catalyst was tested in a fixed bed tubular bench scale reactor using research grade synthetic feed gases (H2:CO being 2:1). Products were analyzed in a GC column fitted with flame ionized detector and the results were compared with Anderson-Schulz-Flory distribution. Liquid product analysis validated the approach used for eggshell catalyst design and synthesis.

The impact of solvent and calcination conditions, on the performance of eggshell catalysts was examined. Solvents such as water and alcohol attach to the silanol groups on the silica gel surface and compete with metal salts during ion exchange and adsorption. The solution properties impact metal dispersion and interaction with metal support. The calcination conditions (static versus dynamic, oxidizing versus reducing atmosphere) also have an impact on metal dispersion and support interaction. Ethanol proved to be a better solvent for enhancing the dispersion due to its surface wetting properties. Direct reduction in dynamic hydrogen provided gradual decomposition of the cobalt precursor thus reducing agglomeration. Both the use of water as a solvent and a static air environment during calcination led to lower dispersion. The back reaction of calcination products (especially H2O) and agglomeration due to thermal expansion were competing phenomenon in a static oxidizing environment. Catalyst characterization revealed that the latter effect was pre-dominant.

Catalyst performance testing was first done with pure gases (H2 & CO) in a fixed bed reactor. Additionally, to examine the technological feasibility and economic viability of producing liquid fuels from biomass via the thermo-chemical route, laboratory scale testing was done using syngas produced by gasification of pine chips. The pine chips were gasified in a tubular entrained flow gasifier operated at MSU and supplied in cylinders. The raw biomass syngas was treated using a series of adsorbents to remove tar, water and other impurities. This pre-treated gas was subjected to Fischer-Tropsch Synthesis (FTS) in a bench scale fixed bed reactor using the eggshell cobalt catalyst developed in our laboratory. Hydrogen was added to attain the 2:1 stoichiometric ratio required for the FTS reaction. The product gases were analyzed using an FTIR gas cell while liquid product was analyzed using a GC/MS HP-5 column. The eggshell catalyst produced fuel preferentially in the range of middle distillates. The activity of FTS catalyst under biomass derived syngas was lower when compared to that under pure surrogates (H2/CO) due to the presence of inert components (such as methane) in the biomass derived syngas

To complement the experimental study, a comprehensive model of FTS catalytic process was developed. This included both catalyst and a fixed bed reactor model. While modeling a catalyst pellet, intra-particle diffusion limitation was taken into account. For a spherical 2mm pellet, eggshell morphology provided highest activity and selectivity. The reactor model was developed by coupling intra-pellet model with inter-pellet model via reaction term. The entire process operation starting with gas injection was considered. Presence of radial temperature profile, due to wall cooling, was confirmed by Mears criterion. Thus for a fixed time duration, a 2-dimensional reactor model, with respect to temperature and concentration, was developed. The safe operational envelopes for a fixed bed reactor, using cobalt catalyst, was narrow 473 < T < 493. The extent of catalyst pore fill changed (i) the radial thermal conductivity (ii) the overall temperature and concentration profile across the bed and (iii) the limits of safe operation without reaction runaway. Finally, hydrocarbon product selectivity also varied during startup. While the catalyst pores were being filled, effluent product mainly composed of lighter, more volatile components. Once the pores are filled, heavier products started to trickle down the bed.

The economics of a large scale production of liquid fuels using this technology was explored using a CHEMCAD model of a large scale process for producing liquid fuel from biomass, a sensitivity study was conducted to determine key process parameters Two different gasification technologies were compared, one that uses only biomass (BTL process) and a second process that supplements the biomass feed with natural gas for meeting energy and hydrogen needs (BGTL process). The basis for the design was 2000 metric tons of dry biomass feed per hour. The breakeven price for synthetic crude oil was estimated at $106/bbl. for the BTL plant, and $88/bbl. for a natural gas assisted BGTL plant using current market prices for raw materials utilities and capital equipment. With the increasing availability, and falling prices of natural gas, the reforming of natural gas will provide a bridge solution in the short term for economical natural gas assisted BTL conversion, thus making it competitive in marketplace.

Download

Share

COinS