Degree Type

Thesis

Date of Award

2009

Degree Name

Master of Science

Department

Mechanical Engineering

First Advisor

Robert C. Brown

Abstract

Fast pyrolysis thermochemically degrades lignocellulosic material into solid char, organic liquids, and gaseous products. Using fast pyrolysis to produce renewable liquid bio-oil to replace crude oil is gaining commercial interest. The production of pyrolysis bio-oil needs to be improved through standardization. Only with standard operational methods can pyrolysis bio-oil be commercially refined into chemicals or transportation fuels. Pretreatments such as washing or torrefaction of biomass prior to pyrolysis are not required to produce high liquid yields, but may improve the end products' yield or quality. Improved quality will offset future processing costs that would otherwise be required. Solid biochar and non-condensable gas products are also formed during fast pyrolysis. Water wash, torrefaction, grinding, and drying pretreatments of biomass were studied to determine how each affects the products of fast pyrolysis.

Three modified central composite experimental designs were developed to study a control and two major biomass pretreatments: washing, and torrefaction for effects of grinding, moisture content or torrefaction, and pyrolysis temperature. A fluidized bed fast pyrolyzer was operated at 0.1 kg/hr to complete the three separate experimental design studies, in a total of sixty tests.

The experimental study was used to develop model equations that express how feedstock pretreatments (grinding and moisture content or torrefaction) and pyrolysis temperature affected the products of fast pyrolysis. In each case, a model equation was derived for the three major studies. Model equations were developed for: biochar yield and composition; bio-oil yield, moisture content, and water insoluble content; and non-condensable gas yield and composition. The results showed pyrolysis temperature was the most significant variable in product modeling. The grind size impacted the extent of decomposition during pyrolysis and the biomass moisture content affected mass balances of low moisture biochar and high moisture bio-oil products.

Results showed that water washing reduced the inorganic mineral content of the biomass but did not eliminate it. A 75% ash reduction in the feedstock was realized from the water wash. Pyrolysis product yields were not significantly affected by the pretreatment. Torrefaction caused non-moisture volatile mass loss in the biomass of 3.8% at 180yC to 15.4% at 250yC during the pretreatment step. The mass loss included the moisture (12 wt. % on a dry basis) as well as other volatile compounds contained in the biomass. The mass reduction caused reduced bio-oil yields of 5% during fast pyrolysis. The bio-oil yield reduction during pyrolysis was realized as an increase in biochar yield. Torrefaction reduced the production of water and light compounds collected in the bio-oil during fast pyrolysis because the compounds were removed during the pretreatment.

The average biochar yield from the tests was 17.6 y 1.5% on a wet basis of the biomass fed. The biochar yield decreased with temperature for all biomass types from 30% to 10% for pyrolysis temperatures from 426yC to 544yC. The biochar elemental analysis showed hydrogen and carbon content varied with pyrolysis temperature. Hydrogen content decreased from 5.5% to 3.5% and carbon increased from 60% to 70% when increasing pyrolysis temperature from 426yC to 544yC.

A fractionated bio-oil collection method was used that collected four separate fractions of bio-oil. The first three fractions had higher heating values (HHV) above 20 MJ/kg and water content below 5 wt. %. The fourth fraction had an average HHV of 6 MJ/kg and an average water content of 58 wt. %. The first three fractions were also more viscous, contained higher amounts of water insolubles, and contained more elemental carbon and less oxygen compared to bio-oil collected in the fourth fraction (elemental carbon contents of 61.4%, 53.6%, 60.1%, and 37.9% and oxygen contents of 31.4%, 39.1%, 33.1%, and 56.0% respectively for bio-oil fraction 1, 2, 3, and 4). Hydrogen, nitrogen, and sulfur content were constant between all four fractions with hydrogen content at 6%, nitrogen content below 1%, and sulfur content below 0.1%.

The total bio-oil mass yield from the tests was 58.7 y 1.3% on a wet basis of biomass fed. The first fraction accounted for 18% of the total bio-oil. It collected compounds that condensed at higher temperatures from 400yC to 100yC. The fraction was close to 50 wt. % insoluble in water, had higher energy density, and had the highest viscosity compared to the other fractions. The fraction exhibited solid properties at room temperature. The second fraction collected compounds that condensed at lower temperatures: 100yC to 80yC. The second fraction accounted for 13% of the total collected bio-oil. The second fraction was an energy dense liquid that contained close to 25 wt. % water insolubles and remained a liquid at room temperature. The third fraction collected the majority of aerosol liquids collected by an electrostatic precipitator. The fraction accounted for 53% of the total bio-oil. The third fraction was composed of close to 45 wt. % water insolubles and had a high viscosity. The fourth fraction collected all remaining compounds that condensed above 10yC. The fourth fraction accounted for 16% of the bio-oil collected in each test. The fraction had a low viscosity, less than 1 wt. % water insolubles, and low energy content due to the increased water content.

The non condensable gasses that were formed during fast pyrolysis were measured with a micro gas chromatograph. The non-condensable gas yield averaged 11.9 y 0.7%. It was determined the total non-condensable gas yield increased with pyrolysis temperature from 10% to 16%. The non-condensable gas carbon monoxide concentration increased and carbon dioxide concentration decreased with increasing pyrolysis temperature. No significant differences in non-condensable gas yield or composition were found between the three biomass pretreatment types studied.

DOI

https://doi.org/10.31274/etd-180810-392

Copyright Owner

Randall Dennis Kasparbauer

Language

en

Date Available

2012-04-30

File Format

application/pdf

File Size

307 pages

Share

COinS