Co-axial turbulent diffusion flames with directed oxygen injection

Oxy-coal combustion technology has been suggested as the most promising strategy for retrofitting conventional coal power plants to generate electric power while capturing carbon dioxide. The current research addresses three issues in oxy-coal combustion, namely: 1- What is the effect of coal composition on the stability of co-axial turbulent diffusion oxy-flames? 2- What are the stability criteria for turbulent diffusion oxy-coal flames in an advanced triple concentric co-axial burner allowing directed streams of pure oxygen to be introduced into the combustion mix? 3- How does minimization of CO2 diluent affect radiant heat flux in the combustion chamber? It is hoped that data produced in this investigation can be used for validation of advanced simulations of the appropriate configurations considered. In order to address Issue #1 listed above, the consequences of differences in coal composition on flame stability for two types of coal in oxy-combustion were explored: Utah Skyline Bituminous and Illinois #6 Bituminous. Differences in flame stand-off distances at equivalent experimental input conditions were interpreted through differences in the structure of the two coals as well as differences in their pyrolysis behavior, as determined by fundamental solid state 13C NMR and Thermal Gravimetric Analysis (TGA), respectively. In addressing Issue #2, the consequences of segregating all the input oxygen into one stream composed of 100% oxygen were determined using the co-axial burners with different oxygen stream configurations. Flame stability, heat flux, and NOx formation measurements were taken to evaluate the differences. Flame stability was quantified through flame probability density functions (PDF) of the stand-off distance (determined using photo-imaging techniques). The PDFs obtained from these simplified prototype configurations led to physical insight into coal flame attachment mechanisms and the significant effects of fine coal particles and their radial transportation by large eddies on flame stability. Finally, in addressing Issue #3, impacts of reducing the amount of injected diluent CO2 (mimicking the minimization of the recycle ratio) on the radiation heat flux were explored. Radiant heat flux, gas temperature, and wall temperature measurements were taken, and a simple radiation model was developed to correlate the average gas temperature and radian heat flux. This study provided a better understanding of the radiation mechanism and the significant effects of soot radiation on the total heat transfer in the next generation of oxy-coal combustion.

CO-AXIAL TURBULENT DIFFUSION FLAMES WITH DIRECTED OXYGEN INJECTION by Dadmehr Rezaei A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Chemical Engineering The University of Utah May 2013 Copyright © Dadmehr Rezaei 2013 All Rights Reserved The Uni v e r s i t y of Utah Graduat e School STATEMENT OF DISSERTATION APPROVAL The dissertation of _______________________ Dadmehr Rezaei____________________ has been approved by the following supervisory committee members: Jost O.L. Wendt , Chair 10-24-2012 Date Approved Eric G. Eddings , Member 10-24-2012 Date Approved Terry A. Ring , Member 10-24-2012 Date Approved Philip J. Smith , Member 10-24-2012 Date Approved Lawrence E. Bool , Member 10-24-2012 Date Approved and by _______________________JoAnn Lighty_______________________ , Chair of the Department of ____________________ Chemical Engineering_________________ and by Donna M. White, Interim Dean of The Graduate School. ABSTRACT Oxy-coal combustion technology has been suggested as the most promising strategy for retrofitting conventional coal power plants to generate electric power while capturing carbon dioxide. The current research addresses three issues in oxy-coal combustion, namely: 1- What is the effect of coal composition on the stability of co-axial turbulent diffusion oxy-flames? 2- What are the stability criteria for turbulent diffusion oxy-coal flames in an advanced triple concentric co-axial burner allowing directed streams of pure oxygen to be introduced into the combustion mix? 3- How does minimization of CO2 diluent affect radiant heat flux in the combustion chamber? It is hoped that data produced in this investigation can be used for validation of advanced simulations of the appropriate configurations considered. In order to address Issue #1 listed above, the consequences of differences in coal composition on flame stability for two types of coal in oxy-combustion were explored: Utah Skyline Bituminous and Illinois #6 Bituminous. Differences in flame stand-off distances at equivalent experimental input conditions were interpreted through differences in the structure of the two coals as well as differences in their pyrolysis behavior, as determined by fundamental solid state 13C NMR and Thermal Gravimetric Analysis (TGA), respectively. In addressing Issue #2, the consequences of segregating all the input oxygen into one stream composed of 100% oxygen were determined using the co-axial burners with different oxygen stream configurations. Flame stability, heat flux, and NOx formation measurements were taken to evaluate the differences. Flame stability was quantified through flame probability density functions (PDF) of the stand-off distance (determined using photo-imaging techniques). The PDFs obtained from these simplified prototype configurations led to physical insight into coal flame attachment mechanisms and the significant effects of fine coal particles and their radial transportation by large eddies on flame stability. Finally, in addressing Issue #3, impacts of reducing the amount of injected diluent CO2 (mimicking the minimization of the recycle ratio) on the radiation heat flux were explored. Radiant heat flux, gas temperature, and wall temperature measurements were taken, and a simple radiation model was developed to correlate the average gas temperature and radian heat flux. This study provided a better understanding of the radiation mechanism and the significant effects of soot radiation on the total heat transfer in the next generation of oxy-coal combustion. iv CONTENTS ABSTRACT................................................................................................................................iii LIST OF TABLES......................................................................................................................ix LIST OF FIGURES.....................................................................................................................x NOMENCLATURE................................................................................................................ xiv ACKNOWLEDGMENTS...................................................................................................... xvi 1. INTRODUCTION.............................................................................................................2 1.1 Background ............................................................................................................... 2 1.2 Oxy-Coal Combustion Technology Description ................................................. 2 1.3 Ignition and Flame Stability in Oxy-Coal Combustion......................................3 1.4 Sulfur and Recycled Flue Gas (RFG)................................................................... 5 1.5 NOX Formation.........................................................................................................7 1.6 Heat Transfer............................................................................................................9 1.7 Co-Axial Jets...........................................................................................................11 1.8 Coal Chemistry.......................................................................................................13 1.9 Previous Work at the University of Utah............................................................15 2. RESEARCH OBJECTIVES, MOTIVATION, AND ORGANIZATION OF THIS DISSERTATION .................................................................................................. 20 3. EXPERIMENTAL ........................................................................................................... 24 3.1 Oxy-Fuel Combustor (OFC).................................................................................24 3.2 Burner Zone............................................................................................................25 3.3 Convection Zone.................................................................................................... 26 3.4 Gas Heater............................................................................................................... 27 3.5 Wide Angle and Narrow Angle Radiometers.................................................... 27 3.6 Wide Angle Radiometer....................................................................................... 28 3.7 Narrow Angle Radiometer....................................................................................30 3.8 Calibration of Radiometers...................................................................................31 3.8.1 Thermal Response of Radiometers..........................................................31 3.8.2 Angular Response of WA Radiometers................................................... 32 3.9 Burner Design.........................................................................................................33 3.9.1 Design of Double Concentric Burner..................................................... 33 3.9.2 Triple Co-centric Annulus Burner Configuration A: Burners with Oxygen Stream in Middle Annular Stream.............................................35 3.9.3 Triple Co-centric Annulus Burner Configuration B: Burners with Oxygen Stream in Centrally Located Pipe..............................................37 3.10 Coal Sample and Analysis....................................................................................37 3.11 Gas Analyzers.........................................................................................................38 3.12 O2 and CO2 Delivery..............................................................................................39 3.13 U-Shaped Tube Water Cooling Heat Exchanger...............................................39 4. METHODOLOGY OF MEASURMENT OF STAND-OFF AND LENGTH OF FLAME............................................................................................................................ 64 5. THE EFFECT OF COAL COMPOSITION ON IGNITION AND FLAME STABILITY IN CO-AXIAL TURBULENT DIFFUSION FLAMES................. 69 5.1 Abstract....................................................................................................................69 5.2 Introduction............................................................................................................. 70 5.3 Coal Selection and Sample Preparation .............................................................. 72 5.4 Oxy-Coal Combustor ............................................................................................. 73 5.5 Gas Analyzers ......................................................................................................... 73 5.6 Burner and Feeder.................................................................................................. 74 5.7 Methodology of Flame Stability Measurement.................................................74 5.8 Combustion Operating Conditions......................................................................75 5.9 TGA and NMR Experiments................................................................................76 5.10 Results......................................................................................................................76 5.10.1 Combustion of Utah Skyline and Illinois #6 Coals at 489°K Preheat Secondary Stream Temperature and Overall 40% Oxygen Concentration ..............................................................................................76 5.10.2 Combustion of Utah Skyline and Illinois #6 coals at 544°K Preheat Secondary Stream Temperature and Overall 40% Oxygen Concentration ..............................................................................................77 5.10.3 Combustion of Utah Skyline and Illinois #6 Coals at 489°K Secondary Stream Temperature with Increasing of Overall Oxygen Concentration in the Secondary Stream.................................................. 77 vi 5.11 TGA Analysis.........................................................................................................78 5.12 Discussion............................................................................................................... 78 5.13 Conclusion.............................................................................................................. 82 5.14 Acknowledgments.................................................................................................. 83 6. NEAR FIELD AERODYNAMICS EFFECTS OF PURE O2 INJECTION IN CO-AXIAL OXY-COAL TURBULENT DIFFUSION FLAMES.......................... 93 6.1 Abstract................................................................................................................... 93 6.2 Introduction............................................................................................................. 94 6.3 Experimental ........................................................................................................... 97 6.3.1 Coal Selection and Sample Preparation.................................................. 97 6.3.2 Oxy-Coal Combustion Furnace............................................................... 97 6.3.3 Methodology of Flame Stability Measurement.....................................98 6.3.4 Burners and Combustion Operating Conditions....................................98 6.3.5 Experimental Methodology for Configuration A: Burners with Oxygen Stream in Middle Annular Stream.............................................99 6.3.6 Experimental Methodology for Configuration B: Burners with Oxygen Stream in Centrally Located Pipe..............................................99 6.4 Results....................................................................................................................100 6.5 Discussion and Conclusions.............................................................................. 102 6.6 Acknowledgments................................................................................................105 7. ADDENDUM TO CHAPTER 6..................................................................................111 7.1 Flame Radiant Heat Flux....................................................................................111 7.2 NOx Formation.................................................................................................... 112 7.3 Additional Information on the Second Hypothesis......................................... 114 7.4 Sample Images of Flame from Configurations A and B Burners................. 115 8. MINIMIZATION OF CO2 IN OXY-COAL COMBUSTION WITH PURE O2 INJECTION IN CO-AXIAL TURBULENT DIFFUSION FLAMES.................. 121 8.1 Introduction...........................................................................................................121 8.2 Coal Selection and Sample Preparation............................................................124 8.3 Burners and Combustion Operating Conditions..............................................124 8.4 Radiation Calculation in Oxy-Fuel Combustion..............................................126 8.4.1 Radiation Calculation Method............................................................... 126 8.4.2 Estimation of Hot Gas Volume Emissivity.......................................... 130 vii 8.4.3 Estimation of Soot Particles Emissivity................................................131 8.5 Results....................................................................................................................135 8.6 Conclusions...........................................................................................................137 9. FUTURE WORK...........................................................................................................145 APPENDICES A. FLAME STAND-OFF DISTANCE PDFS OF CONFIGURATION A ................ 148 B. FLAME STAND-OFF DISTANCE PDFS OF CONFIGURATION B ................ 151 C. DESIGN THEORY OF CONFIGURATIONS A AND B ...................................... 153 D. TEMPORAL CHANGES OF WALL TEMPERATURE AT OF45......................158 E. RADIANT HEAT FLUX DATA WITH STANDARD DEVIATION................. 160 F. WALL TEMPERATURE DATA WITH STANDARD DEVIATION..............162 REFERENCES.......................................................................................................................164 viii LIST OF TABLES 1. Flame aerodynamic conditions: burners with inner annular oxygen stream (Configuration A ) ....................................................................................... 62 2. Flame aerodynamic conditions for burners with central oxygen stream (Configuration B ) .................................................................................................... 62 3. Coals proximate analysis........................................................................................ 62 4. Coals ultimate analysis........................................................................................... 63 5. Coals elemental analysis........................................................................................ 63 6. Constant key parameters........................................................................................ 91 7. Combustion aerodynamic operating conditions for both Utah Skyline and Illinois #6 coal..........................................................................................................91 8. Coal feeding rate based on SR .............................................................................. 92 9. Structural and lattice parameters of coals.............................................................92 10. Fuel ratio of coals.................................................................................................... 92 11. Constant parameter of combustion operating conditions.................................110 12. Flame aerodynamic conditions for burners with inner annular oxygen stream (Configuration A ) .....................................................................................110 13. Flame aerodynamic conditions for burners with central oxygen stream........110 14. Combustion and burners conditions for CO2 reduction test............................ 144 15. Combustion and burner conditions for CO2 reduction test (streams flow rate).......................................................................................................................... 144 16. Measurement positions of thermocouples and radiometers............................ 144 LIST OF FIGURES 1. Oxy Fuel Combustion Schematic.......................................................................... 17 2. NO reduction ratio for various combinations of N2 and N O ............................. 17 3. NO reduction ratio for isolated reduction.............................................................18 4. Effect of flue gas recycle ratio on flame temperature and radiative heat flux............................................................................................................................ 18 5. Comparison of radiation intensity in air- and oxy-firing...................................19 6. Comparison of radiant intensity measurements and gas radiation modeling at 384mm from the burner inlet in the Chalmers furnace. The furnace walls are located at radial distances of 800mm.....................................19 7. Near field zones of double concentric je ts ...........................................................20 8. Oxy-Fuel Combustor (OFC) at the University of Utah...................................... 41 9. Recycle Flue Gas system of OFC at University of Utah....................................42 10. OFC at the University of U tah .............................................................................. 43 11. Flow Diagram of OFC.............................................................................................44 12. O2 and CO2 lines equipped with solenoid valve and mass flow controller..... 45 13. Drawing of Windows..............................................................................................45 14. Heat exchanger of convective zone of OFC.........................................................46 15. Sketch of gas heater electric element....................................................................46 16. Gas heater installed in lin e .....................................................................................47 17. Schematic of OFC................................................................................................... 47 18. Schematic of location of radiometers in OFC......................................................48 19. Schematic of principle of wide angle radiometer................................................48 20. Pictures of wide angle radiometer.........................................................................49 21. Schematic of narrow angle radiometer................................................................. 50 22. Pictures of narrow angle radiometer.....................................................................49 23. Blackbody radiator.................................................................................................. 50 24. Wide angle radiometer calibration on 05-09-2011.............................................. 51 25. Narrow angle radiometer calibration on 05-09-2011 ......................................... 51 26. Angular response of Radiometer #1 (top port).................................................... 52 27. Angular response of Radiometer #2 (middle port)..............................................52 28. Angular response of Radiometer #3 (bottom po rt).............................................52 29. Field of view of wide angle radiometers in OFC................................................53 30. Sketch of double coaxial burner............................................................................ 54 31. Schematic of Configuration A burners................................................................. 55 32. Configuration A Burner.......................................................................................... 55 33. Schematic of Configuration B burners................................................................. 56 34. Configuration B burners used in this study..........................................................57 35. Configuration B burner.......................................................................................... 57 36. Particle size distribution of Utah Skyline and Illinois #6 coals........................58 37. Gas analyzers at the University of Utah combustion research lab ................... 58 38. Carbon dioxide tan k ................................................................................................59 39. Oxygen tank ............................................................................................................. 59 40. Manifolds equipped with flow meters and valves...............................................60 41. Position of thermocouples on top plate................................................................ 60 42. Schematic of cooling heat exchanger....................................................................61 43. U-shape water cooling heat exchanger................................................................. 61 xi 44. Procedure of optical methodology to measure flame stand-off distance [49] ............................................................................................................................ 67 45. Flame stand-off distance in PDF form................................................................. 67 46. Flame length distance in PDF form....................................................................... 68 47. Schematic of OFC (side view with no recycle)................................................... 84 48. Design and drawing of burner................................................................................84 49. Comparison of PDF of stand-off distance for Utah Skyline (Left column) and Illinois #6 (Right column) coals at 489°K preheat temperature................ 85 50. Comparison of PDF of stand-off distance for Utah Skyline (Left column) and Illinois #6 (Right column) coals at 544°K preheat temperature................ 86 51. Comparison of PDF of stand-off distance for Utah Skyline (Left column) and Illinois #6 (Right column) coals at 489°K preheat secondary temperature with increasing of overall oxygen concentration (Figure continues)................................................................................................................. 87 52. TGA analysis for Utah Skyline and Illinois #6 coal in Nitrogen environment.............................................................................................................. 89 53. TGA analysis of the Utah Skyline and Illinois #6 coal in Air environment .... 89 54. 13C CPMAS spectra of the two coals taken with a 1 s pulse delay and a 2 ms contact time. All carbon types are seen in these spectra...............................90 55. 13C CPMAS dipolar dephased spectra of the two coals taken with a 1 s pulse delay, a 2 ms contact time, and a dipolar dephasing time of 42 ^s. These spectra show nonprotonated carbons, methyl groups, and protonated carbons subject to a large degree of molecular motion...................90 56. Stand-off distance and length of the flame for zero oxygen in the directed O2 stream for Configuration A, Case 1, Table 12.............................................106 57. Flame stand-off distance PDFs. Left: Configuration A, Right: Configuration B ..................................................................................................... 107 58. Flame length PDFs. Left: Configuration A, Right: Configuration B .............108 59. Stokes number vs. particle diameter at nozzle exit.......................................... 109 60. Illinois #6 coal particle distribution.....................................................................109 61. Positions of wide angle radiometers....................................................................116 xii 62. Distance from burner tip to wide angle radiometers........................................ 116 63. Radiant heat flux from flames with directed pure oxygen injection (Configuration A ).................................................................................................. 117 64. Radiant heat flux from flames with directed pure oxygen injection (Configuration B ) .................................................................................................. 117 65. NOx formation vs. F02 in center oxygen stream...............................................118 66. NOx formation vs. FO2 in annulus oxygen stream........................................... 118 67. Radiant heat transfer mechanism to coal particles in configuration B........... 119 68. Flame images and change of stand-off distance as a function of FO2: Configuration B ..................................................................................................... 119 69. Comparison of effects of pure O2 stream location on flame stability............120 70. Distances of thermocouples and radiometers from burner.............................. 139 71. Axial wall temperature variation caused by reduction of CO2........................140 72. Radiant heat flux variation due to reduction of CO2 in axial locations of OFC.........................................................................................................................140 73. Calculated total emissivities.................................................................................140 74. Increase of gas temperature in top port (175 mm from burner)......................141 75. Increase of gas temperature in middle port (556 mm from burner)............... 141 76. Increase of gas temperature in bottom port (556 mm from burner)............... 141 77. Comparison of calculated and measured gas temperature at OF40............... 142 78. Comparison of calculated and measured gas temperature at OF45............... 142 79. Comparison of calculated and measured gas temperature at OF50............... 142 80. Comparison of calculated and measured gas temperature at OF60............... 143 xiii NOMENCLATURE f a , The fraction of carbon atoms that are sp2 hybridized (aromaticity). fac, The fraction of carbon atoms that are in carboxyl or carbonyl groups. fa0, The fraction of carbon atoms that are in a carbonyl group (aldehydes and ketones). fa00, The fraction of carbon atoms that are in a carboxyl group (acids, esters, amides). f a,, The fraction of carbon atoms that are sp2 hybridized excluding (corrected aromaticity). faH, The fraction of carbon atoms that are protonated aromatics. faw, The fraction of carbon atoms that are non-protonated aromatics. f a , The fraction of carbon atoms that are aromatic with an oxygen atom attached. f a , The fraction of carbon atoms that are aromatic with a carbon chain attached (also includes biaryl carbons). f f , The fraction of carbon atoms that are aromatic and a bridgehead carbon. f ai, The fraction of carbon atoms that are sp3 hybridized (aliphatic). f « , The fraction of carbon atoms that are aliphatic but not methyls. fah, The fraction of carbons that are aliphatic and methyls f a , The fraction of carbon atoms that are aliphatic and attached to an oxygen atom. Xb, The mole fraction of bridgehead carbon atoms. C, The average aromatic cluster size. ct+ 1, The average number of attachments on an aromatic cluster.P0, The fraction of attachments that don't end in a side chain (methyl group). B.L., The average number of attachments on an aromatic cluster that are bridges or loops (a loop is a bridge back to the same cluster). S.C., The average number of side chains on an aromatic cluster. MW, The average molecular weight of an aromatic cluster including side-chains and bridges. Ms, The average mass of a side chain or one-half of a bridge. xv ACKNOWLEDGMENTS First and foremost, I would like to thank my parents, Reza Mohammadrezaei and Zahra Khoshbin, for their lifelong support in everything, but most significantly, for their encouragement to pursue a higher education. Additionally, my siblings, Daryoush and Roya Mohammadrezaei, were an integral part of my success. I would also like to express my most sincere gratitude to Prof. Jost O. L. Wendt for his continual guidance, patience and support throughout my years in the program. It has been such a pleasure for me to work with Prof. Wendt. Prof. Eric Eddings, Prof. Terry Ring., Prof. Phil Smith, and Dr. Lawrence Bool, who were also much appreciated members of my defense committee. I am also thankful to Prof. Ronald Pugmire and Dr. Mark Solum for their encouragements and technical suggestions as well as for providing outstanding NMR data. The help of Dr. Yuegui Zhou of Shanghai Jiao Tong University in the early stages of the research was unforgettable. Kerry Kelly is also much appreciated for contributing to the success of the integration of the photo-imaging technique and analyzing the length and stability of flame. Working with all of these professionals has been a great pleasure and a learning experience The help of Dana Overacker, Ryan Okerlund, Dave Wagner, and Brian Nelson in the lab is appreciated. Without their indispensable help, the experimentation portion would have been significantly drawn out. Additionally, I would like to thank Charles German, Colby Ashcroft, Taylor Geisler, and Travis LeGrande for their help in the lab. This work is based upon work supported by the Department of Energy under Award Number FC26-08NT0005015 through the Clean Coal Center (UC3). The O2 and CO2 supply during the test were provided at no cost to the project by Praxair Inc. xvii 1. INTRODUCTION 1.1 Background The generation of electricity from fossil fuels has caused the emission of greenhouse gases. Previous research has shown global warming is the result of greenhouse gases. Among the greenhouse gases, CO2 is known as the most dominant contributor. It is believed that greenhouse emissions can be reduced by the use of renewable energy sources. However, until the time these energy sources can be developed adequately to generate the required amount of energy, fossil fuels are the most robust energy sources available. Coal is an abundant resource in the United States. To lower the emission of the greenhouse gases from coal power plants, several technologies have been suggested using amine absorption: 1) CO2 capture from power plants by scrubbers 2) Integrated Gasification Combustion Cycle (IGCC) 3) Chemical Looping Combustion, and 4) Oxy-Fuel Combustion [1]. Abraham et al. [2] in 1982 proposed the technology of oxy-coal combustion to produce CO2 for Enhanced Oil Recovery (EOR) . Wang et al. [3] explored the feasibility of burning pulverized coal in O2/CO2 atmosphere. Oxy-Fuel Combustion technology has been suggested as the most promising strategy for conventional coal power plants to generate electric power [4]. 1.2 Oxy-Coal Combustion Technology Description In oxy-coal combustion, a mixture of pure oxygen with purity of greater than 95% and the recycled flue gas are used for combustion of coal. The product of this combustion 2 is a gas consisting of mostly water and CO2. The recycled flue gas is considered as an inert gas to modulate the combustion temperature as well as distribute the heat of combustion in the furnace [1,4]. The required oxygen is provided from an air separation unit (ASU). A part of the flue gas, not recycling, will go through the desulfurization unit and water condensation unit. After condensation of the moisture, the flue gas is a gas containing nearly 95% CO2. This gas will be liquefied and can be applied for EOR or other industrial purposes. A schematic of oxy-fuel combustion is provided in Figure 1[5]. The differences between air and oxy-combustion are: 1) Combustion medium (O2/N2 in air-firing and O2/CO2 in oxy-firing) 2) Composition of combustion products and particulate concentration 3) Thermal properties such as heat capacity, density, and viscosity as well as radiative properties of the furnace gases 4) Radiant and convective heat flux profiles in oxy-combustion Due to the similarity of the oxy-firing and air-firing systems, oxy-fuel combustion is known as one of the most promising remedies for retrofitting conventional power plans; however, there are issues that need to be resolved: 1) O2 supply 2) Corrosion due to recirculation of flue gas 3) Purity of CO2 in flue gas 4) Heat transfer 5) Need for submodels and simulation of the oxy-fuel combustion 6) Oxy-combustion cost 7) Burner aerodynamics and oxygen injection 3 1.3 Ignition and Flame Stability in Oxy-Coal Combustion When a coal particle is placed in a furnace, it is heated to a temperature at which pyrolysis starts. During this process, the volatile contents of the coal particle get released. The pyrolysis temperature depends on the coal and the heating rate of the particle. The pyrolysis and ignitibility of coal volatile species and their transport from a particle to its surrounding determines whether the ignition of an isolated coal particle occurs either heterogeneously or homogeneously (in gas ignition) [6]. Homogeneous reaction occurs when the volatiles eject early due to the environment conditions, they react with oxygen, and start the ignition before oxygen reaches the particle surface. In the cases where combustion is incomplete or liberation of volatiles is retarded, oxygen can reach the particle surface and have a heterogeneous reaction with the particle. Katalambula et al. [7] studied the effect of the volatile matter cloud surrounding a single coal particle on the ignition mechanism. They found that the amount of volatiles surrounding a coal particle being influenced by natural or forced convection affects the ignition mechanism and temperature. Volatile matter content has a significant effect on the ignition temperature, and is almost negligible under forced convection. Additionally, Katalambula and his colleagues showed that the ignition temperature increases correspondingly with the coal particle size. Kharbat and Annamalai [6] applied a digital image processing technique to investigate the ignition and combustion characteristics of isolated coal particles as well as the interactive combustion of a binary coal array. Their experiment concluded that the ignition of high, medium, and low volatile coals is always heterogeneous, and the release of volatiles and secondary gas ignition happened after 4 heterogeneous ignition. They suggested radiative interaction between the particles might have occurred in group particle ignition [8]. Du and Annamalia [9] applied a theory, known as Thermal Explosion Theory, to define the criterion of heterogeneous ignition. According to Thermal Explosion Theory, ignition starts if the rate of heat release caused by chemical reaction is larger than the rate of heat loss of the coal particle. They derived a method to determine heterogeneous char ignition temperature using Thermal Explosion Theory, and determined that heterogeneous ignition is a great function of coal particle ignition activation energy and the surface temperature of the coal particle. These two factors are highly influenced by the heat of ambient gas, radiant heat transfer from the wall to the particle surface, heat of reaction, and activation energy of particle ignition (this factor is a function of coal composition), the particle heat loss during its ignition. Shaddix and Molina [4] claimed the presence of CO2 retards single particle coal ignition. Particle devolatilization proceeds faster with higher O2 concentrations; however, the existence of CO2 decreases the devolatilization rate due to its negative influence on the mass diffusion of O2 and volatiles in the CO2 environment. Qiao et al. [10] proved the impact of the thermal conductivity of CO2 as another factor for the weaker ignition of coal in O2/CO2 environments. Wall and Buhre et al. [1] in their review present the impact of CO2 on coal ignition retardation as well as improvement of flame stability due to pure O2 injection. Shaddix et al. [11], in a different study, found that O2 concentration decreases ignition delay in a manner that could be used to counteract the retardant effects of CO2 . Kimura et al. [12] showed that the flame propagation speed is much lower in an O2/CO2 atmosphere than in O2/N2 and O2/Ar, and that in O2/Ar, it was the highest. The 5 propagation speed increases as oxygen concentration increases in all cases, as expected. Nozaki et al. injected 15% to 20% of the total oxygen into the furnace directly [13]. His group at IHI found hydrogen cyanide (HCN) formation in the near burner zone increases in O2/CO2 with oxygen injection, and he claimed that devolatilization becomes more active at higher gas temperatures realized near the burner by oxygen injection. 1.4 Sulfur and Recycled Flue Gas (RFG) As mentioned previously, one of the focuses of this research is on the effects of minimization of recycle ratio on the radiant heat flux in oxy-coal combustion. It has been proven that due to the existence of high recycle ratio in oxy-coal combustion, the amount of SO2 in the system increases. One of the major issues in oxy-coal combustion is corrosion resulting from SO2 and SO3 formed due to the flue gas recirculation. Therefore, a brief description regarding sulfur issues in oxy-coal combustion is provided in this section. It is important to note that SO2 emissions in an oxy-case are lower than in an air case in terms of the total sulfur mass output; however, the in-furnace SO2 concentration has been reported to increase significantly during O2/RFG combustion. The SO3 concentration is approximately 2.5 to 3.0 times higher in the oxy-case than in the air case [14]. A consequence of reducing volumetric flow and introducing recycled flue gas during oxy-fuel fired combustion is an increase in SO2 concentration in the flue gas. Ochs et al. [15] calculated a rise in SO2 levels from 200 ppm under air-fired combustion to 900 ppm under oxy-firing. Also, pilot scale experimental results by CANMET, Argonne National Laboratory, and IEA Green House R&D report an increase in SO2 from air-firing to oxy-firing about 2 to 4 times. In addition, there are arguments regarding condensation of H2O 6 prior to recycle [14]. Weller reported wet recycle of flue gas has shown an increase in SO2 compared to dry recycle [16]. Oxy-fuel technique is one of the CCS techniques, and as mentioned before, the captured CO2 can be used for other industrial purposes such as EOR. Therefore, the quality and purity of CO2 becomes an important factor. Li et al. [17] calculated the energy for purification and compression of oxy-fuel gas. They proved the significant effects of SO2 on energy for compression, distillation, and final CO2 purification. In the radiative section of the furnace, the formation of H2 S and COS can contribute to corrosion if the flame is not well managed. In the convective section, the formation of SO3 impacts the ash deposit and enhances the transfer of iron from heat transfer surfaces. Furthermore, it reacts with NO2 and generates sticky compounds of ammonium bisulfate, which also fouls heat exchanger elements. In sections of the furnace where the flue gas temperature is below the dew point of the flue gas, the formation of H2SO4 is very undesirable because of its corrosive potential [14]. Methods have been suggested to control SOx in oxy-fuel combustion such as the use of low sulfur and high calcium coals, limestone injection, sulfur scrubbers, condensers in flue gas line, etc. However, performing any of these changes requires consideration of the cost and energy consumption, which make oxy-fuel combustion less efficient compared to other techniques. In addition, it is reported that generally the fly ash produced in the oxy-case contains slightly more sulfur and the furnace deposits contain significantly more sulfur compared to air case samples [14]. 7 In this study, the effects of directed oxygen injection in triple co-axial turbulent diffusion flames on NOx formation was investigated; therefore, a brief description regarding the history of research on NOx formation in both air and oxy-combustion was provided. In conventional air combustion, the generally accepted path ways for NOx formation are the following three mechanisms [18]. Thermal NOx is the production of reactions between O2 and N2. These reactions are temperature-dependent in a sensitive manner, and at temperatures above 1500 °C, they become more important. Thermal NOx formation reactions are described in detail by the Zeldovich mechanism. N2 + O ^ NO + N N + O2 ^ NO + O N + OH ^ NO + H The nitrogen of fuel exists in the char and volatile matter of fuel. Char nitrogen reacts through heterogeneous reaction chains and converts to either N2 or NO. The nitrogen of volatiles is released during devolatilization, and volatiles further decompose into cyanide and amine species. These intermediate species may react to produce N2 or NO. Prompt NOx is the product of the reaction of hydrocarbon radicals with N2 in the fuel-rich zone of the jet. These products are in the cyanide form that will turn into NO after reacting with O2. Due to the nature of oxy-coal combustion, the amount of the molecular nitrogen existing in the furnace is very small or theoretically zero. Therefore, one can consider only formation of the fuel NOx in oxy-fuel combustion. Not only has carbon 1.5 NOX Formation 8 capture and sequestration been a great motivation to pursue oxy-fuel combustion, but the potential for reducing the NOx emissions from power plants considerably compared to conventional power plants has been one of the key drivers in the oxy-fuel combustion research. There are a number of mechanisms for the reduction of NOx in oxy-fuel combustion presented in different literatures. Bose and Wendt et al. [19] reported the slow devolatilization of the nitrogenous compound in the far post flame plays a critical role in supplying HCN, which ultimately drives reactions forming N2 from the NO formed from the fuel nitrogen before the oxygen disappeared. They found there is an indirect heterogeneous influence on the destruction of NO in which NO is decreased using fuel as reducing agent in the fuel-rich zone. Okazaki et al. [20] observed NO reduction in oxy-fuel combustion. They reported that the reactions of NOx with hydrocarbons are the most significant mechanism in reducing NOx emissions, and lowers NO about 50% to 80% in oxy-fuel combustion. Norman et al. [21] at Chalmers University studied nitrogen chemistry and NOx reduction at high temperatures in oxy-fuel combustion. They found that at temperatures below 1400 °C, the Zeldovich mechanism is not active, and the reburning mechanism is dominant. At higher temperatures, the reverse Zeldovich mechanism dominates the NO reduction. High-temperature combustion only reduces NOx when the nitrogen concentration is low, as during oxygen combustion. The influence of N2 concentration is shown in Figure 2 and Figure 3 [21]. A low equilibrium concentration is obtained at low concentrations of N2, high oxygen purity, substoichiometric combustion zones, and low temperatures. Sufficient conversion rate by the relatively slow Zeldovich mechanism 9 needs longer residence time and high temperatures in the reaction zone, and the temperature profile should be decreasing along the furnace in order to lower the equilibrium concentration of NO. 1.6 Heat Transfer The studies of oxy-coal combustion performed by the Argonne National Lab (ANL) in the 3 MW pilot-scale facility located at the Energy and Environmental Research Corporation showed that with wet recycle, an oxygen concentration of 23.8% through the burners resulted in the same overall heat transfer as air-firing. Also, it was found for an oxy-firing case with dry recycle that an overall oxygen concentration of 27% is required to obtain an equivalent overall heat transfer in both cases [1]. In addition, ANL used a two-dimensional heat transfer and combustion zone models to simulate a 50 MW power plant. This model was used to determine the effect of recycle molar ratio on the heat transfer efficiency of the boiler. The main criterion for determining the optimum recycle ratio was to achieve a similar heat efficiency with both oxy-combustion and air combustion. Their results showed that the optimum dry-recycle molar ratio was calculated to be 2.7. Similarly, the wet-recycle molar ratio was calculated to be 3.2. In order to validate this simulation, an experiment was conducted by Energy and Environmental Research Corporation. The optimum wet-recycle molar ratio was found to be 3.25 and the optimum dry-recycle molar ratio was found to be 2.6 by comparing the overall heat transfer to both the radiative and convective heat transfer surfaces to that achieved with the baseline air-firing operation [1,16,22-25]. Buhre et al. reported that the major contributor of heat transfer from a flame is thermal radiation from water vapor, 10 carbon dioxide, and soot [1,16]. In oxy-fuel combustion with RFG, the concentrations of carbon dioxide and water increase tremendously, leading to a significant change in the heat of radiation. As shown in Figure 4, Smart et al. [26] in 2010 reported that a similar adiabatic combustion temperature to conventional air-firing is obtained at a flue gas recycle ratio of 69% [26]. This value depends on the coal or fuel composition and could fluctuate within the range of 68% to 72%. The studies showed for the Russian Coal, a radiative heat flux profile similar to air-firing can be obtained for recycle ratios between 68% and 72% at 3% exit O2 and between 68% and 75% for 6% exit O2 [26,27]. Andersson and his colleagues [28] at Chalmers University investigated the radiative heat transfer of oxy-coal combustion in a 100 kW oxy-fuel test facility. The flue gas recycle ratio was varied, so that, in principle, the same stoichiometry was kept in all cases, whereas the oxygen fraction in the recycled flue gas mixture ranged from 25 to 29 vol.%. Radial profiles of gas concentration, temperature, and total radiation intensity in the furnace were measured and also calculated using a model developed at Chalmers University. Narrow angle radiometers were applied to measure the radiation heat flux. The model in their study is able to include all three sources of radiation (gas, soot, ash) in coal combustion. Malkmus Statistical Narrow Band Model (SNBM) was applied to model the gas radiation intensity. They found temperature, and thereby the total radiation intensity of the oxy-fuel flames, increased with decreasing flue gas recycle rate. They found that with a 25% overall oxygen concentration in the feed gas, the same radiant heat flux can be achieved (see Figure 5 and Figure 6) [28]. Their results showed particle radiation has the most dominant role in the radiation of oxy-coal flames. 11 In addition, they proved the increase in total gas emissivity from air to oxy-fuel operation with dry RFG only is about 5%; however, it increased more than 20% for the same case with the wet-recycling. Thus, the moisture content of the flue gas is more important than content of carbon dioxide. However, particle radiation plays the most important role in radiation [28,29]. 1.7 Co-Axial Jets Forstall [30] studied the impact of the velocity ratio on turbulent mixing in a co­axial jet containing central and annular streams. He concluded that material diffuses more rapidly than momentum; therefore, the principle independent variable determining the shape of the mixing regime is the velocity ratio. Chigier and Beer focused on the region near the nozzle in double concentric jets. They considered two streams as two different central cores and described how these two streams emerge based on their fluid entrainment due to velocity differences [31-33]. In addition, they provided plots that define the length of each stream before they emerge as a function of velocity ratio for co­axial jets. In another study, Beer [34] reported that when the secondary velocity is low and the density of the secondary fluid is considerably higher than that of the recirculating fluid, more recirculation and less secondary air is entrained in the early part of the primary jet. This means coal particles will be heated faster, and it has a significant effect on the stability of the flame. Durano [35] measured the mean velocity of a co-axial jet for three different velocity ratios. He reported co-axial jets obtain a self-preserving state faster than single round jets. Kawn and Ko [36], in their study, divided the near field zone of double concentric jets into three zones. The nearest zone to the nozzle is called the Initial Merging Zone. The 12 length of the initial zone depends on the core velocity ratio. The next region is the Intermediate Zone. In this zone, the streams are still considered as potential cores. Disparaging of the central initial core is the indication of the end of this region. As shown in Figure 7, the last zone is where the two cores are fully mixed, and the jet is treated as a single jet. Crowe et al. [37] recommended the evaluation of Stokes Number (St) in turbulent two-phase jets. St is defined as the ratio of Tp over Tf where Tp is the response time of the particle and Tf is defined as the fluid time scale, which is a function of characteristic length and velocity of the fluid. At small values of St, the particles follow the fluid and maintain the velocity equilibrium with the fluid. For large values of St, the response times of the particles are greater than the fluid response time. Thus, the particles do not follow the fluid. For the St in the order of one, the particles follow the fluid partially. Kennedy and Moody [38] in their research found St has the highest value in the near field, and it becomes much smaller downstream in the jet. Preliminary high-performance computer simulations [39] using Large Eddy Simulation (LES) suggest that coal particle ignition occurs in clusters of small particles, many of which have been transported radially outwards by large eddies. In a cold flow study, Budilarto et al. [40] used Laser Doppler Velocimetry to show that smaller particles migrated to the edge of co-axial two-phase turbulent jets. Their results showed that the dispersion of particles was enhanced with decreasing particle size and increasing the velocity ratio greater than 1.0. They found that the addition of finer particles leads to an increase in the mean velocity of the coarse particles near the pipe center and, also to a less flat radial distribution of mean velocity of the coarse particles. 13 The first criterion of ignition was determined to be the amount of volatile matter; however, it was proved that a more correct way to determine the combustibility is the index of Fuel Ratio (FR), which is the ratio of fixed carbon content / volatile matter content [1,13,41]. More research showed that the fuel ratio is not always a robust index to anticipate the combustibility and ignition behavior. Oka et al. [42] studied carbon burnout in seven types of coal, and found a general trend that the higher fuel ratio coals have higher carbon burnout. However, some of the coals do not follow this trend. Blending several types of coal is a crucial process to improve the efficiency of power plants. An optimum blend of coal based on the type of power plant can help to reduce costs, meet emission regulations, improve combustion behavior, control ash deposition, enhance fuel flexibility, and extend the acceptability of coals. Research on blending coal is categorized in different aspects. Grinding, flame stability, carbon burnout, slagging and fouling, pollutant emissions, ash disposal, and overall heat transfer and efficiency are the most important aspects of coal compositions and their blends that need to be investigated. The influence of the maceral composition of the coal on the ignition and flame stability has been pursued by Su et al. [43,44] who suggested a criterion, the maceral index, to predict burnout behavior of the coal and blends by including the effects of maceral composition such as liptinite, vitrinite, and inertinite. The burnout of coal or coal blends depends on the amount of volatile mater, physical structure of char, and the diffusion rate of char burning. This relation fundamentally depends on the maceral composition of the coal or blend. Liptinite has the highest hydrogen content and volatile matter, and liptinite is associated with ignitibility and flame stability [45]. In addition, Su 1.8 Coal Chemistry 14 et al. [43] defined an expression for the maceral index (MI) to predict burnout in two different pilot-scale facilities at Australian Coal Industry Research Laboratory (ACIRL) and Energy and Environmental Engineering Research Corporation (EER). the volume percent of inertinite. All of these indices are measured free-mineral based. R represents the mean maximum reflectance and HV is the heating value of the coal. RF presents the reactivity of coal or blend. It was found that MI correlates with the burnout, and can potentially be used to predict ignitability as well. Based on very simple models, in order to obtain a stable flame, coal particles need be heated to temperatures where the heat generation is balanced with the heat loss. The rate of consumption and production are respectively a function of velocity and coal composition. Also, it is found that coals with the same proximate analysis may not have the same ignition and flame stability characteristics. This is because the ignition and flame stability are the results of fast heat release, not only the amount of volatile matter. Hallate et al. [46] researched the dependency of coal oxidation on coal particle size using thermo-gravimetric analysis (TGA). He found the oxidation rate of vitrinite increases with decreasing particle size while oxidation of inertinite is independent from the particle size. It is important to note that all of these studies were conducted in the air atmosphere; therefore, the behavior of coal ignition in an O2/CO2 environment based on its composition and chemistry is not yet well understood. (1) where L is liptinite volume percent, V is the volume percent of vitrinite, and I stands for 15 This work builds on a previous study at the University of Utah, performed by Zhang et al. [47,48]. That study focused on the effects of the following on coal particle ignition and flame stability: 1. The effect of partial pressure of oxygen in the primary stream 2. The effect of partial pressure of oxygen in the secondary stream 3. The effect of preheat temperature of the secondary stream 4. The effect of transport medium Zhang et al. [47] proved that by increasing the partial pressure of oxygen in the primary stream, flame stability increases. The total amount of oxygen in this test was kept constant, and the oxygen was transported consequently from the secondary stream to the primary stream such that partial pressure of oxygen in the primary stream changed from 0.0% to 20.9% in five different cases. It is important to note that the preheat temperature was maintained at T= 489 K. Also, the velocities of both streams were kept constant in order to minimize the effect of jet aerodynamics as much as possible. In order to investigate the effect of preheat temperature, the previous test was repeated; however, the preheat temperature of the secondary steam was changed to T=544 K. The results of this new test and the previous test showed that by changing the secondary stream preheat temperature, the flame stability increases such that even at low oxygen concentration in the primary stream, coal ignition occurs earlier, leading to a more stable flame. In addition, the effect of an increase of oxygen in the secondary stream was explored by increasing the total amount of oxygen in the secondary stream. It is important to note that no oxygen was fed into the primary stream to observe the contrasting effects 1.9 Previous Work at the University of Utah 16 of the presence of oxygen in the secondary stream. The results of this research showed that the increase of oxygen in the secondary stream facilitates coal particle ignition and results in a shorter flame stand-off distance which provides more flame stability. Furthermore, a study was performed on the impact of the transport medium or combustion environment. Two sets of experiments were executed. One was performed in an O2/CO2 environment and the other one was in O2/N2. This experiment was very similar to the first experiment explained above, but in a different environment. The results of this study proved that the ignition of coal particles start at earlier stages in O2/N2, leading to more stable flames with short stand-off distances. It was explained that the concentration of O2 at the surface of the coal particle plays a key role in particle ignition and, due to higher molecular diffusion of O2 in N2 compared to O2 in CO2 , the concentration of oxygen increases earlier, causing a faster ignition and resulting in a more stable flame. The results of these experiments were consistent with the results that Shaddix and his colleagues found parallel to this research [4,11]. 17 Figure 1. Oxy-Fuel Combustion Schematic Figure 2. NO reduction ratio for various combinations of N2 and NO Normalised Adiabatic Flame Temperature 18 0%N2 i% n 2 10% n 2 lOOOppm NO - B -------- A - O 1300 1500 1700 1900 2100 Temperature [°C] Figure 3. NO reduction ratio for isolated reduction Dry Oxyfuel Operation Normalised to Air Operation Peak Radiation Flux, Convective heat transfer and calculated flame temperature Russian coal 60% 65% 70% 75% 80% Effective Recycle Ratio Figure 4. Effect of flue gas recycle ratio on flame temperature and radiative heat flux Normalised Radiative and Convective Heat Flux 19 Figure 5. Comparison of radiation intensity in air and oxy-firing Figure 6. Comparison of radiant intensity measurements and gas radiation modeling at 384mm from the burner inlet in the Chalmers furnace. The furnace walls are located at radial distances of 800mm 20 Figure 7. Near field zones of double concentric jets 2. RESEARCH OBJECTIVES, MOTIVATION, AND ORGANIZATION OF THIS DISSERTATION One of the advantages of oxy-coal combustion is having the oxygen concentration and injection configuration as a crucial variable compared to air combustion. In oxy-coal combustion, especially in simple burners such as IFRF type-0 burners, two streams, namely, the primary stream and secondary stream, exist. The primary stream is utilized to transport pulverized coal using a carrier gas, and the secondary stream is applied as an oxidant stream containing oxygen and carbon dioxide from the recycle line. As mentioned before, coal combustion flames are less stable at 21% oxygen in a CO2 environment as compared to an N2 environment (air combustion). Therefore, it is very important to research the most suitable percentage and configuration of oxygen in order to find the most optimum flame. Previous research has elucidated this aspect of oxy-combustion; however, the impact of the injection of pure oxygen as a segregated stream is not well understood. In this research, the following items are determined to be the main objectives: 1. To extend the results of Zhang et al. [49] with a view to understanding the effects of coal composition on axial flame attachment and detachment under oxy-coal combustion conditions 2. To understand how advanced O2 distribution strategies for co-axial burners in the University of Utah, 100 kW Oxy-Fuel Combustor (OFC) might affect flame 22 stand-off distances and flame stability, using as examples two configurations : namely, one with pure oxygen in an inner annulus and the other with pure oxygen in a central pipe. 3. To determine the effects of these oxygen input strategies not only on flame stability, but also on flame length, heat flux, and NOx formation. 4. To determine the consequences on flame stability, flame length, heat flux, and NOx formation of segregating all the input oxygen into one stream composed of 100% oxygen, using the co-axial burner configuration detailed above. 5. To determine the effects of minimizing the amount of recirculated flue gas on flame stability, flame length, heat flux, and NOx formation, using the co-axial burner configuration detailed above. 6. To contribute to the validation, with uncertainty quantification, of coal jet ignition submodels of the co-axial burner configurations referred to above. 7. To interpret the data obtained using simple mechanistic concepts in order to understand why certain configurations resulted in observed changes in the flame near field aerodynamics, such as stand-off and flame stability, in contrast to simulations, which are outside the scope of this project. This dissertation is organized as follows: first, the experimental equipment used is described in Chapter 3. This is followed by a detailed description of the methodology used to extract quantitative PDFs describing stand-off distances and flame length from photo images obtained. Chapter 5 is based on a paper suitable for publication and entitled, "THE EFFECT OF COAL COMPOSITION ON IGNTION AND FLAME STABILITY IN CO-AXIAL 23 TURBULENT DIFFUSION FLAMES." This portion of the research deals with extending Zhang's work to different coals, and has been extracted from a manuscript with the same title, ready for submittal for publication in Energy and Fuels. It contains its own self-contained conclusions. Chapter 6 is based on another suitable paper for publication, entitled "NEAR FIELD AERODYNAMIC EFFECTS OF PURE O2 INJECTION IN CO-AXIAL OXY-COAL TURBULENT DIFFUSION FLAMES." It contains material taken from a second manuscript that is also ready for submittal for publication in Energy and Fuels. This chapter, therefore, also contains its own conclusion. Chapter 8 contains results from the third phase of this work and focuses on heat flux measurements from systems with minimum flue gas recycle, and for consistency with the preceding chapters, also contains its own conclusions. This chapter is followed by a discussion of future work in Chapter 9. 3. EXPERIMENTAL 3.1 Oxy-Fuel Combustor (OFC) The combustion furnace employed in this research is a nominally 100 kW combustor with the capability for both oxy- and air-firing at the University of Utah. This combustor was initially constructed by Zhang [49] and modified for flue gas recycle by Morris [50]. It was then further modified in this work, to accommodate multiple inlet streams of various compositions into the burner, and to allow possible cooling in the upper combustion chamber. As shown in Figure 8, Figure 9, and Figure 10, the oxy-fuel combustor (OFC) is equipped with a recycle system to provide similar oxy-operating conditions as industrial oxy-coal combustors. As presented in Figure 8, OFC is a down-fired flame combustor. The top section of the furnace is called the burner zone. The position of the burner is on top of this section, and during the combustion process, the flame is located mainly in this part of the furnace. Because this research explores the aerodynamics and radiation of the flame in various conditions, most of the experiments are focused on this zone. The second part of the OFC following the burner zone is the radiant zone. The third part located horizontally in the bottom of the furnace is the convective zone. This section is equipped with water heat exchangers to simulate an environment similar to industrial pulverized coal boilers. There are 26 circular ports along the furnace symmetrically that are applied as probe inputs. K-Type thermocouples are installed along the furnace to measure the temperature of combustion. gaseous products. It is important to note that in this research, the recycle system was not applied. Therefore, instead of recycled flue gas (RFG), pure CO2 was fed from a tank. The qualitative diagram and the details of the piping system used in this study are illustrated in Figure 11. The first valves used in each pipe line are ball valves to stop possible hazard caused by gas leak manually. The second valves are solenoid valves controlled electronically, allowing them to be connected to a central control system. In addition, the flow is measured using a mass flow controller. The mass flow controllers are calibrated separately for each gas. Solenoid valves and mass flow controllers communicate with each other to maintain the desired flow. This control is managed by a control system program called OPTO22 that will be explained in this chapter. A picture of a part of the pipe lines equipped with the valves and mass flow controllers especially for CO2 and O2 is shown in Figure 12. 3.2 Burner Zone The burner zone of the OFC is where the flame is located. Since this study mostly focuses on flame aerodynamics and near burner zone combustion, it was critical to equip the burner zone section with instruments to obtain information about the flame under desired combustion operating conditions with the highest possible control. The burner zone contains four windows located in its quadrants allowing optical access to the flame and the chamber of combustion. Photo images of the flame are captured from these windows. These images potentially provide information regarding the flame stability using the photo-imaging technique that will be discussed later. In addition, having the windows in four quadrants permits applying other techniques such as Particle Image Velocimetry (PIV) or Particle Shadow Velocimetry (PSV). 25 Details of the design of these windows are provided in Figure 13. The burner zone is a part of the furnace that has the highest temperature. To prevent heat loss from this section, 3 inches of fiber board 2600 is employed. The fiber board insulation is formed from a special blend of regular Fiberfrax alumina-silica fibers and is able to tolerate temperatures up to 2600 °F. In order to have a control on the furnace wall temperature in the burner zone, ceramic heaters are applied. These ceramic plates are 6 inches wide, 10 inches long, and have the thickness of ^ inch. In order to cover the surface of the chamber for the purpose of managing the wall temperature uniformly at desired values, 24 ceramic plate heaters supplied with 840 W of power are embedded into the inside surface of the chamber. The heaters are composed of mainly 38% Al2O3 and 60% SiO2. The ceramic plate heaters are purchased from the company Thermcraft. The thermal conductivity of the heaters is 0.22 W/m.K. K-Type thermocouples are attached to the back of the heaters that allow setting the temperature of the wall heaters to desired temperatures using the OPTO22 control system. Six circular ports benefit the burner zone for the purpose of collecting several data of the flame. For instance, radiometer probes have been employed to measure the heat of radiation of the flame in the OFC. The principles of the probes and results of this study are discussed later in this chapter. 3.3 Convection Zone As shown in Figure 11, the convection zone of the OFC consists of four main sections, and each section contains two heat exchangers. A picture of one of these sections is provided in Figure 14. The two heat exchangers shown in this picture are equipped with 26 two K-Type thermocouples in their water outlet stream. The thermocouples are very important for safety. In order to measure the flue gas temperature, one K-Type thermocouple is installed on each section. The convective zone of the OFC is located in the lowest part of the furnace; therefore, ash buildup caused by combustion happens mostly in this area. The cleaning of this zone of the furnace after ten days of experimentation is suggested. The ash buildup causes pressure buildup in the furnace that can potentially create serious hazards. 3.4 Gas Heater As shown in Figure 11, a gas heater is installed in the line of the secondary stream in order to increase the secondary stream gas temperature to desired values. The power of this gas heater is 2 kW. The power of the gas heater was determined after calculating the amount of heat required to heat up a mixture of flue gas containing O2 and CO2 to a temperature of 520 °F. Figure 15 shows an inside view of the heater. The heater consists of an electric element and a K-Type thermocouple. The electric element is located in a metal shell, and the gas enters from the left side of the element and exists from the right. The shell is insulated using insulation blankets and an aluminum shield. The picture of the insulated gas heater is represented in Figure 16. 3.5 Wide Angle and Narrow Angle Radiometers One narrow angle (NA) and three wide angle (WA) radiometers were applied to measure the radiant heat flux of the flame in various combustion operating conditions (see Figure 20 and Figure 21). The design of the wide angle radiometers was dedicated by Praxair, and was built at the University of Utah. The WA probes measure the total radiation coming to the location of measurement, and the NA probe measures the 27 radiation being emitted from a location in the small view of the probe. The WA probe has nearly a full hemispheric view. The view of the wide angle radiometers is discussed in this chapter. This is limited slightly by imperfections in the machining at the inlet orifice for the elliptical element. In addition, the reflectance of the gold coated surface of the ellipsoidal cavity has a great impact on the efficient view angle of the WA radiometers. The narrow angle radiometer view is a small angle of about 5 to 6 degrees that can collect the emitted radiation of radiant subjects in that path. The typical locations of these radiometers are provided in Figure 17 and Figure 18. It is important to note that the location of the probes might change based on the type of experiment. The details of radiometer construction will be discussed further in later chapters. 3.6 Wide Angle Radiometer The WA radiometer measures the total radiation coming to the measurement location from subjects seen in the view zone. Due to the ability of a wide angle radiometer to have a nearly hemispheric view, radiation from the flame and wall can be captured using the WA radiometer (see Figure 20). The schematic of the principle of the wide angle radiometer is shown in Figure 19. A copper ellipsoidal element is located in the front of the probe that collects all the radiation in the radiometer view to the orifice [51]. The orifice is located at the front of the ellipsoidal cavity that collects the radiation from the furnace and flame. The material of the cavity in this probe is from copper, and the inside surface of the cavity is gold coated. The gold coated surface amplifies the collection and reflection of all radiation received in the cavity to the thermopile. In theory, the ellipsoidal cavity can capture all the radiation in an imaginary hemisphere view located at the front of the 28 probe; however, due to the possible roughness of the surface of the cavity, the angle of the view becomes smaller, implying the radiometer can capture only a portion of the radiation in the aforementioned hemisphere. A calibration was performed to obtain the view angles of the radiometers. The detail of the calibration is explained in this chapter. All the radiation coming inside the cavity is focused onto a stainless steel thermopile at the opposite focal point. The thermopile consists of a hemispherical pellet and a cylinder that is connected to the pellet and a cooled mass. All three parts are made of stainless steel. The surface of the pellet is blackened and oxidized in order to absorb nearly 95% to 98% of the radiation. The metal mass is kept cool by the cooling water. There are two constantan wires; one is connected to the pellet (high temperature segment), while the other is connected to the metal mass (low temperature segment). This temperature gradient generates a current in the wires that is proportional to the energy received by the pellet. A picture of one of the wide angle radiometers used in this research is provided in Figure 20. The pipe around the probe plays the role of a water cooling jacket heat exchanger. Water enters the inner side of the jacketed heat exchanger and exits from the outer shell. This configuration lowers the temperature of the cavity as much as possible, minimizing errors associated with radiation from the inner surface of the ellipsoidal cavity. Changes in the flow rate of water do not have a significant effect on the calibration of the radiometer; however, the water flow rate should be high enough to adequately lower the temperature of the cavity. A 2.0 gpm rotameter is installed to measure the water flow rate. The water flow rate is kept constant at 1.25 gpm in this study. In order to prevent ash buildup in the ellipsoidal cavity, CO2 needs to be purged into the bottom of the cavity, carrying the ash out through the orifice. The purged gas circulates inside the cavity and cleans the inner surface. The presence of any ash on the 29 inner surface of the cavity can have a tremendous impact on the reflectance of the cavity, resulting in significant errors in the readings of the radiometers. The amount of purged CO2 has a considerable effect on the calibration of the radiometer due to the high emissivity of CO2 and its participation in radiation. The amount of CO2 purged in the system was kept constant at 20 standard cubic feet per hour (scfh). This value needs to be maintained constant in all the experiments after the calibration. The exit velocity of CO2 from the orifice is calculated to be about 4.1 (ft/s). A 22 (scfh) rotameter was applied in the system to control the amount of CO2. 3.7 Narrow Angle Radiometer The NA radiometer measures radiation being emitted from a location (see Figure 21). In this study, the location was 175 mm below the burner tip. The investigation of the radiation of this region provides a better insight regarding flame stability. The schematic of the narrow angle radiometer is shown in Figure 22. First, the radiation enters the pipe and, after passing through, sees a focusing lens. Radiant light after passing through the lens is focused on a thermistor, inducing a change in resistance as the thermistor temperature changes. Another thermistor is placed next to the other thermistor in order to capture the ambient effects. The combination of the two thermistors creates a Wheatstone bridge circuit. The difference between the two thermistors generates a voltage across the Wheatstone bridge circuit, corresponding to the radiant heat received by the radiometer from a location. In the same manner as the wide angle radiometer, a water heat exchanger is required to lower the temperature of the probe; however, changes in the flow rate of water do not have remarkable effects on the calibration of the radiometer. A 3.5 gpm rotameter was installed to measure the water flow rate, which was kept constant 30 at 1.25 gpm. The purpose of purging CO2 is to prevent any particles or contaminants (condensing ash, tars, etc.) from entering the probes. Therefore, especially in wide angle radiometers, the amount of purging cannot be too low. At 20 scfh, as it was mentioned, the exit velocities for wide angle radiometers are 4.1 ft/s. This value for the same amount of CO2 in a narrow angle radiometer was calculated to be 1.7 ft/s. The amount of purged CO2 in the NA radiometer is not as critical as in the wide angle radiometer, for any particles covering the inner surface of the cavity can create errors in the performance of the WA radiometers. 3.8 Calibration of Radiometers 3.8.1 Thermal Response of Radiometers The calibration of the radiometers is critical, and is required before measurements can be taken. Despite the fact that CO2 is being purged into the system, there is a chance that a small amount of particles can enter the ellipsoidal cavity and settle especially on the edge of the orifice, inner surface of the cavity, or even on the pellet, lowering the reflection efficiency of the radiation considerably. Additionally, due to the heat of the furnace, there is always a chance of physical shocks to the thermopile that can damage the reception of radiation by the thermopile. The same issue may present itself in the orifice and should be taken into consideration. In order to calibrate the radiometers, a blackbody radiator was employed. The blackbody radiator is a thermally insulated and electrically heated graphite tube cavity. The image of the blackbody radiator used in this calibration is shown in Figure 23. The blackbody radiator contains a thermocouple that can regulate the temperature of the enclosure to any desired value. In order to calibrate the radiometer, it should be subjected 31 to radiation from a blackbody radiator set to different temperatures. The results of the calibrations are presented in a linear graph showing voltage signals (mV) as a function of received radiant energy from the blackbody radiator. The energy from the blackbody radiator is only a radiant energy that can be controlled based on the set temperatures of the blackbody radiator. Limitations such as overheating the thermopile, or lack of capability of the blackbody to reach to high temperatures (above 2000 °F), need to be taken into consideration in the calibration process. Figure 24 and Figure 25 show calibrations of both wide and narrow angle radiometers performed on 05-09-2011. As shown, the radiant power is plotted as a function of the voltage received by the radiometers. The amount of radiant power corresponds to the temperature of the blackbody (Q = aT4). Also, it is important to note that the response time of the wide angle radiometers is of the order of nearly one minute, not allowing monitoring of the radiation instantaneously. However, the narrow angle thermistor is more sensitive to temperature differences due to radiation, and the values might change in the order of one second. Therefore, the voltage reported is an average over the time the measurements have been performed. 3.8.2 Angular Response of WA Radiometers The main purpose of this experiment was to find the angular response of the WA radiometers. Additionally, by this test, the field of view of the radiometers was obtained [52]. For this experiment, each radiometer was positioned in front of the blackbody radiator with the radiometer axis parallel to the axial axis of the radiating aperture of the blackbody. The radiometer axis was aligned with the blackbody axis using a protractor. Several angles of 0°, ±15°, ±30°, ±45°, ±60°, ±75°, and ±90° were chosen for the 32 measurements. At each angle, the radiometers were aligned in front of the blackbody radiometer for 10 minutes, and the average of the signals (mV) was counted as the response signal of the radiometer at that specific angle. The results for the three radiometers are presented in Figure 26, Figure 27, and Figure 28. As shown in the figures, the response of the radiometers is normalized based on their highest value, occurring when the radiometers are completely facing the blackbody radiator aperture. Theoretically, the field of view of the WA radiometers is a hemisphere. This assumption is true if the inner surfaces of the radiometers are extremely polished such that all the radiation going inside the cavity reflects onto the thermopile. However, according to the test results, in reality, the optimum field of view of the radiometers is a cone with the angle of 120°. Additionally, one can assume that the cone is a right circular cone due to the symmetry of the angular response of the radiometers. Based on the angular response calibration, the field of view of the radiometers was determined. Figure 29 shows the view fields in the OFC. More experiments were conducted to explore the effect of water flow rate of the radiometers on the response values. The results showed that the water flow rate does not have a considerable effect on the radiometers' performance. However, the flow rate should be set to a minimum value that can keep the ellipsoidal cavity around room temperature. 3.9 Burner Design 3.9.1 Design of Double Concentric Burner Many studies have been performed on double concentric burners [31-34]. A typical double concentric burner contains two streams with the coal carrier stream usually situated in the center of the jet. A double concentric burner consisting of two streams 33 named: 1) Primary stream; 2) Secondary stream, was employed in this study. The primary stream is located in the center of the burner. A carrier gas transports the coal into the chamber from the primary stream. The carrier gas in this study is either CO2 or a mixture of O2 and CO2 . The secondary stream contains a mixture of O2 and CO2 and flows around the primary jet. Achieving a stable flame in oxy-combustion is a crucial issue. Coal particles exiting the burner require to be mixed with O2 in the secondary stream. Meanwhile, it is important for coal particles to have enough residence time in order to reach high temperatures for devolatilization. Thus, it is important to note that in the design of burners, the momentum of the primary stream should be less than the secondary stream. Additionally, a secondary stream with a higher momentum will create recirculation around the jet, assisting the mixing and stability of the flame. In another study, Villermaux and Rehab et al. [53] studied water jets and showed that the potential core length is proportional to the velocity ratio. Also, they recommended that the effects of variable density jets could be accounted for by substituting velocity ratio with the square root of the momentum flux ratio. Work by Favre-Marient and Schetti et al. [54] showed that the core length is proportional to the ratio of the square root of pU2 for variable density nonreacting jets up to a momentum flux ratio of approximately 50 where recirculation in the inner jet stream starts. The ratio of I- for the gases in this study is nearly one; therefore, aJ p 2 it was decided to only focus on the velocity ratio instead of the momentum ratio. In addition, in this scale, it is not possible to conduct the experiments by keeping the momentum ratios constant. In order to minimize the aerodynamic impacts as one of 34 the strong factors of mixing to be able to explore the mixing phenomena, it was decided to maintain the velocity ratios constant. The sketch of the double concentric burner employed in this research is presented in Figure 30. The size of the pipes is chosen in such a way that the velocity ratio of the secondary stream to the primary stream is nearly 2.5. This type of burner was the only type applied in the previous study performed by Zhang in which the focus was mostly on the effect of the mole fraction of oxygen in the primary and secondary streams on flame stability. The results and details of this study are provided elsewhere [49]. 3.9.2 Triple Co-centric Annulus Burner Configuration A: Burners with Oxygen Stream in Middle Annular Stream In order to investigate the effects of directed pure oxygen injection, a pure oxygen stream was dedicated to the previous type of burner. This stream might contain a specific fraction or 100% of all the oxygen required for combustion for the combustion operating condition. The primary stream is a transport medium carrying pulverized coal particles in the chamber and, for the experiments in this design, contained only CO2 , with no oxygen. Directed oxygen and transport streams were maintained at room temperature. The secondary stream is a mixture of O2 and CO2 , and plays a significant role in moderating the combustion temperature as well as entrainment of the flame jet. In order to explore the effects of directed injection of O2, two possible important impacts of O2 have been investigated. The first parameter is the fraction of the total amount of oxygen that could be situated either in the pure oxygen stream or in the secondary stream. The orientation or configuration of the O2 stream can have significant effects on combustion, such as flame 35 stability as well as heat transfer and NOx formation. O2 stream configuration was the second aspect to be explored. It is inevitable that the aerodynamic conditions of a flame jet has big impacts on the flame property; therefore, it was decided to maintain the velocity ratios of all the streams constant in all the operating conditions by changing the streams' pipe diameter. Several burners were built and employed to allow the stream velocities to remain constant. The typical schematic of these burners is shown in Figure 31. As shown in Table 1, as case number increases, O2 is added from the secondary stream to the pure oxygen stream while the overall O2 is maintained constant. It is important to note that in burner number 6, all of the oxygen required is being introduced in the pure directed oxygen stream. The primary stream (Coal + CO2) is located in the center of the burner. The annulus around the primary stream is the directed oxygen stream, and the furthest outer annulus contains the secondary stream. The velocity of the pure oxygen stream and the primary stream are kept approximately equal in order to delay mixing, leading to potentially longer flames with a more uniform heat distribution. The secondary stream velocity is 2.5 times higher than that of the other two streams, creating external recirculation in the upper chamber and maintaining jet entrainment (IFRF Type-0 flame). Multiple burners with various appropriate dimensions allowed the stream velocities to be maintained constant such that aerodynamic mixing effects were unchanged as much as possible, as shown in Table 1. An image of one of the configuration A burners is presented in Figure 32. 36 3.9.3 Triple Co-centric Annulus Burner Configuration B: Burners with Oxygen Stream in Centrally Located Pipe For Configuration B, the burners are also triple concentric burners but now with a pure oxygen stream located in the center pipe. The primary stream that carries the coal particles is located in the inner (middle) annulus. The primary stream contains only CO2 and coal. The outer annulus is the secondary stream. The secondary stream carries O2 and CO2 to the combustion chamber. As for Configuration A, several cases cover different fractions of total oxygen in the oxygen stream located in the center from 0% to 100%. In the very last case, all the oxygen is once again in one segregated stream (now at the center). The typical schematic of Configuration B burners is provided in Figure 33. The velocity ratios of the streams are similar to Configuration A to minimize the aerodynamic differences as shown Table 2. Keeping the velocities similar for all the cases allows the experimental data to be comparable, especially the results obtained from Configurations A and B. An image of Configuration B burners is presented in Figure 34. 3.10 Coal Sample and Analysis Two types of coal were used in this research: Utah Skyline Bituminous and Illinois #6 Bituminous. Illinois #6 was applied in all the studies of this research; however, in order to investigate the effect of coal composition, Utah Skyline Bituminous was utilized in this research as well. The ultimate, proximate, and ash analyses of the coals are provided in Table 3, Table 4, and Table 5. It is important to note that these data are reported based on As Received (AR). Both coals were pulverized and prepared for steady feeding to the combustor. Since the size of the coal particle is one of the indices of coal ignition, both coals were prepared with similar particle size distribution. The particle 37 size distribution was calculated and is summarized in Figure 36. According to the data shown in Figure 36, the mass average diameters of both coals were calculated to be 62 |im and 68.5 |im, respectively, for the Utah Skyline and Illinois #6 coal particles. 3.11 Gas Analyzers In order to collect samples for the gas analyzers, a flue gas probe is located at the end of the convective zone of the OFC. This probe contains a baghouse that allows for the capture of ash particles in the first step. The gas goes through a modified refrigerator with a vacuum pump that condenses the moisture of the flue gas. Moisture can cause damage to analyzers. The gas once again goes through three more layers of filters and another condenser in order to be prepared for gas analyzers. Six types of gas analyzers were employed in this study to permit the measurement of oxygen, carbon dioxide, NOx, and SOx during the combustion process. Figure 37 shows the array of analyzers in the control room of the University of Utah combustion lab. The analyzers applied were the following: • Yokogawa - Zirconia oxygen analyzer ZA8 • Horiba- Paramagnetic oxygen analyzer • California Analytical Instrument - Infrared CO and CO2 gas analyzer: ZRH • California Analytical Instrument - Infrared CO and CO2 gas analyzer: ZRE • Thermal Environmental Instrument - Chemiluminescense NOx analyzer • Horiba - Chemiluminescense NOx analyzer CLA-510SS It is important to note that the lack of filters in the line can cause serious damages to the analyzers. In the same manner, the moisture needs to be removed from the gas sample for the safety of the analyzers. 38 As mentioned before, the OFC at the University of Utah is equipped with a recycled flue gas system; however, required O2 and CO2 in this study were provided from tanks. The opportunity to have O2 and CO2 allowed for better control of the gas flow rates, leading to more accurate measurements. O2 and CO2 gas supply was delivered from tanks donated by Praxair to the University of Utah combustion research lab. The O2 tank has a capacity of 6000 gallons, and the CO2 tank can store up to 400 gallons of carbon dioxide. The O2 tank contains liquid oxygen, and a vaporizer is equipped in the line to convert the liquid oxygen to its gaseous state. Additionally, it is important to note that the O2 tank was installed professionally by Praxair, and the handling of oxygen to the furnace must follow the safety requirements. Pictures of the CO2 and O2 tanks are presented in Figure 38 and Figure 39. 3.13 U-Shaped Tube Water Cooling Heat Exchanger It is important to note that u-shaped heat exchanger tubes were designed only to be applied in the experiment of the reduction of CO2 (minimization of recycled flue gas). As discussed, the OFC is a combustor with the capacity of 100 kW. U-shaped heat exchanger tubes were designed with the aim of performing the experiment at higher coal feeding rates. However, in order to be able to have a correct comparison between data sets, the decision was to conduct the test at the same heating rates as before. Therefore, the heat exchanger tube was never employed in this study. However, for higher heating rates, especially to study the effects of CO2 reduction, lowering the amount of CO2 will increases the temperature of combustion such that it could damage the refractory and the 39 3.12 O2 and CO2 Delivery insulation of the furnace. Therefore, it is recommended to employ a cooling heat exchanger to prevent any possible damage under these conditions. These tubes also simulate a furnace similar to industrial boilers. There are several types of heat exchanger, such as water cooling jacketed heat exchangers, that could be installed either on the furnace wall facing flame or underneath the insulation. However, In order to have a heat exchanger that does not block the windows of the OFC, and compatible with the current design of the furnace, it was decided to choose U-shape heat exchangers. In this work, the heat exchanger consisted of 4 U-shape stainless steel pipes being hung from the top plate of the burner zone. The size of the pipes used in this heat exchanger is %", Sch 40. New plumbing was required to supply water to these heat exchangers. The new plumbing system transports the required water stream from the cooling tower to the top of the furnace to provide water for U-shape heat exchangers. The manifold shown in Figure 40 divides the water pipe into four different streams at the end of each pipe. Each water stream enters into one heat exchanger. Also, four flow meters are provided in the line to measure the flow rate of water. The other manifold on the left-hand side collects the water from the heat exchangers. This water stream flows back to the cooling tower to be cooled. As shown in Figure 41, it was decided to employ hoses to connect the heat exchangers to the manifolds. Also, four 6" long K-Type thermocouples were installed to measure the exit water temperature from each heat exchanger. By knowing the temperature difference and the flow rate, we were able to have a better understanding of the heat transfer to the heat exchangers. The Solidworks drawings and an image of the heat exchanger are provided in Figure 42 and Figure 43. The size annotations in the drawing are reported in inches. 40 41 Figure 8. Oxy-Fuel Combustor (OFC) at the University of Utah 42 Figure 9. Recycle Flue Gas system of OFC at University of Utah 43 Figure 10. OFC at the University of Utah 44 Figure 11. Flow diagram of OFC 45 Figure 12. O2 and CO2 lines equipped with solenoid valve and mass flow controller • y1 ^ j 7.<r i \9 1 > \ \ 17.75" f t » 1 i 17.75" V t 2.0*' Figure 13. Drawing of windows 46 Figure 14. Heat exchanger of convective zone of OFC Figure 15. Sketch of gas heater electric element 47 Figure 17. Schematic of OFC 48 Figure 18. Schematic of location of radiometers in OFC. Figure 19. Schematic of principle of wide angle radiometer 49 Figure 21. Pictures of narrow angle radiometer 50 Figure 22. Schematic of narrow angle radiometer Figure 23. Blackbody radiator 51 250000 200000 CJ J= 150000 5 100000 50000 0 Wide Angle Radiometer Calibration y = 735082x- 22793 R2 = 0.9988 0.1 0.2 mV 0 0.3 0.4 Figure 24. Wide angle radiometer calibration on 05-09-2011 Figure 25. Narrow angle radiometer calibration on 05-09-2011 52 Figure 26. Angular response of Radiometer #1 (top port) Radiometer #2 0° Figure 27. Angular response of Radiometer #2 (middle port) Radiometer #3 0° Figure 28. Angular response of Radiometer #3 (bottom port) 53 Figure 29. Field of view of wide angle radiometers in OFC 54 Figure 30. Sketch of double coaxial burner 55 Figure 31. Schematic of Configuration A burners Figure 32. Configuration A Burner Figure 33. Schematic of Configuration B burners 56 57 Figure 35. Configuration B burner (,>LmnLiLy (%) 58 0 20 +3 60 SO 100 120 14-0 ISO 180 200 Coal particle size (jim) Coal particle size ^ m) Figure 36. Particle size distribution of Utah Skyline and Illinois #6 coals Figure 37. Gas analyzers at the University of Utah combustion research lab 59 Figure 38. Carbon dioxide tank Figure 39. Oxygen tank 60 Manifold of Exiting Water Streams i i ' ? R < i - Water Flow Meter Manifold of Entering Water Streams Figure 40. Manifolds equipped with flow meters and valves Figure 41. Position of thermocouples on top plate 61 Dadmehr Rezaei OFC Heat Exchanger - ' Figure 42. Schematic of cooling heat exchanger Figure 43. U-shape water cooling heat exchanger 62 Table 1. Flame aerodynamic conditions for burners with inner annular oxygen stream (Configuration A) Case F02 inner 0 2 annulus Prim. Vel. (m/s) 0 2 stream Vel. (m/s) Second. Vel. (m/s) Prim. As (cm2) 02 Stream As (cm2) Second. As (cm2) 1 0.0% 6.37 0.0 14.9 1.9604 6.0744 2 23.0% 6.37 6.10 15.8 1.9604 1.0653 5.0665 3 55.0% 6.34 6.40 15.8 1.9604 2.3788 3.9442 4 75.0% 6.38 6.10 14.6 1.9478 3.2475 3.9515 5 85.0% 6.38 6.70 14.6 1.9478 3.5059 3.6361 6 100.0% 6.38 6.50 15.0 1.9478 4.2651 3.0994 Table 2. Flame aerodynamic conditions for burners with central oxygen stream (Configuration B) Case F02 center 0 2 stream Prim. Vel. (m/s) 0 2 stream Vel. (m/s) Second. Vel. (m/s) Prim. As (cm2) 02 Stream As (cm2) w A CSD (o o 1 0.0% 6.37 0.0 14.9 1.9604 6.0744 2 21.0% 6.40 6.28 16.43 1.9801 0.9066 4.9498 3 44.0% 6.32 6.22 16.27 2.0005 1.9604 4.3723 4 64.0% 6.00 6.26 15.8 1.9883 2.9267 3.9442 5 80.0% 6.26 6.27 16.21 1.9883 3.5244 3.4077 6 100.0% 6.30 6.26 16.17 1.9883 4.3825 2.8689 Table 3. Coals proximate analysis Proximate Analysis (As Received Basis) Moisture Ash Volatile Matter Fixed carbon Coal wt% wt% wt% wt% Utah Skyline 3.10 ± 0.07 10.27 ± 1.44 38.70 ± 0.10 47.91 ±1.47 Illinois #6 9.65 7.99 36.78 45.58 63 Table 4. Coals ultimate analysis Ultimate Analysis (As Received Basis) C H 0 N S Coal wt% wt% wt% wt% wt% Utah Skyline 68.44 ± 4.85 ± 0.56 13.14 ± 1.32 ± 0.09 0.45 ± 0.07 Illinois #6 64.67 4.51 8.07 1.12 3.98 Table 5. Coals elemental analysis Elemental Analysis Al Ca Fe Mg Mn P K Si Na S Ti Element Al20 3 Cao Fe20s M g0 M n0 oin2 P K20 S i0 2 Na20 S 0 3 T i02 Utah 13.30 11.57 5.06 2.88 0.05 0.30 1.49 52.37 1.44 2.44 0.65 Ill #6 17.66 1.87 14.57 0.98 0.02 0.11 2.26 49.28 1.51 2.22 0.85 4. METHODOLOGY OF MEASURMENT OF STAND-OFF AND LENGTH OF FLAME In this research, the flame stability is determined based on the flame stand-off distance concept. Flame stand-off distance is defined as the distance between the tip of the burner and the visible part of the flame. The methodology of defining stand-off distance and flame stability using image processing was developed by Zhang et al. at the University of Utah [49]. This method was applied for the purpose of this study to measure the flame length. In order to measure the stand-off distance, a high-speed camera was used to capture the flame images. This camera is an EPIX CM0S camera SV5C10 with XCAP software, providing a friendly interface. During the test, the lens was adjusted such that the camera could capture the entire length of the windows. This setup allowed for measuring both flame length and stand-off distance simultaneously for each image. It is obvious that because of the turbulent nature of co-axial flames, the presence of fluctuations was witnessed. However, the purpose of this technique was not to capture the flame turbulence fluctuations, but to find the flame envelope leading to the measurement of the flame length and stand-off distance. The camera operating conditions chosen for this set of experiments are an 8.3 millisecond (ms) exposure time and 30 frames per second (fps). For every combustion condition, 6000 images were taken. For this exposure time, it was discovered that there are stand-off distance and 65 flame length fluctuations that need to be considered when using this method. It was suggested to include all the fluctuations in the flame images by the Probability Density Function (PDF). In order to define the flame envelope, it is required to define a luminosity intensity threshold. The first step that needs to be considered is that the goal is not to capture the details of the flame structure. However, the purpose is to understand what the envelope of the flame is and where it exists. Therefore, it was decided to set up the camera such that smeared images can be produced. This methodology is applicable for smeared images. As a summary to elucidate the methodology for the flame length and stand-off distance measurement, the following procedures were applied: 1. Convert the original images to gray scale 2. Use the Sobel operator for edge detection. The Sobel operator finds the zones where the intensity gradient of the luminosity has the highest value. The Sobel operator of MatLab has been applied to evaluate the intensity gradient of each pixel of the images based on the brightness and darkness of the pixels. The maximum gradients can define that flame envelope. The intensity values at the maximum gradients were calculated, and their average was determined as the threshold for the whole image. It is important to note that every image has its own threshold. 3. Convert the images to black and white images using the threshold obtained from the previous step. If the luminosity intensity of a pixel has a value smaller than the threshold, the pixel will be assumed to be a black pixel, and if the pixel 66 luminosity intensity is more than the threshold, that pixel is considered to be a white pixel. 4. Measure the stand-off distance, which is the distance from the tip of the burner to where the flame starts. The point that the flame starts is defined as explained above, and it is where the first white pixel is calculated. 5. Calculate the flame length in a similar manner. The flame length is defined as the distance from where the flame starts to where the flame ends. 6. Measures the PDF of the stand-off distance fraction of occurrences (out of 6000 frames) of the "ignition zone" lying between two specified distances. Figure 45 shows examples of flame stand-off distance, and Figure 46 provides flame length reported in PDF form. Flame stability provides results that can be applied in understanding flame ignition and near burner zone mixing. In addition, these data permit the exploration of the effects of different important parameters of combustion such as 02 concentration, wall temperature, preheating temperature, and flame aerodynamics changes on the stability and ignition of coal particles. Furthermore, data obtained from flame length elucidate mixing in co-axial turbulent diffusion flames. In addition, flame length provides information regarding the heat distribution of the flame. The results of these experiments will be contributed to the validation, with uncertainty quantification, of coal jet ignition submodels that can be used in simulations of practical axial coal flames. I'mlMhi lily Itensity ( I /cm ) 67 Figure 44. Procedure of optical methodology to measure flame stand-off distance Standoff Distance (cm) Figure 45. Flame stand-off distance in PDF form Probability Density (1/cm) 68 Longer than Window Figure 46. Flame length distance in PDF form 5. THE EFFECT OF COAL COMPOSITION ON IGNITION AND FLAME STABILITY IN CO-AXIAL TURBULENT DIFFUSION FLAMES Dadmehr Rezaei*, Yuegui Zhou, Jingwei Zhang, Kerry E. Kelly, Eric G. Eddings, Ronald J. Pugmire, Mark S. Solum, Jost O.L. Wendt Corresponding author. Institute for Clean and Secure Energy, Department of Chemical Engineering, University of Utah, 155 South 1452 East, Room 350, Salt Lake City, UT, 84112, USA 5.1 Abstract Past research at the University of Utah on flame stability and stand-off distance was performed for one specific coal, namely a Utah Bituminous Coal. The experiments were carried out in a 100kW pulverized coal test rig with a co-axial turbulent diffusion burner. The purpose of the research described in this paper is to extend the previous work and to explore how coal composition changes affect the following dependencies that control flame stand-off distance and flame ignition, namely (1) the effect of partial pressure of oxygen (PO2) in the primary stream with differing preheat temperatures in the secondary stream and (2) the effect of PO2 in the secondary stream with zero O2 in the primary stream. The results of this new study were designed to extend previously obtained knowledge on the effects of secondary preheat temperature, turbulent mixing, and P0 2 in various streams, from one single coal to other coals of differing compositions. This paper, therefore, explores the effects of coal composition on ignition in oxy-coal, coaxial, turbulent diffusion flames. In this research, the stability and stand-off distance of the flame were studied for the following two types of coal: Utah Skyline Bituminous, Illinois #6 Bituminous. To this end we investigated: 1) the effect of P02 in the primary stream, 2) the effect of P0 2 in the secondary stream, and 3) the effect of preheat temperature in the secondary stream, on flame stand-off distance, using the same photo­imaging methodology described elsewhere. The results of the ignition and flame stability analysis for these two coals under oxy-firing conditions are compared, and the effects of coal composition are elucidated. At 420°C preheat temperature, Illinois coal showed a more stable flame. Flame stability of Illinois coal did not change at 520°C preheat temperature; however, Utah Skyline flame stability increased significantly. Data obtained from TGA analysis and solid state 13C NMR data provide further insight regarding the structural variation in these two coals as well as their pyrolysis behavior. These data justified the difference in the flame behaviors of two coals with similar coal composition in a pilot-scale oxy-fuel combustor. 5.2 Introduction The increase of greenhouse gases, and particularly carbon dioxide, for energy production has resulted in new technologies with lower emissions of NOx, and SOx. These techniques are capable of complying with carbon dioxide capture and sequestration. Oxy-Fuel Combustion technology has been suggested as the most promising strategy for the conventional coal power plants to generate electric power [4]. In oxy-coal combustion, coal burns with pure oxygen instead of air, and the combustion 70 gases are diluted using the recycled flue gas (RFG), which essentially contains C02 and H20. The gas mixture has higher emissivity and subsequently greater heat transfer while at the same time a lower volume. By recycling the flue gas, a gas consisting mainly of C02 and water is produced, ready for C 02 capture and sequestration [1]. 0ne of the difficulties associated with oxy-fuel combustion is delayed ignition. Several studies have shown the effect of the burner aerodynamics and higher momentum flux of primary stream on the flame stability and delayed ignition [13,55]. 0ther research groups have investigated the impact of the C02 atmosphere on coal ignition. Yu Qiao and his colleagues claimed that thermal conductivity of the gas atmosphere surrounding the particle could significantly affect the observed particle ignition [10]. Another study showed that the use of C 02 as diluents will retard devolatilization due to the lower mass diffusion of fuel volatiles in C02 compared to a nitrogen atmosphere [4]. The fuel ratio, defined as fixed carbon content / volatile matter content, is believed to provide an index to predict the combustibility, but it does not always anticipate the real ignition behavior [42,56]. The influence of the maceral composition of the coal on the ignition and flame stability has been pursued by Su who suggested a criterion, the maceral index, which predicts burnout behavior of the coal and blends by including the effects of maceral composition such as liptinite, vitrinite, and inertinite [57]. However, these studies were conducted in an air atmosphere and, hence, the behavior change of coal ignition in an 02/C02 is not yet well understood. Previous work by Zhang [49] on Utah Skyline Bituminous in the 100 kW vertical oxy-fuel combustor at the University of Utah explained the effects of P 02 in the primary stream on ignition and flame stability at different preheat temperature in the secondary 71 stream. In addition, this work provided information in connection with ignition and flame stand-off distance by increasing PO2 in the secondary stream without O2 in the primary stream. However, the influence of coal composition, which is one of the key criteria of combustibility in this set of experiments, is not well known. The present study was conducted to compare the ignition behavior, under similar oxy-fuel operating conditions, of two coals with remarkably similar rank and elemental analysis composition in a 100 kW pilot-scale combustor with a co-axial turbulent diffusion burner operated in a O2/CO2 environment. The results of this study will contribute to the validation, with uncertainty quantification, of coal jet ignition submodels for simulations of axial coal flames. 5.3 Coal Selection and Sample Preparation Two coals of similar rank and elemental composition, except for sulfur content (Utah Skyline HVB and Illinois #6), were chosen in this study. The ultimate, proximate, and ash analyses of the coals are provided in Table 3 and Table 4. Both coals were pulverized and prepared for a steady feeding to the combustor. Since the size of the coal particles is one of the indices of coal ignition, both coals were prepared in a manner to have a similar particle size distribution. The maximum size of the coal particles is 150 ^m. The particle size distribution has been calculated and is summarized in Figure 36. According to data shown in Figure 36, the mass average diameters were calculated to be 62 ^m and 68.5 ^m for Utah Skyline and Illinois #6 coal particles, respectively. Hence, this study could be focused only on the differing effects of coal composition/structure. 72 The Oxy-Fuel Combustor (OFC) consists of three sections, namely 1) Burner zone or the main chamber; 2) Radiant zone, which is located in the bottom of the main chamber; 3) Convective zone, which benefits from the heat exchangers to resemble the industrial furnace conditions (see Figure 47). The dimension of the burner zone is 0.61m I.D, 0.91m O.D. and, 1.22m as the height of the burner zone. The chamber is insulated by 76 mm thick Fiberboards that tolerate high temperatures up to 1700 K. The temperature of the furnace can be monitored by three high-temperature-resistant K-Type thermocouples that are located along the height of the chamber. Also, to have an optical access to the flame regarding optical diagnostic, four quartz windows have been provided on the quadrants of the cylindrical chamber. The chamber of the OFC is equipped with 24 ceramic electric heaters that have been arranged in three rows. The 840 watt heaters permit control of the wall temperature of the chamber in an accurate way. Wall temperature has an important role on ignitibility of coal. Therefore, in this study, the wall temperature for all studies was kept at 1850 °F. 5.5 Gas Analyzers The flue gas probe is located at the end of convection zone of the OFC. This probe is used to monitor the exhaust gas composition during the combustion process using oxygen, carbon dioxide, NOx, and SOx analyzers. The furnace is controlled and monitored by the OPTO22 control system. All the data from the furnace and analyzers have been plotted, and saved on charts during the experiment. This system provided extensive control on the OFC. 73 5.4 Oxy-Coal Combustor The schematic of the burner drawn in Solidworks is shown in Figure 48. The burner consists of: 1) Primary stream; 2) Secondary stream. The primary line is located in the center of the burner. A carrier gas transports the coal into the chamber from the primary stream. The carrier gas in this study is either CO2 or a mixture of O2 and CO2. The secondary stream contains a mixture of O2 and CO2 and flows around the primary jet. A heat exchanger is installed in the secondary stream, which provides the ability to preheat the secondary flow up to 700 K. The flow and temperature of both streams are accurately automated and under control using the OPTO22 control system. Having a steady coal feeding system is one of the most effective indices of flame stability. In this study, it was decided to use a twin-screw coal feeder with a solid conveying eductor. In the eductor system, the primary stream gases get mixed with coal fed by the screw feeder. The eductor provides a steady stream of pulverized coal feeding through the burner. 5.7 Methodology of Flame Stability Measurement Flame stability is determined based on the flame stand-off distance concept. Flame stand-off distance is defined as the distance between the tip of the burner and visible part of the flame. In order to measure stand-off distance, a high-speed camera was used to capture the flame images. The methodology for quantification of flame stability was developed in the University of Utah combustion laboratory [47]. Due to the turbulent nature of the flame, existence of fluctuations in the flame properties is obvious. The camera operating conditions chosen in this set of experiments are 8.3 ms exposure time and 30 fps. For every combustion condition, 6000 images were taken. Since flame 74 5.6 Burner and Feeder attachment is defined by human eye vision, the operating conditions of the camera are set to simulate the detection capabilities of the human eye instead of capturing all the turbulent fluctuation. For this exposure time, all stand-off distance fluctuations are required to be considered. The method suggests including all the fluctuations in the flame images by the Probability Density Function (PDF). 5.8 Combustion Operating Conditions The experiment operating conditions and the burner jet aerodynamic parameters are shown in Table 6 and Table 7. In order to have a robust comparison to investigate the effect of coal composition on ignition and flame stability, the operating conditions were kept constant for each type of coal. However, to maintain the total Stoichiometric Ratio (SR) of the combustion process, it was necessary to change the coal feeding rate based on the SR (see Table 8). In this study, there are three sets of experiments. The experiments of set A and B were designed to research the effect of oxygen in the primary stream. The second objective of this experiment was to investigate the role of turbulent diffusion mixing in the burner jet by increasing the oxygen in the primary stream. The overall oxygen concentration is kept at 40%; however, in each case, the amount of oxygen is deducted from the secondary stream, and added to the primary stream. The flow rates in both the experiments of set A and B are identical. In order to evaluate the effect of secondary stream preheat temperature on the combustibility of the coals, two temperatures were determined. For the experiments of set A, the secondary stream preheat temperature was kept at 489 K, and for the experiments of set B, the temperature was maintained at 544 K. The experiments of set C were devised to look at the importance of overall 75 oxygen concentration in the furnace. In this case, the primary stream oxygen concentration is zero; however, the amount of oxygen in the secondary stream increases. The overall oxygen increases until the attached flame is obtained. The secondary stream temperature in the experiments of set C was kept at 489 K. 5.9 TGA and NMR Experiments Thermo-gravimetric Analysis (TGA) data were obtained in both oxygen and nitrogen environments. The temperature ramp of the TGA was 20 K/min. The same pulverized coal that was used in combustion was applied for the TGA analysis. As it was mentioned, the diameters of the coal particles were 62 |im and 68.5 |im, respectively, for the Utah Skyline and Illinois #6 coals. At the beginning of each experiment, 25 mg of the coal sample was loaded into the pan in the furnace of the TGA. The amount of the purge gas was 100 ml/min for both N2 and Air environments [44]. Each sample was heated from 30°C to 1000°C. The solid-state 13C NMR experiments were conducted as previously described [58]. The coal structural data are given in Table 9, which also contains the lattice structural parameters of the two coals as defined in reference [59]. 5.10 Results 5.10.1 Combustion of Utah Skyline and Illinois #6 Coals at 489 K Preheat Secondary Stream Temperature and Overall 40% Oxygen Concentration The results of the experiments in set A are shown in the PDF plots for Illinois #6 and Utah Skyline coals. It is noted that at these conditions, attached flame was obtained from Illinois #6 coal at PO2=0.144 in the primary stream. However, for Utah Skyline coal, 76 the attached flame was obtained at P02=0.207 in the primary stream. Results are presented in Figure 49. 5.10.2 Combustion of Utah Skyline and Illinois #6 coals at 544 K Preheat Secondary Stream Temperature and 0verall 40% 0xygen Concentration This set of experiments was selected to evaluate the effect of secondary stream temperature on the combustibility of the coals. The temperature of the secondary stream was kept at 544 K during the experiment. The results of this set are provided in Figure 50. According to the results for Utah Skyline, a semi-attached flame was obtained at P02=0.054 and it was fully attached at P02=0.099 in the primary stream. However, Illinois #6 coal had an attached flame at P02=0.144. Also, it is noticeable that the results of the Illinois #6 coal at 544 K preheat temperature did not change considerably compared to the results at 489 K of the secondary stream preheat temperature. 5.10.3 Combustion of Utah Skyline and Illinois #6 Coals at 489 K Secondary Stream Temperature with Increasing of 0verall 0xygen Concentration in the Secondary Stream In this set of experiments, the secondary stream preheat temperature was kept at 489 K. However, to see the effect of overall oxygen on combustion, the amount of oxygen in the secondary stream was increased. It is important to note that in this test, attempts were made to keep the primary stream's aerodynamics constant. This operating condition only permits the study of the influence of overall oxygen concentration on 77 frame stability. Therefore, the amount of oxygen in the primary stream was zero. The results of this test are shown in Figure 51 in PDF form. 5.11 TGA Analysis TGA provides information regarding the devolatilization of both coals and possible relationships to the fundamental structural features of these coals. TA Q500 TGA was applied for this test. Nitrogen and air were chosen as the environmental gases for this experiment. In the nitrogen environment, the weight loss of the coal due to drying and devolatilization in an inert gas is presented in Figure 52. In Figure 53, devolatilization in an air environment presents a more complex picture of the differences in the pyrolysis/combustion behavior of the two coals, indicating that the fundamental structural differences of two coals of a similar rank can be manifested in their reaction/oxidation processes. The temperature ramp for both coals was 20 °C/min and the samples were heated to 1000 °C. 5.12 Discussion According to the results in Figure 49, by increasing the oxygen in the primary stream from the secondary stream, flame stability increases. Better flame stability implies better ignition. The first criterion of ignition is the amount of volatile matter; however, it is believed the more correct way to determine the combustibility is the Fuel Ratio index (FR) which is the ratio of fixed carbon / volatile matter content [1,13,55]. A general trend is observed that the higher fuel ratio corresponds with less carbon burnout. However, this reasoning is not always reliable [42]. The PDF data for set A agree with the fuel ratio indices; however, it is not consistent with the data from the experiments in set B. In 78 addition, the difference between the fuel ratios is not significant enough to allow for sole reliance on this ignition index. Therefore, in order to justify the results, it was decided to use TGA tests. The TGA plots and data are presented in Figure 52 and Figure 53. The fuel ratios of Utah Skyline and Illinois #6 are provided in Table 10. The TGA plots in Figure 52 provide the weight loss of both Utah Skyline and Illinois #6 coals in a nitrogen environment. The first peak in each coal trace indicates the moisture lost, and the subsequent peaks manifest the devolatilization processes of both coals. The pyrolysis pattern of the Illinois #6 coal in a nitrogen environment begins at a lower temperature than that of the Skyline coal but the differential weight loss curves are quite similar. The TGA experiment was also carried out in an air environment and the results are shown in Figure 53. In the air environment, significant differences are noted in the differential weight loss of the two coals. The first peak in both traces (~79°C) corresponds with moisture loss. Four additional inflection points are identified (251, 350, 452, and 610°C) in the differential weight loss of Illinois #6 while only 3 inflection points (296, 450 and 640°C) are evident in the Utah Skyline coal. The differences noted in the pyrolysis behavior can be traced to the fundamental differences in the chemical structure of the two coals, as reported earlier [60-63]. Detailed studies of combustion devolatilization data have been compared on Illinois #6, Montana Rosebud sub-bituminous, and North Dakota Beulah Zap Lignite [61], while comparisons of matched tar-char pairs in rapid pyrolysis of Illinois #6 and Beulah Zap Lignite are described in reference [62]. The details of the differences in the chemical structures are quite similar in a 13C NMR standard cross­polarization magic angle spinning experiment, as one can observe in Figure 54. NMR 79 Spectra acquired by means of dipolar dephasing begin to illustrate major differences in the structure of these two coals (see Figure 55). The details of the structural parameters (referred to as lattice parameters) of the two coals can be obtained as described in reference [44] and these data are presented in Table 9. For purposes of discussion, the critical lattice parameters are the average number of aromatic carbons in the aromatic clusters (C), the coordination number or number of attachments or cross links per cluster (s+1) that consist of either side chains (S.C.), bridges and loops (B.L.), or bi-aryl linkeages to adjacent clusters in the lattice. The molecular weight of an averaged cluster (M.W.) and the average mass of the aliphatic attachments to the aromatic clusters or half of the mass of the cross links between aromatic clusters is represented by Ms. The number of cross links (s+1) per cluster in Illinois #6 suggest that the pyrolysis/oxidation processes would begin liberating light gases prior to the release of comparable amounts of aliphatic-derived components in Utah Skyline coal. The release of aliphatic chain ends is more energetically favored than the breakage of fairly stable aliphatic bridges and loops between aliphatic structure [63]. This early release is noted at approximately 250°C for Illinois #6 while the first major release of light gases is at approximately 300°C in Utah Skyline. Once the lattice begins to break down, a broad range of volatile material is released, including aromatic as well as aliphatic components, and this process appears to start at approximately 350°C in the Illinois #6 and reach a maximum differential weight loss near 450°C. A similar secondary release is not present in the Utah Skyline coal but at 450°C, one would expect the onset of lattice break up for both coals and oxidation of the remaining aromatic structures, which continues from 450°C to approximately 700°C. Hence, the lattice 80 structure data seem to predict the general pyrolysis data in these two coals of similar rank and elemental content. The TGA data are reliable evidences for the combustion behavior in the two types of coal utilized in Experiments A and B. Illinois #6 pyrolysis occurs at lower temperatures than Utah Skyline; therefore, it is seen that in Experiment A, carried out at 489 K secondary stream preheat temperature, flame stability of Illinois #6 develops by increasing the primary stream oxygen concentration. However, this effect is not significant for Utah Skyline coal. Experiment B was performed at 544 K secondary stream preheat temperature. It is notable that only a 55 K increase of temperature in the secondary stream stimulates the devolatilization of Utah Skyline. Comparing Figure 49 and Figure 50, it is shown that Utah coal flame stability increases significantly. However, Illinois #6 is already able to be pyrolyzed at lower temperatures, and increasing the temperature will not have a more considerable influence on the ignitibility of the Illinois #6 coal. As it is shown in Figure 49 and Figure 50, PDF plots of Experiments B-4 and A-4 appear to be quite similar. In Experiment C, the amount of overall oxygen was increased until an attached flame was observed. The amount of oxygen in the secondary stream was increased in each case of Experiment C; however, the concentration of oxygen in the primary stream was kept at zero. Figure 51 shows the influence of increasing overall oxygen is more important for Utah Skyline than Illinois #6. The Utah coal flame has some indication of an attached flame at 42% overall O2 concentration, and it is fully attached at 44% overall concentration of O2. However, there was not any indication of an attached flame for the Illinois #6 coal until the concentration of O2 was increased to 48% and even at 52% of 81 overall O2, a fully attached flame was not witnessed. According to the proximate analysis of both types of coal, the moisture content of Utah Skyline is 3.03% and for Illinois coal is more than three times greater (9.65%). It is important to note that when the overall oxygen concentration increases, the velocity of the burner jet in the secondary stream increases as well; therefore, the residence time decreases. The values of both velocity and residence time for each case of Experiment C has been calculated and tabulated in Table 7. The lack of sufficient residence time for the high moisture content coal retards the rate of both heat transfer and mass transfer for drying and devolatilization of the coal particles. Therefore, the flame stability of Illinois #6 coal lowers compared to Utah Skyline coal, even though O2 concentration was increased. 5.13 Conclusion The studies carried out in the pilot scale oxy-fuel combustor, and TGA analysis on two types of coals of similar rank and elemental composition but different structural composition, reveal th