Upload
ngodiep
View
217
Download
2
Embed Size (px)
Citation preview
PRODUCING FUEL AND SPECIALTY CHEMICALS FROM THE SLOW PYROLYSIS OF
POULTRY DAF SKIMMINGS
by
JARROD GAYDEN SMITH
(Under the direction of K.C. Das)
ABSTRACT
The production of bio-oil via the slow pyrolysis of dissolved air flotation (DAF) skimmings from
poultry processing is described. The raw DAF skimmings were characterized for
physicochemical properties and for thermal behavior (TGA). The bio-oil was produced in a
batch pyrolysis system at varying temperatures between 400 and 700 oC to study the effect of
temperature on product yield. The fatty acids in the bio-oil produced displayed a high degree of
saturation that caused the bio-oil to have poor cold flow properties (high cloud point and
viscosity) so a solvent extraction scheme was devised to extract a bio-oil fraction rich in
unsaturated fatty acids that could be further esterified into a biodiesel and fatty nitriles that could
be further processed into surfactants. This ethyl acetate-soluble fraction demonstrated much
improved cold flow properties as well as lower water content and a higher HHV. The
esterification of this soluble fraction was performed using methanol and sulfuric acid as an acid
catalyst and the formation of fatty acid methyl esters was verified using GC/MS and FT-IR.
INDEX WORDS: Pyrolysis, Triglycerides, Biodiesel, DAF
PRODUCING FUEL AND SPECIALTY CHEMICALS FROM THE SLOW PYROLYSIS OF
POULTRY DAF SKIMMINGS
by
JARROD GAYDEN SMITH
B.S., Mississippi State University, 2003
A Thesis Submitted to the Graduate Faculty of the University of Georgia in Partial Fulfillment of
the Requirements for the Degree
MASTERS OF SCIENCE
ATHENS, GA
2008
PRODUCING FUEL AND SPECIALTY CHEMICALS FROM THE SLOW PYROLYSIS OF
POULTRY DAF SKIMMINGS
by
JARROD GAYDEN SMITH
Major Professor: K.C. Das
Committee: James Kastner Thomas Adams Manuel Garcia-Perez
Electronic Version Approved: Maureen Grasso Dean of the Graduate School The University of Georgia May 2008
DEDICATION
This is dedicated to my lovely wife, Sarah, and to my son, Atticus. You make it all worth
it and I love you very much.
iv
ACKNOWLEDGEMENTS
I’d like to thank my major professor, Dr. K.C. Das, for his support and guidance on this project.
I’d like to thank my committee members: Dr. James Kastner, Dr. Thomas Adams, and Dr.
Manuel Garcia-Perez for their advice on the project and suggestions for my research. Special
thanks to all of those who aided me during my research: Roger Hilton, Brian Bibbens, Kate Lee,
and the University of Georgia Forage and Feed Lab. Special thanks also to Walt Moore, and
David Garrett of Pilgrim’s Pride in Athens, GA for taking time out of their busy workdays to
help me with the project. Lastly, thanks to all my family and friends for their support and
encouragement during this process.
v
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS……………...………………………………………………………..v
LIST OF TABLES…………………………………………………………………………….…vii
LIST OF FIGURES…………………………………..…………………………………………viii
CHAPTER
1 INTRODUCTION………………………………………………………………...1
2 LITERATURE REVIEW…………………………………………………………3
3 PRODUCING FUEL AND SPECIALTY CHEMICALS
FROM THE SLOW PYROLYSIS OF POULTRY DAF SKIMMINGS………..10
4 CONCLUSION…………………………………………………………..………45
vi
LIST OF TABLES
Page
TABLE 1 - Physico-chemical composition of DAF……………………………………………..21
TABLE 2 - Char characterization………………………………………………………………..26
TABLE 3 - Molar composition of gas generated (mol %)………………………………………29
TABLE 4 - Raw bio-oil composition determined via GC/MS………………………………......30
TABLE 5 - Characterization of raw bio-oils…………………………………………………….31
TABLE 6 - Characterization of soluble fractions………………………………………….…….32
TABLE 7 - Cloud point of raw oil and soluble fractions………………………………………..34
TABLE 8 - Yields of soluble fractions…………………………………………………………..35
TABLE 9 - Soluble and insoluble compounds as determined by GC/MS………………………36
TABLE 10 - Comparison of ethyl acetate soluble fraction and raw bio-oil…………………..…37
vii
viii
LIST OF FIGURES
Page
FIGURE 1 - TG and DTG curves of poultry DAF…………………………...………………….22
FIGURE 2 - Temperature profile at 600 oC……………………………………………………...24
FIGURE 3 - Mass rate of gas evolved…………………………………………………………...25
FIGURE 4 - Yields of products………………………………………………………………….25
FIGURE 5 - Chromatogram of Raw Bio-oil………………………………………………….….30
FIGURE 6 - Determination of oxidative onset temperature……………………………………..33
FIGURE 7 - Cloud point determination………………………………………………………….34
FIGURE 8 - Chromatogram of Ethyl Acetate-insoluble fraction………………………………..35
FIGURE 9 - Chromatogram of Ethyl Acetate-soluble fraction………………………………….36
FIGURE 10 - Gas chromatogram of esterified ethyl acetate soluble fraction…………………...37
FIGURE 11 - FT-IR of ethyl acetate-soluble fraction…………………………………….……..38
FIGURE 12 - FT-IR of biodiesel produced from ethyl acetate-soluble fraction………………...38
CHAPTER 1
INTRODUCTION
In the effort to identify feedstocks for the purposes of creating a new biomass economy, it
may be beneficial to look in the area of waste streams from industrial and commercial sources.
This serves a dual purpose. Processing these waste streams could not only produce bio-products
from a source that may not be fully utilized but it may also lessen the strain on other non-
renewable energy sources and resources.
In the poultry processing industry, the skimmings produced from clarifying wastewater
by Dissolved Air Flotation (DAF) are a byproduct of the process that is high in triglycerides and
proteins. DAF skimming are often not beneficially utilized and their disposal often poses unique
challenges. Currently, methods such as land application and landfilling are means of disposal
but all come with significant costs, material handling, environmental, and/or quality of life issues
attached. Additionally, selling to a rendering company is another disposal option that can
occasionally produce an income but this is tied closely to current fat market conditions. High
freight costs due to the high water content of DAF skimmings can often offset any income from
rendering when market conditions are poor. A more favorable alternative would be a method to
convert this waste material into a value-added product.
Currently transesterification into biodiesel is a popular avenue for the conversion of
waste triglycerides but triglyceride feedstocks with a high level of contamination like DAF
skimmings present difficulties as feedstocks for transesterification. An alternate approach would
be to use thermo-chemical conversion to thermally break the bonds that form the triglycerides to
produce fatty acids.
1
Thermo-chemical conversion has long been used as a means to convert waste biomaterial
into usable specialty chemicals and fuels. Thermal (pyrolysis) cracking is a commonly utilized
thermo-chemical conversion technique. Pyrolysis involves subjecting the feedstock to high
thermal stress (typically at 450oC and above) in the absence of oxygen to break down molecules
into smaller components.
The research presented within is focused on the slow pyrolysis of DAF skimmings to
produce bio-oil rich in fatty acids that could then be esterified to produce a biodiesel for fuel
applications and fatty nitriles that can be processed into surfactants.
2
CHAPTER 2
LITERATURE REVIEW
The current state of global energy consumption has raised concerns about not only the
environmental impact of fossil fuel use but also about our growing dependence on a limited
source of non-renewable energy. In the US, the dependence on foreign fossil fuels is of further
concern due to our large consumption. The search for renewable energy sources is an important
step in achieving our energy goals. One avenue for renewable energy is the conversion of
biomass into a fuel.
In the effort to identify feedstocks for the purposes of creating a new biomass economy, it
may be beneficial to look in the area of waste streams from industrial and commercial sources.
This serves a dual purpose. Processing these waste streams could not only produce bio-products
from a source that may not be fully utilized but it may also lessen the strain on other non-
renewable energy resources. A number of other research groups have focused on the idea of
utilizing waste streams, both municipal and industrial, to produce a fuel oil or specialty
chemicals (Karayildirim et al., 2006; Mahadevaswamy and Venkataraman, 1986; Gomez-Rico et
al., 2003; Shinogi and Kanri, 2003).
One possible source of waste streams is from the meat processing industry, particularly
the sludge recovered from wastewater by dissolved air flotation (DAF). The DAF sludge is high
in triglyceride (fat) and protein content. It typically contains around 60-70 mass% of moisture,
around 15 mass% fat (triglycerides) and 15 mass% protein (Cai et al., 1994). The high water
3
content of this waste sludge presents difficulties for its conversion in processing facilities away
from the poultry processing plants.
Render, a trade magazine associated with the National Renderers Association, states in
their 2006 Market Report that the US produced 4.3 million metric tons of fat and grease from the
animal slaughter industry (Swisher, 2007). Much of this waste is landfilled or land applied as
fertilizers (Bailey, 1989) or shipped off to rendering companies for animal feed (Swisher, 2007).
All of these disposal methods present their own problems such as transportation and materials
handling costs as well as environmental and quality-of-life issues. The conversion of this
material into a useable fuel source would help eliminate any disposal problems and would also
serve as a value-added product derived from a waste stream.
Possible conversion processes for these high lipid and protein content waste streams are
summarized in a number of publications (Bridgwater and Peacocke, 2000; McKendry, 2002;
Maher and Bressler, 2007). These methods include both chemical conversion techniques like
transesterification (Lang et al., 2001) and thermochemical conversion techniques such as
gasification, liquefaction, pyrolysis and hydrocracking (Maher and Bressler, 2007).
The production of bio-diesel from vegetable oils and animal fats has received much
attention from the renewable fuel community. Due to the high cost of virgin vegetable oil a
number of researchers are turning to waste vegetable oils like used restaurant cooking oil to
produce biodiesel (Kulkarni and Dalai, 2006; Canakci, 2007; Issariyakul et al., 2007). The
production of biodiesel via the transesterification of waste streams rich in oils and grease has
been documented in past studies by researchers at NREL (Kinast, 2003). Often the
transesterification of these waste materials can be difficult due to the amount of undesirable
4
contaminants in the waste stream, the presence of large amount of water and proteins interfere
with the trans-esterification reactions (Kulkarni and Dalai, 2006; Adebanjo et al., 2007).
Feedstocks with a high saturated fatty acid content typically yield bio-diesels with high clouding
points.
A possible alternative to the transesterification of very dirty high lipid waste streams is
thermal cracking (pyrolysis) or hydrocracking (Rana et al., 2007) which can be performed close
to the rendering facilities.
Most research in the field of pyrolysis has centered on the pyrolysis of lignocellulosic
material but a few researchers have focused on feedstocks that are high in lipids and proteins.
Some of these high lipid and high protein feedstocks that have been the focus of prior research
are used cooking oil (Lima et al., 2004; Adebanjo et al., 2007), animal fats (Adebanjo et al.,
2005), poultry droppings (Mahadevaswamy and Venkataraman, 1986), and algae (Goldman et
al., 1981; Peng et al., 2001; Miao et al., 2004).
Maher and Bressler (2007) compiled a literature review on the available research on the
pyrolysis of triglycerides. The group concluded that the research in the pyrolysis of triglycerides
was lacking when compared to that of lignocellulosic material. Areas that they identified as
needing improvement were “optimization of the reaction conditions to obtain specific reaction
products, understanding the chemistry behind the pyrolysis reaction, and comprehensive
evaluation of the final product properties”.
The oils produced from high lipid and high protein feedstocks have compared favorably
with oils produced from lignocellulosic biomass. Miao et al. (2004a; 2004b) focused on
microalgae as a high lipid and protein feedstock for pyrolysis. The research characterized
5
different microalgae for biomolecular composition, especially lipid and protein composition.
They compared the resulting bio-oil from the pyrolysis of microalgae to literature values for
pyrolysis oil from lignocellulosic material. The microalgae, which was comprised of 65% lipids
and proteins, produced a bio-oil with much lower oxygen content and a much higher heating
value than that produced from lignocellulosic material. In addition, a group of researchers in
Canada pyrolyzed animal fat and found the heating value of the pyrolysis oil to be in the range of
38.5 – 40 MJ/kg as compared to 45 MJ/kg for #2 diesel fuel which is the standard according to
ASTM D975 (Adebanjo et al., 2005).
The focus of this thesis research was on the conversion of poultry processing DAF
skimmings into a crude bio-oil using pyrolysis followed by upgrading the bio-oil. This crude oil
could be further transformed in a final step to produce biodiesel. The approach studied in this
research was to separate the saturated fatty acids from the un-saturated fatty acids using solvent
extraction techniques to produce a bio-oil with low clouding point that could be esterified to
produce a bio-diesel. The saturated fraction with high clouding point could find other
applications in the rendering plants, for example as a source of process fuel. The skimmings
were pyrolyzed at multiple temperatures to evaluate the impact of operational temperature on
product yields and composition. This paper will describe the pyrolysis and fractionation of
resulting bio-oil using solvent extraction and the esterification of the upgraded bio-oil to produce
biodiesel.
6
LITERATURE CITED
Adebanjo, A., M. G. Kulkarni, A. K. Dalai, and N. N. Bakhshi (2007). "Pyrolysis of waste fryer
grease in a fixed-bed reactor." Energy & Fuels 21(2): 828-835.
Adebanjo, A. O., A. K. Dalai, and N. N. Bakhshi (2005). "Production of diesel-like fuel and
other value-added chemicals from pyrolysis of animal fat." Energy & Fuels 19(4): 1735-
1741.
Bailey, K. M. (1989). "Solvent Extraction of Oil and Grease from Poultry Processing
Wastewater." University of Arkansas Masters Thesis.
Bridgwater, A. V. and G. V. G. Peacocke (2000). "Fast pyrolysis processes for biomass."
Renewable and Sustainable Energy Reviews 4: 1-73.
Cai, T., O. C. Pancorbo, and H. M. Barnhart (1994). "Chemical and Microbiological
Characteristics of Poultry Processing By-Products, Waste, and Poultry Carcasses During
Lactic Acid Fermentation." J APPL POULT RES 3(1): 49-60.
Canakci, M. (2007). "The potential of restaurant waste lipids as biodiesel feedstocks."
Bioresource Technology 98(1): 183-190.
Goldman, Y., N. Garti, B. Ginzburg, and M. R. Bloch (1981). "Conversion of halophilic algae
into extractable oils. 2. Pyrolysis of proteins." Fuel 60: 90-92.
7
Gomez-Rico, M. F., I. Martin-Gullon, A. Fullana, J. A. Conesa, and R. Font (2003). "Pyrolysis
and combustion kinetics and emissions of waste lube oils." Journal of Analytical and
Applied Pyrolysis 68-69: 527-546.
Issariyakul, T., M. G. Kulkarni, A. K. Dalai, and N. N. Bakhshi (2007). "Production of biodiesel
from waste fryer grease using mixed methanol/ethanol system." Fuel Processing
Technology 88(5): 429-436.
Karayildirim, T., J. Yanik, M. Yuksel, and H. Bockhorn (2006). "Characterisation of products
from pyrolysis of waste sludges." Fuel In Press, Corrected Proof.
Kinast, J. A. (2003). "Production of Biodiesels from Multiple Feedstocks and Properties of
Biodiesels and Biodiesel/Diesel Blends: Final Report; Report 1 in a Series of 6." NREL
Report.
Kulkarni, M. G. and A. K. Dalai (2006). "Waste cooking oil-an economical source for biodiesel:
A review." Industrial & Engineering Chemistry Research 45(9): 2901-2913.
Lang, X., A. K. Dalai, N. N. Bakhshi, M. J. Reaney, and P.B. Hertz (2001). "Preparation and
characterization of bio-diesels from various bio-oils." Bioresource Technology 80(1): 53-
62.
Lima, D. G., V. C. D. Soares, E. B. Ribeiro, D. A. Carvalho, E. C. V. Cardoso, F. C. Rassi, K. C.
Mundim, J. C. Rubim, and P. A. Z. Suarez (2004). "Diesel-like fuel obtained by pyrolysis
of vegetable oils." Journal of Analytical and Applied Pyrolysis 71(2): 987-996.
Mahadevaswamy, M. and Venkataraman (1986). "Bioconversion of poultry droppings for biogas
and algal production." Agricultural Wastes 18: 93-101.
8
Maher, K. D. and D. C. Bressler (2007). "Pyrolysis of triglyceride materials for the production of
renewable fuels and chemicals." Bioresource Technology 98(12): 2351-2368.
McKendry, P. (2002). "Energy production from biomass (part 2): conversion technologies."
Bioresource Technology 83(1): 47-54.
Miao, X. and Q. Wu (2004). "High yield bio-oil production from fast pyrolysis by metabolic
controlling of Chlorella protothecoides." Journal of Biotechnology 110 85-93.
Miao, X., Q. Wu, and C. Yang (2004). "Fast pyrolysis of microalgae to produce renewable
fuels." Journal of Analytical and Applied Pyrolysis 71(2): 855-863.
Peng, W., Q. Wu, and P. G. Tu (2001). "Pyrolytic characteristics of heterotrophic Chlorella
protothecoides for renewable bio-fuel production." Journal of Applied Phycology 13: 5-
12.
Rana, M.S., V. Samano, J. Ancheyta, and J. A. I. Diaz. (2006), "A review of recent advances on
process technologies for upgrading of heavy oils and redidua." Fuel 86(9): 1216-1231
Shinogi, Y. and Y. Kanri (2003). "Pyrolysis of plant, animal and human waste: physical and
chemical characterization of the pyrolytic products." Bioresource Technology 90: 241-
247.
Swisher, K. (2007). "Market Report 2006." Render Magazine.
9
CHAPTER 3
PRODUCING FUEL AND SPECIALTY CHEMICALS FROM THE SLOW PYROLYSIS OF
POULTRY DAF SKIMMINGS1
1 J. Smith, M. Garcia‐Perez, and K.C. Das. 2008. To be submitted to the Journal of
Analytical and Applied Pyrolysis
10
Abstract
The production of bio-oil via the slow pyrolysis of dissolved air flotation (DAF) skimmings from
poultry processing is described. The raw DAF skimmings were characterized for
physicochemical properties and for thermal behavior (TGA). The bio-oil was produced in a
batch pyrolysis system at varying temperatures between 400 and 700 oC to study the effect of
temperature on product yield. The fatty acids in the bio-oil produced displayed a high degree of
saturation that caused the bio-oil to have poor cold flow properties (high cloud point and
viscosity) so a solvent extraction scheme was devised to extract a bio-oil fraction rich in
unsaturated fatty acids that could be further esterified into a biodiesel and fatty nitriles that could
be further processed into surfactants. This ethyl acetate-soluble fraction demonstrated much
improved cold flow properties as well as lower water content and a higher HHV. The
esterification of this soluble fraction was performed using methanol and sulfuric acid as an acid
catalyst and the formation of fatty acid methyl esters was verified using GC/MS and FT-IR.
Keywords: pyrolysis, triglycerides, biodiesel, DAF
I. Introduction
The current state of global energy consumption has raised concerns about not only the
environmental impact of fossil fuel use but also about our growing dependence on a limited
source of non-renewable energy. In the US, the dependence on foreign fossil fuels is of further
concern due to our large consumption and its impact on trade deficits and national security. The
search for renewable energy sources is an important step in achieving our energy goals. One
avenue for renewable energy is the conversion of biomass into a fuel.
11
In the effort to identify feedstocks for the purposes of creating a new biomass economy, it
may be beneficial to look in the area of waste streams from industrial and commercial sources.
This serves a dual purpose. Processing these waste streams could not only produce bio-products
from a source that may not be fully utilized but it may also lessen the strain on other non-
renewable energy resources. A number of other research groups have focused on the idea of
utilizing waste streams, both municipal and industrial, to produce a fuel oil or specialty
chemicals (Karayildirim et al., 2006; Mahadevaswamy and Venkataraman, 1986; Gomez-Rico et
al., 2003; Shinogi and Kanri, 2003).
One possible source of waste streams is from the meat processing industry, particularly
the sludge recovered from wastewater by dissolved air flotation (DAF). The DAF sludge is high
in triglyceride (fat) and protein content. It typically contains around 60-70 mass% of moisture,
around 15 mass% fat (triglycerides) and 15 mass% protein (Cai et al., 1994). The high water
content of this waste sludge presents difficulties for its conversion in processing facilities away
from the poultry processing plants.
Render, a trade magazine associated with the National Renderers Association, states in
their 2006 Market Report that the US produced 4.3 million metric tons of fat and grease from the
animal slaughter industry (Swisher, 2007). Much of this waste is landfilled or land applied as
fertilizers (Bailey, 1989) or shipped off to rendering companies for animal feed (Swisher, 2007).
All of these disposal methods present their own problems such as transportation and materials
handling costs as well as environmental and quality-of-life issues. The conversion of this
material into a useable fuel source would help eliminate any disposal problems and would also
serve as a value-added product derived from a waste stream.
12
Possible conversion processes for these high lipid and protein content waste streams are
summarized in a number of publications (Bridgwater and Peacocke, 2000; McKendry, 2002;
Maher and Bressler, 2007). These methods include both chemical conversion techniques like
transesterification (Lang et al., 2001) and thermochemical conversion techniques such as
gasification, liquefaction, pyrolysis and hydrocracking (Maher and Bressler, 2007).
The production of bio-diesel from vegetable oils and animal fats has been getting much
attention from the renewable fuel community. Due to the high cost of virgin vegetable oil a
number of researchers are turning to waste vegetable oils like used restaurant cooking oil to
produce biodiesel (Kulkarni and Dalai, 2006; Canakci, 2007; Issariyakul et al., 2007). The
production of biodiesel via the transesterification of waste streams rich in oils and grease has
been documented in past studies by researchers at NREL (Kinast, 2003). Often the
transesterification of these waste materials can be difficult due to the amount of undesirable
contaminants in the waste stream, the presence of large amount of water and proteins interfere
with the trans-esterification reactions (Kulkarni and Dalai, 2006; Adebanjo et al., 2007).
Feedstocks with a high saturated fatty acid content typically yield bio-diesels with high clouding
points.
A possible alternative to the transesterification of very dirty high lipid waste streams is
thermal cracking (pyrolysis) or hydrocracking which can be performed close to the rendering
facilities.
Most research in the field of pyrolysis has centered on the pyrolysis of lignocellulosic
material but a few researchers have focused on feedstocks that are high in lipids and proteins.
Some of these high lipid and high protein feedstocks that have been the focus of prior research
are used cooking oil (Lima et al., 2004; Adebanjo et al., 2007), animal fats (Adebanjo et al.,
13
2005), poultry droppings (Mahadevaswamy and Venkataraman, 1986), and algae (Goldman et
al., 1981; Peng et al., 2001; Miao et al., 2004).
Maher and Bressler (2007) compiled a literature review on the available research on the
pyrolysis of triglycerides. The group concluded that the research in the pyrolysis of triglycerides
was lacking when compared to that of lignocellulosic material. Areas that they identified as
needing improvement were “optimization of the reaction conditions to obtain specific reaction
products, understanding the chemistry behind the pyrolysis reaction, and comprehensive
evaluation of the final product properties”.
The oils produced from high lipid and high protein feedstocks have compared favorably
with oils produced from lignocellulosic biomass. Miao et al. (2004a; 2004b) focused on
microalgae as a high lipid and protein feedstock for pyrolysis. The research characterized
different microalgae for biomolecular composition, especially lipid and protein composition.
They compared the resulting bio-oil from the pyrolysis of microalgae to literature values for
pyrolysis oil from lignocellulosic material. The microalgae, which was comprised of 65% lipids
and proteins, produced a bio-oil with much lower oxygen content and a much higher heating
value than that produced from lignocellulosic material. In addition, a group of researchers in
Canada pyrolyzed animal fat and found the heating value of the pyrolysis oil to be in the range of
38.5 – 40 MJ/kg as compared to 45 MJ/kg for #2 diesel fuel which is the standard according to
ASTM D975 (Adebanjo et al., 2005).
The focus of this thesis research was on the characterization of the physicochemical
properties of poultry processing DAF skimmings and their thermal behavior (TGA) and the
conversion of this material into a crude bio-oil using pyrolysis followed by upgrading the bio-oil.
This crude oil could be further transformed in a second step in the same unit to produce
14
transportation fuels and chemicals. The approach studied in this research was to separate the
saturated fatty acids from the unsaturated fatty acids using solvent extraction techniques to
produce a bio-oil with low clouding point containing unsaturated fatty acids that could be
esterified to produce a bio-diesel and fatty nitriles that could be processed into surfactants. The
saturated fraction with high clouding point could find other applications in the rendering plants,
for example as a source of process fuel. The skimmings were pyrolyzed at multiple temperatures
to evaluate the impact of operational temperature on product yields and composition. This paper
will describe the pyrolysis and fractionation of resulting bio-oil using solvent extraction and the
esterification of the upgraded bio-oil to produce biodiesel along with the verification of the
presence of fatty acid methyl ester via GC/MS and FTIR.
2. Experimental Plan
2.1. - DAF characterization
The DAF skimmings for this project were obtained from the a poultry processing facility
in Athens, GA. The DAF skimmings were treated during processing with sulfuric acid as a pH
adjustment, two ionic polymers (an anionic copolymer of acrylamide and acrylic acid and a
cationic copolymer of acrylamide) and ferric chloride for the purposes of flocculation.
The DAF skimmings were dried at 103oC for 24 hour in a VWR Scientific Products 1370 FM
oven before being pyrolyzed. A portion of this dried DAF was saved and sent to the Forage and
Feed Lab for characterization and the rest was used for the pyrolysis experiment. The elemental
composition (CNS), proximate analysis (crude fat, crude proteins, crude fiber, moisture and ash),
and HHV of all DAF samples were determined at the UGA Forage and Feed Lab.
The CNS was determined in a LECO CNS-2000. Samples are combusted in an oxygen
atmosphere at 1350oC, converting elemental carbon, sulfur, and nitrogen into CO2, SO2, NOx,
15
and N2. These gases are then passed through the IR (infrared) cells to determine the carbon and
sulfur content and a TC (thermal conductivity) cell to determine N2. The fat content was
determined in a Tecator-Soxtec Fat Extractor. A gram of sample was extracted at 105ºC for
approximately 40 minutes with boiling ethyl ether. The amount of fiber was determined using an
Ankom fiber analyzer that determines crude fiber which is the organic residue remaining after
digesting with 0.255N H2SO4 and 0.313N NaOH. The compounds removed are predominantly
protein, sugar, starch, lipids and portions of both the structural carbohydrates and lignin.
The lab analyzed % moisture by drying 2 g of material at 135ºC for 2 h, and weighed
according to the Association of Official Analytical Chemist (AOAC) method 930.15. The % ash
was determined by igniting 2 g of material at 600 ºC for 2 h in a pre-weighed porcelain crucible
and weighing the residue according to AOAC method 942.05. The % protein was determined
using approximately 0.2 g of material in an Elementar™ Rapid N Combustion Nitrogen Analyzer
according to AOAC method 990.03. The gross energy of the biomass was determined in a Parr
1241 Adiabatic Calorimeter.
2.2.- Thermal behavior of DAF
The thermal decomposition behavior of studied DAF samples was determined by
thermogravimetric (TG) in a Mettler Toledo TGA/SDTA851e in accordance to ASTM E 1641.
TGA was performed on the DAF skimmings samples at a heating rate of 8oC/min that
approximates the heating rated of the furnace used for the pyrolysis experiment. Nitrogen at 50
mL/min was the carrier gas used.
2.3. - Pyrolysis tests
The pyrolysis process was performed in a 5.87 liter batch pyrolysis reactor. The empty
reactor vessel was weighed before the biomass was added so that both the weight of the biomass
16
used and the weight of char produced can be determined by difference. The furnace was
outfitted with two ports so that inlet and outlet lines could be connected to the reactor body for
carrier gas and exhaust. A thermocouple was inserted into the reactor vessel so that the internal
temperature of the biomass could be measured during pyrolysis.
The pyrolysis process was conducted at four different final pyrolysis temperatures
ranging from 400-700oC. The vapors were condensed into a set of four condensers arranged
with the first two in parallel and the subsequent two in series. The condensers were weighed
before and after pyrolysis to determine the weight of bio-oil generated by difference.
The non-condensable gases that leave the condensers were vented to the atmosphere. On
one run for each operational temperature, the flow-rate of the non-condensable gases was
measured using a Manostat volumetric flowmeter and a sample of the gas was taken every 50oC.
The bio-oil produced during the pyrolysis process was collected from the condensing
traps. The traps were weighed before and after to determine the weight of the oil produced. The
yield of the oil was then calculated using the weight of the oil produced and the starting weight
of the biomass.
2.4.- Analysis of products
2.4.1. - Gases
Gas chromatography was performed on the collected non-condensable gas samples in an
Agilent 3000A Micro GC. The GC was operated with four different columns each with a
different operating temperature and runtime. Column A was a molecular sieve column operated
at 90oC with a 120 second runtime. Column B was a PLOT-U type column operated at 90oC
with a 150 second runtime. Column C was an OV-1 type column operated at 65oC with a
runtime of 480 seconds. Column D was an alumina column operated at 130oC with a runtime of
17
480 seconds. Helium was used as a carrier gas for columns B, C, and D and argon was used for
column A. The mass rate of non-condensable gasses generated during the pyrolysis process was
determined using the Mnostat volumetric flowmeter and molar composition determined by GC
and the ideal gas law.
2.4.2. - Char
Proximate analyses and CHNS composition was performed on the resulting char. The
CHNS composition was determined in a LECO CHNS932 that was calibrated using sulfa
methazine as a standard. The CHNS composition was determined in accordance to ASTM
D5291. The proximate analysis was performed on a LECO TGA731 according to ASTM
D3176.
2.4.3. - Bio-oils
All of the bio-oils were characterized in the lab for higher heating value, CHNS
composition, viscosity, and water content. For the oil, the higher heating value was determined in
a Parr 1351 bomb calorimeter. The CHNS composition was determined using the same LECO
CHNS932 mentioned previously. The viscosity was measured on a Brookfield DV-I+
Viscometer with an ultra low (UL) viscosity adapter with a 16 mL sample volume capacity and a
304 s/s spindle. The water content was determined in a Mettler Toledo DL31 Karl Fisher
Titrator according to ASTM D-1744. The composition of the bio-oil was characterized using
GC/MS. The GC/MS used was a Hewlett Packard 5890 GC/5971A MS equipped with an EC-5
column (Econo-Cap). The column is 30 m in length, with a 0.25 mm ID, and a 0.25 um film.
The composition of the film is a mixture of siloxanes to create a specific polarity. EC-5 is
intermediate in polarity. During the method used for the raw samples, the program started at
18
50oC for 3 minutes, then 15oC/min to 290oC, and then was held at 290oC for 10 minutes. The
injector and detector were set to 280oC. Helium was used as the carrier gas.
2.5. – Solvent Extraction
The crude bio-oil produced from the pyrolysis of poultry DAF skimmings was in a waxy
(solid) state at room temperature. In order to decrease the viscosity a method was developed to
separate the fraction with the high clouding point (usually the saturated fatty acids) from the
fraction with the low clouding point (usually the unsaturated fatty acids). Several solvent
extraction schemes were attempted to upgrade the bio-oil. The solvent extraction study was used
to determine if this was a viable pathway to separate compounds with high clouding point from
compounds with low clouding points.
All the extraction experiments were carried out as follow: A solvent to bio-oil ratio of
5mL of solvent for every 1g of bio-oil was used in the extraction scheme. Once the solvent was
added to the bio-oil the solution was mixed using a vortex mixer. This ratio corresponded to
those cited in literature on the solvent extraction of fats (Bailey, 1989). The insolubles were
further separated out using vacuum filtration using a Whatman 114V 25 mm filter.
A Buchi Rotovapor R-200 rotovaporator was used to evaporate the solvent out of the solution.
The weight of the recovered soluble fraction was determined by taring the rotovaporator bulb
and then weighing the remaining soluble fraction in the bulb after the solvent was evaporated
out. This was determined to be when the weight loss of the contents of the bulb was less than
1% for a 5-minute interval of rotovaporating. The soluble fraction was recovered and stored in a
refrigerated container for future characterization.
The eight solvents chosen for these studies were hexane, dichloromethane, ethyl ether,
acetone, toluene, ethanol, propanol, and butanol. These solvents were chosen based on the wide
19
range of dielectric constants and their application in industry (Mellan, 1957). Of the eight
solvents tested four did not produce a soluble liquid fraction. These solvents were toluene and
the three alcohols; ethanol, butanol, and propanol. No further testing was done with these four
solvents.
The fractions extracted using four solvents were ranked based on these three criteria
(yield, oxidation onset temperature and water content). From those four solvents acetone was
chosen based on performance. Ethyl acetate and methyl ethyl ketone were also chosen along
with acetone for a second screening of solvents. These two additional solvents were chosen due
to their chemical similarities to acetone. Additional extractions were performed using these three
solvents on bio-oils from the entire range of operating temperatures. These final extractions
were characterized using yield and cloud point as determined by DSC. Additionally, the ethyl
acetate-soluble fraction was characterized by viscosity, HHV, and water content.
The oxidation onset temperature (OOT) was determined as outlined by ASTM E 2009-
02. OOT is a measure of a material’s oxidation stability. The DSC curves were attained using a
Mettler Toledo DSC832 with a heating rated of 10.0oC/min to 350oC and a gas flow rate of
50mL/min of O2. The OOT was determined graphically from the DSC curve of the extracted
fraction.
The cloud point was determined from DSC curves obtained at a heating rate of 10oC/min
and a nitrogen carrier gas flow rate of 50 mL/min in accordance to ASTM D4419. The test was
conducted by increasing temperature from -40oC to 200oC and then back to -40oC.
2.6. – Esterification of raw bio-oil
A sample of an ethyl acetate-soluble fraction of bio-oil was esterified to convert the free
fatty acids using a methanol to bio-oil ratio of 60:1 and an addition of 5wt% of sulfuric acid as
20
catalyst. The mixture was stirred at 60oC for 120 minutes. This procedure was outlined in
(Berrios et al., 2007). Once the esterification process was completed the end product was
analyzed by GC/MS. Additionally, FT-IR was performed to determine the presence of fatty acid
methyl esters (FAMEs) on a Varian 2000 FT-IR using 16 scans in the mid-IR spectral range
(Oliveira et al., 2006).
3. Results
3.1. - DAF Characterization
The results of the characterization of the raw biomass as performed by the Forage and
Feed Lab at the University of Georgia are listed in Table 1. As can be seen from the results of the
characterization, the DAF contained a very high percentage of fat and protein and a high carbon
and nitrogen content.
Table 1: Physicochemical composition of DAF
Carbon (%) 64.61
Hydrogen (%) 8.61
Nitrogen (%) 3.65
Sulfur (%) 0.43
Oxygen* (%) 22.70
Fat (%) 55.91
Protein (%) 25.43
Crude Fiber (%) 1.48
Ash (%) 1.93
Moisture (%) 4.52
HHV (MJ/kg) 32.7
* By difference
21
The fat and protein composition of poultry DAF can differ greatly depending on several
factors but the values generally fall into the ranges of 40-70% fat and 15-40% for protein content
on a solid material basis (Lee, 2003). The fat and protein content for the DAF used for this
experimentation fell within the expected ranges.
3.2. Thermal behavior of DAF
The thermogravimetric (TG) and the differential thermogravimetric (DTG) curve are
shown in Fig. 1. The DTG exhibits two distinct peaks, one occurring between 200-350oC and
the other between 350-450oC with some overlap around 350oC.
Operating Temperature (oC)
0 200 400 600 800
Mas
s %
0
20
40
60
80
100
120
TGADTG
Figure 1: TG and DTG curves of poultry DAF
The first peak in the DTG occurs between around 200-350oC and corresponds to the
volatilization of proteins and polyunsaturated fatty acids(Rodante et al., 1992; Souza et al., 2004;
22
Maher and Bressler, 2007). The second peak overlaps the first and occurs around 350-450oC
with the peak occurring above 400oC. This peak corresponds with the volatilization of
monounsaturated and saturated fatty acids (Souza et al., 2004; Adebanjo et al., 2005; Maher and
Bressler, 2007).
3.3. Pyrolysis tests
The DAF samples were pyrolyzed at 400, 500, 600, and 700oC with three different runs
from each. The temperature profile and the gases released for a typical pyrolysis run with final
temperature of 600oC are shown in Fig. 2 and 3.
When the graphs of the TG/DTG (Fig. 1), temperature profile (Fig. 2), and gas evolution
(Fig. 3) are compared they give a picture of the effects of the pyrolysis process on the DAF
skimmings. The gas generated during the pyrolysis of DAF skimmings (Fig. 3) was mostly
comprised of CO2 that was produced during the thermal decomposition of triglycerides and
amino acids (Rodante et al., 1992; Maher and Bressler, 2007). At 150oC the amount of CO2
increases steadily until it peaks at 400oC where it drops rapidly until it levels off for the rest of
the experiment to 600oC. Additionally, there is a peak on the gas generated curve at 500oC that
does not correspond with a peak on the DTG curve. This peak corresponds to the generation of
C1-C4 parafins, olefins, and alkynes from further decomposition of fatty acids (Adebanjo et al.,
2005; Maher and Bressler, 2007).
After the amount of CO2 in the gas stream begins to decrease at 400oC, the amount of
paraffins and olefins and trace amounts of other C2-C4 compounds increase. Primarily at 500oC
there is an increase of CH4, C2H6, and C3H8 produced (Table 3). These gases are typical
products of the further decomposition of triglycerides (Maher and Bressler, 2007).
23
Figure 4 shows the yields of products at different final pyrolysis temperatures. The
pyrolysis of DAF skimmings yielded a very high yield of oil. The highest yield of oil was at the
600oC where the average oil yield was 78.5%. The lowest average yield of oil was 66.94% at an
operating temperature of 400oC and this result was supported by the fact that there was evidence
of unvolatilized volatile material left in the reactor vessel at 400oC.
Additionally, the results of the TGA data also support the finding that not all of the
volatile material in the DAF skimmings was volatilized. From the TG graph (Figure 1) it can be
seen that the mass of the sample is still decreasing beyond 400oC indicating that some volatile
material remains. The yields of the pyrolysis products as a function of operating temperature are
included in Figure 4.
Operating Time (Min.)
0 20 40 60 80 100 120 140 160
Inte
rnal
Tem
pera
ture
(o C)
0
100
200
300
400
500
600
700
Figure 2 - Temperature profile at 600 oC
24
0.0000
0.0010
0.0020
0.0030
0.0040
0.0050
0.0060
0.0070
0.0080
0.0090
0.0100
0 100 200 300 400 500 600
Operating Temp (C)
Mas
s (g
/min
)
Figure 3: Mass rate of gas evolved
66.94%71.98%
78.15% 77.12%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
400 500 600 700Operating Temperature (oC)
Yiel
d Oil
Char
Gas
Figure 4: Yields of products
25
3.4.- Analysis of Products 3.4.1. - Char Characterization
The results of the characterization of the char are included in Table 2. The oxygen
content was calculated by difference.
Table 2: Char characterization Property 400oC 500oC 600oC 700oC Carbon (%) 74.29 55.13 57.73 61.48
Hydrogen (%) 6.92 3.26 2.4 1.87
Oxygen (%)* 2.35 22.66 17.03 16.49
Nitrogen (%) 7.99 6.84 7.21 6.42
Sulfur (%) 0.06 0.02 0.06 0.02
Moisture (%) 0.90 4.76 3.78 5.06
Volatiles (%) 42.62 15.88 14.36 10.47
Ash (%) 8.39 12.10 15.56 13.72
Fixed Carbon (%) 48.09 67.25 66.30 70.75
*Determined by difference
The values for the characterization of char are close in value to those reported for other
waste sludge feedstocks (Lua and Guo, 1998; Hwang et al., 2007). The nitrogen content of the
char is higher than that of chars reported in literature for chars produced from lignocellulosic
material (0.5-1.5%) and from waste sludge feedestocks (3.5-3.8%) (Hwang et al., 2007). This is
likely due to the high level of proteins in the initial feedstock.
The high volatiles content at 400oC is worth note as it supports the yield data and visual
observations from that operating temperature that the volatile material in the biomass had not
fully volatilized at that temperature. This also causes the values for CHNS to be skewed toward
carbon for 400oC. Between 400-500oC, much of the volatile material had been volatilized and
26
the elemental composition of the char follows a more expected trend where as the temperature of
the experiment increases the produced char becomes more carbonaceous and the percentage of
hydrogen and oxygen decreases with temperature (Sharma and Hajaligol, 2003).
3.4.2. - Gas Characterization
The molar (volumetric) composition of the gas collected during the pyrolysis experiment
was determined using GC. The molar composition in Table 3 is of the gas collected during the
pyrolysis runs from 250-550oC. This range represents the period when the majority of non-
condensable gases were generated.
The mass flow rate of gas as determined from the gas data collected during the
experimentation at 600oC (Figure 3 & Table 3) was used to calculate the total mass of gas
produced by numerical integration. This was done so it could be compared to the mass of gas
produced as determined from the yield of products. Using the gas data collected during the
experiment, the mass of gas generated from the pyrolysis of 1.22 kg of sample was 0.108 kg and
the mass of gas generated as calculated by difference from the product yield for the same
pyrolysis experiment was 0.113 kg, a difference of 0.005 or just under 5%. This was used as a
check of the reliability of the mass flow rate of gas generated data as determined during
experimentation.
27
Table 3: Molar composition of gas generated (mol %) 250 300 350 400 450 500 550 600
Hydrogen 0.00 0.03 0.07 0.54 1.44 4.56 5.92 31.06
Methane 0.66 0.71 1.47 4.18 14.69 30.35 33.16 26.53
CO 2.21 4.63 5.94 11.94 12.33 5.39 2.41 4.77
CO2 96.29 93.75 91.12 73.33 38.29 8.42 10.03 30.13
Ethylene 0.04 0.13 0.20 0.69 2.77 4.59 4.41 0.60
Ethane 0.02 0.09 0.20 1.49 8.39 15.52 15.48 2.03
1,2-Propadiene 0.00 0.00 0.00 1.10 3.37 4.59 4.05 0.41
iso-Butane 0.51 0.22 0.20 0.43 0.71 0.72 0.61 0.20
n-Hexane 0.00 0.00 0.00 0.37 1.35 1.97 1.77 0.70
Propylene 0.17 0.11 0.17 0.95 2.84 3.91 3.47 0.57
Propane 0.00 0.12 0.28 2.73 7.92 11.40 10.55 1.44
trans-2-Butene 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
iso-Butylene 0.00 0.00 0.00 0.09 0.37 0.60 0.57 0.08
1-Butene 0.00 0.02 0.04 0.39 1.57 2.39 2.17 0.31
cis-2-Butene 0.00 0.06 0.13 0.49 0.83 0.78 0.61 0.16
iso-Pentane 0.00 0.00 0.00 0.06 0.27 0.44 0.42 0.06
n-Pentane 0.00 0.03 0.05 0.16 0.29 0.28 0.22 0.06
1,3-Butadiene 0.10 0.10 0.11 0.52 1.66 2.26 1.96 0.42
trans-2-Pentene 0.00 0.00 0.00 0.01 0.06 0.09 0.08 0.01
2-methyl-2-Butene 0.00 0.00 0.00 0.05 0.22 0.38 0.36 0.07
1-Pentene 0.00 0.00 0.03 0.15 0.25 1.08 1.54 0.07
cis-2-Pentene 0.00 0.00 0.01 0.31 0.30 0.07 0.00 0.27
28
3.4.3. - Oil Characterization
Some of the species present in bio-oils were identified using GC/MS. The bio-oil
consisted mostly of saturated and unsaturated fatty acids and fatty nitriles and alcohols (Figure 5
& Table 4). The melting point and boiling point of each compound is included in Table 4.
When compared to the pathway proposed in Idem et al. (1997) the compounds identified in the
raw bio-oil fit the reaction mechanism for the thermal decomposition of saturated and
unsaturated triglycerides. The presence of saturated and unsaturated fatty acids is a result of the
initial cracking of triglycerides. The fatty alcohol in the bio-oil is produced by the
decarbonylation of saturated oxygenated hydrocarbons which is an intermediate step in the
thermal decomposition of triglycerides (Idem et al., 1997). The fatty nitriles are a product of a
reaction of fatty acids with ammonia (Ekinci et al., 1994). The ammonia is a by-product of the
thermal decomposition of the proteins in the DAF skimmings (Rodante et al., 1991). These
nitriles would be undesirable in a fuel application but can be converted into fatty amines that
have industrial applications as surfactants (Ostgard et al., 2007).
The results of this characterization along with values for bio-oils and bio-diesels
produced from other high triglycerides feedstocks are included in Table 5. As can be seen from
the data, the raw bio-oil had an energy density of ~36 MJ/kg and a water content of ~5%. The
cloud point and viscosity are too high for fuel applications, though, and must be upgraded before
they can be used as such.
29
E
D
G
FB H
C A
Figure 5: Chromatogram of Raw Bio-oil
Table 4: Raw bio-oil composition determined via GC/MS
Compound Chemical Description Melting Point (oC)
Boiling Point (oC)
A Ricinoleic Acid Unsaturated Fatty Acid 5.5 245
B Hexadecanenitrile Fatty Nitrile - -
C Z-11-Hexadecenoic Acid Unsaturated Fatty Acid - -
D Palmitic Acid Saturated Fatty Acid 59-63 350-351
E 1-Heptadecanol Fatty Alcohol 56-58 308
F Octadecanenitrile Fatty Nitrile - -
G Oleic Acid Unsaturated Fatty Acid 13-14 360
H Stearic Acid Saturated Fatty Acid 71 360
30
Table 5: Characterization of raw bio-oils
DAF Bio-oils Values for Biodiesel from Literature*
Carbon (%) 73.17 75-76
Hydrogen (%) 11.12 11.5-12.5
Nitrogen (%) 4.25 <0.02
Oxygen (%)* 8.08 11.40
Sulfur (%) 0.12 Below detectable limits
Viscosity (cP) at 40oC 37.80 10-12
Viscosity (cP) at 60oC 16.18 N/A
Water Content (wt%) 4.72 3-6
HHV (MJ/kg) 36.06 38-40
Cloud Point (oC) 61.72 10-15
*(Kinast 2003; Adebanjo et al., 2005)
The carbon, nitrogen, and oxygen content of the pyrolysis oil were all slightly lower than
that of bio-oil and the difference was made up by a higher nitrogen content in the pyrolysis oil
(Kinast, 2003). This is due to the presence of proteins in the DAF feedstock. The heating value
for the raw bio-oil is slightly less than that found in the literature for bio-diesels produced from
cleaner high-triglyceride feedstocks (Kinast, 2003) and the presence of saturated fatty acids in
the oil causes the cloud point and viscosity to be too high for fuel applications. There is a need
to upgrade the oils for these purposes.
3.5. - Solvent Extraction
Initially, eight solvents were evaluated for their ability to extract an oil fraction with a
clouding point in the range specified for traditional bio-diesel (-3-12oC) (Chongkhong et al.,
2007). For the initial panel eight solvents chosen were hexane, dichloromethane, ethyl ether,
31
acetone, toluene, ethanol, propanol, and butanol. Toluene, ethanol, butanol, and propanol failed
to produce a liquid fraction so they were excluded from further evaluation.
The soluble fractions of the other four solvents were characterized by yield of soluble
fraction, water content, and oxidative onset temperature (a measure of the oxidative stability as
an indicator of long-term storagability). This test is important to ensure that the resulting fatty
acid methyl esters will not be very unstable leading to increased tendencies to form gums in
storage tanks.
The oxidation onset temperature (OOT) was determined using DSC curves by taking the
intercept of the baseline and the slope of the peak. Figure 6 is an example of how the OOT was
determined graphically. These results are presented in Table 6. The values for OOT for all
soluble fractions compared favorably to literature values of OOT for bio-diesels (~125oC)
(Dunn, 2005). There wasn’t much difference in performance among the four solvents.
Table 6: Characterization of soluble fractions Yield (%) OOT (oC) Water Content (%)
Acetone 58% 227.26 0.22
Dichloromethane 54% 231.49 0.23
Hexane 50% 230.41 0.34
Ethyl Ether 47% 232.99 0.45
32
Temperature (oC)
50 100 150 200 250 300 350 400
Hea
t Flu
x (m
W)
-0.5
0.0
0.5
1.0
1.5
2.0
Onset: 234.63oC
Figure 6: Determination of oxidative onset temperature
A second trial was performed using acetone along with ethyl acetate and methyl ethyl
ketone, the latter two of which were chosen due to their similar nature to acetone. Additional
extractions were run using acetone, ethyl acetate, and MEK on a random bio-oil from each
operating temperature.
The extracted oils were evaluated using cloud point and yield. The cloud point was
determined graphically from DSC curves as illustrated in Figure 7. The results of the cloud point
analysis are included in Table 7.
The cloud points for all three soluble fractions were fairly comparable. The measured
value for the cloud point for the methyl ethyl ketone soluble fraction at 500oC fell outside of the
reasonable range of the rest of the data points. The sample was rerun and the results were the
same. A summary of the yield data is also presented in Table 8. Additionally, the yield of
33
soluble fraction as a function of the raw DAF is included using 600oC as the basis as it yielded
the highest yield of oil for all operating temperatures.
Table 7: Cloud point of raw oil and soluble fractions Cloud Point oC
400 500 600 700 Raw Oil 62.40 62.32 57.37 64.80
Ethyl Acetate Soluble Fraction 0.85 7.99 2.02 -1.19
Acetone Soluble Fraction -0.27 2.74 -0.99 -1.17
MEK Soluble Fraction 2.07 17.66 1.40 -0.40
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0.0
Temperature (oC)
O nset: -0.99oC
Hea
t flu
x (m
W)
-40 -20 0 20 40 60 80
Figure 7: Cloud point determination
34
Table 8: Yields of soluble fractions
400 500 600 700
Bio-oil Produced at 600oC
Ethyl Acetate Soluble Fraction 43% 62% 63% 35% 49.23%
Acetone Soluble Fraction 42% 40% 51% 27% 39.19%
MEK Soluble Fraction 44% 43% 53% 33% 41.47%
The ethyl acetate soluble and insoluble fractions were characterized using GC/MS. The
chromatograms for both are included in Figures 8 and 9. A summary of the compounds detected
is included in Table 4. Once the GC/MS was performed the absolute area under the curve on the
GC was calculated for each compound and the ratio of soluble to insoluble was taken to
determine in which fraction each compound favored (Table 9).
C
H
G
F
ED
B
A
Figure 8: Chromatogram of Ethyl Acetate-insoluble fraction
35
E
G
B
F
D H
A C
Figure 9: Chromatogram of Ethyl Acetate-soluble fraction
Table 9: Soluble and insoluble compounds as determined by GC/MS
Ethyl Acetate Soluble Ethyl Acetate Insoluble Ricinoleic Acid Palmitic Acid
Hexadecanenitrile Stearic Acid
Z-11-Hexadecenoic Acid 1-Heptadecanol
Oleic Acid
Octadecanenitrile
The results of the GC/MS for the soluble and insoluble fractions confirms the findings in
the lab that the insoluble fraction was made up of saturated fatty acids with high melting points
that caused the solid nature of the bio-oil at room temperature and the high cloud point of the raw
bio-oil.
A summary of the characterization of the ethyl acetate soluble fraction as compared to the
raw bio-oil is included in Table 10.
36
Table 10: Comparison of ethyl acetate soluble fraction and raw bio-oil
Soluble Fraction Raw Bio-oil
Cloud Point (oC) 2.02 57.37
Viscosity at 40oC (cP) 14.30 37.80
Water Content (wt%) 0.134 4.72
HHV (MJ/kg) 39.04 36.06
The ethyl soluble fraction exhibited much better flow properties than that of the raw bio-
oil with a much lower cloud point and a reduced viscosity at 40oC. The ethyl acetate-soluble
fraction also had lower water content and a higher heating value than that of the raw bio-oil
3.6. – Esterification
The chromatogram of the esterified ethyl acetate-soluble bio-oil fraction (Figure 10)
contained peaks of esterified unsaturated fatty acids along with fatty acid nitriles that are present
in the ethyl acetate-soluble fraction. Additionally, there are two new peaks of saturated fatty
acid esters from saturated fatty acids that remained in the soluble fraction.
9-Octadecanoic acid methyl ester
11-Hexadecenoic acid methyl ester
Figure 10: Gas chromatogram of esterified ethyl acetate soluble fraction
37
The FT-IR spectrums for the ethyl acetate-soluble fraction and for the esterified ethyl acetate-
soluble fraction are included in Figure 11 and 12 respectively.
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
5001000150020002500300035004000Wavenumber
Abu
ndan
ce
Figure 11: FT-IR of ethyl acetate-soluble fraction
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
5001000150020002500300035004000Wavenumber
Abu
ndan
ce
Figure 12: FT-IR of biodiesel produced from ethyl acetate-soluble fraction
38
On the FT-IR spectrum of the esterified ethyl acetate-soluble fraction (Figure 12) the presence of
bands occurring between wavenumber 1260-1000, which are absent on the spectrum of the un-
esterified fraction (Figure 11), is consistent with the formation of esters (Silverstein, 2005).
4. Conclusions The pyrolysis of poultry DAF skimmings produced a bio-oil with a high heating value
but that was also high in saturated fatty acids. The level of saturation of the fatty acids causes
the bio-oil to have a high cloud point and viscosity due to the high melting point of the saturated
fatty acids. During the project solvent extraction was used to extract a fraction with a reduced
amount of saturated fatty acids. This fraction showed significant improvement in cold flow
properties with a reduced cloud point and viscosity. There was also improvement in the water
content and higher heating value over the raw bio-oil. The process produced a bio-oil containing
unsaturated fatty acids that could be esterified into a bio-diesel and fatty nitriles that could be
further processed into surfactants.
39
References:
Adebanjo, A., M. G. Kulkarni, A. K. Dalai, and N. N. Bakhshi (2007). "Pyrolysis of waste fryer
grease in a fixed-bed reactor." Energy & Fuels 21(2): 828-835.
Adebanjo, A. O., A. K. Dalai, and N. N. Bakhshi (2005). "Production of diesel-like fuel and
other value-added chemicals from pyrolysis of animal fat." Energy & Fuels 19(4): 1735-
1741.
Bailey, K. M. (1989). "Solvent Extraction of Oil and Grease from Poultry Processing
Wastewater." University of Arkansas Masters Thesis.
Berrios, M., J. Siles, M. A. Martin, and A. Martin (2007). "A kinetic study of the esterification of
free fatty acids (FFA) in sunflower oil." Fuel 86(15): 2383-2388.
Bridgwater, A. V. and G. V. C. Peacocke (2000). "Fast pyrolysis processes for biomass."
Renewable and Sustainable Energy Reviews 4: 1-73.
Cai, T., O. C. Pancorbo, H. M. Barnhart (1994). "Chemical and Microbiological Characteristics
of Poultry Processing By-Products, Waste, and Poultry Carcasses During Lactic Acid
Fermentation." J APPL POULT RES 3(1): 49-60.
Canakci, M. (2007). "The potential of restaurant waste lipids as biodiesel feedstocks."
Bioresource Technology 98(1): 183-190.
Chongkhong, S., C. Tongurai, P. Chetpattananondh, and C. Bunyakan (2007). "Biodiesel
production by esterification of palm fatty acid distillate." Biomass and Bioenergy 31(8):
563-568.
Dunn, R. O. (2005). "Effect of antioxidants on the oxidative stability of methyl soyate
(biodiesel)." Fuel Processing Technology 86(10): 1071-1085.
40
Ekinci, E., S. Turkay, M. Citiroglu, and A. Akar (1994). "Nitrogen compounds in pyrolysis
produced liquids from two Turkish oil shales." Fuel Processing Technology 37(2): 175-
184.
Goldman, Y., N. Garti, B. Ginzburg, and M. R. Bloch (1981). "Conversion of halophilic algae
into extractable oils. 2. Pyrolysis of proteins." Fuel 60: 90-92.
Gomez-Rico, M. F., I. Martin-Gullon, A. Fullana, J. A. Conessa, and R. Font (2003). "Pyrolysis
and combustion kinetics and emissions of waste lube oils." Journal of Analytical and
Applied Pyrolysis 68-69: 527-546.
Hwang, I. H., T. Matsuto, N. Tanaka, Y. Sasaki, and K. Tanaami (2007). "Characterization of
char derived from various types of solid wastes from the standpoint of fuel recovery and
pretreatment before landfilling." Waste Management 27(9): 1155-1166.
Idem, R. O., S. P. R. Katikaneni, and N. N. Bakhshi (1997). "Catalytic conversion of canola oil
to fuels and chemicals: roles of catalyst acidity, basicity and shape selectivity on product
distribution." Fuel Processing Technology 51(1-2): 101-125.
Issariyakul, T., M. G. Kulkarni, A. K. Dalai, and N. N. Bakhshi (2007). "Production of biodiesel
from waste fryer grease using mixed methanol/ethanol system." Fuel Processing
Technology 88(5): 429-436.
Karayildirim, T., J. Yanik, M. Yuksel, and H. Bockhorn (2006). "Characterisation of products
from pyrolysis of waste sludges." Fuel In Press, Corrected Proof.
Kinast, J. A. (2003). "Production of Biodiesels from Multiple Feedstocks and Properties of
Biodiesels and Biodiesel/Diesel Blends: Final Report; Report 1 in a Series of 6." NREL
Report.
41
Kulkarni, M. G. and A. K. Dalai (2006). "Waste cooking oil-an economical source for biodiesel:
A review." Industrial & Engineering Chemistry Research 45(9): 2901-2913.
Lang, X., A. K. Dalai, N. N. Bakhshi, M. J. Reaney, and P. B. Hertz (2001). "Preparation and
characterization of bio-diesels from various bio-oils." Bioresource Technology 80(1): 53-
62.
Lee, J. H. (2003). "RECOVERY PROCESS FOR FOOD SKIMMINGS FROM DISSOLVED
AIR FLOTATION UNITS." Ninth International Animal, Agricultural and Food
Processing Wastes Proceedings of the 12-15 October 2003 Symposium.
Lima, D. G., V. C. D. Soares, E. B. Ribeiro, D. A. Carvalho, E. C. V. Cardoso, F. C. Rassi, K. C.
Mundim, J. C. Rubim, and P. A. Z. Suarez(2004). "Diesel-like fuel obtained by pyrolysis
of vegetable oils." Journal of Analytical and Applied Pyrolysis 71(2): 987-996.
Lua, A. C. and J. Guo (1998). "Preparation and characterization of chars from oil palm waste."
Carbon 36(11): 1663-1670.
Mahadevaswamy, M. and Venkataraman (1986). "Bioconversion of poultry droppings for biogas
and algal production." Agricultural Wastes 18: 93-101.
Maher, K. D. and D. C. Bressler (2007). "Pyrolysis of triglyceride materials for the production of
renewable fuels and chemicals." Bioresource Technology 98(12): 2351-2368.
McKendry, P. (2002). "Energy production from biomass (part 2): conversion technologies."
Bioresource Technology 83(1): 47-54.
Mellan, I. (1957). Handbook of Solvents. New York, Reinhold Pub. Corp.
Miao, X. and Q. Wu (2004). "High yield bio-oil production from fast pyrolysis by metabolic
controlling of Chlorella protothecoides." Journal of Biotechnology 110 85-93.
42
Miao, X., Q. Wu, and C. Yang (2004). "Fast pyrolysis of microalgae to produce renewable
fuels." Journal of Analytical and Applied Pyrolysis 71(2): 855-863.
Oliveira, J. S., R. Montalvão, L. Dahre, P. A. Z. Suarez, and J. C. Rubim (2006). "Determination
of methyl ester contents in biodiesel blends by FTIR-ATR and FTNIR spectroscopies."
Talanta 69(5): 1278-1284.
Ostgard, D. J., R. Olindo, M. Berweiler, S. Roder, and T. Tacke (2007). "The chemoselective
hydrogenation of tallow nitriles to unsaturated 1[degree sign] fatty amines with carbon
modified Ni catalysts." Catalysis Today 121(1-2): 106-114.
Peng, W., Q. Wu, and P. G. Tu (2001). "Pyrolytic characteristics of heterotrophic Chlorella
protothecoides for renewable bio-fuel production." Journal of Applied Phycology 13: 5-
12.
Rana, M. S., V. Sámano, J. Ancheyta, and J. A. I. Diaz (2007). "A review of recent advances on
process technologies for upgrading of heavy oils and residua." Fuel 86(9): 1216-1231.
Robert M. Silverstein and F. X. Webster (2005). Spectrometric Identification of Organic
Compounds. New York, Wiley.
Rodante, F., G. Marrosu, and G. Catalani (1991). "Thermal analysis of some a-amino acids with
similar structures." Thermochimica Acta 194: 197-213.
Rodante, F., G. Marrosu, and G. Catalani (1992). "Thermal analysis of some [alpha]-amino acids
with similar structures." Thermochimica Acta 194: 197-213.
Sharma, R. K. and M. R. Hajaligol (2003). "Effect of pyrolysis conditions on the formation of
polycyclic aromatic hydrocarbons (PAHs) from polyphenolic compounds." Journal of
Analytical and Applied Pyrolysis 66(1-2): 123-144.
43
Shinogi, Y. and Y. Kanri (2003). "Pyrolysis of plant, animal and human waste: physical and
chemical characterization of the pyrolytic products." Bioresource Technology 90: 241-
247.
Souza, A. G. d., J. C. O. Santos, M. M. Conceicao, M. C. D. Silva, and S. Prasad (2004). "A
THERMOANALYTIC AND KINETIC STUDY OF SUNFLOWER OIL." Brazilian
Journal of Chemical Engineering 21(2): 265-273.
Swisher, K. (2007). Market Report 2006. Render Magazine. April 2007: 10-17.
44
45
CHAP TER 4
CONCLUSION
The pyrolysis of poultry DAF skimmings produced a raw bio-oil with a high heating
value (36.06 MJ/kg) and a low water content (4.72%). The raw bio-oil was also high in saturated
fatty acids that cause the bio-oil to have poor cold flow properties as indicated by a high
viscosity at 40oC (37.80 cP) and a high cloud point (61.72oC). These poor cold flow properties
are the result of the high melting points of the fatty acids found in the bio-oil (60-70oC).
Saturated fatty acids are undesirable in a biodiesel feedstock because their poor cold flow
properties would be passed on to the biodiesel.
During the project solvent extraction with ethyl acetate was used to extract a fraction rich
in unsaturated fatty acids. This fraction showed significant improvement in cold flow properties
with a reduced cloud point (2.02oC) and viscosity at 40oC (14.30 cP). There was also
improvement in the water content (0.1%) and higher heating value (39.04 MJ/kg) over the raw
bio-oil.
This unsaturated fatty acid-rich bio-oil fraction was then esterified using an acid catalyst
(sulfuric acid) and methanol to produce fatty acid methyl esters (FAMEs). The presence of these
FAMEs was verified using GC/MS and FTIR.