The Agricultural Biorefinery Innovation Network (ABIN): A Canadian

Agricultural Bioproducts Innovation Program
(ABIP)
The Agricultural Biorefinery
Innovation Network (ABIN):
A Canadian Network for Research
i Green
in
G
Energy,
E
Fuels
F l and
d
Chemicals
Franco Berruti
Network Leader
Institute for Chemicals and Fuels
from Alternative Resources
The University of Western Ontario
London, Ontario, CANADA
$ 8.7 M AGRICULTURAL BIOREFINERY
INNOVATION NETWORK (ABIN) [2008-2011]:
70 researchers from 17 Canadian Institutions
(Academia, Government and Industry)
Vision of ABIN
• to enable Canada to exploit its plentiful
pp
of biomass,,
supplies
• focusing on agricultural (non-food) coproducts, residues, and selected energy
crops
• through research and development of novel
technologies
tec
o og es for
o tthe
e eco
economical
o ca a
and
d
sustainable conversion of such resources
into energy and value-added products,
• and
d tto supportt th
the d
development
l
t off the
th
emerging bio-based economy
Goal of ABIN
Contribute to, and encourage,
sustainable
t i bl d
development
l
t
while strengthening Canada’s rural
economy with the creation of new
businesses and jobs
Key Features of Vision (1)
• Similarly to the developments of
the past century using petroleum
feedstocks, we are focusing on
the BIOREFINERY approach,
where the feedstock is a
sustainable, renewable, and low
value material and a large
spectrum of value-added
products are generated in
Key Features of Vision (2)
• The key elements are:
• full life-cycle assessment,
• sustainability,
sustainability
• environmental preservation
• creation of value and jobs.
Key Features of Vision (3)
•C
Connecting
ti clusters
l t
off expertise
ti
from across Canada and leveraging
synergies,
i
• Sharing and distribution of
k
knowledge
l d related
l t d tto th
the
advancement of biorefining,
• Contribution
C t ib ti tto th
the d
development
l
t off
a vigorous and enduring Canadian
bi
bioeconomy
through
th
h education
d
ti and
d
training of HQP
Participating Institutions
University of Western Ontario
University of Toronto
École Polytechnique de Montreal
University of Northern British
Col mbia
Columbia
University of Guelph
University of Alberta
University of Manitoba
Agri-Therm Limited
Ryerson University
Perth Community Futures
University of Saskatchewan
Saskatchewan Research Council
Université de Sherbrooke
Stormfisher Ltd
Ltd.
University of British Columbia
National Research Council
Agriculture and Agri-Food
Canada
Research Themes:
1. Feedstock Enhancements and Biorefinery
Interface
2. Green Chemicals
3. Green Fuels
4. Green Energy
5. Life-Cycle Assessment and Technology
Integration
6. Knowledge Transfer, Technology Transfer,
Commercialization and Policy
Development
1)) Feedstock Enhancement and
Biorefinery Interface
L. Tabil ((UofS),
) S. Sokhansanjj ((UBC),
) S. Panigrahi
g
((UofS),
) G. Turcotte ((Ryerson),
y
) P.
Krishna (Western), R. Knox (AAFC), N. Huner (Western)
• Research to reduce handling,
g storage
g and
processing costs, and to ensure a steady supply of
agricultural-based lingocellulosic feedstocks required
for biorefineries
PROJECTS:
•
•
•
•
Pre-processing
p
g and densification
Rheology of pre-processed straw
Feedstock genetic optimization
Supercritical CO2 pre-treatment prior to
enzyme hydrolysis
h d l i
2) Green Chemicals
C. Briens (Western), I. Scott (AAFC), F. Berruti (Western), X. Bi (UBC), J. Chaouki (École
Poly), P. Charpentier (Western), E. Chornet (Sherbrooke), A. Dalai (UofS), Y. Dahman
(Ryerson), R. Dutton (Guelph), R. Golden (Agri-Therm), B. McGarvey (AAFC), S. Liss
(Gueph)
• Research on the efficient use of crops or plant
residual materials to generate valuable chemicals
and pharmaceuticals for agricultural, industrial or
medicinal uses (i.e., green chemicals)
PROJECTS:
• Pyrolytic bio-oil production
• Reactor Technology: fluid bed
bed, rotating fluid bed
bed,
microwave
• Chemical identification and biological
g
activity
y of biooils
• Extraction and separation
• Glycerol conversion
• Bio-char
• Monomers and polymers
• Functional biomaterials
3) Green Fuels
A. Dalai (UofS), J. Chaouki (École Poly), R. Ranganathan (SRC), D. Anweiler (SRC), N.
Ellis (UBC), K. Smith (UBC), S. Duff (UBC), P Watkinson (UBC), N. Abatzoglou
(Sherbrooke), E. Chornet (Sherbrooke), C. Briens (Western), F. Berruti (Western), H.
DeLasa (Western), H. Wang (NRC), G. Wolfaardt (Ryerson), Y. Dahman (Ryerson), A.
Lohi (Ryerson)
(Ryerson), G
G. Hill (UofS)
(UofS), JJ. Kozinski (UofS)
(UofS), T
T. ugsley (UofS)
(UofS), C
C. Niu (UofS)
(UofS), B
B.
Roesler (PhibrioChem), D. Bayrock (PhibrioChem), S. Helle (UNBC), W. McCaffrey
(UofA), M. Thomson (UofT)
• D
Develop
l iintegrated
t
t d and
d original
i i l approaches
h ffor th
the
complete utilization of biomass feedstocks to
produce g
p
green fuel p
products
PROJECTS:
• Bio-diesel p
production and application
pp
(nanocatalysts, quality improvement)
• Bio
Bio-oil
oil production and upgrading
• Syn-gas, Hydrogen and Bio-gas
• Bio-ethanol
Bi th
l and
d bi
bio-butanol
b t
l
4) Green Energy
M. Thomson (UofT), A. Dalai (UofS), J. Chaouki (École Poly), C. Briens (Western), F.
) H. Wang
g ((NRC),
) G. Hill ((UofS),
) E. Bibeau ((Manitoba))
Berruti ((Western),
• Integrates fuel cells, pyrolysis, combustion and
biological technologies for heat and power
production into sustainable agricultural cycles.
Research related to production of green energy.
PROJECTS:
•
•
•
•
Pyrolysis bio-oil for heat and power
C b ti off bi
Combustion
biomass iin spouted
t db
bed
d
Direct Liquid Fuel Cells for bio-fuels
Algae growth in CO2 for ethanol and
y
electricity
• Brayton Hybrid Cycle for heat and power
5) Life Assessment and Technology
g
Integration
L. Townley-Smith (AAFC), R. Samson (École Poly), L. Deschenes (École Poly), X. Bi
(UBC), M. Wismer (SRC)
• Life cycle approach:
• integrate the environmental variables
• optimize industrial processes
• minimize the risk of major problems after the
introduction of one of these technologies and
ensure that the new technology does not shift
the problem elsewhere
PROJECTS:
• Biomass inventory
y mapping
pp g and analysis
y
tools integration in LCA
• LC inventory database for biofuel and
bioenergy
• Energy and materials flows in biofuels
production
6) Knowledge Transfer, Technology Transfer,
Commercialization and Policy Development
T. Bansal (Western), D. Cunningham (Western), C. Guilon (StormFisher), B. van Berkel
(StormFisher), R. Golden (Agri-Therm), J. Henhoeffer (PCFDC), D. Lee (AAFC), L . Townley) M. Stumborg
g ((AAFC),
) J. Adams ((Western),
) D. Hewson ((Western),
) J. Kabel
Smith ((AAFC),
(Western), R. Ranganathan (SRC)
• Investigate
g
existing
g knowledge
g networks to identify
y the
key success factors required to develop sustainable
biorefinery clusters in Canada
• Examine the degree political (policy)
(policy), economic and
social factors will influence entrepreneurial firms’
technology development decision-making and
performance.
PROJECTS:
• Existing
g knowledge
g and success factors to
develop sustainable biorefinery clusters
• Influence of political
political, economic and social
factors on entrepreneurial firms’
technology decision-making
technology,
decision making and
performance
ABIN Administration
• Network Lead
Franco Berruti
berruti@eng.uwo.ca
• Network Manager
Chantal Gloor cgloor@eng.uwo.ca
• Financial Administrative Assistant
Pi S
Pina
Sorbara
b
psorbara@eng.uwo.ca
b @
• Administrative Assistant
Christine Ramsden cramsden@eng.uwo.ca
ABIN Governance
• Network Management Committee
Committee, chaired
successively by a Federal Network Lead and the
Recipient Network Lead, with membership
representation
t ti from
f
allll 6 th
themes; meett att least
l
t2
times annually
• Board of Directors with a role to recommend
strategies to heighten the relevance and impact
off the
th workk plans,
l
as wellll as id
identifying
tif i
mechanisms related top sustainability of Network.
Numerical Study of Fast Pyrolysis of Woody Biomass in a GravityDriven Reactor
H.S. Choi1*, Y.S. Choi1, S.J. Kim1
1. Environment and Energy Systems Research Division, Korea Institute of Machinery Materials, Daejeon, South Korea
* Corresponding author: hschoi@kimm.re.kr
INTRODUCTION
Gravity-Driven Reactor
for Fast Pyrolysis
To overcome environmental problems such as CO2 discharge caused by fossil fuels, fast pyrolysis method becomes bright prospect for
thermal conversion of biomass into biocrude-oil, which can be used for heat and power generation and additionally bio-refinery.
In order to design cost-effective fast pyrolysis reactor, it is necessary to increase biocrude-oil yield and, at the same time, to decrease
energy and working materials which are needed for the fast pyrolysis process.
Hence, simple gravity-driven reactor is devised in the present study, which does not demand working fluids and related energy to run
them typically needed for fluidizing techniques.
In the present study, the gravity-driven fast pyrolysis reactor is simplified as an inclined 2-dimensional duct and the flow and thermal
fields of the reactor and furthermore the effects of inclined angle and inlet height for sand are numerically investigated as a starting point
for the optimal design of the reactor and for future industrial application.
MATHEMATICAL FORMULATION (Eulerian-Eulerian Method)
COMPUTATIONAL CONDITIONS
Continuity equations for gas and solid are as follows;
n
t
g
g
g
g vg
R
,
n
sj
t
1
sj
sj v sj
sj
Computational Domain (2- Dimensional)
R sj
1
Momentum equations for gas and solid are as follows;
t
g
gvg
gvgvg
g
τg
Pg
g
n
Fgsj v sj
vg
g
gsj v sj
Height (H)
10 cm
Grid allocation (x,y)
100 x 35
Boundary Conditions
' gsj v g
j 1
sj
sj v sj
sj v sj v sj
sj
sj
n
Pg
S sj
Fgsj v sj
vg
Fsksj v sk
v sj
sj
sj g
Validation for Inclined Chute Flow
k 1
n
RM gsj
gsj v sj
' gsj v g
RM sksj
sjsk v sk
' sjsk v sj
k 1
Energy equations for gas and solid are as follows;
sj
100 cm
j 1
RM gsj
g
Length (L)
gg
n
t
Computational Conditions
Computational Domain
g C pg
sj C psj
Tg
t
Tsj
t
n
vg
Tg
qg
gsj
Tsj
Tg
H rg
Tsj
q sj
gsj
Tsj
Tg
Dirichlet (Vinlet=7.37 cm/s, T inlet=753 K)
Woody biomass inlet
Dirichlet (Vinlet=7.37 cm/s, Tinlet=300 K)
Outlet
Neumann
Wall
Johnson and Jackson, Dirichlet(T wall=753 K)
Particle Density
Wood
0.65 g/cm3
Char
1.0 g/cm3
Sand
2.5 g/cm3
Semi-global Two Stage Reaction Mechanism for
Wood Pyrolysis
j 1
v sj
Sand inlet
H rsj
Species equations for gas and solid are as follows;
t
g
g Yg
g
g Yg v g
Dg
Yg
Rg
,
t
sj
sj Ysj
sj
sj Ysj
v sj
Dsj
Ysj
Rsj
RESULTS AND DISCUSSION
Fig.1.(b) The primary reaction rate for tar (R1)
Fig.1.(a) Gas velocity vector
Fig.1.(d) Mass fraction of non-condensable gas by primary reaction
Fig.1.(e) Density of char by primary reaction
Fig.2.(a) The secondary reaction rate for non-condensable gas (R4)
Fig.2.(b) The secondary reaction rate for char (R5)
Fig.2.(c) Mass fraction of non-condensable gas by secondary reaction
Fig.2.(d) Density of char by secondary reaction
(b)
(a)
(c)
Fig.1.(c) Mass fraction of tar by primary reaction
(d)
Fig.3.(a) Reaction rates at x/H=3, (b) Granular temperature for sand at x/H=3, (c) Reaction rates for R1, (d) Granular temperatures for sand
(a)
(b)
(c)
(d)
(e)
(f)
Fig.4. Reaction rate R1; (a) x/H=1, (b) x/H=3, (c) x/H=9, Reaction rate R4; (d) x/H=1, (e) x/H=3, (f) x/H=9
(Case1: inclination angle of 45◦, Case2: inlet height for sand is increased to 4 times larger than that of case1, Case3: inclination angle of 55◦)
In Fig.1 (a), weak flow-recirculation region appears
upstream near the inlets and toward downstream the gas
flow is developed following the solid flow. For the
primary reaction rate of tar production (R1), the reaction
mainly takes place at very close to the bottom wall and
tar mass fraction is increased downstream. The mass
fraction of non-condensable gas and char density are
increased toward downstream in Figs.1 (d) and (e). In
particular, from Fig.1 (c) and Fig.2 (c), the tar entrained
into the flow recirculation region becomes noncondensable gas by the secondary reaction. Hence, the
length of reactor and the recirculation region should be
carefully considered to reduce the secondary reaction.
In Fig.3, it is noted that the maximum primary reaction
rates are located between the first and second peaks of
the granular temperature at very close to the wall, where
wood particles are heated by hot sand as well as the
heated bottom wall. Although, in general, the magnitude
of granular temperature is known as small compared
with mean particle velocity, the vigorous motions of the
hot sand and wood particles with higher granular
temperature may be helpful to mixing between wood and
sand particles and the consequent heat transfer from
sand to wood. In Fig.4, the case2 shows the highest
values of R1 and R4 compared with others at four
different streamwise positions. It is noted that in Fig.4 (a)
case2 has the highest R1 value and case 1 has the lowest
one, where the peak magnitude of granular temperature
shows the same pattern. Hence, the solid mixing and
consequent heat transfer have a great effect on the fast
pyrolysis reaction.
CONCLUDING REMARKS
In the present study, CFD is applied to the gas-particle flows with pyrolysis reaction in the gravity-driven reactor. To analyze the pyrolysis reaction of the reactor, the
semi-global two stage chemical kinetics having tar cracking mechanism is applied. From the results, it is noted that the vigorous motions of the hot sand particles
with higher granular temperature may be helpful to mixing between wood and sand particles and the consequent heat transfer for fast pyrolysis from sand to wood.
A cooperative program by:
Amber Broch, S. Kent Hoekman
PROCESS OPTIMIZATION
To support a DOE Cooperative Agreement with the Gas Technology
Institute (GTI), DRI is partnering with the University of Nevada, Reno
(UNR), the Renewable Energy Institute International (REII), and
Changing World Technologies (CWT) to demonstrate the viability of
hydrothermal pre-treatment as a method to convert lignocellulosic
biomass into a uniform, densified feedstock that could be easily fed
into a thermo-chemical conversion process to produce syngas,
pyrolysis oils, and other value-added products. DRI is focused on
feedstocks available in the State of Nevada, and is also conducting a
biomass resource assessment within the State.
Biomass feedstocks include a wide variety of materials that exhibit
significant differences in handling characteristics, energy content, and
recalcitrance to conversion -- all factors that must be accommodated
within a biorefinery context. Hydrothermal pretreatment of biomass
promises to produce a uniform solid that can be easily fed to any
thermochemical conversion process.
DRI is collecting and analyzing all products of the HPT process from a
variety of feedstocks (loblolly pine, rice hulls, corn stover, pinion/
juniper chips, and white fir/Jeffery pine chips. The products include:
• Pre-treated solid biomass or “bio-char”
• Condensed liquid
• Gases
Through a comprehensive set of lab analyses, we will perform
complete mass and energy balances of the HPT process. This
includes ultimate and proximate analyses, lignocellulosic composition,
and detailed chemical analysis.
BIOMAX 15
The Biomax 15, manufactured by Community
Power Corp. (CPC),
produces syngas by
gasification of wood
chips. The syngas is
then combusted in an The Biomax 15 produces 15 kW of electrical power by burning
engine/ generator set to syngas from gasification of biomass in a generator.
produce 15 kW of electrical power and provide available heat. We
intend to run the Biomax using pre-treated, Nevada-specific biomass.
PRELIMINARY DATA AND RESULTS
Some preliminary results from HPT of Alabama Loblolly Pine are
shown in Figures 3 & 4 below. The mass of the recovered dry biochar is calculated through moisture measurements. In this case, the
recovered solid is lower than expected, due to uncertainties in the
moisture content of the recovered wet bio-char.
SYNGAS CHARACTERIZATION
Dilution sampling will be used
to collect syngas from:
• raw wood feedstock
HYDROTHERMAL PRETREATMENT (HPT)
HPT transforms lignocellulosic biomass
into a uniform, friable solid with much
higher mass and energy densities than
the parent biomass (Fig. 2)
Approach
Figure 1. Loblolly pine chips before
and after pre-treatment
Biomass is treated in water at
temperatures around 260°C and
equilibrium pressures (~680 psig) for 25 minutes to produce a hydrophobic
solid that is easily dried and pelletized.
Other
products
include
noncondensable gases and condensed
liquid that is mostly water (Fig. 3)
1.8
Technical Accomplishments
Atomic H/C ratio
• The process takes less time
than
conventional
drytorrefaction.
Peat
1.4
Lignite
1.2
0.8
Pretreated Wood
Increased Heating Value
0.6
Pretreated Corn
Stover and Rice Hulls
Raw Corn Stover and
Rice Hulls
Torrefied Wood
0.4
0.2
• The mass of the feedstock
decreases while its energy
content is mostly retained.
Wood
Lignin
Cellulose
Coal
1.0
Anthracite
0
0.2
0.4
0.6
Atomic O/C ratio
Figure 2. HPT lowers O content and increases C
content, making biomass more similar to coal.
0.8
Carbonyl
Sampler
Canister
Sampler
• HPT wood feedstock
• conventionally torrefied
wood feedstock
Tenax (VOC)
Sampler
Dilution
Tunnel
Equipment for sampling and analysis of syngas.
Figure 3. Total material recovery from HPT of Alabama Loblolly Pine. 95% of starting material
is recovered. About 54% of the dry starting material is accounted for. (Based on moisture
content measurement of 78.2% for the wet bio-char)
20 g Gas
619g Condensed Liquid
5.6 g dissolved solid
312 g Wet Biochar 39g dry Biochar (13 g identified)
Monosaccharides
(2.2%)
Polars (0.17%)
CO
CO2
(4%)
(94%)
Other
(2%)
Unidentified
Cations &
Anions
Dry
Bio-Char
Unidentified
(0.02%)
(assumed H2O)
(99.07%)
Ac etic Acid
Dissolved
Solids
0.93%
(0.31%)
Other
Elements
(0.01%)
Unidentified
organics
( 0.34%)
Other
Organic
Acids
(0.09 %)
Moisture in
Biochar
78.2%
Furans
(17.3%)
Lignin
Monomers
Dry Biochar
Identified *
(0.5%)
Hydroxy Acids (4.1%)
Acetic Acid
(1.3%)
Other Polars
(7.2%)
Other
Organic Acids
(0.8%)
* Percentages based on dry biochar: 34.2% identified
Elements
(0.8%)
Figure 4: Chemical analysis of products from HPT processing of Loblolly Pine. 100% of the gaseous
product is identified; 34% of dry bio-char is identified; <1% of the condensed liquid is identified.
Biomass
1.6
• The O content is lowered,
but C content is increased.
To demonstrate the viability of the pre-treatment process, we intend
to use the bio-char as feedstock for a gasifier. DRI is partnering with
UNR’s College of Agriculture, which has acquired a Biomax 15, a
commercial gasifier/ power generation system.
Ultimate analyses of three different raw feedstocks and resulting
biochar produced by the HPT process are summarized below. Note
the increase in energy content and C, and the decrease in O.
Ultimate
Loblolly Pine
Corn Stover
Rice Hulls
Analysis % Feedstock Biochar Feedstock Biochar Feedstock Biochar
C
68.3
48.7
43.2
51.4
43.1
39
5.9
5.3
4.8
H
5.1
4.7
4
N
0.37
0.94
0.4
0.23
0.75
0.26
0.04
0.09
0.06
S
0.03
0.1
0.05
O
25.9
30.7
24
42.1
40.1
35.6
0.39
10.9
20.4
Ash
0.27
14.7
27.9
8511
7207
6650
Dry HV (Btu/lb)
11793
8239
7328
A techno-economic analysis of the pre-treatment process is being
conducted to determine the viability of building a full-scale,
commercial facility in Nevada.
• This analysis incorporates results of the resource assessment
and the mass/energy balances of the pre-treatment process.
• Hydrothermal pre-treatment will be coupled with gasification (for
syngas production) or pyrolysis (for bio-oil production).
• Based upon results of the Nevada biomass resource
assessment, the facility would be located in Eastern Nevada.
This work was performed under a subcontract to the Gas Technology Institute to support
the technical goals of US DOE Cooperative Agreement DE-FG36-01GO11082
DRI Participants : Jay Arnone, Amber Broch, Alan Gertler, Kent Hoekman, Richard Jasoni, Steve
Kohl, Tim Minor, Jeremy Riggle, Curt Robbins, Lycia Ronchetti, Vera Samburova, Dave Sodeman,
Paul Verburg, Barbara Zielinska.
UNR Participants: Chuck Coronella, Victor Vasquez, Wei Yan
REII Participants: Matt Caldwell, Dennis Schuetzle, Greg Tamblyn
Thermogravimetric analysis and devolatilization of wood under nitrogen and steam gas atmospheres.
Igor V. Kolomitsyn, Andriy B. Khotkevych, Donald R. Fosnacht.
Natural Resources Research Institute, University of Minnesota Duluth, Minnesota. 5013 Miller Trunk Hwy., Duluth, MN, 55811.
Percentage by weight, %
Softwood
Hardwood
Hemicelluloses
Xylose
Y = 9.9E(-7)*X + 0.03504039
R2 = 0.999
Y = 1.690E(-6)*X - 0.048736458
R2
Mannose
Y = 9.56E(-7)*X + 0.065521422
R2 = 0.999
Arabinose
Y = 1.017E(-6)*X + 0.032770391
R2 = 0.999
For quantitative analysis, the samples of temperature treated wood were hydrolysed ucording to the procedure [4] and after
chromatographic separation the amount of glucose, xylose, galactose, mannose, arabinose were calculated from the calibration
curve.
Galactoglucomannan (1:1:3)
5-8
Galactoglucomannan (0.1:1:4)
10-15
Glucomannan (1:2 – 1:4)
0
Arabinoglucoroxylan
7-10
15-35
0
250
200
6
2-5
After the exposure time is over, the tube has been placed out of the heater, allowing it to cool down
at ~20 C/min, and then the sample has been unloaded, weighed and analyzed.
Trace
Trace
15-30
40-44
40-44
Lignin
25-35
18-25
Extractives
5-8
2-8
The experiments on devolatilization of various wood samples have been conducted in a stainless steel Fixed Bed tube
reactor. For comparison, some samples have been treated via TGA-like procedure in the specially designed unit. The
products of devolatilization have been maintained using conventional wet lab techniques.
Materials:
All commercial reagents were ACS reagent grade and used without further purification..
330
340
350
360
370
380
°C
The Upscale Thermogravimetric Apparatus (TGA):
The specially designed unit allows to get thermogravimetric plots on the relatively large samples of
material (30 – 50 grams, in case of wood chips).
The sample of wood chips is being loaded as shown, in the Inconel cup, placed on a tip of a longshaft thermocouple. The weight of assembly is being monitored live with 0.02 g accuracy. The
process is typically running at manually adjusted flow rate and at PID-controlled outside wall
temperature. The inside temperature has been monitored separately, using the wireless transmitter
on the top of the thermocouple assembly.
The TGA-tests have been made in nitrogen atmosphere at constant ramping speed within (2.5 – 3.5)
C/min.
8
390
0
6
16
2. Effect of Gas Flow Rate.
100
95
Setup:
Fixed Bed Tube
90
Process Temperature: 300°C
85
Process Time:
Softwood (South Yellow Pine)
Hardwood (Yellow Poplar)
-2
200
250
300
350
400
450
-4
10
-6
0
16
In the typical process conditions the temperature inside of the
sample cup lags within (5 - 7) °C behind the oven. However,
the spontaneous temperature rise takes place when the sample
goes from 300 to 400 °C. The peak on the plot is about 15 °C
high, and is matching the temperature, when decomposition of
the sample is most extensive (see TGA plots at 1.) This effect
confirms the exothermic nature of some reactions, which take
place during pyrolysis of wood biomass.
300
350
20
uRIU
26.408
24.950
23.200
22
24
26
28
30
RID10A
20
20
ivk2809s4 r1 spruce 260C 06-01-2009 10-02-30 PM-Rep3.dat
Retention Time
Spruce T=260 oC
10
0
16
18
10
0
20
22
24
26
28
30
Minutes
Volatiles of spruce wood after treatment at 230 oC.
Rosins
Abundance
1100000
2200000
15.73
15.56
15.46
2000000
1500000
2000000
16.64
16.19
13.30
1000000
For hardwood, the changes of flow rate are effective below GHSV = 50, and
almost no effect has been observed at higher values.
15.77
1400000
RT, min
Name
Phenols, methoxybenzols,
aromatic aldehydes
900000
15.40
15.40
800000
Rosin acids
C18:2
700000
15.73
9.69
C18:0
600000
1200000
1000000
15.38
15.69
13.25
1000000
16.13
Abietic acid
13.62
13.71
16.13
Abietic acid
500000
6.76
5.66
500000
8.67
0
Time-->
7.93
7.38
7.49
10.03
8.98 9.83
12.08
13.66
13.47
8.00
10.00
12.00
14.00
16.00
16.39
8.65
200000
6.00
200000
15.05
400000
8.00
9.00
6.50
8.63
9.43 10.38
12.05
13.43
12.77
13.03
17.38
16.94
6.00
7.00
8.00
9.00
10.00 11.00 12.00 13.00 14.00 15.00 16.00
10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00
Conclusions
3. Effect of the Process Time.
 The decomposition of softwood starts at lower temperature (about 260 oC) compared to the decomposition temperature of
hardwood;
Same, 325 C
85
85
80
80
75
75
Carrier Gas: N2
70
70
GHSV = 200
65
65
60
60
55
55
 The thermochemical reaction of softwood started by the decomposition of the arabinoglucoroxylan in the hemicelluloses.
Setup: Fixed Bed Tube
Softwood (South Yellow Pine)
 Extractibles of spruce wood are separated in the steam atmosphere. Phenols, aromatic aldehydes, fatty acids, fatty alcohols
and non-polar rosins are found in a volatile fraction whereas rosin acids stay in the wood.
 Rosin acids started to decompose in N2 and steam atmosphere at 260 oC.
Hardwood (Yellow Poplar)
10
15
20
25
min. 30
Unlike the case above, the process time effects more on devolatilization of
hardwood, then softwood. The effect of the process time is more evident at lower
temperatures, when pyrolitic processes possess the minor role in overall reaction.
At higher temperatures, when pyrolysis becomes the dominate reaction, the effect of
process time becomes insignificant.
15.00
14.25
17.70
Time-->
7.00
7.38
300000
16.51
Time-->
6.00
15.64
600000
17.00
17.74
18.31
10.19
400000
800000
14.83
14.11
15.07
6.24
5.67
30
min.
0
Abundance
Abundance
In case of softwood, the process of devolatilization shows major dependence on a
gas flow rate. Probably, the volatile components of a softwood have a trend to recondensation or recombination at a solid matrix. Increase of the flow rate
(dilution of the gas media with inert carrier gas) makes this process slow down.
25
18
20
10
TIC: 2209S2B.D
1600000
20
Spruce T=245 oC
Extract of spruce wood after treatment at 230 oC.
1800000
15
30
TIC: 2209S4A.D
* The “Zero GHSV” points were obtained in the TGA arrangement.
10
28
6. GC/MS profile of extractibles of softwood .
TIC: I1509S4A.D
250
26
Minutes
N2, GHSV*
200
Retention Time
20
65
150
24
ivk2809s3 r1 spruce 245C 06-01-2009 07-57-15 PM-Rep2.dat
500
Extract of spruce wood.
100
22
RID10A
Oven, C
0
70
50
20
10 min.
80
75
0
Minutes
-8
The yield of condensable volatile products shows some complex temperature
dependence with a local hump at (340 – 350) °C. At this point, the pyrolysis
reactions begin to dominate in the overall process. This can also confirmed by
comparison of the composition of products.
18
10
2
Surprisingly, both hardwood and softwood samples in the tube reactor (10 min.,
N2, GHSV = 200) show much more weight loss, compared to the same in TGA
arrangement. This difference is more evident at temperatures below 350 °C. At
higher temperatures the weight plots are going to be similar, no matter what
arrangement is employed.
dichloromethane to determine the amount and composition of the volatile matter.
TC
Experimental.
320
Hardwood (Yellow Poplar)
Solid matter, % : T = 305 C
)
310
4
The charcoal from the trap, alongside with the fiber plug have been and extracted in Soxlett with
Glucoronoxylan
300
20
Thermochemical
reaction
of
softwood (White Spruce) started by
hemicelluloses decomposition at 260
oC. In the fixed bed tube reactor the
concentration of arabinose is
decreased to an undetectable level.
This data strongly support the idea
that thermochemical reaction of
hemicelluloses occurred by the
arabinoglucoroxylan decomposition.
uRIU
Softwood (Red Pine)
290
30
26.408
500
21.367
°C
24.942
450
24.942
400
10
23.192
350
0
28
Spruce T=230 oC
23.192
300
Rise, C
10
26
ivk2809s2 r1 spruce 230C 06-01-2009 06-23-21 PM-Rep2.dat
Retention Time
21.367
250
24
12
2
20
200
22
RID10A
500
20
4
30
400
20
Minutes
uRIU
40
300
10
0
18
Hardwood (Aspen)
200
8
16
Softwood (Tamarack)
Oven, C
10
Tube
Fixed Bed Tube Reactor:
0
0
Mannose
uRIU
12
0
The 12” long ½” ID reaction tube is connected to the gas/steam supply on the top, and to the
charcoal trap on a bottom. The sample of wood chips is being loaded as shown, over the glass fiber
plug. The process runs at manually adjusted flow rate and at PID-controlled outside tube
temperature. The inside temperature has been recorded with a separate device. The work pressure
has been kept within (6 – 8) psig – as low as needed to maintain selected flow rate.
19.858
300
14
19.850
TGA
TC
N2
uRIU
16
80
60
25-30
Cellulose (
350
Solid matter, %
= 0.999
10
uRIU
R2 = 0.999
21.375
18
90
Galactose
20
26.375
Y = 9.13E(-7)*X + 0.026891794
Ramping speed: 3.3 C/min.
Arabinose
Xylose
Retention Time
20
400
20
Chromatograms of solid residue of
thermally treated softwood.
ivk2809s1 r1 spruce wood 06-01-2009 04-49-27 PM-Rep2.dat
23.192
Glucose
Setup: Upscale TGA
Load of material: 35 – 45 g
Condensable matter, % (Tube)
100
Softwood profile (Spruce)
RID10A
21.367
HPLC analysis of cellulose and hemicelluloses:
Galactose
Chemicals
Solid matter, %
50
The analysis was performed on a Shimadzu (Shimadzu Scientific Instruments, Inc., Columbia, MD, U.S.A.) liquid
chromatographic system consisting of a Model SCL-10Avp system controller, a Model DGU-14A on-line Degasser, a Model
LC-10ATvp HPLC pump, a Model FCV-10ALvp Low-pressure Gradient Flow Control Valve, a Model SIL – 20A auto
sampler, a Model RID-20A refractive index detector, and a Model CTO-20A column oven. For data acquisition and analysis
the Shimadzu EZStart Ver. 7.2.SP1 was used. The chromatographic column utilized was VA 300/7.8 Nucleogel Sugar Pd2+,
(Macherey-Nagel Inc., Cat # 719530.) Elution was carried out in the isocratic mode at a flow-rate of 0.4 mL/min. HPLC grade
water was used a suitable mobile phase. Elution time was 30 min; column temperature was 80 oC, and injection volume was 50
μL.
Sample, C
500
450
60
Carbohydrate analysis by HPLC was performed by modified procedure from ASTM 1758-01 [4]. Stock solution (5.1 mg/mL)
of reference compounds (glucose, xylose, galactose, mannose, and arabinose) was prepared in water (HPLC grade). Dilutions
were obtained in water to afford the concentration range 0.4 mg/mL to 5.1 mg/mL. The standard solution was injected in
triplicate and the curve was constructed using the average values of the detector response. Calibration curve was as follow (X –
peak area; Y – concentration of monosaccharide (mg/ml)):
Table 1. Major component of wood. [1]
1. Solid and volatile matter at variable process temperature.
19.850
The purpose of this investigation is to develop a pretreatment regimes for various lignocellulosic materials, that allows
more easy access to the individual cellulosic carbohydrates. This carbohydrates may be further converted to liquid fuels
by either bio or thermal conversion methods.
Glucose
70
HPLC chromatographic conditions:
4. Autothermal effects.
19.850
Introduction.
GC/MS analysis was performed using a Hewlett Packard Gas Chromatograph Model 5890, which was equipped with Hewlett
Packard Mass Selective Detector 5970A, and capillary column (Optima-1 12 m x 0.2 mm with film thickness 0.2 mm;
Macherey-Nagel Inc., Cat # 726834.12. ). The program conditions that were used are as follows: column was kept 1 min at
80oC and then heated from 80 oC to 250 oC with a heating rate of 10oC/min, then kept at 250 oC for 10 min. Injector
temperature was held at 300 oC and detector temperature was held at 300oC. Solvent delay: 5 min. Head pressure was 7 psi.
Carrier gas: He. Injection volume: 1 ml. All samples before injection were methylated using a solution of CH2N2 in ether [2]
and then silylated with BSTFA [3].
5. Composition of solid residues of softwood at variable process temperature.
uRIU
Thermogravimetric plots have been measured in range from 393 K to 1023 K for several hard woods (yellow poplar,
aspen) and soft woods (red pine and spruce) under nitrogen gas atmosphere on a 50 g scale. The typical ramping speed
was (2 – 5) K/min. Temperatures for thermal events for each species were recorded. A comparison between hard
woods and soft woods shows that, in the latter case, the decomposition starts at lower temperature. A tubular fixed bed
reactor was used to investigate each thermal event under nitrogen and steam gas atmospheres at the constant
temperature settings. Each sample before thermal treatment was dried at 373 K for 24 hrs. Volatiles and wood
extractives before and after thermal treatment were analyzed using gas chromatography mass spectroscopy (GC/MS)
technique. It was found that at 503 K – 543 K under steam gas atmosphere, extractible phenols, fatty alcohols, and
fatty acids were accumulated in the volatile fraction. The solid residue after thermal treatment was also analyzed. The
concentration of D(+)-xylose, D(+)-mannose, L(+)-arabinose, D(+)-glucose, D(+)-galactose in each sample before and
after thermal treatment at various temperatures was measured by high performance liquid chromatography (HPLC)
technique. These data are used to estimate the concentration of cellulose and hemicellulose in wood samples. The
purpose of this investigation is to develop a lignocellulosic pre-treatment regime that allows more easy access of the
cellulosic sugars for conversion of the materials to liquid fuel by either bio or thermal conversion methods.
26.408
Results and Discussion.
Gas chromatography-mass spectroscopy (GC/MS) method:
uRIU
Abstract.
References
1. Amidon, T. E.; Liu, S., Water-based woody biorefinery. Biotechnology Advances 2009, 27, (5), 542-550.
2. Fieser, L. F.; Fieser, M. Diazomethane. In: Reagents for Organic Synthesis. New-York, John Wiley & Sons, Inc, 1967, p. 191.
3. Supelco. Guide to Derivatization Reagents for GC. Bulletin 909.
4. Standard Test Method for Determination of Carbohydrates in Biomass by High Performance Liquid Chromatography. ASTM
International, E 1758-01. 2007.
INTEGRATED HEAT, ELECTRICITY AND
BIO-OIL PRODUCTION
J. Lehtoa,*, P. Jokelab, Y. Solantaustac, A. Oasmaac
Power, Kelloportinkatu 1 D, PO Box 109, FI-33101 Tampere, Finland
bUPM, Eteläesplanadi 2, PO Box 380, FI-00101 Helsinki, Finland
cVTT, Biologinkuja 5, PO Box 1000, FI-02044 VTT, Finland
*Corresponding author. Tel: + 358 20 14121 Fax: + 358 20 1412 210
E-mail: jani.lehto@metso.com
aMetso
FUEL CHAIN
WORLD’S FIRST INTEGRATED PYROLYSIS PLANT
Metso has built the world’s first integrated pyrolysis pilot
plant in Finland, in co-operation with UPM and VTT. The
related concept covers the entire business chain, from
feedstock purchase and pre-treatment to bio-oil production,
transportation, storage and end use. This project is partly
funded by TEKES, the Finnish Funding Agency for
Technology and Innovation. Integrated pyrolysis pilot plant
is now in operation.
UPM is among the most important users of wood-based raw
materials in Finland. The company plans to exploit the
potential of several commercial pyrolysis plants in terms of
bio-oil production, for its own use as well as for sale to the
market, through current and future boiler investments.
Metso will be able to market pyrolysis solutions to third
parties in the global market. The construction of a
commercial-scale demonstration plant will be planned
based on the results and experiences garnered from the
test runs in 2009 and 2010.
Feedstock processing,
transporting, feeding
Pyrolysis liquid production,
solids removal by centrifugation
Pyrolysis liquid combustion
for CHP in boilers
Forest residue, stumps
On-line moisture
analysis
Gas on-line monitoring
On-line analyses for
water and solids
Water max 28 wt-%
solids <0.05 wt-%
single-phase liquid
QUALITY CONTROL
Standard analyses and novel on-line methods will be used through the
quality control chain
INTEGRATION REDUCES THE COSTS
A fast pyrolysis unit can be integrated with a fluidized bed
boiler. Based on such a concept, the pyrolysis unit utilizes
the hot sand in the fluidized bed boiler as a heat source.
The devolatilized gas compounds are condensed into bio
oil and the remaining solids, including sand and fuel char,
returned to the fluidized bed boiler. In the boiler, the
remaining fuel char and non-condensable gases are
combusted to produce heat and electricity.
ON-LINE QUALITY CONTROL IN USE
Quality follow-up along the entire chain from biomass
processing via pyrolysis to oil use, will both ensure the
production of a consistently high-quality product and help in
avoiding possible problems during production. Standard
and novel on-line methods will be used and further
developed.
A 2 MW fuel fast pyrolysis unit has been integrated with Metso’s 4 MWth
circulating fluidized bed boiler, located at Metso’s R&D Center in Tampere
FIELD TESTS
CONCEPT
FOR
VERIFICATION
THE
BOILER
UPM’s focus is on using bio-oil as a substitute for light and
heavy fuel oil in heating and combined heat and power
plants. Oilon is currently developing a new burner for
pyrolysis oils, to be tested in Finland in 2009.