Hydrodynamic and gasification behavior of coal and biomass ... · different gasification agents....

Hydrodynamic and gasification behavior of coal and biomass fluidized beds and their mixtures

Bahareh Estejab

Dissertation submitted to the faculty of the Virginia Polytechnic Institute and State University in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

In

Mechanical Engineering

Francine Battaglia Javid Bayandor

Scott T. Huxtable Brian Y. Lattimer Gerald H. Luttrell

February 18, 2016 Blacksburg, Virginia

Keywords: Biomass, Binary mixture, Coal, Drag models, Fluidization, Gasification, Co-

gasification, Mixture

Copyright 2016, Bahareh Estejab

Hydrodynamic and gasification behavior of coal and biomass fluidized beds

and their mixtures

Bahareh Estejab

ABSTRACT

In this study, efforts ensued to increase our knowledge of fluidization and gasification behavior of

Geldart A particles using CFD. An extensive Eulerian-Eulerian numerical study was executed and

simulations were compared and validated with experiments conducted at Utah State University. In order

to improve numerical predictions using an Eulerian-Eulerian model, drag models were assessed to

determine if they were suitable for fine particles classified as Geldart A. The results proved that if static

regions of mass in fluidized beds are neglected, most drag models work well with Geldart A particles.

The most reliable drag model for both single and binary mixtures was proved to be the Gidaspow-blend

model. In order to capture the overshoot of pressure in homogeneous fluidization regions, a new modeling

technique was proposed that modified the definition of the critical velocity in the Ergun correlation. The

new modeling technique showed promising results for predicting fluidization behavior of fine particles.

The fluidization behavior of three different mixtures of coal and poplar wood were studied. Although

results indicated good mixing characteristics for all mixtures, there was a tendency for better mixing with

higher percentages of poplar wood.

In this study, efforts continued to model co-gasification of coal and biomass. Comparing the coal

gasification of large (Geldart B) and fine (Geldart A) particles showed that using finer particles had a

pronounced effect on gas yields where CO mass fraction increased, although H2 and CH4 mass fraction

slightly decreased. The gas yields of coal gasification with fine particles were also compared using three

different gasification agents. Modeling the co-gasification of coal-switchgrass of both fine particles of

Geldart A and larger particles of Geldart B showed that there is not a synergetic effect in terms of gas

yields of H2 and CH4. The gas yields of CO, however, showed a significant increase during co-

gasification. The effects of gasification temperature on gas yields were also investigated.

iii

To the loving memory of my Mother, Azam Kolahdooz Esfahani,

who has inspired me every day and who I miss every day.

To my Husband, Siavash, who has supported me during

the whole journey of our lives together.

To my dearest family.

iv

Acknowledgments

Throughout my PhD journey, a lot of people helped me without whom accomplishing this

dissertation was not possible.

First, I would like to thank my PhD advisor, Dr. Francine Battaglia who will remain forever one of

my mentors in life. She provided me with the opportunity to achieve one of my goals in life. She

dedicates her time generously to her students. Her creative advises, helps and supports was always with

me during the course of this research.

I would also like to express my gratitude to Dr. Foster Agblevor and his former student, Guevara Che

Nyendu, who provided me with most of the experimental data presented in this study. I also would like to

express my appreciation to Ben Poe and Jamie Archual for the technical computer supports provided

throughout my dissertation.

I would like to acknowledge members of Crest lab, whose advices, supports and friendship has

always been with me.

I also would like to especially thank my true friends in Blacksburg, who become and remain forever a

part of my family.

v

Contents

Contents ........................................................................................................................................................ v

List of Figures ............................................................................................................................................ viii

List of Tables ............................................................................................................................................... xi

Chapter 1. Introduction ................................................................................................................................. 1

1.1. Background ........................................................................................................................................ 1

1.2. Research objectives and approaches .................................................................................................. 2

1.3. Outline of the dissertation .................................................................................................................. 3

Chapter 2. Literature review ......................................................................................................................... 4

2.1. Fluidization ........................................................................................................................................ 5

2.2. Gasification ........................................................................................................................................ 9

Chapter 3. Numerical Formulation ............................................................................................................. 12

3.1. Governing equations ........................................................................................................................ 12

3.2. Problem description and grid resolution study ................................................................................ 15

Chapter 4. Fluidization ................................................................................................................................ 17

4.1. Fluidization theory ........................................................................................................................... 17

4.2. Drag models .................................................................................................................................... 19

4.2.1. Introduction ............................................................................................................................... 19

4.2.2. Drag models .............................................................................................................................. 20

4.2.2.1. Wen-Yu drag model (1966) ............................................................................................... 20

4.2.2.2. Syamlal-O’Brien drag model (1988) ................................................................................. 20

4.2.2.3. Gidaspow drag model (1994) ............................................................................................. 21

4.2.2.4. Huilin–Gidaspow drag model (2003) ................................................................................ 21

4.2.2.5. Hill-Koch-Ladd (HKL) drag model (2006) ....................................................................... 22

4.2.2.6. BVK drag model (2007) .................................................................................................... 23

4.2.2.7. HYS drag model (2010) ..................................................................................................... 24

vi

4.2.3. Results and discussion .............................................................................................................. 25

4.2.3.1. Single solids phases of coal and poplar ............................................................................. 28

4.2.3.2. Binary Mixtures of Coal and Poplar Wood ....................................................................... 31

4.2.4. Conclusions ............................................................................................................................... 37

4.3. Fluidization behavior and mixing properties of coal-biomass mixture............................................ 38

4.3.1. Coal and poplar fluidization ...................................................................................................... 38

4.3.2. Binary mixtures of coal and poplar fluidization ....................................................................... 41

4.3.2.1. Mixing properties of coal-poplar wood mixtures ............................................................... 46

4.3.3. Conclusions ............................................................................................................................... 50

4.4. New method to model fluidization of Geldart A particles ............................................................... 51

4.4.1. Introduction ............................................................................................................................... 51

4.4.2. Results and discussions ............................................................................................................. 54

4.4.3. Conclusions ............................................................................................................................... 57

Chapter 5. Gasification ............................................................................................................................... 58

5.1. Introduction ...................................................................................................................................... 58

5.1.1. Coal and biomass composition ................................................................................................. 59

5.1.2. Gasification reactions................................................................................................................ 60

5.1.2.1. Steam gasification: C + H2O CO + H2 .......................................................................... 61

5.1.2.2. Carbon dioxide gasification: C + CO2 2CO .................................................................. 62

5.1.2.3. Methanation: C + 2H2 CH4 ............................................................................................ 62

5.1.2.4. Carbon combustion: 2C + O2 2CO and 2CO + O2 2CO2 .......................................... 62

5.1.2.5. Methane oxidation: CH4 + 2O2 CO2 + 2H2O ................................................................. 63

5.1.2.6. Hydrogen-Oxygen reaction: 2H2 + O2 2H2O ................................................................ 63

5.1.2.7. Water-gas shift reaction: CO + H2O CO2 + H2 .............................................................. 63

5.2. Coal gasification .............................................................................................................................. 64

5.2.1. Effects of fluidization velocity on gasification process ............................................................ 64

vii

5.2.2. Effects of particle classifications on gasification process ......................................................... 74

5.2.3. Effects of gasification medium on gasification process ............................................................ 78

5.3. Co-gasification of coal and biomass ................................................................................................ 81

5.3.1. Single gasification reaction ....................................................................................................... 81

5.3.2. Complete set of reactions .......................................................................................................... 85

5.3.3. Complete set of reactions with Geldart A particles .................................................................. 88

5.4. Conclusions ...................................................................................................................................... 89

Chapter 6. Conclusions and Future work .................................................................................................... 91

6.1. Summary .......................................................................................................................................... 91

6.2. Significance and contribution .......................................................................................................... 93

6.3. Future work ...................................................................................................................................... 94

References ................................................................................................................................................... 95

Appendix A ............................................................................................................................................... 105

Appendix B ............................................................................................................................................... 108

viii

List of Figures

Figure ‎3.1. The 2D plane representing the chamber of the cylindrical reactor. .......................................... 16

Figure ‎4.1. (a) Pressure drop versus the gas inlet velocity comparing experiments [90] and simulations,

and (b) the dimensionless bed height for different gas inlet velocities comparing experiments [90] and

simulations for sand. ................................................................................................................................... 19

Figure ‎4.2. Bed height of (a) poplar wood and (b) 70:30 coal-poplar wood mass ratio comparing

predictions for adjusted and unadjusted initial conditions to experiments [90]. For both simulations, the

Gidaspow drag model is used. .................................................................................................................... 27

Figure ‎4.3. Bed height of (a) coal and (b) poplar wood for different drag models compared to experiments

[90]. ............................................................................................................................................................. 29

Figure ‎4.4. Void fraction profiles for (a) coal and (b) poplar wood at h/h0 = 1 and Ug = 9.87 cm/s. ......... 30

Figure ‎4.5. Bed height of (a) 90:10, (b) 80:20, and (c) 70:30 coal-poplar wood mass ratio for different

drag models compared to experiments [90]. ............................................................................................... 32

Figure ‎4.6. Void fraction profiles for (a) 90:10, (b) 80:20 and (c) 70:30 coal-poplar wood mass ratio at

h/h0 = 1 and Ug = 9.87 cm/s. ........................................................................................................................ 33

Figure ‎4.7. Instantaneous void fraction for 70:30 coal-poplar wood mass ratio from 25 – 29 s (t = 1 s) at

Ug = 9.87 cm/s using (a) HKL and (b) Gidaspow-blend drag models ........................................................ 34

Figure ‎4.8. Time-average coal and poplar wood velocity vectors along with velocity magnitude contours

for Ug = 9.87 cm/s for 70:30 mass ratio of coal-poplar wood using (a) HKL and (b) Gidaspow-blend drag

models. ........................................................................................................................................................ 34

Figure ‎4.9. Instantaneous coal vertical velocity for 70:30 coal-poplar wood mass ratio from 25 – 29 s (t

= 1 s) at Ug = 9.87 cm/s using (a) HKL and (b) Gidaspow-blend drag models. ......................................... 36

Figure ‎4.10. Instantaneous poplar wood vertical velocity for 70:30 coal-poplar wood mass ratio from 25 –

29 s (t = 1 s) at Ug = 9.87 cm/s using (a) HKL and (b) Gidaspow-blend drag models. ............................ 37

Figure ‎4.11. Pressure drop versus the gas inlet velocity comparing experiments [90] and simulations for

(a) coal, (b) poplar wood. ............................................................................................................................ 40

Figure ‎4.12. The dimensionless bed height for different gas inlet velocities comparing experiments [90]

and simulations for (a) coal and (b) poplar wood. ...................................................................................... 40

Figure ‎4.13. Instantaneous void fraction contours at t = 30 s for coal at (a) Ug = 4.93 cm/s and (b) Ug =

9.87 cm/s; and (c) void fraction horizontally averaged across the reactor diameter versus axial direction;

for poplar at (d) Ug = 4.93 cm/s and (e) Ug = 9.87 cm/s; and (f) void fraction horizontally averaged across

the reactor diameter versus axial direction. ................................................................................................ 41

Figure ‎4.14. Pressure drop versus the gas inlet velocity comparing experiments [90] and simulations for

mass ratios for (a) 90:10, (b) 80:20 and (c) 70:30 mass ratio of coal-poplar. ............................................. 43

ix

Figure ‎4.15. The dimensionless bed height for different gas inlet velocities comparing experiments [90]

and simulations for mass ratios for (a) 90:10, (b) 80:20 and (c) 70:30 mass ratio of coal-poplar. ............. 44

Figure ‎4.16. Void fraction horizontally averaged across the reactor diameter versus axial direction for Ug

= 9.87 cm/s for 90:10, 80:20 and 70:30 mass ratio of coal-poplar. ............................................................ 45

Figure ‎4.17. Instantaneous volume fraction contours for all mass ratio of coal-poplar using a binary

mixture model with Ug = 9.87 cm/s at t =30 s. .......................................................................................... 47

Figure ‎4.18. Time-averaged axial solids mass flux spatially averaged across the bed width versus the

dimensionless height for Ug = 9.87 cm/s. ................................................................................................... 49

Figure ‎4.19. Time-averaged axial solids mass flux spatially averaged across the bed width versus the

dimensionless height for Ug = 9.87 cm/s. ................................................................................................... 49

Figure ‎4.20. Time-average coal velocity vectors for Ug = 9.87 cm/s for (a) 90:10, (b) 80:20, (c) 70:30

mass ratio of coal-poplar. ............................................................................................................................ 50

Figure ‎4.21. Pressure drop versus the gas inlet velocity comparing experiments [90] and adjusted

simulations based on Umf and Umb for poplar. ............................................................................................. 52

Figure ‎4.22. A sample of the new approach to find accurate bed height (poplar wood, Ug = 9.87 cm/s); (a)

averaged void fraction, and (b) averaged solid pressure, across the reactor diameter versus axial direction.

.................................................................................................................................................................... 54

Figure ‎4.23. (a) Pressure drop and (b) dimensionless bed height versus the gas inlet velocity comparing

experiments [90] and simulations for coal .................................................................................................. 55


experiments [90] and simulations for poplar wood .................................................................................... 56


experiments [90] and simulations for switchgrass ...................................................................................... 56

Figure ‎5.1. The 2D plane representing a center plane of the cylindrical reactor. ....................................... 65

Figure ‎5.2. Mass fraction of (a) CH4, (b) H2, and (c) CO versus time for Geldart B particles. .................. 67

Figure ‎5.3. The instantaneous void fraction contours for six concessive times from 5 – 10 s for case 2. .. 68




Figure ‎5.7. Time-averaged void fraction contours for Geldart B particles. ................................................ 72

Figure ‎5.8. Time-averaged gas temperature contours for Geldart B particles. ........................................... 73

Figure ‎5.9. Time-averaged CO mass fraction contours for Geldart B particles. ......................................... 74

x

Figure ‎5.10. Mass fraction of (a) CH4, (b) H2, and (c) CO versus time for air gasification for Geldart A

particles. ...................................................................................................................................................... 76

Figure ‎5.11. Gas temperature versus time for air gasification for Geldart A particles. .............................. 77

Figure ‎5.12. Time-averaged contours of (a) gas temperature, (b) H2 mass fraction, and (c) CO mass

fraction for Geldart A particles. .................................................................................................................. 78

Figure ‎5.13. Mass fraction of (a) CH4, (b) H2, and (c) CO versus time for CO2 and N2 mediums for

Geldart A particles. ..................................................................................................................................... 80

Figure ‎5.14. Gas temperature versus time for air gasification for Geldart A particles. ............................. 81

Figure ‎5.15. Time-averaged void fraction contours for coal, switchgrass and coal-switchgrass (50:50). .. 83

Figure ‎5.16. Time-averaged reaction rate for CO2 gasification reaction contours for coal, switchgrass and

coal-switchgrass (50:50). ............................................................................................................................ 84

Figure ‎5.17. Time-averaged mass fraction of CO contours for coal, switchgrass and coal-switchgrass

(50:50). ........................................................................................................................................................ 84

Figure ‎5.18. CO mass fraction profiles along the reactor radius for coal, switchgrass, and coal-switchgrass

(50:50). ........................................................................................................................................................ 85

Figure ‎5.19. Time-averaged void fraction contours for different gasification temperature. ....................... 87

Figure ‎5.20. Time-averaged rate of CO2 gasification reaction contours for different gasification

temperature for 50:50 mass ratio of coal-switchgrass................................................................................. 88

Figure A . 1. Combustor geometry. .......................................................................................................... 106

Figure A . 2. (a) Void fraction, (b) gas temperature, (c) cold char temperature, (d) hot char temperature,

(e) CO mass fraction, and (f) CO2 mass fraction contours along with Pannala results [143] comparing

simulations for coal gasification. .............................................................................................................. 107

Figure B . 1. Time-averaged solid temperature contours for (a) coal and (b) switchgrass. ...................... 108

xi

List of Tables

Table ‎3.1. Grid resolution and central processing unit. .............................................................................. 16

Table ‎4.1. Properties for sand, coal, and poplar .......................................................................................... 18

Table ‎4.2. Properties for binary mixture model .......................................................................................... 25

Table ‎4.3. Properties for coal, poplar wood and switch grass ..................................................................... 53

Table ‎5.1. Kinetic constants for two types of coal ...................................................................................... 64

Table ‎5.2. Initial conditions ........................................................................................................................ 66

Table ‎5.3. Annulus and jet velocity for air and solids. ............................................................................... 66

Table ‎5.4. Initial conditions for studying the effects of particle classifications .......................................... 75

Table ‎5.5. Gas yields from coal gasification in different mediums for Geldart A particles ....................... 78

Table ‎5.6. Simulation initial data for co-gasification .................................................................................. 82

Table ‎5.7. Gas yields from coal and switchgrass gasification and co-gasification mixtures. ..................... 87

Table ‎5.8. Gas yields from co-gasification in different temperatures ......................................................... 87

Table ‎5.9. Gas yields from coal and switchgrass gasification and co-gasification of 50:50 mixture. ........ 89

1

Chapter 1. Introduction

1.1. Background

Fluidized beds are widely used in many modern technologies for efficient implementation of various

physical and chemical processes found in the pharmaceutical industry, petroleum industry, and fuel

industry. The nearly isothermal condition, rapid mixing of particles, high heat and mass transfer rates,

large available surface area, and resistance to sudden changes in operating conditions are some inherent

advantages of fluidization technology. In contrast, overarching challenges with fluidized beds are scale-up

from lab to pilot-scale and from one type of particles with specific physical properties to a new type.

In a fluidized bed reactor, a fluid is passed upward through a granular material. At sufficiently high

fluid velocities, when aerodynamic drag forces exerted on particles are equal to the particle gravitational

force, solids behave similar to a fluid. This process is known as fluidization. Gasification, which is an

important process for converting solid fuels into useful gaseous products, can be applied to fluidized beds.

In gasification, the flow is not only responsible for fluidization but also contributes to reaction processes.

Gasification is an old technology that was invented in the 16th century when it was discovered that

heating coal or wood can produce gas. By the mid-18th century, the technology was used in commercial

gas plants for lighting and heating purposes. The modern use of the technology, however, is the

production of fuels and chemicals, which began about 60 years ago with large scale industrial gasification

plants. During the last decade, due to unstable prices of crude oil and geopolitical instabilities of oil-

producing countries, demand to use the technology has increased and several gasification plants are

operating around the world where the amount of synthesis gas produced almost doubled during the past

decade [1].The gasification technologies council expects the worldwide gasification capacity to expand

significantly by 2018 [1].

In gasification, fuel reacts with a controlled amount of oxygen. Gasification technology has a variety

of applications from waste treatment to power production and nuclear and chemical industries. The main

advantage of gasification compared to direct combustion of fuel is that the gaseous products that are

called synthesis gas (syngas) or producer gas are likely more efficient and can be combusted at higher

temperatures. Syngas can be used directly or can be converted into synthetic fuel.

Coal is the common form of the fuel used in gasification plants; however, long-term availability of

coal and its carbon emission level have always been a problem [2,3,4]. Attempts to solve these problems

have drawn attention to biomass as a renewable energy resource. Biomass has a great potential to be

substituted for the Nation’s coal needs that potentially can reduce the amount of produced carbon dioxide

due to its carbon neutrality nature [5,6,7].However using biomass in existing gasifiers without incurring

2

increased costs in retrofits is not possible, and due to its lower calorific value and density compared to

fuels like coal, using biomass alone is not economically efficient [8].

As an effort to solve the common problems of using coal in gasification plants and benefit from

positive aspects of biomass as a fuel, co-gasification of coal and biomass is being considered [2]. To

create high quality producer gas, having enough knowledge about the process is extremely important.

However, a thorough understanding of the nature of coal-biomass gasification is not possible without

adequate knowledge of the fluidization behavior of the mixture. Therefore in this study, there will be

considerable focus on fluidization aspects of the co-gasification process.

1.2. Research objectives and approaches

The main goal of the current research is to computationally model fluidization and gasification of

coal, biomass and their mixtures to better understand their behavior using Multiphase Flow with

Interphase eXchanges (MFiX). The gas and solids phases are modeled using an Eulerian-Eulerian

approach. Both gas and solids are treated as a continuum and the solids phase constitutive equations are

modeled using kinetic theory for granular flows (KTGF) assuming instantaneous binary particle

collisions. In Eulerian modeling, the interaction between the gas phase and solids phases is coupled using

additional closure laws called drag models, which are inserted in the momentum equations.

For the first objective of the study, the fluidization of sub-bituminous coal, hybrid poplar wood and

their mixtures will be modeled. In this regards, the extensive study on particles mixing behavior of coal-

poplar wood mixtures will be performed, which is an important feature that can induce better particle

contact, increase the efficiency of the process and provide a relatively uniform temperature throughout the

fluidized bed. Efforts will include identifying the effects of biomass ratio on mixing behavior with coal.

Using Eulerian-Eulerian models to study mixing behavior will be a big step forward toward minimizing

the scale-up problems when modeling larger reactors. The predictions of bed height, pressure drop, gas

and solids volume fractions, mass flux and particle velocities will be examined and compared to the

experiments whenever possible.

The second objective of the study is to find the best drag model to predict the most reasonable results

for beds containing Geldart A particles, since both coal and biomass employed in this study fall in this

particle classification. Several studies reported over-prediction of bed expansion of fine particles

regardless of the drag model used. However, this study will shed light on the reason underlying this

problem and present the best drag model to use with Geldart A particles. Seven drag models reported in

the literature will be examined for single solid phases of coal and poplar wood, and then will be extended

to coal-poplar wood mixtures. The predictions of bed height, average and instantaneous void fraction and

velocity will be examined.

3

The third objective of the study involves proposing a novel method to improve the simulation results

for Geldart A particles fluidization. Capturing the homogeneous fluidization behavior of Geldart A

particles has been reported as a challenge using Eulerian-Eulerian models. The proposed method will

enable researchers to correctly predict the fluidization behavior of fine powders in all regions of

fluidization. The method will be examined for coal, poplar wood and switchgrass fluidized beds, and bed

height and pressure drop will be studied and validated with experiments.

Finally the last objective of the study is to model coal and switchgrass gasification and their co-

gasification. The co-gasification of coal and biomass has not been studied enough in the literature and

there are still questions that can be answered. Examples include the effects of gas and solid velocities,

particle diameter, gasification agent, and gasification temperature on the syngas products. Therefore,

these parameters will be studied and the synergetic effects of co-gasification on product gas yields will be

investigated.

1.3. Outline of the dissertation

The current study is mainly divided into two parts; the first part focuses on fluidization behavior of

coal and biomass mixtures and the second section concentrates on gasification behavior. In Chapter 2, a

broad literature review on fluidization and gasification is presented. Chapter 3 focuses on the governing

equations, the problem description and a grid resolution study. Chapter 4 tackles the topic of fluidization

from different approaches. At first, the chapter concentrates on finding the best drag model for beds

containing Geldart A particles. Specifically, sub-bituminous coal, hybrid poplar wood and their mixtures

were considered. Next, the numerical model is further validated by comparing the predictions of pressure

drop and bed height with experiments for sub-bituminous coal, hybrid poplar wood and their mixtures,

which are all classified as Geldart A particles. This section is continued by an extensive study on the

fluidization behavior of coal-poplar mixtures followed by particle mixing study for three mass ratios of

coal-poplar wood mixtures. Finally, an attempt to improve the simulation results of Geldart A particle

fluidization is presented. Chapter 5 describes the work related to coal and biomass gasification and the

effects of co-gasifying the materials. A 5-reaction model is validated by comparing the simulation results

with experimental data reported in literature for a coal particle classified as Geldart B. The effects of

velocity on coal gasification are investigated. Also, coal gasification of Geldart A particles and the effects

of the gasification agent on product gas yields are studied. The co-gasification of coal and switchgrass are

considered and the possible synergetic effects of co-gasification are investigated. Furthermore, the effects

of temperature on syngas composition are studied. In chapter 6, the summary and conclusions of the

research presented in this dissertation and possible future extensions to the current research are

mentioned.

4

Chapter 2. Literature review

Fossil fuels have been an unrivaled source of energy for centuries and still provide a majority of

energy demands. Coal as one of the common forms of fossil fuels is the largest domestically produced

source of energy in the United States, which supplies more than 50% of the Nation’s energy needs and

continues to be a more viable option due to unexpected changes in oil and natural gas prices [2,3].

However, as the world’s population and demand for energy have increased, serious problems related to

traditional fossil fuels have extensively arisen. Long-term availability of fossil fuels, their changing prices

and most importantly, their impact on the environment in terms of the undesired pollution, such as carbon

emission levels [4] and other greenhouse gases that has led to concerns about global warming are some

examples. Hence, these all expedite the need to lessen our dependence on fossil fuels and replace it, at

least partially, with a clean and renewable source of energy.

Biomass is being considered as a renewable energy resource, and includes a wide range of materials

from alive or recently living organisms. In recent decades, biomass has attracted attention as a possible

option to be substituted for fossil fuels. Biomass is a carbon neutral energy source since the amount of

produced carbon dioxide from biomass during the combustion is absorbed by energy crops during

photosynthesis [5,6,7]. The issue is that while biomass is environmentally friendly, it cannot be easily

used in existing gasifiers without incurring increased costs in retrofits. Furthermore, gasification of

biomass alone is not economically efficient or cost-effective because of its lower calorific value and

density compared to fuels like coal [8].

As efforts persist for the United States to effectively manage and use fossil fuels, and to address the

issue of sustainable energy sources while ensuring cost effectiveness and National security, co-

gasification of coal-biomass fuel mixtures is receiving attention as a viable option to be exploited. The

use of biomass has advantages in that it can potentially reduce carbon emissions and capitalize on using a

domestic feedstock [3]. Coal-biomass fuel mixtures have great potential to provide cleaner, low-cost and

domestic energy [9]. Biomass can be added to coal prior to gasification to reduce the amount of fossil fuel

required to provide energy for heating processes.

In recent decades, the use of coal-biomass gasification process to obtain producer and synthesis gas

has increased extensively. In order to truly understand gasification, there is a certain need to initially

comprehend the fluidization phenomena, which motivates our focus on properly modelling fluidization

and hydrodynamics. Gas-solids fluidization has widespread industrial importance by itself, however,

issues still persist related to mixing and segregation [10]. The gas bubbles that fluidize the granular

materials can enhance mixing or induce segregation. In addition, the presence of biomass can be

problematic during fluidization. Biomass particles are mostly irregular in shape, size, density, and

5

pliability. Due to these irregularities, biomass cannot be simply fluidized and characteristics of biomass

particles must be considered carefully to prevent particle agglomeration, elutriation, and segregation [11].

Additionally, since biomass materials are mostly waste materials, they can be in the shape of fine

powders, which may create fluidization problems and require extra attention.

To help characterize material fluidization behavior, particles have been classified into four groups

known as the Geldart particle classification [12] using material and gas density differences and mean

particle diameter. Geldart A particles have small diameters (20-100 µm) and/or low densities (< 1.4

g/cm3). The particles in this group expand significantly before bubbles commence and fluidize easily,

where bubble rising velocities are much higher than the gas inlet velocity. Geldart B particles classify a

broad group of materials that have larger diameters (40-500 µm) and medium densities (1.4-4.0 g/cm3).

The particles in this group also fluidize easily, where bubbles form at or slightly greater than the

minimum fluidization velocity. Unlike the two former groups, Geldart C particles that have very small

diameters (20-30 µm) are extremely difficult to fluidize due to strong inter-particle forces. Geldart D

particles have very large diameters (> 600 µm) and typically high densities. The velocities of the rising

bubbles are low and consequently, particle fluidization is difficult [12].

2.1. Fluidization

Knowledge about bed pressure drop and bed height is useful to assess the operation of fluidized beds.

For most materials, with increasing gas velocity at the distributor plate, pressure drop through the reactor

increases linearly until the critical value of velocity, which is called the minimum fluidization velocity, is

reached. Once the bed fluidizes, the pressure drop across the bed remains constant with increasing gas

velocity. Geldart A particles, however, manifest an interval of homogeneous fluidization (non-bubbling

expansion) before the onset of bubbling fluidization where the bed expands significantly. As inlet gas

velocity increases, bubbling or heterogeneous fluidization regime is reached, the bed becomes fully

fluidized and bubbles form. The critical velocity that demarcates the homogeneous and heterogeneous

fluidization regimes is called the minimum bubbling velocity (Umb) [12]. The presence of a homogeneous

fluidization region for Geldart A particles is demonstrated in pressure drop curves by an overshoot of

pressure near the minimum fluidization velocity (Umf). Such a spike in pressure has been reported by

some researchers [13,14,15]. Inter-particle forces, specifically adhesive forces and fluid-particles

interactions, are mostly held accountable as the mechanism underlying the homogeneous fluidization

[16,17,18,19].

Despite the wide application of fluidized beds of Geldart A particles, a thorough understanding of

their fluidization hydrodynamics still lacks and has always been a challenge for researchers. Numerical

models as a potential tool can be used to understand the fluidization behavior and help researchers to

6

design and troubleshoot the industrial units. Two groups of models have been used to model gas-solids

systems, Lagrangian models and Eulerian models.

Lagrangian models, or discrete particle models (DPM), compute the detailed motion of every particle,

including collision and external forces acting on the particles; however due to limitation of computational

resources, only a very small number of particles can be studied and as a result only small systems can be

simulated. Different hydrodynamic Lagrangian models have been proposed by researchers to model fine

particles of Geldart A classification. Here, only a few that focused on the existence of an overshoot of

pressure in fine particles fluidization are presented.

One of the early models was developed by Anderson and Jackson [20], which was a one-dimensional

DPM model. In order to simplify the model [20], the idea of local mean variables for fluid velocity and

pressure, and solids velocities were introduced. Later the model was employed by Jackson [13,14] to

show the effect of wall friction on fluidization behavior, especially, pressure drop overshoot for non-

cohesive materials. Srivastava and Sundaresan [21] found that the model proposed by Anderson and

Jackson [20] was able to capture the fluidization behavior of fine particles in restricted conditions for a

first order compressive yield stress equation. Loezos et al. [22] used an improved version of Anderson

and Jackson’s model [20] and proposed that not only wall friction but also compressive yield stress of the

particles can have dramatic effects on all fluidization-defluidization characteristics including the spike of

pressure drop. Their [22] results were confirmed by Ye et al. [15] who used a 3D DPM model and

proposed that fluidization behavior of fine particles is also influenced by inter-particles Van der Waals

forces.

Ye et al. [15] also studied the effect of sidewalls, particle density and size, gas density and viscosity

on fluidization behavior. They implemented two different wall boundary conditions, no-slip and free-slip,

on the gas phase, which indicated that neither bed height nor pressure drop showed large deviations. They

reported a direct relation between particle density and size, and the predicted ratio of the minimum

bubbling velocity to the minimum fluidization velocity. As particle density and size increased, the

velocity ratio decreased rapidly. Ye et al. [15] reported no relation between gas density and fluidization

behavior, however, an indirect relation between gas viscosity and the predicted value of minimum

fluidization velocity was reported.

The second group of models, Eulerian simulations, considers both gas and solids phases as a

continuum and use additional closure laws called drag models to describe particle-particle and particle-

wall interactions. A good comparison of different governing equations and closure laws used in the

Eulerian CFD models can be found in Van Wachem et al. [23]. Using an Eulerian approach has enabled

researchers to model large systems of realistic sizes and has been used extensively in modeling fluidized

beds by Battaglia [10,11,24]. The Eulerian approach, however, has shown some deficiencies in predicting

7

Geldart A fluidization behavior [23,25,26]. To improve two-fluid models to capture fine particles

fluidization behavior, many attempts have been made [27].

Many studies reported the existence of particle clusters in the gas-solid flow [28,29,30], which is the

result of inter-particles forces that is specifically important for Geldart A particles. As an effort to

improve prediction of Geldart A particles fluidization behavior using two-fluid models, some models

[31,32] based on the cluster movement of particles (cluster-based approach) were proposed that were in a

good agreement with experiments. The main difficulty in these models was to correctly predict the size of

clusters for different applications of the models. To eliminate this complication, some researchers tried to

use the traditional particle-based approach and modify the drag models to capture the importance of

particle clusters in Geldart A fluidization behavior.

McKeen and Pugsley [33] used the computational fluid dynamics code, Multiphase Flow with

Interphase eXchanges (MFIX), and the drag relation presented by Gibilaro et al. [34] to study freely

bubbling fluid catalytic cracking (FCC) beds and compared simulations to experiments. They [33]

reported severely over-predicted bed expansions and proposed that commonly-used models for coarser

particles only with modification on gas-solid drag closure law can be employed to correctly predict the

Geldart A particles fluidization behavior. They [33] confirmed that the reason is due to agglomeration of

particles and tried to solve the problem by using a modified drag law corresponded to a larger particle

diameter by applying a scale factor of 0.2 to 0.3, which leads to smaller drag forces acting on the

particles. The result found to be in a better agreement with experimental data. They [33] claimed that their

finding is in accordance with significant cohesive inter-particle forces of fine particles.

In accordance with the observations and conclusions by McKeen and Pugsley [33], Ye et al. [19,35]

emphasized inter-particles forces and suggested to modify drag model to correctly predict the

hydrodynamic behavior of Geldart A particles using two-fluid models. Ye et al. [35] used the Wen-Yu

drag closure law [36] to study a bed of particles with a diameter of 75 m and a density of 1.5 g/cm3.

They [35] suggested that the drag model must be scaled-down at low gas velocities and scaled-up at high

gas velocities and proposed that the scale factors account for the contribution of cohesion to the excess

compressibility of fine particles.

In contrast with these group of researchers who artificially modified the two-fluid models,

Parmentier et al. [37] investigated the influence of mesh size on fluidization behavior of particles and

found a drastic improvement of Geldart A predictions for sufficiently small cell sizes with no need to

modify the model or drag law. For mesh refinement studies, averaged bed height, axial and radial profiles

of solid volume fraction were considered. They [37] also proposed an empirical law to find the

appropriate mesh size to ensure mesh independent results. In spite of the Geldart A bed height decrement

tendency to match better with experiment as grid mesh size decreases, they [37] observed the drag model

8

dependency of the Geldart A bed height predictions. In this study, two drag models, Wen-Yu [36] and the

drag relation presented by Gibilaro et al. [34], were considered and it was observed that the former drag

law predicted larger final bed height than the later one.

Similarly, Wang et al. [38,39] suggested that the failure of two-fluid models does not stemmed from

the existence of clustering structures and instead, emphasized on the necessity of sufficiently fine grid

size and small time step to correctly predict the fluidization behavior of Geldart A particles using two-

fluid models. A bed of particles with a diameter of 75 m and a density of 1.29 g/cm3 was modeled using

both DPM and two-fluid model and bed height were studied. The predictions of two models agreed well,

however, no experimental results were used to validate the bed height predictions. Wang et al. [39]

studied minimum bubbling velocity values using both DPM model and experiment to validate the two-

fluid model predictions and found reasonable agreement in terms of the Umb dependency on the gas and

particles properties. However, with current and foreseeable computer resources, their method [37,38,39]

is very expensive if even feasible for a large-scale fluidized bed system.

Inherent heterogeneous nature of a bubbling bed of Geldart A particles and presence of particle

clusters over a wide range of length and time scales held responsible for failure of two-fluid models in

capturing fluidization behavior of fine particles by a group of studies. This group of researchers tried to

employ new drag models to count for these structures. Wang [27] emphasized the necessity of a suitable

subgrid-scale (SGS) model to deem the effect of meso-scale structures for successfully describing the

hydrodynamics of a fluidized bed containing Geldart A particles. Lu et al. [40] also concentrated on the

significant effect of meso-scale structures on Geldart A particles performance and proposed that not even

very fine grids are capable of capturing these structures. In contrast, they [40] suggested that using an

energy-minimization multi-scale (EMMS) drag model can correctly predict fluidization behavior of

Geldart A particles. In their study, a bed of particles with a diameter of 75 m and a density of 1.5 g/cm3

was considered, and simulation results of using two drag models, an EMMS and the Gidaspow (hybrid)

drag model, were compared with experiments. Lu et al. [40] reported the failure of the Gidaspow drag

closure law to correctly predict the axial and radial profiles of void fraction, and also solid flux even when

fine grids were used, whereas, EMMS drag model present quantitatively and qualitatively reasonable

predictions even with coarse grids. Benzarti et al. [41] reported good performance of an EMMS drag

model in predicting the bed expansion of Geldart A particles compared to the Gidaspow and the Syamlal

and O’Brien drag models, which severely overestimated the drag force on a FCC bed of particles.

The goal of the first part of the current research is to investigate the fluidization behavior of coal-

biomass particles that are classified as Geldart A using simulations and validate our results with

experiments. Notwithstanding the fact that numerous experimental studies on the fluidization of Geldart

A particles have been conducted [42,43,44,45,46] and numerical studies on FCC particles that are

9

classified as Geldart A have been reported [47,48,33,49,50], a thorough understanding of fluidization

hydrodynamics of fine particles still lacks. To the best of our knowledge, so far there have been no

published numerical studies specific to Geldart A biomass or Geldart A coal particle fluidization as a

mixture. (Experimental studies of larger biomass particle fluidization processes were summarized by Cui

and Grace [51].) As an attempt to truly understand the hydrodynamics of coal-biomass mixtures first part

of this study is dedicated to fluidization.

2.2. Gasification

Despite the significant number of numerical studies on gas-solid flow in fluidized beds, there have

been a very limited number of numerical studies on these flows coupling with chemical reactions.

Therefore, the attempt of gasification section of this study is to better understand the gasification

behavior. Although the gasification section mostly concentrates on coal gasification, it is tried to prepare

a broad literature review on coal-biomass gasification. Notwithstanding the differences between coal and

biomass structure and chemical composition, the model structure and mathematical description to

simulate gasification are not different [52]. In modeling gasification, complicated chemical reactions must

be introduced to the fluidization model through kinetic expression of each reaction. Such expressions

depend upon the specific chemistry of reaction and must be found experimentally.

To model gas-solids flows with chemical reactions, two previously-introduced groups of models,

Eulerian and Lagrangian, have been used by researchers. However, due to the high number of

computational calculations in presence of chemical reactions, the current computational resources and the

computationally expensive nature of Lagrangian models, these models cannot be used successfully for

large systems with chemical reactions. The Eulerian approach, on the other hand, has been used

successfully to model gasification.

Yu et al. [53] used a two-fluid model including the kinetic theory of granular flow (KTGF) to model

coal gasification in a fluidized bed. They [53] reported a good agreement between the gas composition

simulation results and experiments. In their [53] study, reaction rates were determined by the combined

effects of Arrhenius rates and turbulent mixing rates for homogeneous reactions and diffusion rates for

heterogeneous reactions. Wang et al. [54] also developed a 3D two-fluid numerical model to simulate the

coal gasification in a fluidized bed using k-ε turbulent model, KTGF, and considered coal pyrolysis. To

find the chemical reaction rates for homogeneous reactions, Arrhenius-Eddy dissipation reaction rates and

for heterogeneous reactions, Arrhenius-diffusion reaction rates were employed. Results for gas

compositions were compared with experiments and were in good agreement. Kajitani et al. [55] studied

the gasification of brown coal and concentrated on its inherent catalytic effect compared to high-rank

coals. They [55] employed the Langmuir-Hinshelwood (LH) reaction rate equations for gasification and

10

suggested that elementary char gasification and volatile-char interactions consist of free radical absorption

from volatiles, evolution of the char structure by these free radicals, and catalyst volatilization. Their

results on coal conversion and Na concentration of char were in good agreement with experiments.

Gasification of biomass as a replacement for coal has also been studied by some researchers.

Although most studies used Lagrangian models to simulate biomass gasification [56,57], there are some

studies reported the effectiveness of Eulerian models. Gerber et al. [58] employed an Eulerian two-fluid

model using KTGF and Arrhenius reaction rates to simulate wood gasification in a fluidized bed. They

[58] compared the product gas composition, tar concentration, and temperature with experiments and

found a strong influence of thermal boundary conditions on CO and CO2 components of syngas, and

operating conditions on tar content.

The number of studies on coal-biomass mixture gasification is limited and there is an obvious lack of

thorough and accurate understanding of coal and biomass interactions during co-gasification. Some

studies have revealed that coal and biomass mixtures demonstrate synergistic behavior in most aspects

[59,60,61], while others have revealed no synergy or no significant synergy between coal and biomass

[62]. A postulated explanation for the synergistic behavior during co-gasification is hydrogen transfer

between biomass and coal. The process starts with early reduction of biomass, which is due to the weak

covalent bonds in biomass and its higher content of oxygen than coal. Consequently, biomass volatiles

release, break down and form free radicals or react with oxygen present in the system and produce gases.

The free radicals from biomass react with coal and enhance coal decomposition, which release free

radicals from coal. Then, light volatile molecules and the producer gases (from cracking heavy volatiles

that have a high content of hydrogen) react with coal free radicals as hydrogen donors, hence, coal free

radicals are prevented from recombination reactions. As the result, the amount of less-reactive secondary

char in the system is reduced [63,64]. In addition, due to high calorific value of coal volatiles, temperature

in such a system increases, therefore, tar1 cracking and biomass and coal endothermic conversion into

gases is enhanced, which all result in a net increase in volatile yields, higher fuel conversion rates and less

char [64]. In spite of many research that confirms the hydrogen transfer between coal and biomass

[65,66,67,68], there is still lack of thorough understanding of free radical formation and evolution during

pyrolysis, which is the key to the synergistic behavior between coal and biomass. To overcome this

inadequacy, in situ observation of free radicals released during the pyrolysis process is essential. In spite

of considerable numbers of studies on the subject for coal [65,69,70], there is dearth of research on

biomass or biomass components.

1 tar is a high molecular weight gaseous compound that is produced by biomass and coal.

11

The possible synergetic effect during co-gasification can also possibly be explained through catalytic

effects. Due to the high content of alkali (K+ and Na

+) and alkaline earth (Ca

2+) metals in some biomass,

and a clearly known effect of these metals as catalyst on coal gasification, co-gasification of coal and

biomass is expected to benefit and show some catalytic behavior, which has been addressed by some

studies [71,72,73].

To investigate the synergetic effect of co-gasification, different aspects of gasification process have

been studied. Aghalayam et al. [74] studied the activation energy and efficiency of co-gasification process

and found synergistic behavior between coal and biomass. They [74] reported that the activation energy

of a coal and biomass blend is lower than activation energy of either coal or biomass separate beds, which

results in faster reaction time and less gasification agent for coal or biomass mixtures than coal or

biomass. Edreis et al. [61] confirmed a synergetic effect on activation energy and also suggested that co-

gasification of coal and biomass increases the net power outlet compared to coal gasification. They [61]

found that the net power outlet does not increase linearly with biomass blending ratio. Long and Wang

[131] results show a slight improvement of efficiency in co-gasification of coal and biomass process.

Kumabe et al. [130] and Franco et al. [75] observed an increase in energy content of the syngas or

conversion efficiency as biomass mass ratio increases.

The goal of gasification section of this study is to truly understand the nature of coal gasification

behavior and provide enough background for follower researchers to continue the work presented here to

investigate the co-gasification behavior of coal and biomass fuel mixtures.

12

Chapter 3. Numerical Formulation

3.1. Governing equations

In the present study, the code Multiphase Flow with Interphase eXchanges (MFiX) is employed for

all simulations [76]. The gas and solids phases are modeled using an Eulerian-Eulerian approach to

efficiently simulate the physics during fluidization and gasification, which has been used extensively by

Battaglia [e.g.,10,77,78]. Each phase behaves as interpenetrating continua that has its own physical

properties. In order to derive equations that describe the phase interaction, the instantaneous variables are

averaged over a region that is larger than the particle spacing but smaller than the flow domain.

To track the volume occupied by each phase, volume fractions are introduced where g is the gas

volume fraction (void fraction) and s is the solids volume fraction. The volume fractions must satisfy:

1∑1

M

m

smg (1)

where m is the solids component and M is the total number of solids phases in the system. Continuity

equations are solved for the gas phase and solids phases. Considering generation or destruction within the

phases, the continuity equations for the gas and solids phases, respectively, are:

∑1

)( ∇)( ∂

∂gN

n

gnggggg Rut

(2)

∑1

)( ∇)( ∂

∂smN

n

smmsmsmsmsmsm Rut

(3)

where is density, u

is velocity vector, and Rgm and Rsmn are the rate of formation of gas species n and of

solid phase-m, respectively. The gas and solid phases may contain an arbitrary number of chemical

species, N. In both Equations (2) and (3), the terms on the left account for the rate of mass accumulation

per unit volume and the net rate of convective mass flux, respectively. The term on the right consider

interphase mas transfer due to chemical reactions or physical process, such as evaporation. For

fluidization cases with no chemical reactions or physical process, the term on the right of both Equations

(2) and (3) is zero. Gas density can be modeled using the ideal gas law:

ggg TRMWp (4)

Where pg is the gas pressure, MW is the gas molecular weight, R is the universal gas constant and Tg is the

gas temperature.

The momentum equations for the gas phase and mth solids phase, respectively, are:

13

gu-uFp-uuut

gg

M

m

gsmggggggggggg

∑1

)( . ∇.∇)(.∇)(∂

∂

(5)

∑

≠

1

)(- .∇.∇)(. ∇)(∂

∂M

ml

l

mlsmsmgsmgsmgsmsmsmsmsmsmsmsm Igu-uFp-uuut

(6)

The terms on the left hand side represent the rate of momentum increase and net rate of momentum

transfer by advection, respectively. Terms on the right hand side represent contributions of buoyancy

caused by the fluid pressure gradient, the viscous stress tensor g in Eq. (5), the solids phase stress tensor

sm (which is a summation of the pressure and viscous stress tensors) in Eq. (6) and, gas and solids

phases interaction, and gravity. The last term in Eq. (6) is the particle interaction force between the mth

and lth solids phases. In order to find the drag coefficient to calculate the solids-solids momentum transfer,

a simplified version of kinetic theory for granular flow proposed by Syamlal [79] is used.

The gas-solids momentum transfer in Eqs. (5) and (6) is expressed as a drag force and is defined as

the product of Fg that is the coefficient of interphase drag force between the gas and the solids phases and

the slip velocity between the phases, )( gsm u-u

. To develop a relation for Fg, many experimental

methods have been proposed [76,80]. A detail explanation for different drag models will be presented in

section 4.3, where fluidization behavior predictions using different drag models will be compared and the

best drag model for coal, biomass and their mixtures will be proposed. Herein for the rest of this study,

the Gidaspow model [80] is employed and the accuracy of using this model to predict biomass

fluidization was confirmed by Estejab et al. [81]. To calculate the interphase drag force coefficient, the

Gidaspow model uses two correlations depending on the local void fraction coupled with blending

function [82] when switching between correlations [83].

In case of existence of momentum transfer due to mass transfer in cases with chemical reactions or

physical process, another term will be added to Equations (5) and (6),

][ 000 gmsmmm uuR

(7)

where mR 0 is mass transfer from the gas phase to solid phase- m and

00

01

0

0

0m

m

mRfor

Rfor (8)

and mm 00 1 (9).

The conservation of internal energy for the fluid phase and for the mth solids phase is given by:

14

)( gwallwallrggmggg

g

pggg TTHHHqTut

T

C

(10)

rsmgmsmsmsmsm

psmsmsm HHqTut

TC

(11)

where the first term in both Equations (10) and (11) is the conductive heat flux, the second term accounts

for fluid-solids interphase heat transfer, and the third term describes the heat of reaction. The last term in

the Equation (10) accounts for the heat loss to the wall.

The granular temperature for the solids phase is proportional to the granular energy defined as the

specific kinetic energy of the random fluctuating component of the particle velocity. In rapid granular

flow, the kinetic energy of random particle fluctuations is derived from the kinetic energy of the mean

flow [76,84]. It should also be noted that the solids stress tensor sm shown in Eq. (6) is defined in terms

of a plastic regime and a viscous regime, where the viscous regime is a function of θ. Further details of

the constituent relationships can be found in [76].

The resulting transport of granular energy for the mth solids phase is:

gmsmsmsmsmsmmsmsm quut

.:).()(

2

3

(12)

where q

is the diffusive flux of granular energy, is the rate of granular energy dissipation due to

inelastic collision, and gm is the transfer of granular energy between the gas phase and the mth solids

phase.

The species conservation equation for gas and solids phase m, which consider accumulation,

convection, and rate of reactions, are:

gnggngggngg R)=uXρ(ε)+Xρ(εt

.

(13)

smnsmsmnsmsmsmnsmsm R)=uXρ(ε)+Xρ(εt

.

(14)

where Xgn and Xsmn are mass fraction of gas species n, and of solid phase-m, respectively. The species

conservation equations must only be considered when chemical reactions or physical process is involved.

To reduce numerical instabilities, the governing equations are discretized using a finite volume

approach for staggered grids. Discretizations of temporal and spatial derivatives are first-order and

second-order, respectively. To improve the convergence and accuracy of the solution, the convection

terms are discretized using the Superbee method [85]. To solve the governing equations, MFIX uses a

modification of Semi Implicit Method for Pressure Linked Equations (SIMPLE) algorithm proposed by

Patankar [86], which is the most widely used primitive-variable method. The first modification replaces

15

the solids pressure correction equation with a solids volume fraction correction equation, which

incorporates the effect of the solids pressure and helps to stabilize and facilitate convergence of both

loosely and densely packed regions. To increase the execution speeds, a second modification introduces a

variable time-stepping scheme. The number of iterations per time step is monitored, and if convergence

occurs in less than the maximum number of iterations, the time step is doubled. If convergence is not

satisfied, the time step is halved and the calculation is repeated. The process continues until convergence

is satisfied [84].

3.2. Problem description and grid resolution study

Herein in this study, the cylindrical fluidized bed reactor used had a 5.08 cm inner diameter and 30

cm height, illustrated in Figure ‎3.1. Nitrogen gas was fed into the reactor from the base of the distributor

plate and the reactor discharged to the atmosphere [87]. The reactor is modeled in a Cartesian coordinate

system as a 2D plane representing the center-plane of the cylinder. The gas inlet velocity at the bottom of

the reactor, i.e., the inlet of distributor plate, is defined as a uniform velocity and atmospheric pressure is

specified at the exit. To model the gas-wall and particle-wall interactions, no-slip and partial-slip

boundary conditions are used, respectively.

The grid resolution study is performed for the pressure drop of single-component system of sand for

gas inlet velocity Ug = 14.8 cm/s and single-component system of poplar wood for inlet gas velocity Ug =

8.22 cm/s using the geometry that is presented in Figure ‎3.1. The rectangular domain is discretized

uniformly using a coarse mesh (20×100), medium mesh (40×200), and fine mesh (80×400). The cell sizes

along with the central processing unit time are shown in Table ‎3.1. The numerical uncertainty due to

discretization is estimated by analyzing the grid convergence index (GCI) [88]. The GCI values for the

predicted pressure drop for sand are 3% and 1.5% changing from coarse to medium and medium to fine

mesh. To ensure that the grid resolution is acceptable for different particles, a grid resolution study for

poplar-wood is also performed and the GCI value is 3% and 1% for coarse-medium and medium-fine

grids, respectively. Since the medium mesh is less computationally expensive compared to the fine mesh,

the medium mesh will be used for the following cases.

16

Figure 3.1. The 2D plane representing the chamber of the cylindrical reactor.

Table 3.1. Grid resolution and central processing unit.

Mesh 20 100 40 200 80 400

# Cells 2000 8000 32000

Cell size (cm2) 0.254 0.3 0.127 0.15 0.0635 0.075

CPU (s) for sand 1,031 4,079 42,919

CPU (s) for poplar wood 2,270 15,611 107,966

17

Chapter 4. Fluidization

Section 4.2 of this chapter is produced from [81] with permission from Journal of Fluids Engineering:

Estejab, B., Battaglia, F., 2015, “Assessment of Drag Models for Geldart A Particles in Bubbling

Fluidized Beds,” Journal of Fluids Engineering, 138(3), pp. 031105-031105-12.

4.1.Fluidization theory

To predict fluidization, pressure drop can be used to determine expected trends. With increasing gas

inlet velocity, pressure drop increases linearly until the bed begins to fluidize at the minimum fluidization

velocity, Umf. For fine particles of Geldart A classification, after this point, increasing the velocity to

higher values expanded the bed without any visible sign of bubbling until the bed begins to

heterogeneously fluidize at the minimum bubbling velocity, Umb. Typically, after this point, velocity

increment does not affect the pressure drop and the value remains constant. The fluidization pressure drop

can be derived from a force balance neglecting cohesive forces:

AmgP (15)

where mg is the bed weight and A is the reactor cross-sectional area normal to the incoming flow. Bulk

density of the solids phase can be defined using:

0hAm smb sm (16)

where 0h is the bed initial height. The initial gas void fraction (

mf ) can be found using the volume

fractions:

M

m

smmf

1

1 (17)

where smbsm sm (18).

To estimate the minimum fluidization velocity, a simplification of the original Ergun correlation [89] can

be used:

mfg

mfgsp

mf

gdU

1150

32

(19)

where pd is the mean diameter of all solids particles and is the estimated sphericity of the actual

particles. To develop Equation (19), it was assumed that the buoyancy and drag force acting on a particle

are equal.

18

Using the measured experimental data such as mass and initial bed height as the initial conditions to

the simulations, the predicted pressure drop does not always agree with experiments. Kanholy et al. [24]

proposed that the reason is due to regions where material remains static and does not contribute in the

fluidization process. To overcome this difficulty, they correlated experimental void fraction and particle

bulk density, and adjusted mass and initial bed height using experimental pressure drop with Eqs. (15)

and (16). Sphericity was estimated using the experimental values of Umf and mf. from Eq. (19). For the

work herein, the corrected data for all three materials are presented in the simulation columns of Table ‎4.1

(columns 5-7). It is important to note that if all the mass present in the bed participates in the fluidization

process, the required pressure drop to fluidize the bed is higher than the experimental value reported. As

an example, the pressure drop for coal and poplar wood using Eq. (15) is found to be 225 Pa (the initial

mass for both material is the same) instead of 75 Pa and 69 Pa reported by experiments for coal and

poplar wood, respectively. This will be further discussed in Section 4.3.

To validate the fluidized bed modeling and correspondence with the experiments, sand was

considered first because it is a well-characterized material and falls within the Geldart B classification.

Time-averaged pressure drop across the bed for the simulations and experiments versus the inlet gas

velocity are presented in Figure ‎4.1(a). After fluidization at Umf = 6.3 cm/s, the predicted pressure drop

remains constant at 465 Pa, which is consistent with the experiments (458 Pa). To further investigate sand

fluidization, the time-averaged dimensionless height of the sand bed for different gas inlet velocities is

presented in Figure ‎4.1(b). To compare the experimental and simulated bed height, a dimensionless

parameter is defined that is the ratio of expanded bed height to the initial bed height. The simulations

agree well with experiments, which confirm the validity of the simulation modeling.

Table 4.1. Properties for sand, coal, and poplar

Properties Experiment [90] Simulation

Sand Coal Poplar wood Sand Coal Poplar

�̅�𝑝(µm) 251 ± 0.17 62 ± 0.51 151 ± 1.21 251 62 151

𝜓 0.93 ± 0.01 0.95 ± 0.02 0.59 ± 0.05 0.93 0.95 2

𝜌𝑠 (g/cm3) 2.62 ± 0.01 1.38 ± 0.01 0.4 ± 0.02 2.62 1.38 0.4

𝜌𝑏 (g/cm3) 1.51 ± 0.04 0.57 ± 0.01 0.25 ± 0.02 1.51 0.57 0.25

m (g) 100 30 30 94.6 15.5 14.2

𝜀𝑚𝑓 0.42 ± 0.001 0.59 ± 0.005 0.38 ± 0.001 0.42 0.59 0.38

19

(a) (b)

Figure 4.1. (a) Pressure drop versus the gas inlet velocity comparing experiments [90] and simulations, and

(b) the dimensionless bed height for different gas inlet velocities comparing experiments [90] and simulations

for sand.

4.2. Drag models 2

4.2.1. Introduction

In order to accurately predict the hydrodynamic behavior of gas and solid phases using an Eulerian-

Eulerian approach, it is crucial to use appropriate drag models to capture the correct physics. In this study,

the performance of seven drag models for non-reacting fluidization of Geldart A particles of coal, poplar

wood, and their mixtures was assessed. Notwithstanding the notable number of numerical studies to find

the best drag model for larger particles [40,91,92], there is a dearth of information related to drag models

for finer Geldart A particles. Furthermore, previous findings bode badly for using predominately Geldart

B drag models for fine particles, despite their excellent performance predicting larger particle fluidization

behavior. Many of the existing studies on this subject reported over-prediction of bed expansion of fine

particles regardless on the drag model used [25,93,94]. Additionally, to our knowledge, these drag models

have not been tested with a binary mixture of Geldart A particles. Seven drag models considered are:

Beetstra, van der Hoef, and Kuipers (BVK), Holloway, Yin, and Sundaresan (HYS), Hill-Koch-Ladd

(HKL), Gidaspow, Huilin-Gidaspow (Gidaspow-Blend), Syamlal-O’Brien, and Wen-Yu drag models. All

drag models are explained in detail in following.

2 This chapter is produced from [81] with permission from the Journal of Fluids Engineering.

Ug

(c m /s )

p

(Pa

)

0 2 4 6 8 1 0 1 20

1 0 0

2 0 0

3 0 0

4 0 0

5 0 0

E x p e r im e n t

S im u la t io n

Ug

(c m /s )

h/

h0

0 2 4 6 8 1 0 1 20

0 .5

1

1 .5

2

E x p e r im e n t

S im u la t io n

20

The results of our study revealed that if static regions of mass in the fluidized beds are considered,

these drag models work well with Geldart A particles. It was found that drag models derived from

empirical relationships adopt better with Geldart A particles compared to drag models that were

numerically developed. Overall, the Huilin–Gidaspow drag model showed the best performance for both

single solid phases and binary mixtures, however, for binary mixtures, Wen-Yu model predictions were

also accurate.

4.2.2. Drag models

The gas-solids momentum transfer for cases without chemical reaction or physical process that is

shown in Eqs. (5) and (6) is drag force, which for beds with no chemical reaction or physical process is

the product of the interphase drag force coefficient Fg between the gas and the solids phases and the slip

velocity between the two phases )( gsm u-u . To find Fg, many experimental methods have been proposed

[76,79], therefore, dissimilarity between drag models originate from different correlations used to develop

a relation for Fg. In this study, we are trying to find the best drag model proposed for Geldart A particles

and their mixtures. To calculate the interphase drag force coefficient, each model uses its own

correlations, which will be explained next.

4.2.2.1. Wen-Yu drag model (1966)

Wen and Yu [36] developed an empirical correlation to represent the drag coefficient. The model is

valid for all ranges of void fraction, however, defines two drag coefficients for different ranges of

Reynolds number. The drag coefficient is given as:

65.2

75.0-

ggs

p

sgg

DYuWeng u-ud

CF

(20)

where the relationship for drag over a sphere is:

1000≥Re44.0

1000Re)Re(Re15.0124687.0

ggDC

(21)

g

pgsg duu

-

Re (22)

4.2.2.2. Syamlal-O’Brien drag model (1988)

In the Syamlal-O’Brien drag model [95], the coefficient of interphase drag force, gF , is:

21

gs

rp

sgg

Dg u-u

d

CF

2

υ4

3 (23)

where the particle correlation [96] is:

2

υRe

8.463.0

r

DC (24)

the relative velocity correlation [97], r , is:

22)2(Re12.0Re06.0

2

1Re06.0

2

1υ AA-B-Ar

(25)

where

14.4

gA (26)

85.0

85.0≤8.0

65.2

28.1

gg

ggB

(27)

The Reynolds number appearing in Eqs. (24) and (25) is particle Reynolds number and is calculated using

Eq. (22).

4.2.2.3. Gidaspow drag model (1994)

In Gidaspow drag model [80], for void fraction regions less than 0.8, the Ergun equation [89] is used

and is given as:

gs

p

sg

pg

ggs

Ergung uudd

F

-75.1)-1(150

2

(28)

and for void fraction regions equal or larger than 0.8, the Wen-Yu equation [36], shown in Equations.

(20)-(22), are used.

4.2.2.4. Huilin–Gidaspow drag model (2003)

The model developed by Huilin and Gidaspow, known as Gidaspow-blend drag model in MFIX [82], is

similar to the Gidaspow model and uses the same equations. The difference between the two models lies

in an equation that is used to combine the results of the Ergun and Wen-Yu drag models to “blend” or

smooth the transition when the particle characteristics transition from one regime (dilute) to the other

(dense) [83]:

)-()()-1( YuWenggsErgunggsg FFF (29)

22

where the blending function, gs [82] is:

5.0)2.0(75.1150arctan sgs - (30)

4.2.2.5. Hill-Koch-Ladd (HKL) drag model (2006)

The Hill-Koch-Ladd drag model used Lattice-Boltzmann simulations in the development, which required

significant computational effort [98, 99]. The model was developed in an effort to obviate the need for

empirical drag relations, for which there is no consensus on their performance. The original drag model

proposed was modified later to cover a full range of Reynolds numbers and void fractions experienced in

fluidized beds [100], which is used in this study. The drag force model is:

2

218

p

sggg

d

FF (31)

where F is a dimensionless drag factor, which is a piecewise function and is given as:

2

3

2

3 3 1 0 2

0 1

1

2 3

-11 3 / 8 R e 0 .0 1 a n d R e

( 3 8 - )

- 4 ( - )R e 0 .0 1 a n d R e

2

R e

r s r

r s r

r

F

F

F F F F FF F F

F

F F O th e r w is e

(32)

The coefficients are defined as:

4.0≥

)1(

10

4.001.0

)1(

10

16.84.8681.01

14.17)(ln)64135(231)1(

3

332

0

s

s

s

s

s

s

sss

ssss

-

--

-

F

(33)

1.0)6.11(exp00051.011.0

1.0≤01.0402

1

ss

ssF

(34)

23

4.0≥

)1(

10

4.0

)1(

10

41.1503.11681.01

89.17)(ln)64135(231)1(

3

332

2

s

s

s

s

s

s

sss

ssss

-

--

-

F

(35)

0953.0≥)1(/0232.0212.00673.0

0953.003667.0593.0

53

sss

ss

-F

(36)

where and rRe is defined as:

)/)4.0(10(exp ss-- (37)

g

pgsg

r

duu

2

-

Re

(38)

4.2.2.6. BVK drag model (2007)

The drag model developed by Beetstra, van der Hoef, and Kuipers [101], referred to as the BVK model,

also used Lattice-Boltzmann simulations. To control numerical predictions, the model employs Reynolds

numbers and packing fraction. The BVK drag model was suggested for a range of Reynolds numbers up

to 1000. The proposed model for bi-dispersed systems calculates the individual drag force on a particle of

mth solids phase, where:

Re10640118

32

2,εFyε-.yεyε-

d

εμF smsmsms

p

gg

gm

(39)

and

g

sgg

μ

dε-uρ 1Re

(40)

∑

∑2

3

mpmm

mpmm

dn

dn

d (41)

d

dy

pm

m (42)

ss

ε-.-ε

.-

sg

-

g

s

ss

s

ss

.εεε

ε-

.ε.ε-

ε-

ε,εF

2503

34301

2

2

2Re101

Re483

124

Re41305111

1

10Re

(43).

24

4.2.2.7. HYS drag model (2010)

The drag model developed by Holloway, Yin, and Sundaresan [102], known as the HYS model, is an

extension to the drag relation by Yin and Sundaresan [103] proposed for low Reynolds number poly-

dispersed beds. The HYS drag model uses the inertial correction proposed by Beetstra et al. [101] to

enable the effect of finite Reynolds numbers on particle drag by a fluid in poly-dispersed beds. In this

model, it is proposed that Fg must have an extra term that was commonly neglected to account for the

drag force experienced by each solids phase due to the relative motion of other solids phases:

∑≠ lm

lmmlmmgm U-UUF (44)

where Fgm is the fluid-particle drag on the mth solids phase. m and ml are dimensional friction

coefficients for the mth solids phase due to the relative motion of the lth solids phase. m is given by:

*

-2

)1(18

fixedgm

pm

gsms

m F

d

-

(45)

where *

- fixedgmF is the dimensionless drag on a particle of the mth solids phase for moderate Reynolds

numbers, is:

BVKss

s

smixmsfixedgm -

-

yfF

15.11)1(

)1(

10Re,,

2

2

*

- (46)

2/)41(3

343.01

4Re101

Re4.8)1(3)1(

5.11)1(24240

Re413.0

ss -

mix

-

mixss-

s

sss

mixBVK

--

-

(47)

where

dU smix

mix

)-1(Re

(48)

∑

∑

msm

mmsm

mix

U

U

(49)

and d and my are calculated using Eqs. (41) and (42), respectively.

The friction coefficient ml is given by:

)/()/(249.1

),(minlog3131.12 10

llsmms

lsmslpmp

ml

dd

(50)

25

where is the lubrication cut-off and is mostly assumed to be 1 µm, which is roughly the mean free path

of a gas.

4.2.3. Results and discussion

The basic properties and parameters measured in the experiments are shown in Table ‎4.1 for coal and

poplar wood. Considering the particle properties presented in Table ‎4.1, both coal and poplar wood

correspond to Geldart A classification. In this study, binary mixtures of coal-poplar are also studied for

mass ratios of 90:10, 80:20 and 70:30. To simulate the binary mixtures, a model (Sim-B2) is adopted

from Kanholy et al. [24] that considers two solids phases. The model uses the particle diameters presented

in Table ‎4.1 to calculate the mixture properties. To find a characteristic single solids phase diameter and

density the following mixture equations were employed:

1

1

M

i

iiimixmix dXd (51)

1

1

M

i

iimix X (52)

where Xi = mi/M is the mass fraction of component i. Similar to the procedure presented in the

fluidization theory (section 4.1), experimental pressure drop and Eq. (15) are used to adjust the mass.

Mass for each component is then found using Xi.. The initial void fraction is found by substituting the

mixture properties (Eqs. 51 and 52) and the experimental minimum fluidization velocity into Eq. (19) and

solving for mf. The bulk density is then found using Eqs. (17) and (18), and the adjusted initial bed height

is found by substituting the adjusted value of mass into Eq (16). All properties for the simulations for

each mass ratio are presented in Table ‎4.2

Table 4.2. Properties for binary mixture model

Properties

Experiment [90] Simulation

90:10 80:20 70:30 90:10 80:20 70:30 Coal Poplar Coal Poplar Coal Poplar Coal Poplar Coal Poplar Coal Poplar

�̅�𝑝(µm) 62 151 62 151 62 151 62 151 62 151 62 151

𝜓 0.95 0.59 0.95 0.59 0.95 0.59 0.95 2 0.95 2 0.95 2

𝜌𝑠 (g/cm3) 1.38 0.4 1.38 0.4 1.38 0.4 1.38 0.4 1.38 0.4 1.38 0.4

𝜌𝑏 (g/cm3) 0.57 0.48 0.43 0. 51 0.06 0.38 0.1 0.3 0.13

m (g) 27 3 24 6 21 7 16.1 1.8 21 5.3 16.6 7.1

𝑋𝑖 0.9 0.1 0.8 0.2 0.7 0.3 0.9 0.1 0.8 0.2 0.7 0.3

𝜀𝑚𝑓 0.49 0.48 0.45 0.49 0.48 0.45

26

It was discussed in section 2.1 that for Geldart A particles, with increasing gas velocity at the

distributor plate, pressure drop through the reactor increases linearly, while the bed remains relatively

static until reaching the critical velocity, which is called minimum fluidization velocity, Umf. For

velocities just above Umf, particles display an interval of homogeneous fluidization (non-bubbling

expansion) before the onset of bubbling fluidization at minimum bubbling velocity, Umb. After this point,

however, Geldart A particles exhibit heterogeneous fluidization similar to other materials (e.g., Geldart B

particles) and bed height increases continuously. Therefore, knowledge about bed height during the

fluidization process is useful to assess the fluidization behavior. The fluidization experiments for coal and

poplar wood show that minimum fluidization velocity occurs at Umf = 1.6 cm/s and Umf = 3.2 cm/s,

respectively. The onset of heterogeneous fluidization occurs at Umb = 6.6 cm/s and Umb = 8.2 cm/s for coal

and poplar wood, respectively. Deza et al. [78] found that for inlet velocities smaller than four times the

velocity at the point of the fully fluidized region, no turbulence model or 3D modeling is required.

Considering the range of inlet velocities in this study, a turbulence model is not employed.

The fluidization theory (section 4.1) referred to the existence of regions where material remain static

and do not contribute in the fluidization process, and Kanholy et al. [24] proposed a method to correlate

experimental data to adjust the mass in the bed. To illustrate this point, simulations are performed using

the experimental conditions (as is) and using the corrected data (the corrected data for both coal and

poplar wood and their mixtures are presented in the simulation columns of Table ‎4.1 (columns 6-7) and

Table ‎4.2, respectively). Predicted bed height for a single material (poplar wood) and a binary mixture

(70:30 coal-poplar wood) is presented in Figure ‎4.2, comparing the experiments to the simulations

employing the Gidaspow-blend drag model. The simulations that use the experimental conditions, as is,

represent modeling all of the material, which includes the static regions that do not fluidize. The

simulations based on the corrected data have effectively modeled only the mass that fluidizes. The

predicted bed height using measured experimental data over-predicts the bed height, whilst, the

simulations using adjusted experimental data agree well with the experiments. The results confirm the

importance of considering static regions of mass in fluidized beds. In fact, this finding is very important

for the following discussion related to drag models. As mentioned in section 2.1, drag models based on

Geldart B particles were not found to be suitable for Geldart A particles. However, using Eqs. (15) and

(16) to model only the material that fluidizes tempers those original conclusions by others.

27

(a)

(b)

Figure 4.2. Bed height of (a) poplar wood and (b) 70:30 coal-poplar wood mass ratio comparing predictions

for adjusted and unadjusted initial conditions to experiments [90]. For both simulations, the Gidaspow drag

model is used.

‎‎

Ug

(c m /s )

h/h

0

0 2 4 6 8 1 0 1 20 .5

1

1 .5

2

2 .5

3

3 .5

4

4 .5E x p e r im e n t

S im u la t io n w ith o u t s ta t ic re g io n s

S im u la t io n w ith s ta t ic re g io n s

Ug

(c m /s )

h/h

0

0 2 4 6 8 1 0 1 20 .5

1

1 .5

2

2 .5

3

3 .5

4


S im u la t io n w ith o u t s ta t ic re g io n s

S im u la t io n w ith s ta t ic re g io n s

28

4.2.3.1. Single solids phases of coal and poplar

The analysis begins by studying single solid phases of coal and poplar wood to determine how they

behave individually before the binary mixture study. The fluidization experiments show that pressure

drop across the bed in the fully fluidized region is constant, where Δp = 75 Pa and Δp = 69 Pa for coal

and poplar wood, respectively. Simulations predict a similar trend where pressure remains constant in this

region at 79 Pa for coal and 71 Pa for poplar wood. The deviation of simulations from experiments is

trivial for both materials where the relative error for coal is 5.3% and for poplar wood is 2.9%. The

predicted bed heights for coal and poplar wood using different drag models are presented in Figure ‎4.3(a

and b), respectively, and are compared to the fluidization experiments. To compare the experimental and

simulated bed height, a dimensionless parameter is defined that is the ratio of expanded bed height to the

initial bed height. For coal, the predicted bed height using each drag model is closer to experiment at

lower velocities. However, differences in predicted values become more apparent as inlet gas velocity

increases. Based on the predictions for coal, drag models can be roughly divided into two categories: the

BVK, HYS, and HKL drag models (group 1) that over-predict bed height, whereas, the Gidaspow,

Gidaspow-Blend, Syamlal-O’Brien, and Wen-Yu drag models (group 2) that give good predictions,

which are in good agreement with experimental results. For poplar wood, the predicted bed height using

all drag models are very similar and agree well with the experiments. It is interesting to note that for

poplar wood (Figure ‎4.3(b)), category 1 predictions are closer to the experiments. A larger particle

diameter was used to represent poplar wood due to the experimental observation that these biomass

particles tend to agglomerate. The estimate of a larger particle diameter in the computational model may

account for the under-prediction of bed height. However, the trend still persists that category 1 model

predictions are still larger than category 2.

The time-averaged void fraction profiles between 5-30 s (Δt = 0.01 s) at h/h0 = 1 versus the reactor

diameter for coal and poplar wood at a gas velocity of Ug = 9.87 cm/s are presented in Figure ‎4.4(a and

b), respectively. The two categories of drag models are less distinguishable for poplar wood, however, for

coal, the two categories of drag models are more recognizable. The drag models of BVK, HYS, and

Hoch-Hill show similar behavior, whereby void fraction values are larger for this group of models. The

second group of drag models (Gidaspow, Gidaspow-Blend, Syamlal-O’Brien, and Wen-Yu) predict lower

void fractions compared to the first group for both coal and poplar wood.

29

(a)

(b)

Figure 4.3. Bed height of (a) coal and (b) poplar wood for different drag models compared to experiments

[90].

Ug

(c m /s )

h/h

0

0 2 4 6 8 1 0 1 20 .5

1

1 .5

2

2 .5

3

3 .5

4


B V K

H Y S

H K L

G id a s p o w

G id a s p o w -B le n d

S y a m la l-O 'B r ie n

W e n -Y u

Ug

(c m /s )

h/h

0

0 2 4 6 8 1 0 1 20 .5

1

1 .5

2

2 .5

3

3 .5

4


B V K

H Y S

H K L

G id a s p o w



W e n -Y u

30

(a)

(b)

Figure 4.4. Void fraction profiles for (a) coal and (b) poplar wood at h/h0 = 1 and Ug = 9.87 cm/s.

D (c m )

g

-2 -1 0 1 20 .7

0 .7 5

0 .8

0 .8 5

0 .9

0 .9 5

B V K

H Y S

H K L

G id a s p o w

G id a s p o w _ B le n d


W e n -Y u

D (c m )

g

-2 -1 0 1 2

0 .4 5

0 .5

0 .5 5

0 .6

0 .6 5

0 .7

B V K

H Y S

H K L

G id a s p o w



W e n -Y u

31

4.2.3.2. Binary Mixtures of Coal and Poplar Wood

Binary mixtures of coal and poplar wood are modeled for mass ratios of 90:10, 80:20, and 70:30, and

the properties are shown in Table ‎4.2. In fully fluidized region, the measured pressure drop from the

fluidization experiments is 87 Pa, 128 Pa, and 114 Pa for 90:10, 80:20, and 70:30 mass ratios,

respectively. The simulations predict Δp = 87 Pa, 128 Pa, and 117 Pa for three mass ratios, which are in

very good agreement with experiments. For both 90:10 and 80:20 mass ratios, the results are the same but

for the 70:30 mass ratio, the relative error of the predictions is less than 2.7%.

The bed height of coal-poplar wood mixtures using different drag models are compared with

experimental data, as presented in Figure ‎4.5(a, b and c) for 90:10, 80:20, and 70:30 mass ratios,

respectively. Similar to single solid phases, for the 90:10 mass ratio (Figure ‎4.5(a)), the two groups of

drag models defined earlier are still recognizable. Similarly, the first group of drag models consisting of

BVK, HYS, and HKL drag models over-predicts the bed height whereas the second group of drag models

(Gidaspow, Gidaspow-blend, Syamlal-O’Brien, and Wen-Yu) give better predictions that agree well with

experiments. Amongst the first group of drag models, the BVK and HKL drag models predict the lowest

and highest bed height of the three models, respectively. As poplar wood mass ratio increases, the bed

height predicted by the two groups of drag models become more similar, which is similar to the single

solids phase of poplar wood (Figure ‎4.3(b)). For mass ratios of 80:20 and 70:30, all drag models except

the HKL drag model give good predictions for bed height and agree well with experiments. It is important

to mention that for binary mixtures, the HKL drag model always over-predicts the bed height.

Figure ‎4.6(a, b and c) show the time-averaged void fraction profiles between 5-30 s at h/h0 = 1 at Ug =

9.87 cm/s for 90:10, 80:20, and 70:30 mass ratios, respectively. The bifurcation of the drag models into

two categories can be seen for the mass ratio of 90:10 in Figure ‎4.6(a). As poplar wood mass ratio

increases, the predictions of two groups of drag models, except for HKL drag model, become more

similar (Figure ‎4.6(b and c)).

The instantaneous void fraction contours for 70:30 mass ratio at Ug = 9.87 cm/s from 25 – 29 s (t =

1 s) using the Gidaspow-blend and HKL drag models are presented in Figure ‎4.7. Clearly there is a

difference between bed height predictions using each model. The Gidaspow-blend drag model predicts

lower bed height and smaller bubbles, whereas, the bubbles are larger using the HKL drag model, which

results in a higher bed expansion. To further investigate drag model behavior, the time-averaged velocity

vectors along with velocity magnitude contours for coal and poplar wood using HKL and Gidaspow-

blend drag models at Ug = 9.87 cm/s for the 70:30 mass ratio are presented in Figure ‎4.8. For both drag

models, the movement of particles exhibit similar rotational patterns along the side walls for both solids

phases. In the regions near the center of the bed and in the vicinity of the walls and the inlet, the velocity

32

magnitudes are higher using both drag models. The HKL drag model, however, predicts larger regions of

higher velocity magnitude and higher bed heights.

(a) (b)

(c)

Figure 4.5. Bed height of (a) 90:10, (b) 80:20, and (c) 70:30 coal-poplar wood mass ratio for different drag

models compared to experiments [90].

Ug

(c m /s )

h/h

0

0 2 4 6 8 1 0 1 20 .5

1

1 .5

2

2 .5

3

3 .5

4


B V K

H Y S

H K L

G id a s p o w



W e n -Y u

Ug

(c m /s )

h/h

0

0 2 4 6 8 1 0 1 20 .5

1

1 .5

2

2 .5

3

3 .5

4


B V K

H Y S

H K L

G id a s p o w



W e n -Y u

Ug

(c m /s )

h/h

0

0 2 4 6 8 1 0 1 20 .5

1

1 .5

2

2 .5

3

3 .5

4


B V K

H Y S

H K L

G id a s p o w



W e n -Y u

33

(a) (b)

(c)

Figure 4.6. Void fraction profiles for (a) 90:10, (b) 80:20 and (c) 70:30 coal-poplar wood mass ratio at h/h0 = 1

and Ug = 9.87 cm/s.

D (c m )

g

-2 -1 0 1 2

0 .7 5

0 .8

0 .8 5

0 .9

B V K

H Y S

H K L

G id a s p o w



W e n -Y u

D (c m )

g

-2 -1 0 1 2

0 .7

0 .7 5

0 .8

0 .8 5

B V K

H Y S

H K L

G id a s p o w



W e n -Y u

D (c m )

g

-2 -1 0 1 2

0 .6 5

0 .7

0 .7 5

0 .8

0 .8 5

B V K

H Y S

H K L

G id a s p o w



W e n -Y u

34

(a) HKL drag model

(b) Gidaspow-blend drag model

Figure 4.7. Instantaneous void fraction for 70:30 coal-poplar wood mass ratio from 25 – 29 s (t = 1 s) at Ug =

9.87 cm/s using (a) HKL and (b) Gidaspow-blend drag models

(a) HKL drag model (b) Gidaspow-blend drag model

Figure 4.8. Time-average coal and poplar wood velocity vectors along with velocity magnitude contours for Ug

= 9.87 cm/s for 70:30 mass ratio of coal-poplar wood using (a) HKL and (b) Gidaspow-blend drag models.

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )-2 .5 0 2 .5

1

2

3

4

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )-2 .5 0 2 .5

1

2

3

4

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )-2 .5 0 2 .5

1

2

3

4

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )-2 .5 0 2 .5

1

2

3

4

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )-2 .5 0 2 .5

1

2

3

4

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )-2 .5 0 2 .5

1

2

3

4

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )-2 .5 0 2 .5

1

2

3

4

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )-2 .5 0 2 .5

1

2

3

4

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )-2 .5 0 2 .5

1

2

3

4

P o p la r w o o dC o a l

D (c m )-2 .5 0 2 .5

1

2

3

4

D (c m )-2 .5 0 2 .5

1

2

3

4

4 0

3 5

3 0

2 6

2 1

1 6

1 1

7

2

|us|

(c m /s )

C o a l P o p la r w o o d

35

The instantaneous contours of coal and poplar wood vertical velocity for the 70:30 mass ratio at Ug =

9.87 cm/s using the HKL and Gidaspow-blend drag models are presented in Figure ‎4.9 and Figure ‎4.10,

respectively. For both solids, the HKL drag model predicts the movement of particles higher in the reactor

(see Figure ‎4.9(a) and Figure ‎4.10(a)). It can also be noted that there are more regions of higher coal

velocity using the HKL drag model, hence the deviation in predictions of bed height (Figure ‎4.5(c)).

Overall, the drag models in group 1 predict the fluidization behavior of fine particles of Geldart A

more imprecisely. The reason could be due to the numerical nature of the models in category 1. Lattice-

Boltzmann simulations were used to calculate the drag force on a system of fixed particles, either

arranged in arrays or randomly dispersed. Simulations were repeated for numerous cases to represent a

range of Reynolds numbers and solids volume fractions. The calculated drag forces from the Lattice-

Boltzmann simulations were then used to create the drag correlations of category 1 and may not be as

robust for use in continuum-based Eulerian simulations. In contrast, category 2 drag models were derived

from empirical relationships that use void fraction regimes to adjust the formulations, which worked

better with Geldart A particles. The HKL model is the only Lattice-Boltzmann model that specifies

different formulations based on regimes (using solids volume fraction), however, this model consistently

over-predicted bed expansion. Although all drag models of Group 2 are recommended for Geldart A

particles when unfluidized regions are not considered in the mass of the bed, for binary mixtures, the

Gidaspow-blend and Wen-Yu models performed better than the rest. For single solids phase of coal, for

which there were more differences in the predictions, the Gidaspow-blend drag model shows the best

performance.

36

‎‎‎

(a) HKL drag model


Figure 4.9. Instantaneous coal vertical velocity for 70:30 coal-poplar wood mass ratio from 25 – 29 s (t = 1 s)

at Ug = 9.87 cm/s using (a) HKL and (b) Gidaspow-blend drag models.

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

Vc o a l

(c m /s )

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

Vc o a l

(c m /s )

37

(a) HKL drag model


Figure 4.10. Instantaneous poplar wood vertical velocity for 70:30 coal-poplar wood mass ratio from 25 – 29 s

(t = 1 s) at Ug = 9.87 cm/s using (a) HKL and (b) Gidaspow-blend drag models.

4.2.4.Conclusions

Numerical simulations of fluidized beds for coal, poplar wood and their mixtures were evaluated

using seven drag models reported in the literature and results were compared to experiments. To simulate

and analyze gas-solids hydrodynamic behavior of the fluidized bed, an Eulerian-Eulerian multi-fluid

model was used. Results of the study revealed that if static regions of material are removed by adjusting

the mass in the fluidized bed, the commonly-used drag models for Geldart B particles work well with

Geldart A particles. Results manifested into two categories of drag models. The first group included the

BVK, HYS, HKL models and the second group included the Gidaspow, Gidaspow-Blend, Syamlal-

O’Brien and Wen-Yu drag models. It was found that the first category mostly over-predicted bed height,

however, BVK and HYS model predictions agree well with experiments for poplar wood and mixtures

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

Vp o p la r

(c m /s )

D (c m )

h/h

0

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

D (c m )

-2 .5 0 2 .5

1

2

3

4

5

4 0

3 2

2 4

1 6

8

0

-8

-1 6

-2 4

-3 2

-4 0

Vp o p la r

(c m /s )

38

with higher poplar wood mass ratios. The second group gives good predictions for both single solids

phases and their mixtures, however, the Gidaspow-blend model proved to be the more reliable drag model

for both single solid phases and binary mixtures. These conclusions were also substantiated by examining

void fraction profiles to demonstrate particle distribution. Studying the nature of the two groups of

models revealed that the first group used numerical simulations to derive drag force models, whereas the

second group models were derived from empirical correlations.

4.3. Fluidization behavior and mixing properties of coal-biomass mixture

4.3.1. Coal and poplar fluidization

In this section, the predicted fluidization trends for coal and poplar are compared with the

experiments to determine how the materials behave individually before the binary solids mixture study.

Figure ‎4.11 (a and b) shows the pressure drop across the bed for the experiments and simulations for coal

and poplar wood, respectively. The fluidization experiments for coal show an almost linearly increasing

trend for pressure drop until the onset of fluidization at Umf = 1.6 cm/s. Close to Umf, a significant

overshoot of pressure is observed for coal followed by a constant trend where p = 75 Pa, which begins at

Ug = 6.6 cm/s where the bed is fully fluidized. As shown in Section 4.1, if the entire bed of coal fluidizes,

p = 225 Pa. Clearly, the pressure drop required to minimally fluidize the bed was measured

experimentally to be 140 Pa, further substantiating that there are regions in the bed that do not fluidize.

Furthermore, the fully fluidized bed requires one-third of the maximum pressure drop if all the mass

fluidizes. The simulations for coal do not capture the spike in pressure drop but are similar to the

experiments after 6.6 cm/s, where the bed is fully fluidized, after which the predicted pressure drop

remains constant at 79 Pa. A plausible explanation as to why the simulations cannot predict the sharp rise

in pressure may be due to the continuum assumption for the solids phase. Individual particles cannot be

modeled due to the large number of particles in the bed. Thus, averaged quantities are predicted. In turn,

the significant particle-particle interaction may not be accurately represented in the unfluidized state.

The fluidization experiments for poplar wood similarly show an almost linearly increasing trend until

the commencement of fluidization at Umf = 3.2 cm/s. There is a corresponding, albeit slight, overshoot of

pressure near Umf followed by a constant value of p = 69 Pa, which begins at Umb = 8.2 cm/s. Compared

to coal particles, the density of poplar wood particles is lower that could result in slighter overshoot of

pressure for poplar wood. It was found that modeling the poplar wood was more challenging due to the

irregular particle shapes. The computational modeling can be adjusted by using sphericity to create a

smaller spherical particle. However, using the sphericity of 0.6 that was experimentally calculated for

39

poplar wood produced unphysical results. Based on the experimental observations of the poplar wood

fluidization, it was noted that the particles would aggregate (i.e., particle bridging) forming clusters and

gas flow channels. Due to the formation of particle clusters, the computational model was adjusted by

increasing the sphericity to model a larger particle diameter. Doing so produced very good results

consistent with the experiments. Similar to coal, the simulations do not predict the overshoot of pressure

drop near Umf, however, the pressure in the fully fluidized region is predicted accurately at constant

pressure of 71 Pa.

The dimensionless bed height of coal and poplar wood versus gas inlet velocity are presented in

Figure ‎4.12 (a and b) and compared to experiments, which are in a good agreement. As velocity increases,

the bed height of the particles expands, however, the bed height of coal increases more compared to

poplar wood. The reason can be connected to the value of the minimum fluidization velocity where

fluidization of coal occurs at lower velocities than poplar wood.

The instantaneous void fraction contours based on the simulations for two gas inlet velocities, Ug =

4.93 cm/s and Ug = 9.87 cm/s, are presented in Figure ‎4.13(a-b) and (d-e) for coal and poplar wood,

respectively. As the velocity increases, the size of the bubbles and the average bed height increases.

Although for coal (Figure ‎4.13(a) and (b)) the bed is fluidized and bubbles formed for both velocities,

simulations of poplar wood particles (Figure ‎4.13(d) and (e)) predict a weakly-fluidizing bed at the lower

velocity. As the velocity increases, poplar wood fluidizes more readily and bubbles form. The time-

averaged void fractions spatially averaged across the bed width versus the dimensionless height for the

same two velocities for coal and poplar wood is shown in Figure ‎4.13(c) and (f), respectively. As velocity

increases, the slope of the void fraction curve increases as more bubbles rise to the surface. These figures

help to elucidate the amount of gas present in the bed, where there is more gas towards the particle-

freeboard interface due to the presence of more bubbles. Figure ‎4.13(f) also reaffirms that the simulations

predict the onset of heterogeneous fluidization at higher velocities where Ug = 8 cm/s. At the lower

velocity, the void fraction is almost constant throughout most of the bed, however the bed height

increases to 0hh ≈ 1.2, after which there is a slight increase in void fraction.

40

(a) (b)

Figure 4.11. Pressure drop versus the gas inlet velocity comparing experiments [90] and simulations for (a)

coal, (b) poplar wood.

(a) (b)

Figure 4.12. The dimensionless bed height for different gas inlet velocities comparing experiments [90] and

simulations for (a) coal and (b) poplar wood.

Ug

(c m /s )

p

(Pa

)

0 2 4 6 8 1 0 1 20

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

E x p e r im e n t

S im u la t io n

Ug

(c m /s )

p

(Pa

)

0 2 4 6 8 1 0 1 20

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

E x p e r im e n t

S im u la t io n

Ug

(c m /s )

h/

h0

0 2 4 6 8 1 0 1 20

1

2

3E x p e r im e n t

S im u la t io n

Ug

(c m /s )

h/

h0

0 2 4 6 8 1 0 1 20

1

2

3E x p e r im e n t

S im u la t io n

41

(a) Coal, 4.93 cm/s (b) Coal, 9.87 cm/s (c)

(d) Poplar, 4.93 cm/s (e) Poplar, 9.87 cm/s (f)

Figure 4.13. Instantaneous void fraction contours at t = 30 s for coal at (a) Ug = 4.93 cm/s and (b) Ug = 9.87

cm/s; and (c) void fraction horizontally averaged across the reactor diameter versus axial direction; for

poplar at (d) Ug = 4.93 cm/s and (e) Ug = 9.87 cm/s; and (f) void fraction horizontally averaged across the

reactor diameter versus axial direction.

4.3.2. Binary mixtures of coal and poplar fluidization

Binary mixtures of coal-poplar are studied for mass ratios of 90:10, 80:20 and 70:30. All properties

for the simulations for each mass ratio are presented in‎Table ‎4.2. The pressure drop across the bed for the

simulations and experiments versus the gas inlet velocity for the three mass ratios are presented in

Figure ‎4.14. The experiments show that pressure drop increases until the commencement of fluidization at

Umf = 2.7 cm/s, 4.8 cm/s, and 4.5 cm/s for mass ratios of 90:10, 80:20 and 70:30, respectively. This region

is followed by a quite substantial pressure spike for the mass ratio of 90:10 and a slight overshoot of

pressure for the 80:20 mass ratio. In the fully developed region, a constant pressure drop of 87 Pa and 127

Pa is reported for 90:10 and 80:20 mass ratio, respectively. As the poplar wood mass ratio increases to

30%, no overshoot of pressure is reported and the increasing region is trailed by a constant pressure drop

D (c m )

Y/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

40 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

V o id fra c t io n

0 .8 0 .9 10

1

2

3

4U = 4 .9 3 c m /s

U = 9 .8 7 c m /s

D (c m )

Y/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

40 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

V o id fra c t io n

0 .5 0 .6 0 .7 0 .8 0 .9 10

1

2

3

4U = 4 .9 3 c m /s

U = 9 .8 7 c m /s

42

at 114 Pa. Similar to single solids phase predictions, simulations do not predict an overshoot of pressure

drop for 90:10 and 80:20 mass ratios. The predicted pressure drops in the fully fluidized regimes remain

constant at 87 Pa, 128 Pa, and 117 Pa for mass ratios of 90:10, 80:20 and 70:30, respectively, which are

very close to the experimental pressure drop. The pressure drop across the bed for all mixtures is higher

than the single solids phases of coal or poplar (Figure ‎4.11). However, the pressure for the mass ratio of

90:10 is close to the single solids phase pressure. It is interesting to note that the value of Umf increases

and then decreases with increasing mass ratio, however, Umb remains at approximately 8.2 cm/s for the

three cases. A plausible explanation for the changes in Umf is that the mixture ratio of 80:20 represents a

transitional flow as more poplar wood is added to coal, which is explained further in the following text.

The dimensionless bed height for the three mixtures are presented in Figure ‎4.15 (a, b, and c) and

show that as the velocity increases, the average bed height expands for each mass ratio. The experiments

indicate that as the poplar wood mass ratio increases, the bed height increases. The simulations are in

good agreement with the experiments, although a higher bed height is predicted for the 90:10 mass ratio.

It is important to note that the 90:10 mass ratio shows similar bed height trends as was seen in the single

solids phase for coal fluidization (Figure ‎4.12(a)) with a sudden increase of bed height at medium

velocities. As the poplar wood mass ratio increases to 30%, the bed height trends are more similar to the

single solids phase of poplar (Figure ‎4.12(b)), where bed height increases more uniformly. The bed height

of the 80:20 mass ratio may be interpreted as a transitional state from a bed with properties consistent

with coal to a bed with properties more similar to poplar. The sudden increase in bed height at the

minimum fluidization velocity is still obvious for the 80:20 mass ratio, although less dramatic than the

90:10 mass ratio. It is also interesting to note that the bed height of the mixtures is higher than the bed

height when poplar wood fluidized as a single solids phase but lower than the bed height when coal

fluidized as a single solids phase (comparing Figure ‎4.12 and Figure ‎4.15).

43

(a) 90:10 mass ratio of coal-poplar (b) 80:20 mass ratio of coal-poplar

(c) 70:30 mass ratio of coal-poplar

Figure 4.14. Pressure drop versus the gas inlet velocity comparing experiments [90] and simulations for mass

ratios for (a) 90:10, (b) 80:20 and (c) 70:30 mass ratio of coal-poplar.

Ug

(c m /s )

p

(Pa

)

0 2 4 6 8 1 0 1 20

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

E x p e r im e n t

S im u la t io n

Ug

(c m /s )

p

(Pa

)

0 2 4 6 8 1 0 1 20

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

E x p e r im e n t

S im u la t io n

Ug

(c m /s )

p

(Pa

)

0 2 4 6 8 1 0 1 20

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

E x p e r im e n t

S im u la t io n

44

(a) 90:10 mass ratio of coal-poplar (b) 80:20 mass ratio of coal-poplar

(c) 70:30 mass ratio of coal-poplar

Figure 4.15. The dimensionless bed height for different gas inlet velocities comparing experiments [90] and

simulations for mass ratios for (a) 90:10, (b) 80:20 and (c) 70:30 mass ratio of coal-poplar.

Figure ‎4.16 shows the time-averaged void fraction, and coal and poplar wood volume fractions

spatially averaged across the bed width versus the dimensionless height for Ug = 9.87 cm/s for all mixture

ratios. Similarly, as the poplar wood mass ratio increases from 10% to 30%, the slope of the void fraction

curve decreases and the bed height is lower for the 70:30 mass ratio. For the mass ratio of 80:20, the slope

of the void fraction curve is very close to the slope of 70:30 mass ratio, which is consistent with bed

Ug

(c m /s )

h/

h0

0 2 4 6 8 1 0 1 20

1

2

3E x p e r im e n t

S im u la t io n

Ug

(c m /s )

h/

h0

0 2 4 6 8 1 0 1 20

1

2

3E x p e r im e n t

S im u la t io n

Ug

(c m /s )

h/

h0

0 2 4 6 8 1 0 1 20

1

2

3E x p e r im e n t

S im u la t io n

45

height observations. Solid volume fractions show similar results. The 90:10 mass ratio has the highest

fractions of coal at all heights; however, the coal volume fractions for 80:20 and 70:30 mass ratios have

similar values in the vicinity of the freeboard near 0hY = 2.5. Poplar wood volume fraction increases

with increasing mass and is relatively constant throughout the bed, and can be connected to mixing

characteristics, discussed next.

(a)

(b) (c)

Figure 4.16. Void fraction horizontally averaged across the reactor diameter versus axial direction for Ug =

9.87 cm/s for 90:10, 80:20 and 70:30 mass ratio of coal-poplar.

V o id fra c t io n

Y/

h0

0 .7 5 0 .8 0 .8 5 0 .9 0 .9 5 10

0 .5

1

1 .5

2

2 .5

3

3 .5

49 0 :1 0 m a s s r a t io

8 0 :2 0 m a s s r a t io

7 0 :3 0 m a s s r a t io

V o lu m e fra c t io n

Y/

h0

0 0 .0 5 0 .1 0 .1 5 0 .20

0 .5

1

1 .5

2

2 .5

3

3 .5

49 0 :1 0 m a s s r a t io

8 0 :2 0 m a s s r a t io

7 0 :3 0 m a s s r a t io

V o lu m e fra c t io n

Y/

h0

0 0 .0 5 0 .1 0 .1 5 0 .20

0 .5

1

1 .5

2

2 .5

3

3 .5

49 0 : 1 0 m a s s r a t io

8 0 : 2 0 m a s s r a t io

7 0 : 3 0 m a s s r a t io

46

4.3.2.1. Mixing properties of coal-poplar wood mixtures

Mixing characteristics are an important feature that can induce better particle contact, which is

essential for increasing the efficiency of the process and providing a relatively uniform temperature

throughout the fluidized bed. The fundamental mechanism of solids mixing is due to bubble movement in

the bed during fluidization. In the close vicinity of bubbles that rise through the bed, solids are drawn into

the wake of bubbles where mixing occurs due to the generated movement. In addition, rising bubbles

carry a trail of particles upward and release them at the bed surface (freeboard) as they erupt. Particles

then move downward in the region surrounding the bubbles. An overall circulation of particles in the

axial direction of the bed ensues as the result of these phenomena. Meanwhile, in the lateral direction,

sidewise motion of bubbles due to coalescence and interaction of adjacent bubbles and lateral dispersion

of particles due to bubbles burst result in lateral solids mixing [104,105]. Solids mixing quality improves

as inlet gas velocity increases, which is due to intense internal circulation at higher gas velocities [106].

Solids mixing properties of a fluidized bed have been studied experimentally through tracing methods

[107,108,109], and numerically mostly through Lagrangian models [106,110,111]. Mixing properties and

quality of mixing are usually assessed through either establishing the exchange coefficient between the

bubble wake and the emulsion phase or using the degree of particles mixing in the bed. A qualitative

understanding of particle flow patterns can be acquired by studying snapshots of void fraction and solids

volume fraction contours procured from simulations. Using this method, bubbles formation, growth and

coalescence, solids circulations and concentrations can be clearly discerned [106].

To quantify the mixing of particles in a fluidized bed, the mixing index (MI) is one method that is

used [112]. The mixing index is defined as:

%100

T

U

X

XMI (53)

where XU is the mass fraction of jetsam particles in the upper region of the bed and XT is the total mass

fraction of the jetsam particles in the whole bed. Higher values of MI represent better mixing of particles,

whereby MI = 100% is a perfectly mixed bed and MI = 0 is a completely segregated bed. Considering the

upper region of the bed begins at 0hh = 1, the mixing index at Ug = 9.87 cm/s for 90:10, 80:20, and

70:30 mass ratios is 60.69%, 61.92%, and 64.12%, respectively.

The instantaneous gas and solids volume fraction contours for coal and poplar wood at t = 30 s for the

three mass ratio mixtures are presented in Figure ‎4.17 for Ug = 9.87 cm/s. As the poplar wood mass ratio

increases, void fraction contours (Figure ‎4.17(a)) show a more gas dilute bed, where there are regions of

lower void fraction (corresponding to 0.6 that are medium blue) for the 70:30 mass ratio. The size of the

bubbles (red regions in all figures) does not show any significant difference between each mass ratio.

47

Instantaneous solids volume fraction contours (Figure ‎4.17(b-c)) show that the regions of higher

concentration of coal and poplar wood are mostly coincident, which confirms suitable mixing for each

mass ratio.

90:10 mass ratio 80:20 mass ratio 70:30 mass ratio

(a) gas


(b) coal


(c) poplar

Figure 4.17. Instantaneous volume fraction contours for all mass ratio of coal-poplar using a binary mixture

model with Ug = 9.87 cm/s at t =30 s.

D (c m )

Y/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

40 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

D (c m )

Y/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

40 .2 4

0 .2 2

0 .2

0 .1 8

0 .1 6

0 .1 4

0 .1 2

0 .1

0 .0 8

0 .0 6

0 .0 4

0 .0 2

s

D (c m )

Y/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

40 .2 4

0 .2 2

0 .2

0 .1 8

0 .1 6

0 .1 4

0 .1 2

0 .1

0 .0 8

0 .0 6

0 .0 4

0 .0 2

s

48

To further investigate particle mixing behavior at each mass ratio, the time-averaged solids axial and

lateral mass flux and velocity vectors are presented. Axial mass flux of the solids spatially averaged

across the bed width versus the dimensionless height for Ug = 9.87 cm/s for all mass ratios are presented

in Figure ‎4.18. For the 90:10 mass ratio, mass flux is mostly positive and implies a dominant upward

movement of solids particles. The piecewise integration of axial mass flux shows that as the poplar wood

mass ratio increases, the magnitude of the axial mass flux increases, which leads to better axial mixing of

particles. The value of integral for 90:10, 80:20, and 70:30 mass ratios is 9.3, 16.7 and 22.4, respectively.

Figure ‎4.19 shows the lateral mass flux of the solids vertically averaged across the bed height versus the

diameter of the reactor. Similar to the axial mass flux, as the poplar wood mass ratio increases, the

magnitude of the lateral mass flux increases, which implies better lateral mixing of particles. The value of

integral was found to be 1.3, 34, and 81.3 for mass ratios of 90:10, 80:20, and 70:30, respectively.

Therefore increasing the poplar wood mass ratio improves the quality of particles mixing.

The time-averaged coal velocity vectors and volume fraction contours for Ug = 9.87 cm/s for all mass

ratios are presented in Figure ‎4.20 (because of the similar vector pattern of coal and poplar wood, only

coal velocity vectors are presented). For all mass ratios, the movement of particles creates regions of

higher solids concentration in the vicinity of the walls, distributor plate and near the center of the bed;

these regions are blue in the volume fraction contours. The movement of the particles for all mass ratios

exhibits two similar rotational patterns along the side walls near the distributor plate. For all mass ratios, a

portion of the particles create two separate circulations near the center of the bed in the vicinity of the

freeboard. As the poplar mass ratio increases, the solids velocity pattern changes and the central

circulatory structure does not form. For 80:20 and 70:30 mass ratios, the velocity patterns are very

similar. As poplar mass ratio increases, the particle velocities increase and facilitate better mixing

characteristics for higher poplar mass ratios.

49

Figure 4.18. Time-averaged axial solids mass flux spatially averaged across the bed width versus the

dimensionless height for Ug = 9.87 cm/s.

Figure 4.19. Time-averaged axial solids mass flux spatially averaged across the bed width versus the

dimensionless height for Ug = 9.87 cm/s.

m "y

(g /c m2.s )

Y/

h0

-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 60

1

2

3

4 9 0 : 1 0 m a s s ra t io

8 0 : 2 0 m a s s ra t io

7 0 : 3 0 m a s s ra t io

D (c m )

m"

x(g

/cm

2.s

)

-2 -1 0 1 2

-3

-2

-1

0

1

2

3

9 0 : 1 0 m a s s ra t io

8 0 : 2 0 m a s s ra t io

7 0 : 3 0 m a s s ra t io

50

(a) 90:10 mass ratio (b) 80:20 mass ratio (c) 70:30 mass ratio

Figure 4.20. Time-average coal velocity vectors for Ug = 9.87 cm/s for (a) 90:10, (b) 80:20, (c) 70:30 mass ratio

of coal-poplar.

4.3.3. Conclusions

Numerical simulations and experiments of fluidized beds were used to better understand the

fluidization and particle mixing behavior of coal-biomass mixtures. Sand that is classified as Geldart B

particles was used to validate the fluidized bed modeling and correspondence to the experiments. Coal

and poplar wood that both fall into the Geldart A classification and three binary mixtures of coal and

poplar wood for mass ratios of 90:10, 80:20 and 70:30 were studied and compared to the experiments. To

simulate and analyze gas-solids hydrodynamic behavior of the fluidized bed, an Eulerian-Eulerian model

was used. The predictions for pressure drop across the bed and bed height were validated with the

experiments and found to be in a good agreement. To investigate the mixing behavior of mixtures and

complete the investigation, instantaneous void fraction contours, time-averaged mass flux, velocity

vectors and time-averaged void fractions were examined.

Pressure drop for the coal-poplar mixtures were higher than single solids phases for each material,

however, the pressure of the 90:10 mass ratio was close to the single solids phase pressure drop. It was

found that the 90:10 mass ratio exhibited bed height trends similar with the single solids phase of coal,

D (c m )

-2 .5 0 2 .5

1

2

3

4

D (c m )

Y/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

4

D (c m )

Y/h

0

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

4

D (c m )

-2 .5 0 2 .5

1

2

3

4

0 .2 4

0 .2 1 8

0 .1 9 6

0 .1 7 4

0 .1 5 2

0 .1 3

0 .1 0 8

0 .0 8 6

0 .0 6 4

0 .0 4 2

0 .0 2

s

51

and as the poplar wood mass ratio increased to 30%, the bed height trend was similar to the single solids

phase of poplar wood. Although instantaneous solid volume fraction contours indicated good mixing

characteristics for all mass ratio mixtures, solids mass flux and velocity vectors showed a tendency of

better mixing for mixtures with higher poplar mass ratio. The mixing of coal and poplar wood was further

examined to determine the effect on fluidization. A quantitative analysis of the mixing index confirmed

that as the poplar wood mass ratio increased, the quality of mixing improved, with an average mixing

index of 62%. Therefore, reasonable amounts of biomass can be added with coal without adverse effects

of segregation or elutriation, while reducing the use of a fossil fuel.

4.4.New method to model fluidization of Geldart A particles

4.4.1. Introduction

In order to predict the fluidization behavior of Geldart A particles using a standard two-fluid model, a

novel and optimal method based on the bed structure differences in homogeneous and heterogeneous

regions is proposed in this section. As mentioned in Section 2.1, standard two-fluid models show some

deficiencies in predicting the special behavior of Geldart A particles in the homogeneous fluidization

region, where the bed is stable and no bubbles form, but bed height increases. The structure of fluidized

beds of Geldart A particles in the homogeneous regime has been the subject of various studies

[113,114,115]. These studies indicate that inter-particle attractive forces are responsible for the stable

nature of the bed in this region. The attractive forces create a solid-like structure, whereby the particles do

not move within the bed. In a way, inter-particle attractive forces act as an elastic module that sustains the

bed against small disturbances [113,116]. No particle fluctuations or mobility are reported in this

fluidization region, which confirms the larger pressure drop through the bed [117]. Therefore, the bed

expansion in this region is the result of not only the balance between particle weight and drag force but is

partially the result of inter-particle contacts [113,116]. As gas velocity increases, the distance between

particles grows, and consequently inter-particle forces decrease until reaching the point where these

forces diminish. This point is the yield strength of the solid-like granular bed, where non-bubbling

expansion is not observed anymore. After this point, heterogeneous fluidization begins and bubbles form

[17,19,118,119,120]. The behavior of particles in the heterogeneous fluidization region is a well-

documented phenomenon in literature that is observed in all particle fluidization irrespective of their

Geldart classification.

Regardless of the phenomena underlying homogeneous fluidization, the differences between bed

structures in the two fluidization regions suggest a different modeling technique and/or correlation.

52

Limitations of two-fluid models in accurately predicting the fluidization behavior of Geldart A particles

may be a result of neglecting the transition between the homogeneous and heterogeneous regimes. Two-

fluid models treat both solids and fluid as interpenetrating continua and use closure laws to consider fluid-

solids and solid-solid interactions to numerically solve the flow fields. Drag force between two phases

that are caused by velocity difference [76] is an important closure law. Compared to fluid-solids drag

forces, solids-solids drag forces are not well characterized. Some equations have been proposed [76] to

model these interactions that are mostly based on solids-solids interactions in the heterogeneous

fluidization region, e.g., based on kinetic theory of granular flow [79]. Because of the unique bed

structure of Geldart A particles, additional closure laws or correlations must be considered to successfully

predict fluidization behavior in the homogeneous fluidization region.

Kanholy et al. [24] proposed to use the fluidization velocity as the main characteristic velocity in Eq.

19. For fine Geldart A particles, however, two critical velocities play roles in fluidization behavior: Umf

and Umb. The sole use of any of these two velocities cannot correctly predict the behavior of the bed in

both homogenous and heterogeneous regions. Figure ‎4.21 demonstrates this assertion for a bed of poplar

wood particles. Figure ‎4.21 shows that despite the good performance in predicting the behavior of bed in

the heterogeneous region employing either Umf or Umb, neither of these velocities as the critical velocity is

capable of capturing the peak in pressure drop. Using Umf, the predicted critical fluidization velocity is

consistent with the experiments, but overpredicts the pressure. Using Umb, the critical fluidization velocity

is not captured and the predictions indicate bubbling at the experimental value of 8.2 cm/s.

Figure 4.21. Pressure drop versus the gas inlet velocity comparing experiments [90] and adjusted simulations

based on Umf and Umb for poplar.

Ug

(c m /s )

p

(Pa

)

0 2 4 6 8 1 0 1 20

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0E x p e r im e n t

S im u la t io n b a s e d o n Um f

S im u la t io n b a s e d o n Um b

53

To improve the correlation for Geldart A particles to consider dead zone regions, it is proposed herein

to use both the experimental Umf and Umb as the critical velocities to adjust the data in the homogeneous

and heterogeneous fluidization regions, respectively. As mentioned before, a portion of solids in the bed

does not contribute in fluidization. This portion of particles settles in volumes between the distributor

plates holes. In order to consider this portion in simulations, Kanholy et al. [24] proposed a correction that

reduces the initial mass of simulation to only the fluidizing material. To adopt this procedure for Geldart

A particles, the initial data for initial mass and bed height in the heterogeneous region were found using

Umb. Keeping the mass in the bed constant, then Umb was used to find the void fraction of the bed and bed

height in homogeneous region. Employing this method, coal, poplar wood, and switchgrass that are

categorized as Geldart A particles are studied and the efficacy of the modeling approach is ascertained by

validating with experimental pressure drop and bed height. For the work herein, the corrected data for all

materials for both fluidization regions are presented in Table ‎4.3.

Table 4.3. Properties for coal, poplar wood and switch grass

Properties

Experiment [90] Simulation

coal Poplar

wood

Switch

grass coal Poplar wood Switch grass

homog heterog homog heterog homog heterog

�̅�𝑝(µm) 62 151 145 63 63 151 151 145 145

𝜌𝑠 (g/cm3) 1.38 0.4 0.32 1.38 1.38 0.4 0.4 0.32 0.32

𝜌𝑏 (g/cm3) 0.57 0.25 0.19 0.85 0.57 0.28 0.25 0.09 0.19

𝜀𝑚𝑓 0.59 0.38 0.83 0.39 0.59 0.31 0.38 0.72 0.83

To detect the bed height from the simulations and reduce the uncertainty of the readings, solid

pressure curves were employed instead of the traditional method of finding bed height from void fraction

curves. A common approach to find the averaged bed height is by averaging void fraction horizontally

across the reactor diameter and observing the trend versus axial direction; however there is no consensus

on how to read bed height from these graphs. The averaged bed height is then estimated using the void

fraction in the vicinity of the freeboard where the void fraction is almost 1. Such a method can introduce

discrepancy and uncertainty. To find the expanded bed height, this work proposes to observe the trends of

solids pressure averaged across the reactor diameter versus axial direction. Due to the absence of solid

particles in the freeboard, the value of this parameter is zero in this region, which can be used to find the

bed height. Figure ‎4.22 illustrates the method for poplar wood at Ug = 9.87 cm/s. The average value of the

almost horizontal portion of the void fraction curve in Figure ‎4.22(a) shows the dimensional bed height is

about 5.3 cm, which is consistent with experiments; however, this approach is not as straightforward due

to the uncertainty of averaging the slope between 0.9 < g < 1.0. The same value with no hassle can be

54

read from solid pressure curve in Figure ‎4.22(b). Herein, all bed height values presented are found using

this method. It is important to mention that for higher velocities due to the presence of more bubbles at

the bed-freeboard interface, values near zero solid pressure might form a very small round edge. In such

cases, the intersection of the extended pressure line before the zero-pressure location must be considered

as the bed height, as shown with a red line in Figure ‎4.22(b).

4.4.2.Results and discussions

For coal, time-averaged pressure drop across the bed and time-averaged dimensionless bed height

versus the inlet gas velocity for simulations and experiments are presented in Figure ‎4.23 (a) and (b). The

fluidization experiments for coal show an almost linearly increasing trend for pressure drop until the onset

of fluidization at Umf = 1.6 cm/s. Close to Umf, a significant overshoot of pressure is observed followed by

a constant trend where p = 75 Pa. In the region beginning at Umb = 6.6 cm/s, the bed is fully fluidized

and bubbles form. The fluidization experiments present an overall increasing trend of bed height as inlet

gas velocity increases. The increasing trend, however, can be roughly separated into three sections, which

implies different fluidization regions that are shown with red lines in Figure ‎4.23(b). In the first region

where Ug < 1.6 cm/s that includes Umf and the last region (Ug > 6.6 cm/s) that includes Umb, the bed height

increases with a shallower slope compared to the middle region where the homogeneous bed expansion

occurs. The slopes demark the sudden bed expansion during the homogeneous fluidization where the

pressure peaks. Simulations predict an overshoot in pressure drop, however, the predicted value is less

than the experimental value. Simulation predictions for bed height agree well with the experiments.

(a) (b)

Figure 4.22. A sample of the new approach to find accurate bed height (poplar wood, Ug = 9.87 cm/s); (a)

averaged void fraction, and (b) averaged solid pressure, across the reactor diameter versus axial direction.

g

Y(c

m)

0 .9 10

1

2

3

4

5

6

7

8

ps

(P a )

Y(c

m)

2 4 6 8 1 0 1 20

1

2

3

4

5

6

7

8

55

(a) (b)

Figure 4.23. (a) Pressure drop and (b) dimensionless bed height versus the gas inlet velocity comparing

experiments [90] and simulations for coal

Pressure drop across the bed for the experiments and simulations for poplar wood is presented in

Figure ‎4.24(a). The fluidization experiments show the commencement of homogeneous fluidization at Umf

= 3.2 cm/s. There is a corresponding, albeit slight, overshoot of pressure between Umf and Umb, at 8.2

cm/s, followed by a constant value of p = 69 Pa in the heterogeneous fluidization region. Simulations for

poplar wood particles predict a similar trend that is in a good agreement with the experiments. The

dimensionless bed height of poplar wood versus gas inlet velocity is presented in Figure ‎4.24(b) and

compared to experiments, which are in a good agreement. As velocity increases, the bed height of the

particles expands; however, three regions detected in Figure ‎4.23(b) for coal particles are not as obvious

for poplar wood particles. Bed height expansion is smoother and has an almost constant trend throughout

all inlet gas velocities.

Pressure drop across the bed for the experiments and simulations and the dimensionless bed height

versus gas inlet velocity for switchgrass are presented in Figure ‎4.25(a) and (b), respectively. The

experiments for switchgrass show that pressure increases until the onset of homogeneous fluidization at

Umf = 2.8 cm/s, trailed by a constant pressure drop at 102 Pa with one exception at Ug = 6.58 cm/s where

pressure decreases to 98 Pa. This point may be an erroneous reading during the experiments. No

experimental data for higher inlet gas velocities were provided. Similar to the experimental bed height

expansion for coal, experiments for switchgrass show three regions as the bed expands, where the most

significant expansion happens in the heterogeneous fluidization region. For switchgrass, bed height

increases more gradually and mildly than coal in the two other regions as inlet gas velocity increases.

Ug

(c m /s )

p

(Pa

)

0 2 4 6 8 1 0 1 20

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

E x p e r im e n t

S im u la t io n

Ug

(c m /s )

h/

h0

0 2 4 6 8 1 0 1 20

1

2

3

E x p e r im e n t

S im u la t io n

56

(a) (b)


experiments [90] and simulations for poplar wood

(a) (b)


experiments [90] and simulations for switchgrass

Ug

(c m /s )

p

(Pa

)

0 2 4 6 8 1 0 1 20

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

E x p e r im e n t

S im u la t io n

Ug

(c m /s )

h/

h0

0 2 4 6 8 1 0 1 20

1

2

3

E x p e r im e n t

S im u la t io n

Ug

(c m /s )

p

(Pa

)

0 2 4 6 8 1 0 1 20

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

E x p e r im e n t

S im u la t io n

Ug

(c m /s )

h/

h0

0 2 4 6 8 1 0 1 20

1

2

3

E x p e r im e n t

S im u la t io n

57

4.4.3. Conclusions

A novel procedure for correctly predicting the fluidization behavior of fine particle of Geldart A

proposed. The procedure takes into account the difference of the bed structure in the two fluidization

regions of Geldart A particles, where inter-particle forces that act strongly in homogeneous fluidization

region weaken in heterogeneous fluidization regime. Numerical simulations were compared and validated

with experiments for beds of coal, poplar wood, and switchgrass particles using pressure drop and bed

height expansion. The procedure shows promising results for predicting fluidization behavior of fine

particles.

58

Chapter 5. Gasification

5.1. Introduction

Gasification is a thermo-chemical process that includes a set of chemical reactions in which carbon-

containing feedstock is converted into a gaseous mixture called synthetic gas or syngas. The process

occurs in the presence of a controlled amount of oxygen or air at high temperatures (>700°C). Syngas that

mainly contains hydrogen and carbon monoxide can be burned directly to produce electricity or used to

manufacture fertilizers, chemicals, liquid fuels, or pure hydrogen.

Gasification offers numerous advantages compared to competing technologies such as combustion.

Environmental benefits of gasification are one major attraction of the process. Due to higher temperatures

and pressures of the produced syngas compared to exhaust gases formed in combustion, gasification

offers a simpler emission control than combustion. At higher temperatures and pressures, removal of

carbon dioxide, sulfur and nitrogen oxides, and heavy materials such as mercury, cadmium, selenium,

arsenic, etc. are easier. Lower amounts of sulfur and nitrogen oxides result in extremely low SOx and

NOx emissions. Efficient and easy carbon dioxide and sulfur removal adds value to the gasification

process because the captured carbon dioxide can be sold to other industries for different purposes, and

removed sulfur can be converted into marketable sulfuric acid or elemental sulfur. Another environmental

advantage of gasification, especially coal gasification, is lower water requirement versus other coal-based

technologies, which is crucially important for regions and countries that have already reached the critical

levels of water consumption [121,122,123,124,125].

Beside environmental benefits, gasification efficiency is higher than combustion when the power has

more than one use. In a coal gasification power plant, syngas usually performs dual duties in what is

called a combined cycle. In this kind of plant, syngas is burned in a gas turbine to generate electricity, and

then the hot gases produced are employed to generate steam, which is mostly used in a steam turbine

generator. This dual use of power can potentially improve the fuel efficiency by 50 percent or more,

which results in less fuel consumption to generate power and consequently, produces less greenhouse gas

emissions such as carbon dioxide [125].

In spite of gasification advantages, using coal as the gasification fuel has some disadvantages. The

main drawbacks of using coal are concerns about the eventual depletion of usable coal resources. In

addition, carbon dioxide emission from coal gasification is still an issue. Although coal gasification plants

are certainly cleaner and an improvement over other coal-based technologies, which makes CO2 capture

reasonably easy, there is still more CO2 produced than needed. As a remedy to sequester CO2, coal

gasification plants mostly pump and store it underground, although the effectiveness of this process is not

59

certain yet [126]. To address these issues while ensuring cost effectiveness and National security, biomass

as a sustainable source of energy has attracted much attention.

Biomass feedstocks include agricultural waste and crop residues, byproducts of other biological

industries, and organic waste, and are defined as renewable and biological materials that have the

capability of being used directly or indirectly as a fuel [127,128]. Biomass feedstocks have high volatile

content, whereas, quantities of nitrogen, sulfur and mercury are low. Furthermore, their reaction

temperature is lower than traditional fossil fuels such as coal [6]. Hence, biomass theoretically has a

potential to be an efficient source of energy that reduces SOx, NOx and heavy metal emissions. However,

the direct combustion of biomass fuel is problematic and rarely applied due to various issues such as low

heating value, low ash melting points that can cause fouling and agglomeration, and flame instability due

their high reactivity and volatile matter context [129].

In order to benefit from the positive aspects of biomass feedstocks, and as a remedy to public

concern on coal pollution and climate change, co-gasification of coal and biomass is a profitable option.

Coal-biomass fuel mixtures have great potential to provide cleaner, low-cost, domestic energy, which can

improve the H2/CO ratio in produced gas [130]. The carbon neutrality of biomass, which is the equal

amount of produced carbon dioxide from biomass during the combustion and the amount of absorbed

carbon dioxide by energy crops during photosynthesis [5,6,7], potentially reduces the effective carbon

emission in co-gasification processes [130]. The reduction was found to be linearly proportional to the

biomass quantity in the coal mixture [131].

5.1.1. Coal and biomass composition

To truly understand the nature of co-gasification chemistry, gaining enough knowledge of coal and

biomass chemical properties, compositions and their separate gasification chemistry is essential. Coal is a

converted form of buried dead vegetation over a very long period of time under high temperature and

pressure. Coal is a heterogeneous mixture of minerals and plant materials that does not have a single

chemical formula, but a formula that varies depending on the mixture sources. Due to various natural

conversion processes including time, temperature and pressure, coal types are notably different in their

compositions and consequently, in their emitted syngas from the gasification process. The percentage of

fixed carbon, for instance, can vary from 20% to 98% from low to high rank coal and as a result, the

calorific value of coal varies significantly in the range of 15,000 to 6,300 BTU/lb. The price of coal is

directly affected by rank of coal. The best performance of gasification process from an environmental and

operational point of view is with high rank coal that has low ash content. Although different types of coal

contain different percentages of each constituent, all types of coal are mainly made of carbon, but also

60

consist of other elements such as hydrogen, oxygen, sulfur, and nitrogen. Low-rank coals commonly have

higher oxygen and hydrogen contents.

Biomass composed of mostly plants or plant-based materials has broad varieties with different

chemical compositions that are substantially different from coal [132]. Biomass as a lignocellulosic

material is a complicated mixture of mainly cellulose, hemicelluloses, and lignin, and secondarily

extractives and minerals. Cellulose is a long linear chain of linked D-glucose units and is the main

component of green plant cell walls. Hemicellulose is present in almost all plant cell walls and consists of

short chains of sugar units. Lignin is a complex random-structured polymer of aromatic alcohols that is

the strengthening component of plants cell walls [133]. In comparison with coal, biomass feedstocks

contain higher volatile matter and moisture contents, and lower hydrogen, oxygen, nitrogen and sulfur

context. Also considerable amounts of alkali compounds exist in raw biomass feedstock that can

potentially affect the kinetics of gasification. Biomass particles also have much larger aspect ratio and

smaller material density compared to coal particles.

5.1.2. Gasification reactions

Despite the different physical and chemical properties of coal, biomass or their mixtures, the overall

gasification reactions for all fuels are similar. The chemical reactions of gasification can be divided into

four groups: initial stage, gasification, combustion, and water shift reactions. Each group of reactions is

characterized as follows:

Initial stage reactions:

Moisture H2O

Volatile matter Tar + Gases

Tar Fixed Carbon + Gases

Gasification reactions:

C + H2O CO + H2

C + CO2 2CO

C + 2H2 CH4

Combustion reactions

2C + O2 2CO

2CO + O2 2CO2

CH4 + 2O2 CO2 + 2H2O

2H2 + O2 2H2O

Water shift reaction

CO + H2O CO2 + H2

61

Heterogeneous reactions between char and gaseous species are the most important group of reactions

to explain gasification. These reactions are very slow compared to initial stage reactions and

homogeneous (gaseous) reactions, hence, gasification processes are controlled by this group of reactions

[134,135].

Despite the similarity between gasification reactions of different solid fuels, the structure and

chemical properties of the char products vary and the resulting char reactivity is different. During

gasification, char reactivity is defined as the rate of char consumption per unit mass of remaining char.

Woody biomass char has a more amorphous structure and lower density compared to fuels like coal

[136], and contain metal elements [137], which lead to more reactivity of biomass char than coal char

[138]. Therefore during co-gasification, the presence of more reactive biomass char can potentially

increase the char reactivity of the mixture [139,140].

In the remaining study, initial stage reactions are assumed to take place outside of the reactor and hot

gaseous products enter the reactor through a jet. The gas analysis of the entering gas stream is found using

Carbonaceous Chemistry for Continuum Modeling (C3M) software (developed by NETL) for the specific

proximate and ultimate analysis of coal used. In the following subsections, the primary reactions are

explained in more detail. For Pittsburgh No. 8 and Illinois No. 6 coals that will be used in the following

simulations k and E values for all reactions [141] are presented in Table ‎5.1 .

5.1.2.1.Steam gasification: C + H2O CO + H2

Steam gasification is a heterogeneous reaction that takes place throughout the volume of char

particles. This reaction is basically carbon oxidation by steam, which can be supplied as the gasification

agent or can be the steam that is generated during initial stage reactions. Compared to other gasification

reactions, steam gasification is a slow reaction. This reaction controls the maximum temperature of the

reactor; hence, its kinetics is extremely important. The reaction rate for temperatures below 1200 ºC is

defined as:

g

OHOHFCs

TR

Epp

XkR

1*

11 exp12 22

(54)

where

)1632629.17(exp

2

2

*

g

COH

OHT

ppp

(55)

and X is the mass fraction of a solids phase, p is partial pressure, R is the universal gas constant, and

gT is gas temperature. The subscript FC represents fixed carbon. The parameters k and E are the pre-

62

exponential coefficient and activation energy, respectively, which values depend on the type of coal. The

heat of this reaction in assigned to the solid phase [141,142].

5.1.2.2. Carbon dioxide gasification: C + CO2 2CO

The CO2 gasification is somewhat similar to steam gasification. Similarly, CO2 gasification is a

heterogeneous reaction that takes place throughout the volume of char particles. The reaction rate follows

the absorption relation and is defined as:

g

COCOFCs

TR

Epp

XkR

2*

22 exp12 22

(56)

where

)2028292.20(exp

2

2

*

g

CO

COT

pp

(57)

The heat of reaction is also assigned to solids phase [141,142].

5.1.2.3. Methanation: C + 2H2 CH4

Similar to the last two reactions, the methanation reaction takes place throughout the volume of char

particles. It is a slow heterogeneous reaction; however, at high hydrogen partial pressures and at

temperatures above 700 ºC, the rate of reaction is considerably faster. Methanation reaction rate is

defined as:

gHHFCs

TppX

R 8078087.7exp12

*

3 22

(58)

where

)1099943.13(exp

4

2

*

g

CH

HT

pp

(59)

Similarly, the heat of reaction is assigned to the solids phase [141,142].

5.1.2.4. Carbon combustion: 2C + O2 2CO and 2CO + O2 2CO2

Carbon combustion is a very fast reaction that takes place in two steps for heat transfer purposes. In order

to determining the rate of the reaction, a shrinking core model is employed as:

ashfilm

O

kk

PfR

11

21

4

(60)

where

63

61

10

FC

FC

X

Xf (61)

gp

film

Td

Dk

22

)1(292.0 (62)

75.1)1800(26.4 gTD (63)

c

c

Afilmashd

dkk

1

5.2 (64)

31

0

0

)1(

FCFC

AFCc

XX

XXd (65)

The heat of reaction for the first step is assigned to the solids phase and the heat reaction of the second

step is assigned to the gas phase [141,142].

5.1.2.5. Methane oxidation: CH4 + 2O2 CO2 + 2H2O

Methane oxidation is a surface reaction that mostly takes place in the presence of a catalyst.

Therefore, it is typically omitted from gasification reactions.

5.1.2.6. Hydrogen-Oxygen reaction: 2H2 + O2 2H2O

Hydrogen-Oxygen reaction is an extremely fast reaction, in the way that it is assumed no hydrogen

exists in the presence of oxygen. This reaction is usually neglected in gasification models. The rate of

reaction is defined as:

g

gHgOTR

TPTPR16000

exp.06.82.06.821035.03.010

6 22 [141,142]. (66)

5.1.2.7. Water-gas shift reaction: CO + H2O CO2 + H2

The water-gas shift is a quick surface catalyzed reaction. In the reaction rate, a factor representing the

reactivity of ash ( w ) is considered, and is a different value depending on the coal. The reaction rate is

defined as:

KyyyyTR

PfwR HCOOHCO

g

P/

27760exp10877.2

222

)2505.0(5

7

(67)

where

gsA TXf 555391.8exp)1( 00 (68)

gTK 723463061.3exp (69)

64

Table 5.1. Kinetic constants for two types of coal

Type of coal Pittsburgh No. 8 [141] Illinois No. 6 [141]

k1 (1/atms) 930 2250

E1 ( cal/mol) 45000 42000

k2 (1/atms) 930 2250

E2 ( cal/mol) 45000 42000

w 0.0068 0.0155

In Equation 67, y is mass fraction of gas species and P is pressure. In Equation 68, subscript zero and A

represents the initial value ash, respectively. The heat of reaction is assigned to the solids phase

[141,142].

5.2. Coal gasification

5.2.1. Effects of fluidization velocity on gasification process

As the first step of the gasification analysis, coal gasification in a fluidized bed is modeled. (The

gasification model was validated with literature [143] and the results can be found in Appendix A.) The

geometry of the reactor that is presented in Figure ‎5.1 is similar to the geometry used in fluidization

section (Chapter 4), which is shown in Figure ‎3.1, but a jet for solid fuel feeding is implemented at the

bottom of the reactor. The reactor is modeled in a cylindrical coordinate system that represents one-half

of the centerplane using an axisymmetric boundary condition at r = 0. The geometry is discretized into 10

and 60 cells in x and y directions, respectively. In order to simplify the gasification process, it was

assumed that initial stage reactions occur outside of the reactor and hot product gases along with solids

enter the reactor through the central jet with a 0.32 cm diameter and 2 cm wall height. Therefore, the total

number of reactions to be modeled is reduced from seven to five:

Carbon combustion: 2C + O2 2CO

CO2 gasification: C + CO2 2CO

Steam gasification: C + H2O CO + H2

Methanation: C + 2H2 CH4

Water gas-shift reaction: CO + H2O CO2 + H2

65

To account for temperature changes of solids entering the reactor, a pseudo-reaction is considered to

model cold char converting to hot char. A cut-off temperature for char conversion is assumed to be 800

ºC, when cold char entering the reactor is converted to hot char (hot char is previously present in the

reactor).

The gasification agent is air that enters the reactor through the annulus around the central jet. Air

enters the reactor at 27 ºC where its temperature increases in the reactor due to the reactions. The initial

temperature of the reactor is 1000 ºC and the initial bed height is 8 cm that is mostly ash (90% ash and

10% fixed carbon). The initial void fraction of the bed is 0.5 and the diameter and density of both solid

phases (hot char and cold char) are 0.1 cm and 1 g/cm3. Note that the particle characteristic falls in the

Geldart B classification. The composition of both chars is assumed to be 60% fixed carbon and 40% ash.

A summary of all initial conditions is presented in Table ‎5.2. To study the effect of inlet velocity on the

hydrodynamics of the fluidized bed and gasification, five sets of inlet gas and solids velocities are studied

that are presented in Table ‎5.3.

Figure 5.1. The 2D plane representing a center plane of the cylindrical reactor.

66

Table 5.2. Initial conditions

Properties Pittsburgh No. 8 Coal

�̅�𝑝(μm) 1000

𝜌𝑠 (g/cm3) 1

ℎ0 (cm) 8

𝜀𝑚𝑓 0.5

Initial temperature (ºC) 1000

Air inlet temperature (ºC) 27

Table 5.3. Annulus and jet velocity for air and solids.

Properties Annulus (cm/s) Jet (cm/s)

Gas velocity

Case 1

Case 2

Case 3

Case 4

Case 5

4.76

8.22

13

17.75

13

16

25

25

25

39

Solids velocity

Case 1

Case 2

Case 3

Case 4

Case 5

0

0

0

0

0

6

12

12

12

24

For all cases, the mass fractions of CH4, H2, and CO versus time are presented in Figure ‎5.2. All mass

fractions are averaged spatially over the entire reactor height and width. For all cases, mass fraction of

gases increases rapidly with time to reach to a certain value, after which the mass fraction remains almost

constant. Comparing case 1 with the rest of the cases shows almost no fluctuation in gas mass fractions

for case 1, which is due to the lower velocities of this case. For cases 2-4 that have the same solid and gas

jet velocities but different annulus gas inlet velocity, mass fraction of all three gases decreases as gas

velocity increases. Higher annulus gas velocity results in shorter residence times and the gasification

agent has less time to react with bed materials, thus, product mass fraction decreases. Comparing case 3

and 5 that have the same annulus gas velocity but different jet velocities shows increasing mass fraction

67

of all three gases as jet velocities increase. The reason is due to the higher concentrations of solids in the

system at higher jet velocities that increase the reaction rate in the reactor.

The instantaneous mass fraction for CH4 and H2 for case 5 is higher than the rest of the cases. For the

mass fraction of CO, however, case 2 produced the highest amount of CO mass fraction. The reason for

this difference is that CO production is mostly from carbon combustion that depends on the air entering

the system and the residence time to react, which is higher for case 2. CH4 and H2 are the products of

reactions between solid and initial stage gaseous products entering the system through jet, which must be

higher for case 5 that has higher jet velocities.

(a) (b)

(c)

Figure 5.2. Mass fraction of (a) CH4, (b) H2, and (c) CO versus time for Geldart B particles.

T im e (s )

Ma

ss

fra

cti

on

of

CH

4

0 2 4 6 8 1 00

0 .0 0 0 5

0 .0 0 1

0 .0 0 1 5

0 .0 0 2

0 .0 0 2 5

0 .0 0 3

0 .0 0 3 5

0 .0 0 4

0 .0 0 4 5

C a s e 1

C a s e 2

C a s e 3

C a s e 4

C a s e 5

T im e (s )

Ma

ss

fra

cti

on

of

H2

0 2 4 6 8 1 00

5 E -0 5

0 .0 0 0 1

0 .0 0 0 1 5

0 .0 0 0 2

0 .0 0 0 2 5

0 .0 0 0 3

C a s e 1

C a s e 2

C a s e 3

C a s e 4

C a s e 5

T im e (s )

Ma

ss

fra

cti

on

of

CO

0 2 4 6 8 1 00

0 .0 5

0 .1

0 .1 5

C a s e 1

C a s e 2

C a s e 3

C a s e 4

C a s e 5

68

A qualitative understanding of solid flow patterns of the domain can be acquired by studying

instantaneous patterns from simulations. The instantaneous void fraction contours for six consecutive

times from 5 – 10 s (t = 1 s) for cases 2-5 are presented in Figure ‎5.3 - Figure ‎5.6, respectively. The

instantaneous void fraction contours can be used to understand bubble formation, growth and

coalescence, and solid circulation and concentration. The instantaneous void fraction contours for case 1

confirm that although the bed is not fluidized yet, it is at the verge of fluidization and the velocity of this

case is very close to the fluidization velocity, where small bubbles form, but the bed height has not

changed. Since case 1 is not fluidized yet, the instantaneous contours of this case are not presented here.

As inlet velocity increases, the bed fluidizes and bubbles form. Formation of bubbles takes place near the

distribution plate where solids move toward the top of the reactor due to the jet flow. As bubbles rise,

their diameter increases until they meet the freeboard and burst.

Figure 5.3. The instantaneous void fraction contours for six concessive times from 5 – 10 s for case 2.

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

69

‎


r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

70


r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

71


Contours of time-averaged void fraction for all five cases are presented in Figure ‎5.7. Void fraction

contours confirm that case 1 is not fluidized, but it is on the verge of fluidization due to the presence of

some small bubbles in the bed. As inlet gas velocity through the annulus increases from 8.22 cm/s (case

2) to 17.75 cm/s (case 4), the bed height expansion increases from 8 cm to almost 11 cm, 15 cm, and 17

cm, respectively. For case 2-4, there are regions of higher concentration of solids near the walls, in the

vicinity of the inlet and near the center of the reactor. The solid concentration in these regions, however,

decreases considerably as inlet gas velocity increases and particles are displaced. For all three cases (case

2-4), the void fraction increases gradually toward the top of the bed where it reaches to the maximum

value at the freeboard. Comparing case 3 and case 5, the bed height expands slightly as jet velocity

increases, which can be due to interference of jet gas velocity in the facilitation of the fluidization

process. The regions of higher solids concentration are similar for both cases.

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

r (c m )

2 .5

5

1 0

1 5

2 0

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

72

Case 1 Case 2 Case 3 Case 4 Case 5

Figure 5.7. Time-averaged void fraction contours for Geldart B particles.

Figure ‎5.8 shows time-averaged gas temperature contours for all five cases. Gas temperatures are

relatively uniform for all of the cases except case 1 where higher temperature occurs in the freeboard due

to homogenous combustion in this region. Gas temperatures for this case show the occurrence of some

reactions in the bed that increase the local bed temperature. As annulus gas velocity increases from case 2

to case 4, the gas temperature decreases, which is due to the reduction of reactions occurring as was seen

in Figure ‎5.2. The jet velocities do not affect the gas temperature whereas temperatures for case 3 and

case 5 are very similar.

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5 5

5

1 0

1 50 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

73


Figure 5.8. Time-averaged gas temperature contours for Geldart B particles.

The CO mass fraction contours for all five cases are presented in Figure ‎5.9. The CO mass fractions

are more uniform for case 2-5 where mass fraction increases gradually toward the top of the reactor. The

non-uniform mass fraction for case 1 is in accordance with the non-uniform gas temperature where local

reactions take place in the bed. For all cases, the lower values of CO mass fraction are observed near the

annulus. As inlet annulus gas velocity increases from case 2 to 5, the height of this region increases and

areas of low mass fraction penetrate deeper in the bed. Comparing case 3 and case 5, there are more

regions of higher CO mass fraction along the reactor wall for case 5 than are present for case 3.

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5 5

5

1 0

1 51 0 0 0

9 0 0

8 0 0

7 0 0

6 0 0

5 0 0

4 0 0

3 0 0

2 0 0

1 0 0

Tg

(C )

74


Figure 5.9. Time-averaged CO mass fraction contours for Geldart B particles.

5.2.2. Effects of particle classifications on gasification process

As the next attempt in coal gasification modeling, two cases with Geldart A particles are modeled.

For both cases, the diameter and density of coal particles are 62 μm and 1.38 g/cm3 that are classified as

Geldart A particle, unlike the Geldart B particles discussed in Section 5.2.1. For the first case, the rest of

data are the same as section 5.2.1. For the second case of this section, however, the physical data (particle

diameter, density, initial void fraction and bed height) consistent with the data used in the fluidization

section (section 4.1, Table ‎4.1) is employed. Gas and solid velocities of case 2 (presented in Table ‎5.3)

are identical for both cases; therefore, the results are comparable with this case. The initial data for both

cases along with case 2 from section 5.2.1 are presented in Table ‎5.4. For both cases, the geometry of the

reactor is the same as Figure ‎5.1 and the same reactions as section 5.2.1 are modeled. The initial

temperature of the reactor is 1000 ºC and Air as the gasification agent will be employed. After evaluating

the results in this section, for the second case, the results using two other gasification agents, N2 and CO2,

will be compared with the air medium in the next section.

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5 5

5

1 0

1 50 .2 6

0 .2 3

0 .2

0 .1 7

0 .1 4

0 .1 1

0 .0 8

0 .0 5

0 .0 2

XC O

75

Table 5.4. Initial conditions for studying the effects of particle classifications

Properties Case 1 Case 2 Case 3

�̅�𝑝(μm) 62 62 1000

𝜌𝑠 (g/cm3) 1.38 1.38 1.0

Geldart classification A A B

ℎ0 (cm) 8 1.34 8

𝜀𝑚𝑓 0.5 0.59 0.5

Initial temperature (ºC) 1000 1000 1000

Air inlet temperature (ºC) 27 27 27

The mass fractions of CH4, H2, and CO versus time for air as the gasification agent for all three cases

are presented in Figure ‎5.10. All mass fractions are averaged spatially over the entire reactor height and

width. For all three cases, the trend of mass fraction of CH4 and H2 are very similar but differ by one

order of magnitude. For both gaseous compounds in all three cases, mass fraction increases with time.

Referring to Table ‎5.4, case 1 and case 3 have similar initial data except for particle classification.

Comparing these two cases shows that both CH4 and H2 mass fractions have close instantaneous values,

however, mass fraction of both gaseous products is slightly lower for case 1. It is hypothesized that the

larger particle diameter of case 3 expedites the reactions rates that take place, such as steam gasification,

CO2 gasification, and methanation. The product of steam gasification and methanation are H2 and CH4,

respectively; hence the rate of these reactions dictates the mass fraction of these gaseous products. The

rate of steam gasification for case 1 is 0.39610-8

and for case 3 is 0.38710-7

and the methanation rate is

0.12110-9

and 0.16110-9

for case 1 and 3, respectively. Comparing cases 1 and 3 with case 2, where the

bed height is shorter and consequently, the quantity of un-reacting ash in the bed is smaller, higher

gasification rates are observed for case 2, which result in higher mass fraction of CH4 and H2 in the bed.

For CO mass fraction, a rapid increasing trend at earlier times (before 4 s) can be observed. The CO mass

fraction for cases 1 and 2 (Geldart A particles) reduces afterwards, and remains almost constant. However

for case 1, the decreasing trend occurs at later times. For Geldart B particles (case 3), the CO mass

fraction remains almost constant after the rapid increasing trend at earlier times. Comparing the CO mass

fraction of case 1 and case 3 that have similar initial conditions shows higher peak for CO mass fraction

for Geldart A particles. Two of the main sources of CO production are carbon combustion and CO2

gasification, where the former one is an exothermic reaction and the latter one is an endothermic reaction.

The carbon combustion is a fast reaction that happens at earlier times and provides the required energy for

the endothermic reactions [144,145]. In this study, the shrinking core model [146,147] is employed.

Based on Eqs. (60-65), carbon combustion rate has an indirect relation with particle diameter, hence, its

rate is higher for Geldart A particles. (In general, decreasing the particle diameter increases the surface

area, which results in higher reaction rates for surface reactions.) As a result, higher CO mass fraction for

76

case 1 at earlier times is observed. The average rate of carbon combustion at earlier times (between 0-4s)

for case 1 is 0.52810-4

and for case 3 is 0.45210-4

. Consequently, the consumption of the reactable

solids in the bed for smaller particles of Geldart A is faster, for which CO mass fraction decreases for

Geldart A particles after all material in the bed reacts. Afterward, the solids entering the reactor through

the central jet react with gaseous reactants. The carbon combustion rate averaged between 7-10s is

0.73310-6

and 0.36410-4

for case 1 and case 3, respectively. Due to the shorter bed height for case 2, the

reactable materials in the bed finish earlier, consequently, CO production decreases at earlier times

compared to case 1.

(a) (b)

(c)

Figure 5.10. Mass fraction of (a) CH4, (b) H2, and (c) CO versus time for air gasification for Geldart A

particles.

T im e (s )

Ma

ss

fra

cti

on

of

CH

4

0 2 4 6 8 1 00

0 .0 0 1

0 .0 0 2

0 .0 0 3

0 .0 0 4

0 .0 0 5

0 .0 0 6

0 .0 0 7 C a s e 1

C a s e 2

C a s e 3

T im e (s )

Ma

ss

fra

cti

on

of

H2

0 2 4 6 8 1 00

0 .0 0 0 1

0 .0 0 0 2

0 .0 0 0 3

0 .0 0 0 4

C a s e 1

C a s e 2

C a s e 3

T im e (s )

Ma

ss

fra

cti

on

of

CO

0 2 4 6 8 1 00

0 .0 5

0 .1

0 .1 5

0 .2

C a s e 1

C a s e 2

C a s e 3

77

The temperature changes in the reactor versus time that is presented in Figure ‎5.11can be used to

confirm trends observed for CO mass fractions in Figure ‎5.10 (c). Temperatures are averaged spatially

over the entire reactor height and width. The exothermic carbon combustion reaction maintains the initial

temperature until the time where CO mass fraction shows the decreasing trend for case 1 and case 2. After

that the temperature decreases for these two cases. For case 3, however, the temperature remains almost

constant that is consistent with the observed CO mass fraction trends.

The gas temperature and mass fraction contours of H2 and CO for case 2 are presented in Figure ‎5.12.

The overall temperature is lower than the results in Figure ‎5.8, where large regions of lower temperature

occur near the bottom of the reactor. The temperature contours confirm the easy fluidization of material in

the bed, where more reactions happen in higher parts of the reactor. The H2 mass fraction contour shows

that the higher concentrations occur along the jet stream. The CO mass fraction contour range is

considerably lower than the results shown in Figure ‎5.9. Lower mass fraction regions occupy a large area

near the bottom of the reactor and increase gradually toward the reactor exit. The regions of higher mass

fraction are located near the center of the reactor, which can be due to the fact that the bed material

already reacted, and the CO created is from solids fuel entering the reactor, which accumulate along the

reactor center where jet is located.

Figure 5.11. Gas temperature versus time for air gasification for Geldart A particles.

T im e (s )

Te

mp

era

ture

(C)

0 2 4 6 8 1 07 0 0

8 0 0

9 0 0

1 0 0 0

1 1 0 0

C a s e 1

C a s e 2

C a s e 3

78

(a) (b) (c)

Figure 5.12. Time-averaged contours of (a) gas temperature, (b) H2 mass fraction, and (c) CO mass fraction

for Geldart A particles.

5.2.3.Effects of gasification medium on gasification process

The gas yields obtained from coal gasification of case 2 presented in section 5.2.2 (data from

Table ‎5.4) in two other mediums, N2, and CO2, are evaluated and the results for all three gases are

compared in Table ‎5.5. In order to keep consistency between results, all mole fractions are calculated on a

CO2, N2, O2 and H2O free basis. Gasification temperature for all three cases is 1000ºC.

Table 5.5. Gas yields from coal gasification in different mediums for Geldart A particles

Gasification agent CO (%) H2 (%) CH4 (%)

Air 92 2.7 5.3

N2 6.3 31.7 62

CO2 9.1 30.7 60.2

r (c m )

2 .5 5 7 .5

5

1 0

1 5

2 0

1 0 0 0

9 0 0

8 0 0

7 0 0

6 0 0

5 0 0

4 0 0

3 0 0

2 0 0

1 0 0

Tg

(C )

r (c m )

2 .5 5 7 .5

5

1 0

1 5

2 0

0 .0 1 1

0 .0 1

0 .0 0 9

0 .0 0 8

0 .0 0 7

0 .0 0 6

0 .0 0 5

0 .0 0 4

0 .0 0 3

0 .0 0 2

0 .0 0 1

XH 2

r (c m )

2 .5 5 7 .5

5

1 0

1 5

2 0

0 .0 8 5

0 .0 7 5

0 .0 6 5

0 .0 5 5

0 .0 4 5

0 .0 3 5

0 .0 2 5

0 .0 1 5

0 .0 0 5

XC O

79

In general, the gas yields from coal gasification are similar for N2 and CO2 agents but very different

from gas yields for air medium. For air, the major component is CO but for N2 and CO2, the major

component switches from CO to CH4. The H2 molar fraction for N2 and CO2 agents are almost equal but

have a significant difference when compared with the air medium. The reason for these differences could

be due to the presence of sufficient amounts of O2 in the reactor using air, which results in higher rates of

carbon combustion. The sole product of this reaction is CO, which explains the high amount of CO in the

gas yields. Due to the similarity in gas yields from coal gasification in N2 and CO2 mediums, the mass

fraction of CH4, H2, and CO versus time for these two agents are compared together and presented in

Figure ‎5.13. All mass fractions are averaged spatially over the entire reactor height and width. The trend

of CH4 and H2 mass fractions are similar using CO2 and N2 agents, however, the CH4 mass fraction is

higher than H2 mass fraction by one order of magnitude. As time increases, the mass fraction of both

gases increases. It is interesting to note that CO mass fraction is higher using CO2 than N2 as the

gasification agent. The reason is the abundance of CO2 in the reactor using CO2 as an agent, which results

in higher CO2 gasification rates that yield CO as the sole product. At later times, however, CO mass

fraction using CO2 and N2 mediums are almost equal, which could be due to the fact that after a while, all

the bed materials react such that the created CO is from the solids fuel entering the reactor, which is equal

amounts in both cases.

The temperature changes in the reactor versus time for N2 and CO2 agents are presented in

Figure ‎5.14. Temperatures are averaged spatially over the entire reactor height and width. For both

mediums, temperatures decrease as time increases, which is due to the occurrence of endothermic

reactions in the reactor such as steam and CO2 gasification.

80

(a) (b)

(c)

Figure 5.13. Mass fraction of (a) CH4, (b) H2, and (c) CO versus time for CO2 and N2 mediums for Geldart A

particles.

T im e (s )

Ma

ss

fra

cti

on

of

CH

4

0 2 4 6 8 1 00

0 .0 0 2

0 .0 0 4

0 .0 0 6

C O2

m e d iu m

N2

m e d iu m

T im e (s )

Ma

ss

fra

cti

on

of

H2

0 2 4 6 8 1 00

0 .0 0 0 1

0 .0 0 0 2

0 .0 0 0 3

0 .0 0 0 4

C O2

m e d iu m

N2

m e d iu m

T im e (s )

Ma

ss

fra

cti

on

of

CO

0 2 4 6 8 1 00

0 .0 0 0 5

0 .0 0 1

0 .0 0 1 5

0 .0 0 2 C O2

m e d iu m

N2

m e d iu m

81

Figure 5.14. Gas temperature versus time for air gasification for Geldart A particles.

5.3. Co-gasification of coal and biomass

5.3.1. Single gasification reaction

Co-gasification of coal and biomass, unlike coal gasification, has not been studied enough in the

literature. This part of the study is to take a step forward to shed light on this technology, however, due to

the lack of the needed experimental data on reaction rates and thermodynamics properties of biomass, the

limited data found in literature are employed [72,148]. In a paper by Brown et al. [72], CO2 gasification

was investigated and the catalytic effects of using switchgrass were studied. The activation energy and

pre-exponential factor for the CO2 gasification for coal, switchgrass and their 50:50 mixture by

employing an Arrhenius reaction rate were found [72]. The experimental results presented by Brown et al.

[72] show a considerable catalytic effect of switchgrass in the gasification mixture in terms of gasification

rate. The paper did not provide all the required information to numerically model the experiments, so

additional assumptions will be employed herein with care taken to be as consistent as possible with the

experimental conditions [72].

The first step in the experiments [72] was to preheat the bed material to the gasification temperature

using pure N2. After the preheat process, CO2 was substituted for N2 in the gasification experiments. To

be consistent with their procedure in the simulations, the initial condition of the bed is defined as pure N2

and coal is at the gasification temperature of 1000 ºC. However, the boundary condition is defined as

pure CO2 entering the reactor through annulus around the jet (Figure ‎5.1).

T im e (s )

Te

mp

era

ture

(C)

0 2 4 6 8 1 07 0 0

8 0 0

9 0 0

1 0 0 0

1 1 0 0

C O2

m e d iu m

N2

m e d iu m

82

The rest of the required data to numerically model the experiments of Brown et al. were extracted

from the coal gasification case presented in Section 5.2.1. Since only the CO2 reaction rate is provided by

Brown et al. [72], only this reaction is considered; however, subsequent simulations herein will include

additional gasification reactions.

Simulations for switchgrass and coal (Illinois No. 6) are modeled to replicate the experiments of

Brown et al. The gasification agent is CO2, which enters the reactor through an annulus around the jet

with a velocity of 8.22 cm/s (case 2 in Table ‎5.3). The dimensions of the reactor were not reported by

Brown et al. [72], so the geometry presented in Figure ‎5.1 is employed. Diameter and density of coal

particles are 0.1 cm and 1 g/cm3 and for switchgrass particles are 0.1 cm and 0.4 g/cm

3, respectively. A

summary of all initial simulation data is presented in Table ‎5.6. Thermodynamics properties of

switchgrass were found in literature [148] for low temperatures (from 310-350 C). For use in the work

herein, the data is extrapolated to obtain thermodynamics properties up to 1000 C.

Table 5.6. Simulation initial data for co-gasification

Properties Coal (Illinois No. 6) Switchgrass

�̅�𝑝(μm) 1000 1000

𝜌𝑠 (g/cm3) 1 0.4

ℎ0 (cm) 8

𝜀𝑚𝑓 0.5

Initial temperature (ºC) 1000

Inlet air temperature (º C) 27

A comparison between gasification of coal, switchgrass and their 50:50 mixture using time averaged

void fraction contours is presented in Figure ‎5.15. Void fraction contours show that all three beds fluidize.

Due to lower density of switchgrass particles, beds containing larger amounts of switchgrass expand

higher. For both coal and 50% mass ratio of switchgrass, there are regions of solid concentration near the

bottom of the reactor. For coal, solid particles concentration is also high along the reactor walls; however,

for the coal-switchgrass mixture, these regions mostly occur along the reactor centerline.

Figure ‎5.16 shows the CO2 gasification reaction rate contours for gasification of coal, switchgrass and

their mixture. The reaction rate for CO2 gasification is higher for the coal-switchgrass mixture compared

to the bed of coal or switchgrass, which is consistent with the results of Brown et al. [72]. Switchgrass

shows the lowest reaction rates in spite of its higher pre-exponential factor compared to coal. (The

activation energy for both materials is about the same.) The main reason for the different reaction rates of

coal and switchgrass is their different thermodynamics properties that results in different solid

temperature in the bed. Based on the Arrhenius reaction rate formula, lower switchgrass temperatures are

83

the reason for lower reaction rates. (The solid temperature contours for both materials are presented in

Appendix B.) To further investigate the reaction rates of the beds, the spatially averaged reaction rate over

one-half the reactor height (0 to 15 cm) and the entire reactor width for all three beds were calculated. For

the coal-switchgrass mixture, the average reaction rate is 1.510-5

, which is considerably larger (an order

of magnitude) than the beds containing only coal or switchgrass with average values of 7.910-6

and

3.310-6

, respectively. For all three beds, the occurrence of reactions is mostly near the walls and the

centerline of the reactor.

The CO mass fraction contours for gasification of coal, switchgrass and their mixture are presented in

Figure ‎5.17. Since CO is the sole product of CO2 gasification, the CO mass fraction contours must be

consistent with the reaction rate contours, where the coal-switchgrass mixture has the highest CO mass

fraction and switchgrass has the lowest. For all three beds, there are regions of low mass fraction near the

bottom of the reactor that increases gradually toward the reactor exit. The length of these regions,

however, is different for each bed because of the difference in reaction rates. It is also partially due to the

different particle densities where bed materials transport easier for beds with higher mass ratios of

switchgrass.

Coal Switchgrass Coal-switchgrass

‎‎‎Figure 5.15. Time-averaged void fraction contours for coal, switchgrass and coal-switchgrass (50:50).

r (c m )

2 .5 5

5

1 0

1 5

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

r (c m )

2 .5 5

5

1 0

1 5

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

r (c m )

2 .5 5

5

1 0

1 5

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

84


Figure 5.16. Time-averaged reaction rate for CO2 gasification reaction contours for coal, switchgrass and

coal-switchgrass (50:50).


Figure 5.17. Time-averaged mass fraction of CO contours for coal, switchgrass and coal-switchgrass (50:50).

r (c m )

2 .5 5

5

1 0

1 5

2 .2 E -0 5

2 E -0 5

1 .8 E -0 5

1 .6 E -0 5

1 .4 E -0 5

1 .2 E -0 5

1 E -0 5

8 E -0 6

6 E -0 6

4 E -0 6

2 E -0 6

R e a c t io n

r a te

r (c m )

2 .5 5

5

1 0

1 5

2 .2 E -0 5

2 E -0 5

1 .8 E -0 5

1 .6 E -0 5

1 .4 E -0 5

1 .2 E -0 5

1 E -0 5

8 E -0 6

6 E -0 6

4 E -0 6

2 E -0 6

R e a c t io n

r a te

r (c m )

2 .5 5

5

1 0

1 5

2 .2 E -0 5

2 E -0 5

1 .8 E -0 5

1 .6 E -0 5

1 .4 E -0 5

1 .2 E -0 5

1 E -0 5

8 E -0 6

6 E -0 6

4 E -0 6

2 E -0 6

R e a c t io n

r a te

r (c m )

2 .5 5

5

1 0

1 5

0 .0 3 2

0 .0 2 9

0 .0 2 6

0 .0 2 3

0 .0 2

0 .0 1 7

0 .0 1 4

0 .0 1 1

0 .0 0 8

0 .0 0 5

0 .0 0 2

XC O

r (c m )

2 .5 5

5

1 0

1 5

0 .0 3 2

0 .0 2 9

0 .0 2 6

0 .0 2 3

0 .0 2

0 .0 1 7

0 .0 1 4

0 .0 1 1

0 .0 0 8

0 .0 0 5

0 .0 0 2

XC O

r (c m )

2 .5 5

5

1 0

1 5

0 .0 3 2

0 .0 2 9

0 .0 2 6

0 .0 2 3

0 .0 2

0 .0 1 7

0 .0 1 4

0 .0 1 1

0 .0 0 8

0 .0 0 5

0 .0 0 2

XC O

85

Since only one reaction is considered in this section, the product gas yields at the exit are not usable

to study the possible synergetic effects of co-gasification (note that Section 5.3.2 will model the complete

set of reactions). However, the effects of adding switchgrass to coal can be examined using CO mass

fraction profiles along the reactor radius. The data are presented in Figure ‎5.18 and averaged over one-

half of the reactor height for coal, switchgrass, and their mixture. The radial distributions of the profiles in

Figure ‎5.18 demonstrate the synergetic effect of co-gasification on CO gas yields. The CO mass fraction

for co-gasification is higher than the summation of the CO mass fraction from the separate gasification of

coal and switchgrass.

Figure 5.18. CO mass fraction profiles along the reactor radius for coal, switchgrass, and coal-switchgrass

(50:50).

5.3.2.Complete set of reactions

As the next attempt, a co-gasification case using the complete set of gasification reactions presented

in the section 5.2.1 is modeled. The reaction rates and properties for coal and switchgrass for CO2

gasification will use the values reported by Brown et al. [72]. For the other four reactions, there is no

available information for switchgrass or mixtures with coal. Therefore, the reaction rate coefficients for

Illinois No. 6 coal (see Table ‎5.1) will be used to model switchgrass. However, the mass fraction in the

reaction rate formulas (Eqs. 54-69) will be different for coal and switchgrass, therefore the reaction rates

will be different. For this case, all other data for simulation is the same as section 5.3.1, where a summary

of initial data can be found in Table ‎5.6.

r (c m )

Ma

ss

fra

cti

on

of

CO

0 .5 1 1 .5 2 2 .50

0 .0 1

0 .0 2

0 .0 3

0 .0 4

0 .0 5c o a l

s w itc h g r a s s

5 0 :5 0 m ix tu re

86

In order to investigate the synergetic effect of co-gasification of coal and biomass, Table ‎5.7 shows

the product gas yields of coal, switchgrass, 50:50 and 70:30 mass ratios of coal-switchgrass gasification.

Noting that there is no published data for the reaction rate coefficients for the 70:30 mass ratio, the same

reaction rate for the 50:50 mass ratio is employed. To determine if there is a synergistic effect when

mixing coal and switchgrass, an estimate for the product gas yields with 50:50 and 70:30 coal-switchgrass

are calculated using the gas yields of separate beds of coal and switchgrass. Molar fractions of each

product from separate beds of coal and switchgrass in the mixture are added together and then the whole

gas yield of the mixture is estimated on a 100% basis. The gas yields predicted from the co-gasification

simulations show an interaction between coal and switchgrass but the effects are not always positive. For

example, the estimated value of H2 for 70:30 mass ratio is 15.3%, however, 7.5% is predicted from the

simulations. For both H2 and CH4, the predicted gaseous compounds from simulations are lower than the

estimated value assuming separate gasification of coal and switchgrass. However, there are clear

synergetic effects for CO gas product. It seems that the co-gasification decreases the H2 and CH4 gaseous

compounds but increases the CO yield. The observation of synergetic effect for CO gas yields is in

accordance with results of section 5.3.1 when only one reaction was modeled.

To investigate the effects of gasification temperature on co-gasification, three different temperatures

for the 50:50 mass ratio of coal-switchgrass are considered: 700 ºC, 800 ºC, and 900 ºC. For these

simulations, all gasification reactions presented in section 5.2.1 are considered and simulation conditions

except for temperature are presented in Table ‎5.6. The gas yields on a CO2, N2, O2 and H2O free basis

obtained from co-gasification are compared in Table ‎5.8 . Results show that H2 and CH4 are the major

compounds produced at 700 ºC and 800 ºC, however, at 900 ºC, the CO and H2 yields are both significant

and comparable. For all three temperatures, CH4 yields remain the most pronounced gas in the products.

The yields of H2 and CH4 decrease as temperature increases, unlike CO, which increases as temperature

increases.

In order to further investigate the effects of temperature on co-gasification of coal and switchgrass,

the void fraction contours at three gasification temperature are presented in Figure ‎5.19. The void fraction

contours show large regions of solids concentration near the bottom of the reactor. As temperature

increases, these regions shrink due to the higher reaction rates. In general, the solids consumption for co-

gasification of coal-switchgrass increases at higher temperatures. The increase of reaction rates with

temperature is demonstrated in Figure ‎5.20 for the CO2 gasification rate. The reaction rate at different

temperature is an order of magnitude different for each 100ºC temperature increase.

87

Table 5.7. Gas yields from coal and switchgrass gasification and co-gasification mixtures.

Exit gas

products

Coal

(simulation)

Switchgrass

(simulation)

50:50 mass

ratio of

coal-

switchgrass

(simulation)

50:50 mass

ratio of

coal-

switchgrass

(estimated)

70:30 mass

ratio of

coal-

switchgrass

(simulation)

70:30 mass

ratio of

coal-

switchgrass

(estimated)

CO (%) 69.3 21.3 73.2 45.3 77.9 54.9

H2 (%) 10.4 26.6 9.1 18.5 7.5 15.3

CH4 (%) 20.2 52.1 17.7 36.2 14.6 29.8

Table 5.8. Gas yields from co-gasification in different temperatures

Temperature (ºC) CO (%) H2 (%) CH4 (%)

700 6.3 31.7 62

800 9 30.8 60.2

900 26.4 25 48.7

T=700ºC T=800ºC T=900ºC

‎‎‎‎Figure 5.19. Time-averaged void fraction contours for different gasification temperature.

r (c m )

2 .5 5

5

1 0

1 5

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

r (c m )

2 .5 5

5

1 0

1 5

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

r (c m )

2 .5 5

5

1 0

1 5

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

g

88

T=700ºC T=800ºC T=900ºC

‎‎‎‎Figure 5.20. Time-averaged rate of CO2 gasification reaction contours for different gasification

temperature for 50:50 mass ratio of coal-switchgrass.

5.3.3.Complete set of reactions with Geldart A particles

In order to complete the investigation of co-gasification using coal and switchgrass, in this part of the

study, Geldart A particles are investigated. Diameter and density of coal particles are 62 μm and 1 g/cm3

and for switchgrass particles are 145 μm and 0.4 g/cm3, respectively. For this case, all other data for the

simulations are the same as section 5.3.2, where a summary of initial data can be found in Table ‎5.6.

The synergetic effect of co-gasification of coal and biomass, is analyzed using the product gas yields

at the exit. Table ‎5.9 shows the results for coal, switchgrass and 50:50 mass ratios of coal-switchgrass

gasification. The estimate for the product gas yields with 50:50 coal-switchgrass are calculated using the

gas yields of separate beds of coal and switchgrass.

The gas yields predicted from the co-gasification simulations show similar interaction between coal

and switchgrass as presented in section 5.3.2 (Table ‎5.8), where the predicted values from the simulations

for H2 and CH4 are lower than the estimated values assuming separate gasification of coal and

switchgrass. The synergetic effect for CO gas product, however, is clear and even more significant than

using Geldart B particles.

r (c m )

2 .5 5

5

1 0

1 5

4 .0 E -0 7

3 .6 E -0 7

3 .2 E -0 7

2 .8 E -0 7

2 .4 E -0 7

2 .0 E -0 7

1 .6 E -0 7

1 .2 E -0 7

8 .1 E -0 8

4 .1 E -0 8

1 .0 E -0 9

R e a c t io n

r a te

r (c m )

2 .5 5

5

1 0

1 5

4 .0 E -0 7

3 .6 E -0 7

3 .2 E -0 7

2 .8 E -0 7

2 .4 E -0 7

2 .0 E -0 7

1 .6 E -0 7

1 .2 E -0 7

8 .1 E -0 8

4 .1 E -0 8

1 .0 E -0 9

R e a c t io n

r a te

r (c m )

2 .5 5

5

1 0

1 5

4 .0 E -0 7

3 .6 E -0 7

3 .2 E -0 7

2 .8 E -0 7

2 .4 E -0 7

2 .0 E -0 7

1 .6 E -0 7

1 .2 E -0 7

8 .1 E -0 8

4 .1 E -0 8

1 .0 E -0 9

R e a c t io n

r a te

89

Table 5.9. Gas yields from coal and switchgrass gasification and co-gasification of 50:50 mixture.

Exit gas

products

Coal

(simulation)

Switchgrass

(simulation)

50:50 mass

ratio of

coal-

switchgrass

(simulation)

50:50 mass

ratio of

coal-

switchgrass

(estimated)

CO (%) 46.7 41.3 91.6 44

H2 (%) 18 19.9 2.8 19

CH4 (%) 35.3 38.8 5.6 37

5.4. Conclusions

Numerical simulations of gasification in fluidized beds were used to better understand the gasification

of coal, biomass and their mixtures. In this study, a five-reaction model was employed. First, coal

gasification of larger coal particles was modeled for varying annulus and jet velocities. The results

showed that increasing the annulus gas inlet velocity has a reverse effect on the production of CH4, H2,

and CO, which could be due to the shorter residence time of the gasification agent in the bed. Thus, the

gas and solids have less time to react and the gas temperature reduces. Higher annulus gas velocities also

increased the bed expansion. Increasing the jet velocities, on the other hand, had a positive effect on the

products. The reason could be due to higher concentration of solid in the rector that is brought in by

higher jet velocities. The bed height expansion is slightly more as jet velocities increases but gas

temperature in the reactor does not show any significant change.

As a next step, coal gasification of fine particles of Geldart A was investigated. The results showed

promising effects of using Geldart A particles where high CO mass fractions were observed, however, H2

and CH4 mass fractions were low. The gas yields obtained from Geldart A coal gasification in three

mediums: N2, CO2 and air, were compared and results showed a similarity between gas yields in N2 and

CO2 mediums where the major component was CH4. However, the gas yields from air were very different

with major component being CO. The reason could be due to the presence of sufficient amounts of O2 in

the reactor using air agent, which increases the reaction rate of the carbon combustion reaction with high

rate. The sole product of this reaction is CO, which explains the high amount of CO in the gas yields. It

should be noted that for the other two mediums (N2, CO2), carbon combustion does not take place.

Co-gasification of coal and switchgrass was studied, and when data was not available from the

literature, reasonable assumptions were made to model the gasification. The results showed higher bed

expansion for beds containing larger amounts of switchgrass due to lower density of switchgrass particles.

The reaction rate for CO2 gasification was higher for the coal-switchgrass mixture compared to the bed of

coal or switchgrass. Modeling only CO2 gasification, the synergetic effect of co-gasification on CO gas

90

yields was observed, where the CO mass fraction is higher than the summation of CO mass fraction from

separate gasification of coal and switchgrass. Modeling the complete set of the five-reaction gasification

model showed similar results in terms of synergetic effects of co-gasification on CO gas yields. However,

co-gasification decreased the H2 and CH4 gaseous compounds yields. Investigating the effects of

gasification temperature on gas yields using a CO2, N2, O2 and H2O free basis showed pronounced effects

on CO gas yields but negative effects on H2 and CH4 gas yields. Regardless of gasification temperature,

CH4 yields remain the most pronounced gas in the products. It was also found that the solids consumption

and CO2 gasification rate for co-gasification of coal-switchgrass increases at higher temperatures. Co-

gasification of Geldart A particles of coal and switchgrass were modeled and results proved the existence

of the synergetic effect of co-gasification in terms of CO gas yields. The significance of this effect was

found to be more significant using Geldart A particles.

91

Chapter 6. Conclusions and Future work

6.1. Summary

Gas fluidized beds of Geldart A particles are used extensively in different industries, but more work is

needed to thoroughly understand the fluidization behavior of these fine particles. It is well known that as

Geldart A particles are fluidized, there exists a homogenous region where no bubbles form before the

onset of heterogeneous fluidization. In the homogeneous fluidization region, the bed height increases and

pressure spikes before reducing to a constant pressure drop across the bed during heterogeneous

fluidization. In order to broaden our knowledge and design fluidized bed reactors that use fine particles,

CFD can be an extremely powerful tool. In this study, efforts ensued to increase our knowledge of

fluidization behavior of Geldart A particles using the CFD code MFIX. An extensive Eulerian-Eulerian

numerical study of coal and biomass fluidization and their mixtures were executed. Simulations were

compared and validated with experiments provided by experimental group in Utah state university.

To analyze gas-solids hydrodynamic behavior of the fluidized bed, the predictions for pressure drop

across the bed and bed height were validated with experiments. However, the Eulerian-Eulerian model

was not able to predict the overshoot of pressure in the homogeneous region, although good agreement

with experiments was observed. Pressure drop for the coal-poplar mixtures were higher than single solids

phases for each material. It was found that the 90:10 mass ratio exhibited bed height trends similar with

the single solids phase of coal, and as the poplar wood mass ratio increased to 30%, the bed height was

similar to the single solids phase of poplar wood.

Mixing is an important parameter that has significant impact on particle contact and consequently on

the efficiency of the process, heat transfer and providing uniform temperature throughout the fluidized

bed. The fluidization behavior of three different mixtures of coal and poplar wood were studied. To

investigate the mixing behavior of particles, volume fractions, velocities and, mass flux were examined.

Although instantaneous solid volume fraction contours indicated good mixing characteristics for all mass

ratio mixtures, solids mass flux and velocity vectors showed a tendency of better mixing for mixtures

with higher poplar mass ratio. A quantitative analysis of the mixing index confirmed that as the poplar

wood mass ratio increased, the quality of mixing improved, with an average mixing index of 62%.

Therefore, reasonable amounts of biomass can be added with coal without adverse effects of segregation

or elutriation, while reducing the use of a fossil fuel.

In order to improve numerical predictions using an Eulerian-Eulerian model, choosing a suitable drag

model is important. Literature has reported commonly-used drag models, which show excellent

performance for larger particles (e.g., Geldart B and D), do not work well with fine particles of Geldart A.

The results of this study, however, proved that if static regions of mass in fluidized beds are neglected,

92

these drag models work well with Geldart A particles. In this study, predictions using seven different drag

models were investigated by studying the bed height and void fraction. It was shown that the performance

of the drag models could be grouped into two categories. The first group included the BVK, HYS, Hill-

Koch-Ladd models, which mostly over-predicted bed height for a single solids phase, however, BVK and

HYS model predictions agreed well with experiments for coal poplar wood mixtures. The second group

included the Gidaspow, Gidaspow-Blend, Wen-Yu and Syamlal-O’Brien drag models. This group gave

good predictions for both single solids phases and their mixtures, however, the Gidaspow-blend model

proved to be the more reliable drag model for both single and binary mixtures. These conclusions were

also substantiated by examining void fraction profiles to demonstrate particle distribution. Studying the

nature of the two groups of models revealed that the first group used numerical simulations to derive drag

force models, whereas the second group models were derived from empirical correlations.

It is important to mention that none of the studied drag models with used correlation for dead zone

were capable of capturing the overshoot of pressure drop in homogeneous fluidization region. Hence, new

work was proposed to modify the definition of the critical velocity in the Ergun correlation. Considering

the difference of bed structure in two fluidization regions of Geldart A particles where inter-particle forces

that act effectively in homogeneous fluidization region diminish in heterogeneous fluidization regime, the

proposed procedure treats two fluidization regions differently. Pressure drop and bed height for beds of

coal, poplar wood, and switchgrass particles were compared with experiments. The procedure showed

promising results for predicting fluidization behavior of fine particles.

In this study also, efforts continued to model co-gasification of coal and biomass, however, the lack of

required input data causes this study to suffer from taking a final step in this regard. First, coal

gasification of large (Geldart B) particles was compared to coal gasification of fine (Geldart A) particles.

Results showed that using finer particles in coal gasification had a clear effect on gas yields where CO

mass fraction increased, although H2 and CH4 mass fraction decreased. The higher mass fractions of H2

and CH4 can be due to the smaller particles, which increased the reactions rates of steam gasification, CO2

gasification, and methanation. The gas yields of coal gasification with fine particles on a CO2, N2, O2 and

H2O free basis were also compared using three different mediums: air, N2 and CO2. Gas yields were

found to be very similar for N2 and CO2 gasification agents. For air, CO was found to be a major

component where CH4 was the major component for N2 and CO2 agents. The reason for this difference

might originate from the presence of sufficient amounts of O2 in the reactor when air used as the

gasification agent, resulting in higher rates of carbon combustion that solely produces CO as the product.

The results of coal-switchgrass gasification showed a positive effect on the rate of CO2 gasification,

where CO mass fraction was higher for co-gasification compared to the separate gasification of coal or

switchgrass. Co-gasification with all reactions included showed that there is not a synergetic effect in

93

terms of gas yields of H2 and CH4. The gas yields of CO, however, showed a significant increase during

co-gasification for both Geldart A and Geldart B particles of coal and switchgrass. It was also shown that

the yields of H2 and CH4 decreased as gasification temperature increased, whereas CO decreased as

gasification temperature increased.

6.2. Significance and contribution

The main aim of this dissertation was to computationally model the fluidization and gasification of

Geldart A particles of coal, biomass and their mixtures. While bench-scale experiments have been a

possible way to design reactors, there always have been difficulties in scale-ups from bench- and pilot-

scales to real-size reactors. CFD, especially Eulerian-Eulerian models, is a powerful tool that can help

with scale-up, design and performance of real-size reactors, and without incurring financial costs.

However due to the nature of Geldart A particles, especially their mixtures, Eulerian-Eulerian models had

shown some deficiencies in predicting the fluidization behavior of these fine particles. This study is a

significant contribution in addressing these issues and shedding light on the reason of these deficiencies.

It was found that neglecting to consider deadzones in fluidized beds was the main reason for poor model

predictions. Correcting for the unfluidized regions, it was found that most of the commonly-used drag

models that were reported unsuccessful with Geldart A particles were in fact usable with fine particles.

The performance of these drag models were investigated and the best of them in predicting fluidization

behavior of Geldart A particles and their mixtures were introduced. Quality of mixing that has significant

effects on performance of mixtures were also investigated and it was found that increasing biomass

content to 30% increases the quality of mixing. In addition, a procedure to improve predictions for the

homogeneous fluidization of Geldart A particles was proposed, which is another big step forward in

improving the performance of Eulerian-Eulerian models.

This study also contributed to advancing modeling gasification of coal, biomass and their mixtures.

While this part of the study needed some experimental information that was not available and was

assumed, the significant contribution was employing the Eulerian-Eulerian modeling to predict synergetic

effects of co-gasification. The results show a clear synergetic effect in terms of CO gas yields, but

negative interaction between coal and switchgrass in the mixture in terms of CH4 and H2 product yields.

However, the CO mass fraction increase in most of the cases is more significant than H2 mass fraction

decrease. Co-gasification of Geldart A particles of coal and switchgrass reveal CO gas yields, which is in

accordance with coal gasification findings. Coal gasification of fine particles of Geldart A also show

pronounced effects in terms of using Geldart A particles, where CO mass fraction increases, however, H2

mass fraction slightly decreases.

94

6.3. Future work

This study provided further knowledge on fluidization and gasification behavior of coal, biomass and

their mixtures. Although remarkable progress has been made, only two biomass materials were studied in

this work. Due to the nature of fluidized beds where particle properties have a significant impact on

reactor performance, examining other biomass materials and comparing the results to the current study

would be interesting. Testing the technique presented in section 4.3 to improve Eulerian-Eulerian model

performance in capturing homogeneous fluidization behavior with different Geldart A particles would

prove to be valuable, too.

Also in this study, low inlet gas velocities that that induce moderately bubbling fluidized beds were

investigated. It also would be of interest to study higher inlet velocities that cause turbulent and fast

fluidization. These new studies would require turbulence modeling and comparing their results with

experiments would be interesting.

As presented in Chapter 5, there is a lack of data pertaining to reaction rates and properties for

biomass materials; consequently, reasonable assumptions were made to model the gasification. It would

be recommended to experimentally investigate the shortage of data and computationally model the

gasification with this new-found information. Also, data from the literature were only found for one

mixture mass ratio. With data for different mass ratios, it is of interest to find the most efficient mass ratio

of biomass in the mixture in terms of co-gasification synergetic effects on syngas.

95

References

[1] Chris Higman, 2014, “State of the Gasification Industry: Worldwide Gasification Database 2014

Update,” Gasification Technologies Conference, Washington DC.

[2] Baskara Sethupathy, S., 2012, “Hydrodynamics study on gasification of biomass in a fluidized

bed gasifier,” International Journal of Engineering and Technology, 4(1), pp. 316-323.

[3] Annual energy outlook, 2008, Department of Energy/US Energy Information Administration,

DOE/EIA-0383.

[4] Klara, S. M., 2009, “Biomass for thermal energy and electricity: A research and development

portfolio for the future,” Hearing before the Subcommittee on Energy and Enviroment,

Committee on Science and Technology, US House of Representatives, One hundred eleventh

congress, first session, Serial No. 111-56.

[5] Hartmann, D., Kaltschmitt, M., 1999, “Electricity generation from solid biomass via co-

combustion with coal: Energy and emission balances from a German case study,”Biomass

Bioenergy, 16(6), pp. 397–406.

[6] Biagini, E., Barontini, F., Tognotti, L., 2006, “Devolatilization of Biomass Fuels and Biomass

Components Studied by TG/FTIR Technique,” Industrial and Engineering Chemistry Research,

45(13), pp. 4486-4493

[7] Kumar, A., Jones, D.D.,Hanna, M.A., 2009, “Thermochemical Biomass Gasification: A Review

of the Current Status of the Technology,” Energies 2009, 2(3), pp. 556-581.

[8] Prins, M.J., Ptasinski, K.J., Janssen, F.J.J.G., 2007, “From coal to biomass gasification:

Comparison of thermodynamic efficiency,” Energy, 32(7), pp. 1248-1259.

[9] Cuellar A. D., 2012, “Plant Power: The cost of using biomass for power generation and potential

for decreased greenhouse gas emissions,” Master Thesis, Massachusetts Institute of Technology,

Cambridge.

[10] J. Sun, F. Battaglia, 2006, “Hydrodynamic modeling of particle rotation for segregation in

bubbling gas-fluidized beds”, Chem. Eng. Sci. 61(5) pp. 1470 – 1479

[11] Deza, M., Franka, N.P., Heindel, T.J., Battaglia, F., 2009, “CFD Modeling and X-Ray Imaging of

Biomass in a Fluidized Bed”, ASME J. Fluids Eng., 131, 111303.

[12] Geldart, D., 1973, “Types of gas fluidization,” Powder Technology, 7, pp. 285-292.

[13] Jackson, R., 1998, “The nature and role of effective stress in fluidized systems,”

FLUIDIZATION -ENGINEERING FOUNDATION CONFERENCE; 1-14 Fluidization

Conference; 9th, Fluidization, Engineering Foundation Publication, New York.

[14] Jackson, R., 2000, “The dynamics of fluidized particles,” Cambridge University Press,

Cambridge.

[15] Ye, M., van der Hoef, M. A., and Kuipers, J.A.M., 2005, “The effects of particle and gas

properties on the fluidization of Geldart A particles,” Chemical Engineering Science, 60(16), pp.

4567–4580.

http://www.gasification.org/uploads/eventLibrary/2014_Wednesday_Higman.pdf

http://www.gasification.org/uploads/eventLibrary/2014_Wednesday_Higman.pdf

96

[16] Foscolo, P.U., Gibilaro, L.G., 1984, “A fully predictive criterion for the transition between

particulate and aggregate fluidization,” Chemical Engineering Science, 39(12), pp. 1667-1775.

[17] Kobayashi, T., Kawaguchi, T., Tanaka, T., Tsuji, Y., 2002, “DEM analysis on flow pattern of

Geldart's group A particles in fluidized bed,” Proceedings of the World Congress on Particle

Technology 4 (CD-ROM), July 21-25, Sydney, Australia, pp, 178-185.

[18] Pandit, JK., Wang, XS., Rhodes, MJ., 2006, “On Geldart Group A behaviour in fluidized beds

with and without cohesive interparticle forces: A DEM study,” Powder Technology, 164(3), pp.

130-138.

[19] Ye, M., van der Hoef, M.A., Kuipers, J.A.M., 2004, “A numerical study of fluidization behavior

of Geldart A particles using a discrete particle model,” Powder Technology, 139(2), pp. 129-139.

[20] Anderson, T. B., and Jackson, R., 1967, “A fluid mechanical description of fluidized beds:

Equations of motion,” Industrial & Engineering Chemical Fundamentals, 6(4), pp. 527–539.

[21] Srivastava, A., Sundaresan, S., 2002, “Role of wall friction in fluidization and standpipe flow,”

Powder Technology, 124(1-2), pp. 45–54

[22] Loezos, P. N., Costamagna, P., Sundaresan, S., 2002, “The role of contact stresses and wall

friction on fluidization,” Chemical Engineering Science, 57(24), pp. 5123–5141.

[23] Van Wachem, B.G.M., Schouten, J.C., van den Bleek, C.M., Krishna, R., and Sinclair, J.L., 2004,

”Comparative Analysis of CFD Models of Dense Gas-Solid systems,” AIChE Journal, 47(5), pp.

1035-1051.

[24] Kanholy, S. K., Chodak, J., Lattimer, B. Y., Battaglia, F., 2012, “Modeling and Predicting Gas-

Solid Fluidized Bed Dynamics to Capture Nonuniform Inlet Conditions,” Journal of Fluids

Engineering, 134(11), 111303.

[25] Ferschneider, G., Mege, P., 1996, “Eulerian simulation of dense phase fluidized beds,” Revue de

l’institut français du pètrole, 51(2), pp. 301-307.

[26] Krishna, R., van Baten, J.M., 2001, “Using CFD for scaling up gas-solid bubbling fluidized bed

reactors with Geldart A powders,” Chemical Engineering Journal, 82(1-3), pp. 247-257.

[27] Wang, J., 2009, “A review of Eulerian simulation of Geldart A particles in gas- fluidized beds,”

Industrial & Engineering Chemistry Research, 48(12), pp. 5567–5577.

[28] Lackermeier, U., Rudnick, C., Werther, J., Bredebusch, A., Burkhardt, H., 2001, “Visualization

of flow structures inside a circulating fluidized bed by means of laser sheet and image

processing,” Powder Technology, 114, pp. 71-83.

[29] Harris, A.T., Davidson, J.F., Thorpe, R.B., 2002, “The prediction of particle cluster properties in

the near wall region of a vertical riser,” Powder Technology, 127, pp. 128-143.

[30] Mostoufi, N., Chaouki, J., 2004, “Flow structure of the solids in gas–solid fluidized beds,”

Chemical Engineering Science, 59, pp. 4217-4227.

[31] Huilin, L., Qiaoqun, S., Yurong, H., Yongli, S., Ding, J., Xiang, L., 2005, “Numerical study of

particle cluster flow in risers with cluster-based approach,” Chemical Engineering Science,

60(23), pp. 6757-6767.

http://www.sciencedirect.com/science/article/pii/S000925090500494X






http://www.sciencedirect.com/science/journal/00092509

97

[32] Shuyan, W., Zhiheng, Sh., Huilin, L., Long, Y., Wentie, L., Yonlong, D., 2008, “Numerical

predictions of flow behavior and cluster size of particles in riser with particle rotation model and

cluster-based approach,” Chemical Engineering Science, 63(16), pp. 4116-4125.

[33] McKeen, T., and Pugsley, T., 2003, “Simulation and experimental validation of a freely bubbling

bed of FCC catalyst,” Powder Technology, 129(1-3), pp. 139–152.

[34] Gibilaro, L.G., Di Felice, R., Waldram, S.P., 1985, “Generalized friction factor and drag

coefficient correlations for fluid – particle interactions”, Chem. Eng. Sci. 40 (10), pp. 1817–

1823.

[35] Ye, M., Wang, J., van der Hoef, M. A., Kuipers, J. A. M., 2008, “Two-fluid modeling of Geldart

A particles in gas-fluidized beds,” Particuology, 6(6), pp. 540–548

[36] Wen, C. Y.,Yu, Y. H., 1966, “Mechanics of Fluidization”, Chem. Eng. Prog. Symp. Ser. 62, pp.

100-111.

[37] Parmentier, J. F., Simonin, O., and Delsart, O., 2008, “A Numerical Study of Fluidization

Behavior of Geldart B, A/B and A Particles Using an Eulerian Multifluid Modeling Approach,”

Circulating Fluidized Bed Technology IX, Hamburg, Germany.

[38] Wang, J., Van der Hoef, M.A., and Kuipers, J.A.M., 2009, “Why the two-fluid model fails to

predict the bed expansion characteristics of Geldart A particles in gas- fluidized beds: a tentative

answer,” Chemical Engineering Science, 64(3), 622–625.

[39] Wang, J., van der Hoef, M. A., Kuipers, J. A. M., 2010, “CFD study of the minimum bubbling

velocity of Geldart A particles in gas-fluidized beds,” Chemical Engineering Science, 65(12), pp.

3772-3785.

[40] Lu, B., Wang, W., Li, J., 2011, “Eulerian simulation of gas-solid flows with particles of Geldart

groups A, B and D using EMMS-based meso-scale model”, Chemical Engineering Science 66,

pp. 4624-4635.

[41] Benzarti, S., Mhiri, H., and Bournot, H., 2012, “Drag models for Simulation Gas-Solid Flow in

the Bubbling Fluidized Bed of FCC particles”, World Academy of Science, Engineering and

Technology 61.

[42] Simone, S., Harriott, P., 1980, “Fluidization of fine powders with air in the particulate and the

bubbling regions,” Powder Technology, 26(2), pp.161–167.

[43] Rapagna, S., Foscolo, P.U., Gibilaro, L.G., 1994, “The influence of temperature on the quality of

gas fluidization,” International Journal of Multiphase Flow, 20(2), pp.305–313.

[44] Xie, H.Y., Geldart, D., 1995, “Fluidization of FCC powders in the bubble-free regime: effect of

types of gases and temperature,” Powder Technology, 82(3), pp. 269–277.

[45] Lettieri, P., Newton, D., Yates, J.G., 2002, “Homogeneous bed expansion of FCC catalysts,

influence of temperature on the parameters of the Richardson–Zaki equation,” Powder

Technology, 123(2-3), pp. 221–231.

[46] Sidorenko, I., Rhodes, M.J., 2004, “Influence of pressure on fluidization properties,” Powder

Technology, 141(1-2), pp. 137–154.

http://www.sciencedirect.com/science/article/pii/S0009250908002856







98

[47] Benyahia, S., Arastoopour, , H., Knowlton, T.M., Massah, H., 2000, “Simulation of particles and

gas flow behavior in the riser section of a circulating fluidized bed using the kinetic theory

approach for the particulate phase,” Powder Technology, 112(1-2), pp. 24-33.

[48] Han, I., Chung, C., 2001, “Dynamic modeling and simulation of a fluidized catalytic cracking

process. Part II: Property estimation and simulation,” Chemical Engineering Science, 56(5), pp.

1973-1990.

[49] Jiradilok, V., Gidaspow, D., Damronglerd, S., Koves, W. J., 2006, “Mostofi, R., “Kinetic theory

based CFD simulation of turbulent fluidization of FCC particles in a riser,” Chemical Engineering

Science, 61(17), pp. 5544-5559.

[50] Gao, J., Lan, X., Fan, Y., Chang, J., Wang, G., Lu, C., Xu, C., 2009, “CFD modeling and

validation of the turbulent fluidized bed of FCC particles,” AIChE Journal, 55(7), pp. 1680-1694.

[51] Cui, H., and Grace, J.R., 2007, “Fluidization of Biomass Particles: A Review of Experimental

Multiphase Flow Aspects,” Chemical Engineering Science, 62(1-2), pp. 45-55.

[52] Buekens, AG., Schoeters, JG., 1985, “Modelling of biomass gasification,” In: Overend RP, Milne

TA, Mudge KL, editors. Fundamentals of thermochemical biomass conversion. London: Elsevier

Applied Science Publishers, pp. 619–89.

[53] Yu, L., Lu, J., Zhang, X., Zhang, S., 2007, “Numerical simulation of the bubbling fluidized bed

coal gasification by the kinetic theory of granular flow (KTGF),” Fuel 86, pp. 722–734

[54] Wang, X., Jin, B., Zhong, W., 2009, “Three-dimensional simulation of fluidized bed coal

gasification,” Chemical Engineering and Processing, 48, pp. 695–705.

[55] Kajitani, Sh., Tay, H.L., Zhang, S., Li, Ch.Z., 2013, “Mechanisms and kinetic modelling of steam

gasification of brown coal in the presence of volatile–char interactions,” Fuel, 103, pp. 7-13.

[56] Oevermann, M., Gerber, S., Behrendt, F., 2009, “Euler–Lagrange/DEM simulation of wood

gasification in a bubbling fluidized bed reactor,” Particuology, 7, pp. 307–316.

[57] Pepiot, P., Dibble, C.J., Foust, T.D., 2010, “Computational Fluid Synamics Modeling of Biomass

Gasification and Pyrolysis,” National Renewable Energy laboratory, Golden CO 80401, pp. 273-

298

[58] Gerber, S., Behrendt, F., Oevermann, M., 2010, “An Eulerian modeling approach of wood

gasification in a bubbling fluidized bed reactor using char as bed material,” Fuel 89, pp. 2903–

2917.

[59] Sjöström, K., Chen, G., Yu, Q., Brage, C., Rosén, C., 1999, “Promoted reactivity of char in co-

gasification of biomass and coal: synergies in the thermochemical process, “ Fuel 78, pp. 1189–

1194.

[60] Krerkkaiwan, S., Fushimi C., Tsutsumi, A., Kuchonthara, P., 2013, “Synergetic effect during co-

pyrolysis/gasification of biomass and sub-bituminous coal,” Fuel Processing Technology 115, pp.

11–18

[61] Edreis, E.M.A., Luo, G., Li, A., Xu, C., Yao, H., 2014, “Synergistic effects and kinetics thermal

behaviour of petroleum coke/biomass blends during H2O co-gasification,” Energy Conversion

and Management 79, pp. 355–366.












http://www.researchgate.net/researcher/64217058_Jinsen_Gao

http://www.researchgate.net/researcher/35742195_Xingying_Lan

http://www.researchgate.net/researcher/83202724_Yiping_Fan

http://www.researchgate.net/researcher/35738493_Jian_Chang

http://www.researchgate.net/researcher/35737633_Gang_Wang

http://www.researchgate.net/researcher/83350740_Chunxi_Lu

http://www.researchgate.net/researcher/35736725_Chunming_Xu

99

[62] Gil, M.V., Casal, D., Pevida, C., Pis, J.J., Rubiera, F., 2010, “Thermal behaviour and kinetics of

coal/biomass blends during co-combustion,” Bioresour Technology 101, pp. 5601–5608.

[63] Chen, G., Yu, Q., Brage, C., Sjostrom, K., 2000, “Co-Gasification of Birch Wood and Daw Mill

Coal in Pressurized Fluidized Bed Gasifier: The Investigation into the Synergies in the Process,”

Proceedings of the 1st World Conference on Biomass for Energy and Industry, Sevilla, Spain, pp.

1545–1548.

[64] Tchapda, A.H., and Pisupati, S.V., 2014, “A Review of Thermal Co-Conversion of Coal and

Biomass/Waste,” Energies 7, pp. 1098-1148.

[65] Squir, K.R., Solomon, P.R., Carangelo, R.M.,DiTaranto, M.B., 1986, “Tar evolution from coal

and model polymers: 2. The effects of aromatic ring sizes and donatable hydrogens,” Fuel 65, pp.

833–843.

[66] Coughlin, R.W., Davoudzadeh, F., 1986, “Coliquefaction of lignin and bituminous coal,” Fuel 65,

pp. 95–106.

[67] Hayashi, J., Mizuta, H., Kusakabe, K., Morooka, S., 1994, “Flash copyrolysis of coal and

polyolefin,” Energy Fuels 8, pp. 1353–1359.

[68] Mastral, A.M., Mayoral, M.C., Murillo, R., Callen, M., Garcia, T., Tejero, M.P., Torres, N., 1998,

“Evaluation of synergy in tire rubber-coal coprocessing,” Industrial and Engineering Chemistry

Research 37, pp. 3545–3550.

[69] Ahmaruzzaman, M., Sharma, D.K., 2007, “Coprocessing of petroleum vacuum residue with

plastics, coal, and biomass and its synergistic effects,” Energy Fuels 21, pp. 891–897.

[70] Gao, C., Vejahati, F., Katalambula, H., Gupta, R., 2009, “Co-gasification of biomass with coal

and oil sand coke in a drop tube furnace,” Energy Fuels 24, pp. 232-240.

[71] Srivastava, S.K., Saran, T., Sinha, J., Ramachandran, L.V., Rao, S.K., 1988, “Influence of alkali

on pyrolysis of coals,” Fuel, 67, pp. 1683–1684.

[72] Brown, R.C., Liu, Q., Norton, G., 2000, “Catalytic effects observed during the co-gasification of

coal and switchgrass,” Biomass Bioenergy, 18, pp. 499-506.

[73] Cordero, T., Rodriguez-Mirasol, J., Pastrana, J., Rodriguez, J.J., 2004, “Improved solid fuels from

co-pyrolysis of a high-sulphur content coal and different lignocellulosic wastes,” Fuel, 83, pp.

1585-1590.

[74] Aghalayam, P., Baretto, J., Ganesh, A., Kauchali, S., 2008, “Pyrolysis of Sawdust-Lignite

Blends,” 25th International Pittsburgh Conference, Pittsburgh, PA.

[75] Franco, C., Pinto, F., André, R.N., Tavares, C., Dias, M., Gulyurtlu, I., Cabrita, I., 2001,

“Experience Using INETI Experimental Facilities in Co-Gasification of Coal and Wastes,”

Proceedings of the Workshop Co-Processing of Different Waste Materials with Coal for Energy,

Lisbon, Portugal, pp. 238-247.

[76] Syamlal, M., Rogers, W., and O’Brien, T., 1993, “MFIX Documentation: Theory Guide,”

National Energy Technology Laboratory, Department of Energy, Technical Note DOE/METC-

95/1013 and NTIS/DE95000031.

100

[77] Xie, N., Battaglia, F., Pannala, S., 2008, “Effects of using two versus three-dimensional

computational modeling of fluidized beds: Part II, budget analysis,” Powder Technology, 182, pp.

14-24

[78] Deza, M., Battaglia, F., 2013, “A CFD Study of Pressure Fluctuations to Determine Fluidization

Regimes in Gas–Solid Beds,” Journal of Fluids Engineering, 135(10), 101301.

[79] Syamlal, M., 1987, "The Particle-Particle Drag Term in a Multiparticle Model of Fluidization,"

Topical Report, DOE/MC/21353-2373, NTIS/DE87006500, National Technical Information

Service, Springfield, VA.

[80] Gidaspow, D., 1994, “Multiphase Flow and Fluidization: Continuum and Kinetic Theory

Descriptions”, Academy, Boston, MA.

[81] Estejab, B., Battaglia, F., 2015, “Assessment of Drag Models for Geldart A Particles in Bubbling

Fluidized Beds,” Journal of Fluids Engineering, 138(3), pp. 031105-031105-12.

[82] Huilin, L., and Gidaspow, D., 2003, “Hydrodynamics of Binary Fluidization in a Riser: CFD

Simulation Using Two Granular Temperatures,” Chemical Engineering Science, 58(16), pp.

3777–3792.

[83] Lathouwers, D., and Bellan, J., 2000, “Modeling and Simulation of Bubbling Fluidized Beds

Containing Particle Mixtures,” Proceedings of the Combustion Institute, Edinburgh, United

Kingdom, 28(2), pp. 2297–2304.

[84] Syamlal, M., 1998, “MFIX Documentation: Numerical Technique,” National Energy Technology

Laboratory, Department of Energy, Technical Note DOE/MC31346-5824 and DE98002029.

[85] Syamlal, M., 1998, “High Order Discretization Methods for the Numerical Simulation of

Fluidized Beds,” Department of Energy, Technical Note DOE/FETC/C-98/7305 and CONF-

971113.

[86] Patankar, S. V., 1980, “Numerical Heat Transfer and Fluid Flow,” Hemisphere Publishing

corporation, New York.

[87] Nyendu,G.C., 2015, “Non-Catalytic Co-Gasification of Sub-Bituminous Coal and Biomass,”

Master thesis, Utah State University.

[88] Celik, I. B., Ghia, U., Roache, P. J., Freitas, C. J., Coleman, H., and Raad, P. E., 2008,

“Procedure for estimation and reporting of uncertainty due to discretization in CFD applications,"

Journal of Fluids Engineering, 130(7), 078001.

[89] Ergun, S., 1952, "Fluid Flow Through Packed Columns," Chemical Engineering Progress, 48(2),

pp. 89-94.

[90] Battaglia, F., Agblevor, F.A., Klein, M., and Sheikhi, R., 2015, “Investigation of Coal-biomass

Catalytic Gasification using Experiments, Reaction Kinetics and Computational Fluid

Dynamics”, Final Report, DE-FE0005476, National Energy Technology Laboratory, Pittsburgh,

PA.

[91] Lundberg, J., Halvorsen, B.M., “A Review of some existing drag models describing the

interaction between phases in a bubbling fluidized bed”,

101

[92] Loha, C., Chattopadhyay, H., Chatterjee, P., 2012, “Assessment of drag models in simulating

bubbling fluidized bed hydrodynamics”, Chem. Eng. Science 75, PP 400-407.

[93] Bayle, J., Mege, P., & Gauthier, T., 2001, “Dispersion of bubble flow properties in a turbulent

FCC fluidized bed”. In M. Kwauk, J. Li, & W.-C. Yang (Eds.), Fluidization X (pp. 125–132).

New York, USA: Engineering Foundation.

[94] Zimmermann, S. and F. Taghipour, 2005, “CFD Modeling of the Hydrodynamics and Reaction

Kinetics of FCC Fluidized Bed Reactors,” Ind Eng Chem J., vol.44, pp.9818–9827.

[95] Syamlal, M., and O’Brien, T.J., 1988, “Simulation of Granular Layer Inversion in Liquid

Fluidized Beds,” Int. J. Multiphase Flow, 14(4), pp. 473-481.

[96] Garside, J. and Al-Dibouni, M. R., 1977, “Velocity-voidage relationships for fluidization and

sedimentation,” Ind. Eng. Chem. Process Des. Dev., 16(2), pp. 206-214.

[97] Dalla Valle, J. M., 1948, Micromeritics: the technology of fine particles, Pitman, London.

[98] Hill, R.J., Koch, D.L., and Ladd, A.J.C., 2001, “The first effects of fluid inertia on flows in

ordered and random arrays of spheres,” J. Fluid Mech., 448, pp. 213-241.

[99] Hill, R.J., Koch, D.L., and Ladd, A.J.C., 2001, “Moderate-Reynolds-number flows in ordered and

random arrays of spheres,” J. Fluid Mech., 448, pp. 243-278.

[100] Benyahia, S., Syamlal, M., and O’Brien, T.J., 2006, “Extension of Hill–Koch–Ladd drag

correlation over all ranges of Reynolds number and solids volume fraction,” Powder Tech.,

162(2), pp. 166-174.

[101] Beetstra, R., Van Der Hoef, M.A., and Kuipers, J.A.M., 2007, “Drag Force of Intermediate

Reynolds Number Flow Past Mono- and Bidisperse Arrays of Spheres”. AIChE J., 53(2), pp.

489-501.

[102] Holloway, W., Yin, X., and Sundaresan, S., 2010, “Fluid-particle drag in inertial, polydisperse,

gas-solid suspensions”, AIChE J., 56(8), pp. 1995-2004.

[103] Yin, X., and Sundaresan, S., 2009, “Drag Law for Bidisperse Gas-Solid suspensions Containing

Equally Sized Spheres,” Ind. Eng. Chem. Res., 48(1), pp. 227-241.

[104] Woollard, I. N. M., and 0. E. Potter, 1968, “Solids Mixing in Fluidized Beds,” AIChE Journal

14(3), pp. 388-391.

[105] Nguyen, H.V., Whitehead, A.B., Potter, O.E., 1977, “Gas backmixing, solids movement, and

bubble activities in large scale fluidized beds,” AIChE Journal, 23(6), pp. 913-922.

[106] Norouzi, H.R., Mostoufi , N., Mansourpour, Z., Sotudeh-Gharebagh, R., Chaouki, J., 2011,

“Characterization of solids mixing patterns in bubbling fluidized beds,” Chemical Engineering

Research and Design, 89(6), pp. 817–826.

[107] Stein, M., Ding, Y.L., Seville, J.P.K., Parker, D.J., 2000, “Solids motion in bubbling gas fluidized

beds,” Chemical Engineering Science, 55, pp. 5291-5300.

[108] Mostoufi, N., Chaouki, J., 2001, “Local solid mixing in gas-solid fluidized beds,” Powder

Technology, 114, pp. 23-31.








http://www.sciencedirect.com/science/journal/02638762/89/6

102

[109] Glicksman, L., Carr, E., Noymen, P., 2008, “Particles injection and mixing experiments in a one-

quarter scale model bubbling fluidized bed,” Powder Technology, 180, pp 284-288.

[110] Rhodes, M.J., Wang, X.S., Nguyen, M., Stewart, P., Liffman, K., 2001, “Study of mixing in gas-

fluidized beds using a DEM model,” Chemical Engineering Science, 56(8), pp. 2859-2866.

[111] Liu, D.Y., Chen, X.P., Liang, C., Zhao, C.S., 2010, “Solids Mixing in the Bottom Zone of

Fluidized Beds”, Proceedings of the 20th International Conference on Fluidized Bed Combustion,

pp 459-463

[112] Rowe, P. N., Nienow, A. W., and Agbim, A. J., 1972, "A Preliminary Quantitative Study of

Particle Segregation in Gas Fluidised Beds - Binary Systems of Near Spherical Particles,"

Transactions of the Institution of Chemical Engineers, 50, pp. 324-333.

[113] Donsi, G., Formisani, B., Valentino, R., Vopicelli, G., 1984, “The measurement of characteristic

angle of powders in the prediction of their behavior in the gas fluidized state,” Powder


[114] Gilbertson, M.A., Yates, J.G., 1996, “The tilting fluidized bed: A re-examination,” Powder


[115] Valverde, J.M., Castellanos, A., Sanchez Quintanilla, M.A., 2001, “Self-diffusion in a gas-

fluidized bed of fine powder,” Phys Rev Lett, 86(14), pp. 3020-3023.

[116] Rietema, K., 1973, “The effect of interparticle forces on the expansion of a omogeneous gas-

fluidized bed,” Chemical Engineering Science, 28(7), pp. 1493–1497.

[117] Marzocchella, A., and Salatino, P., 2000, “Fluidization of solids with COH at pressures from

ambient to supercritical,” AIChE J., 46(5), pp. 901-910.

[118] Tsinontides, S.C., and Jackson, R., 1993, “The mechanics of gas fluidized beds with an interval

of stable fluidization,” J. Fluid Mech., 255, pp 237- 274.

[119] Loezos, P.N., Costamagna, P., Sundaresan, S., 2002, “The role of contact stresses and wall

friction on fluidization,” Chem. Eng. Sci., 57(24), pp. 5123–5141.

[120] Xu, B.H., Zhou, Y.C., Yu, A.B., Zulli, P., 2002, “Force structures in gas fluidized beds of fine

powders,” Proceedings of World Congress on Particle Technology 4 (CD-ROM), July 21–25,

Sydney, Australia, Paper no. 331.

[121] Simbeck, D., et al., “Coal Gasification Guidebook: Status, Applications, and Technologies,”

Report prepared for EPRI by SFA Pacific, Inc., TR-102034, Dec 1993.

[122] The Future of Coal - An MIT Interdisciplinary Study (Mar 2007)

[123] National Energy Technology Laboratory (http://www.netl.doe.gov)

[124] Gasification and syngas technologies council (GSTC) (http://www.gasification-syngas.org/)

[125] Office of fossil energy (energy.gov)

[126] Phelps, J., Blackford, J., Holt, J., Polton, J., 2015, "Modelling Large-Scale CO2 Leakages in the

North Sea", International Journal of Greenhouse Gas Control, 38, pp. 210–220.

http://link.springer.com/book/10.1007/978-3-642-02682-9



http://www.ncbi.nlm.nih.gov/pubmed/?term=Valverde%20JM%5BAuthor%5D&cauthor=true&cauthor_uid=11290097

http://www.ncbi.nlm.nih.gov/pubmed/?term=Castellanos%20A%5BAuthor%5D&cauthor=true&cauthor_uid=11290097

http://www.ncbi.nlm.nih.gov/pubmed/?term=Sanchez%20Quintanilla%20MA%5BAuthor%5D&cauthor=true&cauthor_uid=11290097

http://www.ncbi.nlm.nih.gov/pubmed/11290097

http://www.sciencedirect.com/science/article/pii/000925097385153X



http://journals.cambridge.org/action/displayJournal?jid=FLM







http://web.mit.edu/coal/The_Future_of_Coal.pdf

http://www.sciencedirect.com/science/journal/17505836/38/supp/C

103

[127] Van de Velden, M., Baeyens, J., Brems, A., Janssens, B., and Dewil, R., 2010, "Fundamentals,

kinetics and endothermicity of the biomass pyrolysis reaction," Renewable Energy, 35(1), pp.

232-242.

[128] Zinoviev, S., Müller-Langer, F., Das , P., Bertero , N., Fornasiero , P., Kaltschmitt , M., Centi ,

G., and Miertus, S., 2010, "Next-Generation Biofuels: Survey of Emerging Technologies and

Sustainability Issues," ChemSusChem, 3(10), pp. 1106-1133.

[129] Werther, J., Saenger, M., Hartge, E.-U., Ogada, T., Siagi, Z., 2000, “Combustion of agricultural

residues,” Progress in Energy and Combustion Science, 26, pp. 1–27

[130] Kumabe, K., Hanaoka, T., Fujimoto, S., Minowa, T., Sakanishi, K., 2007, “Co-gasification of

woody biomass and coal with air and steam,” Fuel 86, pp. 684–689.

[131] Long, H.A., and Wang, T., 2011, “Case studies for biomass/coal co gasification in IGCC

applications,“ Proceedings of ASME Turbo Expo, Vancouver, Canada.

[132] Van Krevelen, D. W., 1950, "Graphical-statistical method for the study of structure and reaction

processes of coal," Fuel (Guildford), 29(12), pp. 269.

[133] Kumar, A., Jones, D.D., and Hanna, M.A., 2009, “Thermochemical Biomass Gasification: A

Review of the Current Status of the Technology,” Energies, 2(3), pp. 556-581.

[134] Xu, Q., 2013, “Investigation of Co-Gasification Characteristics of Biomass and Coal in Fluidized

Bed Gasifiers,” PhD thesis, University of Canterbury.

[135] Everson, R. C., Neomagus, H. W. J. P., Kasaini, H., and Njapha, D. ,2006, "Reaction kinetics of

pulverized coal-chars derived from inertinite-rich coal discards: Gasification with carbon dioxide

and steam," Fuel, 85(7-8), pp. 1076-1082.

[136] Klose, W., and Wölki, M., 2005, "On the intrinsic reaction rate of biomass char gasification with

carbon dioxide and steam," Fuel, 84(7–8), pp. 885-892.

[137] Asadullah, M., Zhang, S., Min, Z., Yimsiri, P., and Li, C.-Z., 2010, "Effects of biomass char

structure on its gasification reactivity," Bioresource Technology, 101(20), pp. 7935-7943.

[138] Kastanaki, E., and Vamvuka, D., 2006, "A comparative reactivity and kinetic study on the

combustion of coal–biomass char blends," Fuel, 85(9), pp. 1186-1193.

[139] Lu, L., Kong, C., Sahajwalla, V., and Harris, D., 2002, "Char structural ordering during pyrolysis

and combustion and its influence on char reactivity," Fuel, 81, pp. 1215-1225

[140] Zhu, W., Song, W., and Lin, W., 2008, "Catalytic gasification of char from co-pyrolysis of coal

and biomass," Fuel Processing Technology, 89, pp. 890-896.

[141] Syamlal, M., Bissett, L. A., 1992, “METC Gasifier Advanced Simulation (MGAS) Model,” U.S.

Department of Energy, Office of Fossil Energy, Morgantown Energy Technology Center.

[142] Wen, C.Y., Chen, H., Onozaki, M., 1982, “USER'S MANUAL FOR COMPUTER

SIMULATION AND DESIGN OF THE MOVING BED COAL GASIFIER,” Department of

Chemical Engineering, West Virginia University, Morgantown, W. Va. 26505

[143] Pannala, S., 2010, “Introduction to CFD Modeling of Fluidized Beds,” Coal Gasification Short

Course, Criciúma, Santa Catarina, Brazil

104

[144] Higman, C., and van der Burgt, M., 2003, “Gasification,” Elsevier: Gulf Professional Publishing.

[145] Ratafia-Brown, J., Manfredo, L., Hoffmann, J., and Ramezan, M., 2002, “Major Environmental

Aspects of Gasification-Based Power Generation Technologies,” by U.S. Department of Energy,

Office of Fossil Energy, National Energy Technology Laboratory.

[146] Wen, C.Y., Chen, H., and Onozaki, M., 1982, "User's Manual for Computer Simulation and

Design of the Moving Bed Coal Gasifier," DOE/MC/16474-1390, NTIS/DE83009533, National

Technical Information Service, Springfield, VA.

[147] Yoon, H., Wei, J., and Denn, M.M., 1978, "A Model for Moving-Bed Coal Gasification

Reactors," AIChE J., 24, pp. 885-903.

[148] Dupont, C., Chiriac, R., Gauthier, G., Toche, F., 2014, “Heat capacity measurements of various

biomass types and pyrolysis residues,” Fuel, 115, pp. 2014.

Appendix A

In order to verify the correctness for modeling reactions, a partial combustor was modeled and the

results are compared with Pannala [143]. The geometry of the combustor is presented in Figure A . 1. A

central tube with a 2 cm diameter and 12.5 cm height is used to feed char (8.8 g/s) and air (20.2 g/s) into

the combustor, while nitrogen (4 g/s) is fed into the reactor through the annular region of combustor. The

top of the reactor has an opening of 3 cm to the atmosphere, where the product gases and unconverted

solids exit.

The gas phase includes four species: O2, CO, CO2, N2 and the solid phase consists of coal with a

density of 1 g/cm3 and particle diameter of 250 µm that is modeled in two phases: hot and cold char. The

cold char is the incoming char at a lower temperature of 300ºC and the hot char is transformed from cold

char after reaching a certain temperature. Each solids phase consists of two species: fixed carbon and ash.

The fluidized bed reactor is modeled using a 2D geometry that represents one-half of the reactor center

plane. The chemical reactions modeled are:

2C + O2 2CO

CO + ½ O2 2CO2

C + CO2 2CO

Cold char Hot char

Void fraction, gas temperature, solids temperature, CO and CO2 mass fractions are compared to

Pannala in Figure A . 2(a - f), respectively. Comparison between results of the current study and Pannala

shows good agreement that validates the modeling techniques.

106

Figure A . 1. Combustor geometry.

107

(a)

(b)

(c)

(d)

(e)

(f)

Figure A . 2. (a) Void fraction, (b) gas temperature, (c) cold char temperature, (d) hot char temperature, (e)

CO mass fraction, and (f) CO2 mass fraction contours along with Pannala results [143] comparing

simulations for coal gasification.

1

0 .9 5

0 .9

0 .8 5

0 .8

0 .7 5

0 .7

0 .6 5

0 .6

0 .5 5

0 .5

0 .4 5

0 .4

g

1 0 5 0

9 8 7 .5

9 2 5

8 6 2 .5

8 0 0

7 3 7 .5

6 7 5

6 1 2 .5

5 5 0

4 8 7 .5

4 2 5

3 6 2 .5

3 0 0

Tg

1 1 6 0

1 0 7 4

9 8 8

9 0 2

8 1 6

7 3 0

6 4 4

5 5 8

4 7 2

3 8 6

3 0 0

Ts

1 2 5 0

1 1 7 0 .8 3

1 0 9 1 .6 7

1 0 1 2 .5

9 3 3 .3 3 3

8 5 4 .1 6 7

7 7 5

6 9 5 .8 3 3

6 1 6 .6 6 7

5 3 7 .5

4 5 8 .3 3 3

3 7 9 .1 6 7

3 0 0

Ts

0 .1 8 1

0 .1 6 2 9

0 .1 4 4 8

0 .1 2 6 7

0 .1 0 8 6

0 .0 9 0 5

0 .0 7 2 4

0 .0 5 4 3

0 .0 3 6 2

0 .0 1 8 1

XC O

0 .2 5 2

0 .2 3 1

0 .2 1

0 .1 8 9

0 .1 6 8

0 .1 4 7

0 .1 2 6

0 .1 0 5

0 .0 8 4

0 .0 6 3

0 .0 4 2

0 .0 2 1

XC O 2

108

Appendix B

The main reason for the different reaction rates of coal and switchgrass modeling CO2 gasification is

their different thermodynamics properties that results in different solid temperature in the bed presented in

Figure B . 1. Based on the Arrhenius reaction rate formula, lower switchgrass temperatures are the reason

for lower reaction rates.

(a) coal (b) switchgrass

Figure B . 1. Time-averaged solid temperature contours for (a) coal and (b) switchgrass.

r (c m )

2 .5

5

1 0

1 5

r (c m )

2 .5 5

5

1 0

1 57 5 0

7 0 0

6 5 0

6 0 0

5 5 0

5 0 0

4 5 0

4 0 0

3 5 0

3 0 0

2 5 0

2 0 0

1 5 0

Ts

(C )

Hydrodynamic and gasification behavior of coal and biomass ... · different gasification agents....

Documents

Transcript of Hydrodynamic and gasification behavior of coal and biomass ... · different gasification agents....