open.metu.edu.tr · Approval of the thesis: CO-FIRING BIOMASS WITH COAL IN BUBBLING FLUIDIZED BED...

CO-FIRING BIOMASS WITH COAL IN BUBBLING FLUIDIZED BED COMBUSTORS

A THESIS SUBMITTED TO THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES

OF MIDDLE EAST TECHNICAL UNIVERSITY

BY

ZUHAL GÖĞEBAKAN

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR

THE DEGREE OF DOCTOR OF PHILOSOPHY IN

CHEMICAL ENGINEERING

SEPTEMBER 2007

Approval of the thesis:

CO-FIRING BIOMASS WITH COAL IN BUBBLING FLUIDIZED BED COMBUSTORS

submitted by ZUHAL GÖĞEBAKAN in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Chemical Engineering Department, Middle East Technical University by, Prof. Dr. Canan Özgen __________ Dean, Graduate School of Natural and Applied Sciences Prof. Dr. Nurcan Baç __________ Head of Department, Chemical Engineering Prof. Dr. Nevin Selçuk Supervisor, Chemical Engineering Dept., METU __________ Prof. Dr. Ekrem Selçuk Co-Supervisor, Metallurgical & Materials Engineering Dept., METU__________ Examining Committee Members: Prof. Dr. Yavuz Samim Ünlüsoy Mechanical Engineering Dept., METU _______________ Prof. Dr. Nevin Selçuk Chemical Engineering Dept., METU _______________ Assist. Prof. Dr. Murat Köksal Mechanical Engineering Dept., Hacettepe University _______________ Prof. Dr. Nurcan Baç Chemical Engineering Dept., METU _______________ Assist. Prof. Dr. Görkem Külah Chemical Engineering Dept., METU _______________

Date: 06.09.2007

I hereby declare that all information in this document has been obtained and

presented in accordance with academic rules and ethical conduct. I also declare

that, as required by these rules and conduct, I have fully cited and referenced

all material and results that are not original to this work.

Name, Last Name : Zuhal Göğebakan

Signature :

iii

ABSTRACT

CO-FIRING BIOMASS WITH COAL

IN BUBBLING FLUIDIZED BED COMBUSTORS

Göğebakan, Zuhal

Ph.D., Department of Chemical Engineering

Supervisor: Prof. Dr. Nevin Selçuk

Co-Supervisor: Prof. Dr. Ekrem Selçuk

September 2007, 275 pages

Co-firing of biomass with coal in fluidized bed combustors is a promising alternative

which leads to environmentally friendly use of coal by reducing emissions and

provides utilization of biomass residues. Therefore, effect of biomass share on

gaseous pollutant emissions from fluidized bed co-firing of various biomass fuels

with high calorific value coals have extensively been investigated to date. However,

effect of co-firing of olive residue, hazelnut shell and cotton residue with low

calorific value lignites having high ash and sulfur contents has not been studied in

bubbling fluidized bed combustors to date.

In this thesis study, co-firing of typical Turkish lignite with olive residue, hazelnut

shell and cotton residue in 0.3 MWt METU Atmospheric Bubbling Fluidized Bed

Combustion (ABFBC) Test Rig was investigated in terms of combustion and

emission performance and ash behavior of different fuel blends.

iv

The results reveal that co-firing of olive residue, hazelnut shell and cotton residue

with lignite increases the combustion efficiency and freeboard temperatures

compared to those of lignite firing with limestone addition only. O2 and CO2

emissions are not found sensitive to increase in olive residue, hazelnut shell and

cotton residue share in fuel blend. Co-firing lowers SO2 emissions considerably

while increasing CO emissions. Co-firing of olive residue and hazelnut shell has no

significant influence on NO emissions, however, reduces N2O emissions. Co-firing

cotton residue results in higher NO and N2O emissions. Regarding to major, minor

and trace elements partitioning, co-firing lignite with biomasses under consideration

shifts the partitioning of these elements from bottom ash to fly ash. No chlorine is

detected in both EDX and XRD analyses of the ash deposits. In conclusion, olive

residue, hazelnut shell and cotton residue can easily be co-fired with high ash and

sulfur containing lignite without agglomeration and fouling problems.

Keywords: Co-firing, biomass, lignite, bubbling fluidized bed combustor.

v

ÖZ

BİYOKÜTLE İLE KÖMÜRÜN KABARCIKLI AKIŞKAN YATAKLI

YAKICILARDA BİRLİKTE YAKILMASI

Göğebakan, Zuhal

Doktora, Kimya Mühendisliği Bölümü

Tez Yöneticisi: Prof. Dr. Nevin Selçuk

Ortak Tez Yöneticisi: Prof. Dr. Ekrem Selçuk

Eylül 2007, 275 sayfa

Biyokütle ve kömürün akışkan yataklı yakıcılarda birlikte yakılması, emisyonları

düşürerek kömürün çevre dostu olarak kullanılmasını ve biyokütle artıklarının

değerlendirilmesini sağlayan umut verici bir alternatiftir. Bu nedenle, biyokütle

miktarının kirletici gaz emisyonlarına etkisi, çeşitli biyokütleler ve yüksek kalorili

kömürleri birlikte yakan akışkan yataklı yakıcılarda günümüze kadar yaygın bir

şekilde incelenmiştir. Ancak, zeytin artığı, fındık kabuğu ve pamuk artığının düşük

kalorili, yüksek kükürt ve kül içeren linyitle kabarcıklı akışkan yataklı yakıcılarda

birlikte yakılmasının etkisi üzerine bir çalışma bulunmamaktadır.

Bu tez çalışmasında, tipik Türk linyiti ile zeytin artığı (prina), fındık kabuğu ve

pamuk artığının birlikte yakılması ODTÜ 0.3 MW ısıl gücündeki kabarcıklı akışkan

yataklı yakıcıda farklı yakıt karışımlarının yanma ve emisyon performanslarına ve

kül davranışına etkisi açısından incelenmiştir.

vi

Elde edilen sonuçlar, birlikte yanmanın yanma verimini ve serbest bölge

sıcaklıklarını kireçtaşı ilaveli linyit yakılmasına kıyasla artırdığını göstermiştir. O2 ve

CO2 emisyonlarının karışımdaki artan zeytin artığı, fındık kabuğu ve pamuk artığı

miktarına duyarlı olmadığı görülmüştür. Birlikte yanma SO2 emisyonlarını azaltırken

CO emisyonlarını artırmıştır. Zeytin artığı ve fındık kabuğu ile linyitin birlikte

yakılması NO emisyonlarında önemli değişikliğe yol açmazken, N2O emisyonlarını

azaltmıştır. Pamuk artığının linyit ile birlikte yakılması NO ve N2O emisyonlarını

artırmıştır. Majör, minör ve iz element dağılımlarına ilişkin olarak, linyit ve söz

konusu biyokütlelerin birlikte yakılması bu elementlerin dağılımlarını alt külden

uçucu küle değiştirmiştir. Hem EDX hemde XRD analizlerinde külde klor

bulunmadığı tesbit edilmiştir. Sonuç olarak, zeytin artığı, fındık kabuğu ve pamuk

artığı yüksek kül ve kükürt içeren linyit ile bitişme ve bırakıntı problemleri

olmaksızın kolayca birlikte yakılabilir.

Anahtar Kelimeler : Birlikte yanma, biyokütle, linyit, kabarcıklı akışkan yataklı

yakıcı

vii

To Yusuf

viii

ACKNOWLEDGMENTS

I wish to express my deepest gratitude to my supervisor, Prof. Dr. Nevin

Selçuk for her guidance, advice and encouragement throughout the development of

this study. I would like to show my appreciation to my co-supervisor Prof. Dr.

Ekrem Selçuk, for his valuable suggestions and support.

I am also grateful to Prof. Dr. Yavuz Samim Ünlüsoy and Asst. Prof. Dr.

Görkem Külah for their precious discussions and comments during the progress of

this work.

I also thank to the AFBC research team members for their support during the

experiments; Dr. Olcay Oymak, Dr. Yusuf Göğebakan, Asst. Prof. Dr. Görkem

Külah, Ümit Kırmızıgül, Mehmet Moralı, Aykan Batu, Dr. Ahmet Bilge Uygur,

Zeynep Serinyel, Güzide Aydın and İlker Soner.

I am very grateful for the technical assistance provided by Kerime Güney for

chemical analyses and Cengiz Tan for SEM and EDX analyses.

Financial supports provided by The Scientific and Technical Research

Council of Turkey (TUBİTAK) through MAG-104M200 project and by Middle East

Technical University (METU) through BAP-2006-07-02-00-01 project are gratefully

acknowledged.

I also want to thank to Cem Özdemiroğlu from Alfer Engineering Co. Ltd. for

construction and installation of the stack of the test rig.

I would like to thank to my family Nazife & Dr. Mehmet Coşkun, Gülizar &

İsmail Göğebakan, Dr. Nezihe & Derya Baltalı for their support and encouragement.

My special thanks go to my mother, Nazife Coşkun for her great kindness,

endless understanding and unshakable faith in me.

Finally, my warmest thanks go to my husband, Dr. Yusuf Göğebakan for his

understanding, endless support and encouragement throughout this study.

ix

TABLE OF CONTENTS

ABSTRACT………………………………………………………………....….. iv

ÖZ……………………………………………………………………..………... vi

DEDICATION…………………………………………………...……………... viii

ACKNOWLEDGEMENTS………………………………………………...…... ix

TABLE OF CONTENTS……………………………………………...………... x

LIST OF TABLES…………………………………………………...…………. xiv

LIST OF FIGURES………………………………………………………..…… xix

CHAPTER

1 INTRODUCTION……………………………………...………………... 1

1.1 General………………………………………………...……………. 1

1.2 Aim and Scope of the Thesis……………………………………….. 4

2 BACKGROUND AND LITERATURE REVIEW…………………….... 6

2.1 General……………………………………………………………… 6

2.2 Biomass Firing in FBC Systems……………………………………. 6

2.2.1 Biomass Characteristics…………………………………....…. 6

2.2.2 Biomass Combustion…………………………………………. 8

2.2.3 Operational Problems Associated with Biomass Ash………… 9

2.2.3.1 Bed Agglomeration…………………..………………… 11

2.2.3.2 Ash Deposition and Corrosion………………………… 13

2.3 Literature Review…………...……………………………………… 17

2.3 2.3.1 Biomass Firing Studies……………………………………….. 17

2.3.2 Biomass and Coal Co-firing Studies………………………….. 45

x

3 EXPERIMENTAL SET-UP, PROCEDURE AND CONDITIONS…….. 63

3.1 General……………………………………………………………… 63

3.2 METU 0.3 MWt ABFBC Test Rig…………………………………. 63

3.2.1 The Combustor………………………………………….……. 66

3.2.2 Air and Flue Gas System………………………………….….. 67

3.2.3 Solids Handling System…………………………………….… 68

3.2.4 Cooling Water System………………………………….…….. 70

3.2.5 Gas Sampling and Analysis System…………………….……. 70

3.2.5.1 Gas Sampling Probe...…………………………….….. 70

3.2.5.2 Gas Sampling and Conditioning System………….….. 73

3.2.5.3 Analytical System………………………………….…. 74

3.2.6 Deposit Sampling System………………………………….…. 77

3.2.7 Instrumentation and Control System……………………….… 77

3.3 Pre-Experimental Modifications……………………………………. 82

3.3.1 Feeding System……………………………………….………. 82

3.3.2 Air and Flue Gas System……………………………….…….. 83

3.3.3 Gas Sampling and Analysis System………………….………. 84

3.3.4 Instrumentation and Control System…………………….…… 84

3.4 Operating Procedures……………………………………………….. 85

3.4.1 Procedures before Cold Start-Up………………………….….. 85

3.4.2 Cold Start-Up………………………………………….……… 85

3.4.3 Procedure during Runs………………………………….……. 87

3.4.4 Shutdown……………………………………………….…….. 87

3.4.5 Post Shutdown………………………………………….…….. 88

3.5 Experimental Conditions…………………………………………… 88

3.5.1 Lignite, Biomass and Limestone Characteristics………….….. 88

xi

3.5.2 Operating Conditions…………………….…………………… 98

4 RESULTS AND DISCUSSION…………………………………………. 101

4.1 General……………………………………………………………… 101

4.2 Particle Size Distributions………………………………………….. 101

4.3 Ash Balance, Split and Discharge Compositions…………………... 107

4.3.1 Ash Balance and Ash Split…………...………………….…… 107

4.3.2 Ash Compositions………………….………………….……… 107

4.3.2.1 Major and Minor Elements……………………………. 107

4.3.2.2 Trace Elements…………………………….…………… 112

4.4 Partitioning of Major, Minor and Trace Elements………………….. 116

4.5 Combustion Efficiencies……………………………………………. 123

4.6 Temperature Profiles……………………………………………….. 126

4.7 Concentration Profiles……………………………………………… 131

4.7.1 O2, CO2 and CO Concentration Profiles……………….……... 131

4.7.2 SO2 Concentration Profiles……………………………….…... 136

4.7.3 NO and N2O Concentration Profiles…………………….……. 136

4.8 Emissions…………………………………………………………… 141

4.9 Agglomeration and Deposit Formation…………………………….. 147

5 CONCLUSIONS…………………………………………………………. 164

5.1 General……………………………………………………………… 164

5.2 Suggestions for Future Work……………………………………….. 166

REFERENCES…………………………………………………………………. 167

APPENDICES

A WORLWIDE CO-FIRING POWER PLANTS………………………….. 183

B TGA RESULTS………………………………………………………….. 196

C POINT VALUES OF MEASUREMENTS……………………………… 199

xii

D CHEMICAL ANALYSES OF ASH STREAMS………………………... 207

E TABULATED SIZE DISTRIBUTIONS………………………………… 211

F SIZE DISTRIBUTION GRAPHS……………………………………….. 237

G CALIBRATION CURVES OF SCREW FEEDERS……………………. 261

H DEPOSIT COMPOSITIONS…………………………………………….. 264

I SCREW FEEDER DESIGN CALCULATIONS………………………... 267

CURRICULUM VITAE………………………………………………………... 273

xiii

LIST OF TABLES

Table 1.1 World electricity generation in 2004 [5]…………………………. 2

Table 2.1 Proximate and ultimate analyses of some selected biomass and coal.………………………………………………………………. 7

Table 2.2 Influence of biomass ash on boiler performance [39]…………….. 10

Table 2.3 Operating conditions and system properties of rice husk firing FBCs……………………………………………………………… 18

Table 2.4 Operating conditions and system properties of straw firing FBCs.. 27

Table 2.5 Operating conditions and system properties of olive residue firing FBCs………………………………………………………... 34

Table 2.6 Operating conditions and system properties of forestry residue firing FBCs………………………………………………………... 39

Table 2.7 Operating conditions and system properties of straw and coal co-firing FBCs…………………………………………………….. 47

Table 2.8 Operating conditions and system properties of olive residue and coal co-firing FBCs……………………………………………….. 52

Table 2.9 Operating conditions and system properties of forestry residue and coal co-firing FBCs…………….…………………………….. 57

Table 3.1 Technical specifications of the baghouse filter…………………… 69

Table 3.2 Relative positions of the gas sampling probes...………………….. 71

Table 3.3 Gas analyzers……………………………………………………... 75

Table 3.4 Backup gas analyzers……..………………………………………. 75

Table 3.5 Relative positions of the thermocouples………………………….. 81

xiv

Table 3.6 Technical specifications of new air compressor………………….. 83

Table 3.7 Fuel analyses……………………………………………………… 90

Table 3.8 Fuel size distributions…………………………………………….. 91

Table 3.9 Fuel ash compositions…………………………………………….. 92

Table 3.10 Chlorine contents of the fuels, %..................................................... 93

Table 3.11 Trace element concentrations in lignites in Runs 2, 5, 8 and 10…. 94

Table 3.12 Trace element concentrations in biomass...………………………. 95

Table 3.13 Characteristics of Beypazarı limestone…………………………… 96

Table 3.14 Trace element concentrations in Beypazarı limestone……………. 97

Table 3.15 Durations of the Runs………………..…………………………… 98

Table 3.16 Operating conditions of Runs 1-10……………………………….. 99

Table 4.1 Ash balance, closure and split…………………………………….. 108

Table 4.2 Trace element concentrations in bottom ash of Runs 2, 5 and 8…. 113

Table 4.3 Trace element concentrations in cyclone ash of Runs 2, 5 and 8.... 114

Table 4.4 Trace element concentrations in filter ash of Runs 2, 5 and 8…..... 115

Table 4.5 Relative enrichment factors of bottom, cyclone and filter ashes…. 117

Table 4.6 Combustion efficiencies…………………………………………... 124

Table 4.7 Flue gas emission data……………………………………………. 142

Table 4.8 Fuel-N to NO conversion…………………………………………. 145

Table 4.9 Emission limits……………………………………………………. 146

Table 4.10 Bed agglomeration index of the fuel blends…………………..….. 148

Table 4.11 Melting temperatures of ternary systems…………………………. 149

Table 4.12 Alkali index of fuel blends………………………………………... 150

Table A.1 Worldwide samples of co-firing experienced pulverized fuel

power plant [13]…..………………………………………………. 184

Table A.2 Worldwide samples of co-firing experienced bubbling fluidized

xv

bed power plants [13].....………..………………………………… 192

Table A.3 Worldwide samples of co-firing experienced circulating fluidized

bed power plants [13].....………..………………………………… 193

Table C.1 Sampling probe readings of O2 concentrations of Runs 1-3, dry mole %....................................................................................... 199

Table C.2 Sampling probe readings of O2 concentrations of Runs 4-6, dry mole %....................................................................................... 199

Table C.3 Sampling probe readings of CO2 concentrations of Runs 1-6, dry mole %....................................................................................... 200

Table C.4 Sampling probe readings of CO concentrations of Runs 1-6, dry mole %....................................................................................... 201

Table C.5 Sampling probe readings of SO2 concentrations of Runs 1-6, dry mole %....................................................................................... 202

Table C.6 Sampling probe readings of NO concentrations of Runs 1-6, dry mole %....................................................................................... 203

Table C.7 Sampling probe readings of N2O concentrations of Runs 1-6, dry mole %....................................................................................... 204

Table C.8 Thermocouple readings of Runs 1-10, °C………………………… 205

Table C.9 Normalized temperatures of Runs 1-10 °C/°C…..………………... 206

Table D.1 Chemical analyses of bottom ashes of Runs 1-4.............................. 207

Table D.2 Chemical analyses of bottom ashes of Runs 5-7.............................. 208

Table D.3 Chemical analyses of bottom ashes of Runs 8-10............................ 208

Table D.4 Chemical analyses of cyclone ashes of Runs 1-10.......................... 209

Table D.5 Chemical analyses of baghouse filter ashes of Runs 1-10............... 210

Table E.1 Particle size distribution of bottom ash of Run 1…………………. 211


xvi








Table E.10 Particle size distribution of bottom ash of Run 10………………. 216

Table E.11 Particle size distribution of cyclone ash of Run 1...………………. 217









Table E.20 Particle size distribution of cyclone ash of Run 10……………….. 226

Table E.21 Particle size distribution of baghouse filter ash of Run 1...………. 227








xvii


Table E.30 Particle size distribution of baghouse filter ash of Run 10.………. 236

Table H.1 Deposit compositions……………………………………………... 264

Table I.1 Comparison of measured and calculated hazelnut shell flow rate... 268

Table I.2 Comparison of measured and calculated olive residue flow rate..... 269

xviii

LIST OF FIGURES

Figure 1.1 Share of total primary energy supply in world in year 2004 [5]…. 2

Figure 1.2 World electricity generation by fuel, (2004-2030) [7]…………… 3

Figure 2.1 Biomass ash compositions……………………………………….. 10

Figure 2.2 Agglomeration of bed material: Type 1- coating-induced, Type 2- melt-induced [40]……………………………………….. 12

Figure 2.3 Corrosion effects of alkalis in biomass combustion and co-combustion systems [48]……………………………………… 16

Figure 3.1 METU 0.3 MWt ABFBC Test Rig………………………………. 64

Figure 3.2 Flow sheet of METU 0.3 MWt ABFBC Test Rig………………... 65

Figure 3.3 Gas sampling probe………………………………………………. 72

Figure 3.4 GASS-II pre-conditioning system………………………..………. 73

Figure 3.5 Schematic of Perma Pure Nafion Membrane Dryer..……………. 74

Figure 3.6 Gas conditioning and analysis system……………………………. 76

Figure 3.7 Deposit sampling probe…………………………………………... 78

Figure 3.8 P & I diagram of METU 0.3 MWt ABFBC Test Rig……………. 79

Figure 3.9 Photographs of biomasses.……………………………………….. 89

Figure 4.1 Size distribution of all solid streams in Runs 1 and 2……………. 102

Figure 4.2 Size distribution of all solid streams in Runs 2, 3, 4 and 5………. 104

Figure 4.3 Size distribution of all solid streams in Runs 2, 6, 7 and 8………. 105

Figure 4.4 Size distribution of all solid streams in Runs 2, 9 and 10..………. 106

Figure 4.5 Bottom ash analyses of Runs 1-10……………………………….. 109

Figure 4.6 Cyclone ash analyses of Runs 1-10……...……………………….. 110

xix

Figure 4.7 Baghouse filter ash analyses of Runs 1-10…...………………….. 111

Figure 4.8 Recovery of major and minor elements in Runs 2, 5 and 8……… 119

Figure 4.9 Recovery of trace elements in Runs 2, 5 and 8…………...……… 120

Figure 4.10 Major and minor elements partitioning of Runs 2, 5 and 8……… 121

Figure 4.11 Trace elements partitioning of Runs 2, 5 and 8……………..…… 122

Figure 4.12 Temperature profiles of Runs 1 and 2……………………………. 127

Figure 4.13 Normalized temperature profiles of Runs 1 and 2.………………. 127

Figure 4.14 Temperature profiles of Runs 2, 3, 4 and 5………………………. 128

Figure 4.15 Normalized temperature profiles of Runs 2, 3, 4 and 5………….. 128

Figure 4.16 Temperature profiles of Runs 2, 6, 7 and 8………………………. 129

Figure 4.17 Normalized temperature profiles of Runs 2, 6, 7 and 8………….. 129

Figure 4.18 Temperature profiles of Runs 2, 9 and 10..………………………. 130

Figure 4.19 Normalized temperature profiles of Runs 2, 9 and 10..………….. 131

Figure 4.20 O2 concentration profiles of Runs 1-6……………………………. 132

Figure 4.21 CO2 concentration profiles of Runs 1-6……….…………………. 133

Figure 4.22 CO concentration profiles of Runs 1-6………..…………………. 135

Figure 4.23 SO2 concentration profiles of Runs 1-6.…………………………. 137

Figure 4.24 NO concentration profiles of Runs 1-6..…………………………. 139

Figure 4.25 N2O concentration profiles of Runs 1-6………………………….. 140

Figure 4.26 Appearance of deposit rings after olive residue/lignite co-firing runs……………………………………………………... 152

Figure 4.27 Appearance of deposit rings after hazelnut shell/lignite co-firing runs……………………………………………………... 153

Figure 4.28 Appearance of deposit rings after cotton residue/lignite co-firing runs……………………………………………………... 154

Figure 4.29 Rate of deposit build-up (RBU, g/m2h)…………………………... 155

xx

Figure 4.30 SEM micrograph of deposit of olive residue/lignite co-firing runs (a)…………………………………………………. 156

Figure 4.31 SEM micrograph of deposit of olive residue/lignite co-firing runs (b)…………………………………………………. 156

Figure 4.32 SEM micrograph of deposit of olive residue/lignite co-firing runs (c)…………………………………………………. 157

Figure 4.33 SEM micrograph of deposit of hazelnut shell/lignite co-firing runs (a)…………………………………………………. 157

Figure 4.34 SEM micrograph of deposit of hazelnut shell/lignite co-firing runs (b)…………………………………………………. 158

Figure 4.35 SEM micrograph of deposit of hazelnut shell/lignite co-firing runs (c)…………………………………………………. 158

Figure 4.36 SEM micrograph of deposit of cotton residue/lignite co-firing runs (a)…………………………………………………. 159

Figure 4.37 SEM micrograph of deposit of cotton residue/lignite co-firing runs (b)…………………………………………………. 159

Figure 4.38 SEM micrograph of deposit of cotton residue/lignite co-firing runs (c)…………………………………………………. 160

Figure 4.39 Deposit compositions…………………………………………….. 161

Figure 4.40 X-ray diffraction patterns of the deposits………………………… 162

Figure B.1 TGA graph of lignite……………………………………………... 196

Figure B.2 TGA graph of olive residue………………………………………. 197

Figure B.3 TGA graph of hazelnut shell..……………………………………. 197

Figure B.4 TGA graph of cotton residue…………………………..…………. 198

Figure F.1 Particle size distribution of lignite fed in Run 1…………………. 237


xxi








Figure F.10 Particle size distribution of lignite fed in Run 10..………………. 242

Figure F.11 Particle size distribution of olive residue fed in Runs 3, 4 and 5... 242

Figure F.12 Particle size distribution of hazelnut shell fed in Runs 3, 4 and 5.. 243

Figure F.13 Particle size distribution of cotton residue fed in Runs 3, 4 and 5.. 243

Figure F.14 Particle size distribution of limestone fed in Runs 2, 3, 4 and 5…. 244

Figure F.15 Particle size distribution of limestone fed in Runs 6, 7 and 8……. 244

Figure F.16 Particle size distribution of limestone fed in Runs 9 and 10..……. 245

Figure F.17 Particle size distribution of bottom ash of Run 1………………… 246









Figure F.26 Particle size distribution of bottom ash of Run 10..……………… 250

Figure F.27 Particle size distribution of cyclone ash of Run 1...……………… 251


xxii








Figure F.36 Particle size distribution of cyclone ash of Run 10.……………… 255

Figure F.37 Particle size distribution of baghouse filter ash of Run 1………... 256









Figure F.46 Particle size distribution of baghouse filter ash of Run 10..……... 260

Figure G.1 Calibration curve for lignite flow rate……………………………. 261

Figure G.2 Calibration curve for olive residue flow rate…………………….. 262

Figure G.3 Calibration curve for hazelnut shell flow rate.…………………… 262

Figure G.4 Calibration curve for cotton residue flow rate…………………… 263

Figure G.5 Calibration curve for limestone flow rate………………………... 263

Figure H.1 EDX analysis graph of cotton residue and lignite co-firing

deposit……………………………………………………………. 265


xxiii

deposit……………………………………………………………. 265


deposit……………………………………………………………. 266

Figure I.1 Schematic description of ash feeder spiral……………………….. 267

Figure I.2 Comparison of measured and calculated hazelnut shell flow rate.. 269

Figure I.3 Comparison of measured and calculated olive residue flow rate… 270

Figure I.4 Schematic description of new bed feeder spiral………………….. 271

Figure I.5 Schematic drawing of the new bed feeder……………………….. 272

xxiv

CHAPTER 1

INTRODUCTION

1.1 General

Today, the demand for electric power continues to rise due to population growth,

technological and economical development. Coal is predicted to be the dominant

fossil fuel for energy production for decades with progressively improving clean coal

technologies and 909 billion tones of proved reserves [1]. Global environmental

impacts of fossil fuels used for power generation is a worldwide concerning topic. To

mitigate effects of fossil fuel combustion, share of renewables in energy production

is increasing. Among the renewable sources, biomass is the lowest risk and capital

required one to be used in energy generation. It is a renewable energy option due to

the fact that it can be considered as CO2-neutral fuel and contributes to the reduction

of SO2 and NOx emissions due to its low sulfur and nitrogen contents. Furthermore,

when burned instead of landfilled, it prevents CH4 release to atmosphere, which is a

more powerful greenhouse gas compared to CO2 [2].

In addition to the above mentioned advantages, biomass is the fourth largest energy

source after oil, coal and gas and that about 10 to 14 % of world’s energy is produced

from biomass. In Figure 1.1, the share of world’s total primary energy supply in year

2004 is shown. Africa obtains about 2/3 of its energy from biomass, Asia about 1/3,

and Latin America about 1/4 [3]. The role of biomass in European countries varies

from nearly 15 % in Finland and Sweden to less than 1 % in the UK. In longer term,

it is anticipated that biomass could contribute to 20 % of the current European Union

primary energy demand [4].

1

Figure 1.1: Share of total primary energy supply in world in year 2004 [5].

Biomass is a very versatile source of energy that can be readily stored and

transformed into electricity and heat [6]. Table 1.1 gives worldwide amounts of

electricity produced from different energy sources in year 2004. As can be seen from

the table, coal is the dominant source of electricity generation and renewable sources

have lower share. However, the share of biomass is expected to rise in the following

years due to increasingly strict legislations on emissions from fossil fuel sources. The

projections on electricity generation till year 2030 given in Figure 1.2 also show the

increase in electricity production from coal and renewable sources.

Table 1.1: World electricity generation in 2004 [5].

Fuel TWh % Coal 6944 39.7 Gas 3419 19.6 Oil 1170 6.7 Hydro 2889 16.5 Nuclear 2738 15.7 Biomass 149.8 0.9 Waste 77.5 0.4 Others 85.1 0.5 Total 17472 100.0

2

Figure 1.2: World electricity generation by fuel, (2004-2030) [7].

However, biomass combustion brings with it some operational problems when

burned alone. The most common problems encountered in industry and utility boilers

are severe fouling, slagging and corrosion which are mainly originated from high

alkali chlorides content of biomass ash. Ash deposits formed on heat transfer

surfaces deteriorate heat transfer and lead to loss in thermal efficiency and corrosion.

These problems in biomass firing combustion systems can be alleviated by either

leaching biomass with chemicals of various strengths to reduce the harmful

components of ash or co-firing biomass with coal. The former alternative was not

found economically feasible. However, the latter represents a near-term, low risk,

low cost, sustainable renewable energy option [8-11]. This has already been

demonstrated in over 100 coal-fired power plants. The capacity of typical power

stations where co-firing is applied varies in the range from 50 MWe to 700 MWe (a

few units are between 5 and 50 MWe). Information on worldwide coal-fired

3

pulverized fuel, bubbling and circulating fluidized bed power plants co-firing

biomass or waste is given in Appendix A. As can be seen from the Appendix, the

majority of co-firing applications are carried out in pulverized fuel-fired power

plants. This is considered to be due to the fact that most of the existing plants are

based on this conventional technology and are forced to comply with the CO2

reduction legislation.

For co-firing high volatile, moisture and low ash content biomass with low volatile,

moisture and high ash content coal, a fuel flexible clean combustion technology is

necessary. Within the available technologies, fluidized bed combustion (FBC)

technology is usually indicated to be the best choice, or sometimes the only choice.

This is confirmed by extensive experimental investigation carried out to date on

technical and economical feasibility and performance of different types of biomass in

FBC technology [12], and also by steadily growing number of industrial applications

(Appendix A [13]).

1.2 Aim and Scope of the Thesis

In Turkey, there exist 9.3 billion tons of indigenous lignite reserves characterized by

low calorific value and high ash and sulfur contents. Absence of studies on co-firing

of indigenous lignite/biomass blends in bubbling fluid bed combustors, on one hand,

and the recent trend in utilization of biomass with local reserves in industry and

utility boilers, on the other, necessitate investigation of co-firing lignite with

biomass.

The most available biomass sources in Turkey are olive residues, hazelnut shells and

cotton residues. Turkey is one of the main olive producers with 662 000 ha of olive

groves and 1 200 000 tons of production in year 2005 [14]. Hazelnut production in

Turkey accounts for 70 % of the worlds’ total production with 584 000 ha plantation

and 530 000 tons of production in year 2005 [15]. Moreover, Turkey is one of the

leader producers of cotton in the world with 546 880 ha plantation, 1 291 180 tons of

cotton lint and 863 700 tons of cotton seed productions in year 2005 [14].

4

Therefore, significant amounts of olive residues, hazelnut shells and cotton residues

are produced in Turkey to be used in co-firing applications.

Olive residue is a specific type of biomass from olive oil production process. It is the

remaining part of olives after pressing and extraction of olive oil. Annual production

of olive residue is about 682 993 tons [16]. Hazelnut shells, produced from crushing

of hazelnuts, have annual production of 453 184 tons [16]. Cotton residues are

produced from cotton oil production process. Cotton is grown mainly for the fibers

or lint, but cotton seeds having high amount of oil content are also very important.

Lint and fiber of the cotton are removed firstly by hand and then mechanically by

ginning process. The remaining part, cotton seed, is then extracted for cotton oil.

Annual production of cotton residues in Turkey is about 593 972 tons [16].

Availability of significant amounts of these residues and indigenous lignite reserves

together with gradual introduction of increasingly restrictive legislations on

emissions makes co-firing option attractive.

Therefore, the objective of this thesis study has been to investigate co-firing of

typical indigenous lignite with olive residue, hazelnut shell and cotton residue in the

METU 0.3 MWt Atmospheric Bubbling Fluidized Bed Combustion (ABFBC) Test

Rig in terms of combustion and emission performance and ash deposition tendencies.

5

CHAPTER 2

BACKGROUND AND LITERATURE REVIEW

2.1 General

Prior to presenting a detailed review of previous studies on biomass firing and

biomass and coal co-firing in fluidized bed combustion systems (FBC), various

issues associated with biomass characteristics and combustion are briefly

overviewed.

2.2 Biomass Firing in FBC Systems

2.2.1 Biomass Characteristics

The term biomass describes carbonaceous materials derived from plants. Biomass

fuels can be classified as agricultural residues, woody residues, dedicated energy

crops and industrial and municipal waste of plant origin [17]. Unlike conventional

fuels, some physical and chemical properties of biomass such as high volatile matter

and moisture content, low bulk density and low ash melting temperature complicate

its processing and combustion [18].

Table 2.1 displays proximate and ultimate analyses of some selected biomass and

coal. As can be seen from the table, biomass fuels are mainly characterized by their

high volatile matter and low ash contents.

6

LHV

(M

J/kg

)

19.5

0

17.2

3

11.7

0

12.3

4

17.7

0

17.9

9

15.9

5

17.3

0

11.7

0

7.99

15.2

0

25.5

9

25.4

0

12.2

6

Ash

8.00

6.22

22.6

0

14.6

0

1.30

5.22

0.88

4.80

0.30

0.32

0.92

18.1

0

14.8

0

42.2

0

Cl

0.01

1.05

0.08

0.00

0.05

0.00

0.00

0.00

0.00

0.00

2

0.00

0.00

0.00

0.00

O

32.8

1

39.7

0

35.3

6

41.1

2

43.8

5

38.6

7

45.7

1

42.0

0

46.0

0

41.3

8

44.1

8

8.25

12.5

0

12.4

0

S 0.30

0.22

0.06

0.04

0.05

0.00

0.67

0.00

0.10

0.10

0.10

0.71

0.60

2.70

N

1.38

0.50

0.30

0.36

0.20

1.23

0.22

0.30

0.20

0.10

0.20

1.34

1.20

1.40

H

5.80

5.36

4.70

5.46

5.97

5.59

5.76

5.50

6.10

6.00

6.10

3.73

3.70

3.20

Ulti

mat

e A

naly

sis (

dry,

wt %

)

C

51.7

0

46.9

5

36.9

0

38.4

2

48.5

8

49.2

9

46.7

6

47.4

0

47.3

0

52.1

0

48.5

0

67.8

7

67.2

0

38.1

0

Ash

7.28

5.74

20.0

2

12.9

9

1.24

4.90

0.77

4.22

0.20

0.15

0.80

17.1

1

14.5

0

36.4

0

FC

18.4

9

14.4

6

25.1

7

18.0

1

28.2

2

10.3

24.0

8

15.0

5

13.3

0

6.76

26.6

0

52.2

4

60.2

0

17.2

0

VM

65.3

0

72.0

5

43.4

1

58.0

0

65.5

4

78.7

0

62.7

0

68.7

3

51.6

0

39.7

9

59.6

0

25.1

6

23.0

0

32.7

0

Prox

imat

e A

naly

sis (

as re

ceiv

ed, w

t %)

Moi

stur

e

8.93

7.75

11.4

0

11.0

0

5.00

6.10

12.4

5

12.0

0

34.9

0

53.3

0

13.0

0

5.49

2.30

13.7

0

Tab

le 2

.1: P

roxi

mat

e an

d ul

timat

e an

alys

es o

f sel

ecte

d bi

omas

s and

coa

l.

Fuel

Bio

mas

s

Oliv

e re

sidu

e [1

9]

Whe

at st

raw

[19]

Ric

e st

raw

[20]

Ric

e hu

sk [2

1]

Bag

asse

[20]

Cot

ton

seed

cak

e [2

2]

Haz

elnu

t she

ll [2

3]

Bar

k [2

4]

Woo

d ch

ips [

25]

Saw

dust

[26]

Pine

seed

shel

l [27

]

Coa

l

Bitu

min

ous c

oal [

28]

Sout

h A

fric

an c

oal [

27]

Lign

ite [2

9]

7

It is possible to burn any type of biomass, but in practice combustion in fluidized bed

combustors is feasible only for biomass with moisture content less than 60 % unless

the biomass is pre-dried [6]. High moisture content can lead to poor ignition and

reduce combustion temperature and hence quality of combustion [18].

Volatile matter content of biomass is greater than that of coal. High volatile content

leads to easy ignition and burning which may cause difficulties to control the

operation.

Biomass bulk densities are lower compared to that of coals which may lead to

transportation, storage, feeding and firing problems. Coal densities typically range

from 900 kg\m3 for low rank coals to 2330 kg\m3 for high density pyrolytic graphite

[30]. Biomass densities range from 100 kg\m3 for straw to 590 kg\m3 for olive

residue.

It is generally unfeasible to reduce biomass size as most of the biomass has fibrous

structure [8]. Blockage and discontinuous dosing problems may occur during feeding

biomass with screw feeders in fluidized bed combustors which can be prevented by

replacing them with pneumatic feeding systems [18].

Most of the biomasses are characterized by low ash contents and hence they require

addition of bed material in order to maintain the bed at a constant level during the

operation [9].

2.2.2 Biomass Combustion

Biomass has similar combustion mechanisms to coal such as drying, devolatilization

and combustion. During devolatilization they undergo thermal decomposition with

subsequent release of volatiles and char and tar formation [18]. Combustion is

composed of two phases which are volatile matter and char combustion phase. Coal

has short volatile and long char combustion phases whereas biomass has long

volatile and short char combustion phases. In biomass combustion devolatilization

8

starts at low temperatures and almost all combustion is completed in this phase.

Devolatilization of high volatile matter content can also result in a highly porous

char, thus accelerating char combustion as well [2].

In fluidized bed combustion of biomass, location of volatiles release and combustion

mainly depend on method of feeding and distribution of combustion air [18]. Over-

bed feeding generally leads to higher freeboard temperatures compared to under-bed

feeding as a consequence of less uniform spreading of the fuel along the combustor

[31]. Under-bed feeding provides better combustion of biomass and leads to lower

CO emissions [32].

Generally, biomass fuels have low melting temperature ash mainly composed of high

alkali oxides and salts which lead to many operational problems during combustion.

In the following part, problems associated with ash content of biomass are

summarized.

2.2.3 Operational Problems Associated with Biomass Ash

Operational problems in biomass combustion are mainly originated from inorganic

components of the fuel. Figure 2.1 shows ash compositions of some selected biomass

and coal. As can be seen from the figure, biomass ashes have higher portions of

potassium and sodium with respect to those of coals.

High sodium and potassium content of biomass ash cause bed agglomeration,

slagging on furnace walls and fouling of heat transfer surfaces which lead to reduced

reliability of electricity production from biomass. The successful design and

operation of a fluidized bed combustor depends on the ability to control and mitigate

these ash related problems [9]. Table 2.2 gives the possible operational problems

related with biomass ash. Biomass ash has relatively low ash fusion temperature

compared to that of coal.

9

Figure 2.1: Biomass ash compositions.

Table 2.2: Influence of biomass ash on boiler performance [39].

Fuel constituents Operational Problems

Alkali content (Sodium, Potassium)

Bed agglomeration Slagging Fouling of heat transfer surfaces Hot corrosion

Chlorine

Hot corrosion Fouling HCl emission Dioxin formation

Heavy metals Emissions Furnace corrosion Ash handling

10

The rate of slag formation on furnace walls tends to increase owing to the reduction

of ash fusion temperature by introduction of biomass ash into the system. The

problems associated with ash characteristics of biomass in fluidized bed combustion

of biomass are summarized in the following sections.

2.2.3.1 Bed Agglomeration

Agglomeration occurs when a part of fuel ash melts and causes adhesion of bed

particles [9]. The factors affecting agglomeration are bed temperature, ash

composition and alkali content of the fuel. Temperature has the most pronounced

effect on agglomeration tendency. Alkaline compounds resulting from high

potassium and sodium in biomass ash have very low melting points. In addition to

adhesion effect of sintered ash, alkali oxides or salts can react with silica compounds

of the bed material according to [18];

2 2 3 2 2 2

2 2 3 2 2 2

4 SiO + K CO K O . 4 SiO + CO (2.1)2 SiO + Na CO Na O . 2 SiO + CO (2.2)

⎯⎯→

⎯⎯→

Potassium and sodium form eutectic mixtures with melting points of 874 and 764 ºC,

respectively. At higher concentrations, melting point of these eutectic mixtures is as

low as 650-700 °C which is well below the normal operation temperature of a

fluidized bed combustor [9].

If sufficient amount of Fe2O3 is present in the ash of the fuels burned, formation rate

of agglomerates may be reduced by [18],

2 3 2 2 2 4

2 3 2 2 2 4

2 3 2 3 2 2 4 2

Fe O + K O K Fe O (2.3)Fe O + Na O Na Fe O (2.4)

Fe O + K CO K Fe O + CO

⎯⎯→

⎯⎯→

⎯⎯→

2 3 2 3 2 2 4 2

(2.5)Fe O + Na CO Na Fe O + CO (2.6)⎯⎯→

11

Figure 2.2 displays extreme agglomeration types. Type 1 is more commonly

observed in commercially operated fluidized bed combustors burning woody type

fuels and results from “coating-induced” agglomeration. Here, a uniform coating is

formed on the surface of the bed material grains. At certain critical conditions like

coating thickness or temperature, neck formation may occur between coatings of

individual grains which initiate agglomeration.

Agglomeration due to melt formation

Sintering of coatings

Bed grain

Bed grain

Gas phase ash

Molten ash particles

Coating

Bed grain

Figure 2.2: Agglomeration of bed material: Type 1- coating-induced,

Type 2- melt-induced [40].

The other type, results from “melt-induced” agglomeration. In this case, bed material

grains are glued together by a melt phase, which roughly matches the chemical

composition of the ash and is produced at normal operating temperature. If, however,

after first neck formation partial de-fluidization of the bed leads to local peak

temperatures, melt formation may occur and a combination of the two extremes can

be recognized in one sample. [40].

12

Agglomeration tendency of the fuel can be estimated by the ratio of iron oxides to

sum of potassium and sodium oxides in the fuel ash. Agglomeration occurs when the

ratio is less than 0.15 [9].

2 3

2 2 (2.7)

%(Fe O )Bed Agglomeration Index (BAI) = %(K O+Na O)

2.2.3.2 Ash Deposition and Corrosion

Slagging, fouling and corrosion of heat transfer surfaces are mostly influenced by

high alkaline, chlorine and low sulfur content of biomass. Ash constituents which

result from combustion processes react with flue gas or with each other to form

variety of compounds. Partially molten ash particles in the flue gas impact and

deposit on the heat transfer surfaces and furnace walls which are termed as fouling

and slagging, respectively.

Fouling tendency of the fuel can be estimated based on different indicators such as

the ratio of alkali metal oxides to silica oxide. The index higher than 1 indicates

severe fouling possibilities [39].

2

2 2 (2.8)%(K O+Na O)Alkali Index (AI) = %(SiO )

As indicated previously, inorganic compounds (mostly sodium, potassium and

chlorine) in easily vaporizing form lower the melting temperature of ash and

consequently expected to provoke serious ash deposition problems. Hot corrosion of

the boiler tubes becomes a problem with fuels having high chlorine. When the fuel

chlorine content exceeds 0.1 %, corrosion may occur on heat transfer surfaces [39].

Chlorine concentration often dictates the amount of alkali vaporized during

combustion more strongly than the alkali concentration in the fuel. In most cases

chlorine appears to play a shuttle role, facilitating the transport of alkali from the fuel

to surfaces where alkali often forms sulfates [41]. Removal of chlorine from fuels by

13

leaching is usually not economical. Corrosive deposits tend to damage the heat

transfer surfaces at metal temperatures >470 ºC and typical vapor temperatures of

modern steam boilers exceed 500 ºC [42].

Superheater corrosion and ash deposition can be reduced in biomass firing FBC

systems by;

• Leaching biomass to reduce its ash content.

• Using alternative bed material other than SiO2 (e.g. Fe2O3).

• Using additives (alumina silicates) to form alkali alumina silicates and

prevent the release of gaseous KCl and NaCl or react with them in the

gas phase and form less corrosive compounds (e.g. kaolin, bauxite).

• Using sulfur containing additives for sulfation of gaseous alkalis, KCl

and NaCl to form less corrosive sulfates, K2SO4 and Na2SO4.

• Co-firing biomass with coal.

Control of the rate of deposit formation in biomass combustion is associated with the

reactions between compounds containing chlorine, sulfur, aluminum and alkali. High

risk alkali chlorine compounds, NaCl and KCl, can be trapped by reactions with SO2

[43] and aluminum silicates [42, 44] to liberate gaseous HCl as shown in below

reactions;

2 2 2 2 4

2 2 2 2 4

2KCl(s)+SO (g)+1/2 O (g)+H O(g) K SO (s)+2HCl(g) (2.9) 2NaCl(s)+SO (g)+1/2 O (g)+H O(g) Na SO (s)+2HCl(g) (2.10)

→→

2 3 2 2 2 2 3 2

2 3 2 2 2 2 3 2

Al O 2SiO (s)+2KCl(s)+H O(g) K O Al O 2SiO (s)+2HCl(g) (2.11)Al O 2SiO (s)+2NaCl(s)+H O(g) Na O Al O 2SiO (s)+2HCl(g) (2.12)

⋅ → ⋅ ⋅⋅ → ⋅ ⋅

Corrosion of heat transfer surfaces can also be caused by acid components in flue

gases. Generally hydrogen sulfide (H2S) and hydrogen chloride (HCl) are the most

harmful gaseous species in flue gases. However, in the absence of reducing

14

conditions, high concentrations or high temperatures required to initiate corrosion

process, they are stable and leave the system with flue gas [45].

The S/Cl ratio in the feedstock has often shown to affect Cl deposition and corrosion.

Theoretically based in the sulfation reaction shown above, S/Cl ratio of 0.5 should be

enough to sulfate all alkali chlorides to alkali sulfates. However, in practice higher

ratios are required due to reactions with calcium compounds which consume

available SO2 from the flue gas. It has been suggested that if the S/Cl ratio of fuel is

less than 2, there is a high risk of superheater corrosion. When the ratio is at least 4,

the blend could be regarded as non-corrosive [46].

Inside the deposits, close to the heat transfer surface where temperature is below 600

ºC KCl is in solid form. Sulfation reaction can occur inside the deposits at anhydrous

conditions and produce free Cl2, instead of HCl, according to reaction [42];

2 2 2 4 22 KCl (s) + SO (g) + O (g) K SO (s) + Cl (g) (2.13)→

Free Cl2 may also form in the reaction between KCl and Fe2O3 which exists on

superheater tube surface according to the following reaction [42];

2 3 2 2 2 4 22 KCl (s) + Fe O (s) + 1/2 O (g) K Fe O (s) + Cl (g) (2.14)→

Chloride is highly reactive with iron and chromium, forming metal chlorides and in

long term these reactions will destroy heat exchange tubes, however, alkali sulfates

formed as result of reactions given in (2.8)-(2.11) are much less harmless to heat

exchange tubes.

Due to differences in elemental composition of biomass and coal, chlorine and alkali

metal behavior during co-firing will be different from both fuels because of the

interactions between volatile elements (sodium, potassium and chlorine) and other

mineral elements. Major ash forming elements of coal (aluminum and silicon) have

significant influence on this behavior [45]. Coal brings within itself protective ash

15

elements to the system which lowers gas phase alkali chlorides. Full-scale co-firing

of coal with high alkali content straw is reported to show little or no chlorine in the

deposits [47]. Protective behavior of ash forming elements in coal is illustrated in

Figure 2.3. In Case 1, biomass is burning alone. The ash of biomass fuels has high

alkaline metal content. When this is associated with high chlorine content, which is

often the case, these elements react to form alkali chlorides. This, in turn, induces

corrosion rates after deposition of these substances on the heat transfer surfaces. In

Case 2, sulfur and aluminum silicates from coal ash are able to form alkali silicates

and alkali sulfates. Hence, chlorine is released as HCl in flue gases and alkali metals

are bound in compounds that have a high melting point and no corroding effect [48].

Biomass

Combustion Biomass and Coal

Co-combustion

Case 1 Case 2

Figure 2.3: Corrosion effects of alkalis in biomass combustion

and co-combustion systems [48].

Due to these operational problems in biomass firing, co-firing of biomass with coal

becomes an attractive alternative from environmental and operational points of view.

16

2.3 Literature Review

The use of biomass in energy generation has importance as global warming is

concerned since biomass firing has the potential to be CO2 neutral. This is

particularly the case with regard to agricultural residues and energy plants which are

periodically planted and harvested [18]. They remove CO2 from atmosphere during

their growth and release again during combustion. Biomass types having high energy

potential include agricultural and forestry residues are rice husk, straw, olive residue,

sugar cane bagasse, pine barks and wood chips. In the following parts, review of

biomass firing and biomass and coal co-firing studies on agricultural and forestry

type biomass residues in FBC systems are presented.

2.3.1 Biomass Firing Studies

Rice husk is one of the high moisture and volatile matter content agricultural

residues. Rice husk alone is difficult to fluidize, but fluidization can be improved by

the addition of bed material. Details and operating conditions of FBC systems

burning rice husks are summarized in Table 2.3.

Combustion of rice husks was investigated in the study of Preto et al. [49] in a pilot

scale bubbling fluidized bed. The operating parameters in the experiments were

temperature, fluidizing velocity, rice husk feed rate and excess air. No feeding

problems were encountered during under bed feeding of rice hulls with screw feeder.

Due to high volatile matter content of rice husks, combustion mainly took place in

freeboard region. Over 97 % combustion efficiency was achieved. At fluidizing

velocities greater than 1.75 m/s, temperature peak was at the top of the freeboard

which opened up the potential of combustible losses. At low fluidizing velocities,

temperature gradient occurred in the bed which indicated improper mixing in bed.

17

Table 2.3: Operating conditions and system properties of rice husk firing FBCs.

Reference Fuel Operating conditions System Bed material

Preto et al. [49]

Rice husks

Bed temperature: 650-900 °C u0: 0.4-2.2 m/s

Feeding rate: 22-70 kg/h Excess air: 30-95 %

BFBC 1 MWt 0.380 ×0. 406 m cross section & 4.8 m height

Sand

Guanyi et al. [11]

Rice husks Bed temperature: 750-850 °C

u0: 1-2 m/s Feeding rate: 20-40 kg/h

CFBC 1 MWt0.784 m2

cross section & 6 m height

Silica sand

550 μm

Armesto et al. [50]

Rice husks Bed temperature: 840-880 °C u0: 1-1.2 m/s

BFBC 0.3 MWt CIEMAT

Sand

Fang et al. [51] Rice husks Bed temperature: 750-850 °C

Excess air coefficient: 1.1-1.2

CFBC 1 MW

0.04 m2 cross section & 6 m height

Silica sand

550 μm

Permchart &

Kouprianov [52]

Rice husk Saw dust

Sugar cane bagasse

Excess air: 20, 40, 60, 80, 100 %

Silica sand 300-500 μm

Kurpianov et al. [53]

Rice husks Sugar cane

bagasse

Bed temperature: 700-800 °C Excess air: 40, 60, 80, 100 %

Feeding rate: 82.8 kg/h

BFBC Conical lower section (40º

cone angle), 0.9 m inner

diameter & 3 m total height

Silica sand

0.45 mm

Albina [32] Rice husks

Under-bed, over-bed feeding Primary to secondary air ratio:

40, 50, 60 % Excess air: 10, 20, 30 %

BFBC multiple-

spout/ spout

0.49 m2 cross section

Silica sand

300-1000 μm

Shimizu et al. [54]

Rice husks Bed temperature: 850 °C u0: 0.39 m/s

BFBC 0.053 m

inner diameter & 1.3 m height

Quartz sand

270 μm, Porous alumina 690 μm

Natarajan et al. [55]

Rice husk Bagasse Cane trash Olive flesh

6 % excess air under normal FBC conditions BFBC

Quartz sand or

lime

Liu et al. [56] Rice husks u0: 0.8 m/s

BFBC 0.255 m2 cross sectional area

Sand

Skrifvars et al. [57]

Rice husks Eucalyptus

bark

EFR flue gas temperature: 900 ºC

Deposit probe temperature: 500 ºC

BFBC 157 MWt & EFR

0.18 m inner diameter & 6 m height

Sand

Bakker et al. [58]

Rice straw Wood

Flue gas temperature: 900 ºC Excess air: 30-50 %

Secondary air: 13-20 % Deposit probe temperature:

500 ºC

Lab scale BFBC

42×10-4 m2 cross section & 0.915 m

height

Alumina-silicate grains

18

However, high ash softening temperature of rice husk ash (~1400 ºC) which have

significant portion of SiO2 ensured acceptable operation. CO emission varied from

200 to 5000 ppm. CO levels showed dependence on both temperature and fluidizing

velocity. The effect of the velocity is obvious as the higher the velocity, the lower is

the residence time. However, effect of temperature was opposite of the expected

behavior which was explained to be the outcome of masking effect of the velocity.

The pollutant emissions ranged from 50 to 150 ppm for SO2 and from 100 to 180

ppm for NOx. The rice husk ash was very fine and easily elutriated out of the bed.

The majority of the ash was collected as cyclone product.

Generally pneumatic feeding systems are preferred to avoid blocking in rice husk

burning fluidized bed combustors. In 1 MWt CFBC system of Guanyi et al. [11]

where ignition and combustion characteristics of rice husks were investigated, the

most serious problem during operation was bridging in the feed bin and blocking

along the screw. Therefore, screw feeder was replaced by pneumatic conveyor to

avoid blockage during fuel feeding. Operating parameters were bed temperature,

fluidization velocity and rice husk feed rate. Secondary air was injected to the system

1.2 m above the distributor to prolong the retention time of rice husk in bed and to

improve combustion. Compared to coal, ignition stage of rice husk was characterized

by lower ignition temperature (340 °C) and easier ignition, fast devolatilization,

intensive combustion and rapid temperature rise and slower temperature drop in

terminal stage and more time for complete burn-out of coke. Similar to the previous

studies, major portion of the rice husk combustion took place in the freeboard region.

About 97 % combustion efficiency was achieved. The emission levels were low with

SO2 ranging from 50 to 150 ppm and NOx ranging from 150-220 ppm.

The effect of temperature and fluidization velocity on emissions and combustion

performance of highly volatile (~74 %) rice husk combustion was investigated by

Armesto et al. [50] in 0.3 MWt CIEMAT bubbling fluidized bed combustor. Due to

high volatile matter content, rice husk was easier to ignite and burn with respect to

coal however; its combustion was rapid and difficult to control. The temperature

range was from 840 to 880 °C and fluidization velocity range was from 1 to 1.2 m/s.

19

Fuel was fed pneumatically from the bottom of the bed. Rice husk was defined as a

problematic fuel during handling and transportation due to its low density. The main

elements of rice husk ash were silica, potassium and phosphorus which could have

detrimental effect on the ash melting properties leading to agglomeration. Increase in

fluidization velocity resulted in reduction of combustion efficiency. Increase in bed

temperature improved the combustion efficiency to about 97 %. CO emissions were

greater than 1000 mg/Nm3. NOx emissions were between 200 and 300 mg/Nm3. Due

to low sulfur content of rice husks SO2 emissions were very low. Potassium and

calcium silicates, which played important role in agglomeration, were detected in the

bed ash, so that during experiments the control of potassium content of the bed was

significant for avoiding agglomeration.

Low CO, SO2 and NOx emissions were reported in the study of Fang et al. [51] who

studied combustion of high volatile content rice husks in 1 MWt circulating fluidized

bed. Rice husk alone was difficult to fluidize but it was well fluidized with silica

sand bed material. Some difficulties such as bridging in the feed bin and blocking at

the outlet occurred in feeding the rice husks with the screw feeder. The situation was

improved by supplying secondary air through the feeding port and by vibrating the

hopper. The ignition temperature of the rice husk was obtained as 340 °C. Due to its

high volatile content, major portion of the rice husk combustion took place in the

freeboard. Secondary air was injected 1.2 m above the distributor plate in order to

improve the combustion. The proper air split was obtained as 7:3 at fluidizing

velocity of 1.2 m/s. Combustion efficiency was 97 %. CO emission varied from 200

to 800 ppm. SO2 emission ranged from 50 to 100 ppm and NOx emission ranged

from 150 to 220 ppm.

The effect of excess air and combustor load on combustion performance and

emissions of some agricultural residues such as rice husk, sawdust and sugar cane

bagasse were investigated by Permchart and Kouprianov [52] in a bubbling fluidized

bed combustor with a conical bed. As received bagasse, was pre-dried from moisture

content of 48.8 % to 14.4 %. Fuel feeding rate and excess air percentage were taken

as independent variables. For maximum and minimum combustor loading, feed rates

20

were 81.5 kg/h and 35 kg/h for sawdust, 82.4 kg/h and 37.3 kg/h for rice husk, 70

kg/h and 31 kg/h for bagasse. The fuels were burnt at 20, 40, 60, 80, 100 % excess

air. The highest temperature in the combustor was observed for saw dust. Despite the

higher heating value of bagasse compared to rice husk, temperature profile of

bagasse was lower due to its high moisture content and smaller feed rate.

Temperatures in the freeboard section were found to have increasing tendency at

higher excess air levels. Increase in excess air form 20 % to 100 %, increased the

temperature at the top of combustor (2.75 m) by 60-80 °C for rice husk and bagasse,

and by 160 °C for saw dust. The maximum oxygen consumption was observed in the

bed section for all the fuels. CO formation rate of rice husk was much greater than

that for other fuels due to its higher fuel ash concentration and coarser char particles

which were 200 μm for rice husk, 10 μm for bagasse and 5 μm for saw dust. These

factors resulted in higher bed hold-up and led to higher values for CO emission from

rice husk combustion. NO emission was expected to originate from fuel-N [18, 59].

Basically, fuel-NO could be formed through combustion of HCN, NH3 with volatile

matter and oxidation of nitrogen retained in the char. These reactions resulted in

rapid NO formation in the bed region. In the upper region of the combustor as

oxygen concentration decreased, NO reduction was seen through its reaction with

NH3, followed by formation of N2 and H2O vapor. During rice husk combustion,

high fuel-N and high amount of coarser ash particles resulted in higher NO formation

in the bed region with respect to saw dust and bagasse. NO was reduced in freeboard

region. N2O formation in biomasses was found negligible. CO emission mostly

depended on the excess air level. When excess air in rice combustion was smaller

than 50 %, CO emission increased however, when the excess air was greater than 50

%, CO emission decreased. For the maximum combustor load and excess air of 50 to

100 %, over 99 % combustion efficiency was achieved when firing saw dust and

bagasse. Lower combustion efficiencies (80-86 %) were obtained for rice husk

combustion owing to high amounts of unburned carbon.

In more recent study of Kuprianov et al. [53] single firing of rice husks and co-firing

of rice husk with sugarcane bagasse were carried out in the same test unit to

investigate the effect of rice husk fraction, excess air on combustion efficiency and

21

emissions. Temperature during single firing of rice husk ranged from 700 to 800 ºC

within the combustor. Increase in the sugarcane bagasse fraction having 48.8 %

moisture content resulted in reduced temperatures at around 550-650 ºC. Increasing

excess air from 40 to 100 % did not show a certain effect on bed temperature. The

axial O2 profiles for various blends were similar to single combustion of rice husk.

CO emission was reduced from 2000 ppm to about 500 ppm with increasing excess

air from 40 to 100 %; however, the effect of excess air on CO emission was

weakened for higher values. NO emission was found to increase with increasing

excess air from about 120 to 200 ppm. Combustion efficiency of single firing rice

husk tests were about 96 % in the excess air range of 40 to 100%. Combustion

efficiency of co-firing of rice husk having greater than 60 % fraction on energy base

was similar to single rice husk firing and thus, found to be feasible.

The effect of excess air, primary to secondary air ratio and over-bed and under-bed

feeding on CO and CO2 emissions from rice husk combustion were investigated in

the study of Albina [32] in the multiple spouted and spouted fluidized beds. High

volatile matter and ash content of rice husks influenced the combustion and emission

characteristics. Due to its lower ignition temperature it was difficult to control the

operation compared to coal combustion. During over-bed feeding in multiple-

spouted bed, combustion efficiency peaked to 92 % at 10 % excess air and reduced

to 83 % when additional excess air was supplied. CO emission during under-bed

feeding was less compared with the emission with over-bed feeding. However, CO2

emission did not change significantly for both methods of feeding. Combustion

efficiency during under-bed feeding ranged from 92 to 94 % showing a peak value at

20 % excess air and resulted in lower CO emission. Highest values were obtained at

the primary to secondary air ratio of 40/60 for both over and under-bed feeding. To

determine the influence of different nozzle designs, some experiments were

performed in spout-fluid bed. CO emissions were lower at 10 % excess air due to

high combustion efficiency. Over or under-bed feeding had no significant influence

on CO and CO2 emissions in the given excess air range. Combustion efficiency

ranged from 95 to 98 % during both over and under-bed feeding. Lower CO

emissions were measured in spouted bed compared to that of multi-spouted bed. This

22

result was attributed to higher combustion efficiency attained by spouted fluidized

bed.

Using alternative bed materials instead of silica sand was considered to improve the

emission performance of biomass combustion systems. The influence of quartz sand

and porous alumina on CO and NOx emissions from combustion of rice husks was

investigated by Shimizu et al. [54] in a batch scale fluidized bed combustor. Quartz

sand was used as a conventional bed material and alternatively porous alumina was

employed. For quartz sand high CO emission was observed at O2 concentration

lower than 6 %. For alumina bed, the increase in CO emission was observed at O2

concentration lower than 4 %. Porous alumina bed was found to suppress CO

emission. The enhanced volatile matter combustion and reduced CO emissions was

explained by hydrocarbon capture of porous particles. The capture of volatile matter

in the pores prolonged its residence time in dense bed, thus enhanced the burn up of

the volatile matter. Another effect of volatile matter capture was, enhanced

horizontal mixing of carbonaceous material in the volatile matter. Carbon in solid

was dispersed horizontally by solids mixing. Both effects were effective in CO

emission reduction. NOx emissions for porous alumina were found to be lower (250-

550 ppm) than that of the quartz sand bed (150-400 ppm). One possible mechanism

of NOx reduction could be carbon retention within the pores which is known to

reduce NOx to N2 under fluidized bed conditions. Thus the increased carbon loading

in the bed material by volatile matter suppressed NOx emission. N2O emissions were

below 10 ppm for both bed materials. During the experiments with silica sand bed

material defluidization took place however, with alumina bed material no such

problem observed.

Prediction of agglomeration tendencies of agricultural residues are mainly obtained

by ASTM standard ash fusion test [60]. However, this test has been reported to be a

poor indicator of ash related problems in fluidized beds [61, 62] but, it is still used

since there is no reliable alternative method. To compare the predictions of ASTM

standard ash fusion test with agglomeration tendencies of residues in fluidized bed

Natarajan et al. [55] performed experiments on rice husk, bagasse, cane trash and

23

olive flesh by a method based on controlled fluidized bed agglomeration test. The

most important parameter for bed agglomeration is the actual bed temperature.

However, the particle temperature in fluidized bed combustion is an unknown

variable which may exceed the bed temperature [63, 64]. To avoid the uncertainty in

the surface temperatures of the particles in the controlled bed agglomeration tests,

experiments were performed without actual combustion but in the similar

combustion atmosphere. In the experiments, the bed was charged with the bed

material (quartz sand or lime). Then fuel was fed to the combustor and burnt with 6

% excess air under normal FBC conditions to produce the required ash content for

further agglomeration test. When 10 % ash is formed in the bed, fuel feeding

stopped and combustion atmosphere is maintained by air pre-heater in order to avoid

localized high temperature around particles and to maintain isothermal conditions in

the bed. Then the bed temperature was raised by 3 °C/min until the agglomeration

was encountered. Agglomeration was indicated by the differential pressure drop in

the bed due to cohesion of ash particles. From the ASTM ash fusion tests, initial

deformation temperatures were obtained as 1600 °C for rice husks, 1200 °C for

bagasse, 900 °C for cane trash and 1100 °C for olive flesh. But during FBC of these

residues significant melting of the ash occurs far below the initial deformation

temperature. Initial agglomeration temperatures were 1009 and 1020 °C for rice

husks, 1020 and 1020 °C for bagasse, 890 and 905 °C for cane trash and 933 and

1020 °C for olive flesh in the case of the bed materials quartz sand and lime,

respectively. The use of lime instead of quartz sand as the bed material improved the

agglomeration temperature. Cane trash and olive flesh showed higher bed

agglomeration tendency than rice husk and bagasse.

Slagging and agglomeration characteristics of rice husks were studied by Liu et al.

[56] in a bubbling fluidized bed combustor. FBC is given as a recommended way for

burning rice husk with excellent heat and mass transfer characteristics. Chemical

reaction kinetics of rice husk was obtained on the basis of thermogravimetric

experiments. Agglomeration and slagging characteristics of rice husks were studied

in crucibles heated in electrical furnace for 2 hours at 950, 1000 and 1050 °C. Ash

fusion temperature of rice husk was measured to be higher than 1500 °C which

24

implied safe operation under FBC conditions. Experiments were carried out with an

accelerated surface area and porosimetry system to examine the microscopic

structure change of rice husk. Pore volume and specific surface area of rice husk

reached maximum at 850 °C but decreased when temperature exceeded 900 °C.

During FBC of rice husk release of volatiles and burning of carbon occurred almost

separately which complicated its burnout. High silica content of rice husk ash (95 %)

resulted in combination of silica crystals with carbon at high temperatures so that

carbon did not oxidize even at 1200 °C. Therefore, high temperature was not

indicative of high efficiency in the case of rice husk. The optimum temperature range

for fluidized bed combustion of rice husk was given from 800 to 850 °C.

Bubbling fluidized bed boilers generally are very suitable for biomass fuels and fuel

mixtures. However, all combustion systems burning biomass are quite sensitive to

fuel ash behavior. Fouling and corrosion of heat exchange surfaces are common

problems in units burning biomass. Slagging and fouling behavior of rice husks

having high ash content (20 wt %) were investigated by Skrivars et al. [57] in pilot

and full scale units. The investigation was based on short term (3-10 h) deposits

samples taken with air-cooled deposit probes in the superheater region of a large

scale (157 MWt) bubbling fluidized bed boiler burning rice husks and eucalyptus

bark. From previous experiences, it was stated that for biomass fired boilers 2-3 h

sampling time was sufficient to give the first indication of fouling. Pilot scale tests

were performed in an entrained flow reactor which was built to simulate conditions

that fly ash particles experience around the superheater area in bubbling fluidized

bed boilers. Thus, flue gas temperature was set to 900 ºC and deposit probe

temperature was set to 500 ºC. Rice husk ash was composed of 95 wt % SiO2 with

the remainder being mostly K, Ca and P and produced coarse fly ash particles. When

fired alone, rice husk particles did not stick onto heat exchange surfaces and caused

no significant fouling or slagging problems in both pilot and full scale tests. In co-

firing tests, the presence of rice husk ash particles in fly ash kept the tube surface

unfouled, even if a fouling fuel such as eucalyptus bark was fired together with rice

husk.

25

Leaching of inorganic constituents from biomass prior to combustion has been

shown to be of substantial benefit in improving combustion properties and reducing

fireside fouling [65, 66]. Rice straw ash content is greater than rice husk ash content

and rice straw ash differs from rice husk ash in that it contains lower silica and

higher potassium. Therefore, it involves higher possibility to cause agglomeration,

slagging and fouling problems. In the study of Bakker et al. [58] rice straw

combustion tests were performed in a laboratory scale fluidized bed to asses fouling,

slagging and agglomeration properties of rice straw due to leaching. Leaching

extracted large amount of alkali metals and chlorine from rice straw ash so that

potassium and chlorine contents were reduced from 16.60 to 1.99 % and 1.15 to 0.03

%, respectively. During leached rice straw experiments, good fluidization is achieved

and no evidence of agglomeration was observed. A slight light brown deposit

accumulated on the air cooled deposit probes. However, in untreated rice straw

combustion tests, extensive bed agglomeration occurred. Inspection of the reactor

revealed that a large section of reactor tube was completely filled with agglomerated

straw ash and bed media. Most of the agglomerates consisted of straw ash and char

with bed particles adhering to the char. A bright white deposit consisting of very fine

particles had accumulated on the air cooled deposit probes, and some larger black

ash particles were adhering to this fine deposit.

Straw is another mostly utilized agricultural residue in fluidized bed combustors.

Similar to rice husk case, addition of silica sand bed material is needed to have better

fluidization during straw combustion. Details and operating conditions of FBC

systems burning straw are summarized in Table 2.4.

Main operational problem is agglomeration during straw combustion. High

potassium content of straw causes formation of agglomerates and eventually

defluidization. In combustion processes, potassium containing compounds are prone

to remain in the bed and form low melting potassium-silicates. The molten ashes coat

the bed material, promoting agglomeration and defluidization in fluidized beds [67].

26

Table 2.4: Operating conditions and system properties of straw firing FBCs.


Lin et al. [68]

Straw

Bed temperature: 805-930 °C

Stoichiometric factor: 0.95-2.58

Quartz sand

Lin et al. [67]

Wheat straw

Bed temperature: 725-930 °C

Stoichiometric factor: 1-2.6

Gas flow rate: 14-28 Nl/min

BFBC 0.068 m inner diameter & 1.2

m height Quartz sand 275-460 μm


Wheat straw Olive flesh

Peat, Sugar cane

bagasse Cane trash

Forest residue Grass Bark RDF

ASTM Ash fusion test Sintering test

Laboratory scale FBC test N.A.

Öhman et al. [70]

Wheat straw Sugar cane

bagasse Forest residue

Grass Bark

Quartz sand

Öhman & Nordin

[71]

Wheat straw Bark

Bed temperature: 760 ºC, The agglomeration test was

at 650 ºC (for straw) Quartz sand & 10 wt % Kaolin (200

µm)

Brus et al. [72]

Wheat straw Olive residue

Bark Peat

Grass

Bed temperature: 800 ºC For straw and grass 650 ºC

40 h operation

BFBC 5 kW

0.1, 0.2 m inner

diameter in bed and

freeboard, respectively

& 2 m height

Quartz sand & blast-

furnace slag (106-125

µm)

Arvelakis et al. [19]

Wheat straw Olive residue Bed temperature: 800 ºC

BFBC 0.0996 m

inner diameter &

2.55 m height

Sand

Nielsen et al. [73]

Straw

Exposure of superheater tube alloys to synthetic

flue gas (6 vol % O2, 12 vol %

CO2, 400 ppmv HCl, 60 ppmv SO2, balance N2) at

550 ºC

Electrically heated oven N.A.

N.A: Not available

27

Temperature has strong influence on agglomeration. Agglomeration tests for straw

were performed by Lin et al. [68] in a laboratory scale bubbling fluidized bed. The

influence of temperature and stoichiometric factor on defluidization time was

investigated. Straw was manufactured as pellets of 1-10 mm to avoid feeding

problems. Agglomeration of the bed material was marked by significant pressure

drop over the bed. Increasing the temperature accelerated the agglomeration time.

Increasing the stoichiometric factor also increased the agglomeration tendency

however; its effect was negligible compared to that of temperature. The ash was

mainly composed of K2O and SiO2 which formed low melting point eutectic

mixtures below 800 °C. The adhesive efficiency of the particles increased

significantly with increasing the temperature after exceeding a critical point and a

low viscous molten phase occurred [74]. This caused the accumulation of the

particles and formation of agglomerates in the bed during higher temperature

combustion at 930 ºC. During the lower temperature experiments at 805 ºC, it was

found that agglomerates have already been formed long before the pressure drop.

When the particles began to agglomerate, they form multi-size particles which tend

to segregate with bigger or heavier particles at the bottom (jetsam) and smaller or

lighter particles at the top (flotsam). In straw combustion, the agglomerates mostly

acted as jetsam particles. Increasing agglomerate levels could cause segregation and

a layer of defluidized agglomerates would form.

The influence of bed temperature, gas velocity, particle size and stoichiometric factor

of bed material on agglomeration tendency during Danish wheat straw (in 1-10 mm

pellet form) combustion were investigated by Lin et al. [67] in the same laboratory

scale BFBC system mentioned above. Defluidization occurred in all the combustion

experiments which were indicated by a sudden decrease in the pressure drop over the

bed. Temperature profile was another indicator of defluidization. When the bed was

in normal fluidization state, bed temperature was very uniform. Just before

defluidization, difference between temperature in the bottom of the bed and 2 cm

above (bed temperature) became larger due to poor mixing. It was noticed that

pressure drop declined slowly before defluidization occurred at relatively low

combustion temperatures, suggesting a segregation of large agglomerates to the

28

bottom of the bed. Temperature influence on defluidization time was obtained to be

very significant. Straw ash started to melt already at 750 °C. Potassium compounds

in straw ash remained in bed and little amount was evaporated at 810-900 °C.

Decrease in temperature led to increasing defluidization time. On the other hand,

increase in temperature caused decrease in ash melt viscosity and coatings were

formed which accelerate defluidization. Doubling the gas flow rate provided longer

defluidization time due to better mixing of the particles and increased force acting on

agglomerates. The better mixing and higher breaking rate of formed agglomerates

prolonged defluidization time about 30 %. Increasing the particle size of the bed

material caused shorter defluidization times because of the particles having smaller

outer surface area that resulted in thicker coating layer. The effect of stoichiometric

factor on agglomeration was not found to be significant.

Agglomeration tendency of ten biomasses during combustion in a fluidized bed

boiler were predicted by three different methods in the study of Skrifvars et al. [69].

ASTM ash fusion test, a sintering test based on compression strength measurements

of ash pellets and a combustion test in a laboratory scale fluidized bed furnace in

which bed agglomeration was achieved in a controlled manner were used to predict

agglomeration tendency. A Danish wheat straw rich in chlorine (3.7 %), olive flesh

from olive oil processing waste, sugar cane bagasse, cane trash mainly the leaves

from sugar cane, reed canary and Lucerne grass, peat, forest residue, bark from

barking process of a pulp mill material and refuse derived fuel (RDF) from source-

separated community waste were used in agglomeration tests. In all cases, except for

wheat straw, the initial deformation temperature was above 1000 ºC, the highest

being 1550 ºC for grass ash. For the wheat straw initial deformation temperature was

900 ºC. Ash from Lucerne grass showed the lowest sintering temperature as 625-650

ºC while peat ash showed the highest sintering temperature as 1000-1100 ºC.

Sintering temperature of wheat straw was about 700 ºC. In combustion tests the bed

agglomeration was detected from bed pressure drop. Agglomeration temperatures of

wheat straw, olive flesh, bagasse and bark were detected as 740, 930, 1020 and 988

ºC, respectively. In all cases the ASTM ash fusion test failed to predict the bed

agglomeration temperature. Many processes in the ash affecting bed agglomeration

29

such as sintering and melting had already started at far lower temperatures than ash

fusion test detected. Although compression strength tests gave more accurate results

than as fusion tests, it also had limits. This test only detected ash particle to ash

particle or ash particle to gas phase interactions. Possible interactions with bed

material were not detected. Adding bed material would mainly give a diluting effect

on ash pellets. The controlled bed agglomeration tests gave the accurate predictions

of possible bed agglomeration problems caused by ash.

In the study of Öhman et al. [70], in-bed behavior of ash forming elements in FBC of

wheat straw, wood, peat, cane trash, grass and bark was examined by SEM/EDS

analyses of samples collected during controlled agglomeration tests. The fuels were

in pellet form having a diameter of 6-8 mm and length of 5-15 mm. Fuels were

combusted in a 5 kW bench scale fluidized bed reactor and compared with bed ash

samples collected from biomass fired full-scale fluidized bed boilers. The

agglomeration test was initialed by loading the bed with a certain ash to bed material

ratio, under normal FBC conditions. Bed temperature was maintained at 760 ºC for

all fuels except for straw. To avoid agglomeration during the ash forming procedure,

650 ºC bed temperature was used for straw. At an ash amount corresponding to 6 wt

% ash in the bed, feeding stopped and operation was switched to external heating.

Several initial runs showed that 1.5 wt % of ash in bed were sufficient for

agglomeration to occur. The onset of agglomeration was determined by pressure and

temperature differences in the bed. The SEM examinations of the bed material

samples showed that a coating with 10-50 µm was formed around the bed particles.

Similar SEM images were observed for 18 MWt CFB boiler samples. From the

elemental analyses of bulk bed samples, silicon was found to dominate in ash

forming elements. Although a major fraction of ash forming elements introduced by

fuel was retained in the bed, they were depleted by some elements such as sulfur and

chlorine. In the case of full-scale boilers, total concentrations of these elements in the

bulk bed were also low. The vaporization during combustion in bench scale unit was

found to increase linearly with increasing bed temperature. In addition, no

vaporization was found during external heating period, even when the temperature

was increased to 960 ºC. Thus, coating characteristics seem to be preserved during

30

this period to determine reliable coating agglomeration temperatures. The results of

SEM/EDS analyses indicated that layers covering the bed particles were

homogeneous, but elemental distributions in the coatings varied significantly

between bed samples of different fuels. Samples collected from firing wood, bark,

cane trash and wheat straw showed that overall compositional distributions of major

fraction of bed particle coatings are mainly limited to ternary system K2O-CaO-SiO2.

The melting behavior was very sensitive to relative amounts of potassium and

calcium in the fuel. Coatings with a relatively high fraction of potassium and lower

fraction of calcium (wheat straw, wood) contained large amount of melt below 900

ºC. On the other hand, coatings with higher calcium and less potassium (bark, cane

trash) did not contain large amount of melt until temperatures well above 900 ºC.

To reduce agglomeration and fouling problems various kinds of mineral additives

can be used for alkali sorption to increase the melting temperature of the system [75].

In a comparative study between kaolinite, bauxite and emalthite, kaolinite was

proved to be the most efficient mineral for alkali sorption [76]. Kaolin addition to

fluidized bed combustion system burning various high alkali and chlorine content

fuels was found to reduce chlorine deposition at typical superheater temperatures of

modern fluidized bed boilers (500 ºC) [42]. The effect of kaolin addition on actual

agglomeration temperature of two troublesome biomass fuels, straw and bark having

high potassium content, and role of kaolin in bed agglomeration were investigated in

the study of Öhman and Nordin [71]. Several initial runs showed that 1.5 % of ash in

bed was sufficient for agglomeration to occur and that 50 % of the added kaolin were

retained in the bed during the experiments. When kaolin, Al2Si2O5(OH)4, was added

to the bed, it was first transformed to metakaolinite, Al2O3.2SiO2, and captured

potassium to form potassium alumina-silicates, K2O.Al2O3.2SiO2, not leaving

sufficiently available potassium oxide to form molten coatings, K2O.CaO.SiO2, on

bed particles. Hence, compositions of the coatings were changed toward higher

melting temperatures mainly because of their decreased potassium content. This

resulted in an increase of initial agglomeration temperature by over 100 °C and 10 °C

for straw and bark, respectively.

31

To prevent agglomeration during combustion of high alkaline fuels different bed

materials other than quartz sand can be used. In the study of Brus et al. [72] iron

blast-furnace slag (BFS) was used as bed material in controlled agglomeration

experiments of bark, olive residue, peat, straw, and reed canary grass. The raw

materials were in pellet form having 6-8 mm size. The initial agglomeration

temperature for quartz sand was about 25 to 60 ºC lower compared to BFS bed

material when bark or olive residue was burned. The formed agglomerates were

found to be more porous and the agglomeration process was extended when BFS bed

material was used. A significant difference in the time until agglomeration was

detected for combustion of reed canary grass at 800 ºC (210 min for sand, 380 min

for BFS), while no significant difference in bed agglomeration was detected for

straw. The bed particles of the quartz sand showed an inner attack layer more often

than those of the BFS. The BFS bed material showed a lower tendency to react with

ash forming elements from the fuel. The outer coating layer had similar thickness

and characteristics for both materials. SEM/EDS analyses of the agglomerates

showed that inner calcium-potassium-silicon attack layer was responsible for

agglomeration process using quartz sand bed material. The composition of

agglomerates formed when using BFS as bed material was similar to outer coating

layer.

Leaching is an efficient way of reducing ash content and agglomeration tendency of

high alkaline content straw ash. Leaching is the treatment of materials with tap water

at room temperature in order to reduce the ash content and change the ash chemistry

to minimize ash melting behavior of the materials [65, 77]. To investigate the

influence of leaching on agglomeration behavior, Arvelakis et al. [19] performed

combustion tests on Danish wheat straw and olive residue (1<L<1.4 mm and L<1

mm) in a laboratory scale bubbling fluidized bed. Fuel was fed to the combustor with

pneumatic transport. In the combustor, limestone with average size varying from 0.5

to 1 mm and density of 2650 kg/m3 was selected as the inert bed material since it

could act as desulphurization agent and have low tendency to react with alkali

metals. At the end of each test, bed material was collected and analyzed with SEM-

EDS technique to define the main inorganic elements forming the agglomerates.

32

Leaching was seen to be insufficient to prevent deposit formation and agglomeration

during fluidized bed combustion of wheat straw. Leaching improved ash thermal

behavior of olive residues securing unproblematic operation of the combustor. No

deposition and agglomeration problem was observed during the tests carried out with

leached olive residues.

Deposition problems such as slagging and fouling are caused by inorganic

constituents of biomass whereas problem with corrosion of superheater tubes is

related to deposition and presence of potassium chloride in the deposits and it is

traditionally avoided by keeping steam temperatures below 450 ºC. Presence of

potassium chloride in the deposits can cause selective chlorine corrosion of

chromium and iron and leaving a nickel skeleton behind. In the study of Nielsen et

al. [73], corrosion measurements were performed in the laboratory to investigate

corrosion of superheater tubes in straw-fired boilers under well defined conditions.

Experiments were conducted to determine the corrosion rate for test metals (ferric

×20 and austenitic AISI 347) which are the commercial superheater alloys. They

were exposed to a synthetic flue gas (6 vol % O2, 12 vol % CO2, 400 ppmv HCl, 60

ppmv SO2, balance N2) in 550 ºC electrically heated ovens for 1 week to 5 months.

The corrosion of the metal was quite uniform and corrosion products were mainly

consisted of iron and chromium oxides. The inner oxide layer contained both iron

and chromium oxides whereas the outer part primarily consisted of iron oxides. On

the top of the oxide layers, a dense and very distinct mixed layer of K2SO4 and iron

oxide (FexOy) was present. Surface of the deposits were covered by K2SO4 and KCl-

K2SO4 mixed layers. The corrosion rate was determined as the thickness of the oxide

layer measured in 3-6 points on each metal sample. Austenitic AISI 347 was shown

to be more resistant toward the corrosion from KCl deposit than ferric×20. A

corrosion mechanism for chlorine corrosion was also suggested in this study which

was based on gaseous chlorine attack where iron and chromium in the metal reacted

with gaseous chlorine and form metal chlorides. High partial pressure of the chlorine

close to the metal was previously believed to be originated from in-deposit sulfation

of KCl and K2SO4. The new part of the corrosion mechanism revealed that KCl

formed a melt with K2SO4 and Fe compounds, and the sulfation was fast in this melt.

33

At low temperatures, solid phase sulfation was slow and metal suffered only from

general oxidation. When metal temperature exceeded the lowest melting temperature

in the KCl/ K2SO4/Fe compounds system, KCl sulfated quickly in the melt,

generating a high partial pressure of Cl2/HCl. This caused accelerated oxidation and

possibly selective corrosion of the metal.

Olive residues are also utilized in fluidized bed combustors. Olive residue is a

specific type of biomass from olive oil production process. Olive residues produced

from pressing of olives and extraction of olive oil, have much lower moisture and oil

contents than the ones produced from centrifuging and then extracting of the oil.

Details and operating conditions of bubbling and circulating fluidized combustors

burning olive residues are summarized in Table 2.5.

Table 2.5: Operating conditions and system properties of olive residue firing FBCs.


Khraisha et al. [78]

Olive cake Bed temperature: 850-900 °C

BFBC 0.15 m inner diameter & 1 m height

N.A.

Topal et al. [79]

Olive cake Bed temperature: 830-870 °C

Excess air ratio: 1.1-2.16 u0: 1.75-2.61 m/s

CFBC 0.125 m

inner diameter &

1.8 m height

Silica sand

Cammarota et al. [80]

Olive husk Bed temperature: 850-900 °C u0: 0.5 m/s

Quartz sand (212-

400 µm)

Scala & Chirone

[81] Olive husk

Bed temperature: 850-900 °C u0: 0.38-0.92 m/s

Excess air ratio: 1.17-2.1

BFBC 0.102 m

inner diameter & 2 m height

Quartz sand (212-

400 µm & 600-850 µm)

N.A.: Not available

Combustion tests of high moisture content olive residue so called olive cake was

fired in the study of Khraisha et al. [78] in a bubbling fluidized bed. The effect of

bed temperature, feeding rate, fluidization velocity and particle size on combustion

34

efficiency and flue gas composition was investigated. Olive cake samples were dried

in oven before feeding into the hopper. The temperature along the bed was fairly

uniform at about 900 ºC which indicated good mixing within the bed. The

combustion efficiency increased with increasing bed temperature due to increased

reaction rate of carbon with oxygen. The combustion efficiency increased with

increasing particle size due to reduction in elutriation losses for larger particles.

When particle size was greater than 600 μm, combustion efficiency stayed almost

constant at about 70 %. The combustion efficiency decreased with increasing fuel

feeding rate and fluidization velocity.

Combustion of olive cake volatiles mainly takes place in the upper regions of the

combustor so that freeboard temperatures are higher with respect to coal combustion.

In the study of Topal et al. [79] combustion characteristics of olive cake was

investigated in a circulating fluidized bed at different excess air ratios changing from

1.1 to 2.16. Feeding rate was 15.5 kg/h and fluidizing velocity was varied in the

range from 1.75 to 2.61 m/s. Freeboard temperatures were measured as 860-900 °C.

CO and hydrocarbon emissions at excess ratio of 1.25 were measured as 3000

mg/Nm3 and 2400 mg/Nm3, respectively. Increase in excess air ratio resulted in a

sharp decrease in both CO and hydrocarbon concentrations. Combustion efficiency

was 98.7 % at excess air ratio of 1.35 (36 %). When excess air ratio was higher than

60 %, CO and hydrocarbon emissions were increased due to insufficient residence

time in bed and incomplete combustion. The optimum value of excess air was given

as 36 %. NOx emission increased slightly with increasing excess air ratio due to lack

of air staging and small residence time in the combustor. Almost zero SO2 emission

was detected during the tests. The ash of the olive cake was mainly composed of

Na2O (50 wt %) and CaO (32 wt %), however no agglomeration of bed material was

noticed during combustion.

Agglomeration mostly occurs during burning olive residues with high potassium

content. The influence of temperature and excess air on agglomeration tendency

during combustion of high potassium content olive husk was investigated by

Cammarota et al. [80] in a stainless steel cylindrical fluidized bed. Fuel particle size

35

was in the range from 20 to 4000 µm. Feeding system consisted of a fuel hopper

mounted over a screw feeder that further delivered powder into a mixing chamber

where a swirled air flow pneumatically conveyed fuel to the bed. Steady combustion

tests were carried out at fluidization velocity of 0.5 m/s. Fuel feeding was started 750

°C. Bed temperature was kept fairly constant at either 850 or 900 °C. Bed pressure

increased with time during the experiment due to accumulation of ash in bed and

absence of a drain flow. Every 15-20 minutes elutriated material collected at the

cyclone was measured and analyzed for carbon concentration. The pressure variance

was used as onset of agglomeration and it was constant at 60 % for both

temperatures. Agglomeration occurred when critical ash content was reached

depending on bed temperature and irrespective of excess air value (10-90 %). Faster

agglomeration occurred with higher temperature and lower excess air. Potassium

enrichment in agglomerates confirmed that sand surface composition reached the

silica-potassium eutectic point which led to extensive melting.

In the study of Scala and Chirone [81], fluidized bed combustion of olive husk in

virgin and exhausted forms were investigated. Olive residue was having high

propensity for unwanted bed agglomeration problems as a consequence of the high

potassium content (26.56 % in exhausted and 41.65 % in virgin olive husk ash). The

influence of temperature, excess air, fluidization velocity and, bed particle size were

investigated. In contrast to other fuels, air-assisted in bed olive husk feeding was

fairly smooth due to high density (1000 kg/m3) and sphericity (0.84) of the particles.

CO, CH4, SO2 and NOx concentrations were measured at the outlet of the combustor

at different excess air ratios. As expected, CO and CH4 concentrations sharply

decreased as the excess air increased. SO2 concentration was relatively high because

the sulfur content of the fuel was considerable. As excess air was increased SO2

concentration decreased due to gas dilution. NOx concentration at the outlet exhibited

a minimum with excess air. Possible explanation of this situation might rely on two

conflicting effects given as tendency of NOx to decrease due to larger gas dilution

and tendency of NOx to increase due to higher oxygen concentration which promoted

nitrogen oxidation. Carbon elutriation rate decreased and combustion efficiency

increased as excess air increased. Total combustion efficiencies higher than 96 %

36

were found for all the experiments, and they were higher than 98 % when excess air

was above 1.5. Faster agglomeration occurred with higher temperature as

agglomeration was enhanced at higher temperatures where low melting point

eutectics at the bed particle surface were easily reached. Lower excess air and higher

fluidization velocity also led to fast agglomeration. When the experiments were

carried out with larger bed particles, defluidization time was almost doubled with

respect to smaller particles. As larger bed particles had more inertia they were

associated with more energetic collisions and hence, adhesion of the particles to form

agglomerates was difficult. Defluidization time for virgin residue was shorter than

the exhausted residue due to its higher alkali content ash. When critical ash content

was reached in the bed, defluidization occurred. Ash amount in bed was strongly

dependent on bed temperature, sand size and husk type.

Sugar cane bagasse was also burned in fluidized beds [52, 53, 61, 69, 70]. Despite

its fibrous nature, low density and high moisture content, it can be fluidized by

mixing with bed material [82]. FBC of sugar cane bagasse was studied by Kuprianov

et al. [83] in a bubbling fluidized bed of 0.9 m inner diameter, 3 m total height and

having conical bed with cone angle of 40°. Since ash content of bagasse was low,

silica sand with particle size of 0.3-0.5 mm was used as bed material to ensure

sustainable ignition and combustion. Bagasse was dried from 48.8 % to 14.4 % by

natural ventilation. Feeding rate of bagasse was 31 and 70 kg/h. The excess air ratio

of 20, 60 and 120 % were tested. In the bed region temperature profiles were found

to be quite uniform and almost independent of excess air. Meanwhile, in freeboard

region temperature profiles were slightly diverged depending on feeding rate and

excess air ratio and simultaneously diminished along the combustor height due to

heat loss across the walls. The effect of combustor load on O2 and CO2

concentrations were negligible. Both NOx and CO axial profiles go through maxima

whose location divided conventionally the combustor volume into the region of the

predominant formation and that of the predominant decomposition of these

pollutants. Higher excess air led to higher NOx emissions. The maximum point of

NOx emission was mainly depending on feeding rate (115 ppm at 70 kg/h and 87

ppm at 31 kg/h) but independent of excess air ratio. The maximum level of CO

37

emission was strongly affected by combustor load as well as excess air ratio. Higher

excess air ratios led to lower CO emissions. Combustion efficiency changed between

96 and 99.7 %. Excess air ratio of 50-60 % was given as the optimum value to secure

high combustion efficiency and low emissions.

Forestry residues such as wood chips, tree barks and pine seed shells are favorable

fuels to be utilized in fluidized beds. Details and operating conditions of fluidized

bed combustion systems burning forestry residues are summarized in Table 2.6.

The ash behaviors of wood chips, peat and coal were investigated in 12 MW

circulating fluidized bed boiler in the study of Skrifvars et al. [84]. Deposit samples

were collected at the cyclone inlet where the flue gas temperature was 850 ºC and

from one location on the convective path where flue gas temperature was 680 ºC

with a sampling time varying from 15 minute to 21 hour. Sampling was carried out

using specially designed surface temperature controlled deposit probes. The probes

were also equipped with a removable ring for later SEM/EDX analyses. The main

ash elements in wood were calcium, potassium and magnesium. No heavy slagging

and fouling was experienced during the tests. Deposits were thin having 0.1-1 mm

thickness. No baghouse samples could be collected during wood combustion tests

since ash amount reaching the bag house was too low. In combustion tests chlorine

was enriched in deposits. In coal and peat firing the main ash components appeared

to be silicates while in wood combustion main components were alkali and alkaline

earth salts. Alkali salts seemed to be present in larger proportions in the deposits

when firing wood than when firing coal or peat, even if the wood itself contained low

amounts chlorine (0.05 wt %) and sulfur (0.03 wt %). The melting behavior or

stickiness of the deposits was evaluated in the temperature range of 500 to 900 °C. It

was assumed that alkali salt fraction was subject to melting and the remainder being

in solid state. In wood combustion, alkali salts appeared in the secondary cyclone

samples and the deposits. In the deposit probe at the convective path, the first

melting point was reached at 564 ºC in the front side and 582 ºC in the back side

deposits.

38

Table 2.6: Operating conditions and system properties of forestry residue firing

FBCs.



Wood chips Peat Coal

Bed temperature: 840-850 °C Feeding rate: 0.50-0.75 kg/s

Excess air ratio: 1.2

Quartz sand

(350 μm)

Lyngfelt & Leckner

[85] Wood chips

Bed temperature: 750-850 °C Total air ratio: 1.2

Combustor load: 50, 70, 100 %

CFBC 12 MW

2.5 m2 cross section &

13.5 m height

Sand

Preto [86] Tree bark

Bed temperature: 780-900 °C Fluidizing velocity: 2.1- 2.4 m/s

Excess air: 20-75 %

CFBC 0.8 MWt

CANMET 0.405 m

inner diameter & 6.7 m height

Sand (210-

500 μm)

Miccio et al. [87]

Pine seed shells

Bed temperature: 850 °C Fluidization velocity:

0.66- 0.97 m/s Excess air ratio: 1.23-1.35

BFBC 200 kWt 0. 37 m inner

diameter & 4.65 m height

Silica sand (300

or 725 μm)

Chirone et al. [88]

Pine seed shells

Bed temperature: 850-900 °C Fluidization velocity:

0.5- 0.7 m/s (bench-scale) 0.7-1.0 m/s (pilot-scale)

Excess air: 30-90 %

BFBC Bench-scale,

0.102 m inner

diameter & 1.625 m height BFBC

Pilot-scale, 0.37 m inner diameter &

4.65 m height

Quartz sand

(300-725 μm)

Tranvik et al. [89]

Wood mixture (40 % sawdust + 25 % bark

+ 25 % forest residue

+ 10 % peat)

Bed temperature: 850 °C CFBC 104 MWt

Quartz sand

(280 μm)

39

The alkali salt part of the deposit was completely molten at 724 ºC on the front side

and 754 ºC on the backside indicating roughly that 20 % of the deposit would be

molten. Fly ash entering the flue gas channel was problematic due to its high alkali

salt content. Although, coal and peat have higher ash forming elements than wood in

their structure, the resulting ash entering the flue gas channel was non-sticky and as a

result it was unlikely to have ash related problems.

Effect of air staging and excess air on NO and CO emissions from fluidized bed

combustion of highly volatile wood chips were investigated in the study of Lyngfelt

and Leckner [85]. In the case of coal combustion NO emissions were known to

depend on both temperature and excess air [90], however, for the case of biomass

combustion; NO emission was found to be insensitive to temperature and decreasing

the excess air had an influence on NO emission which was less pronounced than its

influence on that of coal [91]. By this method NO emissions were reduced, however,

CO emissions were increased due to unburned hydrocarbons. Under normal

conditions in a CFB with a cyclone, CO emission from biomass was lower than that

of coal [92]. As CO was produced from combustion of small char particles in the

cyclone during combustion of coal, this source was less important during combustion

of low char containing biomass. Air staging was also given as another way to reduce

NO emissions. However, conventional air staging in CFB was not effective for NO

reduction in biomass combustion except when secondary air was added as late as the

cyclone outlet for final burnout [92] which was called as reversed air-staging. The

aim of reversed air staging was to lower air ratio at the upper part of the furnace for

N2O reduction, while increasing the air ratio in the lower part for improved sulfur

capture. As for wood combustion sulfur capture was out of interest, the mode of

operation was called as late air staging.

The reference conditions were full-load, 850 ºC bed temperature and normal air

staging, which meant 60 % primary and 40 % secondary air introduced at a height of

2.2 m. Under these conditions CO emission was around 14 ppm and NO emission

was around 104 ppm. Decrease in load resulted in lower emissions. Increasing the

bed temperature from 750 to 850 ºC at full-load had no influence on NO emission,

40

but it had a pronounced reducing effect on CO emission. Reducing total air ratio

from 1.2 to 1.12 resulted in decreased NO emission (27 %) at the expense of high

CO emission. Additional test series that included late air staging at the cyclone exit

were carried out for full and 70 % and half loads. Increase in the flow of air added to

the cyclone exit resulted in an equal reduction in air flow from bottom and the total

air flow was kept constant. As combustor air ratio (air ratio in the riser and cyclone

inlet) was lowered, NO emission decreased and passed through a minimum, while

CO emission increased and went through a maximum.

The maximum reductions in NO emission were 25 % relative to normal air staging.

At the minimum NO emission, CO emission level was about 100 ppm. At 70 % load,

NO emissions decreased by 42 % relative to normal air staging at full-load. At the

minimum NO emission, CO emission level was below 100 ppm. At the half load

maximum NO reduction relative to normal air staging was 65 % and, CO emission

was 750 ppm. At full-load, introduction of secondary air from both cyclone inlet and

outlet resulted in a maximum decrease of 35 % in NO emission relative to normal air

staging and reduced CO emissions somewhat to about 54 ppm. At 70 % load,

introduction of secondary air from both cyclone inlet and outlet resulted in a

maximum decrease of 60 % in NO emission relative to normal air staging and CO

emissions were only few ppm as a consequence of a raised temperature at the exit

chamber. At the half load with introduction of secondary air at both cyclone inlet and

exit, reduction in NO emission was about 64 % but CO emission was increased to

162 ppm due to fall in temperature at the inlet and exit of cyclone. CO emission was

strongly correlated to exit chamber temperature that high temperatures resulted in

lower CO emissions while low combustor air ratios and low temperature resulted in

higher CO emission. At the bed temperature of 850 ºC, the temperature at the

chamber exit was typically 840-850 ºC during full load, 800-820 ºC during 70 % load

and 740-760 ºC during half load.

A wood processing residue, tree bark, was burned in a 0.8 MWt CANMET

circulating fluidized bed combustor by Preto [86]. Large amount of Ca in ash

provided Ca/S molar ratio above 3.8, so there was high intrinsic sulfur capture. Due

41

to high volatile content of the fuel, large part of combustion occurred in gas phase,

either in bed or in freeboard. Combustion efficiency was about 99 %. Bed

temperature was uniform but it increased above the bed. Temperature rise was

depending on several factors but moisture was likely to be an important factor.

Lower moisture content of fuel led to more uniform temperature throughout the

system. Emissions of NOx (100-130 ppm), N2O (<1 ppm) and SO2 (5-20 ppm) were

below the limits [93]. CO levels were greater than the proposed limits due to

fluctuations in the feeding system, variation of the fuel composition and small hang-

ups during the feeding process. Formation of chlorinated hydrocarbons, dioxins and

furans was not an issue for inland wood residues.

Highly volatile fuel particles have greater segregation tendency at the bed surface

during devolatilization. The released volatiles may form bubbles around

devolatilizing particles and lift the coarse particles to the bed surface [94]. So that

significant amount of volatile matter bypasses the bed and burns in freeboard section

and increases temperature in this region. Under-bed or over-bed feeding has

significant influence on segregation characteristics. Miccio et al. [87] worked on this

influence on pine seed shells in a 200 kWt bubbling fluidized bed. Experiments were

performed at different fluidization velocities and excess air ratios. No sorbent was

added to bed as the fuel was sulfur free. CO emission was less than 50 ppm due to

high reactivity of volatile matter and residual char. NO emissions were within the

range of 70 to 140 ppm due to low nitrogen content of the fuel and lower at low

excess air ratios. Carbon elutriation at the cyclones was negligible so that about 99 %

combustion efficiency was achieved. Higher freeboard temperature compared to that

of bed indicated that post combustion of the unburned species occurred in this region

due to insufficient residence time, incomplete mixing of oxygen and combustibles in

bed and segregation of fuel particles in the upper bed zone. Over-bed feeding caused

higher freeboard temperatures compared to under-bed feeding. Difference between

freeboard and bed temperatures was 113 °C at fluidization velocity of 0.97 m/s

during over-bed feeding. Radial profiles of gaseous species in the splashing zone

showed significant differences between under and over-bed feeding. Changes in

measured radial concentration along combustor diameter were related to the

42

segregation of combustible matter. This mechanism was enhanced by adopting

under-bed feeding as a consequence of less uniform spreading of fuel particles along

the combustor.

Formation of particle agglomerates in the bed was the most likely occurred

phenomena during woody biomass combustion. Extensive bed agglomeration

problems were experienced during combustion of these fuels as a consequence of

enrichment of sodium and potassium on the sand particles, in conjunction with high

temperature spots near burning char particles. Agglomeration behavior of pine seed

shells was investigated in the study of Chirone et al. [88] in a bench and a pilot scale

bubbling fluidized bed combustor. The runs were ended when agglomeration of the

bed occurred, as indicated by a jump in the temperature and pressure drop profiles

within the bed. Combustion efficiencies were higher than 96.8 % for the experiments

carried out in bench-scale fluidized bed, and higher than 99.9 % in the pilot-scale

fluidized bed. Both facilities eventually ended with bed defluidization. Upon

defluidization, temperatures in the lower section of the bed tended to decrease, while

those in the upper section of the bed increased due to segregated fuel combustion at

the upper section. Agglomerate formation was enhanced at higher temperatures,

where it was easier to reach low-melting point eutectics at the inert particle surface.

Using larger sand particle size resulted in doubled defluidization times with respect

to the corresponding experiments where the smaller sand was used. This was

considered to be due to higher inertia of larger particles associated with more

energetic collisions making the formation of agglomerates more difficult. Combustor

scale exerted a relevant role that defluidization time was 2.5 times higher in the

larger facility under similar operating conditions due to different bed-fluid dynamics

and presence of horizontal cooling tubes. It was also given that irrespective of the

excess air value and fluidization velocity, well-defined ash content was necessary to

defluidize the bed, depending on the bed temperature and scale of the combustor.

Deposition of bed material during woody biomass firing was investigated in the

study of Tranvik et al. [89] in a 104 MWt wood firing circulating fluidized bed

combustor. Bed material deposits collected from cyclone and riser were analyzed by

43

SEM/EDX. The deposits from bottom ash were also sieved to particle sizes of 125,

355, 500, 710, 1000 and 1400 μm and analyzed with respect to potassium content.

The content of potassium oxide increased with increasing particle size. Bed particle

layers were found in all bed samples. The accumulated material around the bed

particles consisted of two layers. The inner layers were thicker and more

homogeneous; the outer layers were thinner and more particle rich. Inner layer on the

quartz bed particles mainly consisted of silica, calcium, potassium and oxygen. The

composition of the outer layers was more complex and also similar to ash

composition. The samples were crashed mechanically and divided into shell and

kernel. Kernel was mainly concentrated with silica, and most of the other compounds

were enriched in the shell. For prevention of deposition of bed material in cyclones

of wood fired FBCs, the use of non-quartz bed material was suggested.

Co-combustion with coal or addition of chemical materials is known to be practical

and economic treatments to reduce sintering caused by alkaline compounds. The

effect of these treatments is to increase the melting point of biomass ash. In the study

of Llorente et al. [95] limestone was used as the bed material to eliminate tendency

of bed material agglomeration and sintering that normally occurs in plants that

operate with the traditional silica bed material. Combustion tests were performed in a

1 MWt BFBC having 1.1 m inner diameter and 4 m height. Brassica (vegetable oil

residue, bulk density=92 kg/m3), thistle (energy crop, bulk density=220 kg/m3) and

almond shells (agro-industrial biomass, bulk density=400 kg/m3) were fired with

either silica or limestone bed material. Materials were dried before feeding to

combustor. During the tests primary to secondary air ratio was 60/40 and excess air

was 60 %. Carbon conversion efficiency was greater than 98 % in the combustor.

The combustion tests using thistle and almond shell generated agglomerates with

silica sand bed material. Round agglomerates were formed near biomass feeding

point in the thistle combustion test, probably due to high temperature at this point of

silica bed. In the case of almond shells, homogeneous agglomerates formed in a

narrow film on the top of the silica bed, which caused poor fluidization and

shutdown of the combustion test 2 hours after its start-up, as a consequence of higher

residence time of almond shells in this area with respect to other bio-fuels with lower

44

density and higher porosity. Agglomerates were not observed when limestone was

used as the bed material indicating the positive effect of this material to reduce or

even avoid the ash agglomeration. An explanation of the good behavior of limestone

bed material compared with silica bed material was due to the dilution effect which

was caused by adsorption of alkaline salts on the surface of the pores of the material

[75].

As shown in the studies, biomass combustion offers many advantages on reduction

of gaseous emissions. However, it brings many operational problems. To avoid those

operational problems related with high alkali content of biomass, co-firing of these

residues with coal is the most simple and effective alternative. Blending biomass

with coal offers higher combustion efficiencies compared to biomass combustion

only and lower emissions compared to coal combustion only. Moreover, it provides

controllable disposal of residue. Following section reviews the studies on co-firing of

various biomass and coal types.

2.3.2 Biomass and Coal Co-firing Studies

The agricultural residues mostly co-fired with coals in fluidized bed combustors are

rice husk, straw, olive residue, and forestry residues are woodchip, bark and sawdust.

Rice husks can be co-fired with coal to reduce the gaseous emissions from coal

combustion. The influence of co-firing of coal with rice husk (0-15 wt %) and

woodchips (0-25 wt %) on N2O and NOx emissions was investigated in a bubbling

fluidized bed in the study of Shen et al. [96]. The results revealed that co-firing

decreased N2O and NOx emissions. As wood chips and rice husks have high volatile

matter, their addition to the fuel blend resulted in larger release of volatiles in the

lower part of the bed. These volatiles consumed most of the oxygen and form a lean

zone in this area to restrict N2O and NOx formation. Increasing the rice husk share in

the fuel blend from 5 to 15 % reduced NOx and N2O emissions from 180 to 145 ppm

and 150 to 100 ppm, respectively. Reduction rate of N2O emission decreased with

45

increasing biomass share. Also increasing the temperature resulted in higher NOx and

lower N2O emissions.

There are various studies on co-firing of straw with coal. Details of the straw and

coal co-firing in fluidized bed combustion systems and their operating conditions are

summarized in Table 2.7.

Straw has low melting point ash and firing straw alone in fluidized beds results in

unscheduled shutdowns of combustors due to agglomeration of bed material and

fouling of heat transfer surfaces. Co-firing coal and straw reduces the fouling and

agglomeration tendency because of the higher melting point of coal ash and its

ability to react with potassium [97].

Agglomeration is a direct result of stickiness of bed material. Tendency of

agglomeration during co-firing of straw and coal was investigated by Lin and Dam-

Johansen [98] in laboratory scale fluidized bed. Quartz sand was used as bed

material. The main parameter reflecting the agglomeration tendency was

defluidization time. The influence of temperature and coal fraction on defluidization

time was investigated. Defluidization was indicated by a sudden decrease of pressure

drop over the bed to a low level. Defluidization time was extended when straw was

co-fired with coal. Combustion temperature had the most pronounced effect on

defluidization time. The influence of coal fraction was found to be insignificant at

high temperatures, but it became very sensitive at low temperatures. Defluidization

time increased sharply with decreasing temperature when temperature was below a

critical value. This critical temperature was 875 °C at coal fraction of 20-30 wt % in

the fuel blend. It was about 100 ºC higher than the melting temperature of ash from

straw firing only.

To predict the melting behavior of ash components, co-firing experiments were

carried out with coal and straw (50 wt %) in a 30 kWt bubbling fluidized bed in the

study of Sandelin et al. [99]. Mixing the fuels did not lead to a mean value of the fuel

properties or a linear relationship for the behavior of the fuels.

46

Table 2.7: Operating conditions and system properties of straw and coal co-firing

FBCs.

Reference Fuel Blend Operating conditions System Bed material

Lin & Dam-

Johansen [98]

Wheat straw pellets

(70-80 wt %) & Polish Coal

Bed temperature: 870-945 °C Under-bed feeding by screw

feeder

BFBC Lab scale

0.068 m inner diameter &

1.135 m height

Quartz sand

(275-460 μm)

Sandelin et al. [99]

Straw (50 wt %)

& Coal


feeder

BFBC 30 kWt

0.108 m inner diameter

& 3 m height

Quartz sand

(300-600 μm)

Lin et al. [97]

Coal (36 %) Woodchips

(40 %) Straw (24 %).

Bed temperature: 825-870 °C MgO was used as additive in

lab scale tests.

CFBC 20 MWt

3.4 m2 cross section & 15 m

height, CFBC

80 MWt BFBC

Lab scale

Quartz sand, fuel

ash & limestone

Glazer et al. [100]

Straw pellets (20, 50 % on energy basis)

& coal


feeder

CFBC 25 kWt

0.080 m inner diameter & 5 m

height

Silica sand

(300-600 μm)

Kassman et al. [101]

Straw pellets (10, 20, 30 %

on energy basis) &

bituminous coal

Bed temperature: 850-865 °C Primary/total air ratio : 44 %

U0 : 4.8 m/s

CFBC 12 MW

2.25 m2 cross section &

13.6 m height

Started with silica

sand, then

gradually replaced by coal ash and

lime

Okasha [102]

Rice straw (50, 66% ) &

bitumen - CaO pellets

Over-bed feeding Bed temperature: 750-875 °C

in pilot scale BFBC, Bed temperature: 800 °C &

U0 : 3 m/s in lab scale BFBC

BFBC Lab scale


0.8 m height BFBC

Pilot scale 0.3 m inner diameter &

3.3 m height

Silica sand

(250-500 μm)

47

Significant amount of energy was released by coal through char combustion. In

contrast, high and significant amount of energy was released from straw during

combustion of pyrolysis gases. Consequently, burning behavior of the fuel mixture

influenced the behavior of the char residue, where also the greatest amount of ash

forming elements was found. No agglomeration was observed in the bed. The spent

bed material was completely depleted of chlorine and sulfur. Significant amounts of

potassium, calcium, sodium and aluminum were captured in bed ash.

Agglomeration, deposition, corrosion and emission behavior of co-firing straw (0-60

%, thermal basis) with coal was studied by Lin et al. [97] in full scale 20 MWt

circulating fluidized bed boiler, 80 MWt circulating fluidized bed boiler and a

laboratory scale bubbling fluidized bed combustor.

The 20 MWt boiler was a demonstration plant so called multi-circulating fluidized

bed boiler. The unique feature of the boiler was the pre-separator which was

integrated with the particle cooler embedded as a superheater and an evaporator. The

separated particles prevented the heat transfer surface from direct contact with flue

gas and potentially avoided corrosion. Low fluidization velocity in the coolers

minimized the erosion risk. The tests in this boiler were focused on the local

atmosphere (oxidizing and reducing) and alkali concentration in the gas phase in the

combustor and pre-separator. Fuel blend was composed of 36 % coal, 40 % wood

chips and 24 % straw on thermal basis. The alkaline measurements in gas phase

showed that potassium concentration in the combustor section varied between 0.16

and 2.3 ppm (v/v). In the pre-separator, the potassium vapor concentration was in the

range from 0.06 to 3.7 ppm. The combustor exit had the highest potassium

concentration, while, in the pre-separator part, potassium concentration is the highest

at the top. The high potassium concentration in the top part of the pre-separator was

due to high char concentration in this region, resulting in more release of potassium

during char combustion.

In 80 MWt CFBC [103], co-firing straw with coal resulted in considerably higher

corrosion rates compared to coal firing only. Variation in boiler load had a large

48

impact on operating conditions of the boiler, as well as the chemical reactions in the

combustor, loop-seal and convective pass. When the boiler load decreased,

temperature in all parts of the boiler decreased and the condition in the loop seal

shifted from predominantly reducing to more oxidizing. Decrease in particle re-

circulation rate led to lower elutriation rate of char particles to loop-seal, and more

oxidizing conditions there. Deposition tests in the loop-seal showed that the deposits

were only formed at the full load during co-firing of coal and straw. It appeared that

half of the deposit was potassium chloride. Small amount of char transported to the

loop-seal was believed to be the primary source of potassium chloride. Complete

combustion of the char with a small amount of aeration in the loop-seal led to high

concentration of potassium chloride, forming deposits.

In the same study, laboratory scale experiments were carried out with Polish coal and

wheat straw pellets. A mixture of MgO particles and sand was used as the bed

material (66.7 wt % MgO, 33.4 % sand). Agglomeration tendency of SiO2-MgO was

lower than that of quartz sand, but higher than pure MgO, suggesting silica

containing compounds were responsible for agglomeration tendency of bed material.

Using MgO as an additive reduced the agglomeration tendency, but did not eliminate

it. Fine particles were collected in the cyclone and large agglomerates were found in

the bed material. This meant that agglomeration and attrition/abrasion occurred

simultaneously in the bed material. Significant decrease in SO2 emission was

detected at straw share of 60 % on thermal basis. Potassium in straw was capable of

SO2 capture. Using MgO as additive resulted in higher NO emissions.

Chlorine behaves as a shuttle for potassium transportation to the particle surface

before release as potassium chloride to the gas phase. Influence of operating

conditions and fuel composition on the release of alkali compounds to the gas phase

during combustion and co-combustion of high alkali straws with low sulfur content

coal was investigated by Glazer et al. [100] in a 25 kWt circulating fluidized bed.

The amount of total gas-phase sodium and potassium compounds in the flues gases

were measured with excimer laser induced fluorescence. The results showed that the

release of gaseous species was depended on the fuel composition, K/Cl and K/Si

49

ratios in the fuel. A synergic effect of co-combustion of straw with coal led to a

strong decrease in gaseous alkali concentrations. The part of the alkali metals

released from the straw to the gas phase interacted with clay minerals in the coal to

form alkali–alumina-silicates. Potassium and sodium concentrations were much

lower in co-firing tests compared to straw firing tests. Small additions of straw to

coal led to dramatic increase in gaseous alkali content in the flue gas.

Alkali related problems in straw and coal co-firing was investigated in the study of

Kassman et al. [101] in 12 MW circulating fluidized bed boiler. Straw pellets were

co-fired with Polish bituminous coal. The tests were focused on variation of straw to

coal in combination with different feeding rates of limestone. Chlorine content of

straw and coal were 0.36 % and 0.32 % on dry and ash free basis, respectively. Gas

phase alkali chlorides were measured by means of an in-situ alkali chloride monitor

(IACM). SEM/EDX was used for analyses of collected deposits on the steel deposit

rings. Alkali chlorides existed in the gas phase when the flue gas temperature was

higher than 650 °C. IACM instrument measured the sum of sodium and potassium

chloride concentrations on-line. In the first part of the experiments coal was used as

fuel with increasing lime supply. Potassium chloride level was generally low and

independent of SO2 concentrations. HCl level was high and decreased somewhat by

increasing lime supply. In the second part, straw and coal co-firing experiment was

carried out with increasing lime supply. Alkali supply was equal to a fraction of

straw of 17-19 % on energy basis. Addition of straw pellets increased potassium

chloride concentration with approximately one order of magnitude (from ~2 to ~20

ppm). In the third part, straw fraction in the fuel blend was increased with a constant

lime supply. The lime supply was equal to a Ca/S molar ratio of 6.2-6.9. Potassium

chloride concentration increased with increasing straw share. At straw share of 29 %

on thermal basis, about 50 ppm potassium chloride concentration was measured. This

level was believed to be much too high with respect to potential problems fouling

and super heater corrosion [104]. No clear relationship was seen between increased

lime supply and potassium chloride.

50

Deposits were collected on steel rings during a period of 4 hours using a temperature

controllable probe held at 500 °C. The deposits were analyzed for chlorine, sodium,

potassium, alumina, silica, sulfur and calcium by EDX. In the test series where straw

was co-fired at a constant ratio with increasing lime supply, chlorine was present in

all the deposits. Excess supply of lime led to capture of chlorine due to formation of

calcium chloride in the deposits. In the series carried out with constant lime supply

and increasing straw share, significant amounts of potassium and chlorine were

found in the deposits. Increasing the straw ratio in the fuel blend increased chlorine

concentration in the deposits drastically.

Co-firing reduces SO2 emissions due to low sulfur content of biomass and intrinsic

sorbent capability of biomass ash. In the study of Okasha [102], enhancement of

sulfur retention of rice straw-bitumen pellets by integrating CaO within the pellets

was investigated either by a batch operation in a laboratory scale bubbling fluidized

bed or a continuous operation in a pilot scale bubbling fluidized bed. A series of

batch combustion tests were carried out with feeding a single pellet at different molar

ratios of built-in Ca/S. SO2 concentration was measured throughout the burning time

of entire pellet. SO2 concentration decreased greatly with increasing Ca/S ratio in

the pellet. Sulfur retention was less efficient during devolatilization stage as volatiles

intensively released and stayed a very short time within the pellet. On the other hand,

sulfur retention was very efficient during char combustion due to relatively lower

rate of combustion in this stage. SO2 emission was close to zero at char burning stage

at Ca/S molar ratio of 0.75. Sulfur retention efficiency increased with increasing the

share of straw in the pellet. Increasing the Ca/S ratio within the pellet increased NOx

emissions slightly. This was attributed to the catalytic effect of CaO promoting NOx

formation [105, 106]. Increasing the bed temperature intensified the rate of volatile

release and reduced sulfur retention. In continuous operation sulfur retention was

observed at steady state. At built-in Ca/S ratio of 0.75, SO2 emission was reduced to

65 ppm which corresponded to 96 % retention efficiency. Evidently, the retention

efficiency under continuous operation was considerably higher than that of batch

tests considering the same Ca/S ratio.

51

Olive residues have also been co-fired with coals in FBC units. Details of the olive

residue and coal co-firing FBC systems and their operating conditions are

summarized in Table 2.8.

Table 2.8: Operating conditions and system properties of olive residue

and coal co-firing FBCs.


Armesto et al. [107]

Alpeorujo (10, 20, 25 wt %)

& Puertollano

lignite (<1, <5 mm)

Bed temperature : 830, 850, 870 °C

Fluidization velocity: 0.7-1 m/s Under/over bed feeding

by screw feeder

N.A.


Foot cake (10, 15, 20, 25 wt %, dry basis)

& lignite/ anthracite

Bed temperature : 830, 850, 870 °C

Fluidization velocity: 0.7-1 m/s Under bed feeding by screw

feeder

BFBC 0.1 MWt


& 3 m height

N.A.

Cliffe & Patumsawad

[109]

Olive oil waste

(0, 10, 20 wt %) & coal

Bed temperature: 830 to 940 °C Excess air: 50, 90 % Pneumatic feeding

BFBC 10 kWt


& 2.3 m height

Sand (850 μm)

Atımtay & Topal [110]

Olive cake (25, 50, 75 wt %) & Tunçbilek

lignite

Bed temperature: 850-865 °C Excess air ratio: 1.1-2.16

u0: 1.76-2.55 m/s

CFBC 0.125 m inner diameter & 1.8

m height

Silica sand

(560 μm)

N.A.: Not available

Alpeorujo is a specific type of residue from olive oil production having high

moisture (50-70 %), volatile matter and alkaline content. Difficulties such as bed

agglomeration, slagging on furnace walls and fouling of heat transfer surfaces arise

during combustion of Alpeorujo due to high potassium and sodium content in its ash.

There were some experiences on burning alpeorujo although continuous failure of

the bed during fluidization had determined severe constraints for commercial

operation. The feasibility of co-combustion of Alpeorujo with lignite was

investigated in the study of Armesto et al. [107] in CIEMAT 0.1 MWt atmospheric

bubbling fluidized bed combustor. Operating parameters in the co-firing tests were

52

furnace temperature, fluidization velocity, biomass share, coal particle size and

position of feeding system. Higher combustion efficiencies were obtained when fuel

is fed from the bottom. Increase in biomass share did not show significant effect on

combustion efficiency when fuel was fed from the bottom. CO2 emissions were

changing within the range of 12.6-13.6 % during feeding from bottom. The

placement of feeding had strong influence on CO emission. CO emissions were

about 3000 mg/Nm3 when fed from the top and reduced to about 200 mg/Nm3 when

fed from the bottom. The effect of alpeorujo share in the fuel mixture on CO

emission was very little. Decrease in particle size led to increase in CO emission.

Contrary to the pervious studies, increasing biomass share resulted in higher SO2

emissions. At fluidizing velocity of 0.7 m/s, increasing biomass share from 10 to 20

% led to increase in SO2 emissions from 1018 to 1211 mg/Nm3 and it stayed constant

through 25 % share. However, at 1 m/s fluidizing air, SO2 emission continued to

increase to 1646 mg/Nm3 with higher share of alpeorujo. This behavior was

explained according to the ash characteristics. The ashes obtained during the tests

with minor SO2 emissions had major content of unburned fuel and the presence of

sulfur (as pyritic sulfur) in these ashes were significant. The NOx emissions

decreased when biomass content in the mixture increased. This was attributed to the

lower nitrogen content of the fuel blend and its higher volatile matter content.

Increasing the biomass share from 10 to 25 wt % reduced NOx emission from 274 to

219 mg/Nm3 at 0.7 m/s fluidizing velocity. At 25 wt % biomass share, increasing the

furnace temperature from 850 to 880 °C resulted in higher NOx emissions. When fuel

blend is fed from top of the boiler, higher NOx emission was detected. During the test

runs cyclone, baghouse and bed material samples were collected for chemical

analysis. The main components were SiO2 and Al2O3 which were the main ash

components in the Puertollano coal ash. Chlorine was found to concentrate in

baghouse ash. Potassium components were concentrated in cyclone ash due to the

low temperature of the combustor at this zone, 350 °C, which was below the

condensation temperature of majority of potassium compounds. The content of

potassium in cyclone ash increased and in the baghouse ash decreased with

increasing alpeorujo share in the fuel mixture.

53

Another investigation on feasibility of co-firing olive residue with coal was carried

out by Armesto et al. [108] with a specific type of olive oil residue so called foot

cake which was rich in moisture (66.4 %) and volatile matter (25 %). Foot cake was

co-fired with both lignite and anthracite. The effect of coal type, biomass share,

operating temperature and fluidizing gas velocity on combustion efficiency and flue

gas concentrations of O2, CO, CO2, SO2, NO and N2O were investigated in the

bubbling fluidized bed pilot plant same as given in the previous study. HCl in the

flue gas was measured discontinuously. The combustion efficiency of foot cake-

lignite blends was obtained to be higher than the efficiency of foot cake-anthracite

blends. This was attributed to the higher volatile matter content of lignite compared

to anthracite. In the case of foot cake-anthracite blends, there was a decline in

freeboard temperature causing high levels of unburned carbon. The combustion of

volatiles was completed in the bed but the particles did not appear to have long

enough residence time for combustion. In the case of foot cake-lignite blends, the

decline in freeboard temperatures was less. Increasing the operating temperature

from 800 to 850 °C resulted in higher conversion of fuel-N to NOx from 30 to 50 %

Higher temperatures led to higher combustion rates and radical concentrations [111].

On the other hand, fuel-N to N2O conversion slightly decreased with increasing

temperature. The effect of biomass share in the mixture did not show significant

influence on combustion efficiency. SO2 emissions from combustion of foot cake-

lignite blends were significantly higher than the SO2 emissions from combustion of

foot cake-anthracite blends due to the difference in sulfur contents of fuel blends.

Increasing the biomass share from 10 to 20 wt % reduced SO2 emissions from 1050

to 850 mg/Nm3 for lignite blend and from 500 to 400 mg/Nm3 for anthracite blend.

The major NOx emissions were obtained during the tests using anthracite-foot cake

blends due to lower volatile matter content of anthracite. On the contrary, the major

N2O emissions were on the tests with lignite-foot cake blends. This result was

opposite to what was expected since N2O conversion was known to increase with

decreasing volatile matter. This could be resulted from reduced flame temperature

caused by high moisture content of foot cake. HCl emissions were close to 1

mg/Nm3. This was attributed to the chlorine retention by calcium in the ash.

54

Similar to foot cake, olive oil waste having high moisture content (60 %) was co-

fired with coal in the study of Cliffe and Patumsawad [109]. The experiments were

carried out in a 10 kWt bubbling fluidized bed. Operating temperature was changed

in the range from 830 to 940 °C. Excess air was changed between 50 and 90 %. Fuel

was fed pneumatically to the bed surface from a sealed hopper through an inclined

feeding pipe and the flow rate was controlled by a screw feeder. Biomass share

greater than 20 wt % caused bed temperature to drop so that combustion could not be

sustained. TG and DTG burning profiles showed that maximum burning rate of olive

oil waste was about 450 °C [112]. The combustion efficiencies of fuel blends having

0, 10 and 20 wt % of olive oil waste were ranging from 91 to 95 %, 85 to 93 %, and

88 to 93 %, respectively. The combustion efficiency of 20 wt % olive oil waste

mixed with coal was slightly higher than that of 10 % olive oil waste mixed with

coal. This was explained by the fact that when the moisture content of the fuel

increased, devolatilization time was extended due to delayed gas evolution from the

fuel. Therefore, fuel mixture had more time in bed to burn. Increased concentration

of olive oil waste in the fuel mixture led to less carryover. Significant combustion

took place in freeboard. Freeboard temperatures increased by addition of 10 wt %

olive oil waste to the coal. CO emissions were changing between 100 and 350 ppm at

6 % O2 in flue gas. CO emissions increased with increasing excess air.

Co-firing of olive residue so called olive cake with lignite was carried out by Atımtay

and Topal [110] in a laboratory scale circulating fluidized bed. Due to high Na2O (50

wt %) content of the olive cake ash, potential operation problems such as fouling of

heating surfaces and agglomeration of the bed material was expected during

operations. The influence of excess air ratio on combustion characteristics and

emissions was investigated in a wide range from 1.1 to 2.16. Biomass share during

the combustion tests were 25, 50 and 75 wt %. Silica sand and ash were used as bed

materials. The densities of olive cake, coal and sand were given as 591, 1374 and

1730 kg/m3, respectively. Superficial gas velocity was changing between 1.76 and

2.55 m/s. Olive cake addition to the coal firing system at all proportions led to an

increase in combustion efficiency up to a certain excess air ratio and further increase

in excess air led to lower combustion efficiencies. Combustion shifted to the upper

55

sections of the combustor. At 25 wt % olive cake co-firing, about 97 % combustion

was obtained. SO2 emission was about 2600 mg/Nm3 and did not change with

increasing excess air ratio. However, at the excess air ratio of 1.50, a sharp decrease

was observed in CO and hydrocarbon emissions from 3100 to 169 mg/Nm3 and from

1960 to 36 mg/Nm3, respectively. NOx emission was almost constant at 220-260

mg/Nm3. During 50 wt % olive cake co-firing, 98 % combustion efficiency was

achieved at the excess ratio of 1.5. SO2 emission was almost constant at around 1600

mg/Nm3. Similarly up to excess air ratio of 1.5, sharp decrease in CO and

hydrocarbon emissions were observed from 4290 to 360 mg/Nm3 and from 2240 to

62 mg/Nm3, respectively. At 75 wt % olive cake co-firing, 98.2 % combustion was

achieved at the excess air ratio of 1.6. SO2 emission was about 1250 mg/Nm3. Until

the excess air ratio of 1.6, sharp decrease in CO and hydrocarbon emissions were

observed from 5800 to 150 mg/Nm3 and from 2750 to 120 mg/Nm3, respectively.

The optimum excess air ratio was given as 1.51-1.60.

Apart from agricultural residues, forestry residues are co-fired in fluidized beds.

Details of the forestry residue and coal co-firing FBC systems and their operating

conditions are summarized in Table 2.9.

To demonstrate the technical feasibility of fluidized bed as a clean technology for the

combustion of low grade fuels and coal/forestry residues Armesto et al. [113]

performed co-combustion experiments in either circulating or bubbling fluidized bed.

A forestry waste, pine chips, high ash content refuse coal and a low grade, high

sulfur content lignite were used in the experiments. Limestone was injected to the

systems for sulfur retention. In circulating fluidized bed, combustion efficiencies

were changing between 90 and 100 %. Refuse coal/biomass blend gave lower

efficiency. As Ca/S ratio was increased, sulfur retention rate increased. At 50 %

biomass share and using refuse coal, increasing the Ca/S ratio from 1.5 to 3,

decreased SO2 emission from 2298 to 898 mg/Nm3 and CO emission from 659 to

380 mg/Nm3. NO emission stayed almost constant at around 300 mg/Nm3.

56

Table 2.9: Operating conditions and system properties of forestry residue

and coal co-firing FBCs.



Pine chips (0-80 %, 0-100 %) & Refuse

coal

Bed temperature: 830-860 °C &

800-850 °C u0: 5-6.2 m/s &

0.6-0.7 m/s

CFBC 0.2 m inner

diameter & 6.5 m height BFBC 1 MWt

1.14 m inner diameter & 4.6 m total height

N.A.

Adanez et al. [114]

Pine bark (0-100 %) &

Lignite/South African coal

Bed temperature: 800-900 °C u0: 4-6 m/s

Excess air: 15-25 % Secondary air: 10-35 %

CFBC 0.3 MWt


6.5 m height

N.A.

Gayan et al. [115],

Pine bark & Sub-

bituminous coal/Spain

Lignite

VTT Bed temperature: 850 °C

Excess air: 18-25 %, Secondary air: 50 %

Biomass share: 50, 60 wt % CIEMAT

Bed temperature: 800-900 °C Excess air: 10-30 %,

Secondary air: 10-35 % Fluidization velocity: 4-6 m/s Biomass share: 0-100 wt %

CFBC VTT: 0.1 MWt

0.2 m inner diameter & 6.5

m height CFBC

CIEMAT: 0.3 MWt

0.17 m inner diameter & 8 m

height

VTT Sand

(100-300 μm)

CIEMAT Sand

(300-500 μm)

Gani & Naruse [116]

Sawdust (50 wt %) & low

grade coal Furnace temperature: 800 ºC

Electrically heated drop-tube furnace

N.A.

Leckner & Karlsson

[117]

Wood chips/ sawdust

(0-100 %) & bituminous

coal

Bed temperature: 850 °C Fluidization velocity: 7.7 m/s

CFBC 12 MW

1.7×1.7 m2 cross section & 13.5 m height

Silica sand & coal ash mixture

Kakaras et al. [118]

Waste wood (0-50 %) &

lignite

Under bed pneumatic feeding Excess air ratio: 1.3-2.1

BFBC 0.095 m inner

diameter & 1.2 m height

Sand (CaCO3)

(500-1000 μm)

Yrjas et al. [119]

Forest residues,

wood chips (0-100 wt %)

& coal

N.A. CFBC 550 MWt

N.A.

Orjala et al. [120]

Bark, sawdust, peat & coal N.A.

CFBC 150 MWt

CFBC 50 kWt

N.A.

N.A.: Not available

57

In bubbling fluidized bed, combustion efficiencies were in the range of 96.9 to 99.5

%. In co-firing refuse coal/biomass blends with 50 % biomass share between Ca/S

ratio of 3.7 and 6.7, no significant change in sulfur retention rate was observed. At 58

% biomass share and using refuse coal, increasing the Ca/S ratio from 3.7 to 6.7

decreased SO2 emissions from 568 to 485 mg/Nm3 and increased CO emission from

199 to 353 mg/Nm3.

Circulating fluidized bed process was found to be a little more efficient than

bubbling fluidized bed process due to higher combustion efficiency and more

efficient utilization of limestone. Increasing the proportion of biomass improved the

efficiency and environmental impact.

Combustion efficiency of pine barks during co-firing with two different kinds of coal

(South African (SA) coal and high sulfur content lignite) was investigated by Adanez

et al. [114] in 0.3 MWt circulating fluidized bed. The effect of operating conditions

such as biomass share (0-100 %), temperature (800-900 °C), excess air (15-25 %),

air velocity (4-6 m/s) and percentage of secondary air (10-35 %) were studied. For

both coals carbon combustion efficiency is increased with increasing biomass share.

Increasing pine bark share from 0 to 100 % increased combustion efficiency of SA-

pine bark and lignite-pine bark blends from 96 to 99.5 % and 98.2 to 99.6 %,

respectively. Higher reactivity of lignite/pine bark blends resulted in higher

combustion efficiencies with respect to South African coal/pine bark blends.

Temperature rise from 800 to 900 ºC resulted in increased combustion efficiencies

for both blends. The solid circulation flow rate increased when gas velocity increased

so that the flow rate of solid losses by the cyclone increased. This acted on the mean

residence time of char particles in bed which decreased the combustion efficiency for

both blends. On the other hand, increasing excess air increased the mean oxygen

concentration in the bed, thus increased the carbon combustion efficiency. Also,

increasing the percentage of secondary air generated a reducing zone in the lower

part of the combustor and therefore reduced the combustion rate.

58

Similar results were obtained in the co-combustion experiments performed with pine

bark and two different coals in 0.1 MWt VTT and 0.3 MWt CIEMAT circulating

fluidized bed combustion pilot plants. Gayan et al. [115] co-fired pine barks with

South African sub-bituminous coal and high sulfur Spanish lignite. In CIEMAT

secondary air was introduced 1.5 m above the distributor plate. Pine bark and coal

were fed to the boiler simultaneously from two hoppers mounted on a balance

through a screw feeder. In VTT pilot plant, secondary air with 50 % share was

introduced at 2 m height. Fuels were fed from separate hoppers and mixed in a screw

feeder. Moisture content of pine barks were decreased from 37 to 11 % during

storage and grinding. The particle size of pine barks were less than 3 mm in VTT

pilot plant, however, in CIEMAT combustor it ranged to 30 mm. The influence of

different operating conditions was investigated in CIEMAT combustor. They were

biomass share, combustor temperature, fluidization velocity, excess air and

secondary air ratio and size distribution of the feed. In VTT pilot plant the influence

of biomass share (0-100 wt %) on combustion efficiency was studied by burning sub-

bituminous coal. In CIEMAT combustor (850 ºC bed temperature, 25 % excess air,

24 % secondary air and 5 m/s gas velocity) increase in biomass share resulted in

increased combustion efficiency of pine bark-lignite and pine bark-sub-bituminous

coal blends from 97.5 to 99 % and from 96 to 99 %, respectively.

In VTT facility (850 ºC bed temperature, 30 % excess air, 40 % secondary air and

2.3 m/s gas velocity), increase in biomass share of pine bark-sub-bituminous coal

blend resulted in increased combustion efficiency from 99.3 to 99.8 %. Increasing

the bed temperature led to increase in combustion efficiency and decreased carbon

concentration in bottom region due higher reaction rates. Also higher efficiencies and

lower char concentrations were obtained for lignite blends due to higher reactivity of

this coal compared to South African coal. Increase in excess air led to higher mean

oxygen concentration in bed which positively affected combustion efficiency. On the

other hand, increase in the percentage of secondary air produced lower oxygen

concentration in bed therefore decreased the combustion efficiency.

59

In general co-firing is expected to lower SO2 and NOx emissions due to lower sulfur

and nitrogen content of the fuel blend; however especially in woody biomass

combustion increasing biomass share results in higher NOx emissions. In the sawdust

and low grade coal co-firing experiments performed by Gani and Naruse [116], NO

and N2O concentrations were found to be higher than coal combustion only even

when the fuel blend had much lower nitrogen content than coal. Such results were

also reported by Leckner and Karlsson [117] who studied NOx emissions from

fluidized bed combustion of high volatile wood chips/sawdust and bituminous coal

mixtures (0-100 wt %) in a 12 MWt CFB boiler. The extreme cases with 100 %

wood or 100 % coal showed that NO emission from wood combustion is higher than

that of coal despite higher nitrogen content of coal relative to wood. A small addition

of coal to wood however, yielded higher emissions of NO than wood combustion

only. Increasing the coal fraction led to higher amount of char in the bed which had

reducing effect on NO emission. This was explained by a much higher release of NO

from coal combustion than wood, caused by high nitrogen content of coal. For

woody biomass, NO rapidly formed in bed section and remained constant along the

combustor, however during coal combustion, formed NO in bed was reduced along

the combustor by reduction reactions of high amount of coal char [121]. When small

fraction of coal was added, reduction in NO was small. Increasing the coal fraction

resulted in higher amount char in the bed which has reducing effect on NO emission.

The NO emissions were measured within the range of 50 to 100 ppm (6 % O2). The

N2O emission mostly originated from the coal. Wood addition had a slight reducing

effect on N2O emission. CO emission was found to increase steadily with increasing

the coal fraction in the blend. SO2 emissions were increased from 20 to 400 ppm

proportional to addition of coal since the sulfur content of wood is negligible.

Agglomeration, slagging and fouling problems occur especially during high

potassium and sodium content woody biomass combustion. Biomass ash aggravates

agglomeration by reacting with silica bed material and forming low melting point

compounds which brings about defluidization. Before performing experiments some

theoretical study can be done to investigate ash deposition tendency. To predict

agglomeration tendency of low ash, high combustible content uncontaminated wood,

60

railway sleepers and demolition wood before co-firing with low quality Ptomais

lignite in a laboratory scale bubbling fluidized bed reactor, Kakaras et al. [118] used

base to acid ratio to estimate fouling and slagging tendency of the fuels. If base to

acid ratio is greater than 0.7, the fuel was considered to have low fouling potential.

Base to acid ratio of lignite, uncontaminated wood and railway sleepers were 1.4, 3.9

and 6.3, respectively. However; the alkaline metals sodium and potassium can form

combinations with low fusibility temperatures altering the base to acid influences.

Biomass addition to the system improved the ignition and due to higher combustible

content and heating value, steady state was achieved faster. Increasing waste wood

share in the fuel blend decreased CO emission in the system. In uncontaminated

wood co-firing experiments, increasing biomass share from 0 to 50 % decreased CO

emission from 1000 to 300 ppm in excess air range of 1.3-1.6. Higher excess air

ratios (1.7-2.1) led to higher CO emissions changing from 1300 to 700 ppm by

increasing uncontaminated wood share in the fuel mixture from 0 to 50 %. In all co-

combustion experiments, SO2 emission was reduced due to low sulfur content of the

waste wood and also greater sulfur retention was observed in 20 % thermal input of

demolition wood in the blend due to high CaO content in its ash. Furthermore, low

ash and sulfur content of waste wood contributed to minimization of ash

agglomerates formation during co-combustion of low quality lignite. Slight decrease

in NOx emission during co-combustion was mainly attributed to lower nitrogen

content of waste wood. N2O emission was also reduced from 200 to 125 ppm when

increasing waste wood share from 0 to 50 % in the fuel blend.

The dominating elements in bottom and fly ash samples were calcium, iron and

manganese in the ash samples. Bottom ashes had higher calcium concentration due to

lower volatility of this element [122] and calcite sand used as inert bed material.

High concentration of iron was attributed to presence of iron in lignite. The most

remarkable difference among various fuel blends was the high alkaline metal

concentration in the mixtures with higher wood waste contribution and mainly the

potassium concentration, when uncontaminated wood was added in 50 %. Waste

wood addition in blend up to 30 % thermal input was found to be feasible. Higher

wood share in fuel blend resulted in agglomeration problems.

61

Chlorine deposition was investigated in the study of Yrjas et al. [119] in 550 MWt

circulating fluidized bed boiler co-firing bark, wood chips and forest residue (0-100

wt %) with coal. The short-term deposits were collected on the surface of air cooled

probes with detachable rings from two locations where the flue gas temperatures

were 730 and 530 ºC. Advanced fuel analysis method was carried out to obtain

reactive sodium, potassium and calcium amounts which could react with sulfur to

form sulfates instead of reacting with chlorine to form alkali chlorides. This method

was based on selective leaching by water, ammonium acetate and hydrochloric acid.

Ash forming elements leached out with water and ammonium acetate represented the

more reactive species, while the rest and the leached out with the acid represented the

less reactive species. It was stated that sulfur released from the fuel blend reacted

with the sodium and potassium to form alkali sulfates and this prevented alkali

chloride formation and thus fouling. Chlorine left the system with flue gas in the

form of HCl instead of depositing. Lower amounts sulfur and higher amounts

reactive calcium compounds of biomass fuels resulted in lower SO2 emissions.

Therefore, increasing the biomass share indirectly affected the chlorine deposition

risk. However, all the sodium, potassium and calcium in the fuel blend were not

reactive towards sulfur. The rates of deposit build-up were below 12 g/m2h, indicated

no severe deposit problems.

Similar to the previous study, Orjala et al. [120] stated that highly reactive alkali

metals and chlorine in biomass ash reacted with sulfur and alumina silicates and

chlorine was mostly released as HCl. Even a small amount of chlorine in biomass

could result in potassium and sodium chloride deposition on boiler heat transfer

surfaces. To avoid such problems, co-firing was given as an effective way of

diminishing the corroding tendency of biomass ash since coal brought within itself

protective elements to the combustor. By co-firing, even low sulfur containing coal

could significantly reduce the corrosion risk in the combustor [123].

62

63

CHAPTER 3

EXPERIMENTAL SET-UP, PROCEDURE

AND CONDITIONS

3.1 General Experiments were carried out in the Middle East Technical University 0.3 MWt

Atmospheric Bubbling Fluidized Bed Combustion (ABFBC) Test Rig. Test rig was

originally constructed and operated for the investigation of combustion and in-situ

desulfurization characteristics of low quality Turkish lignites. Therefore, the existing

test rig was modified for co-firing of biomass and coal. Modified test rig is described

in detail in section 3.2 and pre-experimental modifications and work carried out

before the experiments are given in section 3.3. The modifications were carried out

within the scope of a research project, MAG 104M200, financed by the Scientific

and Technical Research Council of Turkey (TÜBİTAK).

3.2 METU 0.3 MWt ABFBC Test Rig The test rig in its present form is displayed in Figure 3.1 whereas its flow sheet is

given in Figure 3.2. As can be seen from the flow sheet, the test rig basically consists

of the following sub-systems:

• air and flue gas system

• modular combustor

• fuel and limestone feeding systems

• ash removal systems

64

Figure 3.1: METU 0.3 MWt ABFBC Test Rig

65

Figu

re 3

.2: F

low

shee

t of M

ETU

0.3

MW

t AB

FBC

Tes

t Rig

66

• cooling water system

• instrumentation and control system

3.2.1 The Combustor

The main body of the test rig is the modular combustor formed by five modules of

equal dimensions. Modular structure of the combustor is intended to provide

flexibility in burning various fuels by addition or removal of heating surfaces. Each

module has an internal cross-section of 0.45 × 0.45 m2 and 1 m height. Inner walls of

each module are refractory lined with firebricks with a thickness of 6 cm. Outer walls

of the refractory bricks are insulated with insulation bricks with thickness of 20 cm.

Further insulation is provided by leaving an air gap of 6 mm between the outer wall

of insulation brick and the inner wall of the steel construction of each module. The

first and fifth modules from the bottom are referred as bed and cooler, respectively,

and the ones in between are referred as freeboard modules.

The bed module provides an expanded bed height of 1 m. It contains 6 water-cooled

U-tubes (25 mm OD, stainless steel) providing 0.35 m2 of cooling surface, 5 ports for

thermocouples, 4 ports for gas sampling probes, one port for LPG distributor and two

ports for feeding fuel/limestone mixtures. One of the feeding ports is 22 cm and the

other is 85 cm above the distributor plate.

In the freeboard and cooler modules, there are 6 ports for gas sampling probes and 9

ports for thermocouples. There exists a water-cooled tube bundle consisting of 11

tubes (26.7 mm OD, carbon steel) with 14 passes installed across the cross-section of

the cooler module providing 4.3 m2 cooling surface in the cooler module for cooling

the stack gases before leaving the combustor.

67

3.2.2 Air and Flue Gas System

Fluidizing air is supplied by a forced draft (FD) fan. Fluidizing air fed by the FD fan

enters the bottom of the windbox through a pipe of 6.5 m long and 7.8 cm ID on

which a manual gate valve, an automatic butterfly valve and a vortex flow meter are

installed. The design of the wind box allows the installation of bed ash removal

system as shown in Figure 3.1. It is a mobile wind box supported by four wheels and

a distributor plate is placed on the top. Air supplied to the wind box by means of the

pipe of 7.8 cm ID diverges to the full cross-section of the combustor at the distributor

plate located 1.4 m above the entrance port. Sieve type distributor plate contains 412

holes, each 4.5 mm in diameter, arranged in a triangular pattern. Within the bed

module air mixes with fuels and limestone for effective combustion and sulfur

capture.

Flue gases and elutriated fines leaving the bed surface enter the freeboard. Sufficient

freeboard height is provided to permit burnout of elutriated lignite fines and

combustible gases.

After leaving the freeboard, flue gases pass through the cooler module to cool the hot

combustion gases. Flue gases leaving the modular combustor enter the cyclone and

then the baghouse filter to leave the elutriated particles before passing through

induced draft (ID) fan to exit from the stack. As the temperature of the flue gases

entering the baghouse filter is limited by the maximum operating temperature of the

bag material which is 260 °C, two alternative systems are provided for the safe

operation of the baghouse filter: A bypass line between the cyclone and the ID fan

and an air dilution system to reduce the flue gas temperature at the inlet to the bag

filter through a slide valve if the temperature exceeds the upper operating limit of the

bag material.

The pipes carrying the flue gases before and after the baghouse filter are 14.0 and 5.3

m long, respectively, and have an ID of 15.3 cm. There is also a bypass pipeline

68

between the cyclone and the ID fan. It has an ID of 12.8 cm and length of 14.5 m.

The outlet of the baghouse filter joins this pipeline 4.2 m before the ID fan.

An orifice plate with a bore diameter of 8.05 cm was installed at the stack gas line

before ID fan to measure the flow rate of the flue gases. The pressure drop across the

orifice plate is measured by means of pressure transmitter. Knowing the temperature

and pressure of the flue gases passing through the orifice plate, the signal from the

transmitter is utilized in the control system to yield molar flow rate. After the orifice

plate, flue gas passes through an automatic butterfly valve and then enters to ID fan.

Exit of the ID fan is connected to the stack having dimensions of 20 cm ID and 16 m

height.

3.2.3 Solids Handling System

Lignite, biomass and limestone are stored in three separate silos and conveyed into

the hoppers of water-cooled screw feeders at controlled flow rates via pre-calibrated

volumetric feeders placed under their respective silos. The lignite/biomass/limestone

mixture can be continuously fed to the bed through water-cooled screw feeders either

22 cm or 85 cm above the distributor plate. Both screw feeders are operated at

controlled speed in such a way that there is no accumulation of feed material in the

hopper. In order to prevent backflow of combustion gases from the combustor water-

cooled screw feeders have gas tight connections.

Bed ash is withdrawn from the bed through 5 cm diameter, 1.1 m long water-cooled

ash removal pipe. Some of the bed ash is disposed and the rest is stored to provide

bed inventory when required. Bed ash drain rate is adjusted from the DCS to obtain

the desired bed pressure drop and hence the expanded bed height. Bed ash particles

are collected in a continuously weighted ash storage bin.

The majority of the elutriable fines produced from solids in the bed and those fed

within the solid streams are captured by the cyclone, having dimensions of 45 cm

OD and 2.12 m height. Cyclone catch particles pass through an air lock (i.e. a rotary

69

valve) and fall onto a diverter. Depending on the position of the diverter, particles are

either discharged from the system to a continuously weighted ash storage bin for

experiments without recycle or flow back to the combustor for re-firing. During co-

firing experiments, the elutriated particles collected by the cyclone were directly

discharged from the system, i.e., recycle was not carried out.

In order to catch fine particles of fly ash (dp ≤ 40 μm) leaving the cyclone, jet-pulse

type baghouse filter with a 100 % collection efficiency for particles greater than 1

μm was utilized. Technical specifications of the baghouse filter are summarized in

Table 3.1.

Table 3.1: Technical specifications of the baghouse filter.

Bag material P84-Polyimide

Weight of the bag material 500 g/m2 (± 5 %)

Thickness of the bag material 1.6 mm (± 0.2 mm)

Maximum operating temperature of the bag material 260 °C

Collectable particle size ≥ 1 μm

Bag diameter 0.16 m

Bag height 2.0 m

Collection area of one bag 1.0 m2

Total number of bags 20

Air to cloth ratio (max.) 0.9 m/min

Maximum pressure drop through the filter 150 mm H2O

Ash collection hopper angle 70°

Pulse interval 17 s

Pulse cycle 68 s

Pulse air pressure 6 barg

During the operation, a filter cake is built up at the outer surface of the bags which in

turn becomes a principal collection medium. As the filter cake gets thicker with time,

70

a pulse of compressed air is directed into the bag from the open top which causes a

shock wave to travel down its length dislodging the filter cake from the outer surface

of the bag. A unique aspect of the pulse jet system is the use of a wire cage in each

bag to keep it from collapsing during normal filtration. The bag hangs from the tube

sheet. A series of parallel pulse jet pipes are located above the bags with each pipe

row having a solenoid valve. This allows the bags to be pulsed clean one row of five

bags at a time. Filter cake cleaned off the surface fall into a hopper and is discharged

to fly ash collecting container. There are two fly ash containers each having a volume

of 0.13 m3. During filtration of flue gases if one container gets full, the maximum

level device located just above the ash discharge opening gives alarm by lighting the

level warning light located on control panel, and the container full of ash is replaced

with the other one after closing the ash discharge opening by leak proof slide valve.

3.2.4 Cooling Water System

Cooling water required for the test rig is passed through a magnetic conditioner and

is then divided into two streams, one for the in-bed tube bundles, and the other for

the tube bundle in the cooler module. Heat transfer areas provided by the bed and

cooler modules are 0.30 m2 and 4.3 m2, respectively. The cooling water in bed enters

lower header and leaves the bed through the upper header. The cooling water for the

cooler module enters the upper header and flows downward to provide counter-

current flow to the up flowing flue gases. Water flow rates are adjusted by means of

either a manual or a pneumatic control valve located at the drain of each stream to

maintain maximum exit temperature of about 60 °C.

3.2.5 Gas Sampling and Analysis System

3.2.5.1 Gas Sampling Probe

In order to acquire spatially resolved gas composition data from the combustor, the

combustion gas is extracted from the symmetry axis of the combustor by using gas

sampling probes which are fabricated for in-situ extractive gas sampling.

71

The details of the probe construction are shown in Figure 3.3. The probe is water-

cooled to ensure structural strength at the high temperatures encountered in the test

rig. The body of the probe consists of 4 coaxially oriented stainless steel tubes with

the outermost one being the cooling-shroud and the innermost one being the suction

tube. The two remaining tubes are for guiding the flow of cooling water in the probe.

Cooling water enters from the shroud from the rear of the probe, travels to the tip in

between shroud and the adjacent tube, turns back over the 2nd innermost tube, and

discharged. The suction tube is heated by means of a variable DC power supply and

is isolated from the adjacent one with teflon pipe. A ceramic filter was located at the

tip of the probe for in-situ filtering of the particulates. The filter itself is held in place

by a stainless steel perforated plate and a fixing nut. Once through the ceramic filter,

the dust-free combustion gas travels through the suction tube which is maintained at

130 °C by the combined effects of electrical heating and the water cooling. Relative

positions of gas probes on the combustor are given in Table 3.2.

Table 3.2: Relative positions of gas sampling probes.

Probe No Distance above the distributor plate, cm

P10 26

P9 56

P8 69

P7 85

P6 123

P5 183

P4 291

P3 344

P2 419

P1 500

72

Figu

re 3

.3: G

as sa

mpl

ing

prob

e

73

3.2.5.2 Gas Sampling and Conditioning System

Once through the probe, the sampled combustion gas is passed through a solenoid

valve and sent to the gas conditioning and analysis system of the test rig by means of

sample line. The sample line itself is maintained at 130 °C by means of a variable

DC power supply so that no water, sulfuric acid or hydrocarbons would condense

along the sampling interface. In addition, all lines and fittings in contact with the gas

sample are made of stainless steel to prevent interferences due to gas adsorption or

heterogeneous reactions.

Figure 3.4: GASS-II Pre-conditioning System

Gas is sampled at a rate which is small enough to cause minimal interference to the

combustion system. After passing through the probe, sample gas is transported

through the heated stainless steel line to a sample pre-conditioning system (GASS-II)

to remove particulate, mist and water vapor from the gas stream. The picture of the

pre-conditioning system is shown in Figure 3.4. The First step in conditioning is

filtering out particulates and aerosols by passing the sample through 1 µm filter

having borosilicate glass filter element with a fluorocarbon binder. Collected liquid

mists are removed from the system periodically by an automatic drain filter.

74

After filtering, sample passes through ammonia scrubber which avoids deposition of

ammonia salts if ammonia is present in the sample stream. Sample stream then

passes through Nafion membrane dryer. As the sample enters the dryer, flow splits

into a number of small diameter Nafion membrane tubes arranged in a parallel

bundle. The membrane selectively removes water vapor by a process of permeation

distillation. Water vapor travels through the tubing walls driven by the difference in

partial water vapor pressure on the opposing sides of the membrane. As the sample

flows from inlet to outlet of the dryer, water is continuously removed. There is a

countercurrent flow of dry purge gas in to the dryer to provide a medium for water

vapor to be carried away. The schematic of the Perma Pure Nafion membrane tubes

are shown in Figure 3.5.

Figure 3.5: Schematic of Perma Pure Nafion Membrane Dryer

After the drier, sample gas is passed through the cooler and particulate filters for

removal of submicron sized particulates then pumped to the analyzers via a teflon-

coated diaphragm-type sample pump for species concentration measurements. After

the measurement of species concentrations, sample gas is vented to the atmosphere.

3.2.5.3 Analytical System

The on-line continuous gas analyzers with which the test rig is equipped are listed in

Table 3.3. Analyzers except Bailey SMA 90 are used for measuring concentrations

of species O2, CO, CO2, NO, N2O and SO2 along the combustor and also at the

75

cyclone exit on dry basis. Bailey SMA 90 measures temporal variation of O2 and CO

on wet basis at the combustor exit.

There is also an alternative route for sampled gas in case of a fault in the analytical

system. Components of this backup system are given in Table 3.4. Details of the gas

conditioning and analysis system are shown in Figure 3.6.

Table 3.3: Gas analyzers.

Instrument Species Sensor type Range

ABB Advanced Optima Magnos 106 O2 Paramagnetic 0-10/0-25 % by vol.

ABB Advanced Optima Uras 14

CO CO2 NO N2O

NDIR

0-5 % by vol. 0-20 % by vol. 0-1000/0-2000 ppm 0-500/0-1000 ppm

Siemens Ultramat 6 SO2 NDIR 0-1 % by vol.

Bailey SMA 90 O2 CO

Zirconium oxide Catalytic RTD

0-25 % by vol. 0-2 % by vol.

Table 3.4: Backup gas analyzers.

Instrument Species Sensor type Range

Leeds & Northrup O2 Paramagnetic 0-15 % by vol.

Anarad AR 600 CO CO2

IR 0-5 % by vol. 0-20 % by vol.

Servomex 1491 NO/NOx Chemiluminescence 0-0.2 % by vol.

76

Figu

re 3

.6: G

as c

ondi

tioni

ng a

nd a

naly

sis s

yste

m

77

3.2.6 Deposit Sampling System

A deposit sampling probe is designed and constructed for the investigation of ash

fouling during biomass co-firing runs. Deposit sampling probe is placed at the top of

the freeboard for simulating the conditions in the superheater region of a

commercial-scale boiler where flue gas temperature is about 850-880 ºC. The surface

temperature of the probe is measured using K-type thermocouple embedded into the

outer surface of the probe near the detachable ring in order to control surface

temperature of the removable ring. A constant probe surface temperature of 500 °C is

maintained during the experiment by adjusting the flow rate of the cooling air.

A removable stainless steel deposit ring having dimensions of 15 mm OD and 20

mm length is attached to deposit probe for sampling the ash deposits in this region.

Deposit samples are collected on the surface of air-cooled deposition probe during

each biomass co-firing run. Schematic description of deposit sampling probe is given

in Figure 3.7.

3.2.7 Instrumentation and Control System

Figure 3.8 shows the process and instrumentation (P&I) diagram of the test rig. As

can be seen from the figure the test rig is extensively instrumented for research

purposes. Instrumentation and analytical systems can be divided into following

categories:

• Data acquisition and control system

• Solid flow control and monitoring

• Air and gas flow control and monitoring

• Cooling-water flow control and monitoring

• On-line continuous gas analyzers

• Pressure sensors

• Temperature sensors

• Solids analyses

78

Figu

re 3

.7: D

epos

it sa

mpl

ing

prob

e

79

Fi

gure

3.8

: P&

I dia

gram

of M

ETU

0.3

MW

t AB

FBC

Tes

t Rig

80

The test rig is equipped with a data acquisition and control system namely Bailey

INFI 90. Real time process data is monitored, manipulated, collected and analyzed

with the aid of control software called Bailey LAN-90 Process Control View

installed on an IBM compatible PC 486 computer running under QNX operating

system. The control system scans the signals coming from all of the instruments

attached to it in a fraction of a second and reports and logs their averages discretely

for 30 seconds of intervals. An uninterruptible power supply is connected to Bailey

INFI 90 and PC in order to enable proper shut-down in case of a electricity cut-off by

preventing corruption of data logged.

Fuel and sorbent feed rates are controlled manually by adjusting the fuel feeder or

sorbent feeder control dial from the computer. The flow rates of fuel and sorbent are

normally set to such values that provide desired excess air and Ca/S molar ratio,

respectively. Cyclone ash and bed ash are collected in respective bins and their flow

rates are followed by load cells placed under respective bins.

The volumetric flow rate of air is measured by a vortex flow meter and adjusted with

an automatic butterfly valve driven by a computer controlled pneumatic actuator. In

order to achieve conversion from volumetric to molar flow, a static pressure tap and

a temperature sensor is placed downstream of the vortex flow meter. The flow rate of

air is normally set to a value to achieve the desired superficial velocity in the

combustor. In order to achieve almost neutral pressure on the bed surface, the flow

rate of exhaust gases is adjusted with an automatic butterfly valve driven by a

computer controlled pneumatic actuator.

In order to measure flow rates of cooling-water flowing through bed and cooler

bundles, two orifice plates are located up streams of their lower and upper headers,

respectively. The pressure drops across the orifice plates are measured by means of

pressure transmitters. The signals from the transmitters are interpreted in the control

system to yield mass flow rate of the cooling-water flowing through in-bed and

cooler bundles. There exist two pneumatic control valves installed on the

downstream of upper and lower headers of bed and cooler bundles, respectively, to

81

adjust the cooling-water flow in each bundle. The flow rates of cooling-water in bed

and cooler bundles are normally set to a value which provides exit water temperature

in the range 40-60 °C.

Pressure sensors are used for measuring differential and gauge pressures at various

positions on the test rig. Measured differential pressures are the pressure drops over

orifice plates, bed and distributor plate pressure drop, and gauge pressures are the

pressure at the bed surface and pressure of air feed at the downstream of the vortex

flow meter.

Table 3.5: Relative positions of thermocouples.

Thermocouple No Distance above the distributor plate, cm

TC1 25

TC2 44

TC3 73

TC4 73

TC5 97

TC6 133

TC7 154

TC8 226

TC9 257

TC10 285

TC11 330

TC12 361

TC13 425

TC14 500

Spatial and temporal variations of gas temperatures along the height of the

combustor are measured by means of thermocouples of K type (Chromel-Alumel)

82

with grounded junction. The tips of the thermocouples are on the symmetry axis of

the combustor. The axial positions of thermocouples are given in Table 3.5. The

temperature of air feed at the downstream of vortex flow meter and temperatures of

cooling water at the exits of bed and cooler bundles are measured by resistance

thermocouples of type Pt-100.

3.3 Pre-Experimental Modifications 3.3.1 Feeding System

Biomass feeding was problematic with the available feeding system which was

originally designed for coal feeding. The feeding system of the test rig utilizes

volumetric feeders, i.e. screw feeders to control the fuel and sorbent flow rates. As

biomass fuels have much lower bulk density compared to that of coal, their flow

rates are much higher compared to those of coals. Moreover, biomass fuels have

higher particle sizes (up to 25 mm) compared to that of coals (up to 10 mm). To

mitigate the feeding problems, new bed screw feeders were designed and constructed

by taking into account the available data. Details of bed screw feeder design

calculations are given in Appendix I.

As test rig was originally designed for coal firing only, there is no additional silo for

biomass storage. Therefore, it was decided to use the existing ash silo for biomass

storage during the experiments. Upon this solution, existing ash silo and its screw

feeder was tested for suitability to biomass feeding. Three biomasses, i.e. olive

residues, hazelnut shell and cotton residue were filled into the ash silo and maximum

and minimum achievable flow rates were measured. It was observed that existing

screw feeder cannot provide the minimum biomass flow rate required during the

experiments for all biomasses and that during hazelnut shell and cotton residue

feeding tests, bridging in the silo occurred causing the stoppage of biomass flow into

the screw feeder. To alleviate the minimum flow rate problem existing electric motor

with its gear drive was replaced to decrease its rotation rate from 140 rpm to 14 rpm.

To sort out the bridging problem of silo, inner surface of the existing silo was

83

covered by Teflon layer and silo was hanged up by chains and a vibrator was added.

This new arrangement provided easy flow of biomasses, especially for hazelnut shell

and cotton residue.

3.3.2 Air and Flue Gas System

Instrumentation air used in the test rig is supplied from the compressor located in the

Unit Operations Laboratory of Chemical Engineering Department. During the

experimental preparations, it was observed that compressed air contains too much

rust particles, moisture and oil which cannot be eliminated by the filters.

Investigations showed that existing compressor requires extensive maintenance and

repairs. Therefore a new air compressor was purchased, installed and connected to

the test rig. Technical specifications of the compressor are given in Table 3.6.

Table 3.6: Technical specifications of new air compressor.

Brand Lupamat LKD 61-555

Maximum pressure 8 barg

Operating pressure 6 barg

Tank volume 500 L

Free air flow 1454 L/min

Number of cylinders 3

Motor Power 7.5 kW

The stack of the existing test rig was too short (~ 3 m) and leading to pollution

around the Unit Operations Laboratory during the start-up as the baghouse filter can

not be taken into operation till the flue gas temperature reaches a value at which no

condensation takes place inside the flue gas lines. Therefore, existing stack of the test

rig was removed and a taller stack was installed. The new stack has dimensions of 20

cm ID and 16 m height.

84

During the experimental preparations, it was observed that ID fan has too much noise

and vibration leading to vibration of whole flue gas line. Upon inspection and

various measurements, the manufacturer found that vibration is caused by the

unbalanced fan blades. Manufacturer made the balance adjustment of fan blades and

the vibration problem was sorted out.

3.3.3 Gas Sampling and Analysis System

Perma Pure drier used in the gas conditioning system of the existing unit was found

to fulfill its service life. Therefore, a new gas conditioning system, namely Perma

Pure GASS-II Pre Conditioning System, was bought directly from the USA and

incorporated in the existing system and commissioned. The new gas conditioning

system is fully automatic and also has its own dry air generator.

A new gas analyzer, ABB Advanced Optima, was bought for on-line measurement of

O2, CO, CO2, NO, N2O on dry basis and incorporated into the existing gas analysis

system. As there is no place in the existing gas analyzer cabinet, a new gas analyzer

cabinet was also purchased and necessary connections to the existing gas analyzer

system were made. Gas analysis system is arranged in such a way that both old and

new analyzers can be used for gas analysis.

3.3.4 Instrumentation and Control System

During the experimental preparations, it was observed that after a certain time no

value is shown on the computer connected to the DCS. At first, cable connections

were checked. However, investigations revealed that one of the cards of the DCS,

i.e., Bailey Infi 90 INBIM02 Bus Interface Module, had a problem. As the

manufacturer of the DCS system, Bailey Controls, is not available now as it was

acquired by the ABB, a used card was supplied from a plant in Turkey. It was

installed and the problem was alleviated.

85

As there was no N2O measurement in the original gas analysis system of the test rig,

a new 4-20 mA signal connection to DCS was made from the new ABB Advanced

Optima gas analyzer. In addition, DCS software was modified to incorporate N2O

measurement.

3.4 Operating Procedures 3.4.1 Procedures before Cold Start-Up

Each experiment requires extensive preparation work which can be summarized as

follows:

1. Preparation of limestone (crushing and sieving)

2. Filling of coal, limestone and ash silos.

3. Calibration of screw feeders of coal, biomass and limestone.

4. Checking the calibration of orifice meters.

5. Checking and cleaning of the gas sampling probes.

6. Checking the operation of the gas sampling and conditioning system.

7. Checking and cleaning of the thermocouples.

8. Checking and cleaning of the initial heating system.

9. Checking the bags of baghouse filter for leakages.

10. Assembly of combustor and windbox.

11. Checking the test rig for air leakages.

12. Calibration of gas analyzers.

13. Checking the operation of DCS system

3.4.2 Cold Start-Up

In order to be able to heat up the combustor during cold start-up, a perforated pipe

LPG burner is provided which is located 15 cm above the distributor plate. LPG is

provided by three LPG tubes with 45 kg capacity each stored outside the laboratory.

The line delivering the LPG to the burner is equipped with a needle valve to adjust

86

the flow rate of LPG and a solenoid valve to activate or deactivate the burner. In

order to speed up the heating of the bed charcoal is also provided at controlled flow

rate.

The normal procedure for start-up from cold conditions is as follows:

1. Power on the airlock under the cyclone.

2. Establish a fluidization air flow of about 0.4 m/s (corresponding to 12

kmol/h or 1.5 m/s at 850 °C) and an induced draft at a minimum level at

cold conditions.

3. Establish cooling water flows to bed and cooler bundles, gas sampling

probes, screw feeders and bed ash removal pipe.

4. Open bypass line for the stack gases not to pass through baghouse filter.

5. Establish the start-up LPG burner.

6. About 20 minutes after LPG burning, charge the combustor with bed

material from a previous run corresponding to about 20 cm H2O pressure

drop in the bed.

7. Establish charcoal flow to help the heating of the bed. Initially, feed

charcoal intermittently and wait till temperature in the bed reaches to 700

°C.

8. Commence feeding coal at a rate of approximately 50 kg/h, meanwhile

follow the stack gas O2/CO analyzer and bed temperatures to check the

start of coal combustion.

9. If temperatures do not increase or O2 concentration at the stack does not

decrease, stop coal feeding and continue with preheating the bed and

observe an evidence for ignition of coal fed.

10. If coal is ignited, start feeding ash and coal, and stop LPG and charcoal.

11. Close bypass valve and take baghouse filter into operation.

12. Increase air flow rate to have superficial velocity of approximately 2 m/s.

13. When the bed depth reaches its expected (or planned) steady-state value,

as indicated by bed pressure drop, stop ash feeding and commence

removing bed material to maintain constant bed depth.

87

14. Adjust the air and coal feed rates to produce the desired steady-state

operating conditions.

3.4.3 Procedure during Runs

Owing to the adequate instrumentation of the test rig, full data collection is achieved

with little effort during the runs. The following parameters are logged by means of

the control and data acquisition system to the PC from the beginning of the start-up

procedure:

The temporal variations of fluidizing air and stack gas flow rates, their temperatures

and static pressures, superficial air velocity at average bed temperature, pressure drop

through the distributor plate and that through the bed, pressure at bed exit, oxygen

and carbon monoxide concentrations at the exit of the combustor, rate of coal and

limestone feed and bed ash drain, weights of the carryover and bed ash bins, bed and

freeboard cooling-water flow rates and their inlet and exit temperatures, and

temporal variation of gas temperatures through the combustor. Spatial variation of

gaseous species concentrations are started to be logged at steady state which is

reached in about 5 h after the start-up. The steady state is deduced by following the

variation of gas temperatures along the combustor and O2/CO concentrations at the

exit of the combustor with respect to time.

The only manually recorded parameter is the biomass flow rate which is adjusted

through a digital motor frequency controller with the aid of a calibration curve.

3.4.4 Shutdown

By the end of gas sampling from the combustor, dry air is passed through sampling

lines and analyzers for a few hours so that no sulfuric acid or hydrocarbons would

condense. Stopping of the fuel feeding shuts down the operation of test rig. Air flow

rate is decreased not to lose part of bed material by elutriation. Water flow rates of

both bed and freeboard coolers are decreased and the bed material is discharged

88

rapidly and stored in a bin. The carryover bin is removed and an empty bin is loaded

over the load cell. Water and air cooling continues for about 2 hours and this is

followed by switching off the fans and letting the system to cool down for another 24

hours.

3.4.5 Post Shutdown

Once the system is cooled completely the wind box is detached and bed material on

the distributor plate is collected and weighted. Representative samples of coal,

biomass, limestone, bottom ash, cyclone ash and baghouse filter ash taken during

runs are analyzed for size distribution and chemical analyses.

3.5 Experimental Conditions

3.5.1 Lignite, Biomass and Limestone Characteristics

Experiments were carried out with Çan lignite. Biomasses utilized in the experiments

were supplied from the local markets and were transported to the laboratory in bags

of 40 kg. Representative samples from coals and biomasses (Figure 3.9) were then

subjected to sieve analyses and proximate and ultimate analyses. Size distributions of

fuels, limestone and bottom ash samples were obtained by sieve analyses method

whereas size distributions of cyclone and baghouse filter ashes were obtained by

Malvern Mastersizer 2000 particle size analyzer. Proximate analyses were performed

by using LECO TGA-701. Ultimate analyses were carried out with LECO CHNS-

932. Calorific values of the fuels are measured by using AC-500 bomb calorimeter.

The results of these analyses together with the calorific values and bulk densities are

summarized in Tables 3.7 and 3.8, respectively. As can be seen from these tables,

lignite is characterized by high ash content (∼25 %) and high total sulfur contents (∼4

%). Ash constituents of lignites and biomasses are shown in Table 3.9. With regard

to ash composition, lignite ash is mainly composed of acidic oxides whereas olive

residue ash is mainly composed of basic oxides.

89

Olive Residue

Hazelnut Shell

Cotton Residue

Figure 3.9: Photographs of biomasses.

90

Cot

ton

Res

idue

6.93

5.38

75.5

7

12.1

4

364

46.7

9

6.48

4.40

36.2

3

0.32

5.78

0.32

17.4

Haz

elnu

t Sh

ell

7.62

1.46

73.0

4

17.8

9

320

49.7

7

5.86

0.56

42.1

5

0.08

1.58

0.11

17.5

Oliv

e R

esid

ue

6.07

4.24

75.6

9

14.0

0

591

50.2

2

6.38

1.72

37.0

3

0.14

4.51

0.14

18.1

Run

10

Lign

ite

17.4

7

24.2

9

31.4

4

26.8

0

905

40.0

4

3.84

0.98

21.4

2

4.29

29.4

3

4.35

13.4

Run

9

Lign

ite

17.0

5

27.0

6

30.9

7

24.9

3

905

39.8

7

3.89

0.88

18.2

8

4.46

32.6

2

4.46

12.5

Run

8

Lign

ite

17.1

4

27.4

6

30.3

6

25.0

4

905

41.9

2

4.01

0.96

15.6

4

4.33

33.1

4

4.33

12.5

Run

7

Lign

ite

16.0

5

24.3

6

32.1

7

27.4

2

905

42.2

2

4.23

1.01

19.3

5

4.17

29.0

2

4.17

13.9

Run

6

Lign

ite

17.1

9

25.2

9

31.2

2

26.3

0

905

46.4

7

4.26

1.15

13.9

5

3.63

30.5

4

3.63

13.4

Run

5

Lign

ite

16.7

5

23.8

9

32.0

4

27.3

3

905

44.8

3

4.00

1.20

17.6

6

3.61

28.7

0

3.66

14.0

Run

4

Lign

ite

16.6

0

22.6

8

32.5

2

28.2

0

905

43.8

1

3.86

1.16

20.6

4

3.34

27.1

9

3.92

14.6

Run

3

Lign

ite

16.9

8

24.5

6

31.5

8

26.8

8

905

45.4

2

4.37

1.11

15.8

5

3.67

29.5

8

3.67

13.6

Run

2

Lign

ite

16.4

8

26.7

4

31.0

5

25.7

4

905

44.9

3

4.09

1.14

13.9

6

3.86

32.0

2

4.07

13.3

Run

1

Lign

ite

16.3

5

28.7

8

29.7

9

25.1

7

905

44.6

0

3.95

1.09

11.9

7

3.98

34.4

1

4.17

12.3

Tab

le3.

7: F

uel a

naly

ses.

Prox

imat

e A

naly

sis

(As r

ecei

ved

basi

s)

Moi

stur

e, %

Ash

, %

Vol

atile

mat

ter,

%

Fixe

d ca

rbon

, %

ρ Bul

k, k

g/m

3

Ulti

mat

e A

naly

sis

(Dry

bas

is)

C, %

H, %

N, %

O, %

(by

diff

eren

ce)

S C

ombu

stib

le, %

Ash

, %

S To

tal,

%

LH

V, M

J/kg

91

Cot

ton

Res

idue

0.00

0

1.66

8

6.35

1

31.6

47

12.5

23

7.59

6

14.5

61

7.89

6

7.90

3

2.98

6

5.49

5

1.37

4

Haz

elnu

t Sh

ell

0.00

0

9.04

0

64.4

54

13.8

74

5.84

1

5.15

7

0.63

3

0.30

7

0.18

9

0.05

5

0.45

2

Oliv

e R

esid

ue

0.00

0

0.12

8

11.9

66

21.5

10

20.8

40

13.5

99

9.31

4

9.32

7

6.17

7

7.13

9

Run

10

Lign

ite

0.00

0

1.09

1

3.78

8

5.15

5

6.15

2

16.3

08

13.3

40

20.9

97

11.4

08

6.18

5

6.44

1

3.71

6

5.42

0

Run

9

Lign

ite

0.00

0

0.88

0

7.18

6

6.39

9

5.93

5

13.7

84

10.6

86

18.0

22

11.0

99

6.58

0

7.28

0

4.63

5

7.51

5

Run

8

Lign

ite

0.00

0

1.26

7

5.05

9

7.10

1

7.81

0

18.7

50

13.2

00

17.7

78

9.06

9

4.92

9

5.17

3

3.49

5

6.36

7

Run

7

Lign

ite

0.00

0

1.29

3

8.13

8

8.86

2

9.68

5

23.4

05

12.9

80

14.5

30

6.16

3

3.16

6

3.25

8

2.93

1

5.58

8

Run

6

Lign

ite

0.00

0

0.95

4

8.14

5

11.7

64

13.4

20

27.0

03

13.1

24

11.4

75

3.94

9

1.93

1

1.91

8

2.12

0

4.19

8

Run

5

Lign

ite

0.00

0

0.45

8

4.45

0

7.98

2

10.3

80

26.0

56

14.0

99

15.2

52

6.14

2

3.16

7

3.29

8

3.95

0

4.76

6

Run

4

Lign

ite

0.00

0

0.87

9

3.89

7

9.83

7

11.4

44

29.5

48

14.2

33

13.3

21

4.83

9

2.25

9

2.49

8

3.43

1

3.81

5

Run

3

Lign

ite

0.00

0

0.50

4

3.98

6

7.81

6

10.1

67

25.3

01

14.4

38

15.4

62

6.24

8

3.24

2

3.54

1

3.96

0

5.33

4

Run

2

Lign

ite

0.00

0

2.03

3

8.43

4

9.23

6

10.4

13

23.3

78

13.2

65

14.1

60

5.55

3

2.68

7

2.55

5

2.68

1

5.60

6

Run

1

Lign

ite

0.00

0

1.07

8

3.04

2

4.89

6

6.10

3

16.8

59

13.5

65

21.4

78

10.8

51

5.82

3

5.55

3

4.09

4

6.65

9

Tab

le3.

8: F

uel s

ize

dist

ribut

ions

.

SIEV

E O

PEN

ING

, mm

19.0

00

16.0

00

12.7

00

8.00

0

6.30

0

4.75

0

3.35

0

2.00

0

1.00

0

0.50

0

0.35

5

0.18

0

0.10

6

0.00

0

92

Cot

ton

Res

idue

0.00

0.81

4.95

10.8

3

14.7

7

0.00

10.2

9

57.5

1

0.85

Haz

elnu

t Sh

ell

2.28

2.59

7.11

38.8

4

6.60

5.50

7.40

27.8

6

1.81

Oliv

e R

esid

ue

31.1

9

5.29

5.17

17.5

2

2.51

2.64

5.21

27.9

5

2.52

Run

10

Lign

ite

50.1

1

22.5

7

11.4

6

7.79

0.55

4.24

1.51

0.18

1.58

Run

9

Lign

ite

55.5

5

21.3

5

11.7

1

5.42

0.53

2.71

1.05

0.20

1.48

Run

8

Lign

ite

51.9

1

21.8

3

12.1

5

7.92

0.58

2.31

1.60

0.33

1.37

Run

7

Lign

ite

51.4

3

22.8

0

12.2

8

7.26

0.52

2.23

1.59

0.26

1.63

Run

6

Lign

ite

50.0

2

23.8

1

12.0

3

8.19

0.63

2.43

1.05

0.24

1.59

Run

5

Lign

ite

52.7

0

19.1

4

11.2

6

7.78

0.52

5.23

1.62

0.21

1.54

Run

4

Lign

ite

55.2

6

20.4

3

10.4

2

7.50

0.59

2.63

1.34

0.19

1.63

Run

3

Lign

ite

51.3

3

21.8

4

10.5

9

9.06

0.63

3.49

1.24

0.27

1.55

Run

2

Lign

ite

56.5

6

17.4

9

10.9

9

9.21

0.57

2.05

1.45

0.31

1.38

Run

1

Lign

ite

57.2

9

19.6

7

12.0

5

4.85

0.82

2.00

1.58

0.21

1.53

Tab

le3.

9: F

uel a

sh c

ompo

sitio

ns.

As o

xide

s, %

Silic

a, S

iO2

Alu

min

um, A

l 2O3

Ferr

ic, F

e 2O

3

Cal

cium

, CaO

Mag

nesi

um, M

gO

Sulfu

r, SO

3

Sodi

um, N

a 2O

Pota

sssi

um, K

2O

Tita

nium

, TiO

2

93

Chlorine content of biomasses analyzed by using ED-XRF (Spectro Xepos) is given

in Table 3.10. Chlorine content of lignite was obtained by EDX (Energy Dispersive

X-Ray Analysis) method in JSM-6400 Electron Microscope (JEOL).

Table 3.10: Chlorine content of fuels, %.

Olive Residue 0.11

Hazelnut Shell 0.02

Cotton Residue 0.05

Lignite 0.06

The morphologies and crystal sizes of the synthesized samples were observed by

JSM-6400 (JEOL) Scanning Electron Microscope. The micrographs of SEM were

taken in the magnification range of 1500 to 5000 times. During the SEM study, EDX

method was applied in order to get elemental composition of the deposit samples.

EDX analyses of the samples were also performed by JSM-6400 Electron

Microscope (JEOL). Ash deposits were identified qualitatively from their X-ray

diffraction patterns taken by a 100 kV Philips twin tube X-ray diffractometer

(PW/1050) using CuKα radiation.

For determination of trace element concentrations in coal, biomass and limestone a

microwave-assisted acid digestion followed by inductively coupled plasma optical

emission spectrometry and mass spectrometry (ICP-OES and ICP-MS, respectively)

were applied. A combination of nitric, hydrochloric and hydrofluoric acid was

employed in microwave oven followed by boric acid addition for removal of the

hydrofluoric acid from the reaction mixture.

The microwave digestion was carried out by using Anton Paar Multiwave 3000 oven.

ICP-OES and ICP-MS measurements were performed using Perkin Elmer Optima

4300 DV and Perkin Elmer DRC II, respectively. Trace element concentrations of

lignite and biomasses used in Runs 2, 5, 8 and 10 are given in Tables 3.11 and 3.12,

respectively.

94

Table 3.11: Trace element concentrations in lignites in Runs 2, 5, 8 and 10.

Element (mg/kg)

Run 2 Lignite

Run 5 Lignite

Run 8 Lignite

Run 10 Lignite

Detection limit

ICP-MS

As 46.1±0.9 41.6±0.5 52.7±0.3 65.7±0.5 0.01

Ba 89.5±2.7 98.0±1.1 94.8±0.8 108.4±0.9 0.008

Cd 0.108±0.002 0.106±0.004 0.114±0.002 0.185±0.008 0.002

Co 5.03±0.08 5.48±0.08 5.57±0.16 5.93±0.08 0.002

Li 23.9±1.0 23.1±0.7 28.8±0.7 27.0±0.9 0.008

Mo 3.70±0.01 4.02±0.08 4.39±0.08 4.09±0.08 0.002

Pb 16.26±0.54 16.59±0.08 12.29±0.16 21.38±0.16 0.017

Sb 0.48±0.04 0.58±0.02 0.414±0.002 0.49±0.02 0.02

Se <0.18 <0.18 <0.18 <0.18 0.18

Sn 2.41±0.04 4.22±0.12 1.98±0.04 3.60±0.32 0.008

Tl 0.777±0.004 0.832±0.004 0.581±0.004 0.604±0.080 0.003

ICP-OES

Cr 6.6±0.1 13.2±0.1 6.4±0.1 6.6±0.3 0.007

Cu 54.0±0.3 57.1±0.2 49.0±1.0 52.7±1.7 0.01

Mn 110.0±1.0 114.0±0.5 96.0±1.0 106.2±0.3 0.0014

Ni <0.015 5.3±0.3 <0.015 <0.015 0.015

P 75.9±2.1 50.3±1.3 47.1±1.5 58.5±3.0 0.08

V 112.1±0.5 121.9±2.0 114.0±1.0 111.4±1.1 0.009

Zn 39.3±1.3 39.9±0.3 34.7±0.1 34.7±0.7 0.006

95

Table 3.12: Trace element concentrations in biomass.

Element (mg/kg)

Olive residue

Hazelnut shell

Cotton residue

Detection limit

ICP-MS

As 0.72±0.02 0.29±0.02 0.12±0.01 0.01

Ba 14.7±0.1 18.2±0.2 2.27±0.02 0.008

Cd 0.017±0.001 0.040±0.001 0.022±0.001 0.002

Co 0.527±0.005 0.380±0.005 0.472±0.004 0.002

Li 0.776±0.005 0.118±0.005 0.136±0.004 0.008

Mo 0.25±0.01 0.13±0.01 1.61±0.01 0.002

Pb 3.24±0.01 3.53±0.04 1.22±0.02 0.017

Sb 0.144±0.001 0.066±0.001 0.028±0.001 0.02

Se <0.18 <0.18 <0.18 0.18

Sn 0.616±0.006 0.110±0.005 1.455±0.014 0.008

Tl 0.016±0.001 <0.003 <0.003 0.003

ICP-OES

Cr 8.98±0.11 1.82 ±0.03 2.23 ±0.01 0.007

Cu 15.3±0.2 7.9±0.2 11.3±0.1 0.01

Mn 26.2±0.2 106.3±0.4 17.1±0.1 0.0014

Ni 5.1±0.1 1.97±0.02 2.8±0.1 0.015

P 907±10 146±3 7361±97 0.08

V 2.91±0.02 1.05±0.03 5.3±0.4 0.009

Zn 14.8±0.1 22.7±0.2 32.2±0.1 0.006

96

Limestone utilized in the firing tests was supplied by Park Thermic, Electric Industry

and Trade, Inc. and originates from Acıbaşı limestone quarry, 10 km away from the

Çayırhan Thermal Power Plant. Limestone delivered to the laboratory had a particle

size below 6 cm. It was subjected to size reduction by crushing it in a jaw-crusher

and a hammer mill consecutively. Crushed limestone was sieved through a 1.18 mm

sieve and top product was crushed again by hammer mill. Particles under the sieve

were utilized in the experiments. A representative sample from limestone was

subjected to sieve and chemical analyses and the results are summarized in Table

3.13. Trace element concentrations in limestone are given in Table 3.14.

Table 3.13: Characteristics of Beypazarı limestone.

Size Distribution Chemical Analysis (wet)

Size (mm) Weight (%) Component Weight (%)

1.000 – 1.180 13.01 Moisture 0.69

0.850 – 1.000 5.09 CaCO3 88.92

0.710 – 0.850 6.01 MgCO3 6.44

0.600 – 0.710 10.85 SiO2 2.91

0.500 – 0.600 3.95 Na2O 0.15

0.425 – 0.500 10.07 K2O 0.08

0.355 – 0.425 6.45 Al2O3 0.39

0.180 – 0.355 16.82 Fe2O3 0.43

0.106 – 0.180 10.13 LOI 42.43

0.000 – 0.106 17.63 d50: 0.41 mm

97

Table 3.14: Trace element concentrations in Beypazarı limestone.

Element (mg/kg) Limestone Detection limit

ICP-MS

As 10.4±0.4 0.01

Ba 97.7±1.6 0.008

Cd 0.062±0.001 0.002

Co 1.08±0.02 0.002

Li 17.1±0.2 0.008

Mo 0.97±0.01 0.002

Pb 5.20±0.05 0.017

Sb 0.302±0.014 0.02

Se <0.18 0.18

Sn 2.22±0.04 0.008

Tl 0.168±0.009 0.003

ICP-OES

Cr 9.2 ±0.1 0.007

Cu 11.1±0.1 0.01

Mn 47.6±0.1 0.0014

Ni 5.9±0.1 0.015

P 59±1 0.08

V 9.3±0.1 0.009

Zn 8.3±0.3 0.006

98

3.5.2 Operating Conditions

In order to investigate the effect of biomass share on emission performance of the

test rig, a total of 10 runs without/with limestone addition were carried out at several

biomass shares. In Run 1, coal is burned without limestone and biomass addition

whereas in Run 2, coal is burned with limestone addition. In Runs 3 to 5, coal is

burned with limestone addition at several olive residue shares, i.e., 15, 30 and 50 wt

%, respectively. In Runs 6 to 8, coal is burned with limestone addition at several

hazelnut shell shares, i.e., 11, 30 and 42 wt %, respectively. In the last two runs,

Runs 9 and 10, coal is burned with limestone addition at several cotton residue

shares, i.e., 30 and 41 wt %, respectively. Runs were carried out consecutively by the

same team. The total duration of operation was about 40 hours. Durations of the runs

are given in Table 3.15. As can be seen from the table the last run was of the shortest

duration due to feeding problems encountered at the beginning of the run. In all the

runs, the lignite was burned in its own ash due to its high ash content. Table 3.16 lists

the operating conditions of the runs. During the runs, parameters other than biomass

share were tried to be maintained constant. Feed point location was 0.22 m above the

distributor plate for all runs. Runs reported in this thesis study refer to the experiment

coded as 070317.

Table 3.15: Durations of the runs.

Run 1 5 hours 52 minutes






Run 7 1 hour 56 minutes



Run 10 22 minutes

99

Run

10

35.7

25.2

41

48

12.9

2.7

0.0

18.0

0.0

15.0

15.1

21

2.0

857

843

1.15

Run

9

46.0

19.7

30

37

16.7

2.7

5.5

17.3

0.0

14.1

14.5

10

1.9

860

849

1.15

Run

8

32.4

23.3

42

50

13.9

3.3

2.2

12.1

1.9

14.0

14.3

22

1.9

854

835

1.10

Run

7

41.0

17.2

30

35

14.1

2.8

2.5

14.5

1.4

14.0

14.1

21

1.9

853

832

1.16

Run

6

54.3

7.0 11

14

18.6

3.2

5.5

17.1

1.0

14.2

14.4

20

1.9

857

831

1.22

Run

5

30.2

28.8

49

55

11.2

3.3

1.5

11.0

1.7

14.0

14.3

28

1.9

852

849

1.10

Run

4

40.9

18.8

31

36

13.9

2.9

3.6

12.5

1.3

14.0

14.2

23

1.9

846

832

1.14

Run

3

56.6

10.0

15

19

19.1

3.1

8.0

16.6

0.8

14.0

14.5

18

1.9

860

839

1.18

Run

2

68.7

0.0 0 0 22.4

2.7

8.3

19.4

1.2

14.0

14.6

21

1.9

848

817

1.12

Run

1

76.5

0.0 0 0 0.0 0 6.9

14.2

0.4

16.0

16.8

23

2.2

894

866

1.02

Tab

le 3

.16:

Ope

ratin

g co

nditi

ons o

f Run

s 1-1

0.

Coa

l flo

w ra

te, k

g/h

Bio

mas

s flo

w ra

te, k

g/h

Bio

mas

s sha

re, w

t %

Bio

mas

s sha

re (o

n th

erm

al b

asis

), %

Lim

esto

ne

flow

rate

, kg/

h

Ca/

S m

olar

ratio

(bas

ed o

n to

tal S

)

Bot

tom

ash

flow

rate

, kg/

h

Cyc

lone

ash

flow

rate

, kg/

h

Bag

hous

e fil

ter a

sh fl

ow ra

te, k

g/h

Air

flow

rate

, km

ol/h

Flue

gas

flow

rate

, km

ol/h

Exce

ss a

ir, %

Supe

rfic

ial v

eloc

ity, m

/s

Ave

rage

bed

tem

pera

ture

, °C

Ave

rage

free

boar

d te

mpe

ratu

re, °

C

Bed

hei

ght,

m

100

Run

10

63

13

2

4802

3170

13.0

31.2

27.0

Run

9

65

13

2

3301

3281

12.9

39.4

26.9

Run

8

60

12

2

4073

2423

13.0

34.2

30.7

Run

7

63

12

2

3526

2840

12.9

37.2

27.7

Run

6

66

12

2

3094

2634

13.0

40.4

29.2

Run

5

61

12

2

3165

2691

12.9

39.0

29.6

Run

4

63

12

2

3378

2884

12.9

37.0

27.7

Run

3

65

11

2

3435

2917

12.9

37.0

27.7

Run

2

63

11

2

2842

2767

13.0

40.3

28.1

Run

1

54

11

2

3629

1792

12.9

35.5

41.2

Tab

le 3

.16:

Ope

ratin

g co

nditi

ons o

f Run

s 1-1

0 (c

ont’d

).

ΔPB

ed, c

m H

2O

ΔPD

istri

buto

r pla

te, c

m H

2O

Bed

surf

ace

pres

sure

, cm

H2O

Bed

coo

ling

wat

er fl

ow ra

te, k

g/h

Free

boar

d co

olin

g w

ater

flo

w ra

te, k

g/h

Coo

ling

wat

er i

nlet

tem

p., °

C

Bed

coo

ling

wat

er o

utle

t tem

p., °

C

Free

boar

d co

olin

g w

ater

out

let t

emp.

, °C

101

CHAPTER 4

RESULTS & DISCUSSION

4.1 General

This thesis study is based on experimental data collected from a research project for

the investigation of combustion and gaseous emission characteristics of various

biomasses co-fired with typical low quality Turkish lignite having high ash and

sulfur contents. Combustion tests were carried out on the 0.3 MWt ABFBC test rig

located in the Chemical Engineering Department of Middle East Technical

University.

The effect of biomass type and share on combustion and emission performance was

analyzed with respect to the particle size distributions, ash split and partitioning of

major, minor and trace elements in ash streams, combustion efficiency, temperature

and concentration profiles, emissions and deposit formations on heat exchange

surfaces. Lignite was co-fired with olive residues/hazelnut shells/cotton residues, at

olive residue shares of 15, 31 and 49 wt %, hazelnut shell shares of 11, 30, 42 wt %

and cotton residue shares of 30 and 41 wt % in their own ashes. During the tests

parameters other than biomass share were tried to be maintained constant.

4.2 Particle Size Distributions

Particle size distributions of the inlet and outlet streams for runs burning lignite

without and with limestone addition are shown in Figures 4.1.

102

Figu

re 4

.1: S

ize

dist

ribut

ion

of a

ll so

lid st

ream

s in

Run

s 1 a

nd 2

.

103

As can be seen from the figure, particle size decreases in the following order; lignite,

bottom ash, cyclone ash and baghouse filter ash, as expected. Addition of fine

limestone in Run 2 is seen to decrease the particle size of cyclone ash compared to

cyclone ash of Run 1. This can be attributed to the fine particles of limestone

elutriated with the gas. Particle size distributions of inlet and outlet streams in

biomass co-firing runs are illustrated together with those of lignite firing with

limestone run in Figures 4.2-4.4. In the figures, the latter case has no biomass and

hence can be taken as a reference case. Figure 4.2 displays the particle size

distributions of runs with olive residues and the reference case. Introduction of olive

residues results in coarser cyclone ash particles with respect to lignite combustion

with limestone addition (Run 2) despite finer particle size of olive residues compared

to that of lignite. This is mainly due to the introduction of low bulk density olive

residue (591 kg/m3) compared to that of lignite (905 kg/m3) which leads to elutriation

of less dense coarse particles to cyclone. In addition, decrease in limestone flow rate

with biomass addition may reduce the fines fraction in this region. Increasing the

share of olive residue within the fuel feed from 15 to 49 % has almost no influence

on particle size distributions of the ash streams.

Comparison between particle size distributions of ashes without and with hazelnut

shells is shown in Figure 4.3. As can be seen from the figure the effect is only

noticeable in cyclone ash for the largest share of hazelnut in the fuel blend. This may

be considered to be due to the combined effect of lower bulk density (320 kg/m3)

ands coarser particle size of hazelnut shells compare to those of lignite.

Figure 4.4 shows particle size distributions of inlet and outlet solid streams for co-

firing tests carried out with cotton residues. Trends are similar to those of hazelnut

shells.

104

Figu

re 4

.2: S

ize

dist

ribut

ion

of a

ll so

lid st

ream

s in

Run

s 2, 3

, 4 a

nd 5

.

105

Figu

re 4

.3: S

ize

dist

ribut

ion

of a

ll so

lid st

ream

s in

Run

s 2, 6

, 7 a

nd 8

.

106

Figu

re 4

.4: S

ize

dist

ribut

ion

of a

ll so

lid st

ream

s in

Run

s 2, 9

and

10.

107

4.3 Ash Balance, Split and Discharge Compositions

4.3.1 Ash Balance and Ash Split

Ash balance over the combustor for all runs is tabulated in Table 4.1. The closures

and ash splits between bottom and fly ashes are also presented. As can be seen from

the table, ash recovery rates for Runs 1-10 are consistent. Compared to the ash

recovery rates reported in the literature [124, 125] and considering the difficulty in

closing the total solid mass balances over the fluidized bed combustors, the closure

of the total ash balances is acceptable in all runs. About 70 % of the ash is recovered

from fly ash during lignite combustion runs and addition of biomass results in further

increase in ash split to fly ash. In the runs carried out with olive residues, increasing

olive residue share from 15 to 49 wt % leads to increase in ash split to fly ash from

69 to 89 %. Similar trend is also obtained for the co-firing runs performed with

hazelnut shells. Increasing the hazelnut shell share from 11 to 42 wt % results in

increasing ash split to fly ash from 77 to 87 %. Ash split to fly ash in co-firing

cotton residues with lignite at 30 wt % cotton residue share is obtained as 76 %, and

increasing the share of cotton residue to 42 wt % leads to about 100 % ash split to fly

ash. Enhanced shifting of ash split to fly ash by biomass addition is mainly due to the

lower bulk density of the biomass fuels compared to that of lignite leading to more

elutriation of particles from the combustor.

4.3.2 Ash Compositions

4.3.2.1 Major and Minor Elements

Ash compositions of bottom, cyclone and baghouse filter ashes are shown in Figures

4.5-4.7, respectively. Examination of the compositions of ash streams reveals similar

concentrations of compounds in bottom, cyclone and baghouse filter ashes. As can

be seen from figures, addition of biomass reduces SiO2, Al2O3 and Fe2O3

concentrations in ashes irrespective of biomass type.

108

Run

10

8.

68

1.35

7.

44

17

.47

0.00

18

.00

0.00

18.0

0

103

100

Run

9

12

.45

1.06

9.

60

23

.11

5.53

17

.25

0.00

22.7

8 99

76

Run

8

8.

90

0.34

8.

01

17

.25

2.16

12

.09

1.93

16.1

8 94

87

Run

7

10

.00

0.25

8.

12

18

.37

2.50

14

.47

1.41

18.3

8

100

76

Run

6

13

.74

0.10

10

.71

24.5

5 5.

51

17.1

1 1.

01

23

.63 96

77

Run

5

7.

21

1.22

6.

82

15

.25

1.50

10

.95

1.71

14.1

6 93

89

Run

4

9.

26

0.80

8.

03

18

.09

3.55

12

.45

1.31

17.3

1 96

80

Run

3

13

.91

0.42

11

.03

25.3

7 7.

97

16.6

1 0.

75

25

.33

100

69

Run

2

18

.37

0.00

12

.92

31.2

9 8.

28

19.3

6 1.

20

28

.84 92

71

Run

1

22

.01

0.00

0.

00

22

.01

6.91

14

.15

0.39

21.4

5 98

68

Tab

le 4

.1: A

sh b

alan

ce, c

losu

re a

nd sp

lit.

Inpu

t (kg

/h)

C

oal A

sh

B

iom

ass A

sh

So

rben

t

T

otal

Sol

ids I

n O

utpu

t (kg

/h)

Bot

tom

Ash

Cyc

lone

Ash

B

agho

use

Ash

T

otal

Sol

ids O

ut

Clo

sure

, %

Ash

Spl

it, %

109

Figu

re 4

.5: B

otto

m a

sh a

naly

ses o

f Run

s 1-1

0.

110

Figu

re 4

.6: C

yclo

ne a

sh a

naly

ses o

f Run

s 1-1

0.

111

Figu

re 4

.7: B

agho

use

filte

r ash

ana

lyse

s of R

uns 1

-10.

112

As these compounds are mainly originated from lignite, reduction of lignite share in

the fuel blends results in lower concentrations. On the other hand, CaO and SO3

concentrations are increased from Runs 1 to 10 due to limestone addition and sulfur

capture within the bed and increase in inherent CaO fraction by addition of biomass.

Na2O and K2O concentrations in all the ash streams were obtained to be low and

insensitive to biomass addition. This is indicative of absence of problems associated

with basic oxides of biomass during co-firing.

4.3.2.2 Trace Elements

Trace element analyses are carried out for the run with lignite firing with limestone

addition (Run 2) and olive residue and hazelnut co-firing runs with highest biomass

share in the fuel feed (Runs 5 and 8, respectively) by ICP-OES (Cr, Cu, Mn, Ni, P,

V, Zn) and ICP-MS (As, Ba, Cd, Co, Hg, Li, Mo, Pb, Se, Sb, Sn, Tl) in order to

reflect the effect of biomass addition and biomass type on bottom, cyclone and

baghouse filter ash compositions. Trace element compositions of bottom, cyclone

and baghouse filter ashes are given in Tables 4.2, 4.3 and 4.4, respectively.

Concentration of Se (<0.18 mg/kg) could not be measured as its concentration was

below the detection limits of the instrument. Hg concentration could not be

quantified by ICP techniques because it is volatilized during the sample preparation

procedure [126].

Trace element concentrations can also be described in terms of relative enrichment

factors which describe the behavior of elements in bottom and fly ashes. Relative

enrichment factor (RE) is defined by Meij [127] as;

( )( )

( )Element concentration in ash % Ash content of fuelRE = × (4.1)

Element concentration in fuel 100

113

Table 4.2: Trace element concentrations in bottom ash of Runs 2, 5 and 8. Element (mg/kg) Run 2 Run 5 Run 8 Detection

limit

ICP-MS

As 55.1±0.8 50.6±0.5 43.0±0.1 0.01

Ba 549.7±5. 8 407.3±8.0 389.6±1.7 0.008

Cd 0.782±0.040 0.576±0.004 0.599±0.040 0.002

Co 12.74±0.29 11.51±0.25 12.60±0.13 0.002

Li 71.4±0.9 44.8±0.9 39.1±1.0 0.008

Mo 3.76±0.08 5.00±0.12 4.56±0.13 0.002

Pb 53.71±0.79 41.98±0.29 45.00±0.33 0.017

Sb 3.04±0.04 2.08±0.04 1.80±0.08 0.02

Se <0.18 <0.18 <0.18 0.18

Sn 5.86±0.04 4.98±0.12 3.77±0.12 0.008

Tl 1.22±0.04 0.94±0.04 1.073±0.004 0.003

ICP-OES

Cr 37.5±0.2 38.6±0.2 32.9±0.1 0.007

Cu 17.1±0.2 14.4±0.5 16.5±0.2 0.01

Mn 319.0±3.0 275.0±1.0 275.0±1.0 0.0014

Ni 16.6 ± 0.6 14.7 ± 0.5 13.2 ± 0.5 0.015

P 1139 ± 7 832 ± 5 676 ± 5 0.08

V 76.4±0.6 77.0±1.0 83.5±0.7 0.009

Zn 45.3±0.6 37.3±0.5 35.6±0.9 0.006

114

Table 4.3: Trace element concentrations in cyclone ash of Runs 2, 5 and 8. Element (mg/kg) Run 2 Run 5 Run 8 Detection

limit

ICP-MS

As 95.1±1.0 60.6±1.0 77.5±1.4 0.01

Ba 313.6±6.5 269.7±4.3 230.4±4.1 0.008

Cd 0.199±0.004 0.178±0.004 0.166±0.004 0.002

Co 15.68±0.37 14.04±0.04 14.16±0.12 0.002

Li 69.0±1.0 70.1±1.2 52.6±2.0 0.008

Mo 8.78±0.25 6.55±0.08 5.94±0.21 0.002

Pb 27.55±0.42 19.25±0.17 23.01±0.58 0.017

Sb 1.02±0.04 1.04±0.04 0.82±0.04 0.02

Se <0.18 <0.18 <0.18 0.18

Sn 1.62±0.04 1.30±0.04 1.25±0.04 0.008

Tl 1.333±0.040 1.099±0.004 1.293±0.004 0.003

ICP-OES

Cr 22.0±0.2 24.9±0.5 21.2±0.3 0.007

Cu 31.1±0.1 28.7±0.4 22.1±0.2 0.01

Mn 251.0±1.0 203.9±1.8 180.0±0.9 0.0014

Ni 11.3 ± 0.9 11.3 ± 0.8 8.3 ± 0.2 0.015

P 129 ± 8 563 ± 10 132± 5 0.08

V 212.6±0.6 141.3±2.0 130.0±1.0 0.009

Zn 30.1±0.5 28.9±0.4 26.2±0.6 0.006

115

Table 4.4: Trace element concentrations in filter ash of Runs 2, 5 and 8. Element (mg/kg) Run 2 Run 5 Run 8 Detection

limit

ICP-MS

As 565.8 ± 1.4 369.4 ± 3.6 372.8 ± 1.8 0.01

Ba 329.3±3.2 355.5±4.1 348.9±5.7 0.008

Cd 0.389±0.004 0.422±0.004 0.405±0.004 0.002

Co 26.83±0.54 26.31±0.41 27.68±0.54 0.002

Li 151.1±3.4 189.3±1.9 177.6±4.5 0.008

Mo 26.05 ± 0.21 19.04 ± 0.12 19.91 ± 0.25 0.002

Pb 32.70±0.21 25.94±0.25 21.73±0.46 0.017

Sb 2.647±0.004 2.49±0.08 2.16±0.04 0.02

Se <0.18 <0.18 <0.18 0.18

Sn 2.94±0.08 3.95±0.04 2.72±0.004 0.008

Tl 1.26±0.04 1.62±0.04 1.53±0.04 0.003

ICP-OES

Cr 33.0±0.2 37.6±0.4 35.3±0.1 0.007

Cu 119.9±0.9 127.3±2.0 117.6±0.8 0.01

Mn 395.0±1.0 413.0±3.0 543.0±2.0 0.0014

Ni 24.7 ± 0.2 33.2 ± 0.3 28.4 ± 0.4 0.015

P 128±3 1828 ± 17 746 ± 3 0.08

V 461.0±2.0 426.0±5.0 449.0±1.0 0.009

Zn 77.0±1.1 75.1±0.9 78.0±0.5 0.006

116

Trace elements can be classified according to their volatility behavior as Class I,

Class II and Class III elements [127]. Class I elements are defined as elements that

do not vaporize during combustion. Their enrichment factors are about 1 in bottom

and fly ashes and hence, they are equally distributed between bottom and fly ash.

Class II elements on the other hand, vaporize in the boiler and condense within the

installation. The relative enrichment factor of bottom ash is less than 0.7 because

elements originally present in the vapor phase have no chance to condense on bottom

ash particles. As, Ba, Cd, Co, Cr, Cu, Mn, Mo, Ni, P, Pb, Sb, Sn, Tl, V and Zn

belong to Class II elements. Class III elements such as Hg, Se and Cl are very

volatile and they hardly condense in the boiler. Their enrichments factors are

generally much smaller than 1 [127, 128].

The calculated enrichment factors of trace elements in bottom, cyclone and baghouse

filter ash streams are given in Table 4.5. As can be seen from the table, volatile

elements As, Ba, Co, Cr, Cu, Mn, Mo, Ni, Pb, Sb, Sn, Tl, V and Zn are mostly

enriched in fly ash of Runs 2, 5 and 8. These findings are in accordance with the

previous studies on coal firing [129, 131] and biomass and coal co-firing [132]. The

relative enrichment factors of As, Co, Cu, Mn, Mo, Tl, and V are found to be lower

than 0.7 in all bottom ashes and hence fall into Class II elements which are enriched

in fly ash [127].

4.4 Partitioning of Major, Minor and Trace Elements

Investigation of partitioning behavior of 9 major and minor (Al, Ca, Fe, K, Mg, Na,

S, Si, Ti) and 18 trace elements (As, Ba, Cd, Co, Cr, Cu, Li, Mn, Mo, Ni, P, Pb, Sb,

Se, Sn, Tl, V, Zn) is carried out for the run with lignite firing with limestone addition

(Run 2) and olive residue and hazelnut shell co-firing runs with highest biomass

share in the fuel feed (Runs 5 and 8, respectively) in order to reflect the effect of

biomass addition and biomass type on elements partitioning in bottom, cyclone and

baghouse filter ashes.

117

Filte

r A

sh

2.42

1.11

0.97

1.62

1.57

0.74

2.11

1.08

1.52

6.83

1.67

0.50

1.60

0.45

0.90

1.34

0.52

Cyc

lone

A

sh

0.50

0.73

0.40

0.83

0.94

0.14

0.62

0.36

0.45

2.00

0.30

0.53

0.61

0.21

0.76

0.39

0.18

Run

8

Bot

tom

A

sh

0.28

1.24

1.44

0.74

1.46

0.10

0.46

0.52

0.35

3.17

1.52

1.04

1.34

0.63

0.63

0.25

0.24

Filte

r A

sh

2.88

1.05

1.14

1.45

0.57

0.59

2.62

0.98

1.48

1.08

0.66

0.43

1.14

0.27

0.63

1.13

0.46

Cyc

lone

A

sh

0.47

0.79

0.48

0.78

0.38

0.13

0.97

0.48

0.51

0.37

0.20

0.32

0.48

0.09

0.43

0.37

0.18

Run

5

Bot

tom

A

sh

0.40

1.20

1.56

0.64

0.58

0.07

0.62

0.65

0.39

0.48

0.30

0.70

0.96

0.34

0.37

0.20

0.23

Filte

r A

sh

3.28

0.98

0.96

1.42

1.34

0.59

1.69

0.96

1.88

0.00

0.45

0.54

1.47

0.33

0.43

1.10

0.52

Cyc

lone

A

sh

0.55

0.94

0.49

0.83

0.89

0.15

0.77

0.61

0.64

0.00

0.45

0.45

0.57

0.18

0.46

0.51

0.21

Run

2

Bot

tom

A

sh

0.32

1.64

1.94

0.68

1.52

0.09

0.80

0.78

0.27

0.00

4.01

0.88

1.69

0.65

0.42

0.18

0.31

Tab

le 4

.5: R

elat

ive

enric

hmen

t fac

tors

of t

race

ele

men

ts in

bot

tom

, cyc

lone

and

filte

r ash

es.

As

Ba

Cd

Co

Cr

Cu Li

Mn

Mo

Ni P Pb

Sb

Sn

Tl V

Zn

118

The imbalances are observed in the species mass flow rates. These imbalances could

result from analyzing very small quantity of sample from large quantities of highly

heterogeneous matter, very low concentrations of elements under investigation,

analytical errors and experimental errors in preparing the samples for ICP-OES and

ICP-MS analyses or a combination of all [133].

Figure 4.8 shows the recovery rates of major and minor ash components for Run 2, 5

and 8. As can be seen from the figure, the closure is fairly good for most of the

components. Recovery rates of trace elements are given in Figure 4.9. The

imbalances are observed in the species mass flow rates. These imbalances could

result from analyzing very small quantity of sample from large quantities of highly

heterogeneous matter, very low concentrations of elements under investigation,

analytical errors and experimental errors in preparing the samples for ICP-OES and

ICP-MS analyses or a combination of all [133].

Major and minor ash component partitioning of bottom, cyclone and baghouse filter

ashes produced from Runs 2, 5 and 8 are displayed in Figure 4.10. For each ash

component first, second and third columns refer to Runs 2, 5 and 8, respectively. As

can be depicted from the figure, the partitioning of major and minor ash components

follow the ash split between the bottom ash and fly ash. For the experiments and the

test rig under consideration ash splits to fly ash are found as 71 %, 89 % and 87 %

for Run 2, Run 5 and Run 8, respectively. It can be noted that addition of biomass

shifts the partitioning of major and minor elements from bottom ash to fly ash.

Although, species mass balances over the fluidized bed combustor could not be

closed, some light could be shed to the fate of trace elements in fluidized bed

combustion of high ash content low quality lignites by investigating their partitioning

behavior.

Figure 4.11 illustrates the partitioning of trace elements in bottom, cyclone and filter

ashes for Runs 2, 5 and 8, respectively. For each ash component first, second and

third columns refer to Runs 2, 5 and 8, respectively.

119

Figu

re 4

.8: R

ecov

ery

of m

ajor

and

min

or e

lem

ents

in R

uns 2

, 5 a

nd 8

.

120

Figu

re 4

.9: R

ecov

ery

of tr

ace

elem

ents

in R

uns 2

, 5 a

nd 8

.

121

Figu

re 4

.10:

Maj

or a

nd m

inor

ele

men

ts p

artit

ioni

ng o

f Run

s 2, 5

and

8.

122

Figu

re 4

.11:

Tra

ce e

lem

ents

par

titio

ning

of R

uns 2

, 5 a

nd 8

.

123

The major proportions of As, Cr, Ni, V, Zn were recovered in the cyclone ash for all

runs. This finding complies with the data for coal combustion given in [131, 134]. As

can also be seen from the figure, Ba, Cu, Sn and Mo were mostly recovered in fly

ash. The capture of As in fly ash could be attributed to partial condensation of these

high volatile species due to the low operating temperatures (~350°C) both at the exit

of the combustor due to presence of cooler and in the cyclone [135].

Comparisons between the trace element partitioning of the runs with and without

biomass addition reveal that addition of olive residue and hazelnut shell enhances the

partitioning of As, Ba, Co, Cr, Cu, Li, Mn, Mo, Ni, Pb, Tl, V and Zn to fly ash. Also,

co-firing shifts the partitioning of Cd, P, Sb and Sn from bottom to fly ash. The

reason behind the shift is explained by porous char of biomass [2, 136] leading to

increase in surface area of ash particles over which condensation of vaporized

elements can take place. Li is a non-volatile element; however, it may vaporize with

surrounding alumina silicates and become a part of fly ash. As, Ba, Co, Cr, Cu, Mo,

Ni, Pb, Tl, Sn, V and Zn belong to Class II elements which vaporize and condense in

cooler regions, addition of biomass enhances their partitioning to fly ash.

4.5 Combustion Efficiencies

The effects of operating parameters upon the magnitude of combustibles loss are

investigated by analyzing all solid streams in terms of their carbon contents and CO

emission in the flue gas. The fractional combustibles loss for each run is calculated

as the ratio of the heat loss due to CO emission in the flue gas and unburned

combustibles in bottom, cyclone and baghouse filter ashes to the potential heat of

combustion of coal feed.

Combustion efficiencies calculated from fractional combustibles losses are shown in

Table 4.6. Inspection of the table reveals that combustion efficiencies are very high

(~97 %) for the reactive lignite under consideration despite the absence of cyclone

ash recycle. Combustion efficiency reduces from 97 % (Run 1) to 96 % (Run 2) with

addition of limestone.

124

Run

10

35

.7

3202

11

4310

25.2

41

51

1046

05

2189

15

0.00

1.

87

0 18

.00

3.97

56

00

0.

00

2.41

0

15.1

0 0.

0258

26

3

5863

97

Run

9

46

.0

2989

13

7512

19.7

41

51

8177

5

2192

87

5.53

2.

09

905

17.2

5 4.

30

5814

0.00

2.

26

0 14

.50

0.04

39

431

7150

97

Run

8

32

.4

2985

96

709

23.3

41

85

9751

1

1942

19

2.16

1.

91

323

12.0

9 3.

61

3420

1.93

4.

13

624

14.3

0 0.

0222

21

5

4581

98

Run

7

41

.0

3312

13

5800

17.2

41

85

7198

2

2077

82

5.07

1.

64

653

14.4

7 3.

80

4312

1.41

5.

45

602

14.1

0 0.

0223

21

3

5781

97

Run

6

54

.3

3209

17

4237

7.0

4185

29

295

2035

32

5.51

1.

28

551

17.1

1 3.

50

4699

1.01

4.

79

379

14.4

0 0.

0219

21

3

5843

97

Run

5

30

.2

3343

10

0950

28.8

43

12

1241

86

2251

36

1.50

1.

80

212

10.9

5 3.

50

3001

1.71

9.

34

1251

14.3

0 0.

0516

49

9

4963

98

Run

4

40

.8

3475

14

1797

18.8

43

12

8106

6

2228

62

3.55

1.

67

464

12.4

5 2.

75

2681

1.31

10

.48

1076

14.2

0 0.

0239

23

0

4452

98

Run

3

56

.6

3244

18

3596

10.0

43

12

4312

0

2267

16

7.97

1.

52

946

16.6

1 3.

17

4120

0.75

5.

37

316

14.5

0 0.

0222

21

8

5600

98

Run

2

68

.7

3165

21

7421

0.0

0.0

0.0

2174

21

8.28

1.

62

1049

19.3

6 4.

31

6542

1.20

5.

74

540

14.6

0 0.

0212

20

9

8340

96

Run

1

76

.5

2943

22

5171

0.0

0.0

0.0

2251

71

6.91

1.

31

711

14.1

5 5.

05

5601

0.39

8.

83

270

16.8

0 0.

0347

39

4

6975

97

Tab

le 4

.6: C

ombu

stio

n ef

ficie

ncie

s. In

put:

Coa

l Fee

d R

ate,

kg/

h

Coa

l LH

V, k

cal/k

g

Coa

l Ene

rgy

Inpu

t, kc

al/h

B

iom

ass F

eed

Rat

e, k

g/h

B

iom

ass L

HV

, kca

l/kg

B

iom

ass E

nerg

y In

put,

kcal

/h

T

otal

Ene

r gy

Inpu

t, kc

al/h

Out

put:

Bot

tom

Ash

Dis

char

ge F

low

Rat

e, k

g/h

U

nbur

ned

Car

bon,

%

B

otto

m A

sh E

nerg

y Lo

ss, k

cal/h

C

yclo

ne A

sh D

isch

arge

Flo

w R

ate,

kg/

h

Unb

urne

d C

arbo

n, %

Cyc

lone

Ash

Ene

rgy

Loss

, kca

l/h

Bag

hous

e A

sh D

isch

arge

Flo

w R

ate,

kg/

h

Unb

urne

d C

arbo

n, %

Bag

hous

e A

sh E

nerg

y Lo

ss, k

cal/h

St

ack

Gas

Flo

w R

ate,

km

ol/h

Stac

k ga

s CO

, %

St

ack

Gas

Ene

rgy

Loss

, kca

l/h

T

otal

Ene

r gy

Los

s, kc

al/h

Com

bust

ion

Eff

icie

ncy,

%

125

Reduction in combustion efficiency of about 1 % is in agreement with the findings of

previous study burning similar lignite under similar conditions in the same test rig

[150]. Decrease in efficiency with limestone addition is due to the combined effect of

introduction of a cooler solid which leads to loss of sensible energy from the system

and the net energy loss results from endothermic decomposition reaction of CaCO3

and exothermic formation reaction of CaSO4 at Ca/S molar ratio of 3.

Comparisons of efficiencies of runs with biomass co-firing (Runs 3-10) with that of

base run (Run 2) reveals that combustion efficiency increases with biomass addition.

This increase results from the high volatile matter content of biomass (~75 % on as

received basis) as high volatile matter of biomass rapidly burns and results in highly

porous char accelerating the char combustion as well [2, 136]. Increase in

combustion efficiency with biomass addition is in accordance with previous studies

[108, 114, 115, 118, 155]. During co-firing of olive residue with lignite efficiency

remained constant at 98 % for all runs irrespective of olive residue share.

Insensitivity of combustion efficiency to olive residue share for under-bed feeding

was also reported previously [107].

During co-firing of hazelnut shell, combustion efficiency remains constant at 97 %

up to 30 wt % hazelnut shell in the fuel blend and increases to 98 % with increase of

hazelnut shell share to 42 wt % in the fuel blend. Co-firing of cotton residue with

lignite resulted in 97 % combustion efficiency irrespective of cotton residue share in

the fuel mixture.

126

4.6 Temperature Profiles

Temperature measurements taken during experiments carried out with lignite

combustion with and without limestone addition are displayed in Figure 4.12.

Inspection of the temperature profiles for the experiments with and without limestone

addition shows that temperature decreases considerably in both bed and freeboard

with addition of limestone, as expected. The fall in gas temperature toward the exit is

due to the presence of the cooler in the final module.

In order to see the effect of limestone addition on freeboard temperatures in Run 1

and 2, measured bed and freeboard temperatures are normalized by dividing

freeboard temperatures to the average value of bed temperature of the respective

runs. Figure 4.13 displays normalized temperature profiles of Run 1 and 2. As can be

seen from the figure, addition of limestone shows no significant effect on

temperature difference between bed and freeboard.

The effect of olive residue addition on temperature profiles are illustrated in Figure

4.14. In the figure OR refers to olive residue. Comparisons show that temperatures

slightly increase especially in the freeboard region with increasing olive residue

share in the fuel feed. This is considered to be due to the high volatile content of

olive residue (~76 % on as received basis) and is in accordance with the findings of

previous studies [109, 117]. Normalized temperature profiles of Runs 2, 3, 4 and 5

are shown in Figure 4.15. As can be seen from the figure, increasing olive residue

share leads to reductions in temperature difference between bed and freeboard and

provides more uniform temperature profile across the combustor.

The effect of hazelnut shell addition in the fuel fed on temperature profiles is

illustrated in Figure 4.16. In the figure HS refers to hazelnut shell. As can be seen

from the figure, hazelnut shell addition results in slightly higher temperatures in the

freeboard compared to those of lignite combustion; however, temperatures almost

stay constant along the freeboard irrespective of the hazelnut shell share. Normalized

temperature profiles of Runs 2, 6, 7 and 8 are displayed in Figure 4.17.

127

Figure 4.12: Temperature profiles of Runs 1 and 2.

Figure 4.13: Normalized temperature profiles of Runs 1 and 2.

128

Figure 4.14: Temperature profiles of Runs 2, 3, 4 and 5.

Figure 4.15: Normalized temperature profiles of Runs 2, 3, 4 and 5.

129

Figure 4.16: Temperature profiles of Runs 2, 6, 7 and 8.

Figure 4.17: Normalized temperature profiles of Runs 2, 6, 7 and 8.

130

As can be seen from the figure, addition of hazelnut shell lowers the temperature

difference between bed and freeboard temperatures compared to lignite combustion;

however, increasing hazelnut shell share from 11 to 42 wt % has no significant effect

on the temperature difference between bed and freeboard.

The effect of cotton residue addition on temperature profiles is shown in Figure 4.18.

In the figure CR refers to cotton residue. As can be seen from the figure, cotton

residue addition results in higher freeboard temperatures with respect to lignite

combustion. Freeboard temperatures almost stay constant along the freeboard at

cotton residue shares of 30 and 41 wt %. Normalized temperatures of Runs 2, 9 and

10 are shown in Figure 4.19. Similar to hazelnut shell co-firing runs, increase in

cotton residue share has almost no influence on temperature difference between bed

and freeboard.

Figure 4.18: Temperature profiles of Runs 2, 9 and 10.

131

Figure 4.19: Normalized temperature profiles of Runs 2, 9 and 10.

4.7 Concentration Profiles

4.7.1 O2, CO2 and CO Concentration Profiles

Measured concentrations of O2 and CO2 along the combustor with and without

limestone addition and with increasing biomass shares are illustrated in Figures 4.20

and 4.21. Inspection of the figures shows significant changes in concentrations across

the bed rather than the freeboard. O2 and CO2 concentrations of the first run which

was carried out without limestone addition differs from the ones carried out with

limestone and biomass additions in that higher O2 and lower CO2 concentrations are

measured at the bed exit. These profiles indicate that majority of the combustibles

are burned in bed due to high residence times provided by under-bed feeding and

sufficient bed height.

132

Figu

re 4

.20:

O2 c

once

ntra

tion

prof

iles o

f Run

s 1-6

.

133

Figu

re 4

.21:

CO

2 con

cent

ratio

n pr

ofile

s of R

uns 1

-6.

134

Combustion continues in freeboard rather slowly. Thermogravimetic analysis graphs

of biomasses and lignite given in Appendix B show that devolatilization occurs at

low temperatures in biomass (~350 °C) compared to lignite (~450-500 °C) and

therefore, high amount of volatiles release takes place in bed during co-firing runs.

About 92 and 90 % of combustion takes place across the bed in Run 2 and Runs 3-5,

respectively. Decrease of in-bed combustion by 2 percentage points is considered to

be due to introduction of higher volatile matter with biomass leading to passage and

combustion of volatile matter in freeboard. Increasing the share of olive residue

within the fuel feed from 15 to 49 wt % has almost no effect on O2 and CO2

concentrations. Hazelnut shell addition, in Run 6, also shows no significant influence

on O2 and CO2 concentration profiles.

Figure 4.22 displays CO concentration profiles. CO concentration increases across

the bed and gradually decreases in freeboard. This is due to progressive burning of

CO along the freeboard. Biomass addition to lignite leads to higher CO

concentrations due to introduction of higher volatile matter to the combustor [50,

52]. Increasing biomass share leads to further increase in CO concentrations.

CO concentrations of hazelnut shell co-firing (11 wt % share) are lower than that of

olive residue co-firing (15 wt % share) due to lower volatile matter content of

hazelnut shells compared to that of olive residues (~73 % and ~76 % on as received

basis, respectively).

Discrepancy between CO concentration at the bed exit of Run 4 and that of other

runs is attributed to problems in gas sampling due to bubbles bursting in the splash

zone which influences the measurements and also problems in the gas sampling line

due to high particle concentrations from dense bed which requires frequent purging.

135

Figu

re 4

.22:

CO

con

cent

ratio

n pr

ofile

s of R

uns 1

-6.

136

4.7.2 SO2 Concentration Profiles

SO2 concentration profiles of the Runs 1-6 are shown in Figure 4.23. As can be seen

from the figure, major changes in SO2 concentration takes place across the bed. SO2

concentration is found to be highest in Run 1 where lignite combustion is carried out

without limestone addition due to high sulfur content of lignite under consideration.

Addition of limestone in Run 2 results in lower SO2 concentrations across the bed

compared to Run 1.

SO2 concentrations continue to rise from bed exit to combustor exit. Increase in SO2

concentration along the freeboard is due to progressive release of SO2 by lignite as

well as to insufficient gas residence time for sulfur capture. SO2 profiles of the runs

carried out with biomass addition fall below the SO2 profiles of lignite firing runs.

This is an expected outcome as the sulfur content of biomass under consideration is

an order of magnitude less than that of lignite. This finding is also confirmed by

previous studies [101, 102, 107, 108, 113, 119]. SO2 concentrations of hazelnut shell

co-firing (11 wt % share) is lower than that of olive residue co-firing (15 wt % share)

due to lower sulfur content of hazelnut shells compared to that of olive residues (0.08

wt % and 0.14 wt % on dry basis, respectively).

4.7.3 NO and N2O Concentration Profiles

When a fuel enters the hot fluid bed, volatile nitrogen compounds release and some

nitrogen compounds remains in the solid char. Volatile nitrogen form species such as

NH3 and HCN which are known as the precursors of NO and N2O, respectively

whereas char nitrogen oxidizes directly to NO [137, 138]. As coal has low volatile

content, NO formation is more significant during char combustion. On the other

hand, biomass has high amount of volatiles so that NO and N2O formation is more

significant during volatile combustion rather than char combustion.

137

Figu

re 4

.23:

SO

2 con

cent

ratio

n pr

ofile

s of R

uns 1

-6.

138

Variations in NO concentrations are illustrated in Figure 4.24. As can be seen from

the figure, NO concentrations rise across the bed and remain almost constant along

the freeboard during lignite combustion with and without limestone addition.

Limestone addition in Run 2 results in higher NO concentration at the bed exit. This

is an expected outcome as reactive limestone is known to have catalytic effect on

oxidation of NH3 to NO [105, 106, 139-141].

NO concentrations remain almost constant along the freeboard in co-firing runs.

There are several factors affecting NO concentrations during combustion. Increase in

volatile matter content of the fuels lead to lower NO concentrations [108, 142, 143].

NO concentration can also be reduced by char in the presence of CO [104, 144, 145]

and hence, high CO concentrations along the freeboard may contribute to NO

reduction. Therefore, biomass addition is expected to reduce NO concentrations due

to higher volatile content and higher CO concentrations in the combustor during co-

firing runs. Reductions in NO concentration, on the other hand, are compensated by

introduction of radicals (H and OH) from hydrogen-rich biomass (6.38 wt % and

5.86 wt % on dry basis, for olive residues and hazelnut shells, respectively) during

volatile combustion. These radicals act on HCN to NH3 conversion and lead to

increase in NO concentrations [111, 137]. In addition, higher freeboard temperatures

may enhance NO formation [137, 146]. Heterogeneous catalytic effects of biomass

ash may also be significant in the oxidation of the volatiles. The selectivity in HCN

oxidation towards NO is strongly enhanced by Ca, K and Na in biomass ash [154].

Regarding N2O concentrations displayed in Figure 4.25, it can be noted that fairly

low concentrations prevail along the freeboard even for 100 % lignite combustion.

This is attributed to addition of limestone due to high sulfur content of lignite.

Reduction in N2O concentrations due to limestone addition is also confirmed by

previous studies [105, 106, 139-141]. Increase of olive residue share in fuel feed

decreases N2O concentrations even further. This can be due to the combined effect of

increasing temperature [96, 137, 140, 147] and volatile matter content [148] and also

higher radical concentrations (H and OH) [137, 148] in the fuel feed with increasing

biomass share. Formed N2O can also be destroyed in the presence of these radicals to

form N2 [137, 148].

139

Figu

re 4

.24:

NO

con

cent

ratio

n pr

ofile

s of R

uns 1

-6.

140

Figu

re 4

.25:

N2O

con

cent

ratio

n pr

ofile

s of R

uns 1

-6.

141

In addition to those, calcium, potassium and sodium contents of biomass ash have

catalytic effects on N2O decomposition [148]. Hazelnut shell addition (11 wt %)

leads to lower N2O concentration compared to that of olive residue addition (15v wt

%) due to higher nitrogen content of olive residues (1.72 wt % on dry basis)

compared to that of hazelnut shells (0.56 wt % on dry basis).

4.8 Emissions

Emissions of the Runs 1-10 measured downstream of cyclone are shown in Table

4.7. Inspection of the table reveals that CO, O2 and CO2 emissions of Runs 1 and 2

are similar to each other so that limestone addition has almost no effect on emissions

of these species. Increase in olive residue share from 15 wt % to 49 wt % results in

higher CO emissions compared to lignite firing runs due to high volatile matter

content of olive residues (~76 wt % on as received basis) compared to that of lignite.

O2 and CO2 emissions stay almost constant during olive residue co-firing runs.

Hazelnut shell co-firing shows no significant effect on CO, O2 and CO2 emissions.

Insensitivity of CO emissions to hazelnut shell addition is attributed to its lower

volatile matter content compared to that of olive residues (~73 % and ~76 % on as

received basis, respectively). Cotton residue addition on the other hand, leads to

higher CO emissions compared to those of Runs 1-8. This is attributed to the

combined effect of high volatile matter content of cotton residues (~76 % on as

received basis) and lignite combustion with low amount of limestone addition before

cotton residue co-firing runs due to problems in feeding of the residue. Cotton

residue co-firing shows no significant effect on O2 and CO2 emissions.

CO2 emissions from these tests were also evaluated from the view point of climate

change issue. As can be seen from the CO2 emission values in Table 4.7, CO2

emissions are not affected at all by biomass co-firing. However, as biomass fuels are

CO2 neutral, their contribution to CO2 emission can be considered to be negligible.

Therefore, CO2 emissions can be reduced by the proportion of biomass in fuel feed

for co-firing runs [115, 149].

142

Run

10

5.1

590

558

697

15.3

14

.5

48

6 45

9 13

12

83

285

269

552 60

57

112

Run

9

3.3

903

764

955

17.2

14

.6

92

5 78

3 22

40

76

185

157

321 43

37

72

Run

8

5.1

489

460

575

15.0

15

.0

21

4 20

1 57

6 92

218

205

421 3 3 5

Run

7

5.0

494

465

581

16.0

15

.0

26

9 25

3 72

3 92

228

214

439 6 5 11

Run

6

4.9

503

468

585

16.2

15

.0

29

4 27

3 78

0 92

235

219

449 7 7 13

Run

5

5.1

771

728

910

15.7

14

.8

22

8 21

5 61

6 89

232

219

449 5 4 8

Run

4

5.1

550

519

649

15.6

14

.8

32

1 30

3 86

6 88

253

239

490 11

10

20

Run

3

4.2

532

475

594

16.6

14

.9

44

4 39

7 11

36

88

230

206

422 16

14

28

Run

2

4.8

496

459

574

15.9

14

.8

74

4 68

9 19

70

84

246

228

467 22

20

39

Run

1

5.1

506

477

596

14.8

13

.9

43

46

4097

11

717

8 229

216

443 25

23

45

2 L

imes

tone

Tab

le 4

.7: F

lue

gas e

mis

sion

dat

a.

Com

bust

or O

utle

t

O2,

%

CO

con

tent

, ppm

C

O c

onte

nt1 , p

pm

CO

em

issi

on1 , m

g/N

m3

CO

2 con

tent

, %

CO

2 con

tent

1 , %

SO2 c

onte

nt, p

pm

SO2 c

onte

nt1 , p

pm

SO2 e

mis

sion

1 , mg/

Nm

3 SO

2 ret

entio

n2 , %

NO

con

tent

, ppm

N

O c

onte

nt1 , p

pm

NO

em

issi

on1 , m

g/N

m3

N2O

con

tent

, ppm

N

2O c

onte

nt1 , p

pm

N2O

em

issi

on1 , m

g/N

m3

1 Cor

rect

ed to

6 %

O2

143

SO2 emission is reduced drastically by addition of limestone to lignite in Run 2.

Biomass addition leads to further decrease in SO2 emissions due to negligible sulfur

contents of biomass. Increasing olive residue share from 15 to 49 wt % decreases

SO2 emissions considerably. SO2 emissions from hazelnut shell co-firing runs are

measured to be lower than that of olive residue co-firing runs due to lower sulfur

content of hazelnut shells (0.08 wt %, on dry basis) compared to that of olive

residues (0.14 wt %, on dry basis). Increase in hazelnut share from 11 to 42 wt %

leads to lower SO2 emission due to lower amount sulfur in the fuel feed. SO2

emissions of cotton residues co-firing runs are higher than olive residue and hazelnut

shell co-firing runs due to the same reason explained for high CO emissions.

Sulfur retention efficiencies obtained in all tests are also tabulated in Table 4.7. As

can be seen from the table, as high as 84 % retention efficiency is obtained when

high sulfur content lignite is burned with limestone addition despite the absence of

cyclone ash recycle. This is attributed to increased residence time resulting from

under-bed feeding rather than the reactivity of the limestone utilized as the same

limestone resulted in 69 % retention efficiency when the fuel/limestone mixture was

fed 85 cm above the grid and burned under similar conditions [150].

Co-firing of olive residue results in higher sulfur retention efficiencies compared to

that of lignite (88 % and 84 %, respectively). Increase in olive share in the fuel blend

to 49 wt % leads to further increase in sulfur retention (89 %) due probably to the

additional sulfur capture by the inherent CaO content of biomass [102, 151-153]. Co-

firing of hazelnut shells results in 92 % sulfur retention efficiency irrespective of the

share of hazelnut shell. Increasing the share of cotton residue from 30 to 41 wt %

leads to increase sulfur retention efficiency from 76 to 83 %.

NO emissions stay almost constant during olive residue co-firing runs due to

compensation of high amounts of nitrogen in the fuel blend with coal char and high

volatile matter content. Increasing the hazelnut shell share in the fuel feed leads to

slightly lower NO emissions due to low nitrogen content of hazelnut shell. Co-firing

cotton residue with 30 wt % in the fuel blend leads to lower NO emissions, however,

144

increasing the share to 41 wt % results in higher NO emissions due to increase in

nitrogen content of the fuel blend.

Fuel nitrogen to NO conversion is an important parameter for estimation of NO

emissions. Two different approaches are followed in the literature for determination

of this parameter. The first relates the fuel nitrogen to NO emission to fuel char

combustion [138, 155, 156], the second is just based on NO emission relative to fuel

nitrogen [117]. Both definitions are given below.

2

NO

CFuelCO CO

NFuel

(4.2)CFuel-N to NO conversion, % = 100

N .M × C +CC .M⎛ ⎞ ⎛ ⎞⎜ ⎟ ⎜ ⎟⎜ ⎟ ⎝ ⎠⎝ ⎠

×

where2NO CO CO C ,C and C are the concentrations of NO, CO and CO2 and, CFuel and

NFuel are the carbon and nitrogen contents, respectively. MN and MC are the atomic

masses of nitrogen and carbon, respectively.

(4.3)Nitrogen in the emitted NOFuel-N to NO conversion, % = 100Nitrogen in the fuel ×

Previous studies have shown that fuel nitrogen to NO conversion is a function of

H/N weight ratio [145, 157, 158] as well as the volatile matter content of the fuels

[108, 142, 143] in addition to the operating conditions such as bed temperature,

excess air ratio, etc. In an attempt to see the validity of these correlations for the

results obtained in this thesis study, fuel nitrogen to NO conversions calculated by

using Equations (4.2) and (4.3) together with volatile matter content and H/N ratio

the fuel blends are tabulated in Table 4.8.

145

Table 4.8: Fuel nitrogen to NO conversion.

Volatile matter of fuel feed,

% as received

H/N wt ratio of the fuel feed

Conversion, % (Equation 4.2)

Conversion, % (Equation 4.3)

Run 1 29.8 3.6 7.4 6.8

Run 2 31.1 3.6 7.1 6.8

Run 3 38.2 3.9 5.9 6.0

Run 4 45.9 3.5 5.5 5.8

Run 5 53.4 3.6 4.9 5.3

Run 6 35.8 4.1 7.2 7.4

Run 7 44.4 5.5 7.8 8.9

Run 8 48.3 6.2 8.2 10.1

Run 9 44.3 2.3 2.4 2.9

Run 10 49.5 2.0 3.3 4.0

As can be seen from the table, conversion for olive residue co-firing runs shows no

dependency on H/N ratio but only on the volatile matter content of the fuel blend. On

the other hand, hazelnut shell co-firing runs reveal that conversion increase with H/N

ratio despite increasing volatile matter content of the fuel blend. Cotton residue co-

firing results show no dependency on both parameters. It is worth nothing that

among the three biomasses, cotton residue yields the smallest fuel nitrogen to NO

conversion. These results reveal that predictive accuracy of fuel nitrogen to NO

conversion relationship with volatile matter content and H/N ratio of the fuels is very

poor for systems co-firing fuels with different characteristics.

Inspection of N2O emissions given in Table 4.7 shows that N2O emissions follow a

reducing trend in lignite firing and olive residue co-firing runs. Addition of hazelnut

shells leads to further reduction in N2O emissions; however, cotton residue addition

results in higher emissions due to higher nitrogen content of cotton residue (4.4 wt %

on dry basis).

Although the emissions from a pilot scale test rig are not indicative of the emissions

from commercial size units due to much shorter gas residence times, an attempt was

146

made to find out emission performance of the tests carried out in this study. For co-

firing units emission limits have to be re-defined for different fuel blends. The

emission limits for fuel blends under consideration is calculated by the expression

given in European Directive [159] as;

3

33

Thermal input of biomass (kJ/kg)

× Emission limit of biomass (mg/Nm )

Thermal input of coal (kJ/kg)+

× Emission limit of coal (mg/Nm )Emission limit (mg/Nm ) =

Thermal input

⎧ ⎫⎛ ⎞⎪ ⎪⎜ ⎟⎪ ⎝ ⎠⎪⎨ ⎬

⎛ ⎞⎪ ⎪⎜ ⎟⎪ ⎪⎝ ⎠⎩ ⎭ (4.4)

of biomass (kJ/kg)+Thermal input of coal (kJ/kg)⎧ ⎫⎨ ⎬⎩ ⎭

Gaseous emission limits set by Turkish and European Directives for co-firing units

under consideration are presented in Table 4.9.

Table 4.9: Emission limits set by Turkish and European Union Directives.

Turkish Directive[160] European Union Directive [159]

CO mg/Nm3

SO2 mg/Nm3

NOx mg/Nm3

SO2 mg/Nm3

NOx mg/Nm3

Run 1 200 2000 800 850 400

Run 2 200 2000 800 850 400

Run 3 249 1658 724 726 400

Run 4 294 1347 655 614 400

Run 5 343 1008 579 492 400

Run 6 237 1741 743 757 400

Run 7 290 1376 661 625 400

Run 8 331 1096 599 524 400

Run 9 297 1329 651 608 400

Run 10 324 1140 609 540 400

147

Comparisons between the emission limits given in Table 4.9 and experimental

emissions listed in Table 4.7 shows that CO emissions are above the limits due to

both lack of air staging and short gas residence times in the test rig. SO2 emissions

are found to be lower than the limits set by Turkish Directive except cotton residue

co-firing runs (Runs 9 and 10) due to prolonged lignite firing with insufficient

limestone addition caused by biomass feeding problems encountered before Runs 9

and 10. NO emissions are found to be lower than Turkish emission limits; however,

SO2 and NO emissions are higher than the limits set by European Union Directive.

4.9 Agglomeration and Deposit Formation

Combustion processes with biomass, especially with herbaceous biomasses which

have high alkaline contents are prone to experience agglomeration, fouling and

corrosion [161]. Ash constituents from combustion processes react with flue gas or

with each other to form variety of compounds having low ash melting temperature.

Co-firing biomass with coal on the other hand, reduces risk of operational problem

by introducing the protective compounds such as alumina-silicates in coal which

increases the melting point of the alkaline compounds in the biomass ash [9-11, 42,

45, 47].

Agglomeration tendency is mostly related with melting behavior of the components

in bed material. A bed agglomeration index which expresses the ratio of iron oxides

to sum of potassium and sodium oxides in the fuel ash has been developed, relating

ash composition to agglomerations in fluidized beds [9]. Agglomeration is predicted

to occur when the ratio is less than 0.15. The agglomeration index calculated for fuel

blends in Runs 1-10 are given in Table 4.10. As can be seen from the table, all blends

have low agglomeration tendency.

148

Table 4.10: Bed agglomeration index of fuel blends.

Bed agglomeration index: 2 3

2 2

%(Fe O )BAI =

%(K O+Na O)

Run 1 6.8

Run 2 6.2

Run 3 4.3

Run 4 2.5

Run 5 1.6

Run 6 7.8

Run 7 4.6

Run 8 3.8

Run 9 1.7

Run 10 1.0

Evaluation of agglomeration tendency is also carried out with extracting the melting

behavior data of bottom ash components from ternary phase diagrams. Table 4.11

gives the melting temperatures of bottom ash components of co-firing runs carried

out with highest biomass share (Runs 5, 8 and 10) obtained from different ternary

phase diagrams [162]. As can be seen from the table, melting temperatures of the bed

particles are very high to form agglomerates under fluidized bed combustion

conditions and therefore, during co-firing runs carried out with olive residues,

hazelnut shells and cotton residues, no agglomeration is detected within the bed

material.

149

Table 4.11: Melting temperatures of ternary systems.

Melting Temperature, °C

Ternary System Run 5 (49 wt %

olive residue)

Run 8 (42 wt %

hazelnut shell)

Run 10 (41 wt %

cotton residue)

SiO2-Al2O3-CaO ~1400 Gehlenite

~1440 Gehlenite

~1750 Ca2SiO4

SiO2-Al2O3-K2O ~1700 Mullite

~1700 Mullite

~1700 Mullite

SiO2-Al2O3-Na2O ~1700 Mullite

~1700 Mullite

~1700 Mullite

SiO2-Al2O3-MgO ~1680 Mullite

~1700 Mullite

~1600 Mullite

SiO2-CaO-Fe2O3 ~1470

Pseudo- wollastonite

~1600 Ca2SiO4

~1900 Ca2SiO4

SiO2-CaO-MgO ~1450

Pseudo- wollastonite

~1500 Pseudo-

wollastonite

~1470 Pseudo-

wollastonite

SiO2-CaO-K2O ~1475

Pseudo-wollastonite

~1600 2CaO.SiO2

~1600 2CaO.SiO2

Gehlenite: 2CaO.Al2O3.SiO2,

Mullite: 3Al2O3.2 SiO2,

Pseudo-wollastonite: CaO.SiO2

150

Deposit formation on the heat exchange tubes is another important problem

challenging biomass combustion performance. The feasibility of biomass and coal

co-firing partly depends on its impact on ash deposition since deposits interfere with

operation and eventually lead to corrosion or blockage of gas paths [163].

For predicting the fouling tendency of coals the ratio of sum of potassium and

sodium contents to silica content of the fuel is often used as an index. The ratio

higher than 1, usually indicate fouling problems [11]. Calculated values of base to

acid ratios for fuel blends are given in Table 4.12. As can be seen be seen from the

table, addition of biomass results in no significant increase in the ratio to form

fouling problems.

Table 4.12: Alkali index of fuel blends.

2 2

2

%(K O+Na O)Alkali Index =

%(SiO )

Run 1 0.03

Run 2 0.03

Run 3 0.05

Run 4 0.08

Run 5 0.13

Run 6 0.03

Run 7 0.05

Run 8 0.06

Run 9 0.13

Run 10 0.24

151

As fuel composition and ash melting behaviors are generally insufficient for

prediction of deposit formation and composition [42], an air cooled deposit testing

probe with a detachable ring is designed and constructed for investigation of deposit

formation during co-firing biomass with coal. Deposit probe was placed within port 2

at the combustor height of 4.19 m. The probe was exposed to particle laden flue gas

for about 8 h during olive residue co-firing, 4.5 h during hazelnut shell co-firing and

about 1 h during cotton residue co-firing. Those periods were considered sufficient as

deposit testing probes exposed for 1 to 3 h in full scale boilers can provide reliable

data for prediction of deposit composition [42, 164]. The surface temperature of the

probe is set to 500 °C to represent the surface temperature of the superheater material

at the hottest part of superheater of fluidized bed boilers [164].

Appearance of the deposit rings collected after the co-firing experiments with olive

residue, hazelnut shell and cotton residue are shown in Figures 4.26-4.28,

respectively. As can be seen from the figures no deposit formation took place in the

wind side which implies low risk for fouling for the biomasses under consideration.

Olive residue co-firing runs leads to higher deposit formation compared to hazelnut

shell and cotton residue co-firing runs. Relatively high fouling rate of olive residue

co-firing runs is also confirmed by the calculated values of rate of deposit build-up

(RBU). RBU is defined as the amount of deposit collected per projected surface area

of the probe per unit time (g/m2 h) [24, 165]. RBU values of co-firing runs carried

out with different biomasses are illustrated in Figure 4.29. As can be depicted from

the figure, deposits of olive residue co-firing runs have the higher RBU value than

those of hazelnut shell and cotton residue co-firing runs. Coal is considered to be

non-fouling fuel, therefore, this result may probably due to the higher share of the

olive residue (49 wt %) in the fuel blend compared to shares of hazelnut shell (42 wt

%) and cotton residue (41 wt %) in the fuel blends. RBU value exceeding 20 g/m2h is

usually taken a sign of slagging and fouling problems in the measured area of the

boiler [165, 166]. As all the RBU values are lower than 20 g/m2h, fuel blends can be

considered as non-fouling.

152

Figure 4.26: Appearance of deposit rings after olive residue/lignite co-firing runs.

153

Figure 4.27: Appearance of deposit rings after hazelnut shell/lignite co-firing runs.

154

Figure 4.28: Appearance of deposit rings after cotton residue/lignite

co-firing runs.

155

RB

U (g

/m2 h

)

0

2

4

6

8

10

12

14

Olive residue/lignite Hazelnut shell/lignite Cotton residue/lignite

Figure 4.29: Rate of deposit build-up (RBU, g/m2h)

After each set of the co-firing run, removable ring was separated from the deposit

probe and deposits were scraped off. The deposit powders are coated with carbon

tape and subjected to SEM/EDX analysis. The morphology and size of deposit

crystals observed by SEM at different magnifications are shown in Figures 4.30-

4.38.

As can be seen in the Figures 4.30-4.32, deposits crystals of olive residue co-firing

runs are observed to have irregular shape in the size range of 1 to 3 µm. In Figures

4.33-4.35, deposits from hazelnut shell co-firing runs are observed to have similar

morphology to that of deposits of olive residue co-firing runs. Crystals having

diameter <1 µm and crystals having diameter at around 2-3 µm are observed.

Deposits from cotton residue co-firing runs shown in Figures 4.36-4.38 are observed

to have highly angular structure. The deposit grains are within the size range of 1-5

µm. No agglomerated or molten particles are observed in the deposits.

156

Figure 4.30 SEM micrograph of deposit of olive residue/lignite co-firing runs (a).

Figure 4.31: SEM micrograph of deposit of olive residue/lignite co-firing runs (b).

157

Figure 4.32 SEM micrograph of deposit of olive residue/lignite co-firing runs (c).

Figure 4.33: SEM micrograph of deposit of hazelnut shell/lignite co-firing runs (a).

158

Figure 4.34: SEM micrograph of deposit of hazelnut shell/lignite co-firing runs (b).

Figure 4.35: SEM micrograph of deposit of hazelnut shell/lignite co-firing runs (c).

159

Figure 4.36: SEM micrograph of deposit of cotton residue/lignite co-firing runs (a).

Figure 4.37: SEM micrograph of deposit of cotton residue/lignite co-firing runs (b).

160

Figure 4.38: SEM micrograph of deposit of cotton residue/lignite co-firing runs (c).

The compositions of the deposits analyzed by EDX are shown in Figure 4.39.

Inspection of the figure reveals high concentrations of silicon, calcium, sulfur, iron,

and aluminum in the deposits. Silica and alumina are known as protective coal ash

elements which increase the melting temperatures of the alkalis from biomass ashes

[9-11, 42, 45, 47]. High concentrations of calcium and sulfur suggest formation of

calcium sulfates in the deposits. Potassium concentration in the deposits of olive

residue co-firing runs is found to be higher than potassium content in the other

deposits. This may be attributed to the larger exposure time in olive residue co-firing

test. On the other hand, chlorine concentrations are almost zero in all the deposits

suggesting no formation of alkali chlorides. Therefore, potassium in the deposits

forms potassium sulfates instead of potassium chloride which is a highly fouling

compound leading to corrosion of the superheater tubes of boilers. These findings are

also confirmed with XRD analysis. XRD analysis graph of the deposits of olive

residue and hazelnut shell co-firing are shown in Figure 4.40.

161

Figu

re 4

.39:

Dep

osit

com

posi

tions

.

162

Figu

re 4

.40

: X-r

ay d

iffra

ctio

n pa

ttern

s of t

he d

epos

its.

163

As can be seen from the figure, the deposits are highly crystalline and mainly

composed of calcium sulfate, as estimated from EDX results. Also, in the deposits of

olive residue co-firing runs, potassium sulfate is observed as a minor phase with

calcium sulfate major phase.

As chlorine is not detected in both EDX and XRD analyses of the deposits, the fuel

blends under consideration can be denoted to have low-fouling and corrosion

propensity. Species balance on chlorine carried out for highest share of olive residue

and hazelnut shell co-firing runs (Runs 5 and 8) reveals that 92 % and 86 % of

chlorine is recovered in ash in Runs 5 and 8, respectively. The partitioning of

chlorine in bottom and fly ashes reveal that 100 % and 83 % of chlorine is captured

in fly ashes of Runs 5 and 8, respectively.

164

CHAPTER 5

CONCLUSIONS

5.1 General

In this study, combustion and emission performance of typical Turkish lignite co-

fired with olive residue, hazelnut shell and cotton residue at several shares were

investigated by burning them in their own ashes in 0.3 MWt ABFBC test rig. The

following conclusions were reached under the observations of this study:

• Compared to lignite firing, co-firing of lignite with olive residue, hazelnut

shell and cotton residue results in formation of coarser particles in cyclone

ash.

• SiO2, Al2O3, Fe2O3 concentrations decrease and CaO and SO3 concentrations

increase in bottom, cyclone and baghouse filter ashes during co-firing of

olive residue, hazelnut shell and cotton residue.

• Co-firing shifts major and minor elements partitioning from bottom ash to fly

ash and enhances the partitioning of trace elements to fly ash.

• Co-firing of olive residue and cotton residue increases the combustion

efficiency to 98 % and 97 %, respectively irrespective of biomass share in the

fuel feed. During hazelnut shell co-firing, combustion efficiency is increased

to 98 % with increasing hazelnut shell share to 42 wt %.

• Co-firing of biomasses with lignite leads to slightly higher freeboard

temperatures.

• Co-firing of olive residue results in almost constant O2, CO2 and NO

concentrations, but higher CO and lower N2O and SO2 concentrations across

the freeboard with increasing olive residue share from 15 wt % to 49 wt %.

165

• Co-firing of hazelnut shell results in almost constant O2, CO2 and CO, and

lower SO2 and N2O concentrations across the freeboard at 11 wt % hazelnut

shell share.

• Co-firing of olive residue, hazelnut shell and cotton residue shows no

significant influence on total CO2 emissions, however reduces net CO2

emissions.

• Co-firing of biomasses with lignite leads to higher CO emissions.

• During olive residue co-firing tests, NO emissions stay almost constant while

SO2 and N2O emissions reduce with increasing olive residue share. Hazelnut

shell co-firing results in lower NO, SO2 and N2O emissions with increasing

hazelnut shell share. Cotton residue co-firing, on the other hand, leads to

higher NO and N2O emissions with increasing cotton residue share from 30

wt % to 40 wt %.

• No agglomeration of bed material and fouling problems on heat exchange

surfaces occur during co-firing of lignite with olive residue, hazelnut shell

and cotton residue.

In conclusion, co-firing of high ash and sulfur content, low calorific value Turkish

lignite with olive residue, hazelnut shell and cotton residue is found to be technically

feasible in fluidized bed combustors in terms of combustion and emission

performance, agglomeration and fouling. However, considering the seasonal

availability of these biomasses co-firing application of olive residue, hazelnut shell

and cotton residue with lignite will be more feasible in FBC power plants which are

installed close to the production areas of these residues for reducing pollutant

emissions.

5.2 Suggestions for Future Work

Based on the experience gained in the present study, the following recommendations

for future extension of the work are suggested.

• NO and CO emissions could be lowered by means of staged combustion.

166

• Co-firing experiments could be carried out with recycle of cyclone ash to

provide longer residence time for better utilization of limestone.

• A parametric study with respect to Ca/S ratio and temperature could be

carried out on co-firing of lignite with olive residue, hazelnut shell and cotton

residue for determination of optimum operating conditions.

167

REFERENCES

[1] BP Statistical Review of World Energy June 2007,

http://www.bp.com/statisticalreview, (Last accessed date: 26.08.07). [2] Sami M., Annamalai K., Wooldridge M., “Co-firing of Coal and Biomass

Fuel Blends”, Progress in Energy and Combustion Science, Vol. 27, 171-214, 2001.

[3] Sheinkopf K., “Global Biomass Opportunities”, Independent Energy (USA),

Vol. 23, No. 8, 50-53, 1993. [4] Grassi G., “The Biomass Research and Development Program of the

Commission of the European Communities”, Klass D. L. (Editor), Energy from Biomass and Wastes, Institute of Gas Technology, Chicago, IL, USA, Vol. 16, 15-37, 1993.

[5] International Energy Agency Statistics 2004,

http://www.iea.org/Textbase/stats/index.asp, (Last accessed date: 26.08.07). [6] Advance Techniques for Generation of Energy from Biomass and Waste,

http://www.ecn.nl/fileadmin/ecn/units/bio/Overig/pdf/Biomassa_voordelen.pdf, (Last accessed date: 26.08.07).

[7] International Energy Outlook 2007,

http://www.eia.doe.gov/oiaf/ieo/graphic_data_ electricity.html., (Last accessed date: 26.08.07).

[8] Baxter L., Koppejan J., “Biomass-Coal Co-combustion: Opportunity for

Affordable Renewable Energy”, Euro Heat & Power, Vol. 1, 2004. [9] Bapat D.W, Kulkarni S.V., Bhandarkar V. P., “Design and Operating

Experience on Fluidized Bed Boiler Burning Biomass Fuels with High Alkaline Ash”, in Proceedings of 14th International Conference on Fluidized Bed Combustion (Ed. Preto F.D.S.), Vol. 1, 165-174, ASME, Vancouver, Canada, May 11-14, 1997.

[10] Hughes E.E., Tillman D.A., “Biomass Co-firing: Status and Prospects

1996”, Fuel Processing Technology, Vol. 54, 127-142, 1998. [11] Guanyi C., Mengxiang F., Zhongyang L., Xuantian L., Zhenglun S., Kefa

C., Mingjiang N., Experimental Research on Rice Husk Combustion in

http://www.bp.com/statisticalreview

http://www.iea.org/Textbase/stats/index.asp

http://www.eia.doe.gov/oiaf/ieo/graphic_data_ electricity.html

168

CFB Boiler and The Design of a 35 T/H Rice Husk Fired CFB Boiler, in Proceedings of 14th International Conference on Fluidized Bed Combustion (Ed. Preto F.D.S.), Vol. 1, 175-181, ASME, Vancouver, Canada, May 11-14, 1997.

[12] Anthony E.J., “Fluidized Bed Combustion of Alternative Solid Fuels:

Status, Success and Problems of the Technology, Progress in Energy and Combustion Science, Vol. 21, 239–268, 1995.

[13] “International Energy Agency Co-firing Database”, http://www.ieabcc.nl/.

(Last accessed date: 26.08.07).

[14] Statistical Yearbook of Turkey 2006, http://www.turkstat.gov.tr/yillik/yillik_ing.pdf, (Last accessed date: 26.08.07).

[15] http://www.fiskobirlik.org.tr/, (Last accessed date: 26.08.07). [16] Exploitation of Agricultural Residues in Turkey, http://www.agrowaste-

tr.org/ablife/index.php?lang=en/, (Last accessed date: 26.08.07).

[17] Easterly J.L., Burnham M., “Overview of Biomass and Waste Fuel Resources for Power Production”, Biomass and Bioenergy, Vol. 10, 79-92, 1996.

[18] Werther J., Saenger M., Hartge E.U., Ogada T., Siagi Z., “Combustion of

Agricultural Residues”, Progress in Energy and Combustion Science, Vol. 26, 1-27, 2000.

[19] Arvelakis S., Vourliotis P., Kakaras E., Koukios E.G., “Effect of Leaching

on the Ash Behavior of Wheat Straw and Olive Residue During Fluidized Bed Combustion”, Biomass and Bioenergy, Vol. 20, 459-470, 2001.

[20] Tomishige K., Asadullah M., Kunimori K., “Syngas Production by Biomass

Gasification Using Rh/CeO2/SiO2 Catalysts and Fluidized Bed Reactor”, Catalysis Today, Vol. 89, 389-403, 2004.

[21] Kuprianov V. I., Janvijitsakul K., Permchart W., “Co-firing of Sugarcane

Bagasse with Rice Husk in a Conical Fluidized Bed Combustor”, Fuel, Vol. 85, 434-442, 2006.

[22] Özbay N., Pütün A.E., Uzun B. B., Pütün E., “Biocrude from Biomass:

Pyrolysis of Cottonseed Cake, Renewable Energy, Vol. 24, 615-625, 2001. [23] Midilli, A. , Murat Dogru M., Howarth C.R., Ayhan T., “Hydrogen

Production from Hazelnut Shell by Applying Air-blown Downdraft Gasification Technique”, International Journal of Hydrogen Energy, Vol. 26, 29-37, 2001.

http://www.ieabcc.nl/

http://www.turkstat.gov.tr/yillik/yillik_ing.pdf

http://www.fiskobirlik.org.tr/

http://www.agrowaste-tr.org/ablife/index.php?lang=en/

http://www.agrowaste-tr.org/ablife/index.php?lang=en/

169

[24] Theis M., Skrifvars B. J., Hupa M., Tran H., “Fouling Tendency of Ash

Resulting from Burning Mixtures of Bio fuels. Part 1: Deposition Rates”, Fuel, Vol. 85, 1125-1130, 2006.

[25] Senneca O., “Kinetics of Pyrolysis Combustion and Gasification of Three

Biomass Fuels”, Fuel Processing Technology, Vol. 88, No. 1, 87-97, 2007. [26] Kouvo P., Backman R., “Estimation of Trace Element Release and

Accumulation in the Sand Bed during Fluidized Bed Co-combustion of Biomass, Peat and Refused-derived Fuels”, Fuel, Vol. 82, 741-753, 2003.

[27] Scala F., Chirone R., “Fluidized Bed Combustion of Alternative Fuels”,

Experimental Thermal and Fluid Science, Vol. 28, No. 7, 691-699, 2004. [28] Selcuk N., Gogebakan Y., Altindag H., “Co-firing of Steam Coal with High

Sulfur Content Lignite in a Bubbling Fluidized Bed Combustor”, in Proceedings of 18th International Conference on Fluidized Bed Combustion (Ed. Jia L.), FBC2005-067 in CD-ROM, ASME, Toronto, Ontario, Canada, May 22-25, 2005.

[29] Altindag H., Gogebakan Y., Selcuk N., “Sulfur Capture for Fluidized-Bed

Combustion of High-Sulfur Content Lignites”, Applied Energy, Vol. 79, 403-424, 2004.

[30] Boavida D., Abelha P., Gulyurtlu I., Cabrita I., “Co-Combustion of Coal

and Non-Recyclable Paper and Plastic Waste in a Fluidized Bed Reactor”, Fuel, Vol. 82, 1931-1938, 2003.

[31] Miccio F., Scala F., Chirone R., “Fluidized Bed Combustion of a Biomass

Fuel: Comparison between Pilot Scale Experiments and Model Simulations”, Journal of Heat Transfer, Vol. 127, 117-122, 2005.

[32] Albina D. O., “Emissions from Multiple-Spouted and Spouted Fluidized

Beds Using Rice Husks as Fuel”, Renewable Energy, Vol. 31, 2152-2163, 2006.

[33] Nielsen H. P., Baxter L. L., Sclippab G., Morey C., Frandsen F. J.,

“Deposition of Potassium Salts on Heat Transfer Surfaces in Straw Fired Boilers: A Pilot Scale Study”, Fuel, Vol. 79, 131-139, 2000.

[34] Bakker R. R., Jenkins B. M., Williams R. B., “Fluidized Bed Combustion

of Leached Rice Straw”, Energy & Fuels, Vol. 16, 356-365, 2002. [35] Miles T. R., Miles T. R. Jr., Baxter L. L., Bryers B. M., Jenkins B. M.,

Oden L. L., “Alkali Deposits Found in Biomass Power Plants: A Preliminary Investigation of Their Extent and Nature”, National Renewable Energy Laboratory, Golden, CO, 1995.

170

[36] Demirbas A., Ilten N., “Fuel Analysis and Thermochemical Processing of

Olive Residues”, Energy Sources, Vol. 26, 731-738 2004. [37] Arvelakis S., Gehrmann H., Beckman M., Koukios E. G., “Studying Ash

Behavior of Agricultural Residues Using Thermal Analysis”, Journal of Thermal Analysis and Calorimetry, Vol. 72, 1019-1030, 2003.

[38] Robinson A. L. , Junker H. , Baxter L. L., “Pilot-scale Investigation of the

Influence of Coal-Biomass Co-firing on Ash Deposition”, Energy & Fuels, Vol. 16, No. 2, 343-355, 2002.

[39] Hulkkonen S., Fabritius M., Enestam S., “Application of BFB Technology

for Biomass Fuels: Technical Discussion and Experiences from Recent Projects”, in Proceedings of 17th International Conference on Fluidized Bed Combustion (Ed. Pisupati, S.), FBC2003-132 in CD-ROM, ASME, Jacksonville, Florida, USA, May 18-21, 2003.

[40] Visser H. J.M., Van Lith S.C., Kiel J.H.A., “Biomass Ash–Bed Material

Interactions Leading to Agglomeration in FBC”, in Proceedings of 17th International Conference on Fluidized Bed Combustion (Ed. Pisupati, S.), FBC2003-150 in CD-ROM, ASME, Jacksonville, Florida, USA, May 18-21, 2003.

[41] Baxter L.L., Miles T.R., Miles Jr. T.R., Jenkins B. M., Milne T., Dayton D.,

Bryers R.W., Oden L.L., “The Behavior of Inorganic Material in Biomass-Fired Power Boilers: Field and Laboratory Experiences”, Fuel Processing Technology, Vol. 54, 47-78, 1998.

[42] Aho M., “Reduction of Chlorine Deposition in FB Boilers with Aluminum

Containing Additives”, Fuel, Vol. 80, 1943-1951, 2001.

[43] Nielsen H. P., Frandsen F.J., Dam-Johensen K., “Lab-Scale Investigations of High Temperature Corrosion Phenomena in Straw-Fired Boilers”, Energy & Fuels, Vol. 13, 1114-1121, 1999.

[44] Coda B., Aho M., Berger R., Hein K. R. G, “Behavior of Chlorine and

Enrichment of Risky Elements in Bubbling Fluidized Bed Combustion of Biomass and Waste Assisted by Additives”, Energy & Fuels, Vol. 15, 680-690, 2001.

[45] Lisa K., Lu Y., Salmenoja K., “Sulfation of Potassium Chloride at

Combustion Conditions”, Energy & Fuels, Vol. 13, 1184-1190, 1999. [46] Salmenoja K., “Field and Laboratory Studies an Chlorine Induced

Superheater Corrosion in Boilers Fired with Biofuels”, Ph.D. Thesis, Faculty of Chemical Engineering, Åbo Akademi University, Turku, 102, 2000.

http://www.osti.gov/energycitations/searchresults.jsp?Author=Robinson,%20A.L.

http://www.osti.gov/energycitations/searchresults.jsp?Author=Junker,%20H.

http://www.osti.gov/energycitations/searchresults.jsp?Author=Baxter,%20L.L.

171

[47] Hansen L. A., Frandsen F. J., Dam-Johansen K., Sorensen H. S., Skrifvars

B. J., “Characterization of Ashes and Deposits from High-Temperature Coal-Straw Co-firing”, Energy and Fuels, Vol. 13, 803-816, 1999.

[48] Biomass Co-firing: An Efficient Way to Reduce Greenhouse Gas

Emissions,http://ec.europa.eu/energy/res/sectors/doc/bioenergy/cofiring_eu_bionet.pdf, (Accession date: 25/08/07).

[49] Preto F., Antony E.J., Desai D.L., Friedrich F.D., “Combustion Trials of

Rice Hulls in a Pilot Fluidized Bed”, in Proceedings of 9th International Conference on Fluidized Bed Combustion (Ed. Mustonen J.P.), Vol 2, 1123-1128, ASME, Boston, Massachusetts, USA, May 3-7, 1987.

[50] Armesto L., Bahillo A., Veijonen K., Cabanillas A., Otero J., “Combustion

Behavior of Rice Husk in a Bubbling Fluidized Bed”, Biomass and Bioenergy, Vol. 23, 171-179, 2002.

[51] Fang M., Yang L., Chen G., Shi Z., Luo Z., Cen K., “Experimental Study

on Rice Husk Combustion in a Circulating Fluidized Bed”, Fuel Processing and Technology, Vol. 85, 1273-1282, 2004.

[52] Permchart W., Kouprianov V.I., “Emission Performance and Combustion

Efficiency of a Conical Fluidized-Bed Combustor Firing Various Biomass Fuels”, Biosource Technology, Vol. 92, 83-91, 2004.

[53] Kuprianov V. I., Janvijitsakul K., Permchart W., “Co-firing Sugar Cane

Bagasse with Rice Husk in a Conical Fluidized-Bed Combustor”, Fuel, Vol. 85, 434-442, 2006.

[54] Shimizu T., Nemoto T., Tsuboi H., “Rice Husk Combustion in a BFBC

Using Porous Bed Material”, in Proceedings of 18th International Conference on Fluidized Bed Combustion (Ed. Jia L.), FBC2005-28 in CD-ROM, ASME, Toronto, Ontario, Canada, May 22-25, 2005.

[55] Natarajan E., Öhman M., Gabra M., Nordin A., Lilidahl T., Rao A.N.,

“Experimental Determination of Bed Agglomeration Tendencies of Some Common Agricultural Residues in Fluidized Bed Combustion and Gasification”, Biomass and Bioenergy, Vol. 15, 163-169, 1998.

[56] Liu H., Lin Z., Liu D., Wu W., Combustion Characteristics of Rice Husk in

Fluidized Beds, in Proceedings of 13th International Conference on Fluidized Bed Combustion (Ed. Heinschel K.J.), Vol. 1, 615-618, ASME, Orlando, Florida, USA, May 7-10, 1995.

[57] Skrifvars B. J., Yrjas P., Kinni J., Siefen P., Hupa M., “The Fouling of Rice

Husk Ash in Fluidized-Bed Combustion. 2. Pilot-Scale and Full-Scale Measurements”, Energy & Fuels, Vol. 19, 1512-1519, 2005.

172

[58] Bakker R.R., Jenkins B. M., Williams R. B., “Fluidized Bed Combustion of

Leached Rice Straw”, Energy & Fuels, Vol. 16, 356-365, 2002.

[59] Tillman D.A., Rossi A.J., Kitto.W.D., “Wood Combustion: Principles”, Processes and Economics, New York: Academic Press , 1981.

[60] Standard Test Method for Fusibility of Coal Coke Ash, Designation: D

1857-87, 1994.

[61] Wall T.F., Creelman R.A., Gupta R.P., Gupta S., Colin C., Lowe A., “Coal Ash Fusion Temperature- New Characterization Technique, and Associates with Phase Equilibria, In Application of Advanced Technology to Ash-Related Problems in Boilers”, Ed. Baxter L., and Desollar R., Waterville Valley, 541-556, 1995.

[62] Coin C.D.A., Kahraman H., Reifenstein A.P., “An Improve Ash Fusion

Test”, in Proceeding of Eng. Found. Ash Conference, Waterville Valley, 1995.

[63] Halder P.K., Basu P., “The Temperature of Burning Carbon Particles in a

Fast Fluidized Bed of Fine Particles”, Chemical Engineering Communication, Vol. 104, 245-255, 1991.

[64] Gerald P.H., Huggins F.E., Dunmyre G.R., “Investigation of High

Temperature Behavior of Coal Ash in Reducing and Oxidizing Atmosphere”, Fuel, Vol. 60, 585-597, 1981.

[65] Hansen L. A., Frandsen F. J., Dam-Johansen K., Sorensen H. S., Skrifvars

B. J., “Characterization of Ashes and Deposits from High-Temperature Coal-Straw Co-firing”, Energy and Fuels, Vol. 13, 803-816, 1999.

[66] Biomass Co-firing: An Efficient Way to Reduce Greenhouse Gas

Emissions,http://ec.europa.eu/energy/res/sectors/doc/bioenergy/cofiring_eu_bionet.pdf. (Last accessed date: 26.08.07).

[67] Lin W., Johansen K.D., Frandsen F., “Agglomeration in Bio-fuel Fired

Fluidized Bed Combustors”, Chemical Engineering Journal, Vol. 96, 171-185, 2003.

[68] Lin W., Krusholm G., Dam-Johansen K., Musahl E., Bank L.,

Agglomeration Phenomena in Fluidized Bed Combustion of Straw, in Proceedings of 14th International Conference on Fluidized Bed Combustion (Ed. Preto F.D.S.), Vol. 2, 831-837, ASME, Vancouver, Canada, May 11-14, 1997.

[69] Skrifvars B. J., Öhman M., Nordin A., Hupa M., “Predicting Bed

Agglomeration Tendencies for Biomass Fuels Fired in FBC Boilers: A

173

Comparison of Three Different Prediction Methods”, Energy & Fuels, Vol. 13, 259-363, 1999.

[70] Öhman M., Nordin A., Skrifvars B. J., Backman R., Hupa M., “Bed

Agglomeration Characteristics during Fluidized Bed Combustion of Biomass Fuels”, Energy & Fuels, Vol. 14, 169-178, 2000.

[71] Öhman M., Nordin A., “The role of Kaolin in Prevention of Bed

Agglomeration during Fluidized Bed Combustion of Biomass Fuels”, Energy & Fuels, Vol. 14, 618-624, 2000.

[72] Brus E., Öhman M., Nordin A., Boström D., “Bed Agglomeration

Characteristics of Biomass Fuels Usinf Blast-Furnace Slag as Bed Material”, Energy & Fuels, Vol. 18, 1187-1193, 2004.

[73] Nielsen H. P., Frandsen F. J., Dam-Johansen K., “Lab-Scale Investigations

of High-Temperature Corrosion Phenomena in Straw-Fired Boilers”, Energy & Fuels, Vol. 13, 1114-1121, 1999.

[74] Senior C.L., Srinivasachar S., “Viscosity of Ash Particles in Combustion

Systems for Prediction of Particle Sticking”, Energy & Fuels, Vol. 9, 277-283, 1995.

[75] Steenari B. M., Lindqvist O., “High-Temperature Reactions of Straw Ash

and the Anti-Sintering Additives Kaolin and Dolomite”, Biomass & Bioenergy, Vol. 14, 67-76, 1998.

[76] Uberoi M., Punjak W. A., Shadman F., “The Kinetics and Mechanism of

Alkali Removal from Flue Gases by Solid Sorbents”, Progress in Energy and Combustion Science, Vol. 16, No. 4, 205-211, 1990.

[77] Koukios E. G., Arvelakis S., NTUA Group, FIBEGAR Final Report,

European Commission, JOULE Program (JOR3-CT95-0021), 1998.

[78] Khraisha Y.H., Hamdan M.A., Qalalweh H.S., “Direct Combustion of Olive Cake Using Fluidized Bed Combustor”, Energy Sources, Vol. 21, 319-327, 1999.

[79] Topal H., Atımtay A.T., Durmaz A., “Olive Cake Combustion in a

Circulating Fluidized Bed”, Fuel, Vol. 82, 1049-1056, 2003. [80] Cammarota A., Chirone R., Scala F., “Bed Agglomeration during Fluidized

Bed Combustion of Olive Husk”, in Proceedings of 18th International Conference on Fluidized Bed Combustion (Ed. Jia L.), FBC2005-098 in CD-ROM, ASME, Toronto, Ontario, Canada, May 22-25, 2005.

174

[81] Scala F., Chirone R., “Characterization and Early Detection of Bed Agglomeration during the Fluidized Bed Combustion of Olive Husk”, Energy & Fuels, Vol. 20, 120-132, 2006.

[82] Rasul M.G., Rudolph V., “Fluidized Bed Combustion of Australian

Bagasse”, Fuel, Vol. 79, 123-130, 2000. [83] Kuprianov V.I., Permchart W., Janvijitsakul K., “Fluidized Bed

Combustion of Pre-Dried Thai Bagasse”, Fuel Processing Technology, Vol. 86, 849-860, 2005.

[84] Skrifvars B.J., Backman R., Hupa M., Sfiris G., Åbyhammar T., Lyngfelt

A., “Ash Behavior in a CFB Boiler during Combustion of Coal, Peat or Wood”, Fuel, Vol. 77, 65-70, 1998.

[85] Lyngfelt A., Lecknere B., “Combustion of Wood-Chips in Circulating

Fluidized Bed Boilers- NO and CO Emissions as Functions of Temperature and Air-Staging, Fuel, Vol. 78, 1068-1072, 1999.

[86] Preto F., “Combustion of Wood Processing Residues in a Circulating

Fluidized Bed”, in Proceedings of 17th International Conference on Fluidized Bed Combustion (Ed. Pisupati, S.), FBC2003-171 in CD-ROM, ASME, Jacksonville, Florida, USA, May 18-21, 2003.

[87] Miccio F., Scala F., Chirone R., “FB Combustion of a Biomass Fuel:

Comparison between Pilot Scale Experiments and Model Simulations”, in Proceedings of 17th International Conference on Fluidized Bed Combustion (Ed. Pisupati, S.), FBC2003-033 in CD-ROM, ASME, Jacksonville, Florida, USA, May 18-21, 2003.

[88] Chirone R., Miccio F., Scala F., “Mechanism and Prediction of Bed

Agglomeration during Fluidized Bed Combustion of a Biomass Fuel: Effect of the Reactor Size”, Chemical Engineering Journal, Vol. 123, 71-80, 2006.

[89] Tranvik A. C., Öhman M., Sanati M., “Bed Material Deposition in

Cyclones of Wood Fuel Fired Circulating Fluidized Beds (CFBCs)”, Energy & Fuels, Vol. 21, 104-109, 2007.

[90] Wòjtowicz M. A., Pels J. R., Moulijn J. A., “Combustion of Coal as a

Source of N2O Emission”, Fuel Processing Technology, Vol. 34, 1-71, 1993.

[91] Leckner B., Karlsson M., “Gaseous Emissions from Circulating Fluidized

Bed Combustion of Wood”, Biomass & Bioenergy, Vol. 4, 379-389, 1993.

175

[92] Lyngfelt A., Åmand L. E., Leckner B., “Reversed Air Staging - A Method for Reduction of N2O Emissions from Fluidized Bed Combustion of Coal”, Fuel, Vol. 77, 953-959, 1998.

[93] Environment Canada 1989, Reference Method for Source Testing:

Measurement of Releases of Semi-Volatile Organic Compounds from Stationary Sources, EPS1/RM/2, 1989.

[94] Fiorentino M., Marzocchella A., Salatino P., “Segregation of Fuel Particles

and Volatile Matter during Devolatilization in a Fluidized Bed Reactor - II. Experimental”, Chemical Engineering Science, Vol. 52, 1909-1922, 1997.

[95] Llorente M. J. F., Cuadrado R. E., Laplaza J. M. M., García J. E. C.,

“Combustion in a Bubbling Fluidized Bed with Bed Material of Limestone to Reduce Biomass Ash Agglomeration and Sintering”, Fuel, Vol. 85, 2081-2092, 2006.

[96] Shen B.X., Mi T., Liu D.C., Feng B., Yao Q., Winter F., “N2O Emission

under Fluidized Bed Combustion Condition”, Fuel Processing Technology, Vol. 84, 13-21, 2003.

[97] Lin W., Johnsson J.E., Jensen A.D., Dam-Johansen K., Combustion of

Biomass in Fluidized Beds- Problems and Some Solutions Based on Danish Experiences, in Proceedings of 17th International Conference on Fluidized Bed Combustion (Ed. Pisupati, S.), FBC2003-124 in CD-ROM, ASME, Jacksonville, Florida, USA, May 18-21, 2003.

[98] Lin W., Johansen K.D., “Agglomeration in Fluidized Bed Combustion of

Biomass-Mechanisms and Co-Firing with Coal”, in Proceedings of 15th International Conference on Fluidized Bed Combustion (Ed. Reuther R.B.), FBC99-0120 in CD-ROM, ASME, Savannah, Georgia, May 16-19, 1999.

[99] Sandelin K., Coda B., Backman R., Berger R., Hein K.R.G., “Melting of

Ash Components When Co-Firing Coal, Straw and Paper Sludge under Bubbling Fluidized Bed Conditions”, in Proceedings of 16th International Conference on Fluidized Bed Combustion (Ed. Geiling, D. W.), FBC01-129 in CD-ROM, ASME, Reno, Nevada, USA, May 13-16, 2001.

[100] Glazer M.P., Schürmann H., Monkhouse P., Jong W., Spliethoff H., “Co-

Combustion of Coal with High Alkali Straw, Measuring of Gaseous Alkali Metals and Sulfur Emissions Monitoring”, in Proceedings of 8th International Conference on Circulating Fluidized Beds”, (Ed. Cen, K.), in CD-ROM, Hangzhou, China, May 10-13, 2005.

[101] Kassman H., Anderson C., Högberg J., Ǻmand L. E., Davidsson K., “Gas

Phase Alkali Chlorides and Deposits during Co-Combustion of Coal and Biomass”, in Proceedings of 19th International Conference on Fluidized

176

Bed Combustion (Ed. Winter F.), FBC2006-99.F in CD-ROM, Vienna, Austria, May 21-24, 2006.

[102] Okasha F., “Enhancing Sulphur Self-Retention by Building-in CaO in

Straw-Bitumen Pellets”, Fuel Processing Technology, Vol. 88, 401-408, 2007.

[103] Hansen P.F.B., Lin W., Dam-Johansen K., “Chemical Reaction Conditions

in a Danish 80 MWth CFB-Boiler Co-Firing Straw and Coal”, in Proceedings of 14th International Conference on Fluidized Bed Combustion (Ed. Preto F.D.S.), Vol. 2, 831-837, ASME, Vancouver, Canada, May 11-14, 1997.

[104] Henderson P., Kassman H., Andersson C., “The Use of Ammonium

Sulphate in Waste-Fired Boilers to Reduce Superheater Fouling and Corrosion and NOx emissions, Proc Swedish Finnish Flame Days, Borås, Sweden, 2005.

[105] Johnsson J. E., “Formation and Reduction of Nitrogen Oxides in Fluidized-

Bed Combustion”, Fuel, Vol. 9, 1398-1415, 1994. [106] Leckner B., Amand L. E., “Emissions from a Circulating and Stationary

Fluidized Boiler: A Comparison”, in Proceedings of 9th International Conference on Fluidized Bed Combustion (Ed. Mustonen J. P.), 891-897, ASME, Boston, Massachusetts, USA, May 3-7, 1987.

[107] Armesto L., Veijonen K., Bahillo A. Cabanillas A., Plumed A., Salvador

L., “Co-Combustion of Coal and Biomass Wastes in Fluidized Bed”, in Proceedings of 16th International Conference on Fluidized Bed Combustion (Ed. Geiling D.W.), FBC01-0044 in CD-ROM, ASME, Reno, Nevada, USA, May 13-16, 2001.

[108] Armesto L., Bahillo A., Cabanillas A., Veijonen K., Otero J., Plumed A.,

Salvador L., “Co-Combustion of Coal and Olive Oil Industry Residues in Fluidised Bed”, Fuel, Vol. 82, 993-1000, 2003.

[109] Cliffe K.R., Patumsawad S., “Co-Combustion of Waste from Olive Oil

Production with Coal in a Fluidised Bed”, Waste Management, Vol. 21, 49-53, 2001.

[110] Atimtay A.T., Topal H., “Co-Combustion of Olive Cake with Lignite Coal

in a Circulating Fluidized Bed”, Fuel, Vol. 83, 859-867, 2004. [111] Hamalainen J.P., Aho M.J., “Formation of Nitrogen Oxides from Fuel-N

through HCN and NH3: A Model-Compound Study”, Fuel, Vol. 73, 1894-1898, 1994.

177

[112] Vitolo S., Petarca L., Bresci B., “Treatment of Olive Oil Industry Wastes”, Biosource Technology, Vol. 67, 129-137, 1999.

[113] Armesto L., Cabanillas A., Bahillo A., Segovia J.J, Escalada R., Martinez

J.M., Carrasco J.E., “Coal and Biomass Co-Combustion on Fluidized Bed: Comparison of Circulating and Bubbling Fluidized Bed Technologies”, in Proceedings of 14th International Conference on Fluidized Bed Combustion (Ed. Preto F.D.S.), Vol. 1, 301-311, ASME, Vancouver, Canada, May 11-14, 1997.

[114] Adànez J., Diego L.F., Gayàn P., Garcio-Labiano F., Cabanillas A., Bahillo

A., “Co-Combustion of Biomass and Coal in Circulating Fluidized Bed. Modeling and Validation”, in Proceedings of 17th International Conference on Fluidized Bed Combustion (Ed. Pisupati, S.), FBC2003-064 in CD-ROM, ASME, Jacksonville, Florida, USA, May 18-21, 2003.

[115] Gayàn P., Adànez J., Diego L.F., Garcia-Labino F., Cabanillas A., Bahillo

A., Aho M., Veijonen K., “Circulating Fluidized Bed Co-Combustion of Coal and Biomass”, Fuel, Vol. 83, 277-286, 2004.

[116] Gani A., Naruse I., “NO and N2O Formation/Decomposition Characteristics

during Co-Combustion of Coal with Biomass”, in Proceedings of 18th International Conference on Fluidized Bed Combustion (Ed. Jia L.), FBC2005-024 in CD-ROM, ASME, Toronto, Ontario, Canada, May 22-25, 2005.

[117] Leckner B., Karlsson M., “Emissions from Circulating Fluidized Bed

Combustion of Mixtures of Wood and Coal”, in Proceedings of 12th International Conference on Fluidized Bed Combustion, (Ed. Rubow L.), Vol. 1, 109-115, ASME, San Diego, California, USA, May 9-13, 1993.

[118] Kakaras E., Vourliotis P., Grammelis P., Co-combustion of Lignite with

Waste Wood in a Lab-scale Fluidized Bed, in Proceedings of 16th International Conference on Fluidized Bed Combustion (Ed. Geiling D. W.), FBC2001-004 in CD-ROM, ASME, Reno, Nevada, USA, May 13-26, 2001.

[119] Yrjas P., Skrifvars B.J., Hupa M., Roppo J., Nylund M., Vainikka P.,

“Chlorine in Deposits during Co-firing Biomass, Peat and Coal in a Full Scale CFBC Boiler”, in Proceedings of 18th International Conference on Fluidized Bed Combustion (Ed. Jia L.), FBC2005-097 in CD-ROM, ASME, Toronto, Ontario, Canada, May 22-25, 2005.

[120] Orjala M., Kärki J., Lind T., Oksa M., Vainikka P., “Deposit Formation

and Hot Corrosion in Biomass Firing- Challenges and Problem Solving in Pilot and Full Scale CFB”, in Proceedings of 8th International Conference on Circulating Fluidized Bed, (Ed. Cen, K.), in CD-ROM, Hangzhou, China, May 10-13, 2005.

178

[121] Leckner B., Karlsson M., “Gaseous Emissions from CFB Combustion of

Wood”, Biomass and Bioenergy, Vol. 4, 379-389, 1993.

[122] Manninen H., Perkio A., Palonen J., Peltola K., Ruuskanen J., “Trace Metal Emissions from Co-Combustion of Refuse Derived and Packaging Derived Fuels in a Circulating Fluidized Bed Boiler”, Chemosphere, Vol. 32, 2457-2469, 1996.

[123] Rupar K., Sanati M., “The Release of Terpenes during Storage of

Biomass”, Biomass and Bioenergy, Vol. 28, 29-34, 2005. [124] Åmand L. E., Westberg H. M., Karlsson M., Leckner B., Coda B., Hocquel

M., Berger R., Hein K. R. G., Feng X., Abul Milh M., “Co-Combustion of Dried Sewage Sludge and Coal - The Fate of Heavy Metals”, in Proceedings of 16th International Conference on Fluidized Bed Combustion (Ed. Geiling, D. W.), FBC01-108 in CD-ROM, ASME, Reno, Nevada, USA, May 13-16, 2001.

[125] Åmand L. E., Leckner B., “Metal Emissions from Co-Combustion of

Sewage Sludge and Coal/Wood in Fluidized Bed”, Fuel, Vol. 83, 1803-1821, 2004.

[126] Brendon B. M., Dugwell D. R., Kandiyoti R., “Partitioning of Trace

Elements during Combustion of Coal and Biomass in a Suspension-Firing Reactor”, Fuel, Vol. 81, 2002.

[127] Meij R., “Trace Element Behavior in Coal-Fired Power Plants, Fuel

Processing Technology, Vol. 39, 199-217, 1994.

[128] Xu M., Yan R., Zheng C., Qiaro Y., Han J., Sheng C., “Status of Trace Element Emissions in a Coal Combustion Process: A Review”, Fuel Processing Technology, Vol. 85, 215-237, 2003.

[129] Demir I., Hughes R. E., Maris P. J. D., “Formation and Use of Coal

Combustion Residues from Three Types of Power Plants Burning Illinois Coals”, Fuel, Vol. 80, 659-1673, 2001.

[130] Klika Z., Bartonova L., Spears D. A.,” Effect of Boiler Output on Trace

Element Partitioning during Coal Combustion in Two Fluidised-Bed Power Stations”, Fuel, Vol. 80, 907 -917, 2001.

[131] Åmand L. E., Leckner B., “Metal Emissions from Co-Combustion of

Sewage Sludge and Coal/Wood in Fluidized Bed”, Fuel, Vol. 83, 1803-1821, 2004.

179

[132] Grammelis P., Skodras G., Kakaras E., “Effects of Biomass Co-firing with Coal on Ash Properties. Part 1: Characterization and PSD”, Fuel, Vol. 85, 2310-2315, 2006.

[133] Brown J. A. R., “Overview of Trace Element Partitioning in Flames and

Furnaces of Utility Coal-Fired Boilers”, Fuel Processing Technology, Vol. 39, 139-157, 1994.

[134] Selçuk N., Gogebakan Y., Gogebakan Z., “Partitioning Behavior of Trace

Elements during Pilot-Scale Fluidized Bed Combustion of High Ash Content Lignite”, Journal of Hazardous Materials, Vol. B137, 1698-1703, 2006.

[135] Gulyurtlu I., Lopes M. H. Abelha P., Cabrita I., Oliveira J. F. S., “The

Study of Partitioning of Heavy Metals during Fluidized Bed Combustion of Sewage Sludge and Coal, in Proceedings of 17th International Conference on Fluidized Bed Combustion (Ed. Pisupati, S.), FBC2003-079 in CD-ROM, ASME, Jacksonville, Florida, USA, May 18-21, 2003.

[136] Senneca O., “Kinetics of Pyrolysis, Combustion and Gasification of Three

Biomass Fuels”, Fuel processing Technology, Vol. 88, 87-97, 2007.

[137] Winter F., Wartha C., Hofbauer H., “NO and N2O Formation during the Combustion of Wood, Straw, Malt Waste and Peat”, Bioresource Technology, Vol. 70, No. 1, 39-49, 1999.

[138] Konttinen J., Hupa M., Kallio S., Winter F., Samuelsson J., “NO Formation

Tendency Characterization for Biomass Fuels”, in Proceedings of 18th International Conference on Fluidized Bed Combustion (Ed. Jia L.), FBC2005-78025 in CD-ROM, ASME, Toronto, Ontario, Canada, May 22-25, 2005.

[139] Liu H., Gibbs B. M., “The influence of Calcined Limestone on NOx and

N2O Emissions from Char Combustion in Fluidized Bed Combustors”, Fuel, Vol. 80, 1211-1215, 2001.

[140] Li Z, Lu Q., Na Y., “N2O and NO Emissions from Co-firing MSW with

Coals in a Pilot Scale CFBC”, Fuel Processing Technology, Vol. 85, 1539-1549, 2004.

[141] Zijlma G. J., Gerritsen.G. W., Van Den Bleek C. M, “NOx Formation and

Reuction with NH3 in Fluidized Bed Combustion - The Influence of the O2 Concentration on the Kinetics”, in Proceedings of 15th International Conference on Fluidized Bed Combustion (Ed. Reuther R.B.), FBC99-0168 in CD-ROM, ASME, Savannah, Georgia, May 16-19, 1999.

[142] Shimizu T., Tachiyama Y., Souma M., Inagaki M., “Emission Control of

NOx and N2O of Bubbling Fluidized Bed Combustion”, in Proceedings of

180

11th International Conference on Fluidized Bed Combustion, ASME, 695-700, 1991.

[143] Selçuk N., Batu A., Oymak O., “NOx Emissions from Combustion of High

Sulfur Lignite in an ABFBC Test Rig”, in Proceedings of 17th International Conference on Fluidized Bed Combustion (Ed. Pisupati, S.), FBC2003-109 in CD-ROM, ASME, Jacksonville, Florida, USA, May 18-21, 2003.

[144] Molina A., Eddings E. G., Pershing D.W., Sarofim A.F., “Char Nitrogen

Conversion: Implications to Emissions from Coal-Fired Utility Boilers”, Progress in Energy and Combustion Science, Vol. 26, 507-531, 2000.

[145] Chyang C. S., Wu K. T., Lin C. S., “Emissions of Nitrogen Oxides in a

Vortexing Fluidized Bed Combustor”, Fuel, Vol. 86, 234-243, 2007.

[146] Valk M, “Atmospheric Fluidized Bed Coal Combustion Research Development and Application”, Elsevier, 1995.

[147] Boemer A., Braun A., Renz U., “Emission of N2O from Four Different

large Scale Circulating Fluidized Bed Char Combustion”, in Proceedings of 12th International Conference on Fluidized Bed Combustion, ASME, San Diego, California, USA, 585-597, 1993.

[148] Liu D., Wang J. H., Chen H. P. Zhang S. H., Huang L., Lu J. D., “Emission

Control of N2O by Co-combustion of Coal and Biomass and Narrow Pulse Corona Discharge”, in Proceedings of 15th International Conference on Fluidized Bed Combustion (Ed. Reuther R.B.), FBC99-0064 in CD-ROM, ASME, Savannah, Georgia, May 16-19, 1999.

[149] Baxter L., “Biomass-Coal Co-Combustion: Opportunity for Affordable

Renewable Energy”, Fuel, Vol. 84, 1295-1302, 2005.

[150] Selcuk N., Gogebakan Y., Harmandar H., Altindag H., “Effect of Recycle on Fluidized-Bed Combustion and Emission Characteristics of High-Sulfur Lignite” Combustion Science and Technology, Vol. 176, 959-975, 2004.

[151] Spliethoff H., Hein K. R. G, “Effect of co-combustion of biomass on

emissions in pulverized fuel furnaces”, Fuel Processing Technology, Vol. 54, No. 1-3, 189-205, 1998.

[152] Nordin A., “Optimization of Sulfur Retention in Ash when Co-Combusting

High Sulfur Fuels and Biomass Fuels in a Small Pilot Scale Fluidized Bed”, Fuel, Vol. 74, 615-622, 1995.

[153] Kokko A., Nylung M., “Biomass and Coal Co-Combustion in Utility Scale

- Operating Experience of Alholmens Kraft”, in Proceedings of 18th International Conference on Fluidized Bed Combustion (Ed. Jia L.),

181

FBC2005-035 in CD-ROM, ASME, Toronto, Ontario, Canada, May 22-25, 2005.

[154] Löffler G., Wargadalam V. J., Winter F., “Catalytic Effect of Biomass Ash

on CO, CH4 and HCN Oxidation under Fluidized Bed Combustor Conditions”, Fuel, Vol. 81, 711-717, 2002.

[155] Liu D. C., Wu Z. S., Shen B. X., Lin Z. J., “The Relative Importance of

Char and Volatile Nitrogen on Formation of Nitrous Oxides and Nitric Oxides”, Energy & Fuels, Vol. 13, 1252-1254, 1999.

[156] Goel S. K., Morihara A., Tullin C. J., Sarofim A. F., “Effect of NO and O2

Concentration on N2O Formation during Coal Combustion in a Fluidized Bed”, in Proceedings of 25th International Symposium on Combustion, 1051-1059, 1994.

[157] Hämäläinen J. P., Aho M. J., Tummavuori J. L., “Formation of Nitrogen

Oxides from Fuel-N through HCN and NH3: A Model-Compound Study”, Fuel, Vol. 73, 1894-1898, 1994.

[158] Aho M. J., Rantanen J. T., “Emissions of Nitrogen Oxides in Pulverized

Peat Combustion between 730 and 900 °C”, Fuel, Vol. 68, 586-590, 1989.

[159] Directive 2001/80/EC of the European Parliament and of the Council of 23 October 2001 on the limitation of emissions of certain pollutants into the air from large combustion plants, Official Journal of the European Communities, L 309, 27/11/2001.

[160] Endüstriyel Kaynaklı Hava Kirliliğinin Kontrolü Yönetmeliği, 7 Ekim

2004, Resmi Gazete Sayı 25606, Çevre ve Orman Bakanlığı.

[161] Llorente M. J. F., Cuadrado R. E., Laplaza J. M. M., Garcìa J. E. C., “Combustion in Bubbling Fluidised Bed with Bed Material of Limestone to Reduce Biomass Ash Agglomeration and Sintering”, Fuel, Vol. 85, 2081-2092, 2006.

[162] Osborn E. F., Muan E.F., “Phase Diagrams of Oxide Systems”, American

Ceramic Society and the Edward Orton Jr., Ceramic Foundation, 1960.

[163] Ǻmand L. E., Leckner B., Eskilsson D., Tullin C., “Deposits on Heat Transfer Tubes during Co-combustion of Biofuels & Sewage Sludge”, Fuel, Vol. 85, 1313-1322, 2006.

[164] Enestam S., “Prediction of Ash Behavior & Deposit Formation in FBC of

Biofuel Mixtures”, in Proceedings of 18th International Conference on Fluidized Bed Combustion (Ed. Jia L.), FBC2005-78137 in CD-ROM, ASME, Toronto, Ontario, Canada, May 22-25, 2005.

182

[165] Skrifvars B. J., Yrjas P., Hupa M., Aho M., Silvennoinen J., Eteläaho R.,

Kouki J., Saari K., “Fireside Deposit Formation in Biomass Fired FBC: A Comparison between Tests performed in Three Significantly Different Sized Combustors”, in Proceedings of 17th International Conference on Fluidized Bed Combustion (Ed. Pisupati, S.), FBC2003-074 in CD-ROM, ASME, Jacksonville, Florida, USA, May 18-21, 2003.

[166] Skrifvars B. J., Yrjas P., Laurén T., Hupa M., Dittrich M., “The ǺBo

Akademi Database- Ash Behavior Measurements in Full Scale Boilers”, in Proceedings of 18th International Conference on Fluidized Bed Combustion (Ed. Jia L.), FBC2005-78140 in CD-ROM, ASME, Toronto, Ontario, Canada, May 22-25, 2005.

183

APPENDIX A

WORLDWIDE CO-FIRING POWER PLANTS

Tables A.1, A.2 and A.3 refer to coal-fired pulverized fuel, bubbling and circulating

fluidized bed power plants, respectively. Power plant names written in bold refer to

the plants having commercial co-firing operation. Tests have been performed with

every commercially significant fuel type (lignite, sub-bituminous coal, bituminous

coal, petroleum coke) with every major category biomass (herbaceous and woody

fuel types generated as residues and energy crops) [8]. As can be seen from Table

A.1, the majority of co-firing applications are carried out in pulverized fuel power

plants. Table A.2 and A.3 show that mostly woody type biomass is co-fired with

coals in bubbling and circulating fluidized bed boilers.

184

Co-

fired

Fue

l

Plan

tatio

n fo

rest

w

aste

and

gre

en

was

te

Woo

d w

aste

Woo

d w

aste

Woo

d w

aste

(s

awdu

st, s

havi

ngs)

Woo

d w

aste

(fre

sh

saw

dust

)

Woo

d w

aste

(p

lant

atio

n sa

wm

ill

resi

due

and

cons

truct

ion

and

dem

oliti

on w

aste

tim

ber)

Woo

d w

aste

(s

awdu

st, s

havi

ngs)

Woo

d w

aste

Woo

d ch

ips

Prim

ary

Fuel

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Hea

t/w

t %

-/5

N.A

.

-/5

-/5

-/5

-/5

N.A

.

N.A

.

3/-

Out

put

MW

e

1040

4×12

5

2×66

0

4×50

0

2×66

0

2×50

0

4×35

0

2×35

0

124

Bur

ner

Type

N.A

.

N.A

.

T-fir

ed

T-fir

ed

Wal

l- fir

ed

T-fir

ed

N.A

.

N.A

.

T-fir

ed

Ow

ner

Wes

tern

Pow

er

CS

Ener

gy

Del

ta E

lect

ricity

Mac

quar

ie

Gen

erat

ion

Del

ta E

lect

ricity

Del

ta E

lect

ricity

Stan

wel

l C

orpo

ratio

n

Taro

ng E

nerg

y

Ver

bund

A

ustri

an

Hyd

ropo

wer

AG

Plan

t Nam

e

Muj

a

Swan

bank

B

Val

es P

oint

Lid

dell

Mt P

iper

Wal

lera

wan

g

Stan

wel

l

Taro

ng

St. A

ndrä

Loca

tion

Col

lie

Ipsw

ich,

SE

Que

ensl

and

Lake

Mac

quar

ie,

New

cast

le, N

ew

Sout

h W

ales

Lidd

ell,

New

So

uth

Wal

es

Lith

gow

, New

So

uth

Wal

es

Lith

gow

, N

ew S

outh

W

ales

Roc

kham

pton

Taro

ng

St. A

ndrä

Cou

ntry

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tria

Tab

le A

.1: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

pul

veriz

ed fu

el p

ower

pla

nts [

13].

Con

tinen

t

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tralia

Aus

tralia

Euro

pe

N.A

. : N

ot a

vaila

ble.

185

Co-

fired

Fue

l

Bar

k, sa

wdu

st, w

ood

chip

s

Woo

d ch

ips

Stra

w

Stra

w

Stra

w/m

isca

nthu

s

N.A

.

Stra

w

Stra

w

Woo

d, st

raw

Woo

d

Sew

age

slud

ge

Prim

ary

Fuel

Pulv

eriz

ed

Polis

h ha

rd c

oal

Pulv

eriz

ed

coal

Pulv

eriz

ed

coal

Pulv

eriz

ed

coal

Pulv

eriz

ed

coal

Pulv

eriz

ed

coal

Pulv

eriz

ed

coal

Pu

lver

ized

co

al

Lign

ite

Pulv

eriz

ed

Coa

l Pu

lver

ized

C

oal

Hea

t/w

t %

3/-

N.A

.

20/-

20/-

N.A

.

N.A

.

N.A

.

N.A

.

-/7

N.A

.

N.A

.

Out

put

MW

e

137

540

150

350

250

150

108

100

350

75

Bur

ner

Type

T-fir

ed

T-fir

ed

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

Wal

l-fir

ed

N.A

.

N.A

.

Ow

ner

Ver

bund

A

ustri

an

Hyd

ropo

wer

A

G

Elec

trabe

l

Mid

kraf

t Po

wer

Co

Mid

kraf

t Po

wer

Co

Elsa

m

E2

Mid

kraf

t

N.A

.

N.A

.

N.A

.

N.A

.

Plan

t Nam

e

Bio

coco

mb

Rui

en

Stud

stru

pvae

rket

#1

Stud

stru

pvae

rket

#

4

N.A

.

Ave

døre

N.A

.

Bay

ernw

erke

A

G

VEA

G

VEA

G

Saar

berg

wer

ke A

G

Loca

tion

Zeltw

eg

Rui

en

Aar

hus

Aar

hus

Am

ager

Cop

enha

gen

Esjb

erg

Bav

aria

Lübb

enau

Mag

debu

rg

Saar

berg

Cou

ntry

Aus

tria

Bel

gium

Den

mar

k

Den

mar

k

Den

mar

k

Den

mar

k

Den

mar

k

Ger

man

y

Ger

man

y

Ger

man

y

Ger

man

y

Tab

le A

.1: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

pul

veriz

ed fu

el p

ower

pla

nts [

13] (

cont

’d).

Con

tinen

t

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

N.A

. : N

ot a

vaila

ble.

186

Co-

fired

Fue

l

Woo

d

Woo

d pe

llets

, oliv

e w

aste

Peat

, woo

d

Sew

age

slud

ge

Ker

nels

, pap

er

slud

ge, s

hells

, fib

ers

Was

te w

ood

Pape

r slu

dge

Bio

mas

s pel

lets

(p

aper

slud

ge,

com

post

resi

due)

Pulv

eriz

ed w

ood

Prim

ary

Fuel

Pu

lver

ized

C

oal

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Hea

t/w

t %

N.A

.

100/

-

N.A

.

-/4

N.A

.

-/8

N.A

.

N.A

.

N.A

.

Out

put

MW

e

180

279

320

600

403

600

600

2×51

8

602

Bur

ner

Type

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

Ow

ner

Fortu

m

Pow

er &

H

eat A

B

Upp

sala

En

ergi

e A

B

NU

ON

EPZ

Esse

nt

Esse

nt

EON

Elec

trabe

l

Plan

t Nam

e

Vas

tham

nsve

rt C

HP

Has

selb

yvae

rket

Upp

sala

Ene

rgie

Hem

weg

cent

rale

8

Bor

ssel

e 12

Am

erce

ntra

le 9

Am

erce

ntra

le 8

Maa

svla

ktec

entra

le

1+2

Gel

derl

and

Loca

tion

Hal

sing

bour

gi

Stoc

khol

m

Upp

sala

Am

ster

dam

Bor

ssel

e

Gee

rtrui

denb

erg

Gee

rtrui

denb

erg

Maa

svla

kte,

R

otte

rdam

Nijm

egen

Cou

ntry

Swed

en

Swed

en

Swed

en

The

Net

herla

nds

The

Net

herla

nds

The

Net

herla

nds

The

Net

herla

nds

The

Net

herla

nds

The

Net

herla

nds

Tab

le A

.1: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

pul

veriz

ed fu

el p

ower

pla

nts [

13] (

cont

’d).

Con

tinen

t

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

N.A

. : N

ot a

vaila

ble.

187

Co-

fired

Fue

l

- Cer

eal r

esid

ues

Sew

age

slud

ge

Var

ious

Var

ious

Var

ious

Var

ious

Var

ious

Var

ious

Woo

d

Woo

d

Woo

d

Woo

d

Prim

ary

Fuel

Pu

lver

ized

C

oal

Pulv

eriz

ed

Coa

l Pu

lver

ized

C

oal

Pulv

eriz

ed

Coa

l Pu

lver

ized

C

oal

Pulv

eriz

ed

Coa

l Pu

lver

ized

C

oal

Pulv

eriz

ed

Coa

l Pu

lver

ized

C

oal

Pulv

eriz

ed

Coa

l Pu

lver

ized

C

oal

Pulv

eriz

ed

Coa

l Pu

lver

ized

C

oal

Hea

t/w

t %

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

Out

put

MW

e

1980

2034

2400

4000

1960

2035

1995

2010

1000

1455

1200

2000

2100

Bur

ner

Type

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

Ow

ner

Edf

Pow

erge

n

Scot

tish

Pow

er

Dra

x Po

wer

Brit

ish

Ener

gy

AEP

AEP

Pow

erge

n

Inte

rnat

iona

l Po

wer

Inno

gy

Scot

tish

Pow

e r

EdF

Inno

gy

Plan

t Nam

e

Wes

t Bur

ton

Kin

gsno

rth

Lon

gann

et

Dra

x

Eggb

orou

gh

Ferr

ybri

dge

Fidd

lers

Fer

ry

Rat

cliff

e

Rug

eley

Abe

rthaw

Coc

kenz

ie

Cot

tam

Did

cot

Loca

tion

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

Cou

ntry

UK

UK

UK

UK

UK

UK

UK

UK

UK

UK

UK

UK

UK

Tab

le A

.1: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

pul

veriz

ed fu

el p

ower

pla

nts [

13] (

cont

’d).

Con

tinen

t

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

N.A

. : N

ot a

vaila

ble.

188

Co-

fired

Fue

l

Woo

d

Woo

d

RD

F

Urb

an w

ood

was

te, p

etro

leum

co

ke

Saw

dust

, tre

e tri

m

Woo

d ch

ips

Woo

d re

sidu

e,

will

ow

Shre

dded

pal

let

woo

d w

aste

Switc

hgra

ss

Was

te w

ood

Prim

ary

Fuel

Pulv

eriz

ed C

oal

Pulv

eriz

ed C

oal

Pulv

eriz

ed C

oal

Pulv

eriz

ed C

oal

Pulv

eriz

ed C

oal

Pulv

eriz

ed C

oal

Pulv

eriz

ed C

oal

Pulv

eriz

ed C

oal

Pulv

eriz

ed C

oal

Pulv

eriz

ed C

oal

Hea

t/w

t %

N.A

.

N.A

.

N.A

.

-/10

-/13

-/30

20/-

-/12

-/12

N.A

.

Out

put

MW

e

970

1085

75

160

100

108

90

120

60

100

Bur

ner

Type

N.A

.

N.A

.

N.A

.

Cyc

lone

Fron

t Fi

red

T-fir

ed

T-fir

ed

Fron

t-w

all-

fired

cy

clon

e

T-fir

ed

N.A

.

Ow

ner

Pow

erge

n

Inno

gy

N.A

.

NIP

SCO

Sout

hern

C

ompa

ny/

Geo

rgia

Pow

er

Com

pany

New

Yor

k St

ate

Elec

tric

& G

as

Nia

gara

M

ohaw

k Po

wer

C

orp.

Nor

ther

n St

ates

Po

wer

Com

pany

Sout

hern

C

ompa

ny/

Ala

bam

a Po

wer

C

ompa

ny

N.A

.

Plan

t Nam

e

Iron

brid

ge

Tilb

ury

Am

es

Mun

icip

al

Bai

ley

Gen

erat

ing

Stat

ion

# 7

Ham

mon

d G

ener

atin

g St

atio

n #

1

Gre

enid

ge

Gen

erat

ing

Stat

ion

# 6

Dun

kirk

St

eam

St

atio

n

BL

Stat

ion

# 1

Gad

sden

St

eam

pla

nt

#2

Geo

rgia

Po

wer

Loca

tion

N.A

N.A

.

Am

es, I

owa

Che

ster

ton,

In

Coo

sa,

Geo

rgia

Dre

sden

, N

ew Y

ork

Dre

sden

, N

ew Y

ork

Engl

and

Gad

sden

, A

laba

ma

Ham

mon

d

Cou

ntry

UK

UK

USA

USA

USA

USA

USA

USA

USA

USA

Tab

le A

.1: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

pul

veriz

ed fu

el p

ower

pla

nts [

13] (

cont

’d).

Con

tinen

t

Euro

pe

Euro

pe

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

N.A

. : N

ot a

vaila

ble.

189

Co-

fired

Fue

l

Var

ious

gro

und

woo

d

Var

ious

gro

und

woo

d

Chi

pped

railr

oad

ties/

PR

B

Urb

an w

ood

was

te/S

hosh

one

coal

/PR

B b

lend

RD

F

Switc

hgra

ss

Switc

hgra

ss

Saw

dust

Seed

cor

n, so

y be

ans

Prim

ary

Fuel

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Hea

t/w

t %

-/3

-/3

-/5

-/20

2/-

-/15

2.5/

-

-/20

1/-

Out

put

MW

e

190

138

840

469

350

2×50

650

272

450

Bur

ner

Type

T-fir

ed

Wal

l-fire

d

Supe

r-C

ritic

al

Cyc

lone

Cyc

lone

Wal

l-fire

d

Wal

l-fire

d

T-fir

ed

Cyc

lone

Cyc

lone

Ow

ner

Rel

iant

Ene

rgy

Rel

iant

Ene

rgy

Kan

sas c

ity

(MO

) Pow

er &

Li

ght

Nor

ther

n In

dian

a Pu

blic

Ser

vice

C

ompa

ny

Lake

land

El

ectri

c

Mad

ison

G

as &

Ele

ctric

C

ompa

ny

IES

Util

ities

Inc.

TVA

Otte

r Tai

l Pow

er

Co.

Plan

t Nam

e

Shaw

ville

G

ener

atin

g St

atio

n #

3

Shaw

ville

G

ener

atin

g St

atio

n #

2

La C

ygne

G

ener

atin

g St

atio

n #

1

Mic

higa

n C

ity

Gen

erat

ing

Stat

ion

# 12

Lake

land

El

ectri

c #

3

Blo

unt S

treet

Ottu

mw

a G

ener

atin

g St

atio

n #

1

Alle

n (T

.H)

Foss

il Pl

ant

Big

Sto

ne

Plan

t # 1

Loca

tion

John

stow

n,

Penn

sylv

ania

John

stow

n,

Penn

sylv

ania

Kan

sas C

ity,

Kan

sas

Lake

M

ichi

gan,

In

dian

a

Lake

land

, Fl

orid

a

Mad

ison

, W

isco

nsin

Mar

shal

ltow

n,

Iow

a

Mem

phis

, Te

nnes

see

Milb

ank,

So

uth

Dak

ota

Cou

ntry

USA

USA

USA

USA

USA

USA

USA

USA

USA

Tab

le A

.1: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

pul

veriz

ed fu

el p

ower

pla

nts [

13] (

cont

’d).

Con

tinen

t

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

N.A

. : N

ot a

vaila

ble.

190

Co-

fired

Fue

l

Sand

er d

ust

Woo

d ch

ips

Har

dwoo

d sa

wdu

st

Rai

lroad

ties

Shre

dded

ra

ilroa

d tie

s

Saw

dust

Saw

dust

from

pa

llets

Prim

ary

Fuel

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l Pu

lver

ized

C

oal

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Hea

t/w

t %

1/-

-/20

-/5

25/-

N.A

.

-/12

36/-

Out

put

MW

e 25

0,

319,

48

0,

490

165

180

75

170

32

46

Bur

ner

Type

Opp

osed

fir

ed

Wal

l fire

d

T-fir

ed

T-fir

ed

T-fir

ed

Wal

l fire

d

T-fir

ed

Ow

ner

Geo

rgia

Po

wer

C

ompa

ny

Sant

ee C

oope

r

TVA

Illin

ois P

ower

C

ompa

ny

Duk

e Po

wer

C

ompa

ny

Rel

iant

ene

rgy

Sout

hern

C

ompa

ny/

Sava

nnah

El

ectri

c &

Po

wer

C

ompa

ny

Plan

t Nam

e

Har

lee

Bra

nch

Gen

erat

ing

Stat

ion

Jeff

erie

s G

ener

atin

g St

atio

n #

3 &

# 4

Kin

gsto

n Fo

ssil

Plan

t # 5

Ver

mili

on P

ower

St

atio

n #

1 Le

e (W

.S) S

team

St

atio

n #

3

Sew

ard

Gen

erat

ing

Stat

ion

# 12

Kra

ft/

Riv

ersi

de

Plan

ts #

2

Loca

tion

Mill

edge

ville

, A

tlant

a,

Gea

orgi

a M

onck

s C

orne

r,

Sout

h C

arol

ina

Oak

ridge

, Te

nnes

see

Oak

woo

d,

Illin

ois

Pelz

er, S

outh

C

arol

ina

Pitts

burg

h,

Penn

sylv

ania

Port

Wen

twor

th,

Geo

rgia

Cou

ntry

USA

USA

USA

USA

USA

USA

USA

Tab

le A

.1: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

pul

veriz

ed fu

el p

ower

pla

nts [

13] (

cont

’d).

Con

tinen

t

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a N

orth

A

mer

ica

Nor

th

Am

eric

a

Nor

th

Am

eric

a

N.A

. : N

ot a

vaila

ble.

191

Co-

fired

Fue

l

Was

te p

aper

sl

udge

Was

te w

ood

Kiln

drie

d w

ood/

pe

trole

um c

oke/

PR

B b

lend

Pape

r pel

lets

Rai

lroad

ties

Saw

dust

Prim

ary

Fuel

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Pulv

eriz

ed

Coa

l

Hea

t/w

t %

1/-

N.A

.

5/-

5/-

7/-

5/-

Out

put

MW

e

250

54

560

165

175

182

Bur

ner

Type

T-fir

ed

Cyc

lone

Cyc

lone

Cyc

lone

Fron

t wal

l fir

ed

Ow

ner

Tri-S

tate

G

ener

atin

g &

Tr

ansm

issi

on

Ass

ocia

tion,

Inc.

Nor

ther

n St

ates

Po

wer

Tam

pa E

lect

ric

Com

pany

Ass

ocia

ted

Elec

tric

Coo

pera

tive,

In

c.

TVA

Plan

t Nam

e

Esca

lant

e G

ener

atin

g St

atio

n #

1

SEPC

O

Kin

g (A

llen

S.)

Gen

erat

ing

stat

ion

# 1

Gan

non

(F.J.

) G

ener

atin

g St

atio

n #

3

Thom

as H

ill

Ener

gy C

ente

r

# 2

Col

bert

Foss

il Pl

ant #

1

Loca

tion

Prew

itt, N

ew

Mex

ico

Sava

nnah

Still

wat

er,

Min

neso

ta

Tam

pa,

Flor

ida

Thom

as H

ill

Res

ervo

ir,

Col

umbi

a,

MO

Tusc

umbi

a,

AL

Cou

ntry

USA

USA

USA

USA

USA

USA

Tab

le A

.1: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

pul

veriz

ed fu

el p

ower

pla

nts [

13] (

cont

’d).

Con

tinen

t

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

Nor

th

Am

eric

a

N.A

. : N

ot a

vaila

ble.

192

Co-

fired

Fue

l

Peat

, bar

k, o

il

Peat

, woo

d ch

ips,

bark

, oil

Woo

d an

d pa

per

was

te

Peat

, woo

d w

aste

Peat

, woo

d w

aste

, H

FO

Bar

k, sl

udge

, fib

er w

aste

Pe

at, w

ood

was

te,

HFO

Woo

d w

aste

, pea

t, oi

l

Peat

20

%,

pelle

ts 8

0%

Woo

d w

aste

, pea

t, oi

l

Woo

d, re

fuse

der

ived

fu

el (R

DF)

Prim

ary

Fuel

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l (f

orm

erly

)

Coa

l

Coa

l

Hea

t/w

t %

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

100/

-

N.A

.

80/-

Out

put

MW

t

155

218

36

17.5

, 24

20

60

20

100

25

120

N.A

.

Out

put

MW

e

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

35

N.A

.

N.A

.

18

Ow

ner

Oce

an S

ky C

o

PT In

had

Kia

t Pu

lp &

Pap

er

UPM

Kym

men

e C

ompa

ny

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

Skel

lefte

a K

raft

N.A

.

Taco

ma

Publ

ic

Util

ities

Plan

t Nam

e

N.A

.

N.A

.

Lohj

a Pa

per M

ill

Out

okum

pu O

y

N.A

.

Rau

ma

Pape

r Mill

Sein

ajok

i Ene

rgy

Nyk

oppi

ng E

nerg

y

Hed

esby

n

Sode

rene

rgi A

B

City

Of T

acom

a St

eam

Pla

nt N

o. 2

Loca

tion

N.A

.

N.A

.

Lohj

a

Out

okum

pu

Piek

sam

aki

Dis

trict

Hea

ting

Rau

ma

Sein

ajok

i

Nyk

oppi

ng

Skel

lefte

a

Söde

rtälje

Taco

ma,

W

ashi

ngto

n

Cou

ntry

Indo

nesi

a

Indo

nesi

a

Finl

and

Finl

and

Finl

and

Finl

and

Finl

and

Swed

en

Swed

en

Swed

en

USA

Tab

le A

.2: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

bub

blin

g flu

idiz

ed b

ed p

ower

pla

nts [

13].

Con

tinen

t

Asi

a

Asi

a

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Nor

th

Am

eric

a

N.A

. : N

ot a

vaila

ble.

193

Co-

fired

Fue

l

Slud

ge

RD

F

Lign

ite, g

as, o

il,

woo

d

Lign

ite, o

il, w

ood

Lign

ite, w

ood,

se

wag

e sl

udge

Stra

w

Peat

, woo

d, sl

udge

Peat

, RD

f, w

ood

Peat

, woo

d w

aste

Peat

, bar

k, sa

wdu

st

Lign

ite, w

ood

was

te,

oil,

gas

Prim

ary

Fuel

Coa

l

Lign

ite

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Hea

t/w

t %

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

Out

put

MW

t

N.A

.

N.A

.

38

55

94

60

240

98

18

22

84

Out

put

MW

e

N.A

.

20

N.A

.

N.A

.

N.A

.

17

85

N.A

.

N.A

.

8 N.A

.

Ow

ner

Ba

Yu

Pape

r

N.A

.

N.A

.

N.A

.

N.A

.

Mid

kraf

t Pow

er

Co

Kai

nuun

Voi

ma

Oy

IVO

Kuh

mon

Lam

po

Oy

Lies

ka

Etel

a-Sa

von

Ener

gia

Plan

t Nam

e

N.A

.

Prov

inci

al E

lect

ricity

A

utho

rity

of T

haila

nd

Solv

ay O

ster

reic

h

Patri

a Pa

pier

&

Zells

toff

Lenz

ing

AG

.

Gre

naa

Co-

Gen

erat

ion

Plan

t

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

Loca

tion

Peik

ang

Chi

ang

Mai

Eben

see

Fran

tsch

ach

Lenz

ing

Gre

na

Kaj

aani

Kok

kola

Kuh

mo

Lies

ka

Mik

keli

Cou

ntry

Taiw

an

Thai

land

Aus

tria

Aus

tria

Aus

tria

Den

mar

k

Finl

and

Finl

and

Finl

and

Finl

and

Finl

and

Tab

le A

.3: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

circ

ulat

ing

fluid

ized

bed

pow

er p

lant

s [13

].

Con

tinen

t

Asi

a

Asi

a

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

N.A

. : N

ot a

vaila

ble.

194

Co-

fired

Fue

l

Peat

, bar

k, w

ood

Peat

, woo

d w

aste

Peat

, slu

dge,

bar

k

Woo

d, R

DF

Coa

l was

te, w

ood

was

te

Peat

, woo

d

Woo

d, b

ark

Var

ious

was

tes

Woo

d

Woo

d, p

eat

Woo

d, p

eat,

bark

, w

ood

was

te, o

il

Woo

d, o

il

Prim

ary

Fuel

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Hea

t/w

t %

90/-

100/

-

N.A

.

N.A

.

N.A

.

N.A

.

90/-

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

Out

put

MW

t

550

160

26

50

15

55

80

125

80

25

43

Out

put

MW

e

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

9.6

40

N.A

.

N.A

.

N.A

.

N.A

.

Ow

ner

Alh

olm

ens K

raft

N.A

.

Rau

ma

Mill

Sand

e Pa

per

Mill

A/S

N.A

.

N.A

.

Stor

a En

so L

td

Bris

ta K

raft

AB

N

orrk

opin

g K

raft

Nuk

opin

g En

ergi

verk

O

ster

sund

s Fj

arrv

arm

e

Cal

edon

ian

Pape

r plc

Plan

t Nam

e

Piet

arsa

ari

N.A

.

N.A

.

N.A

.

Hun

osa

pow

er

stat

ion

Ave

sta

Ener

give

rk

Stor

a En

so F

ors

Mill

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

Loca

tion

Rau

hala

hti m

unic

ipal

C

HP

plan

t

Rau

ma

Sand

e

La P

ered

a

Alv

esta

Fors

Mär

sta

Nor

rkop

ing

Nuk

opoi

ng

Öst

ersu

nd

Scot

land

Cou

ntry

Finl

and

Finl

and

Finl

and

Nor

way

Spai

n

Swed

en

Swed

en

Swed

en

Swed

en

Swed

en

Swed

en

UK

Tab

le A

.3: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

circ

ulat

ing

fluid

ized

bed

pow

er p

lant

s [13

] (co

nt’d

).

Con

tinen

t

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

Euro

pe

N.A

. : N

ot a

vaila

ble.

195

Co-

fired

Fue

l

RD

F fr

om w

aste

pa

per &

pla

stic

s

Slud

ge

Ant

hrac

ite, w

ood

Tire

s

Woo

d, o

il

Ant

hrac

ite, w

ood,

oil

Prim

ary

Fuel

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Coa

l

Hea

t/w

t %

40/-

N.A

.

N.A

.

N.A

.

N.A

.

N.A

.

Out

put

MW

th

125

168

149

260

132

Out

put

MW

e

35

65

N.A

.

N.A

.

76

N.A

.

Ow

ner

Slou

gh E

stat

es

Sout

heas

t Pap

er

Bla

ck R

iver

Pa

rtner

s U

DG

Nia

gara

G

oody

ear

Rum

ford

Cog

en

Co.

P.H

.Gla

tfelte

r C

o

Plan

t Nam

e

Slou

gh H

eat a

nd

Pow

er L

td.

N.A

.

N.A

.

N.A

.

Rum

ford

Cog

en

Co.

Sprin

g G

rove

Pa

per M

ill

Loca

tion

Slou

gh

Dub

lin, G

A.

Fort

Dru

m

Nia

gara

Fal

ls

Rum

ford

, Mai

ne

Sprin

g G

rove

, Pe

nnsy

lvan

ia

Cou

ntry

UK

USA

USA

USA

USA

USA

Tab

le A

.3: W

orld

wid

e sa

mpl

es o

f co-

firin

g ex

perie

nced

circ

ulat

ing

fluid

ized

bed

pow

er p

lant

s [13

] (co

nt’d

).

Con

tinen

t

Euro

pe

Nor

th

Am

eric

a N

orth

A

mer

ica

Nor

th

Am

eric

a N

orth

A

mer

ica

Nor

th

Am

eric

a

N.A

. : N

ot a

vaila

ble.

196

APPENDIX B

TGA GRAPHS OF FUELS

Time, min

0 10 20 30 40 50 60 70 80 90 100

Wei

ght,

%

0

10

20

30

40

50

60

70

80

90

100

Tem

pera

ture

, °C

0

100

200

300

400

500

600

700

800

900

1000

WeightTemperature

Figure B.1: TGA graph of lignite.

197

Time, min

0 10 20 30 40 50 60 70 80 90 100

Wei

ght,

%

0

10

20

30

40

50

60

70

80

90

100

Tem

pera

ture

, °C

0

100

200

300

400

500

600

700

800

900

1000

WeightTemperature

Figure B.2: TGA graph of olive residue.

Time, min

0 10 20 30 40 50 60 70 80 90 100

Wei

ght,

%

0

10

20

30

40

50

60

70

80

90

100

Tem

pera

ture

, °C

0

100

200

300

400

500

600

700

800

900

1000

WeightTemperature

Figure B.3: TGA graph of hazelnut shell.

198

Time, min

0 10 20 30 40 50 60 70 80 90 100

Wei

ght,

%

0

10

20

30

40

50

60

70

80

90

100

Tem

pera

ture

, °C

0

100

200

300

400

500

600

700

800

900

1000

WeightTemperature

Figure B.4: TGA graph of cotton residue.

199

APPENDIX C

POINT VALUES OF MEASUREMENTS

Table C.1: Sampling probe readings of O2 concentrations of Runs 1-3, dry mole %. No Height Run 1 Run 2 Run 3

1 5.00 4.76 4.85 4.36

2 4.19 - - -

3 3.44 - 4.69 4.32

4 2.91 - 4.64 4.28

5 1.83 - 5.24 4.86

6 1.23 7.50 6.11 6.01

Table C.2: Sampling probe readings of O2 concentrations of Runs 4-6, dry mole %. No Height Run 4 Run 5 Run 6

1 5.00 5.02 5.12 5.62

2 4.19 - - -

3 3.44 5.10 5.27 4.69

4 2.91 5.20 4.85 4.90

5 1.83 5.40 5.63 5.52

6 1.23 5.36 6.68 5.87

200

Run

6

15.5

0

-

16.3

8

15.9

7

15.3

4

14.6

0

Run

5

15.6

4

-

15.4

7

15.7

6

14.8

9

13.5

1

Run

4

15.7

1

-

15.6

2

15.5

4

15.4

0

15.2

7

Run

3

16.5

3

-

16.4

7

16.3

6

15.7

7

14.1

5

Run

2

16.0

7

-

16.3

0

16.3

5

15.3

7

14.2

2

Run

1

15.1

5

- - - -

11.9

3

Hei

ght,

m

5.00

4.19

3.44

2.91

1.83

1.23

Tab

le C

.3: S

ampl

ing

prob

e re

adin

gs o

f CO

2 con

cent

ratio

ns o

f Run

s 1-6

, dry

mol

e %

.

No 1 2 3 4 5 6

201

Run

6

500 - 517

630

1516

3214

Run

5

838 - 896

1900

3696

5478

Run

4

568 - 567

591

891

948

Run

3

570 - 607

831

2083

4814

Run

2

500 - 497

541

1136

3228

Run

1

525 - - - -

2913

Hei

ght,

m

5.00

4.19

3.44

2.91

1.83

1.23

Tab

le C

.4: S

ampl

ing

prob

e re

adin

gs o

f CO

con

cent

ratio

ns o

f Run

s 1-6

, dry

mol

e pp

m.

No 1 2 3 4 5 6

202

Run

6

201 - 273

321

178

72

Run

5

224 - 193

256

95

12

Run

4

315 - 272

239

231

252

Run

3

331 - 380

457

271

147

Run

2

630 - 499

553

427

79

Run

1

4507

- - - -

3211

Hei

ght,

m

5.00

4.19

3.44

2.91

1.83

1.23

Tab

le C

.5: S

ampl

ing

prob

e re

adin

gs o

f SO

2 con

cent

ratio

ns o

f Run

s 1-6

, dry

mol

e pp

m.

No 1 2 3 4 5 6

203

Run

6

224 - 249

245

266

277

Run

5

220 - 252

254

293

325

Run

4

248 - 253

250

257

261

Run

3

236 - 251

260

288

320

Run

2

246 - 267

269

290

297

Run

1

227 - - - - 271

Hei

ght,

m

5.00

4.19

3.44

2.91

1.83

1.23

Tab

le C

.6: S

ampl

ing

prob

e re

adin

gs o

f NO

con

cent

ratio

ns o

f Run

s 1-6

, dry

mol

e pp

m.

No 1 2 3 4 5 6

204

Run

6

10

- 8 9 14

13

Run

5

4 - 5 7 10

7

Run

4

11

- 11

13

16

15

Run

3

16

- 16

17

21

11

Run

2

23

- 22

24

24

21

Run

1

26

- - - - 27

Hei

ght,

m

5.00

4.19

3.44

2.91

1.83

1.23

Tab

le C

.7: S

ampl

ing

prob

e re

adin

gs o

f N2O

con

cent

ratio

ns o

f Run

s 1-6

, dry

mol

e pp

m.

No 1 2 3 4 5 6

205

Run

10

855

857

857

855

861

864

858

851

853

845

822

808

700

322

Run

9

859

860

860

859

863

868

862

856

861

853

830

816

705

320

Run

8

852

854

854

853

858

861

855

845

846

836

810

795

706

312

Run

7

851

852

853

851

856

859

853

840

841

832

806

790

702

313

Run

6

855

857

857

856

859

861

855

839

839

829

804

788

700

313

Run

5

849

851

853

851

857

864

859

857

861

853

830

816

722

316

Run

4

844

846

847

845

850

855

850

842

843

833

808

794

696

310

Run

3

859

860

861

859

863

867

861

848

848

838

812

796

693

316

Run

2

847

848

848

847

850

851

845

826

824

814

788

772

679

312

Run

1

892

893

894

893

900

903

896

879

874

862

834

816

719

339

Hei

ght,

m

0.25

0.44

0.73

0.73

0.97

1.33

1.54

2.26

2.57

2.85

3.30

3.61

4.25

5.00

Tab

le C

.8: T

herm

ocou

ple

read

ings

of R

uns 1

-10,

°C

No 1 2 3 4 5 6 7 8 9 10

11

12

13

14

206

Run

10

1.00

1.00

1.00

1.00

1.00

1.01

1.00

0.99

1.00

0.99

0.96

0.94

0.82

0.38

Run

9

1.00

1.00

1.00

1.00

1.00

1.01

1.00

1.00

1.00

0.99

0.96

0.95

0.82

0.37

Run

8

1.00

1.00

1.00

1.00

1.00

1.01

1.00

0.99

0.99

0.98

0.95

0.93

0.83

0.37

Run

7

1.00

1.00

1.00

1.00

1.00

1.01

1.00

0.99

0.99

0.98

0.95

0.93

0.82

0.37

Run

6

1.00

1.00

1.00

1.00

1.00

1.00

1.00

0.98

0.98

0.97

0.94

0.92

0.82

0.37

Run

5

1.00

1.00

1.00

1.00

1.01

1.01

1.01

1.01

1.01

1.00

0.97

0.96

0.85

0.37

Run

4

1.00

1.00

1.00

1.00

1.00

1.01

1.00

0.99

1.00

0.98

0.95

0.94

0.82

0.37

Run

3

1.00

1.00

1.00

1.00

1.00

1.01

1.00

0.99

0.99

0.97

0.94

0.93

0.81

0.37

Run

2

1.00

1.00

1.00

1.00

1.00

1.00

1.00

0.97

0.97

0.96

0.93

0.91

0.80

0.37

Run

1

1.00

1.00

1.00

1.00

1.01

1.01

1.00

0.98

0.98

0.96

0.93

0.91

0.80

0.38

Hei

ght,

m

0.25

0.44

0.73

0.73

0.97

1.33

1.54

2.26

2.57

2.85

3.30

3.61

4.25

5.00

Tab

le C

.9: N

orm

aliz

ed te

mpe

ratu

res o

f Run

s 1-1

0, °C

/°C

.

No 1 2 3 4 5 6 7 8 9 10

11

12

13

14

207

APPENDIX D

CHEMICAL ANALYSES OF ASH STREAMS

Table D.1: Chemical analyses of bottom ashes of Runs 1-4.

Run 1 Run 2 Run 3 Run 4

Loss on ignition, % 1.31 1.62 1.52 1.67 Components as oxides, wt %

Silica, SiO2 47.88 32.74 24.68 19.81 Aluminum, Al2O3 23.69 16.09 14.53 15.39 Ferric, Fe2O3 8.85 5.68 5.56 6.39 Calcium, CaO 8.97 28.30 34.67 36.14 Magnesium, MgO 1.17 1.60 1.99 2.05 Sulfur, SO3 4.83 11.99 14.88 15.47 Sodium, Na2O 1.83 1.23 1.25 1.97 Potasssium, K2O 1.24 1.17 1.19 1.23 Titanium, TiO2 1.54 1.20 1.23 1.56

208


Run 5 Run 6 Run 7

Loss on ignition, % 1.80 1.28 1.64 Components as oxides, wt %

Silica, SiO2 28.75 14.65 26.14 Aluminum, Al2O3 13.28 7.59 9.17 Ferric, Fe2O3 5.49 3.83 4.45 Calcium, CaO 31.25 46.14 37.88 Magnesium, MgO 1.80 2.76 1.94 Sulfur, SO3 15.61 21.71 17.09 Sodium, Na2O 1.19 1.70 1.20 Potasssium, K2O 1.27 0.76 1.08 Titanium, TiO2 1.37 0.87 1.06


Run 8 Run 9 Run 10

Loss on ignition, % 1.91 2.09 1.87 Components as oxides, wt %

Silica, SiO2 27.52 21.23 24.62 Aluminum, Al2O3 12.42 8.01 7.94 Ferric, Fe2O3 4.88 3.69 4.30 Calcium, CaO 37.93 45.60 39.95 Magnesium, MgO 1.34 1.35 1.93 Sulfur, SO3 12.58 17.53 18.08 Sodium, Na2O 1.01 0.94 1.15 Potasssium, K2O 1.10 0.78 0.89 Titanium, TiO2 1.22 0.86 1.13

209

Run

10

3.97

17.5

2

2.45

6.89

53.1

0

1.89

15.9

3

0.36

1.10

0.76

Run

9

4.30

22.9

3

6.28

7.54

43.4

7

2.01

14.8

1

0.61

1.37

0.98

Run

8

3.61

31.5

1

7.96

6.98

31.4

8

2.06

15.9

4

1.46

1.31

1.30

Run

7

3.80

21.9

1

9.40

8.01

38.7

1

2.12

15.6

4

1.82

1.17

1.22

Run

6

3.50

24.8

3

6.77

7.64

40.6

0

2.92

13.6

6

1.72

1.00

0.86

Run

5

3.50

19.5

8

7.51

6.67

45.1

0

4.63

12.6

5

1.60

1.43

0.83

Run

4

2.75

25.9

2

8.76

5.67

38.9

3

5.05

12.4

6

1.35

0.99

0.88

Run

3

3.17

29.7

1

7.75

6.18

33.0

1

4.51

15.5

6

1.18

0.81

1.29

Run

2

4.31

23.2

7

8.04

7.40

39.9

6

4.61

13.4

7

1.36

0.97

0.91

Run

1

5.05

51.6

1

20.9

2

10.4

8

7.41

0.69

4.05

2.34

0.77

1.73

Tab

le D

.4: C

hem

ical

ana

lyes

of c

yclo

ne a

shes

of R

uns 1

-10.

Loss

on

igni

tion,

%

Com

pone

nts a

s oxi

des,

wt %

Silic

a, S

iO2

Alu

min

um, A

l 2O3

Ferr

ic, F

e 2O

3

Cal

cium

, CaO

Mag

nesi

um, M

gO

Sulfu

r, SO

3

Sodi

um, N

a 2O

Pota

sssi

um, K

2O

Tita

nium

, TiO

2

210

Run

10

2.41

32.1

6

5.98

17.5

2

22.9

1

1.54

15.8

1

1.97

1.07

1.05

Run

9

2.26

31.1

5

6.72

14.6

1

28.3

5

1.40

13.5

2

1.86

0.97

1.41

Run

8

4.13

27.6

0

8.42

14.6

8

27.4

6

1.95

15.2

7

1.55

1.02

2.04

Run

7

5.45

29.5

5

8.78

14.8

0

25.2

4

2.54

14.7

5

1.67

1.12

1.56

Run

6

4.79

22.3

4

10.2

9

15.9

5

27.5

8

3.24

15.6

9

1.98

1.26

1.66

Run

5

9.34

22.2

1

8.49

14.2

9

27.8

0

3.11

18.7

9

2.37

1.19

1.76

Run

4

10.4

8

26.5

3

9.22

13.1

8

28.2

5

2.76

15.9

4

1.57

0.86

1.68

Run

3

5.37

25.7

4

6.60

12.5

9

32.4

6

2.01

16.6

7

1.72

1.00

1.20

Run

2

5.74

32.1

9

9.69

12.6

1

22.9

5

1.40

17.8

0

1.14

0.51

1.72

Run

1

8.83

45.8

2

14.6

8

17.3

6

9.49

0.80

7.93

1.57

0.47

1.88

Tab

le D

.5: C

hem

ical

ana

lyes

of b

agho

use

filte

r as

hes o

f Run

s 1-1

0.

Loss

on

igni

tion,

%

Com

pone

nts a

s oxi

des,

wt %

Silic

a, S

iO2

Alu

min

um, A

l 2O3

Ferr

ic, F

e 2O

3

Cal

cium

, CaO

Mag

nesi

um, M

gO

Sulfu

r, SO

3

Sodi

um, N

a 2O

Pota

sssi

um, K

2O

Tita

nium

, TiO

2

211

APPENDIX E

TABULATED SIZE DISTRIBUTIONS

Table E.1: Particle size distribution of bottom ash of Run 1.

ASTM MESH #

SIEVE OPENING, mm

DIFFERENTIAL WEIGHT, %

CUMULATIVE WEIGHT, %

5/16 8.000 0.000 0.000 1/4 6.300 0.003 0.003

4 4.750 0.123 0.127 6 3.350 4.140 4.266 10 2.000 6.572 10.838 18 1.000 47.066 57.904 35 0.500 37.191 95.095 45 0.355 4.578 99.673 80 0.180 0.102 99.776 140 0.106 0.048 99.824 PAN 0.000 0.176 100.000

212


ASTM MESH #

SIEVE OPENING, mm



5/16 8.000 0.000 0.000 1/4 6.300 0.580 0.580

4 4.750 1.458 2.038 6 3.350 9.269 11.307 10 2.000 11.480 22.788 18 1.000 36.110 58.898 35 0.500 31.478 90.376 45 0.355 8.795 99.172 80 0.180 0.550 99.722 140 0.106 0.089 99.811 PAN 0.000 0.189 100.000


ASTM MESH #

SIEVE OPENING, mm



5/8 16.000 0.000 0.000 1/2 12.700 0.287 0.287

5/16 8.000 0.000 0.287 1/4 6.300 0.510 0.797

4 4.750 0.929 1.726 6 3.350 8.052 9.778 10 2.000 11.454 21.232 18 1.000 37.571 58.803 35 0.500 33.463 92.266 45 0.355 7.567 99.833 80 0.180 0.154 99.988 140 0.106 0.012 100.000 PAN 0.000 0.000 100.000

213


ASTM MESH #

SIEVE OPENING, mm



5/16 8.000 0.000 0.000 1/4 6.300 0.693 0.693

4 4.750 1.472 2.166 6 3.350 10.001 12.167 10 2.000 11.805 23.971 18 1.000 30.786 54.758 35 0.500 32.934 87.692 45 0.355 11.265 98.957 80 0.180 0.866 99.823 140 0.106 0.177 100.000 PAN 0.000 0.000 100.000


ASTM MESH #

SIEVE OPENING, mm



5/16 8.000 0.000 0.000 1/4 6.300 0.628 0.628

4 4.750 1.755 2.383 6 3.350 10.238 12.621 10 2.000 11.295 23.915 18 1.000 28.815 52.730 35 0.500 33.934 86.664 45 0.355 12.307 98.972 80 0.180 0.848 99.820 140 0.106 0.180 100.000 PAN 0.000 0.000 100.000

214


ASTM MESH #

SIEVE OPENING, mm



5/16 8.000 0.000 0.000 1/4 6.300 0.533 0.533

4 4.750 0.670 1.202 6 3.350 2.930 4.132 10 2.000 3.134 7.266 18 1.000 19.696 26.962 35 0.500 51.820 78.782 45 0.355 20.240 99.022 80 0.180 0.533 99.554 140 0.106 0.171 99.725 PAN 0.000 0.275 100.000


ASTM MESH #

SIEVE OPENING, mm



5/16 8.000 0.000 0.000 1/4 6.300 0.883 0.883

4 4.750 2.297 3.180 6 3.350 10.811 13.991 10 2.000 10.225 24.215 18 1.000 25.919 50.134 35 0.500 35.159 85.293 45 0.355 13.165 98.458 80 0.180 0.900 99.358 140 0.106 0.313 99.671 PAN 0.000 0.329 100.000

215


ASTM MESH #

SIEVE OPENING, mm



5/16 8.000 0.000 0.000 1/4 6.300 1.210 1.210

4 4.750 2.283 3.494 6 3.350 10.481 13.974 10 2.000 10.390 24.364 18 1.000 26.559 50.923 35 0.500 36.107 87.031 45 0.355 12.243 99.273 80 0.180 0.339 99.613 140 0.106 0.047 99.660 PAN 0.000 0.340 100.000


ASTM MESH #

SIEVE OPENING, mm



5/16 8.000 0.000 0.000 1/4 6.300 1.013 1.013

4 4.750 2.340 3.353 6 3.350 10.147 13.500 10 2.000 9.530 23.030 18 1.000 25.887 48.917 35 0.500 37.141 86.057 45 0.355 13.340 99.398 80 0.180 0.371 99.768 140 0.106 0.036 99.804 PAN 0.000 0.196 100.000

216


ASTM MESH #

SIEVE OPENING, mm



5/16 8.000 0.000 0.000 1/4 6.300 0.866 0.866

4 4.750 2.671 3.537 6 3.350 10.960 14.497 10 2.000 9.775 24.272 18 1.000 26.182 50.454 35 0.500 42.894 93.348 45 0.355 5.964 99.311 80 0.180 0.244 99.555 140 0.106 0.127 99.682 PAN 0.000 0.318 100.000

217

Table E.11: Particle size distribution of cyclone ash of Run 1.

Size, µm CUMULATIVE WEIGHT, % Size, µm CUMULATIVE

WEIGHT, % 1261.915 0.000 28.251 86.920 1124.683 0.110 25.179 87.730 1002.374 0.410 22.440 88.490 893.367 1.130 20.000 89.210 796.214 2.390 17.825 89.900 709.627 4.510 15.887 90.560 632.456 7.510 14.159 91.200 563.667 11.310 12.619 91.830 502.377 15.770 11.247 92.440 447.744 20.720 10.024 93.030 399.052 25.980 8.934 93.610 355.656 31.370 7.962 94.180 316.979 36.710 7.096 94.730 282.508 41.860 6.325 95.270 251.785 46.710 5.637 95.780 224.404 51.200 5.024 96.270 200.000 55.290 4.477 96.730 178.250 58.960 3.991 97.150 158.866 62.240 3.557 97.520 141.589 65.160 3.170 97.850 126.191 67.760 2.825 98.100 112.468 70.090 2.518 98.310 100.237 72.210 2.244 98.530 89.337 74.130 2.000 98.750 79.621 75.910 1.783 98.970 70.963 77.550 1.589 99.170 63.246 79.070 1.416 99.370 56.368 80.480 1.262 99.550 50.238 81.780 1.125 99.710 44.774 82.980 1.002 99.830 39.905 84.090 0.893 99.930 35.566 85.110 0.796 99.990 31.698 86.050 0.710 100.000

218



WEIGHT, % 1002.374 0.000 22.440 78.370 893.367 0.030 20.000 80.030 796.214 0.150 17.825 81.640 709.627 0.380 15.887 83.190 632.456 0.860 14.159 84.700 563.667 1.780 12.619 86.150 502.377 3.200 11.247 87.540 447.744 5.050 10.024 88.840 399.052 7.280 8.934 90.060 355.656 9.870 7.962 91.190 316.979 12.770 7.096 92.210 282.508 15.930 6.325 93.120 251.785 19.310 5.637 93.920 224.404 22.860 5.024 94.630 200.000 26.520 4.477 95.250 178.250 30.230 3.991 95.790 158.866 33.940 3.557 96.270 141.589 37.610 3.170 96.700 126.191 41.190 2.825 97.100 112.468 44.650 2.518 97.470 100.237 47.960 2.244 97.820 89.337 51.130 2.000 98.160 79.621 54.130 1.783 98.480 70.963 56.980 1.589 98.780 63.246 59.670 1.416 99.070 56.368 62.200 1.262 99.330 50.238 64.600 1.125 99.560 44.774 66.860 1.002 99.750 39.905 69.010 0.893 99.890 35.566 71.050 0.796 99.980 31.698 72.990 0.710 100.000 28.251 74.860 25.179 76.640

219



WEIGHT, % 893.367 0.000 20.000 86.900 796.214 0.010 17.825 87.850 709.627 0.450 15.887 88.780 632.456 1.700 14.159 89.690 563.667 3.760 12.619 90.580 502.377 6.620 11.247 91.440 447.744 10.230 10.024 92.280 399.052 14.520 8.934 93.070 355.656 19.350 7.962 93.810 316.979 24.560 7.096 94.510 282.508 29.970 6.325 95.150 251.785 35.400 5.637 95.740 224.404 40.680 5.024 96.270 200.000 45.680 4.477 96.750 178.250 50.300 3.991 97.190 158.866 54.480 3.557 97.580 141.589 58.210 3.170 97.930 126.191 61.500 2.825 98.250 112.468 64.400 2.518 98.540 100.237 66.950 2.244 98.800 89.337 69.220 2.000 99.030 79.621 71.240 1.783 99.240 70.963 73.080 1.589 99.420 63.246 74.770 1.416 99.590 56.368 76.320 1.262 99.720 50.238 77.770 1.125 99.840 44.774 79.130 1.002 99.930 39.905 80.420 0.893 100.000 35.566 81.630 31.698 82.770 28.251 83.870 25.179 84.920 22.440 85.920

220



WEIGHT, % 893.367 0.000 20.000 87.470 796.214 0.020 17.825 88.380 709.627 0.630 15.887 89.270 632.456 2.050 14.159 90.140 563.667 4.230 12.619 90.980 502.377 7.210 11.247 91.790 447.744 10.950 10.024 92.580 399.052 15.380 8.934 93.320 355.656 20.380 7.962 94.020 316.979 25.760 7.096 94.680 282.508 31.330 6.325 95.280 251.785 36.890 5.637 95.840 224.404 42.270 5.024 96.350 200.000 47.310 4.477 96.820 178.250 51.920 3.991 97.240 158.866 56.040 3.557 97.630 141.589 59.670 3.170 97.970 126.191 62.830 2.825 98.290 112.468 65.590 2.518 98.570 100.237 68.010 2.244 98.830 89.337 70.160 2.000 99.060 79.621 72.090 1.783 99.260 70.963 73.860 1.589 99.440 63.246 75.500 1.416 99.600 56.368 77.020 1.262 99.740 50.238 78.460 1.125 99.850 44.774 79.810 1.002 99.930 39.905 81.080 0.893 100.000 35.566 82.290 31.698 83.430 28.251 84.510 25.179 85.540 22.440 86.520

221



WEIGHT, % 893.367 0.000 20.000 87.380 796.214 0.020 17.825 88.290 709.627 0.760 15.887 89.180 632.456 2.290 14.159 90.030 563.667 4.540 12.619 90.870 502.377 7.590 11.247 91.680 447.744 11.390 10.024 92.460 399.052 15.890 8.934 93.200 355.656 20.940 7.962 93.910 316.979 26.370 7.096 94.570 282.508 31.970 6.325 95.190 251.785 37.530 5.637 95.760 224.404 42.870 5.024 96.290 200.000 47.850 4.477 96.770 178.250 52.370 3.991 97.210 158.866 56.390 3.557 97.600 141.589 59.920 3.170 97.960 126.191 62.980 2.825 98.290 112.468 65.650 2.518 98.580 100.237 67.980 2.244 98.840 89.337 70.060 2.000 99.070 79.621 71.940 1.783 99.280 70.963 73.670 1.589 99.460 63.246 75.280 1.416 99.610 56.368 76.800 1.262 99.740 50.238 78.240 1.125 99.850 44.774 79.600 1.002 99.940 39.905 80.900 0.893 100.000 35.566 82.130 31.698 83.290 28.251 84.390 25.179 85.440 22.440 86.430

222



WEIGHT, % 2000.000 0.000 35.566 68.770 1782.502 0.050 31.698 70.760 1588.656 0.130 28.251 72.650 1415.892 0.230 25.179 74.450 1261.915 0.350 22.440 76.180 1124.683 0.480 20.000 77.860 1002.374 0.590 17.825 79.490 893.367 0.700 15.887 81.080 796.214 0.810 14.159 82.640 709.627 0.930 12.619 84.160 632.456 1.080 11.247 85.640 563.667 1.450 10.024 87.050 502.377 2.150 8.934 88.390 447.744 3.180 7.962 89.640 399.052 4.590 7.096 90.790 355.656 6.430 6.325 91.830 316.979 8.690 5.637 92.750 282.508 11.370 5.024 93.580 251.785 14.410 4.477 94.310 224.404 17.770 3.991 94.950 200.000 21.360 3.557 95.530 178.250 25.120 3.170 96.050 158.866 28.970 2.825 96.530 141.589 32.850 2.518 96.980 126.191 36.690 2.244 97.400 112.468 40.440 2.000 97.800 100.237 44.060 1.783 98.190 89.337 47.520 1.589 98.550 79.621 50.810 1.416 98.890 70.963 53.910 1.262 99.200 63.246 56.820 1.125 99.470 56.368 59.530 1.002 99.700 50.238 62.070 0.893 99.870 44.774 64.450 0.796 99.980 39.905 66.680 0.710 100.000

223



WEIGHT, % 2000.000 0.000 35.566 64.820 1782.502 0.180 31.698 67.020 1588.656 0.470 28.251 69.170 1415.892 0.850 25.179 71.270 1261.915 1.250 22.440 73.330 1124.683 1.640 20.000 75.340 1002.374 1.990 17.825 77.320 893.367 2.300 15.887 79.240 796.214 2.590 14.159 81.120 709.627 2.940 12.619 82.940 632.456 3.400 11.247 84.680 563.667 4.060 10.024 86.330 502.377 4.990 8.934 87.870 447.744 6.240 7.962 89.280 399.052 7.830 7.096 90.560 355.656 9.770 6.325 91.710 316.979 12.030 5.637 92.720 282.508 14.580 5.024 93.610 251.785 17.340 4.477 94.380 224.404 20.270 3.991 95.050 200.000 23.310 3.557 95.630 178.250 26.420 3.170 96.150 158.866 29.550 2.825 96.630 141.589 32.670 2.518 97.060 126.191 35.760 2.244 97.470 112.468 38.800 2.000 97.860 100.237 41.780 1.783 98.230 89.337 44.680 1.589 98.580 79.621 47.490 1.416 98.910 70.963 50.220 1.262 99.210 63.246 52.860 1.125 99.480 56.368 55.400 1.002 99.700 50.238 57.860 0.893 99.870 44.774 60.250 0.796 99.980 39.905 62.560 0.710 100.000

224



WEIGHT, % 1124.683 0.000 20.000 89.670 1002.374 0.060 17.825 90.500 893.367 0.490 15.887 91.290 796.214 1.490 14.159 92.040 709.627 3.030 12.619 92.760 632.456 5.270 11.247 93.450 563.667 8.260 10.024 94.110 502.377 12.010 8.934 94.740 447.744 16.440 7.962 95.340 399.052 21.420 7.096 95.930 355.656 26.790 6.325 96.490 316.979 32.310 5.637 97.040 282.508 37.790 5.024 97.570 251.785 43.040 4.477 98.070 224.404 47.910 3.991 98.540 200.000 52.330 3.557 98.960 178.250 56.240 3.170 99.330 158.866 59.670 2.825 99.640 141.589 62.670 2.518 99.850 126.191 65.280 2.244 99.990 112.468 67.610 2.000 100.000 100.237 69.730 89.337 71.690 79.621 73.550 70.963 75.340 63.246 77.080 56.368 78.740 50.238 80.340 44.774 81.840 39.905 83.250 35.566 84.550 31.698 85.740 28.251 86.840 25.179 87.850 22.440 88.790

225



WEIGHT, % 2000.000 0.000 35.566 69.990 1782.502 0.020 31.698 71.860 1588.656 0.040 28.251 73.650 1415.892 0.070 25.179 75.370 1261.915 0.110 22.440 77.050 1124.683 0.150 20.000 78.690 1002.374 0.200 17.825 80.310 893.367 0.240 15.887 81.920 796.214 0.350 14.159 83.510 709.627 0.680 12.619 85.070 632.456 1.310 11.247 86.590 563.667 2.280 10.024 88.040 502.377 3.630 8.934 89.400 447.744 5.400 7.962 90.660 399.052 7.600 7.096 91.810 355.656 10.210 6.325 92.830 316.979 13.180 5.637 93.730 282.508 16.460 5.024 94.520 251.785 19.980 4.477 95.200 224.404 23.660 3.991 95.790 200.000 27.410 3.557 96.300 178.250 31.150 3.170 96.750 158.866 34.830 2.825 97.160 141.589 38.380 2.518 97.530 126.191 41.790 2.244 97.880 112.468 45.030 2.000 98.210 100.237 48.100 1.783 98.520 89.337 51.020 1.589 98.810 79.621 53.800 1.416 99.090 70.963 56.460 1.262 99.350 63.246 59.000 1.125 99.570 56.368 61.420 1.002 99.750 50.238 63.730 0.893 99.890 44.774 65.930 0.796 99.980 39.905 68.010 0.710 100.000

226



WEIGHT, % 893.367 0.000 20.000 89.640 796.214 0.030 17.825 90.440 709.627 0.950 15.887 91.220 632.456 2.770 14.159 91.970 563.667 5.360 12.619 92.700 502.377 8.840 11.247 93.400 447.744 13.160 10.024 94.080 399.052 18.210 8.934 94.740 355.656 23.820 7.962 95.370 316.979 29.760 7.096 95.970 282.508 35.790 6.325 96.560 251.785 41.660 5.637 97.110 224.404 47.180 5.024 97.640 200.000 52.220 4.477 98.140 178.250 56.690 3.991 98.600 158.866 60.570 3.557 99.010 141.589 63.890 3.170 99.370 126.191 66.730 2.825 99.660 112.468 69.160 2.518 99.860 100.237 71.280 2.244 99.990 89.337 73.180 2.000 100.000 79.621 74.930 70.963 76.570 63.246 78.140 56.368 79.640 50.238 81.070 44.774 82.430 39.905 83.690 35.566 84.870 31.698 85.970 28.251 86.980 25.179 87.920 22.440 88.800

227

Table E.21: Particle size distribution of baghouse filter ash of Run 1.


WEIGHT, % 14.159 0.000 1.002 84.720 12.619 0.050 0.893 86.990 11.247 0.490 0.796 89.020 10.024 1.580 0.710 90.860 8.934 3.270 0.632 92.510 7.962 5.590 0.564 94.000 7.096 8.560 0.502 95.320 6.325 12.170 0.448 96.480 5.637 16.380 0.399 97.460 5.024 21.130 0.356 98.250 4.477 26.320 0.317 98.860 3.991 31.820 0.283 99.320 3.557 37.500 0.252 99.650 3.170 43.220 0.224 99.880 2.825 48.860 0.200 100.000 2.518 54.290 2.244 59.440 2.000 64.220 1.783 68.610 1.589 72.580 1.416 76.150 1.262 79.350 1.125 82.190

228



WEIGHT, % 14.159 0.000 1.002 86.470 12.619 0.070 0.893 88.420 11.247 0.600 0.796 90.170 10.024 1.810 0.710 91.750 8.934 3.670 0.632 93.180 7.962 6.210 0.564 94.490 7.096 9.440 0.502 95.670 6.325 13.360 0.448 96.710 5.637 17.930 0.399 97.610 5.024 23.060 0.356 98.350 4.477 28.620 0.317 98.910 3.991 34.480 0.283 99.350 3.557 40.480 0.252 99.660 3.170 46.460 0.224 99.880 2.825 52.270 0.200 100.000 2.518 57.800 2.244 62.940 2.000 67.640 1.783 71.860 1.589 75.600 1.416 78.890 1.262 81.760 1.125 84.270

229



WEIGHT, % 14.159 0.000 1.002 86.950 12.619 0.050 0.893 88.910 11.247 0.320 0.796 90.650 10.024 1.440 0.710 92.210 8.934 3.220 0.632 93.600 7.962 5.650 0.564 94.860 7.096 8.800 0.502 95.980 6.325 12.650 0.448 96.960 5.637 17.170 0.399 97.800 5.024 22.270 0.356 98.490 4.477 27.840 0.317 99.010 3.991 33.730 0.283 99.410 3.557 39.800 0.252 99.700 3.170 45.870 0.224 99.890 2.825 51.800 0.200 100.000 2.518 57.460 2.244 62.740 2.000 67.580 1.783 71.930 1.589 75.800 1.416 79.180 1.262 82.140 1.125 84.710

230



WEIGHT, % 15.887 0.000 1.125 85.550 14.159 0.380 1.002 87.560 12.619 1.370 0.893 89.350 11.247 2.880 0.796 90.960 10.024 4.970 0.710 92.410 8.934 7.670 0.632 93.740 7.962 10.950 0.564 94.940 7.096 14.810 0.502 96.030 6.325 19.190 0.448 96.990 5.637 24.030 0.399 97.820 5.024 29.240 0.356 98.490 4.477 34.710 0.317 99.010 3.991 40.320 0.283 99.410 3.557 45.940 0.252 99.690 3.170 51.450 0.224 99.890 2.825 56.750 0.200 100.000 2.518 61.750 2.244 66.370 2.000 70.580 1.783 74.360 1.589 77.720 1.416 80.680 1.262 83.280

231



WEIGHT, % 14.159 0.000 1.002 86.930 12.619 0.210 0.893 88.840 11.247 1.120 0.796 90.550 10.024 2.590 0.710 92.090 8.934 4.690 0.632 93.480 7.962 7.430 0.564 94.740 7.096 10.830 0.502 95.870 6.325 14.870 0.448 96.870 5.637 19.480 0.399 97.730 5.024 24.600 0.356 98.430 4.477 30.120 0.317 98.970 3.991 35.890 0.283 99.380 3.557 41.770 0.252 99.680 3.170 47.630 0.224 99.880 2.825 53.320 0.200 100.000 2.518 58.730 2.244 63.770 2.000 68.370 1.783 72.520 1.589 76.210 1.416 79.450 1.262 82.290 1.125 84.760

232



WEIGHT, % 12.619 0.000 0.893 88.030 11.247 0.070 0.796 89.900 10.024 0.900 0.710 91.570 8.934 2.350 0.632 93.070 7.962 4.430 0.564 94.420 7.096 7.230 0.502 95.630 6.325 10.730 0.448 96.690 5.637 14.930 0.399 97.600 5.024 19.760 0.356 98.340 4.477 25.120 0.317 98.910 3.991 30.890 0.283 99.350 3.557 36.900 0.252 99.660 3.170 42.990 0.224 99.880 2.825 49.010 0.200 100.000 2.518 54.820 2.244 60.280 2.000 65.330 1.783 69.900 1.589 73.990 1.416 77.600 1.262 80.760 1.125 83.510 1.002 85.910

233



WEIGHT, % 14.159 0.000 1.002 86.160 12.619 0.080 0.893 88.210 11.247 0.080 0.796 90.030 10.024 0.970 0.710 91.660 8.934 2.470 0.632 93.120 7.962 4.600 0.564 94.450 7.096 7.460 0.502 95.640 6.325 11.020 0.448 96.700 5.637 15.280 0.399 97.600 5.024 20.170 0.356 98.340 4.477 25.600 0.317 98.910 3.991 31.420 0.283 99.340 3.557 37.480 0.252 99.660 3.170 43.610 0.224 99.880 2.825 49.650 0.200 100.000 2.518 55.450 2.244 60.910 2.000 65.920 1.783 70.460 1.589 74.490 1.416 78.030 1.262 81.130 1.125 83.810

234



WEIGHT, % 14.159 0.000 1.002 86.860 12.619 0.050 0.893 88.800 11.247 0.320 0.796 90.520 10.024 1.410 0.710 92.060 8.934 3.140 0.632 93.450 7.962 5.520 0.564 94.710 7.096 8.620 0.502 95.840 6.325 12.420 0.448 96.840 5.637 16.900 0.399 97.710 5.024 21.970 0.356 98.410 4.477 27.530 0.317 98.960 3.991 33.440 0.283 99.370 3.557 39.530 0.252 99.670 3.170 45.640 0.224 99.880 2.825 51.620 0.200 100.000 2.518 57.320 2.244 62.640 2.000 67.510 1.783 71.890 1.589 75.760 1.416 79.150 1.262 82.090 1.125 84.640

235



WEIGHT, % 14.159 0.000 1.002 86.950 12.619 0.140 0.893 88.820 11.247 0.910 0.796 90.500 10.024 2.370 0.710 92.030 8.934 4.520 0.632 93.420 7.962 7.370 0.564 94.690 7.096 10.950 0.502 95.830 6.325 15.190 0.448 96.850 5.637 20.050 0.399 97.710 5.024 25.400 0.356 98.420 4.477 31.100 0.317 98.970 3.991 37.020 0.283 99.380 3.557 43.000 0.252 99.680 3.170 48.870 0.224 99.890 2.825 54.510 0.200 100.000 2.518 59.820 2.244 64.730 2.000 69.170 1.783 73.150 1.589 76.670 1.416 79.760 1.262 82.470 1.125 84.850

236



WEIGHT, % 14.159 0.000 1.002 86.350 12.619 0.060 0.893 88.380 11.247 0.580 0.796 90.190 10.024 1.760 0.710 91.820 8.934 3.590 0.632 93.280 7.962 6.080 0.564 94.600 7.096 9.270 0.502 95.780 6.325 13.120 0.448 96.810 5.637 17.610 0.399 97.690 5.024 22.650 0.356 98.410 4.477 28.130 0.317 98.960 3.991 33.900 0.283 99.380 3.557 39.820 0.252 99.680 3.170 45.740 0.224 99.890 2.825 51.520 0.200 100.000 2.518 57.040 2.244 62.200 2.000 66.940 1.783 71.230 1.589 75.060 1.416 78.450 1.262 81.430 1.125 84.050

237

APPENDIX F

SIZE DISTRIBUTION GRAPHS

Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 1 Lignite

Figure F.1: Particle size distribution of lignite fed in Run 1.

238

Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 2 Lignite


Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 3 Lignite


239

Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 4 Lignite


Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 5 Lignite


240

Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 6 Lignite


Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 7 Lignite


241

Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 8 Lignite


Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 9 Lignite


242

Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 10 Lignite


Sieve opening, mm

0 1 2 3 4 5 6 7

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Olive residue

Figure F.11: Particle size distribution of olive residue fed in Runs 3, 4 and 5.

243

Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Hazelnut shell

Figure F.12: Particle size distribution of hazelnut shell fed in Runs 6, 7 and 8.

Sieve opening, mm

0 2 4 6 8 10 12 14 16 18 20

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Cotton residue

Figure F.13: Particle size distribution of cotton residue fed in Runs 9 and 10.

244

Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 2 Limestone

Figure F.14: Particle size distribution of limestone fed in Runs 2, 3, 4 and 5.

Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 6 Limestone

Figure F.15: Particle size distribution of limestone fed in Runs 6, 7 and 8.

245

Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 10 Limestone

Figure F.16: Particle size distribution of limestone fed in Runs 9 and 10.

246

Sieve opening, mm

0 1 2 3 4 5 6 7 8

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 1 Bed drain ash

Figure F.17: Particle size distribution of bottom ash of Run 1.

Sieve opening, mm

0 1 2 3 4 5 6 7 8

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 2 Bed drain ash


247

Sieve opening, mm

0 2 4 6 8 10 12 14 16

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 3 Bed drain ash


Sieve opening, mm

0 1 2 3 4 5 6 7 8

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 4 Bed drain ash


248

Sieve opening, mm

0 1 2 3 4 5 6 7 8

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 5 Bed drain ash


Sieve opening, mm

0 1 2 3 4 5 6 7 8

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 6 Bed drain ash


249

Sieve opening, mm

0 1 2 3 4 5 6 7 8

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 7 Bed drain ash


Sieve opening, mm

0 1 2 3 4 5 6 7 8

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 8 Bed drain ash


250

Sieve opening, mm

0 1 2 3 4 5 6 7 8

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 9 Bed drain ash


Sieve opening, mm

0 1 2 3 4 5 6 7 8

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 10 Bed drain ash


251

Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 1 Cyclone ash

Figure F.27: Particle size distribution of cyclone ash of Run 1.

Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 2 Cyclone ash


252

Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 3 Cyclone ash


Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 4 Cyclone ash


253

Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 5 Cyclone ash


Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 6 Cyclone ash


254

Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 7 Cyclone ash


Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 8 Carryover ash


255

Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 9 Cyclone ash


Sieve opening, mm

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 10 Cyclone ash


256

Sieve opening, mm

0.000 0.005 0.010 0.015

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100

Run 1 Baghouse filter ash

Figure F.37: Particle size distribution of baghouse filter ash of Run 1.

Sieve opening, mm

0.000 0.005 0.010 0.015

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100



257

Sieve opening mm

0.000 0.005 0.010 0.015

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100



Sieve opening, mm

0.000 0.005 0.010 0.015

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100



258

Sieve opening, mm

0.000 0.005 0.010 0.015

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100



Sieve opening, mm

0.000 0.005 0.010 0.015

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100



259

Sieve opening, mm

0.000 0.005 0.010 0.015

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100



Sieve opening, mm

0.000 0.005 0.010 0.015

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100



260

Sieve opening, mm

0.000 0.005 0.010 0.015

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100



Sieve opening, mm

0.000 0.005 0.010 0.015

Res

idue

, %

0

10

20

30

40

50

60

70

80

90

100



261

APPENDIX G

CALIBRATION CURVES

Frequency, Hz

0 5 10 15 20 25 30 35

Lign

ite fl

ow ra

te, k

g/h

0102030405060708090

100110120130140

y = 4.304 xR2 = 1.000

Figure G.1: Calibration curve for lignite flow rate.

262

Frequency, Hz

0 10 20 30 40 50 60 70 80 90

Oliv

e re

sidu

e flo

w ra

te, k

g/h

0

10

20

30

40

50

60

70

80

90

100

110

y = 1.250 xR2 = 1.000

Figure G.2: Calibration curve for olive residue flow rate.

Frequency, Hz

0 10 20 30 40 50 60 70 80 90 100 110

Haz

elnu

t she

ll flo

w ra

te, k

g/h

0

10

20

30

40

50

y = 0.466 xR2 = 1.000

Figure G.3: Calibration curve for hazelnut shell flow rate.

263

Frequency, Hz

0 10 20 30 40 50 60

Cot

ton

resi

due

flow

rate

, kg/

h

0

10

20

30

40

50

y = 0.787 xR2 = 0.999

Figure G.4: Calibration curve for cotton residue flow rate.

Frequency, Hz

0 10 20 30 40

Lim

esto

ne fl

ow ra

te, k

g/h

0

10

20

30

40

50

60

70

80

y = 2.377 xR2 = 0.999

Figure G.5: Calibration curve for limestone flow rate.

264

APPENDIX H

DEPOSIT COMPOSITIONS

Table H.1: Deposit compositions.

Element, wt %

Olive residue/ lignite co-firing

deposit

Hazelnut shell/ lignite co-firing

deposit

Cotton residue/ lignite co-firing

deposit

Na 3.12 2.00 1.72

Mg 3.70 2.13 2.51

Al 5.05 7.62 5.68

Si 16.79 27.93 23.95

P 1.28 0.00 4.54

S 20.21 19.00 15.10

Cl 0.27 0.00 0.00

K 11.59 1.62 4.12

Ca 27.76 25.01 25.66

Ti 0.45 0.97 0.94

Mn 0.78 0.00 0.00

Fe 9.00 13.72 15.78

265

Figure H.1: EDX analysis graph of olive residue and lignite co-firing deposit.

Figure H.2: EDX analysis graph of hazelnut shell and lignite co-firing deposit.

266

Figure H.3: EDX analysis graph of cotton residue and lignite co-firing deposit.

APPENDIX I

SCREW FEEDER DESIGN CALCULATIONS

Biomass feeding was problematic with the available feeding system which was

originally designed for coal feeding. As biomass fuels have much lower bulk density

compared to coal bulk density, their flow rate was much lower compared to that of

coals. To mitigate the feeding problems, a new bed screw feeder is designed by

taking account of the available data. To check the design approach existing ash

feeder is utilized. The results of the calculations and comparisons are as follows.

Ash Feeder Design Check

The flow rate of available spiral ash feeder is checked with the theoretical and design

values. The calculation procedure for volumetric and mass flow rates are given

below.

L = 0.046 m

R2 = 0.029 m R1 = 0.010 m

0.019 m

Figure I.1: Schematic description of ash feeder spiral.

267

( )

( )

2 22 1

3

1

2

2 2

Q = Π R -R L N 60

where

Q : volumetric flow rate (m /h)R : inside radius (m)R : spiral radius (m)L : pitch lenght (m)N : revoltion per minute (rpm)

Q = Π 0.029 -0.010 0.046 N 60 = 0.006425 N

m = ρ

⋅ ⋅ ⋅ ⋅

⋅ ⋅ ⋅ ⋅

i

i

i

i

Bulk BulkQ = 0.006425 Nρ⋅ ⋅ ⋅i

⋅

Bulk densities for hazelnut shell and olive residue are measured as 260 and 600

kg/m3, respectively. By utilizing the above expression, mass flow rates of both

biomasses are calculated and then compared with the measured values of hazelnut

shell and olive residue flow rate. Results are given in Tables I.1 and I.2 and Figures

I.2 and I.3.

Table I.1: Comparison of measured and calculated hazelnut shell flow rate.

Frequency, Hz Measured Flow Rate, kg/h

Calculated Flow Rate, kg/h Difference, %

0.0 0.0 0.0 0.0

30.0 13.2 14.0 6.3

40.0 18.0 18.7 3.9

50.0 22.0 23.4 6.3

60.0 27.0 28.1 3.9

75.0 33.0 35.1 6.3

85.0 36.0 39.8 10.4

90.0 38.0 42.1 10.8

268

0

5

10

15

20

25

30

35

40

45

0 20 40 60 80 100Frequency, Hz

Flow

rate

, kg/

h

MeasuredCalculated

Figure I.2: Comparison of measured and calculated hazelnut shell flow rate.

Table I.2: Comparison of measured and calculated olive oil residue flow rate.

Frequency, Hz Measured Flow Rate, kg/h

Calculated Flow Rate, kg/h Difference, %

0.0 0.0 0.0 0.0

10.0 12.3 10.8 12.2

20.0 24.5 21.6 11.9

30.0 37.0 32.4 12.5

40.0 49.2 43.2 12.2

50.0 61.0 54.0 11.5

60.0 74.0 64.8 12.5

75.0 96.0 81.0 15.7

269

0

20

40

60

80

100

120

0 10 20 30 40 50 60 70 80

Frequency, Hz

Flow

rate

, kg/

hMeasuredCalculated

Figure I.3: Comparison of measured and calculated olive residue flow rate.

As can be seen from the above tables and figures, theoretical and measured values

are in good agreement. The differences between the theoretical and measured values

are acceptable considering the motor frequency controller accuracy which is around

10 %. Following the verification of design approach, over bed and under bed screw

feeders design have been performed.

Bed Screw Feeder Design

In the figure given below the dimensions of new bed screw feeders are shown. The

main constraints in the design of new feeder are the available space for installation,

cooling jacket and availability of the materials. To maximize the diameter of the

feeder spiral for feeding of large biomass particles, wall thicknesses are reduced. To

provide easy flow of materials and also increased durability, stainless steel (SS-304)

was used in the manufacturing of new feeders. New feeder design is then checked

with 45 % hazelnut shell and 55 % lignite blend having 500 kg/m3 bulk density.

270

L = 0.05 m

R2 = 0.025 m R1 = 0.010 m

0.015 m

Figure I.4: Schematic description of new bed feeder spiral.

( )( )

2 22 1

2 2

3Bulk blend

Bulk blend

Q = Π R -R L N 60

Q = Π 0.025 -0.010 0.050 N 60 = 0.004948 N

N = 50 rpmρ = 500 kg/m

m Q = 0.004948 500 50 123.7 kg/h

m 97 kg/h (from stoichiometric combustion calculat

ρ

⋅ ⋅ ⋅ ⋅

⋅ ⋅ ⋅ ⋅ ⋅

= ⋅ ⋅ ⋅

=

=

i

i

i i

iion)

As 123.7 > 97 kg/h, which is obtained from stoichiometric combustion calculations,

the new feeder is suitable for co-firing processes. Figure I.5 shows drawing of the

new bed feeder.

271

Figu

re I.

5: S

chem

atic

dra

win

g of

the

new

bed

feed

er.

272

273

CURRICULUM VITAE

PERSONAL INFORMATION Surname, Name : Göğebakan, Zuhal Nationality : Turkish (TC) Date and Place of Birth : 21 March 1979, Denizli Marital Status : Married Phone : + 90 312 2104387 Fax : + 90 312 2102600 E-mail : [email protected]

EDUCATION Degree Institution Year of Graduation PhD METU, Chemical Engineering 2007 MS METU, Chemical Engineering 2003 BS METU, Chemical Engineering 2001 High School Anatolian High School, Denizli 1997

WORK EXPERIENCE Year Place Enrollment 2001-2007 METU, Chemical Engineering Research Assistant 2000 July Deniz Textile, Denizli Intern Engineering Student

ACADEMIC EXPERIENCE Assistantship to some undergraduate courses:

Chemical Engineering Design

Mathematical Modeling in Chemical Engineering

mailto:[email protected]

274

Fundamentals of Heat and Mass Transfer

Novel Topics in Separation Processes

Chemical Engineering Laboratories I, II, III

Chemical Engineering Economics

Industrial Organization and Management

Introduction to Chemical Engineering

Assistantship to some graduate courses:

Combustion Technology

Advanced Heat Transfer

Fluidization

• Attended to “2nd Chemical Engineering Conference for Collaborative Research in Eastern Mediterranean” in Ankara, Turkey in 20-24 May 2001.

• Attended to “3rd Chemical Engineering Conference for Collaborative Research in Eastern Mediterranean” in Thessaloniki, Greece in 13-15 May 2003.

• Attended to “18th International Conference on Fluidized Bed Combustion” in Toronto, Ontario, Canada in 22-25 May 2005.

AREAS OF EXPERTISE

• Fluidized bed combustion

• Co-firing of biomass and coal

• Deposit formation in FBC FOREIGN LANGUAGES Advanced English, Basic German COMPUTER SKILLS

• Software packages: Microsoft Office (Word, Excel, PowerPoint, FrontPage,Visio), AutoCAD, Sigma Plot, MathCAD.

• Programming Languages: Visual Basic.

275

REFEREED PAPERS IN CONFERENCE PROCEEDINGS

[1] Coskun Z., Yucel H., Culfaz A., “Synthesis and Characterization of ZSM-35”, 3rd Chemical Engineering Conference for Collaborative Research in Eastern Mediterranean, Thessaloniki, Greece, in CD-ROM (2003).

[2] Selcuk, N., Gogebakan, Y., and Gogebakan, Z., “Partitioning of Trace Elements during Fluidized Bed Combustion of High Ash Content Lignite”, Proceedings of the 18th International Conference on Fluidized Bed Combustion (Ed. Jia L.), ASME, Toronto, Ontario, Canada, Paper No: 58 in CD-ROM (2005).

[3] Gogebakan Z., Gogebakan Y., Selcuk N., “Co-firing of Olive Residue with Lignite in Bubbling FBC”, accepted to be published in Proceedings of 5th Mediterranean Combustion Symposium, Monastir, Tunisia, September 9-13, (2007).

REFEREED PAPERS IN JOURNALS

[1] Selcuk, N., Gogebakan, Y., and Gogebakan, Z., “Partitioning Behavior of Trace Elements during Pilot Scale Fluidized Bed Combustion of High Ash Content Lignite”, Journal of Hazardous Materials B137 (2006) 1698-1703.

[2] Gogebakan Z., Yucel H. and Culfaz A., “Crystallization Field and Rate Study for the Synthesis of Ferrierite”, Journal of Industrial and Engineering Chemistry Research 46 (2007) 2006-2012.

RESEARCH PROJECTS

[1] Selçuk N., Göğebakan Y., Göğebakan Z., Uygur A. B., Moralı M., Co-firing of Biomass with Coal in Bubbling Fluidized Bed Combustors, TÜBİTAK MAG-104M200 (Continuing).

HOBBIES Traveling, Shopping, Movies, Formula 1, Painting.

open.metu.edu.tr · Approval of the thesis: CO-FIRING BIOMASS WITH COAL IN BUBBLING FLUIDIZED BED...

Documents

Transcript of open.metu.edu.tr · Approval of the thesis: CO-FIRING BIOMASS WITH COAL IN BUBBLING FLUIDIZED BED...