"a comprehensive model of biomass pyrolysis" final

"A COMPREHENSIVE MODEL OF BIOMASS PYROLYSIS" Phase I

FINAL REPORT

For the Period

May 15, 1996 - February 12, 1997

Award Number: 96-33610-2675

Sponsored by:

United States Department Of Agriculture

Small Business Innovation Research Program

Program Official: Charles F. Cleland

Director, SBIR Program

Michael A. Serio, Principal Investigator

Marek A. W6jtowicz, Co-Principal Investigator

Yonggang Chen

Sylvie Charpenay

Anker Jensen *

Rosemary Bassilakis

Kerry Riek

ADVANCED FUEL RESEARCH, INC.

87 Church St., East Hartford, CT 06108

Phone: (860)528-9806

(860)528-0648

Fax: E-mail: [email protected]

http://www.online-ftir.com/afr.htm

*Department of Chemical Engineering

Technical University of Denmark

2800 Lyngby, Denmark

DOA-523001 Final report

2112197

1. PROJECT SUMMARY

1.1 Purpose of the Research - The objective of this work is to adapt an eXlstmg coal pyrolysis model and make it suitable for the pyrolysis of biomass. The soundness of this approach is based on numerous similarities between biomass and coal. There are important differences, however, which preclude direct application of the coal model. 1.2 Description of Work - This work involved: 1) selection of a set of materials representing the main types of biomass; 2) development of a biomass classification scheme; 3) development of a modeling approach based on modifications of a coal pyrolysis model; 4) calibration of the model for a set of standard materials against pyrolysis data taken over a range of heating rates; 5) validation of the model against pyrolysis data taken under other (higher) heating rate conditions. 1.3 Research Results - A number of biomass samples were considered based on the abundance and availability of agricultural and forestry feedstocks and waste materials in the United States. Six samples were obtained from the National Institute of Standards and Technology'S Standard Reference Materials Program which included microcrystalline cellulose (C), sugar cane bagasse (B), wheat straw (WS), com stalk (CS), softwood Pinus radiata (PR), and hardwood Populus de Itoides (PD). These six samples were supplemented with two samples of the pure component materials: lignin (L) and hemicellulose (HC). To understand the chemical structure of these feedstocks, the starting materials were characterized by several methods including trace elemental analysis, ultimate analysis, proximate analysis, Field Ionization Mass Spectrometry (FIMS), solvent extraction and solvent swelling. This information was complemented by a compilation of relevant literature data. In addition pyrolysis kinetic information was obtained for overall weight loss and the evaluation of individual volatile species using a TG-FTIR system. A biomass classification scheme was developed based on a van Krevelen diagram (plot of HlC versus OIC atomic ratio). The kinetic and composition files for the FG-DVC model were determined using a

close neighbor approximation, in which the properties of a given biomass material are interpolated among the corresponding properties of its neighbors (reference samples) in the van Krevelen diagram. The use of this approximation was supported by similarities in the pyrolysis results from the TG-FTIR and FIMS experiments. The kinetic parameters for individual gas species were determined based on a first order reaction with a distribution of activation energies. Usually, only a single functional group pool was used to describe the evolutions of the major volatile species. Model predictions were compared with the yields and composition of pyrolysis products obtained for non­ reference samples and samples pyrolyzed under significantly different heating conditions. In general, good agreement was obtained. 1.4 Potential Applications - Over the next decade there will be a renewed emphasis on the production of chemicals and liquid fuels from biomass, the use of agricultural wastes as feedstocks, and the development of crops and farming techniques optimized for fuel production. In view of the tremendous diversity of biomass feedstocks, a great need exists for a robust, comprehensive model that could be· utilized to predict yields, composition, and properties of pyrolysis products as a function of feedstock characteristics and process conditions. The commercial applications of the proposed model involve biomass process design, development, and scale-up. Since pyrolysis constitutes an jmportant step in most biomass-conversion processes, the market for the product would be extensive. Most engineering firms use sophisticated hydrodynamic models for reactor design, and our pyrolysis model will be incorporated into this type of code (FLUENT) as a submodel.

2

2112/97

DOA·523001 Final repon

A COMPREHENSIVE MODEL OF BIOMASS PYROLYSIS TABLE OF CONTENTS I. PROJECT SUMMARy .................................................................................................................2

. 2. EXECUTIVE SUMMARy ............................................................................................................4

2.1 2.2 2.3 2.4

Statement of the Problem or Opportunity ............................................................................. .4

The Innovations .....................................................................................................................4

Phase I Technical Objectives ..................................................................................................5

Summary of Phase I Accomplishments .................................................................................. .5

3. DEGREE TO WHICH PHASE I OBJECTIVES HAVE BEEN MET ...............................................6

3.1 3.2 3.3 3.4 3.5

Introduction ...........................................................................................................................6

Background........................................................................................................................... 10

Phase I Results ...................................................................................................................... 15

Estimate Of Technical Feasibility ..........................................................................................3 5

Explanation of Cost Data .....................................................................................................3 5

4. POTENTIAL APPLICATIONS ..................................................................................................3 5

4.1 4.2 4.3 4.4

Technical, Economic, and Social Benefits ..............................................................................3 5

Comparison of Proposed Approach to Existing Technology ................................................. 3 6

Outline of Phase II .............................................................................................................. .3 6

Potential for Transition to Phase III ................................................................................... .3 7

5. ACKNOWLEDG·MENTS........................................................................................................... .3 7

6. RECIPIENTS OF REPORT ....................................................................................................... .3 7

7. REFERENCES ........................................................................................................................... .38

8. APPENDIX................................................................................................................................41

3

DOA-52300 I Final report

2112/97

2. EXECUTIVESUNrndARY 2.1

Statement of the Problem or Opportunity

The future outlook for biomass pyrolysis is quite promising, despite the fact that petroleum prices should remain low for at least the next decade [1]. The political and environmental benefits of using biomass will continue to provide impetus to the development of biomass pyrolysis processes (indigenous supply, low sulfur, no net CO2 , biodegradable, etc.). The current level of activity is high (about 650 activities were identified in a recent study [2]), and international in scope_ Over the next several years there will be a renewed emphasis on the production of chemicals from biomass with minimal upgrading, and on the use of agricultural wastes as feedstocks_ Research and development activities will also continue on the production of liquid fuels and chemicals from biomass, and on the development of crops and farming techniques optimized for fuel or chemical production. In view of the anticipated rapid development of pyrolysis-based processes, and because of the tremendous diversity of biomass feedstocks, a great need exists for a robust, comprehensive model that could be utilized to predict yields, composition and properties of pyrolysis product. Such a model would constitute a valuable tool in process development and scale-up. No such model is currently available. The specific problem addressed in this study was related to the current inability to predict a prwrl gas, liquid, and solid products on the basis of known feedstock characteristics and process conditions (temperature, pressure, heating rate, residence time, etc.). This makes it difficult, of course, to design a process in such a way so that the most desirable slate of products is manufactured. This, in turn, leads to appreciable, and often unnecessary, costs of product upgrading. The model, the feasibility of which was demonstrated in Phase I, will greatly alleviate all the above difficulties.

2.2 The Innovations

The innovation of this project involves modifYing the currently available advanced coal devolatilization model, and making it suitable for modeling pyrolysis of agricultural and forestry feedstocks. The soundness of this approach is based on numerous similarities between biomass and coal, such as the fact that coal was formed from biomass, and that coal can actually be viewed as petrified biomass. There are important differences, however, which preclude direct application of the coal model to pyrolysis of biomass, and which made it necessary to create a separate model dedicated to biomass conversion. As a result of work done over the past ten years on coal pyrolysis, Advanced Fuel Research, Inc. (AFR) has unique expertise and capabilities in experiments and models related to the pyrolysis of macromolecular and polymeric materials [3-21]. The first capability is the development of a general model for coal pyrolysis that combines a functional group (FG) model for gas evolution and a statistical depolymerization, vaporization, and cross-linking (DVC) model for tar and char formation [11,12]. The FG model currently predicts gas yields from the functional group sources in the coal using rank-dependent kinetics. The DVC model simulates coal as a macromolecular network and includes the processes of depolymerization, cross-linking, and transport. The model accurately predicts volatile yields, extract yields, cross-link densities, fluidity, and tar (oil) molecular weight distributions. The variations with pressure, devolatilization temperature, rank, and heating rate are also accurately reproduced. The second important capability is the fact that AFR has developed a novel system for performing thermogravimetric analysis combined with concurrent FTIR analysis of evolved gases (TG-FTIR). The instrument simultaneously measures the weight loss of a solid during heating, and the composition of the volatile species using FT-IR transmission spectroscopy [22,23]. This information on volatile species evolution is important in modeling the solid-phase transformations 4

DOA-52300 1 Final

'"PO"

2112/97

which involve gas evolution and cross-linking. The gas evolution also affects the porosity of the solid, which has a significant effect on the physical properties of the materials. The ability to have these measurements done in a single instrument will provide a rich data set for model development and validation. 2.3 Phase I Technical Objectives The objective of the Phase I program was to create a predictive biomass-pyrolysis model incorporating a detailed understanding of the physical and chemical processes occurring during pyrolysis of agricultural feedstocks. Key to this research was the application of measurement and modeling techniques previously developed to study the thermal decomposition of coal to the thermal decomposition of biomass. This was accomplished in the following three tasks. Selection and Chara'C1erization of the Starting Materials - Six biomass samples were selected for the project. To understand the chemical structure of these feedstocks, the starting materials were characterized by several methods including trace elemental analysis, ultimate analysis, proximate analysis, Field Ionization Mass Spectrometry (FIMS, at SRI International), solvent extract and solvent swelling, as well as a literature review. Model Development - In this task, the TO-FTIR method was used to perform a series of pyrolysis experiments at different heating rates with the samples selected in Task 1. This provided mechanistic and kinetic data for the FG-DVC model. Volatile products (gas and tar) were fully characterized by on-line FT-IR measurements of gas concentrations in the effluent stream. Model Validation - Model predictions were compared with the yields and composition of pyrolysis products obtained under different conditions from those used to determine model parameters (e.g., different heating rates). Selected literature data were used for model validation. It is expected that the overall results of Phases I and II will: (1) establish procedures for the quantitative analysis of biomass by FT-IR, TG-FTIR, and FIMS; (2) produce a model relating the yields and composition of pyrolysis products to the characteristics of biomass feedstocks; (3) provide predictive capabilities for the development and scale-up of biomass-conversion processes, and for better tailoring of such processes to favor a particular distribution of products. 2.4 Summary of Phase I Accomplishments The project objectives were accomplished in three tasks: (1) selection and characterization of several materials representing the main types of biomass; (2) model development based on laboratory measurements; (3) model validation.

The Phase I research has demonstrated the feasibility of applying AFR's modeling and characterization approach developed for coal-pyrolysis processes to create a robust, comprehensible biomass-pyrolysis model. Model predictions are based on feedstock characteristics and process conditions, such as temperature, pressure, heating rate, etc. The specific accomplishments of the Phase I program are as follows: • Selected samples based on literature review of representative biomass materials • Compiled literature data on composition analysis of these samples • Subjected samples to additional analyses (TG-FTIR, solvent swelling, solvent extraction, FIMS, elemental analysis, trace element analysis, ultimate analysis, and proximate analysis) • Developed a biomass classification scheme • Developed a modeling approach based on modification of a coal pyrolysis model. • Calibrated the model for a set of standard materials against pyrolysis data taken over a range of heating rates. • Validated model using pyrolysis data taken under other (high heating rate) conditions 5

DOA-523001 Final report

2112/97

3. DEGREE TO WmCH PHASE I OBJECTIVES HAVE BEEN MET

3.1 Introduction Th~

term biomass refers to material of terrestrial plant origin. The amount of biomass produced has been estimated to be 170 billion tons per year, of which about 70% is from forests [24]. Consequently, wood is the most important biomass component which could potentially be diverted to other uses. Other materials which are available in large quantities include agricultural residues such as cereal straws and cornstalks. An estimate of the amount of biomass available in the U.S. on an annual basis for conversion to fuels or chemicals is 2 billion tons [25]. The conversion of 20% of this material would provide an energy equivalent of 6.5 x 10 15 BTU, roughly 10% of the U.S. annual energy needs [24]. Wood is a mixture of three groups of polymers, cellulose, hemicellulose, and lignin. Representative structures are shown in Figure 1. Cellulose is the largest component (-45% of the dry weight) and consists of an ordered -array of high molecular weight glucose polymer chains. Hemicellulose is a disordered array of several sugar polymers which represents 20-25% of the dry weight of wood. Lignin (20-25%) is a complex amorphous polyphenol polymer which serves as a binder for the cellulose fibers. Wood also contains materials which can be extracted with organic solvents. The amount and composition of the extractives depends on the type of tree and the part of the tree from which the wood has been derived. The extractives can vary from 5-25% of the dry weight of the wood. Some of the earliest commercial products obtained from wood were extractives (pitch and resin) obtained from tapping pine trees [26]. These were called naval stores since they were commonly used for the treatment of wooden ships. The relative amounts of the principal biomass components is also dependent on the type of wood (hardwood or softwood). Other plant materials contain these same major components but in different proportions. A summary of the approximate chemical composition of selected biomass resources is given in Table 1 [27]. Table 1. Approximate Chemical Composition of Selected Biomass Resources, wt% [27]

Compounds Glucose Xylose Mannose Galactose Arabinose Glucuronic acid Acetyl groups Lignins Crude protein Ash Water souble extractives

Wood of Hardwoods 45-55 15-25 0.5-3.0 0.3-1.0 0.3-0.5 2.0-5.0 2.0-4.0 19-28 0.5-0.7 2.0-5.0

Wood of Softwoods 45-55 5.0-7.0

Wheat Straw

Newspaper

1.0-14 0.5-1.5 2.0-4.0 1.0-1.5 27-34

40 21 1.0-2.0 1.7 1.0-2.0 1.5 2.2 18

42-52 4.9-5.3 4.9-6.2 0.5-1.0 0.6-1.1

Office Paper 71-75 6.5-8.9 2.7 0 0

25.5

0.5

0.5-0.7 2.0-5.0

8.0 8.0-12

0.5-3.5

7.7-15

10-12

Corn Kernel 0.07-0.17 6.2

Cassava Roots 82-93 0.1-1.1

1.0 4.2 0.8 0.2 8-14 1.1-3.9

2.1-6.2 0.9-24

The pyrolysis of biomass has been the subject of numerous studies as summarized in review articles [28-33] and collected papers [34-36]. Biomass pyrolysis is similar to coal pyrolysis in that both produce a mixture of char, tar and gases. However, it is an oversimplification to suggest that pyrolysis of biomass is similar to pyrolysis of a very young coal. One of the principal differences is that coal is predominantly an aromatic material while in biomass the aromatic component (lignin) is a relatively minor constituent (-20%) when compared to the cellulose and hemicellulose constituents. In addition, because of the fossil nature of coal, the mineral matter constituents which have been incorporated influence the pyrolysis behavior in a different way than for the case of

6

2112/97

DOA-S23001 Final TeJlOI1

0",",-

Hemicellulose (partial structure)

R = CHO or CH2 0H R' = OH or OC ""'"

OH

R'

Lignin (partial structure) Figure 1. Principal polymeric components of plant materials [37]: (a) cellulose; (b) hemicellulose (partial structure); (c) lignin (partial structure). R = CHO or CH20H; R' = OH or OC-.

7

DOA-S23001 Final report

2112/97

biomass. Besides the lower aromatic content, an important chemical difference between biomass and coal is a much higher oxygen content for the former. The oxygen is present as ether, hydroxyl, carboxyl, aldehyde, and ketone functionalities which decompose to produce oxygenated gases upon pyrolysis (CO, CO2, H 20) in yields similar to those produced from low rank coals (5-10 dry wt.% for CO2 and H 20, 5-15 dry wt. % for CO). However, the tar (liquid) yields are much higher than those produced from low-rank coals (40-50% versus 10-20% on a dry basis) The increased tar yield comes primarily at the expense of char which is much lower for biomass « 10%) than for low rank coals (40-50%). Apparently, the depolymerization of biomass is the predominant pyrolytic reaction while for coal the depolymerization reactions compete with cross-linking reactions, thus producing the high yields of char. Most of the char formed from biomass pyrolysis is derived from the lignin component which is closest to low rank coal in its chemical composition. An approximation which is sometimes made is that the three main components of biomass (cellulose, hemicellulose, lignin) behave independently during pyrolysis [38]. However, it has been shown in our work and in the literature that the yields cannot always be predicted based on a knowledge of the pure component behavior. The tars produced from biomass pyrolysis consist of a mixture of oxygenates in the case of primary tars produced at low temperature or flash heating conditions. As the extent of secondary reactions is increased by exposing the primary volatiles to higher temperatures and/or longer residence times, the tar cracks to produce additional gases (mainly CO) and the tar composition becomes less oxygenated and more aromatic. These changes in tar composition have been summarized in Table 2 from reference [39]. As the severity of the secondary reactions increases further, the polycyclic aromatics produced eventually lead to soot formation. A summary of the global mechanisms for primary and secondary pyrolysis of biomass by Evans and Milne [40] is shown in Figure 2. Table 2. Chemical Components in Biomass Tars [39]. Conventional Flash Pyrolysis 4S0-S00°C Acids Aldehydes Ketones Furans Alcohols Complex Oxygenates Phenols Guaiacols Syringols Complex Phenolics

High-Temperature Conventional Flash Pyrolysis Steam Gasification 600-6S0°C 700-800°C Benzenes Naphthalenes Acenaphthylenes Phenols Catechols Fluorenes Naphthalenes Phenanthrenes Biphenyls Benzaldehydes Phenanthrenes Phenols Naphthofurans Benzofurans Benzaldehydes Benzanthracenes

High-Temperature Steam Gasification 900-1000°C Naphthalene Acenaphthylene Phenanthrene Fluoranthene Pyrene Acephenanthrylene Benzanthracenes Benzopyrenes 226 MW PAHs 276 MW PAHs

As in the case of coal, the yield and distribution of products from the pyrolysis of biomass depends on other variables besides the final temperature and holding time. These include heating rate, total pressure, ambient gas composition, and the presence or absence of catalysts [41-43]. An overview of these effects has been given by Shafizadeh [44]. The trends are similar to those observed for coal, i.e., higher yields of volatiles are observed at higher heating rates, lower pressures, and in the presence of hydrogen. These trends can be interpreted in terms of whether the process variable favors secondary cracking or recombination reactions.

8

DOA-S23001 Final report

2112197

Primary Oils Vapors and Gases

H 20, C02 CO

.

•

Olefins, Aromatics CO, C02' H 2 , H 20 ~

o~

l

j

~fIl

~

Liquids

~

fIl

£

H ~

PNA CO, H 2, CO, H 2, CO2, ' ' ' CO2, H 20 H2 0 ,CH4

1

...

--'"

~e

10

Tertiary Oils (Tars)

Secondary Oils



High

Primary Oil Liquids

~

..

Pressure

Condensed Oils

(Phenols,

Aromatics)

j

r Solids

Biomass

-

Charcoal

*

Coke

Pyrolysis Severity

* • Soot

~

Figure 2. Biomass pyrolysis global mechanisms [40].

DOA-523001 Final report

2112/97

The modeling of the physiochemical processes occurring III the pyrolysis of wood has been undertaken primarily by the fire research community. A comprehensive review of modeling approaches to the chemical and physical processes occurring during biomass pyrolysis has recently been published [45]. Other reviews are available which focus more on the modeling of chemical kinetics [28]. Examples of some of the modeling approaches that have been used can be found in references [34-36]. Unlike in the case of coal, no comprehensive model of whole biomass pyrolysis has yet been developed. 3.2 Background Modeling of Biomass Pyrolysis - Historically, the quantitative modeling of biomass pyrolysis has proceeded along similar lines as for coal [1]. One approach is based on the approximation that the three main components of biomass (cellulose, hemicellulose, and lignin) behave independently during pyrolysis (see, e.g., Nunn et al. in reference [34]). Consequently, yields can be predicted based on a knowledge of the pure component behavior. The shortcoming of this technique is that it cannot account for possible interactions between the biomass components. A more phenomenological approach was successfully used by Serio et al. [3] in modeling the pyrolysis of lignin. Their model was focused on observed evolution rates of major products rather than on the many individual reactions leading to the evolution of each· product. The model of Serio et at. [3] and that of Petrocelli and Klein (discussed in reference [34]) involve a statistical treatment of lignin pyrolysis reactions.

Within the limited scope of the Phase I project, the network features of the FG-DVC model and similar models (such as the one proposed by Petrocelli and Klein) were de-emphasized in favor of the originalfunctional group (FG) description of gas and tar evolution [4,5/. While this is a more empirical approach, there is flexibility in the model to add more details of the whole biomass pyrolysis chemistry as these become known and can be represented in a tractable mathematical expression. In Phase II, the network approach (Le., a DVC-like model) is expected to provide more accurate predictions of the serial reactions that appear to take place during biomass pyrolysis. Statistical Network Models - The important processes in the early stages of pyrolysis are polymerization/depolymerization, cross-linking and gas formation, and these early processes determine the composition of the products [46,47]. The geometrical structure of a polymer (whether it is chain like or highly cross-linked) controls how it reacts under otherwise identical chemical reactions. One, therefore, requires statistical models based on the geometrical structure to predict the reactions of a polymer. At AFR, we have developed such statistical models to describe the thermal decomposition of coal. Many recent studies have proposed that coal can be thought of as having a macromolecular network structure to which concepts of cross-linked polymers can be applied [46-55]. These concepts have been employed tp understand and model such properties of coal as: i) the insolubility; ii) the equilibrium swelling and penetration of solvents; iii) the viscoelastic properties; iv) similarities between the parent coal and products of hydrogenolysis, or mild oxidation; v) cross-linking during char formation [56,57]; and vi) the formation of coal tar in pyrolysis [4,6,7]. With the success of these concepts in describing coal properties, it is logical to extend macromolecular network concepts to describe completely the coal thermal decomposition behavior. This has been done at AFR by applying statistical methods to predict how the network behaves when subjected to thermally induced bond breaking, cross-linking, and mass transport processes [8-13]. The general model developed at AFR to describe coal thermal decomposition is called the FG-DVC model. It was described in several publications [11,12]. In developing the model, extensive experimental work was done with synthetic polymers to allow the study of bond breaking and mass 10

DOA·S2)OOI Final report

2112197

transport in chemically clean systems [4,8,9,14-16]. The model combines two previously developed models, a functional group (FG) model [4,5,17] and a depolymerization-vaporization-cross-linking (DVC) model [8,9-16]. The DVC subroutine is employed to determine the amount and molecular weight of macromolecular fragments. The lightest of these fragments evolves as tar. The FG subroutine is used to describe the gas evolution and the elemental and functional group compositions of the tar and char. Cross-linking in the DVC subroutine is computed by assuming that this event is correlated with CO2 and CH4 evolutions predicted in the FG subroutine. The yield of rapidly released CO2 (which is related to coal rank and weathering) is the factor that controls the thermosetting or thermoplastic behavior of coals. For coals which exhibit thermoplastic behavior, the fluidity is limited by the cross-linking associated with the evolution of methyl groups. The FG Model - The Functional Group (FG) model has been described in a number of publications [4,5,18]. It permits the detailed prediction of the composition of volatile species (gas yield, tar yield and tar functional group and elemental composition) and of char (elemental and functional group composition). The original version employed coal-independent rates for the decomposition of individual functional groups in the coal and char to produce gas species. The approach was successful in predicting the evolution profiles and amounts of the volatile products [12,19]. More recently, it was found that rank-dependent kinetic rates for bridge breaking and cross-linking were required to make accurate predictions of coal fluidity, as discussed in references [20] and [21]. The FG model is based on the parallel, independent evolution of functional groups, and first-order reactions are used along with a distribution of activation energies for each evolving species. Functional groups can also be released from the coal molecule attached to molecular fragments which evolve as tar. (This is the main feature that differentiates the FG model from other parallel-reaction models.) Systematic, rank-dependent kinetics for functional-group decomposition are used, and this kinetic information is derived from TG-FTIR data. The model predictions have been compared to experiments on the pyrolysis of a Pittsburgh Seam coal [5,56,58,59], a North Dakota (Beulah, Zap) lignite [5,10] and the Argonne Premium coal samples [19,20]. In all cases, good agreement between model predictions and experimental data was found. The DVe Model - A simple example of the DVC model is shown in Figure 3. In this model, the parent coal is represented as a two-dimensional network of monomers (condensed ring clusters) linked by strong ("non-breakable") and weak ("breakable") bridges. The latter labile bridges are hydrogen providers. The monomers are represented by circles with molecular weights shown in each circle. The molecular weight distribution of the monomers is assumed to be Gaussian. The monomers are linked to form unbranched oligomers by breakable and non-breakable bridges (shown as horizontal single or double lines, respectively in Figure 3a). Cross-links are added (as vertical double lines in Figure 3a) to connect the oligomers so that the molecular weight between cross-links corresponds to the value reported in the literature for coals of similar rank [60]. The cross-links form the branch points in the macromolecule. The yield of unconnected "guest" molecules (the extracts) depends on the cross-link density. A histogram showing the molecular-weight distribution that was created by randomly picking monomers to form oligomers and randomly cross-linking them is presented at the right of Figure 3a. The distribution is divided into a pyridine soluble portion below 3000 amu (light shading), and a pyridine insoluble portion above 3000 amu (dark shading). The independent parameters of the DVC model are: the starting fraction of unbreakable bonds, the starting cross-link density, and the fraction of bonds (labile or unbreakable) per monomer.

11

DOA·523001 Final report

2/12/97

a. Starting Polymer 20 Pyridine Soluble

Pyridine Insoluble

O'---+--t="'-f-'=~""""-"''I-'''-f-'----J

50

4050

Molecular Weight (AMU)

h. During Tar Formation

Tar

Pyridine Soluble

Pyridine Insoluble

4050 Molecular Weight (AMU)

c. Char Formed Tar

Pyridine Soluble

Pyridine Insoluble

4050 Molecular Weight (AMU) T a r " Char

I'R~S:I

Char

@

Figure 3. Representation of a coal molecule in the DVe simulation and corresponding molecular­ weight distribution. In the molecule, the circles represent monomers (ring clusters and peripheral groups). The molecular weight shown by the number is the molecular weight of the monomer including the attached bridges. The single-line bridges are breakable and can donate hydrogen. The double-line bridges are unbreakable and do not donate hydrogen. The molecular-weight distributions of the coal, tar and chars are shown as a histogram at the right. The histogram is divided into tar and char with pyridine-soluble and insoluble fractions.

12

DOA·523001 Final report

2/12/91

Figure 3b shows the molecule during pyrolysis. The rates for bridge breaking and cross-linking are determined from the FG modeL Some labile (breakable) bridges have broken, other labile bridges have been converted to unbreakable bridges by the abstraction of hydrogen to stabilize the free radicals and new cross-links have been formed. To determine the change of state of the computer molecules during a time step, the number of cross-links formed is determined using the FG subroutine (see below), and passed to the DYC subroutine. These cross-links are distributed randomly throughout the char, assuming that the cross-linking probability is proportional to the molecular weight of the monomer. Then the DYC subroutine breaks the appropriate number of bridging bonds and calculates the quantity of tar (i.e., room-temperature condensables) evolved for this time step using the internal and external transport equations. The result is the coal molecule representation and the molecular weight distributions shown in Figure 3b. The lighter "tar molecules," which leave the particle according to the transport equations, are cross hatched in Figure 3b. A fraction of the donatable hydrogen is used to stabilize the free radicals formed by bridge breaking, creating two new methyl groups per bridge and the same fraction of breakable bridges is converted into (unbreakable) double bonds. Figure 3c shows the final char which is highly cross-linked with unbreakable bridges and has no remaining breakable bridges, and thus no donatable hydrogen. The histogram now shows only tar and pyridine insoluble fractions. The extractables have been eliminated by tar formation and cross­ linking. The output of the DYC subroutine is the molecular weight distribution in the coal, its time dependent transformation during devolatilization and the evolution of tar determined by the transport of the lighter components. The DVC model was originally developed using the Monte Carlo method [12], but has been later modified using a modified percolation theory to perform the network statistics [61]. FG-DVC Model Parameters - The implementation of the FG-DVC model for a complex material like biomass requires the specification of several parameters, some of which can be constrained by the known structural units and some of which are constrained by experimental characterization data. A flow diagram of the model inputs and outputs is given in Figure 4. The basic idea is to calibrate the model using simple small scale experiments like TG-FTIR, FIMS, and elemental analysis, and then use the model to make predictions for conditions where experimental data are not readily available. This approach has been highly successful for coal where the FG-DVC model to make predictions for pyrolysis at very high heating rates by means of a model that was validated using low heating rate TG-FTIR, pyrolysis-FIMS, and fluidity data [12].

Modeling of Lignin Pyrolysis - While the model had not previously been used for whole biomass samples, simulations of lignin pyrolysis were performed to compare model predictions with experimental results obtained from TG-FTIR and pyrolysis-FIMS analyses [3]. These comparisons are used to further refine the polymer composition and/or kinetic files in an iterative fashion, as shown in Figure 4. The calibrated model can then be used to make predictions with reasonable confidence for conditions where experimental data are not available such as high heating rates. The network parameters (fraction of unbreakable bonds, initial cross-link density, and starting fraction of bonds per monomer) were chosen for raw lignin to match the experimental value of pyridine extractables (100%), the amount of tar from TG-FTIR experiments (-50%), and the values of the molecular weight distribution: Mn:::: 800 amu, Mw "" 1600-2000 amu with an overall range of 400-5000 amu. In the model, the actual value of pyridine extractables (which is approximated by the sum of masses less than 3000 amu) has been slightly modified from 100% to 85% in order to take into account the overall range of molecular weights, which goes up to 5000 amu. Two cases, case A and case B, were considered in ALC lignin simulations in order to predict different aspects of the available experimental data.

13

•

DOA-523001 hnaL repon 2/12/97

INPUT DATA

Balance

Helium Out

Window Purge

MODEL • Monomers • Bridges • Crosslinks • % Breakable Bonds

• FWlctional Group Pools • Cross linking Events

j

Fl'·IR Bench

Sample

PREDICTIONS

1----------, High Heating 1

I

Rate Predictions 1.._--______

1I

1

~



_

0.6

.~ c 05 ~ .

CO2

~O~

~

>=

0.1 U 0.0

a "

~

0.8 0.7

~

~

a..

40

,41

4

12.0

.•

- -

:g ~

iii I­

1

Figure 5. Comparison

of FG-DVC model predictions (solid lines) with pyrolysis data (circles) for ALC lignin at 30 °C/min (Case A). (Serio etai., 1994)

sl'

o~

0

I

­ j~

~

,~

0

8 .13,- 1::1

0.2 1

e

if .­

10.0

6.0

l!

400

~

200

!. 300

4.0

c

tattI /, . '''!ilAa • ft

.­

••

Ilijtllulllllt ••

_.

i-....J

_.

32 36 40

"'"XI~:;j'::i •"malm,n.1 1000 9D'Oliil!!t,,!Ja~ i • i i 900 BOO

700 U .,; 600 :; 500

Slm.39.S4% Exp.45.73%

• ft

..

0.4

U 0.0

._-

Tar

2.0 0.0 .IQg;gm o 4 ft

Slm.5.49%

Exp.6.58%

~

a:

:g

S! B.O

tP.

co

I~

9l

5~:J~ o



~ O.B

~y

~ 0.2

~

,,>

Figure. 6. Schematic diagram of the TG-FTIR apparatus.

1.0 .~

Sim. 2.49% Exp. 2.58%

~ 0.4 ~ O~

!;;­

~

I-

u; z w

.1 dl.lU 1.kiI.

4.8

I-

w

1.1. .

e

9

Z

100

••

No.Av.MW - 231. A22806.SUM 2810 505°C POPLAR SAWDUST (BIOMASS #6)

b >­ Ui

Wt.Av.MW _ 252. l3C-Corrected

u; z w

3.2

3.2

I­

I-

3!: z Q

3!: z

Q

1.6

1.6

0.0+--...,...--.... 100 o

O.o+-_,..L..~U~UL._IIM 200 300 MASS (Oa)

400

o

500

100

200 300 MASS (Oa)

400

Figure 16. FIMS-analysis data for (a) cellulose; (b) corn stalk; (c) wheat straw; (d) sugar cane bagasse; (e) Populus deltoides; and (f) Pinus radiata.

25

500

DOA·523001 Final

'epo"

2/12197

100

Herbaceous I.... . 300 400 500

< ..., 100 200

j

0.50

NO. 00

8

600 700 BOO

t F E e - .,.e=.

0

100

200

5.50 r-[-r-r--r-.......,....-r--r-r-.......,~--.--r-..

=

o

2

I

o

10

•

••• ••

"'C

>= o ()

(\I

400

600

800

1000

1200

1400

o

400

1600

2

.,

~

• •

"'C

eft.

• ••• • •

800

1000

1200

1400

1600

• Figure 24. The comparison of char, CO2, CO, CH 4 and H 2 0 experimental data (Nunn et al. [77]) and model predictions using the kinetic input files and product yields for Populus deltoides. The heating rate is 1000 Kls.

•

••

"'C

.~.

Q)

>='¢

600

e

"'C

o o 3:

..z­ Peak Temperature, K

3,------------------------,

-

•.,• •

5

Q)

(j)

1600

• ••• • • • •

d

c 10

•

1400

20,-------------------------------~

12~---------------------------~

6

1200

o

I

()

400

600

800

1000

1200

1400

1600

Peak Temperature, K

34

DOA·523001 Final report

2112197

Results Model Validation - Model predictions were compared with the yields and composition of pyrolysis products obtained under different conditions from those used to determine model parameters (e.g., different heating rates). An example is shown in Figure 24 for the high heating rate data of Nunn et al. (1000 Kfsec) [77]. Experimental data for sweet gum hardwood are compared with model predictions generated using the kinetic input files for PD. The results are reasonable for the overall weight loss (Figure 24) and CO2, but they are less good for the other gas species. It should be pointed out, however, that even in the cases of apparent poor yield predictions (e.g., Figure 24c,d,e), the predicted rates of evolution are often quite good. In other words, a good prediction of the experimental data could be obtained if the total yields of specific species were corrected without changing the kinetic parameters. In addition to the simplifications inherent in the predictive model, and the differences in the composition of the two materials (i.e., PO and the wood used in the work by Nunn et al. [77]), the observed differences in predicted and experimental yields could result from uncertainties in the time­ temperature history in the heated-grid reactor and/or secondary reactions which alter the product distribution. Future work in Phase II will involve refinement of the model to include a robust kinetic scheme, secondary reactions, mineral effects, heat and mass-transfer effects, and an improved classification scheme to account for composition differences. Summary - The model was used to predict pyrolysis behavior at conditions significantly different from those that were used for model development (high heating rates). The predictions were compared with the available literature data., and reasonably good agreement was found. 3.4

Estimate Of Technical Feasibility

The FG-OVC coal pyrolysis model has been successfully applied to biomass. The Phase I results indicate that it is feasible to model the pyrolysis behavior of different biomass feedstocks using kinetic and composition parameters approximated by selected data-base materials (standards). The approximation scheme is currently based on the feedstock elemental composition. Refinements of the proposed scheme and further validation of the model will be carried out in Phase II. 3.5 Explanation of Cost Data

The major difference in the original budget versus the actual budget was an increase in the amount spent on FIMS analysis at SRI International from $2400 to $3000, and an increase in the amount of money spent on other analyses at Huffman Laboratories from $500 to $)032. These additional expenses were due to a decision to increase the sample set from four to six samples and to do proximate and trace-element analyses in addition to ultimate analysis. These increases were offset by a reduction in the number of consulting days from 5 to 2.5, by agreement with Prof. Suuberg. 4. POTENTIAL APPLICATIONS 4.1 Technical, Economic, and Social Benefits

Technical, Economic and Social Benefits - It is expected that the overall results of Phases I and II will: (1) produce a model relating the characteristics of biomass feedstocks to the yields and composition of pyrolysis products; (2) provide predictive capabilities for the development and scale­ up of biomass-conversion processes, and for better tailoring of such processes to favor a particular distribution of products; (3) establish procedures for the quantitative analysis of biomass by FT-IR, TG-FTIR, and FIMS.

35

DOA-52300 I Final report

2112197

Results Model Validation - Model predictions were compared with the yields and composition of pyrolysis products obtained under different conditions from those used to determine model parameters (e_g_, different heating rates). An example is shown in Figure 24 for the high heating rate data of Nunn et al. (1000 Klsec) [77]. Experimental data for sweet gum hardwood are compared with model predictions generated using the kinetic input files for PD. The results are reasonable for the overall weight loss (Figure 24) and CO2, but they are less good for the other gas species. It should be pointed out, however, that even in the cases of apparent poor yield predictions (e.g., Figure 24c,d,e), the predicted rates of evolution are often quite good. In other words, a good prediction of the experimental data could be obtained if the total yields of specific species were corrected without changing the kinetic parameters. In addition to the simplifications inherent in the predictive model, and the differences in the composition of the two materials (Le., PD and the wood used in the work by Nunn et al. [77]), the observed differences in predicted and experimental yields could result from uncertainties in the time­ temperature history in the heated-grid reactor and/or secondary reactions which alter the product distribution. Future work in Phase II will involve refinement of the model to include a robust kinetic scheme, secondary reactions, mineral effects, heat and mass-transfer effects, and an improved classification scheme to account for composition differences. Summary - The model was used to predict pyrolysis behavior at conditions significantly different from those that were used for model development (high heating rates). The predictions were compared with the available literature data., and reasonably good agreement was found. 3.4 Estimate Of Technical Feasibility The FG-DVC coal pyrolysis model has been successfully applied to biomass. The Phase I results indicate that it is feasible to model the pyrolysis behavior of different biomass feedstocks using kinetic and composition parameters approximated by selected data-base materials (standards). The approximation scheme is currently based on the feedstock elemental composition. Refinements of the proposed scheme and further validation of the model will be carried out in Phase II. 3.5 Explanation of Cost Data The major difference in the original budget versus the actual budget was an increase in the amount spent on FIMS analysis at SRI International from $2400 to $3000, and an increase in the amount of money spent on other analyses at Huffman Laboratories from $500 to $1032. These additional expenses were due to a decision to increase the sample set from four to six samples and to do proximate and trace-element analyses in addition to ultimate analysis. These increases were offset by a reduction in the number of consulting days from 5 to 2.5, by agreement with Prof. Suuberg. 4. POTENTIAL APPLICATIONS 4.1 Technical, Economic, and Social Benefits Technical, Economic and Social Benefits - It is expected that the overall results of Phases I and II will: (1) produce a model relating the characteristics of biomass feedstocks to the yields and composition of pyrolysis products; (2) provide predictive capabilities for the development and scale­ up of biomass-conversion processes, and for better tailoring of such processes to favor a particular distribution of products; (3) establish procedures for the quantitative analysis of biomass by FT -JR, TG-FTIR, and FIMS.

35

DOA·523001 Final repon

2/12/97

Results ofthe proposed research should be ofgreat value to the forestry industry, the utilities, and - if biofuels are to become reality - to the biofuels industry. In addition, the model would also benefit efforts to use thermochemical conversion of biomass to produce chemicals and materials. It can also find applications in reprocessing agricultural wastes. Since biomass is the most abundant natural resource in the U.S., its efficient and clean utilization is of great practical interest. The improved, and thus more competitive, biomass-conversion technologies could also open attractive export opportunities for u.s. companies. This in turn would lead to the improvement of the American trade balance. Estimated Total Cost of Approach Relative to Benefits - The costs to the USDA for this project are relatively small ($250,000) when compared to the large potential economic impact in terms of job creation in the industries that are using biomass to produce fuels, chemicals, materials or electrical power. These products currently represent a large sector of the U.S. economy and the replacement of conventional feedstocks by biomass will become an important component over the next decade. A Phase III agreement for an equivalent amount is being negotiated with FLUENT, Inc., which would significantly leverage the USDA investment if the Phase II objectives are met The Phase II investment will also be leveraged by the significant investments in pyrolysis modeling technology that have already been made at AFR by GRI, DOE, and NSF. Effect on Policy Issues - The ability to successfully and comprehensively model biomass pyrolysis will encourage public and private efforts to pursue thermochemical conversion of biomass for the production of high-value products. It will also make it easier to use agricultural waste streams in combustion and gasification devices. 4.2 Comparison of Proposed Approach to Existing Technology Competition - We are not aware of any competing commercial product Some organizations have developed internal models for their own processes. Several biomass kinetic models have been developed and published in the open literature as means of correlating data and determining mechanisms. However, these are fairly simple, and are usually applied to a few species over a narrow range of conditions. Attempts to develop component models (e.g., lignin, cellulose, hemicellulose) have met with limited success, and the component composition is often unknown and/or difficult to determine. Competitive Advantage - Our approach is general for all types of biomass and does not depend on detailed information except for elemental analysis and, perhaps, proximate analysis. We also will have an advantage related to the integration of our model into a leading CFD code, FLUENT (about 2000 licenses, or 60% of the world market). We also have an advantage of the significant prior experience in pyrolysis modeling, the effort that has been supported by both the U.S. government and the private industry. 4.3 Outline of.Phase II The objective of the Phase II program is to complete the development of a predictive biomass­ pyrolysis model incorporating a detailed understanding of the physical and chemical processes occurring during pyrolysis of agricultural biomass feedstocks. The results from Phase I certainly show this approach is promising. The proposed program will rely on the predictive capability of AFR's unique FG-DVC modeling approach, and the application of the TG-FTIR method to provide a better understanding of thermal degradation of biomass. The project objectives will be accomplished in five tasks: (I) selection and characterization of an expanded test sample suite; (2) model refinements; (3) validation of the stand-alone model; (4) integration with a CFD code; and (5) evaluation of model and data methodology for commercial applications.

36

DOA·523001 Final ,epon

2112/91

4.4 Potential for Transition to Phase III Commercial Applications - The commercial applications of the proposed model involve biomass process design, development, and scale-up. Since pyrolysis constitutes an important step in most biomass-conversion or combustion processes, the market for the product would be extensive; about 650 activities in biomass conversion were identified in a recent study [2]. Most engineering firms use sophisticated hydrodynamic models for reactor design, and our pyrolysis model will be incorporated into this type of code as a submodel. Our experience with the coal-pyrolysis model shows that this is perhaps the most effective way of marketing the code. Accordingly, AFR has approached FLUENT, a major computational fluid dynamics (CFD) software developer, to secure the follow-on funding for Phase III. It is believed that marketing AFR's biomass model in conjunction with a widely-accepted CFD code will give AFR a strong competitive advantage. A stand-alone version of the model will also be offered for the use at research organizations and engineering firms. Sales of TG-FTIR equipment, analytical services for biomass characterization, and FT-IR process monitoring equipment would be pursued through On-Line Technologies, Inc., an AFR affiliate.

5. ACKNOWLEDGMENTS The support for this work came from the U.S. Department of Agriculture under Research Agreement No. 96-33610-2675. A research scholarship for Anker Jensen was provided by the Technical University of Denmark. The authors also acknowledge helpful discussions with Professor Eric M. Suuberg of Brown University.

6. RECIPIENTS OF REPORT Prof. Eric M. Suuberg Division of Engineering Brown University 182 Hope Street Baros-Holley, Room 253, Box D Providence, RI 02912 Dr. Thomas A. Milne Principal Chemist (EMERITUS) Center for Renewable Chemical Technologies & Materials National Renewable Energy Laboratory 1617 Cole Boulevard Golden Colorado, 80401-3393 Dr. Dipankar Choudhury FLUENT Product Manager Centerra Resource Park 10 Cavendish Court Lebanon, NH 03766-1442 Mr. John Curran Environmental Resources Services Suite 1600 204 N. Robinson Oklahoma, OK 73102 Dr. A. C. Buchanan, III Chemistry Division Oak Ridge National Laboratory 37

DOA·S23001 Final report

2112/97

PO box 2008, MS-6197 Oak Ridge, TN 37831-6197 Dr. John G. Reynolds Lawrence Livermore National Lab. PO Box 808, L-365 Livermore, CA 94551 Mr. Klaus H. Oehr Vice-President Chemical Research & Development DynaMotive 3760 Westbrook Mall Vancouver, BC V6S 2L2 Dr. Ripudaman Malhotra Group Leader, Carbon Materials SRI International 333 Ravenswood Ave. Menlo Park, CA 94025 7. REFERENCES

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Serio, M. A., W6jtowicz, M. A. and Charpenay, S., "Pyrolysis", in Encyclopedia of Energy Technology and the Environment (A. Bisio and S. G. Boots, Eds), John Wiley & Sons, New York, pp. 2281-2308, 1995 Bridgwater, A.V., Biomass, 22 (1-4), 279 (1990). Serio, M.A., Charpenay, S., Bassilakis, R., and Solomon, P.R., "Measurement and Modeling of Lignin Pyrolysis", Journal of Biomass & Bioenergy, Vol. 7, Nos. 1-6, 107 (1994). Solomon, P.R. and Hamblen, D.G., in Chemistry of Coal Conversion, (R.H. Schlosberg, Editor), Plenum Press, NY, Chapter 5, p. 121, (1985). Serio, M.A., Hamblen, D.G., Markham, J.R., Solomon, P.R., Energy and Fuels, 1(2), l38, ( 1987). Solomon, P.R., New Approaches in Coal Chemistry, ACS Symp. Series 169, ACS, Washington, DC, 1981, pp 61-7l. Solomon, P.R. and King, H.H., Fuel, 63, 1302, (1984). Solomon, P.R., Squire, K.R., Carangelo, R.M., Proc, Int. Con! Coal Science, Sydney, p. 945, 1985. Solomon, P.R. and Squire, K.R., ACS Div. Fuel Chem. Prepr. 30, (4), 347, (1985). Solomon, P.R., Hamblen, D.G., Deshpande, G.V. and Serio, M.A., Coal Science and Technology 11, (J.A. Moulijn, K.A. Nater, and H.A.G. Chermin, Eds.), Elsevier, Amsterdam, p. 601, (1987). Solomon, P.R., Hamblen, D.G., Carangelo, R.M., Serio, M.A., and Deshpande, G.V., Combustion and Flame, 1988, 71, 137. Solomon, P.R., Hamblen, D.G., Carangelo, R.M., Serio, M.A., and Deshpande, G.Y., Energy & Fuels, 2, 405, (1988). Solomon, P.R., Hamblen, D.G., Yu, Z.Z., and Serio, M.A., Fuel 69, 754, (1990). Solomon, P.R., Squire, K.R., Carangelo, R.M., ACS. Div. Fuel Chem. Prepr. 29(1),10, (1984). Squire, K.R., Solomon, P.R., Carangelo, R.M., and DiTaranto, M.B., Fuel, 65, 833, (1986). Squire, K.R., Solomon, P.R., DiTaranto, M.B., and Carangelo, R.M., ACS Div. Fuel Chem. Prepr., 30, (1), 386, (1985). Solomon, P.R., Serio, M.A., Carangelo, R.M., and Markham, J.R., Fuel, 65, 182, (1986). Solomon, P.R., Hamblen, D.G., and Yu, Z.Z., "Network Models of Coal Thermal Decomposition," ACS Div. Fuel Chem. Prepr., 34(4) 1280 (1989).

38

DOA·523001 Final repon

2/12/97

19 Serio, M.A., Solomon, P.R., Yu, Z.Z., Deshpande, G.V., Hamblen, D.G., ACS Div. Fuel Chem. Prepr. 33(3), 91, (1988). 20 Serio, M.A., Solomon, P.R., Yu, Z.Z., and Deshpande, G.V., "An Improved Model of Coal Devolatilization", Presented at the Int. Conference on Coal Science, Japan (10/1989). 21 Solomon, P.R., Best, P.E., Yu, Z.Z., and Deshpande, G.V., "A Macromolecular Network Model for Coal Fluidity," ACS Div. Fuel Chem. Prepr., 34(3), 895, (1989). 22 Carangelo, R.M., Solomon, P.R., and Gerson, D.J., Fuel, 66, 960 (1987). 23 Whelan, J.K., Solomon, P.R., Deshpande, G.V., Carangelo, R.M., Energy and Fuels, 2, 65 (1988). 24 Goldstein, I. S., in Emerging Technologies for Materials and Chemicals from Biomass (R. M. Rowell, T. P. Schultz, and R. Narayan, Eds.), ACS Symposium Series No. 476, American Chemical Society, Washington, DC, pp. 332-338 (1992). 25 Goldstein, I. S., in Organic Chemicals from Biomass (Boca Raton, Ed.), CRC Press (1981). 26 Austin, G.T., Shreve's Chemical Process Industries, 5th Edition, McGraw-Hili, New York pp. 602-612 (1984). 27 Grohmann, K., Wyman, C.E., and Himme~ M.E., in Emerging Technologies for Materials and Chemicals from Biomass, ACS Symposium Series No. 476, R.M. Rowel~ T.P. Schultz, and R. Narayan, Eds.,American Chemical Society, Washington, DC, pp. 354-392 (1992). 28 Antal, M.J., Jr., in Adv. in Solar Energy (Boer, K.W., Duffie, J.W., Eds.), American Solar Energy Society, Boulder, CO, 1,61 (1984); 2,175 (1985). 29 Soltes, EJ., in Biomass Energy Development, W.H. Smith, Ed., Plenum Press, New York, 321 (1986). 30 Shafizadeh, F., in Proc. Specialists' Workshop on Fast Pyrolysis of Biomass (J. Diebold, Ed.), SERI/CP-622-1096, Solar Energy Research Institute, Golden, CO, 79 (1980). 31 Shafizadeh, F., in The Chemistry of Solid Wood (R. M. Rowell, Ed.), Advances in Chemistry Series No. 207, American Chemical Society, Washington, DC (1984). 32 Bridgwater, A.V., and Cottam, M.-L., Energy and Fuels, 6, (2), 113 (1992). 33 Elliott, D.C., Beckman, D., Bridgwater, A.V., Diebold, J., Severt, S.B., and Solantusta, Y., Energy and Fuels, 5, 399 (1991). 34 Overend, R. P. Milne, T. A. , and Mudge, L.K., Eds., Fundamentals of Thermochemical Biomass Conversion, Elsevier, (1985). 35 Soltes, E.J., and Milne, T.A., Eds., Pyrolysis Oils from Biomass: Producing, Analyzing and Upgrading, ACS Symposium Series No. 376, American Chemical Society, Washington, DC (1988). 36 Bridgwtter, A.V, Kuester, J.L., (Eds.), Research in Thermochem ical Biomass ConveJSion, Elsevier Applied Science, London (1988). 37 S. Borman, Chern. Eng. News 19-22 (Sept. 10, 1990). 38 Nunn, T .R., Howard, J.B., Longwell, J.P., and Peters W.A., Fundamentals of Thermochem ical Biomass ConveJSion, R.P. Overend, T .A. Milne, and L.K. Mudge, (Eds.), Elsevier Applied Science Publishers, New York pp. 293-314 (1985). 39 Elliott, D.C., in Pyrolysis Oils from Biomass: Producing Analyzing and Upgrading, ACS Symposium Series No. 376, E.J. Soltes and T.A. Milne, Eds., American Chemical Society, Washington, DC, pp. 55-65 (1988). 40 Evans, R. J. and Milne, T. A., SERIIPR-234-3026, Solar Energy Research Institute, Golden, CO, 1986 41 Essig, M. G., Richards, G. N., Schenck. E. M., "Mechanisms of Formation of the Major Volatile Products from the Pyrolysis of Cellulose," in Cellulose and Wood Chemistry and Technology, C. Schuerch, Ed., J. Wiley & Sons, New York, 1989 42 Pan W., Richards, G. N., "Influence of Metal Ions on Volatile Products of Pyrolysis of Wood," J Anal. Appl. Pyrolysis, 16, 117-126, 1989

39

DOA-S23001 Final report

2/12/97

43 Pan W., Richards, G. N., "Volatile Products of Oxadative Pyrolysis of Wood: Influence of Metal Ions," J Anal. Appl. Pyrolysis, 17,261-273, 1990 44 Shafizadeh, F., in Fundamentals of Thermochemical Biomass Conversion, R.P. Overend, T.A. Milne, and L.K. Mudge, (Eds)., Elsevier Applied Science Publishers, New York (1985). 45 Di Blasi, C., Prog. Energy Combust. Sci., 19, 71 (1993) 46 Van Krevelen, D.W., Coal, Elsevier, Amsterdam, (1961). 47 Green, T.K., Kovac, J., and Larsen, J.W., Fuel, 63, 935 (1984) 48 Green, T.K., Kovac, J., Larsen, J.W., in Coal Structure, R.A. Meyers, Ed., Academic Press, (1982). 49 Lucht, L.M. and Peppas, N.A., Fuel, 66, 803, (1987). 50 Lucht, L.M., Larsen, J.M., and Peppas, N.A., Energy & Fuels, 1, 56, (1987). 51 Larsen, J.W.,ACS Fuel Chem. Div. Preprints, 30, (4), 444, (1985). 52 Green, T., Kovac, J., Brenner, D., and Larsen, J., Coal Structure, (R.A. Meyers, Ed.), Academic, NY, p 199, (1982). 53 Hall, PJ., Marsh, H., and Thomas, K.M., Fuel, 67, 863, (1988). 54 Sanada, Y. and Honda, H., Fuel, 45,295, (1966). 55 Suuberg, E. M., Otake, Y., Langner, M., Leung, K. T. and Milosavljevic, 1., Energy & Fuels., 8, 1247, (1994). 56 Suuberg, E.M., Lee, D., and Larsen, J.W., Fuel, 64, 1668, (1985). 57 Suuberg, E.M., Unger, P.E., and Larsen, J.W., Energy & Fuels, 1, 305, (1987). 58 Suuberg, E.M., Peters, W.A., and Howard, J.B., 17th Symp. (Int) on Combustion, 1979, p. 117. 59 Fong, W.S., Peters, W.A., and Howard, J.B., Fuel, 65, 251, (1986). 60 Nelson, J.R., Fuel, 62, 112, (1983). 61 Hamblen, D. G., Yu, Z. Z., Charpenay, S., Serio, M. A. and Solomon, P. P., Proc. 7th Int. Con! on Coal Science 2, 401 (1993). 62 "Production of Carbon Materials from Lignin," DOE SBIR project, DE-FG05-93ER81593, Phases I and II, 1993-1996. 63 St. John, G.A., Butrill, Jr., S.E. and Anbar, M., ACS Symposium Series, 71, p. 223, (1978). 64 Varhegyi, G., Antal, M. J., Jr., Szekely, T., Till, F., Jakab, E. and Szabo, P., Energy & Fuels, 2, 273 (1988). 65 Squire, K.R., and Solomon, P.R., Characterization of Biomass as a Source of Chemicals, Final Report, National Science Foundation, Contract # CPE-8107453 (1983). 66 Koufopanos, C. A., Maschio, G. and Lucchesi, A., Can. J Chem. Eng., 67, 75 (1989). 67 Ward, S. M. and Breslaw, J., Comb. Flame, 61, 261 (1985). 68 Miller, R. S. and Bellan, J., submitted to Combust. Sci. Tech. (1996). 69 Solomon, P.R., Hamblen, D.G., Serio, M.A., Yu, Z.Z. and Charpenay, S., 'A characterization Method and Model for Predicting Coal Conversion Behaviour', Fuel, 72, 469-488, 1993 70 Zhao, Y., Serio, M. A., Bassilakis, R. and Solomon, P. R., Twenty-Fifth Symposium (International) on Combustion, The Combustion Institute, Pittsburgh, PA, 1994, pp. 553-560. 71 Solomon, P. R., Hamblen, D. G., Serio, M. A., Yu, Z. Z., and Charpenay, S., "A Characterization Method & Model for Predicting Coal Conversion Behavior," Fuel, 72, (4), 489 (1993). 72 Varhegyi, G., Szabo, P. and Antal, M. J., "Reaction kinetics of the thermal decomposition of cellulose and hemicellulose in biomass materials," in Advances in Themochemical Biomass Conversion (A. V. Bridgwater, Ed.), vol. 2, Blackie Academic & Professional, London, 1992 73 Van Heek, K. H. and Juentgen, H., Ber. Bunsenges. Phys. Chem., 72, 1223 (1968). 74 Milosavljevic, I. and Suuberg, E. M., Ind Eng. Chem. Res., 34, 1081 (1995). 75 Reynolds, J.G., and Burnham, A.K., Energy & Fuels, 11,88 (1997). 76 Antal, M. J., Jr. and Varhegyi, G., Ind Eng. Chem. Res., 34, 703 (1995). 77 Nunn, T. R., Howard, J. B., Longwell, J. P. and Peters, W. A., Ind Eng. Chem. Process Des. Dev., 24, 836 (1985).

40

APPENDIX

BIOMASS DATABASE

41

2/12/97

DOA-523001 Final report (Appendix)

L==--=t-~---------------Source

·t- - -·­ . . ,3!tt:::

I

10 11

____ _ Cellulose

_ 44.44

jCellUIOse fellljlose

6.17

Wood

------~·--I

-----.-~--

Bark/unknown Larch/OK Beech/OK Spruce/OK Spruce/OK Pine/OK ________ ~ Beech/OK Spruce/OK --_._--------

-~.---------

--- - - - - - - - - - - - - - - ­

_~_ Sp~~~IDK-==-- -

_!1~~_a~ternredmall!e~.

13 15 19 19 19

0.00 0.03

Beech Beech Balsam fir OOU!;!~sJi~ ____

Eastern hemlock

_A_Sh,-cl~U:;::::~:ara~~~C\f-Ju--Fc,d.;rJcell~~;~_J~:~~~:Ci,~a~i:~~_~JExtractives_

50.70

:~:::

:::~

0.09

0.62 0.20 0.65

~~;-~r-=-~-~~-I~~02=10.00

-I

6.00 6.30

__

42.c.~6

__ Q..~

0.40 0.30 ---I 0.25

0.00

O.()() __ ~9~r

-.--

~

6:6~_

6:6~

24.00

--'-~~

82.77

t--------'l~~-----+-- ---+--

16.98 15.55

-

J---l--=-==--t-= -

86.38

+ _ __________

0.48 0.20

j

-=1

-.

_

0.09

0.03

1.94

_I

0.50

45.3710.00 -1------

r--

0.63 2.04 1.48

+---1-···­ --I---=-t=~=I----~ - ... ~-------

n.r 5.40

40.16

--

--+-----­

------

-

--_. ------~

-~--

-_._­

-+-­

~t---------I-

24.20

-~----+------

1

---\

----------­

24.70

9.79

----=1==-­

---­

.- --~----~~~I-===------t

1­

8.77

91.23

24.20

-------1---- -------

______

2.00 3.10

27.80

--- ------

5.00 7.20

J___0.9~_1

- 1­

_3~._0~_ ~::~6~=-

39.00

---j~---

_~ _

0.00 0.00

--i--­

-t------il---·

:~ I ~~~ I-

0.00

----~.----

46.70

I

0.00 0.00

1.74 0.30 0.70

0.00

~----

-f---

6.65 6.61

0.00

1­

- . ~-.•_.1 ~__10_0. 00

0.42 0.05 0.02 -----------1-- ~----__+ ----j------- --­ 0.15 0.05_L_0.0=! -1------ ~-----0_05 I 0.01

0.17 0.01

0.14 0.02 0.41 ___ ().~ _0.01

_0.10 0.02 0.01 1_____

0.27 0.03 0.01

0.32 0.04 0.03 ---O~58-- 0~07 0.09

48~0_6,10_ -.i.~()_

4862_ 4_S99

+-_ __--+__

4.01

-1-- - - - - . - -

H_

5.07 42.87 49.70 5.90 44.17 -_._-- --- - -----"- '"-48.70 5.90 45.17 50.20 6.00 43.63 501)0 59o____ 43~01 _§.l.OO _6:..~0__ 42.77 49.90 5.60 44.19 52.10 5.00 42.25 50.50 --5.()()42.66

100.00 100.00 100.00

0.00 7.28

--+-----------­

1-

::;~ __ 6~~

__

51.57 -------- --_.

51.10 46.80

0.00 0.00

r:!-':t -:~i-!:: .:!= 44.22

6.00__ ~s.~ 5.60_ ~4_5_.8_0 6.30_ 42.40

48~0

0.00 0.01

0.05~~

---=- =-

,__~_ Spruce/OK__ 7 Silver birch --------------.-----­ --~--

-­

~-

43.76

6.07

50.77

-~--SutJ
r J

.. -

F""" Coo,, """:'

K. Raveendran et al.

. . F._.~. f=r.a. n._d~e~a.n.ell.·S. p. e