Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Coal pyrolysis : chemistry and kinetics in a flow reactor McCafferty, Timothy P. 1984

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1984_A7 M32.pdf [ 9.56MB ]
Metadata
JSON: 831-1.0058901.json
JSON-LD: 831-1.0058901-ld.json
RDF/XML (Pretty): 831-1.0058901-rdf.xml
RDF/JSON: 831-1.0058901-rdf.json
Turtle: 831-1.0058901-turtle.txt
N-Triples: 831-1.0058901-rdf-ntriples.txt
Original Record: 831-1.0058901-source.json
Full Text
831-1.0058901-fulltext.txt
Citation
831-1.0058901.ris

Full Text

COAL PYROLYSIS: CHEMISTRY AND KINETICS IN A FLOW REACTOR By TIMOTHY P. McCAFFERTY B.Sc. (Hon.) The University of Western Ontario, 1981 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF CHEMICAL ENGINEERING We accept this thesis as conforming to the required s tandard THE UNIVERSITY OF BRITISH COLUMBIA September, 1984 © Timothy P. McCafferty, 1984 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree a t the U n i v e r s i t y o f B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and study. I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e copying o f t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the head o f my department o r by h i s or her r e p r e s e n t a t i v e s . I t i s understood t h a t copying or p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l gain s h a l l not be allowed without my w r i t t e n p e r m i s s i o n . Department of CKei-wv S ^ ^ e e r l t o ^ The U n i v e r s i t y o f B r i t i s h Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 Date Oc4oW is,i<*e,t< DE-6 (3/81) ABSTRACT Coal conversion technologies offer the potential to supplement and improve the supply of l i q u i d f u e l s . Coal pyrolysis involves the thermal degradation of coal in an i n e r t atmosphere to generate l i q u i d s ( t a r ) , gases and char. The ultimate commercial v i a b i l i t y of such a conversion process must examine a l l of the quantitative (yield) and q u a l i t a t i v e aspects of the products. A theoretical and experimental program was undertaken to study the variables a f f e c t i n g the product y i e l d and s e l e c t i v i t i e s from coal p y r o l y s i s . A review was made of a va r i e t y of pyrolysis reactions of pure compounds possessing the same chemical functional groups as those i d e n t i f i e d i n coal organic matter. A generalized hydrocarbon reaction flow diagram was developed to consolidate the reaction p o s s i b i l i t i e s i n coal pyrolysis and to attempt to derive an & p r i o r i k i n e t i c model. This led to r a t i o n a l i z a t i o n s of observed trends and predictions which guided the experimental work undertaken. Experimentally, a complete lab scale pyrolysis system was designed and constructed. Investigations were directed at the e f f e c t s of temperature, vapour residence time, coal loading, reactor area/volume ra t i o s and oxygen on the product yields from the pyrolysis of Balraer coal from southeastern B r i t i s h Columbia. To assess the value of each product a number of a n a l y t i c a l procedures were developed. Solvent p a r t i t i o n i n g , s o l u b i l i t y , preparative and high performance l i q u i d chromatography and infrared spectroscopy were used on the tars - i i i -obtained. Chars were characterized by thermal gravimetric analysis. Pyrolysis off-gases were analyzed by gas chromatography. Highest tar yields were observed when dilute phase loading, short residence times (< 1 s) and intemediate temperatures (600-800°C) were employed. In a l l cases, however, improvements in liquid total tar yield were entirely attributable to the greater yield of undesirable asphaltenes. The yields of desirable tar components, aromatics and saturates, remained essentially unchanged over the entire temperature, time and loading range. Increasing the reactor area/volume ratio had a detrimental effect on tar yield but a slightly advantageous impact on the selectivity of aromatics. Char quality, as measured by the content of residual volatiles, showed a constant decrease with increasing pyrolysis temperature. In terms of overall product value, total tar yield is not the most reliable indicator of commercial v i a b i l i t y . Rather, some considerations of product quality is required. Based on the combined experimental and theoretical consideration a number of recommendations are made for improved reactor design and future pyrolysis research. - i v -TABLE OF CONTENTS Page A b s t r a c t i i Table of Contents i v L i s t of Tables v i L i s t of Figures v i i L i s t of Photographs x 1. I n t r o d u c t i o n 1 2. Coal P r o p e r t i e s and Coal Chemistry 3 3. P y r o l y t i c Methods 9 4. Trends i n Coal P y r o l y s i s 14 5. P y r o l y s i s Chemistry 17 6. P y r o l y s i s K i n e t i c s 38 7. P y r o l y s i s System Design and Construction 48 8. Reactor Operations 69 9. A n a l y t i c a l Methods..... 73 10. Experimental Results and D i s c u s s i o n 97 a. Coal Feeder Performance 97 b. System Temperature P r o f i l e s 98 c. R e p r o d u c i b i l i t y of Results and A n a l y s i s 100 d. Coal Loading 103 e. Residence Time and Temperature E f f e c t s 110 f. Extended Temperature Range P y r o l y s i s 123 g. Extended Residence Time Range P y r o l y s i s . . . . . . . 130 h. Reactor Comparisons 134 i . Oxygen 142 j . Thermal A n a l y s i s of P y r o l y s i s Chars... 143 k. I n f r a r e d Spectra of Hexane Ins o l u b l e s 146 11. Conclusions and Recommendations 150 - v -Page Nomenclature 153 Bi b l i o g r a p h y 156 Appendices A. Mechanistic P y r o l y s i s Chemistry and K i n e t i c s 162 B. Coal Feeder Development 176 C. T r i v i a l Names and Structures 185 D. Summary of Results f o r P y r o l y s i s Experiments.......... 187 E. Heat Transfer and Coal P a r t i c l e Temperature P r o f i l e s . . 199 F. POSA D i s t r i b u t i o n s i n some Petroleum Fuels 201 - v i -LIST OF TABLES Page 1. D i s t r i b u t i o n of Some Elements i n Coal 4 2. Oxygen D i s t r i b u t i o n i n Coal Organics 6 3. P y r o l y s i s Reactors 11 4. Design C r i t e r i a 49 5. Proximate and Ultimate A n a l y s i s of Balmer Coal 71 6. HPLC Methods f o r Hydrocarbon A n a l y s i s 76 7. HPLC Conditions 82 8. HPLC Retention Times and Detector Response Factors 84 9. S t a t i s t i c a l A n a l y s i s of R e p r o d u c i b i l i t y Runs #43 and #44 102 B-1 Feeder C h a r a c t e r i s t i c s 172 D-l Summary of R e s u l t s . . . . . 183 E - l P a r t i c l e Temperature Rise Times 200 F - l POSA D i s t r i b u t i o n s i n some Petroleum Fuels 201 - v i i -LIST OP FIGURES Page 1. P y r o l y t i c Methods: Operations and Variables 10 2. T y p i c a l Coal Pyrolysis Product Yields 15 3. Reaction Pathways i n Coal Pyroly s i s 37 4. Coal Feeder Body 52 5. Coal Conveying Tube 56 6. Exploded View - Long RTD Reactor 59 7. Exploded View - Short RTD Reactor 60 8. Tar Recovery System 66 9. P y r o l y s i s System Flowsheet 67 10. HPLC Backflush method 80 11. HPLC - Column Performance Comparison 80 12. Coal Tar Chromatogram 85 13. Cross Correlation Results - Spectra of Bulk Fractionation Cuts... 88 14. TGA Temperature Program 92 15. TGA Results: Balmer Coal 92 16. TGA Results: Char #44 from Balmer Coal 92 17. Pyroly s i s Preparation and Analysis Flow Diagram 94 18. Reactor Temperature P r o f i l e s 99 19. E f f e c t of Coal Solids Loading - Tar. Y i e l d 104 20. E f f e c t of Coal Solids Loading - Product D i s t r i b u t i o n 106 21. E f f e c t of Coal Solids Loading - Net Product Yields 108 22. E f f e c t of Coal Solids Loading - Aromatics S e l e c t i v i t y 109 - v i i i -Page 23. Temperature Versus Residence Time Matrix - 0.5 seconds, tar y i e l d I l l 24. Temperature Versus Residence Time Matrix - 1.0 seconds, tar y i e l d 112 25. Temperature Versus Residence Time Ma t r i x - 2.0 seconds, ta r y i e l d 113 26. Temperature Versus Residence Time Matrix - Overlay Composite of Tar Y i e l d s 115 27. Temperature Versus Residence Time Matrix - 0.5 seconds, Product D i s t r i b u t i o n 117 28. Temperature Versus Residence Time Matrix - 1.0 seconds, Product D i s t r i b u t i o n 118 29. Temperature Versus Residence Time Matrix - 2.0 seconds, Product D i s t r i b u t i o n 119 30. Temperature Versus Residence Time Matrix - 0.5 seconds, Net Product Y i e l d s 120 31. Temperature Versus Residence Time Matrix - 1.0 seconds, Net Product Y i e l d s 121 32. Temperature Versus Residence Time Matrix - 2.0 seconds, Net Product Y i e l d s . . . . . 122 33. Temperature Versus Residence Time Matrix - Overlay Composite: Net Product Y i e l d of Asphaltenes 124 34. Temperature Versus Residence Time Matrix - Overlay Composite: Net Product Y i e l d of Aromatics 125 35. Extended Temperature Range P y r o l y s i s - Tar Y i e l d . . 127 36. Extended Temperature Range P y r o l y s i s - Product D i s t r i b u t i o n 128 37. Extended Temperature Range P y r o l y s i s - Net Product Y i e l d s . . . 129 38. Extended Residence Time Range P y r o l y s i s - Tar Y i e l d 131 39. Extended Residence Time Range P y r o l y s i s - Product D i s t r i b u t i o n 132 - i x -Page 40. Extended Residence Time Range Pyrolys i s - Net Product Yiel d s 133 41. Reactor Comparison - Tar Y i e l d 136 42. Reactor Comparison - Product D i s t r i b u t i o n , Short RTD Reactor.... 138 43. Reactor Comparison - Product D i s t r i b u t i o n , Long RTD Reactor 139 44. Reactor Comparison - Net Product Yi e l d s , Short RTD Reactor 140 45. Reactor Comparison - Net Product Y i e l d s , Long RTD Reactor 141 46. Residual V o l a t i l e s vs. Pyroly s i s Tar Y i e l d 144 47. Residual V o l a t i l e s vs. Pyroly s i s Temperature 145 48. Infrared Spectra of Hexane Insolubles - Residence Time E f f e c t 147 49. Infrared Spectra of Hexane Insolubles - Temperature E f f e c t . . 149 A-l K i n e t i c Simulation Reaction Pathways 162a B-1 E f f e c t of Rotor on Feed Rate 174 B-2 Rotor RPM vs. Feed Rate 176 B-3 Feed Curves - Feed Rate vs. Coal Mass 177 B-4 Coal Feeder C a l i b r a t i o n Curves 180 - x -LIST OF PHOTOGRAPHS Page 1. Coal Feeder Body and Feeder Tubes 53 2. Coal Feeder Body - Disassembled 53 3. Coal Feeder Body - Top View... 54 4. Coal Feeder Body - In s t a l l e d 54 5. Coal Conveying Tube 57 6. Pyr o l y s i s Reactors 57 7. Pyrolys i s Furnace 61 8. Total Reactor System: Side View 62 9. Total Reactor System: Front View, Open-Configuration 63 10. Total Reactor System: Closed Configuration 68 - 1 -1. INTRODUCTION It i s well recognized that the very operation of our i n d u s t r i a l society depends on a steady supply of appropriate f u e l s . Yet i t i s somewhat surp r i s i n g that very l i t t l e attention i s given to ensuring their long term supply. The events of the 1970's showed how dangerously dependent our society i s on p a r t i c u l a r fuels supplied l a r g e l y by a very small number of producers. Given that some 90% of the recoverable and useable f u e l i n North America i s i n the form of coal, i t i s appropriate that greater u t i l i z a t i o n of this resource i s being encouraged. Where not useable i n i t s s o l i d form, coal conversion technologies offer the promise of supplementing more conventional supplies of hydrocarbon l i q u i d and gases. In B r i t i s h Columbia, as i n many other places, the d r i v i n g force behind the coal conversion research i s rather easy to comprehend. Currently an estimated 76% of a l l l i q u i d hydrocarbons used i n the province must be brought i n as crude o i l or refined products. On the other hand, an estimated 39,800 m i l l i o n tonnes of coal are held within the province's boundaries (1). A host of coal conversion technologies for recovering l i q u i d s , gaseous, and s o l i d fuels e x i s t on both a developmental and a commercial scale. In terms of the l o c a l economy, an opportunity exists for f i l l i n g the l i q u i d f u e l demand by applying any of a number of the conversion technologies. The work i n this study i s a step towards a greater understanding of coal technology. Although a pa r t i c u l a r conversion method i s - 2 -considered, much of this thesis should be applicable i n other avenues of coal research. One of the more promising coal conversion processes i s p y r o l y s i s , in one of i t s various forms. Pyrolysis may be defined as the chemical decomposition or other chemical change brought about by the action of heat. In the majority of cases this translates to the degradation of organic compounds or materials i n an i n e r t (oxygen-free) atmosphere. The p r i n c i p a l advantage of this means of generating hydrocarbon l i q u i d s from coal i s i t s inherent s i m p l i c i t y , allowing less design complexity, support i n f r a s t r u c t u r e and operational l i m i t a t i o n s than competing ( d i r e c t or i n d i r e c t ) l i q u e f a c t i o n techniques. This i s p a r t i c u l a r l y important when one considers that a stand-alone operation for generating l i q u i d s i s most u n l i k e l y . Integrated f a c i l i t i e s to handle a l l of the s o l i d , gaseous and l i q u i d fuels i s the only r e a l i s t i c a l t e r n a t i v e . This would be possible by f i t t i n g a l i q u e f a c t i o n f a c i l i t y as a front end operation to a t r a d i t i o n a l pipeline or power u t i l i t y . In this regard, f l e x i b i l i t y and low cost are e s s e n t i a l i f one i s to simultaneously exploit the l i q u i d fuels market without jeopardizing the regulated u t i l i t y . In the present study, the nature of p y r o l y t i c products from co a l , i n terms of both quantity (yield) and composition, w i l l be examined. Before presentation of these r e s u l t s , a discussion of the methods and observations of other researchers i n the f i e l d i s i n order. - 3 -i 2. COAL PROPERTIES AND COAL CHEMISTRY Despite a l l the completed work, a thorough understanding of c o a l composition and pr o p e r t i e s i s a long way o f f . This i s understandable when one considers the heterogeneous nature of c o a l . Coal i s p r i m a r i l y a f i x e d carbon matrix w i t h i n which i s included both i n o r g a n i c and organic compounds. Whether t h i s matrix i s i n the form of g r a p h i t i c sheets, an adamantyl network, or extended aromatic lamellae i s a matter of controversy. This carbon framework i s l a r g e l y u nreactive except under g a s i f i c a t i o n or combustion c o n d i t i o n s . Within t h i s carbon s t r u c t u r e are a v a r i e t y of ino r g a n i c and organometallic minerals and metal complexes ( 2 ) , l a r g e l y i n the form of clays [Al2Si20 5(OH) 1+] , s i l i c a (SiO-2) and some s u l f i d e s (FeS2). The d i s t r i b u t i o n of the elements and the qua n t i t y of minerals v a r i e s g r e a t l y w i t h the type and o r i g i n of the c o a l . Representative elemental d i s t r i b u t i o n s are shown i n Table 1. The e f f e c t s of these m a t e r i a l s on the thermal degradation i s poorly understood. Most important from the point of view of generating petroleum s u b s t i t u t e f u e l s i s the organic m a t e r i a l present i n c o a l . Due to the d i f f i c u l t y of removing coal organics from t h e i r n a t u r a l entangled s t a t e . w i t h i n the carbon matrix, i t i s very d i f f i c u l t to e l u c i d a t e t h e i r s t r u c t u r e and composition. Two methods are used. F i r s t l y , many i n v e s t i g a t o r s have looked at the precursor organics l a i d down when c o a l i f i c a t i o n began ( 5 ) . These i n c l u d e polysaccharides, waxes, terpenes, p o l y k e t i d e s , tannins and l i g n i n s . Coal s t r u c t u r e may then be - 4 -Table 1 D i s t r i b u t i o n of Some Elements i n Coal and Peat (3) Coal Peat L i g n i t e Sub- Bituminous A n t h r a c i t e Bituminous Carbon 60 65-75 75 80-87 93-94 Hydrogen 6 5-6 4-5 4-5 3-4 Oxygen 30-45 20-30 10-20 5-15 2-3 S u l f u r + Nitrogen 1-2 1-2 1-3 1-3 1-3 Trace Metals Concentration Range (whole coals) (4) (ppm) • B e r y l l i u m 0.1 — 6.0 Boron 10 - 220 Chromium 5 - 60 Cobalt 1 - 16 Copper 5 - 40 Gal l i u m 0.1 - 6.0 Germanium 1.0 - 35 Lanthanum 0.5 - 30 Molybdenum 0.5 - 15 N i c k e l 5 - 65 Titanium 50 - 1200 Vanadium 5 — 70 Y t t r i u m 2 -30 [others a l s o present: i r o n , phosphorus, calcium, magnesium, potassium, sodium, c h l o r i n e . . .] - 5 -elucidated by monitoring the changes in these functional entities as the progression from peat to coal occurs. The alternative approach is to examine the decomposition products from liquefaction or pyrolysis and attempt to piece the structure back together from the fragments (6). By either of these techniques, one finds that coal organics are largely molecularly disperse aromatics. Alkyl side chains on these base structures are usually short (methyl, ethyl) and few in number (7). Heteroatoms, principally oxygen are contained in carboxylic acids, ethers, phenolics and quinone functional groups (8,9). The distribution among these varies greatly with coal rank. Low rank coal, possessing higher oxygen content, has a greater percentage of phenolic and acid functionality. As coal rank is increased, oxygen is lost by the more facile decarboxylation of acids as well as ether condensation and oxidation of phenols to quinones. This is portrayed in the distribution presented in Table 2. Organic sulphur is present largely as sulphides and thiophenols (thiols) with some as thiophenes. Nitrogen, a minor component, is present in heterocyclic pyridines, quinolines or bound in organometallic ligands such as porphyrins. As discussed in subsequent Chapter 4, the functional groups present and their quantity are extremely important when considering the changes that may occur in liquefaction or pyrolysis. • For maximizing liquid production, one requires considerable knowledge of both this native chemical structure as well as the transport and chemical processes that occur during conversion. Quantitatively, these factors affect the YIELD of liquids or tar (as - 6 -" Table 2 Oxygen D i s t r i b u t i o n i n Coal Organics Sub-Bi tuminous Bi tuminous Hyd r oxy1, phenoli c 5.6 wt. % 2.4 Carboxylic Acid 4.4 0.7 Ke t one, quinone 1.0 0.4 Ether 0.9 2.4 (Oxygen i s also present as inorganic complexes and water). - 7 -w e l l as s o l i d char and gases) that may be obtained from a given c o a l . Q u a l i t a t i v e l y , these f a c t o r s a f f e c t the SELECTIVITY towards production of premium grade f u e l s . C l e a r l y , the v i a b i l i t y of any conversion process must consider both the y i e l d and s e l e c t i v i t y c r i t e r i a . When a substance as heterogeneous as co a l undergoes thermal degradation two p h y s i c a l processes occur. F i r s t l y , those compounds possessing s u f f i c i e n t vapour pressure and d i f f u s i v i t y to escape from the coal matrix are d r i v e n o f f ( d e v o l a t a l i z a t i o n ) . General c o r r e l a t i o n s a p p l i c a b l e f o r vapour pressure ( P ) , such as the Antoine equation have the form: . „SAT . B In P = A - T + C and those f o r d i f f u s i v i t y from a porous s o l i d : P 0 E E 1 '47 'zi«' ' fi A $ 10 3 x T 1' 7 5 [ ( M A + M B)/(M A M g ) ] 1 7 2 where — - , / 3 ,/3 2 navy/ + ( E v ) * / J r A B where x i s the i n t e r n a l p o r o s i t y of the p a r t i c l e , r i s the average pore r a d i u s ; T i s the t o r t u o s i t y ; the molecular weight of d i f f u s i n g gases; v the atomic d i f f u s i o n volumes of the gases. In both cases, increased v o l a t a l i z a t i o n (hence increased y i e l d ) i s expected with increased temperature. Note al s o the strong dependence on p a r t i c l e s i z e , p o r o s i t y - 8 -and molecular weight. The second and consecutive process occurring, and the most crucial to this study, is the chemical change that occurs with thermal degradation. The degree of chemical reaction depends on the activation energy (E a) of the reaction considered, the concentration of the reactants and the temperature. Some reactions operate to generate liquids and gases, others to consume; the balance between these is not easily predicted. A great deal more is said of this in subsequent chapters. - 9 -3 . PYROLYTIC METHODS Thermal degradation i s practiced widely i n both i n d u s t r i a l and research applications. Because pyrolysis i s such a general technique an extremely wide range of methods has been developed for a va r i e t y of feedstocks. The complexity of the pyrolysis reactor i t s e l f varies with the nature of the product desired or the type of information required. In general, these methods involve the operations and variables shown in Figure 1. In examining a va r i e t y of pyrolysis systems one finds that a rough c l a s s i f i c a t i o n may be made of a limited number of these variables and operations. Table 3 characterizes a number of the systems reported in the l i t e r a t u r e according to some of the more important c r i t e r i a . This l i s t i s not meant to be exhaustive, but does demonstrate the wide range of conditions u t i l i z e d . As shown, not a l l methods are aimed at work with coal s p e c i f i c a l l y or even with commercial applications. Many of those l i s t e d were developed to study the pyrolysis chemistry of various pure compounds on a small batch scale. The results of these studies are p a r t i c u l a r l y useful in gaining an understanding of the reactions occurring in complex substances such as coal. A number of i n t e r e s t i n g features many of which have been incorporated in the pyr o l y s i s design presented below, can be seen from this l i s t . Temperature r i s e times for the sample are governed by heat transfer conditions within the reactor. P a r t i c l e size, reactor temperatures and gas Prandtl number are major considerations for coal - 10 -Figure 1 Pyrolytic Methods: Operations and Variables pyrolysis reactor Temperature pressure geometry packing contact time materials flow regime carrier gas Sample Sample Feed System 1° PI r Zone 2° P^  I Zone Pyrolyte Quench Product Recovery Static I Analysis Solid, liquid, gas v o l a t i l i t y reactivity Batch, continuous sample size, sample loading, pressure, temperature Thermal Energy Supply heating load heating rate heat transfer Dynamic Analysis species l i f e time thermal quench chemical quenching (trapping) product stabilization gas, liquid, solid selectivity, relative & operational yield - 11 -Table 3 Pyrolysis Reactors Reactor Batch of Sample Size Temperature* Vapor Residence Configuration Continuous or Feed Rate Rise Times Contact Time Reference C l a s s i f i c a t i o n Carbonization Proximate Analysis Destructive D i s t i l l a t i o n Thin Film Curie Point Therraobalance Thermogravlme t r i e Analysis Moving Bed VF Py VLPF Py Laser Photolysis MS Bombardment re tort tubular coa ted filament boat, ribbon flui d i z e d bed spouted bed c i r c u l a t i n g hed en trained st i r r e d r i s e r tubular tubular sample plug kg mg mg/s - t/h mg/s - t/h mg/s - t/h mg/s - t/h any any mg mg mg mg slow slow fast Intermediate fast fast fast fast Intermediate very fast very fast very fast very fast N.A. min. < 10" 3 s < 10" 2 s 10" 5 s 10" 1 s < 10" 3 s < 10 s < 10" 3 s 10 11,12 14,15,16, 17,18 13 19,20,21,22 23 ,24 ,25.26 27 28,29,30, 31,32.33 34 35 C l a s s i f i c a t i o n of Temperature Rise Times Slow In termedlate Past Very Fast Heating Rate CC/s) . < 1 5 - 100 10 3 - 105 > 10 Heat Up Time for Typical P a r t i c l e Used  minu tr»s 10 s to several minutes 10 ms to several seconds < 1 ms VF Py - vacuum flash pyrolysis VLPF Py - very low pressure flash pyrolysis MS - mass spectroscopy - 12 -pyrolysis. (See Appendix E: Heat transfer and coal particle tempera-ture profiles.) As well, this author has carried out some preliminary heat transfer calculations which demonstrate a significant effect of coal char thermal conductivity on particle core temperatures. Moreover, in the case where coal particles contact another surface, unsteady state conduction can significantly increase particle heating rates. Results from the literature on the effect of heating rate on coal pyrolysis yields are rather divergent and no clear statement can be made. Once a vapour or gas has been generated, the duration of time i t spends within the reactor (residence time) governs the degree to which that vapour may undergo various reactions. Hot gases emitted from a pyrolysis furnace must be thermally quenched to halt or retard vapour phase reactions. Depending on the precision required in determining the time-at-temperature, a number of methods may be used. For small scale flash pyrolysis units, cryogenic traps are most common. Chemical quenching or trapping may also be employed by injection of a reactant into the downstream section of the hot zone and so bring about in situ derivatization of the short-lived species present. Methanol or analine i s most commonly used for this purpose. In large scale reactors cooling jets of water or inert gases are frequently used in conjunction with condensers or heat exchangers. Construction materials for high temperature reactors have been shown to have a significant effect on the overall product yields and se l e c t i v i t i e s . Taylor and Kulich (36) demonstrated accelerated rates of olefin decomposition in metal reactors over so-called wall-less and quartz reactors. Albright and Yu (37) demonstrated accelerated coke - 13 -formation and higher conversions of acetylene, butadiene and benzene pyrolysed in an Inconel-800 reactor over that in aluminized Incoloy and quartz reactors. Similar effects may be expected in the pyrolysis of coal. To that end, reactor surface areas and materials of construction may influence char, liquid and gas selectivity. This makes comparisons between different studies d i f f i c u l t . Diluent gases supplied to purge tars or to provide a desired hydrodynamic condition (fluidized, spouted, entrained . . .) may alter pyrolysis results in two ways. First l y , the thermal conductivity of some gases such as hydrogen and helium are much higher than that of nitrogen, carbon dioxide or light hydrocarbons (CH^, C2H2, C 2H 6). Consequently, higher heat transfer rates and temperature rise times may be expected for the former gas environments. As well, some gases especially hydrogen (but also oxygen, olefins, H2S, N0X) can react with thermally excited hydrocarbons to alter the product. Use of hydrogen as the pyrolysis gas (hydropyrolysis) allows improved liquid and alkane yields. Pressurized hydropyrolysis reactors allow even greater improvements due to the kinetics of the hydrogenation process. With this brief introduction to both coal and general pyrolytic methods i t is now appropriate to begin a more detailed study of coal pyrolysis i t s e l f with emphasis on the chemical transformations and kinetics. t - 14 -4. TRENDS IN COAL PYROLYSIS Figure 2 depicts the major trends in product yields with temperature in coal pyrolysis. This graph, which illustrates qualitatively the yields of a variety of volatile gases, tar and char, is a composite picture derived from a number of studies. Actual yields or variations with temperature depend on the coal used, reactor configuration, hydrodynamic condition, gaseous environment and other factors. A number of observations may be made from this above figure. Tar yields typically peak at a maximum value of 20 to 35 weight percent of the dry ash-free basis coal weight. With gas yields added to this the major product is not volatiles but rather, the solid char. This char is essentially fixed carbon and mineral matter. In an integrated pyrolysis f a c i l i t y this char must be either burned or gasified. With this in mind, heating value, igni t a b i l i t y and physial structure of the char cannot be ignored in judging the overall f e a s i b i l i t y of such a coal utilization process. Overall tar yield maxima generally occur at 600 to 650°C. Some evidence that tars produced at different temperatures are not compositionally alike has been presented (38). Tars generated at relatively low temperatures (400-500°C) are generally higher in heteroatoms (oxygen, nitrogen, sulfur) than those generated at higher temperatures. Hydrogen to carbon ratios of tars show a steady decrease with temperature from a value of 1.2 at 400°C to 0.8 at 900?C. A l l of these factors suggests that the overall yield of liquids is not a proper - 15 -Figure 2 TYPICAL COAL PYROLYSIS PRODUCT YIELDS TEflPERATURE (C) [see references: 13, 19, 23, 74']. - 16 -tool with which to judge the commercial value of the tars, but that quality and up-gradeability measures must also be considered. The y i e l d of d i f f e r e n t species of gases shows a very strong temperature dependence. Since very l i t t l e gas i s evolved u n t i l a s u f f i c i e n t l y high temperature i s reached (> 400°C) one must conclude that chemical reactions r e s u l t i n g i n gas elimination must be occurring rather than desorption. The temperatures at which xeach gas i s evolved suggests the degree of thermal a c t i v a t i o n required to i n i t i a t e the gas generating reactions. Again, gas y i e l d and composition are important to any economic assessment. The use of off-gases as preheating f u e l , as a feedstock for upgrading to pipeline gas, or for reforming to produce hydrogen for hydrotreating of the l i q u i d s are a l l possible. 5. PYROLYSIS CHEMISTRY The o v e r a l l results noted above are the consequences of many events occurring at the molecular l e v e l . Chemical k i n e t i c s and thermodynamics of reactions i n the gas and s o l i d phases ultimately determine the product d i s t r i b u t i o n and y i e l d s . In this section we examine some of the reactions one would expect to find in the p y r o l y s i s of c o a l . For this study two things are required: from the previous section the t y p i c a l d i s t r i b u t i o n of functional groups and structures found i n coal serve to i d e n t i f y the types of reactants involved. Next, pyro l y s i s of pure compounds possessing these functional e n t i t i e s indicates the products and pathways for these reactants. I t w i l l become obvious as the reaction pathways are examined that such a study i s u n l i k e l y to provide the means of determining d e f i n i t i v e l y the ov e r a l l results from the pyrolysis of a material such as co a l . The objectives of this section are more q u a l i t a t i v e : demonstrating typical reactions, how they occur and what factors influence them. Pyrolysi s of fixed carbon i t s e l f i s r e l a t i v e l y uneventful. Compounds ph y s i c a l l y or chemically adsorbed on the matrix can be thermally desorbed with r e l a t i v e l y low energies. Carbon i t s e l f remains r e l a t i v e l y unreactive except at very high temperatures (> 800°C) where g a s i f i c a t i o n reactions may occur. The most important chemistry involves the thermal decomposition of the organics and mineral matter. These reactions may occur within the s o l i d matrix or in the gas phase once the compound i s l i b e r a t e d . Uniraolecular, f i r s t order reactions, where a - 18 -c o a l organic decomposes on i t s own, are expected to be most s i g n i f i c a n t i n the e a r l y stages of a p y r o l y s i s . As more r e a c t i v e components are generated, bimolecular and other secondary reactions become important. The p y r o l y s i s of such a simple m a t e r i a l as ethane (C2H6) demonstrates the complexity of these r e a c t i o n systems: A H3C-CH3.. »»H 2C = CH 2 + CH4 + H 2 + H 3C-CH2-CH2-CH3 ethylene n-butane The most widely accepted mechanism f o r t h i s r e a c t i o n was proposed by Rice and H e r z f e l d (39): A H3C-CH3 H3C • + «CH3 unimolecular i n i t i a t i o n •CH 3 + C2H6 A, »CHi4 + *C2H5 •C2H5 ^ — * H 2C=CH 2 + H* free r a d i c a l chain r e a c t i o n s A H* + C2H6 -=-* 'C2H5 + H2 2C2H5 ~ > H3C-CH2-CH2-CH3 termination I t must be remembered that once the concentration of the products increases a new set of re a c t i o n s i s then p o s s i b l e : CH 4 . A,» C 2H 6 + H 2 (40) 700°C CH 2=CH 2 •> HC=CH + H 2 4- H3CHC=CH2 690°C (41) HCECH H3C-C=CH + «c=G-c.=crt - 19 -H3C-CH=CH2 » H2C=C=CH2 + H 3C-C =CH (42) allene As well, bimolecular reactions may occur between reactants, intermediates and products: •CH3 + H2C=CH2 ^ H3C-CH2-CH3 H2C=CH-CH3 + H 2 •CH3 + HC=CH -A* H3C-C=CH >- H3C-C=CH + H 2 + CH3 •CH3 + H3C-CH=CH2 • A C H 4 + H2C-CH=CH2 The complexity of pyrolysis reactions becomes immediately apparent. From the above discussion, a number of observations may be made: Pyrolyses involve mainly free radical reactions. Other reactive intermediates, including carbenes, carbocations or anions may also be involved in thermolysis but radicals are the dominant species. Radicals are created mainly by the homogeneous cleavage of a chemical bond. These intermediates are usually very reactive species which react readily with a wide range of materials. Some radicals, usually polyaromatic, are reasonably stable, but these are fa i r l y rare. As well, pyrolyses are very complex. With a free radical intermediate, a l l of the following courses of action are available as secondary processes: 1. unimolecular fragmentation to produce a stable molecule (olefin, ketone) and a less stable radical (see above) - 20 -(CH 3) 3 CO H3C-CO-CH3 + «CH3 2. act as a propagator or chain carrier in further chain reactions by: (i) hydrogen abstraction ( i i ) substitution •CH3 + Et-S-Et — E f + CH3-S-Et 3. addition to olefins or other multiple bonds causing chain branching and alkylation 5 . radical-radical coupling reactions that are terminal. 4. radical-radical reactions that are not terminal. 6. disproportion tion: (i) intermolecularly H 1 R-CH-CH-CH-CH-CH- R • R-Crt-CH-CH-CU -• T. X 2. 2. 2. Z. ( i i ) intramolecularly CH CH, f +• <b vA q> A VI - 21 -In most of the above reactions, the ac t i v a t i o n energies are very low. Because of this, major changes i n composition can occur within brief periods ( 1 0 - 3 seconds) of time. Also, with so many competing processes, steady state r a d i c a l fluxes could develop. Key intermediates such as hydrogen (H«), hydroxyl («0H) and carbon centred r a d i c a l s (R*) could presumably reach this condition. These may then become important l i m i t i n g reagents i n cer t a i n reactions. Contrasting this r e a c t i v i t y i s the apparent d i f f i c u l t y i n i n i t i a t i n g most p y r o l y t i c reactions. Most py r o l y s i s reactions do not occur u n t i l temperatures i n excess of 300 VC are reached. This suggests very high a c t i v a t i o n energy b a r r i e r s to reaction i n i t i a t i o n ; chemical bonds must be cleaved before r a d i c a l s are created. Consider the following reactions and the respective bond d i s s o c i a t i o n energies (43): BDE (KJ mol - 1) 435 ± 5 + H- 460 ± 10 H £ C ~ CH"-CH ^  H2C=CH-CH2 + H« 373 ± 5 2 .CH3 368 ± 5 368 + 5 22 -or — cr' 2 8 9 ± 1 0 285 ± 5 CO,H The easier the bond i s to break the lower the temperature required to initia t e the reaction. The more stable the radical product, especially those radicals involving delocalized electrons (benzylic, a l l y l i c ) the easier the reaction. Other functional groups can yield heteroatora radicals: -r- «OH cr — o 460 or0 +• H 356 ± 20 CH3-S-S-CH3 2 CH3 S 249 C 2H 5 - Sn(C 2H 5) 3 * Sn(C 2H 5) 3 + C 2H 5 239 HO-O-H -^ 2 OH 213 C 2H 5 - Zn(C 2H 5) * C 2 H 5 + Zn(C 2H 5) 200 H 2 »- 2H. 43{ H2S Hi + -SH 378 ± 4 o-- 23 -R e c a l l i n g the f u n c t i o n a l groups present i n c o a l , s i m i l a r r e a c t i o n s to- the ones above can be expected. As temperatures i n c r e a s e , generation of a v a r i e t y of r a d i c a l s occurs, i n p r o p o r t i o n to t h e i r r e s p e c t i v e a c t i v a t i o n energies and substrate c o n c e n t r a t i o n . These f a c t o r s lead to the complexity of p y r o l y s i s r e a c t i o n s . The primary process ( r a d i c a l generation, i n i t i a t i o n ) i s very energy i n t e n s i v e and produces a m u l t i p l i c i t y of r a d i c a l s at a given temperature. The secondary processes ( c h a i n , r a d i c a l and termination r e a c t i o n s ) are f a s t and r e q u i r e l i t t l e a c t i v a t i o n . The d i s c u s s i o n so f a r i s s u f f i c i e n t to e x p l a i n many of the c o a l p y r o l y s i s products noted p r e v i o u s l y . C l i p p i n g of short chain a l k y l fragments from s u b s t i t u t e d aromatics creates a f a i r l y s t a b l e aromatic r a d i c a l and a h i g h l y mobile a l k y l r a d i c a l . Having high d i f f u s i v i t y , a l k y l r a d i c a l s such a methyl (»CH3), can a b s t r a c t hydrogen from other organics r e s u l t i n g i n very s t a b l e methane. The same pathway could give e t h y l , propyl or higher r a d i c a l s . The f a c t that very l i t t l e of the higher alkanes are produced i s expected since l a r g e r chain a l k y l s u b s t i t u t e d aromatics are not the predominant coal species. A l k y l r a d i c a l s may a l s o a l k y l a t e an unsaturated hydrocarbon, a feature which can be very d e s i r a b l e i n i n c r e a s i n g the value of tar products as motor f u e l feedstocks. The hydrogen r a d i c a l f o l l o w s a s i m i l a r route as most of a l k y l r a d i c a l s , e s p e c i a l l y hydrogen a b s t r a c t i o n to produce hydrogen gas ( H 2 ) . Hydrogenation of an unsaturated compound i s a l s o p o s s i b l e , as i s hydrogen reduction v i a termination of a hydrocarbon r a d i c a l (R*). These r e a c t i o n s r e s u l t i n a decrease i n entropy and are not as - 24 -thermodynamically favorable as the r e a c t i o n s that give a net increase i n gaseous products. K i n e t i c a l l y , many of these r e a c t i o n s i n v o l v e two reagents that are present i n low concentrations (H* and R* or unsaturated RHXO) and so progress at much slower rates unless an abundance of hydrogen r a d i c a l s can be generated. This i s the idea behind h y d r o p y r o l y s i s , to increase the steady s t a t e hydrogen r a d i c a l c o n c e n t r a t i o n by thermal s p l i t t i n g of H 2. Of great i n t e r e s t and p o t e n t i a l l y the greatest danger to hydrocarbon p y r o l y s i s i s the hydroxyl r a d i c a l (»0H). K i n e t i c a l l y t h i s species i s very r e a c t i v e and has a very high d i f f u s i v i t y . The p r e f e r r e d 4 route of d e s t r u c t i o n of hydroxyl r a d i c a l s i s to a b s t r a c t hydrogen and give s o - c a l l e d " p y r o l y t i c water." Thermodynamically t h i s i s an extremely favorable r e a c t i o n , being p r a c t i c a l l y i r r e v e r s i b l e . In c o a l processing, hydrogen i s at a premium. Therefore, any reactant that can s t r i p hydrogen i s very damaging to the o v e r a l l process. Other hydrogen a b s t r a c t i o n s (such as w i t h a l k y l r a d i c a l s ) are not nearly so f i n a l . R e c a l l i n g the data presented e a r l i e r , i t w i l l be remembered that the low rank coals were r i c h i n hydroxyl sources ( p h e n o l i c s , quinones,) and are considered more r e a c t i v e . This may seem chemically c o n t r a d i c t o r y . I t must be r e c a l l e d though that s p l i t t i n g o f f hydroxyl r a d i c a l s i s one of the more d i f f i c u l t bond cleavages. Many other b e n e f i c i a l r e a c t i o n s w i l l be i n i t i a t e d before hydroxyl r a d i c a l s are released. While hydroxyl r a d i c a l s are c l e a r l y a major detriment, they could, under c e r t a i n circumstances become a very u s e f u l reactant to - 25 -a s s i s t i n obtaining value-added hydrocarbons. While not e n t i r e l y understood, pyrolysis of lower alkanes i n quartz and "wa l l - l e s s " reactors with minute concentrations of oxygen present have demonstrated improvements i n alkene (and other unsaturates) y i e l d and s e l e c t i v i t y at k i n e t i c a l l y enhanced rates. I f , i n the pyrolysis of coal, alkene yields can be improved, this would be extremely s i g n i f i c a n t i n increasing commercial v i a b i l i t y . F i r s t l y , any alkene produced i s a premium product having immediate markets i n the petrochemical industry. Secondly, alkenes are frequently used i n t y p i c a l o i l upgrading operations as al k y l a t i n g agents. A l k y l a t i o n of most petroleum f r a c t i o n s r e s u l t s i n a more valuable product for ultimate use i n gasolines, d i e s e l f u e l , j e t f u e l and l i g h t f u e l o i l s . If in s i t u alkene y i e l d s could be increased, the p o t e n t i a l for use of coal off-gases as petrochemical feedstocks or for " s e l f - a l k y l a t i n g " reactions, would be very b e n e f i c i a l . Both Martin et a l . (44) and Taylor et_ a l . (45) have proposed mechanisms involving oxygens as a chain c a r r i e r i n hydrocarbon py r o l y s i s chain reactions. I n i t i a t i o n remains the same (cleavage of C-H and C-C bonds) but oxygen enters the propogation steps as i n : • C2H5 + 0 2 »- C2H1+ + H0 2» H* + 0 2 H0 2 H0 2 + C 2H 6 H 20 2 + C 2H 5 H 20 2 2H0« - 26 -H<5 + C 2H 6 — H 2 0 + (52H5 H» + 0 2 — H 0 « +0: 0: + C 2H 6 —»- C 2H 5 + OH Note that with the additional chain carriers («0H and 0:) there is the potential for more hydrocarbon radicals. Secondary reactions of hydro-carbon radicals could presumably lead to greater hydrocarbon degradation and production of smaller fragments particularly unsaturates. Metal surfaces that have not been oxidized can act as effective radical scavengers for «02H and »0H radicals. This may account for some of the catalytic and reactor surface effects noted in many studies. Removal of •OH radicals may retard kinetic chain branching. This would not be beneficial to coal volatiles that need degradation induced by radicals. On the other hand, this metal deactivation could be of benefit to coals that are lean in hydrogen by removing these radicals that strip away the valuable and already scarce hydrogen. The chemistry of sulphur is very much like that of oxygen. Because of this, many of the above oxygen-based reactions have sulphur analogs. Hydrogen abstraction, radical termination and chain reactions occur with *SH radicals as they do with «0H. However, one significant difference has been noted in the previous pages. Unlike water, any hydrogen sulfide (H2S) present may be homolytically cleaved to give a hydrogen (H«) and HS« and re-enter the reaction pathways. The biggest source of sulphur radicals is likely to be from transition metal sulphides which are known to liberate sulphur species upon heating. - 27 -Sulphur functional groups present i n coal organics behave i n a manner si m i l a r to their oxygen counterparts and most often end up i n gaseous products (CS 2, COS, H 2S, S 0 2 , HRS). Coking and char producing reactions are very poorly understood. Two schools of thought have conceptualized char formation. The f i r s t involves extensive "polymerization." While a misnomer, the idea i s that reactive organics such as o l e f i n s or free r a d i c a l s react with each other to produce an ever lengthening molecule. This process occurs u n t i l the molecular i s of i n f i n i t e dimensions, e n t i r e l y stripped of hydrogen and heteroatoms and e s s e n t i a l l y a fixed carbon matrix. The a l t e r n a t i v e idea involves nearly the opposite: a molecule i s continually excited to higher e l e c t r o n i c energy l e v e l s and ultimately fragments down i n t o atomic carbon. A l l hydrogen and heteroatoms are l i b e r a t e d and carbon atoms re-arrange to form the f a m i l i a r carbon matrix. While these two views are not e a s i l y resolvable, this author favors the l a t t e r mechanism. Large molecular weight fragments (MW > 1000) have not been i d e n t i f i e d i n coal or coal tar to any meaningful extent. As well, the build up of molecular l i n k s implies a high degree of mobility of the "monomers" involved. In a s o l i d tortuous char p a r t i c l e this mobility i s not a v a i l a b l e so as to allow the rates of coking near those observed. I t i s more l i k e l y that an organic species trapped inside a char pore, without any other substrates available with which to react, w i l l be forced to react intra-molecularly and ultimately o s c i l l a t e i t s e l f into free atoms or very small reactive fragments. The association of small unsaturated compounds with metal atoms i s well documented and i s - 28 -consistent with the observed enhancements in coking seen when metals are present (46). As coal is essentially free of any non-conjugated double bonds (free olefins), char formation must proceed via a radical mechanism. This could be either a radical-radical termination or radical-olefin or olefin-olefin additions once these secondary products are formed. These reactions must compete with other radical and unsaturate reactions and so char formation could be reduced by any agents that could assist the competing reactions. For instance, reductions in the quantity of transition metals present, hydrogen or very low pressures would reduce char formation. One of the principal methods of improving yields and quality of pyrolysis products is by the use of a hydrogen atmosphere. As noted, hydrogen in coal organics is very valuable but is easily abstracted by a variety of radical species. By using hydrogen as the pyrolysis gas these reactions become s t a t i s t i c a l l y less likely, while the reverse reaction of hydrogenation is more li k e l y . Hydrogenation of organics can occur in three ways. The f i r s t involves transition metal hydride complexes which react with carbon-carbon multiple bonds to directly add hydrogen without any radical intermediate. This is a commonplace heterocatalytic mechanism in hydrocarbon processing. Alternatively hydrogen may react via a radical route. Carbon or oxygen centred radicals could directly react with gaseous hydrogen to give a hydrogen radical and a reduced form of the parent structure. Hydrogen i t s e l f may also be thermally homolytically cleaved to give two hydrogen radicals. These may then add directly to any unsaturated organic or become - 29 -involved i n a r a d i c a l - r a d i c a l termination, both of which add hydrogen to the organic substrate. Obviously any means of adding hydrogen to coal organics i s advantageous. I t i s also more economic to do this in s i t u rather than as an upgrading measure after tars are generated. By using hydrogen gas to terminate hydroxyl, sulphur and char-precursor reactants, the undesirable products of "pyrolysis water" and coke may be d r a s t i c a l l y reduced. I t i s more d i f f i c u l t to add hydrogen to tars and impossible to rehydrogenate char downstream once i t has been abstracted i n the py r o l y s i s process. Most of the f r e e - r a d i c a l mechanisms discussed thus far have involved some combination of steps involving horaolytic bond d i s s o c i a t i o n , atom abstraction, and r a d i c a l addition, i n the i n i t i a t i o n and propagation phases of the reaction. In addition, d i s c r e t e electron transfer processes may occur between organic molecules and t r a n s i t i o n metal complexes. In t h i s , an electron i s added or removed from a diamagnetic organic molecule to generate a r a d i c a l , r e s u l t i n g i n oxidations or reduction of a metal atom. Such electron transfer reactions are commonly encountered because of the many t r a n s i t i o n metals having more than one stable oxidation state. Metals then often act as c a t a l y s t s , reagents or i n the i n i t i a t i o n s of free r a d i c a l reactions by this electron transfer process. The decarboxylation of d i a c y l peroxides i s a t y p i c a l example (47): O - 30 -Decarboxylation of acids and subsequent ligand transfer reactions have also been observed (48). ^ ^ - ^ j f 0 H 4. PbdOAo\ J L ^ UW ) R- -v Pt>6v)X + do2 o R 0 In coal, transition metal ligands such as sulfides or carbonates could be liberated in this fashion. These may then become significant features In the overall reaction pathways. Depending on the quantity, oxidation state, ligands, metal atom and proximity to organic reactants, such transition metal reactions could have profound effects on the overall degradation of organics. The probability and course of organo metallic themolysis reactions is poorly understood. Considering the oxidation states available, the types of metal complexes present and the possibility of association, radical and electron transfer reactions, metals greatly Increase the overall complexity of these reactions. As is frequently the case in heterocatalytic studies, the best that can be - 31 -done is to note any catalytic effect or inhibitions caused by an added organometallic reagent. The results are very often purely empirical, matrix specific and mechanistically d i f f i c u l t to comprehend. As coals from different formations or seams of the same formation may contain different quantities and types of metals, their pyrolysis products may be dramatically affected. Two very important classes of pyrolytic reactions do not involve identifiable radical intermediates or transition states. These reactions involve a variety of intra-molecular rearrangements, or extremely fast intra-molecular radical or three centred transition state reactions, the details of which are not important to this thesis. Both reactions are very significant in terms of the products seen from coal functional group reactions. The f i r s t of these are the elimination reactions. In a generally faci l e reaction (hence low temperature) that is strongly favoured by entropic considerations, the bulk of the gases (CO, CO2, H2S, SO2) are liberated from coal. Some typical examples are: (49) COR.' (50) o (51) - 32 -(52) (53) (54) (56)(SS) (57) (58) (59) C O * O (60) - 33 -(61) In the above reactions, aromatic nuclei remain largely i n t a c t while small fragments are eliminated. This process greatly improves the quality of the tars produced for use as f u e l s . Many of these reactions are affected by the presence of heterocatalytic metal atoms. In some cases, the favourable elimination w i l l i n i t i a t e complete degradation of the parent molecule, as seen i n the reactions above. Such eliminations are important i n breaking down larger molecules to more v o l a t i l e fragments. This process may attribute to the observed higher " r e a c t i v i t y " of lower rank coals. The greater the content of heteronucleic species, the more pronounced the degradation, p a r t i c u l a r l y at lower temperatures. The second major category of reactions that do not involve r a d i c a l intermediates i s that of retrogression or retro Diels-Alder reactions. These occur i n a concerted mechanism to l i b e r a t e a diene and dienophile molecule. These reactions are reversible and are affected strongly by met a l l i c c a t a l y s t s . For organics l i k e coal tar, which are extremely aromatic, retro Diels-Alder reactions provide the route to the breakdown of large aromatics and the construction of l i g h t e r aromatis v i a smaller o l e f i n i c u n i t s. Some typ i c a l reactions are: 0 O (62) O o - 34 -CH III f ^ l . ^ r f - ^ _ ^ III I , u . (63) O ^  III * I - «L HC = Ct4 0 II GH + 1* (64) (65) (66) (67) (68) Such reactions are very d i f f i c u l t to study. The derivative dienes are much more reactive than the parent structures and so react r e a d i l y i n a v a r i e t y of mechanisms to produce an extremely complex product d i s t r i b u t i o n that masks the primary processes. The r e v e r s i b i l i t y of the reactions also causes confusion. The forward - 35 -reaction of dienes and olefins provide a means of expanding aromatic structures by adding back the small fragments produced. Because of this reversibility, olefin fragments are not generally observed in significant concentration in coal pyrolysis of gases until very high temperatures. Thermodynamically this i s the expected result: the small increase in entropy in creating the fragments will only become important at higher temperatures. Free olefins or dienes are also very reactive with any free radical species. Because of this a similar low concentration steady state level would be expected for olefins in the gas phase. The above discussion outlines the bulk of the chemistry expected in pyrolysis of coal compounds. As can be seen, the chemistry is complex. In Figure 3, a simplified reactions pathway flow diagram and index of terminology has been developed by this author. In this, a l l of the above reactions are diagramatically represented. This pathway chart provides a valuable tool in understanding the effects of coal type, composition and reactant gases on the products obtained. While developed for coal, the flow sheet is also applicable to a l l other hydrocarbon mixtures in both gas, solid and fluid reactions. This diagram by i t s very nature is not meant to encompass a l l reactions possible, and certainly some significant ones were inadvertently omitted. This is believed to be the f i r s t real attempt to put the myriad of reactions into a useable form. It is hoped that, after exploring the various pathways, this chart will assist researchers in understanding pyrolysis chemistry. - 36 -Index to Figure 3 Reaction Pathways i n Hydrocarbons Pyrolysis Index to o v e r a l l reaction pathways RHXO general hydrocarbon, neutral R = carbon structure, "H = hydrogen, X = heteroatom (nitrogen, s u l f u r ) , 0 = oxygen RH general hydrocarbon, neutral hydrogen enriched or heteroatoms removed RX general hydrocarbon, neutral heteroatom enriched C0 x oxide gases (CO, C0 2) HX reduced form of heteroatom r a d i c a l (H 20, H 2S) H2 hydrogen gas X2 dimer of heteroatom radicals (H2O2, H2S2) R2 high molecular weight hydrocarbon product R» hydrocarbon r a d i c a l (carbon or oxygen centred) H* hydrogen r a d i c a l X* heteroatom r a d i c a l (*0H, *SH) 0 superoxide r a d i c a l olefins/unsaturates alkanes, alkynes, dienes, reactive carbon-carbon multiple bonds (including some aromatics) char fixed carbon, unreactive A — — reaction pathway for material A A j bimolecular reactions of A and B B products thermal or other energy input - 38 -6 . PYROLYSIS KINETICS In terms of designing reactors for ultimate commercial exploitation, chemical kinetics is of the utmost importance. It has been observed by numerous authors that coal pyrolysis reactors operate in a chemical kinetically-controiled regime, far removed from equilibrium. Mass and heat transfer are obviously of importance, but chemical constraints are the principal factors in pyrolysis. Some theoretical aspects of pyrolysis kinetics are revealed in the previous section. Initiation of thermal degradation involves a sizeable energy input. Once started, secondary reactions tend to require very l i t t l e energy. By examining pyrolysis from a mechanistic point of view as above, the overall sequence can be broken down into a series of elementary steps. This method allows a direct translation into familiar chemical kinetic expressions. For simplification the familiar Arrhenius expression is used: For the elementary step aA + bB — c C the reaction rate is given by R = ll£l = k l A ] a [ B ] b at where the rate constant for this elementary step is given by k = A exp (-E /RT) or A' T N exp(-E /RT) - 39 -w i t h A a constant (frequency f a c t o r ) and E a the r e s p e c t i v e a c t i v a t i o n energy. I n i t i a t i o n r e a c t i o n s (homolytic cleavage, r e t r o g r e s s i o n , e l i m i n a t i o n ) have high a c t i v a t i o n energies (120 to 440 KJ/mol), and hence r e q u i r e high temperatures before an appreciable ra t e of r e a c t i o n i s observed. The m a j o r i t y of i n i t i a t i n g r e a c t i o n s are unimolecular. The rate of decomposition of a given hydrocarbon (RHXO) then approximates: _dJM0l = k[RHX0] at = AT N exp(-E ±/RT) • [RHXO] Where T i s the temperature of the reactant ("n" i s u s u a l l y about 1/2). In a l l cases, temperature v a r i e s with time, the gradient depending on the heat t r a n s f e r to the coal p a r t i c l e . This time dependance i s u l t i m a t e l y r e f l e c t e d i n the e f f e c t s of residence time and p a r t i c l e temperature r i s e times on y i e l d and s e l e c t i v i t y . As shown p r e v i o u s l y , a number of i n i t i a t i n g unimolecular r e a c t i o n s may occur simultaneously. The temperature r i s e time then governs the degree to which each r e a c t i o n occurs while the coal v o l a t i l e i s i n the hot zone of the r e a c t o r . At low heating r a t e s , lower a c t i v a t i o n r e a c t i o n s dominate ( r e t r o g r e s s i o n , e l i m i n a t i o n ) u n t i l temperatures that are high enough to s t a r t other r e a c t i o n s are reached. At higher heating r a t e s , a l l r e a c t i o n s ( i n c l u d i n g cleavage r e a c t i o n s ) could occur simultaneously a f t e r a very - 40 -short period of time. Predi c t i n g the outcome of so many simultaneous and consecutive reactions under d i f f e r e n t heating regimes i s extremely d i f f i c u l t . Bimolecular reactions are involved in the major secondary processes. These include r a d i c a l abstractions, chain, termination and coupling reactions. For bimolecular elementary steps, such as the termination of a carbon centred hydrocarbon r a d i c a l by hydrogen, the rate expression i s given by: R = -k[R-] [H-] where the rate constant k i s as above, with the respective a c t i v a t i o n energy value used. Both r a d i c a l s w i l l be present at very low l e v e l s . The a c t i v a t i o n energy involved, tends to be very low (0-50 KJ m o l - 1 ) , and so an appreciable rate may s t i l l be observed despite the low concentration. A large number of secondary reactions, including r a d i c a l - r a d i c a l and r a d i c a l - o l e f i n reactions, have s i m i l a r l y low a c t i v a t i o n energies. Product d i s t r i b u t i o n from these reactions w i l l then depend to a large extent on the gas phase residence time and reactant concentrations. Residence time may be stated in two ways: to t a l residence time and time-at-temperature. Total residence time i s the period between sample introduction, heat-up, reaction and quenching. This includes the time-dependent heat-up period discussed above. Time-at-temperature i s the period a reactant spends in a certain temperature range. Heat - 41 -transfer to and from the reaction system are then considered independently. For very high heat-up rates, the two residence time values may be considered the same. Short residence times may allow in i t i a t i o n but only minor degrees of secondary reaction. Longer times allow extended secondary processes. Since these latter reactions are very fast, times the order of milliseconds could drastically affect the observed product. In typical pure compound mechanistic pyrolysis studies, residence times in the order of 10 second are needed to observe the sequence of intermediate and derivative products. In much of the work done to date on coal, such times have not been achieved. Most coal work has centered on reactions with residence times of the order of several seconds. Over such a period of time, considerable secondary conversion could occur. From the reaction scheme given earlier i t is evident that there are both constructive (tar-yielding) and destructive (char, gas-yielding) reactions. These occur simultaneously and consecutively from the same reactant. Predicting the balance between these processes over a given period of time requires detailed chemical and kinetic information. Even qualitative predictions are therefore d i f f i c u l t . For a more complete examination of this, part of this research was directed at a theoretical prediction of pyrolysis products. By combining the previous reaction diagram and estimated reaction rates and concentrations, an &_ priori kinetic simulation was attempted. Unfortunately, conceptual and time constraints prevented completion of this endeavor. For a more complete discussion the reader i s referred to Appendix A, "Pyrolysis Kinetic Simulation." - 42 -A few of the reaction classes do warrant some discussion. Hydropyrolysis reactions are of two types: free radical addition and heterocatalytic hydrogenation. For the latter, hydrogen is added directly to an unsaturate via adsorption-desorption from a metal atom. Variations in coal minerals or added catalysts could affect this mechanism of hydrogenation. Such reactions can occur at relatively modest temperatures due to the low activations energies involved. Hydrogenated products then open up more channels to secondary reactions. The alternative free radical induced reaction involves the reaction of a hydrogen radical with either a hydrocarbon radical or the neutral parent. Of course, the hydrogen radical must f i r s t be created. Cleavage of the H-H bond in hydrogen requires in the order of 435 KJ mol - 1. This is extremely energy intensive and so requires very high, temperatures to be intiated to any significant degree. Hydrogen radical terminations and chain reactions must compete with numerous other destructive reactions. From such a hydrogen deficient state (H/C ratios » 0.8), starting coal organics must gain large amounts of hydrogen before reaching near-premium products (H/C ratio > 1.5) (71). Kinetically to achieve this requires very high hydrogen partial pressure and extreme temperatures so as to increase the hydrogenation chain length over other competing reactions. This may explain the general result of greatly improved selectivities for gas phase hydropyrolysis at elevated temperatures (> 800°C). There are numerous routes to char. The reaction pathway includes - 43 -both the high molecular weight route and the fragmentation routes discussed e a r l i e r . The reac t i o n s may be represented as: R-' + RHXO R' - RHXO R.' + R . •*,„ R-R» C 1 0 0 H 1 8 °5 * 2 H 2 ° + 3 C 0 + C 9 7 H m *-97C + 7H 2 / Note that a l l char formation r e a c t i o n s are bimolecular and i n most cases i n v o l v e low a c t i v a t i o n energy p o t e n t i a l s . Char producing intermediates may be "deactivated" by enhancing other r e a c t i o n s , notably fragmentation and hydrogenation. The d i f f u s i v i t y of these intermediates w i t h i n the s o l i d c o a l p a r t i c l e and the gas phase to a great extent determines whether coking takes place or not. By p r o v i d i n g reactants with high r e a c t i v i t i e s and d i f f u s i v i t i e s (such as H«)» coking may be reduced. A l t e r n a t i v e l y , compounds trapped w i t h i n the i n t e r s t i c i e s of a co a l p a r t i c l e may have no recourse but to fragment i n t o smaller more mobile fragments. Fragmentation and r e t r o - D i e l s Alder r e a c t i o n s are low energy routes. By keeping a modest temperature (500-700°C) these may be favoured f o r the case of the encapsulated v o l a t i l e . I f higher temperatures are encountered, the rates of H-abstraction may become s u b s t a n t i a l and hence allow greater char formation. This mechanism may i n p a r t e x p l a i n why tar y i e l d maxima are generall y i n the 600°C range. - 44 -Of great commercial significance to p y r o l y s i s reactor design i s the k i n e t i c e f f e c t of loading. Loading may be defined as the quantity of coal fed to the reactor per unit of p y r o l y s i s gas. C l e a r l y as the amount of coal feed i s increased, so does the concentration of hydrocarbon. Technically, e f f e c t s of coal type, p a r t i c l e size and temperature should be factored in to give the true value of hydrocarbons ava i l a b l e i n the gas phase. To a commercial scale pyrolyser the important question i s the o v e r a l l e f f e c t of solids loading on tar y i e l d . The goal i s c l e a r l y to maximize throughput as far as i s economical based on the o v e r a l l tar y i e l d . Chemically, many reactions involve the parent hydrocarbons (RHXO). These include the unimolecular f r e e - r a d i c a l i n i t i a t i o n s as well as the numerous bimolecular r a d i c a l chain reactions. A l l else being equal, the i n t i a t i n g reactions indicate that tar y i e l d should vary d i r e c t l y with coal loading. Assuming some i n i t i a t i o n , the secondary processes show a contrary r e s u l t . As loading increases, the detrimental abstractions and coking reaction are increased as more hydrocarbon becomes available for attack by r a d i c a l s . The desirable i n i t i a t i o n reactions have a c t i v a t i o n energies that are an order of magnitude greater than most of these detrimental secondary e f f e c t s . For these reasons, any p o t e n t i a l benefit of increased loading would l i k e l y be quickly erased by these much faster secondary reactions. By t h i s theory then, high throughput reactors (that do not have coal feed rates that a f f e c t the o v e r a l l p y r o l y s i s temperature) should be poorer performers compared with d i l u t e loaded ones. - 45 -The r e a c t i o n pathway approach a l s o allows one to make rough estimates of the rates and energetics of pyr o l y s e s . For short residence time systems, the o v e r a l l process may be modelled as a two-stage process i n v o l v i n g two d i s t i n c t r e a c t i o n regimes. These are the s o - c a l l e d primary and secondary r e a c t i o n s . As stat e d e a r l i e r , these primary r e a c t i o n i n v o l v e a c t i v a t i o n energies much greater than the secondary ones. This allows us to make a very rough approximation as i s fr e q u e n t l y done i n cases where one r e a c t i o n i n a sequence i s e n e r g e t i c a l l y much more i n t e n s i v e than the others. In t h i s , the primary process i s taken as the rate-determining step (RDS) and so governs the o v e r a l l r e a c t i o n s seen. A l l other steps are ignored as they are u l t i m a t e l y dependent on the primary r a t e . The o v e r a l l r e a c t i o n r a t e i s then given by: R a t e R D S = k« [RHXO] k = A q e x p ( - E a > R D S / R T ) The o v e r a l l a c t i v a t i o n energy can be represented by a mole weighted average according to the d i s t r i b u t i o n of each c l a s s of primary r e a c t i o n (homolytic cleavage, r e t r o g r e s s i o n , hetero c a t a l y t i c cleavage, e l i m i n a t i o n ) . T y p i c a l values f o r the a c t i v a t i o n energies f o r each of the r e a c t i o n s are discussed i n Appendix A. L i g n i t e s and sub-bituminous coals have a preponderance of f u n c t i o n a l groups amenable to the low energy e l i m i n a t i o n and r e t r o g r e s s i o n routes. Bituminous and a n t h r a c i t e coals have fewer options open and must react mostly by cleavage - 46 -r e a c t i o n s . By f i t t i n g experimental y i e l d versus temperature r e s u l t s to a general i z e d f i r s t order Arrhenius expression, many i n v e s t i g a t o r s have derived an o v e r a l l a c t i v a t i o n energy value. From t h i s d i s c u s s i o n , both bituminous and sub-bituminous coals should have r e l a t i v e l y high a c t i v a t i o n energies, the l a t t e r being somewhat lower. This p r e d i c t i o n i s c o n s i s t e n t with the observed experimental trends. While a misnomer, l i g n i t e s and sub-bituminous coals are g e n e r a l l y considered more " r e a c t i v e " than the higher rank c o a l s . As a q u a l i t a t i v e l a b e l , t h i s would seem appropriate according to the above energetic c o n s i d e r a t i o n s . Many studies have been performed i n an attempt to derive c o a l k i n e t i c expressions using simple r e a c t i o n models and the Arrhenius-type expression. A f t e r the above d i s c u s s i o n , i t should not be too s u r p r i s i n g to d i s c o v e r that t h i s backward-looking e m p i r i c a l approach has had l i t t l e success. Pyrolyses have been modelled as a f i r s t order system (72), a d i s t r i b u t i o n of p a r a l l e l (73) f i r s t order r e a c t i o n s or (74) s i n g l e r e a c t i o n s followed by a set of consecutive r e a c t i o n s (75). The general form of these studies i s to f i t the rat e of v o l a t i l e s generated (dV/dt) as a f i r s t order r a t e f u n c t i o n of the decomposing components ( v o l a t i l e matter). The expressions used are u s u a l l y p e r t u r b a t i o n s of the expression: Rate = = k(V - V ) dt p where Vp i s the q u a n t i t y of v o l a t i l e s generated a f t e r time t and V 0 0 the v o l a t i l e s generated at some i n f i n i t e time. The value of V 0 0 may be - 47 -greater or less than that obtained from the standardized proximate v o l a t i l e matter procedure, depending on the method employed. The range of a c t i v a t i o n energies obtained by these empirical approaches vary anywhere from 30 to 400 KJ mole - 1 (76). The broad range should not come as a surprise considering the range of energies involved i n the reactions occurring and the span of residence times, coal types and reactors used. These o v e r a l l values f i r s t include the mole weighted average of each type of reaction available at a given temperature. Factored on top of this i s the time weighting of these energies as the system passes through the primary and secondary processes. C a t a l y t i c and mass transfer considerations and experimental error must then be added as w e l l . I t should become apparent that a universal k i n e t i c expression for coal pyrolysis undoubtedly requires a great deal of both the theoretical and empirical input before emerging as a useful modelling t o o l . - 48 -7 - PYROLYSIS SYSTEM DESIGN AND CONSTRUCTION Having discussed the ove r a l l chemical and k i n e t i c considerations i n coal p y r o l y s i s , the next stage i s the design and construction of a working reactor. In this section we discuss the design c r i t e r i a and the design ultimately chosen. The variables chosen as central to this work were the pyrolysis temperature and residence time. Temperatures as high as 1000°C had to be attainable. Coal preheating was to be t o t a l l y avoided. As for residence time, the range from milliseconds to several seconds would be desirable. To this end a simple tubular, lab scale reactor apparatus was chosen. This would allow each of these operating variables to be attainable without major c a p i t a l or operating expense. A lab-scale model would allow a greater scope to be studied i n the limited time a v a i l a b l e . With such a wide range of operating conditions many other variables are affected. The coal feed rate, i n order to maintain a constant loading throughout the time and temperature range, must s i m i l a r l y be capable of accurate and consistent feed rate over three orders of magnitude. The purging gas system and heater must operate to allow the f u l l residence time range while being able to maintain a constant temperature. S i m i l a r l y the quenching and trapping of tars and off-gases must operate at the same e f f i c i e n c y over the entire ranges. Both accurate quantitative feeding and product recovery are e s s e n t i a l . Table 4 outlines the major objectives that were imposed on the system design. The system chosen involves the rapid heating of coal - 49 -Table 4 Design Criteria Feed System Feed rate Feed materials _ 2 10 to several grams per minute ± 10% quantitative mass balance not limited to use with coal introduction of other reactant gases possible Pyrolysis Reactor Temperature Residence time Pressure range Other to 1000°C ± 1% no coal preheating gas preheating to allow flash coal heat-up capable of withstanding thermal stresses milliseconds to several seconds ± 10% high vacuum to slight positive pressure durability ease of fabrication Recovery System Product recovery Tar trapping char hold up in reactor off-gas sampling and retention quantitative tar recovery cryogenic trapping, rapid cool down rate of products thermal and/or chemical quenching possible Product Analysis dynamic or static analysis fractionation of tars, gases, char Operational fast turnaround/cycle time low operational requirements (coal, power, gases) - 50 -p a r t i c l e s (100 ym) by the preheated p y r o l y s i s i n e r t gas. The char i s r e t a i n e d i n the r e a c t o r tube i n l e t while the gases and tars are swept out at the given temperature and vapour residence time i n the secondary r e a c t i o n tube length. Rapid quenching and product recovery f o l l o w immediately. The system i s comparable i n operation to other tubular reactors and s i m i l a r to many s i n g l e stage f l u i d bed r e a c t o r s . Two p a r t i c u l a r design d e t a i l s deserve some d i s c u s s i o n . Residence times i n the m i l l i s e c o n d range are very d i f f i c u l t to a t t a i n . Two routes are p o s s i b l e : i f the pressure i n a system i s reduced, the mean free path (MFP) of the remaining molecules i s increased. Once the pressure i s decreased to the extent that the MFP i s of the same order of magnitude as the reactor dimensions, flow passes from a laminar to a t r a n s i t i o n regime. I f higher vacuum i s a t t a i n e d (< 10 atmosphere), molecular flow occurs. At elevated temperatures molecular v e l o c i t i e s of 100 to 200 m/s are observed; consequently low residence times are a v a i l a b l e (77). This route i s common f o r l a b o r a t o r y scale research on primary p y r o l y s i s pathways where residence times must match the very high chemical k i n e t i c rates i f r e a c t i v e intermediates and products are to be observed. For commercial scale a p p l i c a t i o n s , the low throughput a v a i l a b l e by the high vacuum route i s p r o h i b i t i v e . Instead, high flow r a t e s of d i l u e n t gases are required to f o r c e the vapour through a r e a c t o r of minimum i n t e r n a l volume. Long residence times are a t t a i n e d by reduced d i l u e n t flow f o r atmospheric or p r e s s u r i z e d reactors or by a change i n reactor geometry f o r the high vacuum case. For t h i s study, the small s c a l e used makes e i t h e r route e a s i l y a v a i l a b l e . For - 51 -experiments using a high vacuum a vacuum-tight feeder, reactor and cold trap were designed. Two exchangeable pyrolysis reactors were constructed having different aspect ratios (D/L) to allow a range of low residence times. Both of the reactors operate in the same way, the difference between them being the arrangement of the reactor tube. Both have the same internal area but have different internal volumes. The low volume (short residence time) reactor employs a longer tubular section of smaller diameter than the larger volume (long residence times) reactor. The choice of two exchangeable reactors also allows an investigation into the effects of reactor surface area on product yield and distribution, the reactors having different area to volume rations (A/V). If heterocatalytic, coking or other surface reactions are occurring, the reactors could give different product results, a l l else being equal. The design and construction of each of the systems components is dealt with below. A. Coal Feeder The coal feeder is fashioned after that designed by Scott and Piskorz (78). The feeder body and parts are depicted in Figure 4 and Photographs 1, 2 and 3. The a l l stainless cylindrical body has a removable top sealed with quick f i t flange with Viton o-ring seal. The side taps are welded-in ultra-torr adapters with o-ring seals for the feed tube put-through, nitrogen inlet and manometer. The entire - 52 -Figure 4 COAL FEEDER BODY - 53 -Photograph 1 - Coal feeder body and feeder tubes Photograph 2 - Coal feeder body - disassembled Photograph 4 - Coal feeder - I n s t a l l e d - 55 -const ruct ion was leak tested to 1 0 " 4 t o r r . The c y l i n d r i c a l body i s loaded with approximately 300 g of coal f ines s ized to a mean diameter of 100 vim. By means of a vacuum-tight rotary gland the contents are s t i r r e d at a f ixed ra te . The feeder tubes, passing through the lower part of the feeder along a 1/2 chord and c lose to the impe l l e r , have small holes d r i l l e d i n the i r mid-points. By passing a flow of n i t rogen through a r e s t r i c t i o n and then through the feed tube, coal i s drawn through the hole at a constant rate and conveyed along the tube and i n to the reactor i n a stream of cool n i t rogen . Connection of the feeder to the reactor i s shown in Photograph 4. The top seals on the feeder body are for a pressure equa l i za t ion l i n e and the system's pressure manometer. In the event of plugging in the feed tube or reac tor , the manometer i s connected with a conducting bridge alarm to warn the operator . Eva luat ion of the feeder i s discussed i n Sect ion 10 and Appendix B. B. Coal Conveying Tube Coal conveyed from the feeder passes along the coa l conveying tube and in to the top of the v e r t i c a l l y or iented reac tor . This component i s depicted i n Figure 5 and Photograph 5. The connection between the feeder and the conveying tube i s v i a the u l t r a- to r r tee adapter. The flow of cool n i t rogen can be on e i ther of the coal feed or by-pass mode through this tee. A l l tubing i s 304 s t a in l ess s t ee l and a l l j o i n t s were leak tested. Photograph 5 shows the f lange and coal i n l e t to the reactor (before the tube i n su l a t i on and coo l ing jacket were - 57 -Photograph 6 - P y r o l y s i s r e a c t o r s - 58 -i n s t a l l e d ) . This design allows for continuous and consistent feeding of the coal fines without coal preheating. This and a l l flanges used i n the assembly of the reactor are made of st a i n l e s s s t e e l with Del-Seal knife edge seals and copper gaskets between each set of flanges. C. Pyrolysis Reactors The pyrolysis reactors are shown i n Figures 6 and 7. The entire reactor as shown, i s v e r t i c a l l y suspended i n a c y l i n d r i c a l furnace. The reactors are designed of 304 st a i n l e s s s t e e l with the same knife-edge seal flanges. The reactors have b a s i c a l l y three zones. Coal i s fed down from the converging tube into the primary zone. This section employs a fine screen c y l i n d r i a l d i s t r i b u t o r . Coal i s captured on a screen i n s e r t at the bottom of the primary cylinder as the hot pyrolysis gas, having passed upwards through the preheat c o i l s , permeates through the side of the d i s t r i b u t o r and purges the vapours and tar into the secondary downcomer tuber. The reactor control thermocouple i s situated j u s t under the screen i n s e r t . The two reactors employ the same dimensions in the primary zone, but d i f f e r i n the diameter and length of the downcomer. The elbow at the bottom of the reactor takes the gases out of the hot zone and into the quenching and trapping system. At the e x i t a two-sided flange (not shown) may be added to allow i n - s i t u d e r i v a t i z a t i o n or diluent quenching and ex i t temperature measurement. Downstream, a flange and metal-to-glass adapter leads to the cyogenic trap and recovery t r a i n . The reactors have a volume of 44.9 (short RTD reactor) and 54.9 3 2 cm (long RTD) with the same in t e r n a l surface areas of 142 cm . This - 60 -Figure 7 EXPLODED VIEW: SHORT RTD REACTOR P h o t o g r a p h 7 - P y r o l y s i s f u r n a c e Photograph 8 - T o t a l r e a c t o r system: side view - 63 Photograph 9 - T o t a l r e a c t o r system: f r o n t view, open c o n f i g u r a t i o n - 64 -gives (A^/y) r a t i o s of 3.16 and 2.59 c m - 1 r e s p e c t i v e l y . The t h e o r e t i c a l range of residence times a v a i l a b l e i s 0.01 to 10 s with the flow contro l and vacuum system employed. The actual range obtained was 0.25 to 4 s. the reactor was safe ly operated for extended periods of time at up to 9 0 0 ° C without evidence of leakage or thermal s tress f a i l u r e . D. Furnace The reactors were enclosed i n a s p l i t - s h e l l c y l i n d r i c a l furnace as shown i n Photograph 7. Two furnace sect ions were constructed to accommodate the two s izes of r eac tor . As i t turned out, a s ing le sec t ion (27 cm x 5 cm I . D . ) was s u f f i c i e n t over the en t i re temperature range. The she l l s themselves are res i s tance heaters encased i n a ceramic body and surrounded by F i b r e f r a x i n s u l a t i o n . The heater sec t ion operates at 115 V and 920 W up to a maximum temperature of 1 2 0 0 ° C . E. Control Gas flows were c o n t r o l l e d by using Matheson rotometers with p r e c i s i o n needle va lves . Furnace contro l was accomplished by a thermoelectr ic o n - o f f / p r o p o r t i o n a l c o n t r o l l e r to maintain reactor temperatures at ± 1 ° C . The feeder rotor was rotated at a constant rate by a p r o p o r t i o n a l c o n t r o l l e r with o p t i c a l measurement of the r o t a t i o n a l speed. - 65 -F. Cryogenic Trapping of Tars A var i e t y of cryogenic cold traps and vapor scrubbers were attempted before a r e l i a b l e system was selected. The f i n a l design (Figure 8) employs an i n i t i a l glass cold trap (3 8.5 cm I.D. x 30 cm) immersed i n a dewar containing a s o l i d CO2 s l u r r y i n methanol (-78.5°C). Cellulose f i l t e r tissue packed loosely i n the dra f t tube of this trap greatly assisted i n trapping the tar. The temperature of the gases i n this trap was measured continuously. From this trap, the gas flowed to the base of a 60 cm x 5 cm I.D. v e r t i c a l gas scrubber packed with 5 mm ceramic berl saddles and the extraction l i q u i d , an azeotrope of methanol and dichloromethane. The temperature i n this column i s also measured. Gases from the scrubber proceed to the off-gas manifold system. G. Off-Gas Manifold System F i n a l discharge of the gaseous products from the reactor was v i a a glass manifold shown in Figure 9. The manifold allows for continuous gas sampling or storage of a gas sample under either atmospheric or vacuum conditions. An i n - l i n e c e l l u l o s e f i l t e r was used to detect the bypassing of any tar from the trapping system. The entire assembly i s shown in photographs 8, 9 and 10. The second cold trap shown was ultimately replaced by the previously-described column. Other minor modifications are not shown in these pictures. The f i n a l flowsheet diagram for the system appears i n Figure 9. - 66 -Figure 8 TAR RECOVERY SYSTEM TC 6 Solvent Make-Up TCCI z z r / C e l l u l o s e packing \ To reactor Cryogenic trap L i q u i d e x t r a c t i o n column Overflow d r a i n Drain Figure 9 Rotor PYROLYSIS SYSTEM FLOW SCHEMATIC Feed Switching Valve Feed O Feeder i Bypass 7 7 n Alarm p y r o l y s i s reactor primary zone secondary zone cryogenic ^ trap Optional quench Nitrogen Optional 6 reactant gas Off gas manifold 0 KJ Bleed To Vacuum Gas sampling L i q u i d e x t r a c t i o n column ON - 68 -Photograph 10 - T o t a l r e a c t o r system: close d c o n f i g u r a t i o n - 69 -8. REACTOR OPERATION The f i n a l procedural and a n a l y t i c a l operations used for each run are depicted i n Figure 17. Each run was c a r r i e d out i n the same manner and documented continuously during operat ion . Before a p a r t i c u l a r run was s t a r t e d , the appropriate flow and feeding condit ions were se l ec ted . For a des ired residence time (T ) at a given temperature ( T ) , the flow ( F Q ) of n i trogen was ca lcu la ted according to „ , 2 9 8 . 1 6 W V N , . 3 , - K F q = ( ~—)(—) .60 ( i n cm min ) For most runs, the s o l i d s loading (G, mass of coal s o l i d s per u n i t volume of expanded p y r o l y s i s gas) was kept constant. The des ired feed rate of coal (M) was then determined by: * " F o ( 2 9 0 6 > ' G The feed rate i s obtained by s e l e c t i n g the appropriate feed tube (by hole s ize) to use with the coal feeder. Each tube of a d i f f e r e n t hole s i ze was c a l i b r a t e d and feed rates corre la ted as a funct ion of flow through that tube (Fpp) as: M = a + b In F „ n - 70 -Once the tube i s s e l e c t e d , the flow rate i s determined from the above equation from which the flow of gas through the preheater c o i l s (F^-jO i s calcuated by: F = F - F HT o FD I t i s d e s i r a b l e that the p r o p o r t i o n of unheated ni t r o g e n ( F F D ) to heated nitrogen be kept s m a l l . The percentage unheated gas i s c a l c u l a t e d as: FFD X 1 0 0 FFD % unheated = -= — — ; — = - — x 100% FD HT F o This value was kept under 10% by s e l e c t i n g the appropriate feed tube that could r e l i a b l y maintain the desired coal feed r a t e . The coal chosen f o r t h i s study was a B r i t i s h Columbian coal from the Balmer f i e l d . This c o a l was chosen f o r a v a r i e t y of reasons. Previous work on t h i s coal at UBC by i n v e s t i g a t o r s i n a v a r i e t y of c o a l u t i l i z a t i o n methods research should allow comparison of r e s u l t s . The coal was a l s o r e a d i l y a v a i l a b l e i n a useable form. The low s w e l l i n g index of t h i s coal was a l s o d e s i r a b l e f o r use i n the feed system employed. The proximate and u l t i m a t e a n a l y s i s (74) of the coal used i s given i n Table 5. The c o a l was crushed i n a tumbling b a l l m i l l and s i z e d to obtain the -120 +200 U.S. standard sieve s i z e f r a c t i o n - 71 -Table 5 Proximate and Ultimate A n a l y s i s of Balmer Coal Proximate A n a l y s i s Ash 11.59% Moisture 1.20% V o l a t i l e matter 20.27% Fixed carbon 67.45% Sulphur 0.22% C a l o r i f i c value 31,473 KJ/kg Ultimate A n a l y s i s (%) Ash 11.59 Carbon 78.36 Hydrogen 4.14 Nitrogen 1.00 Sulphur 0.22 Oxygen (by d i f f e r e n c e ) 4.64 - 72 -( p a r t i c l e diameters: mean 100 ym, minimum: 74 ym, maximum: 149 ym). This f r a c t i o n was then re-sieved and stored under nitrogen u n t i l use. A series of experiments was performed to study the e f f e c t s of coal loading, temperature, time and reactor geometry on operational y i e l d and c h a r a c t e r i s t i c s of the tars produced. For a given run the y i e l d of tar was determined by the weight of tar recovered per u n i t weight of coal on a dry ash free basis. Following c h a r a c t e r i z a t i o n , as described i n Section 9 the product d i s t r i b u t i o n and net product y i e l d s for c e r t a i n families of hydrocarbons were determined. - 73 -9. ANALYTICAL METHODS Complete evaluation of coal pyrolysis products as a p o t e n t i a l f u e l or petrochemical feedstock requires information on the composition of the product. This information i s most appropriately obtained by applying a n a l y t i c a l procedures developed for conventional petroleum products. In this work the primary methods for tar analysis include solvent extraction, solvent s o l u b i l i t y , preparative l i q u i d - s o l i d p a r t i t i o n chromatography, high performance l i q u i d chromatography (HPLC), and in f r a r e d spectroscopy (IR). P y r o l y s i s gas analysis was attempted using gas chromatography (GC). The coals used and chars generated were also analysed by thermal gravimetric analysis (TGA). The analysis presented i s by no means exhaustive, but does provide some i n d i c a t i o n of the s u i t a b i l i t y of coal pyrolysis as a means of obtaining petroleum a l t e r n a t i v e s . Solvent E x t r a c t i o n and Solvent Fractionation Tars generated i n the reactor are quenched and trapped in the recovery system. To determine how best to q u a n t i t a t i v e l y recover the tars from the trap a series of experiments was conducted in solvent extraction and s o l u b i l i t y . F i r s t l y , c r i t e r i a were established to designate a suitable tar solvent. This solvent must be capable of d i s s o l v i n g a l l of the tar samples encountered and be non-reactive with the tar components. As well, a high vapour pressure was desired so that - 74 -the solvent could be easily evaporated without stripping any of the lighter components from the coal liquid. With these cri t e r i a several pure solvents, binary mixtures, binary azeotropes and ternary azeotropes were obtained and tar solubility tests performed. From this work the binary azeotrope consisting of 92.7 weight percent dichloromethane (methylene chloride, CH2CI2) and 7.3 weight percent methanol was chosen. This mixture has a normal boiling point of 37.8°C (CH3OH 64.5°C, CH2CI2 39.7°C), an estimaed Hildebrand solubility parameter of 6 = 10.22 and a remarkable abi l i t y to dissolve a l l of the tars encountered even at extremely high concentrations. In a l l subsequent work the azeotrope was d i s t i l l e d in batch form from laboratory grade solvents and stored over 4 A molecular sieves un t i l use. This same solvent was used in the packed column extraction stage of the pyrolyser recovery system. As a matter of definition, coal tar liquids are taken to be that fraction of the total tar generated that is soluble in the above solvent system. The next stage of the tar fractionation involves precipitation of the highly functionalized, high molecular weight and very non-volatile asphaltenes. In this work, asphaltenes are definied as that fraction by weight of the total tar that is not soluble in hexane at room temperature and at a solvent-to-tar weight ratio of approximately 1000:1. It is recognized that other normal alkane solvents or solvent systems have been used in petroleum analysis for asphaltenes, a l l giving similar results. To avoid confusion these solids w i l l hereafter be - 75 -r e f e r r e d to as hexane i n s o l u b l e s . The remainder of the t a r , the hexane s o l u b l e s , are then s u i t a b l e f o r the next main a n a l y t i c a l procedure, high performance l i q u i d chromatography. High Performance L i q u i d Chromatography (HPLC) Methods The general technique of l i q u i d - s o l i d p a r t i t i o n chromatography i s w e l l documented (80). This procedure i s commonly used f o r the se p a r a t i o n of homologs, isomers and f u n c t i o n a l l y d i f f e r e n t molecules of l i q u i d s and/or n o n - v o l a t i l e s o l u t e s , according to t h e i r d i f f e r i n g a f f i n i t y f o r the column packing ( s t a t i o n a r y phase) and the e l u t i n g solvent (mobile phase). The same p r i n c i p l e s apply to both bulk s c a l e t r a d i t i o n a l preparative column chromatography and the more recent microscale separations using HPLC. The l a t t e r technique has enjoyed a recent increase i n p o p u l a r i t y due to advances i n sample d e t e c t i o n , data a c q u i s i t i o n , system design and c o n t r o l . With t h i s , HPLC can o f f e r a number of advantages over t r a d i t i o n a l l i q u i d chromatography. These in c l u d e increased speed and r e s o l u t i o n of s e p a r a t i o n , higher s e n s i t i v i t y , reusable packed columns and a l e s s l a b o r i o u s method. In t h i s work both preparative and high performance p a r t i t i o n l i q u i d chromatography were c a r r i e d out to evaluate c o a l tar components. Some d i s c u s s i o n of both methods i s r e q u i r e d . From a survey of the a v a i l a b l e recent l i t e r a t u r e one may f i n d a broad range of HPLC techniques amenable to petroleum c h a r a c t e r i z a t i o n . The most notable appear i n Table 6. Because of the newness of the technique, no p a r t i c u l a r method has assumed the p o s i t i o n of a Table 6 HPLC Methoda for Hydrocarbon Analysis Separation Families or Compounds Column Packing* Mobile Phase Me thod Detectors UV IR Analysis Time Relative Resolution Reference Hydrocarbon Group Types S.A.R.A. NH 2 hexane BF - Yes 14 min very good 81* Hydrocarbon Group Types S.A.R.A. S i l i c a hexane BF 200 Yes 65 good 82 Hydrocarbon Group Types P.O.S.A. CN hexane BF 200 Yes 60 good 83* Hydrocarbon Group Types P.O.S.A. CN and S i l i c a hexane BF 200 Yes 65 good 84 Coal Tar P.O.S.A. NH 2 heptane - 254 Yes - good 85 Asphalts P.O.S.A. NH2 hexane BF 254 Yes - f a i r 86 Coal Liquids Saturates 1,2,3 ring aromatics NH 2 hexane — Yes 8 good 87* 190-360°C O.S.A. Pe t r o l . Fraction Alumina hexane BF - Yes 25 f a i r 88 Crude O i l s S.A. S i l i c a hexane BF - Yes - good 89 O i l s P.O.S.A. Kaolin & S i l i c a hexane - 254 Yes 25 good 90 Coal Liquids S.A S i l i c a hexane + CH2CL2 - 254 Yes 10-25 good 91 S.A.R.A. = Saturates, aromatics, resins, asphaltenes P.O.S.A. = polars, o l e f i n s , saturates, aromatics BF = backflush method to elute polars, resins or possibly aromatics NH2 =• chemically bonded propylamine s i l i c a CN = chemically bonded cyanopropyl s i l i c a * = promising techniques In the current work - 77 -standardized a n a l y s i s . Journals of a n a l y t i c a l and t e s t i n g procedures have not yet issued g u i d e l i n e s . With t h i s i n mind, work was peformed to assess the a p p l i c a b i l i t y of a number of the more promising methods f o r f r a c t i o n a t i o n of c o a l t a r s . The o b j e c t i v e of these tes t s was to f i n d a widely a p p l i c a b l e method g i v i n g the best component r e s o l u t i o n p o s s i b l e i n a reasonable period of time. S t a r t i n g w i t h such a non-homogenous mixture as a c o a l t a r , i t was decided that compound r e s o l u t i o n would not be p o s s i b l e ; r a t h e r a broad based "group type" a n a l y s i s would be best. S o - c a l l e d SARA ( s a t u r a t e s , aromatics, r e s i n s , asphaltenes) and POSA ( p o l a r s , o l e f i n s , s a t u r a t e s , aromatics) separations have been i n v e s t i g a t e d i n the HPLC mode and are f a i r l y w e l l e s t a b l i s h e d as preparative methods (92, 93). Bulk SARA f r a c t i o n a t i o n s allow a determination of the weight percent of each group type. HPLC, which does not operate on a mass-sensitive d e t e c t i o n mode, cannot d i r e c t l y y i e l d the same in f o r m a t i o n . A c r o s s - c o r r e l a t i o n of the separations would be r e q u i r e d i f HPLC r e s u l t s were to be used f o r weight percent c l a s s i f i c a t i o n s . This r e q u i r e s any HPLC method to be amenable to c a l i b r a t i o n and to be r e p r o d u c i b l e . Hydrocarbon group type analyses that hold promise are i n d i c a t e d w i t h an a s t e r i s k i n Table 6. Note that most methods in v o l v e the use of hexane as mobile phase and polar m i c r o p a r t i c u l a t e a n a l y t i c a l packings as s t a t i o n a r y phase. Detection of component peaks as they e l u t e i s c a r r i e d out by e i t h e r or both of a r e f r a c t i v e index d i f f e r e n c e or u l t r a v i o l e t absorbance det e c t o r s . There are two p r i n c i p a l d i f f e r e n c e s i n the - 78 -methods given. F i r s t l y , d i f f e r e n c e s i n the p o l a r i t y of the column packing a f f e c t the separation. Highly polar a c t i v a t e d s i l i c a and alumina columns provide e x c e l l e n t r e s o l u t i o n of components but t y p i c a l l y r e q u i r e much longer e l u t i o n times. As column p o l a r i t y i s decreased by d e r i v a t i z a t i o n to chemically bonded propyl amine (NH 2) or l e s s polar propyl n i t r i l e (CN), the r e s o l u t i o n i s compromised f o r increased speed of a n a l y s i s . Next, the methods d i f f e r i n t h e i r treatment of the ( r e s i n ) polar compounds. Because of t h e i r strong r e t e n t i o n on the column packings, polars may be determined e i t h e r by d i f f e r e n c e once the aromatics, o l e f i n s and saturates have f u l l y e l uted or may be backflushed from the top of the column by flow r e v e r s a l . The technique of column ba c k f l u s h i n g i s depicted i n Figure 10. For a n a l y t i c a l completeness t h i s l a t t e r method i s preferred since i t allows f o r d e t e c t i o n of a l l components and reduces the e r r o r s inherent i n the p o l a r s - b y - d i f f e r e n c e method. F i n a l l y , depending on the nature of the sample, the mobile phase used and the column packing, i t may not always be p o s s i b l e to q u a n t i t a t i v e l y recover a l l of the sample from the column. I t has been observed (94) that high p o l a r i t y packings such as FeCl3 treated s i l i c a or a c t i v a t e d s i l i c a may i r r e v e r s i b l y adsorb some resinous or polar compounds and r e s u l t i n l e s s than complete sample recovery. With t h i s i n mind, c a l i b r a t i o n of s i l i c a or alumina HPLC columns becomes somewhat suspect. This adsorption problem may be a l l e v i a t e d by use of e i t h e r a l e s s polar packing or a more polar mobile phase. S i l i c a and alumina columns also s u f f e r due to f a c i l e d e a c t i v a t i o n by contaminant water. A c t i v a t i o n by heating under dry purge i s then required f o l l o w i n g which - 79 -retention time and c a l i b r a t i o n c h a r a c t e r i s t i c s must be redone. Such i s not the case for n i t r i l e and amine columns which t y p i c a l l y exhibit much longer useful l i v e s . With the above consideration i n mind, t r i a l s were performed to develop a rapid SARA method for use on the coal tar hexane solubles. Comparisons were made i n i t i a l l y of two columns, a s i l i c a chemically bonded propyl n i t r i l e and a propyl amine column (see Table 6). In both cases o l e f i n s and saturates coelute f i r s t and may be detected by refractome try. Following t h i s , aromatics elute p a r t i a l l y resolved and are detected by u l t r a v i o l e t absorbance at 254 nm. Polars are then backflushed from the top of the column and elute as a single peak, detected by u l t r a v i o l e t absorbance. As shown by Figure 11, the more polar NH2~10 column re s u l t s i n improved baseline-resolved aromatic peaks, but at the expense of a doubled run time in the analysis of a t y p i c a l coal tar sample. This improved resolution manifests i t s e l f i n a s i g n i f i a n t l y lower run-to-run r e p r o d u c i b i l i t y e r r o r . With the CN-10 column, integration of the aromatic peak signal agrees within ± 3% on consecutive analysis of the same sample, whereas the error using the NH2-10 column i s reduced to ± 0.25%. The 20 minute difference in run time was a small price to pay for such a dramatic improvement i n performance. A l l subsequent work was therefore carried out using an NH2-10 column. Flow optimization was carried out by v i s u a l l y comparing retention times and resolution of representative coal tar solubles. With this a flow of 2.0 mL m i n - 1 ( r e s u l t i n g in a column pressure of - 80 -Figure 10 HPLC BACKFLUSH OPERATION N O R M A L M O D E B A C K F L U S H M O D E 1 0 /A S a m p l i n g L o o p Si« P o r t S a m p l i n g V a l v a O a c k f l u a h V a l v a Figure 11 HPLC - COLUMN PERFORMANCE COMPARISON 2 mL m i n - 1 hexane, UV 254 NM, 30°C, Forestburg coal t a r (95) - 81 -20-22 bar) was found most suitable. Backflushing of the polar group was then i n i t i a t e d at this flow rate after 18 minutes. Temperature programming, column modifiers or wavelength adjustments were not carried out to further refine the method. Standards prepared to check for degradation of the columns were prepared i n i t i a l l y , but deteriorated before use. No apparent loss i n re s o l u t i o n or performance was observed over the course of this study. The same procedure as outlined above was used throughout for the analysis of a l l coal tars. Table 7 gives a summary of the relevant HPLC conditions. HPCL Retention Times and Compound Studies A t y p i c a l chromatogram using the above method on a hexane soluble f r a c t i o n of a tar generated i n this study appears as Figure 12. Saturates and o l e f i n s elute and appear as a very small peak i n the r e f r a c t i v e index response. The poor resolution and universal means of detection of these components does not allow their chemical i d e n t i f i c a t i o n . Further separations made by s p l i t t i n g this f r a c t i o n off to a more polar s i l i c a column as recommended by Alfredson (96) are of marginal use in this case: the quantity of sample being so small and the detector noise and d r i f t so large any further c l a s s i f i c a t i o n of this f r a c t i o n i s not p r a c t i c a l by l i q u i d chromatography. Compound i d e n t i f i c a t i o n could best be made by i s o l a t i n g this f r a c t i o n and subjecting i t to gas chromatography (GC), mass spectroscopy (MS) or combined GC-MS. This procedure has been carried out by numerous other researchers (97,98). - 82 -Table 7 Comparison of HPLC Colums for Fractionation of Hexane Solubles of Coal Tar Apparatus: Sample: Temperature: Valve switching: Refractometer: U l t r a v i o l e t absorbance detector: NH2-IO column: Varian V i s t a 5000 LC CDS 401 D i g i t a l data a c q u i s i t i o n 10 pL loop 30°C Pneumatically actuated i n j e c t o r and 6 port flow switching valve Varian RI-3 10 pL c e l l volume 200 x 10" 6 r e f r a c t i v e index units f u l l scale Varian UV-100 5 nm spectral bandwidth 254 nm, 0.05 AU/MV 4 mm I.D. x 30.0 cm length 10 ym propyl amine bonded s i l i c a assurance report: 4220 t h e o r e t i c a l plates capacity factor K' = 9.1 CN-10 column: 4 mm I.D. x 30.0 cm length 10 pm propyl n i t r i l e bonded s i l i c a 9746 theoretical plates, k' =3.4 - 83 -The aromatic f r a c t i o n , on the other hand, i s reasonably w e l l resolved i n t o broad peaks. These bands are c h a r a c t e r i s t i c f o r a low r e s o l u t i o n technique such as HPLC and represent c l a s s e s of compounds possessing s i m i l a r o v e r a l l s t r u c t u r e . For example, a l k y l s u b s t i t u t e d aromatics e l u t e at nearly the same time as the parent unsubstituted aromatic. The broad peaks seen r e s u l t from t h i s c o - e l u t i o n of f u n c t i o n a l l y s i m i l a r m a t e r i a l s . I t i s p o s s i b l e to make t e n t a t i v e i d e n t i f i c a t i o n s of component peaks by comparing the r e t e n t i o n times of the pure compounds. Once a peak has been i d e n t i f i e d i n t h i s way, i t i s a l s o p o s s i b l e to obtain some i n d i c a t i o n of the r e l a t i o n s h i p between the detector response and concentration. According to the Beer-Lambert law (Ajj = eCL) the absorption measured i s a f u n c t i o n of both the amount of substance present and i t s e x t i n c t i o n c o e f f i c i e n t . I t i s important to keep t h i s point i n mind when converting detector response to mass. By i n j e c t i n g a sample of known concentration of a given aromatic d i s s o l v e d i n hexane using the method o u t l i n e d p r e v i o u s l y , the r e s u l t s i n Table 8 are obtained. Tentative i d e n t i f i c a t i o n of c o a l aromatics may be made by comparison to Figure 12 and are marked with a a s t e r i s k i n Table 8. P o l a r compounds, being eluted as a s i n g l e peak are obviously unresolved. Table 8 includes a wide range of compounds that e l u t e as polars using the method o u t l i n e d . This l i s t includes a v a r i e t y of h e t e r o c y c l i c n i t r o g e n , oxygen and s u l f u r compounds. As w e l l , p r e l i m i n a r y s t u d i e s suggest that very large condensed aromatic s t r u c t u r e s (dibenzoanthracene, benzoperylene) are l i k e l y included as Table 8: HPLC Retention Times and Detector Response F a c t o r s * Compound Retention Time, Min. Response** Factor x 10" e(99) A M A X ( l U B ) Solvent Saturates Pentane Dodecane 2.93 2.82 1.02 1.12 (RI detector) (RI detector) Aromatics Benzene Toluene Xylene T e t r a l i n Napthalene Fluorene Phenanthrene Anthracene 1 ,1'-binapthyl Pyrene Triphenylene Chrysene 3.20 3.21 3.21 3.28 4.02 4.83 5.84 5.85 6.66 7.27 10.56 11.02 13.39 1.96 x 10' 8.54 x 10 : 1.13 x i o : 2.47 x 10' 1.70 x 10 : 4.80 x 10: 3.12 x i o ; 2.70 x 10' 163 3470 19953 64600 20000 11800 92000 254 258 266 250 250 251 257 MeOH MeOH CC11+ MeOH MeOH .MeOH MeOH i oo -o-i P o l a r s P-benzoquinone 34 Anthraquinone 34 2,4-dimethyl phenol 34 8-Napthol 34 Fluoren-9-one 34 Dibenzofuran 34 Carbazole 34 P y r i d i n e 34 1.25 x 10' 2.44 x 10 1.18 x 10 1.79 x 10 20400 19900 250 256 MeOH MeOH *See Appendix C " T r i v i a l Names and Structures of Polynuclear Aromatics and Other Coal Tar Components. detector counts **Response Factor = — — 7 — — T - K v concentration (g mL ; - 85 -CHART SPEED 1.8 CM--I1IN A T T E N : 32 ZERO : 10X 1 M1N/TICK Figure 12 COAL TAR CHROMATOGRAM CHART SPEED 1.6 CM/FUN A T T E N : 8 2 E R 0 : 1 6 * 1 MIN/TICK Refractive index detector response UV absorption detector response Conditions: 2.0 mL min - 1, NH2-10 column, 30°C, 254 NM (32X), ARIs 200 x 10" 6 (8X) - 86 -polars. The classes of compounds chosen as representative polars have been taken from a variety of sources in the l i t e r a t u r e that have aimed s p e c i f i c a l l y at i d e n t i f i c a t i o n of chemical constituents (100,101,102). Again, use of GC, MS, and GC-MS techniques in tandem with the HPLC work i n this study would be helpful in confirming i d e n t i f i c a t i o n of component f r a c t i o n s . HPLC Calibration and Cross-Correlation Table 8 depicts c l e a r l y the major obstacle i n the use of HPLC as a routine quantitative technique. If one requires the standard weight percent breakdown of a given petroleum or tar i t i s necessary to convert back from detector response to concentration. Since each compound present shows a d i f f e r e n t response factor, this i s by no means an easy task. In high resolution methods, such as GC, a compound-by-compound response factor c a l i b r a t i o n i s possible and frequently performed. With a low r e s o l u t i o n method such as HPLC this technique i s not generally a v a i l a b l e . I t i s therefore necessary to do c a l i b r a t i o n by c r o s s - c o r r e l a t i o n with t r a d i t i o n a l bulk scale preparative chromatography. I t i s advisable to caution at this point that combinations of two low resolution techniques such as t h i s , are l i k e l y to generate only quasi-quantitative r e s u l t s at best. We are faced by this l i m i t a t i o n in the current work. Preparative f r a c t i o n a t i o n of coal tar hexane solubles was caried out according to the method outlined by Hirsch et a l . of the U.S. Bureau of Mines (103). A scaled-down version using a column of 80 cm length - 87 -with 2 cm internal diameter was employed. The bottom 28 cm of the column was packed with fully activated Alumina (Fischer Scientific, 160 g, neutral 80-200 mesh, 20 h. at 300°C). On top of this a 20 cm plug of activated s i l i c a was added (Fischer Scientific, 50 g, 230-400 mesh, grade 60, 20 h. at 300°C). The column was rinsed thoroughly with dried (4A molecular seive) HPLC grade hexane before use. In order to best ensure a representative separation, the sample chosen for the preparative scale chromatogram was obtained by mixing a l l of the hexane solubles generated in the course of the pyrolysis experiments. These samples were stored diluted in hexane and frequently purged with nitrogen. Before the bulk separation was performed the sample was filt e r e d , then evaporated under vacuum (P » 10" torr) at no greater than 50°C for less than 30 minutes to remove a l l solvents. A weighed sample was taken, diluted 1:10 in hexane and suspended in solution by immersion in an ultrasonic bath before being applied to the top of the column in a narrow band. Elution of the sample was carried out by a series of step gradients from pure hexane (300 mL) to 95% hexane in toluene (300 mL) to 30% hexane in toluene (3000 mL). Fractions were collected in 100 mL aliquots and analyzed by the HPLC-POSA routine described above. Use of a long wavelength ultraviolet lamp allowed for easy visualization of the column performance owing to the b r i l l i a n t fluorescence of the aromatic bands. Figure 13 depicts the progress of the analysis as seen from the HPLC results. Once the cut point between aromatics and resins/polars was reached, the column was drained, blown out and extracted with methanol, methylene chloride and tetrahydrofuran - 88 -Figure 12 TOTAL TAR SPECTRUM 32X r i i i i i i i i i i i i i i i i i i 1111111111 11111111 n 1111 II 1111 Aromatics Polars —» Figure 13 CROSS-CORRELATION RESULTS: SPECTRA OF BULK FRACTIONATION CUTS. - 89 -to a clear wash. The combination of rinses from this were evaporated under vacuum at no more than 60°C in a tared flask, to give the t o t a l weight of polars. S i m i l a r l y a l l the aromatics aliquots (to number 26) were combined, evaporated and weighed. The combined total masses recovered was 104% of the o r i g i n a l sample. The d i s t r i b u t i o n was 55.7 weight percent polars and 44.3 weight percent aromatics. A sample of the same tar solubles was subjected to HPLC f r a c t i o n a t i o n . By r a t i o of the detector responses for polars and the sum of a l l aromatics, the d i s t r i b u t i o n i s 54.5% polar and 45.4% aromatics. Within the range of experimental error then, the d i s t r i b u t i o n as given by HPLC detector response i s the same as that given by the weight percent bulk separation. Ideally, a number of cross-correlation experiments should have been performed at a va r i e t y of aromatics-to-polars r a t i o s to check this r e s u l t . Due to"time and sample constraints, this was not c a r r i e d out. In a l l subsequent discussions, POSA d i s t r i b u t i o n s are given i n weight percents by d i r e c t use of the HPLC detector response r a t i o s . Infrared Spectroscopy A very l i m i t e d study of the hexane insoluble f r a c t i o n of coal tar was carried out. A l l samples were dried under vacuum at 40°C for 8 hours before use. Spectra were taken using a Perkin Elmer model 710A as a Nujol n u l l between sodium chloride plates. - 90 -Gas Chromatography (GC) Off-gases generated in the course of pyrolysis were collected by way of the reactor gas manifold and sampling cylinder (see Figure 9). Gas analysis was then performed using a Varian Model 6000 gas chromatograph. Gas is i n i t i a l l y injected onto a Porapak-N and a molecular sieve (13 x) column in series for the f i r s t 1.75 minutes and the last 5 minutes of the 12 minute program. Detection is by both thermal conductivity (He) and flame ionization to obtain H2, C02, ethylene, ethane, acetylene oxygen, nitrogen, CO, and methane in order of elution. Calibration was carried out against a prepared gas standard. Unfortunately, when pyrolysis experiments were in progress the above system was undergoing repairs. An attempt was made to store gas samples in 100 mL precision sampling gas syringes. However, leakage was found to have occurred to a significant extent in only two days as evidenced by the presence of oxygen in the chromatograms. Hence quantitaive gas analysis was not possible. H2, C02, ethylene, ethane, acetylene, CO and methane were detected in the samples examined. Some developmental work on the use of a fused s i l i c a capillary column for tar analysis was also carried out. Unfortunately, this was not completed sufficiently by the time of writing. Thermal Gravimetric Analysis (TGA) Both coals and pyrolysis chars were subjected to thermal gravimetric analysis (TGA). A Perkin Elmer TGS-2 furnace and controller - 91 -was used i n conjunction with an integrated data a c q u i s i t i o n system. The TGA program involved measurement of percent weight loss over the temperature-time p r o f i l e as depicted in Figure 14. Solids were prepared by crushing gently and sieve s i z i n g to obtain the -120 +200 mesh (dp =• 100 um) size f r a c t i o n . The p a r t i c l e s were held under vacuum for one hour, then immediately purged with nitrogen for at le a s t another hour. A l l were stored i n septa sealed v i a l s and again purged with nitrogen for 15 minutes before use. Samples weighing in the order of 8 mg were transferred quickly into a platinum microcrucible and placed in 3 1 the TGA balance and furnace assembly. After some time i n a 1 cm s~ stream of nitrogen, the temperature program was started and data acquired. A l l TGA spectra were analysed for moisture and residual v o l a t i l e s . I t i s believed that char moisture i s picked up from the atmosphere during storage. Residual v o l a t i l e s are expressed as the additional percent weight loss from the tangent after elimination of moisture to the tangent l i n e defining the end of the d e v o l a t a l i z a t i o n as shown in Figures 15 and 16. I t was found i n the course of study that sample mass (between 3 and 15 mg) had very l i t t l e e f f e c t on the analysis. P a r t i c l e s i z e , purge rate and heating rates were not investigated. Of c r i t i c a l importance, though, was the removal of even trace amounts of oxygen in either the sample or apparatus. This was evidenced i n a non-reproducible r e s u l t or the long, t a i l i n g appearance of the weight loss p l o t . The complete coal, char and tar l i q u i d s recovery, work-up and analysis i s presented i n Figure 17. I t i s important to stress that the - 92 -(J Ul tr 3 t— < U I Q. Ul Figure 14 TGA TEMPERATURE: PROGRAM TIME/MIN-100.00 X U Ul 80.00 COAL:SAMP3 09. BIX WT. X CHANGE. 18.02 WTi 10.3221 mg 81. 782 Figure 15 TGA RESULTS: BALMER COAL 100.00 CD »—< Ul :* 90.00 RUN44t SAMP2 , 74X WT. X CHANGE. 5.42 WTi 11.7426 mg 04. 3IX Figure 16 TGA RESULTS: CHAR #44 - 93 -r e s u l t s obtained are only as good as the methods used to obtain them. E f f o r t s were made to adhere r i g i d l y to this standardized flow diagram for a l l p y r o l y s i s runs performed. It i s believed that the procedures developed give a r e a l i s t i c impression of the quantitative and q u a l i t a t i v e nature of the pyrolysis products generated. - 94 -Part I Figure 17 Pyrolys is Preparation and Analysis Flow Diagram PRELIMINARY Calculate: - required feed rate - temperature - residence times - n2 flow rates REACTOR clean and dry reactor J check copper gasket seals \ assemble reactor primary zone i assemble remaining parts 4 leak tes t a l ign assembly in furnace \ attach N 2 flow l ines begin N 2 purge attach cryogenic trap J leak test a l l parts i plug in thermocouple leads and power l ines i adjust N 2 flow to calculated leve ls i star t furnace heating i when set point temperature is reached 4 s ta r t rotor on feeder, — set to desired RPM FEEDER RECOVERY SYSTEM choose feed tube i n s t a l l tube I preliminary feed rate check load coal into feeder record feeder weight I attach to framework and motor attach N 2 flow l ines and manometer {' attach to reactor feed l ine c assemble manifold, vent and vacuum k assemble l i qu id extract ion column rinse column repeatedly with Methanol/CH 2 Cl 2 i n s t a l l ce l lu lose packing i> attach column to cryogenic trap t leak tes t begin cool down of dewars and column record measurements for time t = 0 temperatures (reactor, furnace, column, trap) - 95 -Figure 17 continued - Part 2 START RUN switch bypass valve to begin coal feed \f check for proper operation add solvent to column If required t maintain overall pressure drop to < 7 cm Hg l< monitor temperature at uniform Intervals L-c o l l e c t gas sample at mid-point of total run time TERMINATE RUN turn coal feed off V maintain bypass N 2 flow lr maintain reactor at set point temperature for < 15 minutes V reduce flow to minimum reduce temperature of reactor Ir proceed to complete shutdown 1 PRODUCT RECOVERY recovery char from primary zone V weigh recovered char \ weigh feeder ( f i n a l ) V go to TGA char analysis disassemble reactor check for improper operating drain extraction column V rinse exhaustively V add to get total tar recovered V go to tar and gas fractionation and analysis \ rinse a l l reactor parts rinse exhaustively with aid of ultrasonic bath - 96 -Figure 17 continued - Part 3 FRACTIONATION AND ANALYSIS tota l tar recovered in azeotrope i vacuum f i l t r a t i o n to remove char I r inse f i l t r a t i o n unit \ recover so lu t ion . Make up tota l volume to 250.0 mL quant i tat ive ly transfer two (2) 100.00 mL of solut ion to weighed round bottom f lasks \f rotary evaporate to. dryness at < 60°C under vacuum Ir record weight change |r add hexane to give 1000:1 solvent/tar ra t io RESULTS X tota l tar y i e ld = A weight x 2 5 Q * coal feed weight ( in dupl icate) immerse in u l t rasonic bath I weigh mi l l ipore f i l t r a t i o n disk ^ f i l e r and recover disk and so lut ion \f dry f i l t e r d isk , record weight change Ir hexane solubles to HPLC • analysis i determine weight % polars , _ aromatics, saturates v u J i i. n A weight f i l t e r I A r i % hexane insolubles = =—A :—r— x 100 total tar weight ( in duplicate) Production d i s t r i bu t i on ; wt. % aromatics, wt. %. polars Net product y ie lds -(wt. % product d i s t r ibut ion ) x (wt. % tota l tar y ie ld) x 1 0 - 2 ( a l l in duplicate) - 97 -1 0 . EXPERIMENTAL RESULTS AND DISCUSSION Following the construction and testing of the apparatus, experiments were performed to investigate the e f f e c t s of coal loading, temperature, residence time, reactor geometry and trace oxygen on the y i e l d of tar and tar components. In this section the results of these experiments are presented. The observations are discussed and compared i n r e l a t i o n to the chemical and k i n e t i c considerations given e a r l i e r . A. Coal Feeder Performance A detailed discussion of the performance of the coal feeder i s given i n Appendix B. Only the f i n a l r esults are discussed b r i e f l y here. Coal feed rates from 0.03 to greater than 3.0 g/min were obtained using the design shown. The feeding rate was observed to be very consistent over time at a l l feed rates. The coal density in the converying stream was uniform over the entire range. The feed rate was observed to be dependent on the rotor configuration, rotor speed, tube angle, mass of coal i n the feeder, tube hole size and the nitrogen flow rate through the feed tube. The feed rate (M) was found to vary with flow rate (F) at a constant rotor speed according to the equation M = a log F + b where a, b are constant. When a desired feed rate of coal i s required, the appropriate tube could be e a s i l y placed through the feeder and the flow rate adjusted according to the loading and percent unheated gas - 98 -c r i t e r i a discussed e a r i e r . So long as the pressure i n the reactor was kept near atmospheric, the actual feed rate was within 5% of the calculated feed rate. In the course of the pyr o l y s i s runs, the mean feed rate deviation encountered was 18% from the desired value. The deviations were both p o s i t i v e and negative with the sum of the deviations being 4.5% lower than the desired feed rate t o t a l for a l l runs. Considering the very low feed rates used, the pressure changes i s the course of a run and the systematic er r o r s , such a deviation i s to l e r a b l e . Greater feed rate control would be possible with better pressure c o n t r o l . B. System Temperature P r o f i l e s The temperature at various points along the system was measured during a l l of the runs. The temperature p r o f i l e i n the reactor i t s e l f appears i n Figure 18. This graph depicts the r a t i o of the actual temperature (T) to the set point temperature (T p) plotted against the f r a c t i o n of the gas residence time that has passed at the point the temperature was measured. As can be seen, the temperature f a l l s off quite sharply as the gas enters the secondary and e x i t tubes of the reactor. The flow of hot gases down the tube i s e n t i r e l y laminar over the e n t i r e range (see Figure 18)which should a s s i s t i n keeping the heat loss to a minimum. The temperature decrease i s more pronounced i f the downcomer tube i s not insulated, and so a l l work was done by keeping the tube jacketed by multilayer i n s u l a t i o n . The residence time i n this reactor cannot be equated with the time-at-temperature, and comparisons to other systems must therefore include this f a c t o r . - 99 -Figure 18 REACTOR TEflPERflTURE PROFILES Residence Time 0.08 0.16 0.2A 0.32 0.40 OAS 0.56 0.64 0.7 2 0.80 0.92 1.00 FRACTION OF RESIDENCE TlflE PASSED - 100 -Once a run i s i n i t i a t e d and coal feeding begins, the temperature generally r i s e s a few degrees from the set point but returns to the desired temperature within a few minutes. This e f f e c t i s believed to r e s u l t from minute quantities of oxygen remaining i n the system and within the coal i t s e l f causing s l i g h t combustion. If the temperature r i s e was greater than 0.75% of the set point temperature, the r e s u l t s were rejected as being suspect. Once the hot gas escapes the reactor, the gas attains low temperature i n an extremely short time. The mean temperature of gas i n the cryogenic trap was 25°C with a range of 18 to 29°C. The walls of the trap were -75°C over the f u l l range of residence times studies. The temperature of the l i q u i d extraction column averaged 1°C with range of -10°C to + 10°C. In a l l , the system performed well i n terms of rapid heating of coal p a r t i c l e s and vapours and subsequent quenching. No signs of thermal stress or i r r e v e r s i b l e adsorption of tar onto any components were observed. C. R e p r o d u c i b i l i t y of Results and Analysis To test the r e p e a t a b i l i t y and r e p r o d u c i b i l i t y of the feeder, pyrolyser, recovery system and analysis, a series of runs was performed using the same residence times, temperature, coal loading and reactor geometry. The re s u l t s from two i d e n t i c a l runs appear as #43 and #44 i n the summary of r e s u l t s , Table D - l , Appendix D. For a residence time of 1.0 seconds at 700°C i n the long residence time reactor (short, wide secondary zone) the results compare favorably. Recall that for each run, the tar y i e l d and product d i s t r i b u t i o n s are determined by duplicate - 101 -a n a l y s i s . The des i r e d coal l o a d i n g f o r these runs was 5.40 x IO-'' kg 3 c o a l per m of hot gas. As i n d i c a t e d i n Table 9, the a c t u a l l o a d i n g obtained was w i t h i n 5% of the desired value and the loa d i n g from one run to the next was w i t h i n 8% of the mean value. Considering the d e v i a t i o n i n observed feed rates noted e a r l i e r an 18% d e v i a t i o n i n loading could be expected. While i t would have been d e s i r a b l e to keep loading w i t h i n a narrower range, the problems of pressure c o n t r o l and other systematic e r r o r s d i d not al l o w s o l u t i o n to t h i s problem e a r l y enough. Tar y i e l d s over the four analyses averaged 10.1% on a moisture and ash-free basis w i t h i n a 3% d e v i a t i o n . Considering the number of steps i n the a n a l y s i s and procedure (see Figure 15) t h i s r e s u l t i s extremely good. This y i e l d value i s of the order observed by J a r a l l a h (19) using a spouted bed py r o l y s e r on the same c o a l . Char and gas y i e l d s , as o u t l i n e d e a r l i e r , were not a v a i l a b l e due to a n a l y t i c a l c o n s t r a i n t s . C l e a r l y these would have been d e s i r a b l e to perform an o v e r a l l mass balance. From the tar obtained, f r a c t i o n a t i o n by so l v e n t s o l u b i l i t y and subsequent HPLC allows a determination of the product weight % d i s t r i b u t i o n of aromatics, polars and hexane i n s o l u b l e s . Saturates (and o l e f i n s ) were not observed i n q u a n t i f i a b l e amounts. In each f r a c t i o n , the percent d e v i a t i o n was l e s s than 10%. Considering that both solvent f r a c t i o n a t i o n and HPLC are low r e s o l u t i o n methods, t h i s order of magnitude of d e v i a t i o n i s acceptable. One i n t e r e s t i n g f e a t u r e that presents i t s e l f from these analyses i s the much lower d e v i a t i o n i n the run-to-run sum of polars plus hexane i n s o l u b l e s . Throughout a l l the - 102 -Table 9 S t a t i s t i c a l Analysis of Reproducibility Runs #43 and 44 Loading (g cm~ ) Calculated actual a % a 5.40 x 10~ 5 5.11 x 10" 5 0.41 x 1 0 - 5 8.0 Tar Y i e l d (wt. % m.a.f.) mean a % a 10.09 0.30 3.0 Product d i s t r i b u t i o n (wt. %) Mean a % a hexane insolubles 53.6 4.0 7.5 polars 26.5 2.6 9.9 aromatics 20.0 2.0 10.0 polars + insolubles 80.1 2.0 2.5 a = standard deviation % a = standard deviation as a percent of the actual r e s u l t or mean value - 103 -analyses, the deviation i n the sum was lower than that observed for each component. By conventional error anaysis the deviation should be approximately equal to the sum of each component's deviation. What this seems to indicate i s that polars and hexane insolubles may not represent d i s t i n c t or even somewhat unique f r a c t i o n s , but are one. Their s o l u b i l i t y in hexane may be very low and would be expected to be very temperature dependent. As such, day-to-day fluctuations in work-up storage temperatures could a f f e c t the r e l a t i v e proportion of each observed, though the tot a l would remain the same. It has been frequently observed that solutions of hexane solubles tend to form p r e c i p i t a t e s when cooled or stored. This may be a manifestation of this s o l u b i l i t y property rather than evidence of any degradative or chemical change. In a l l , the methods and apparatus seem to operate f a i r l y well i n giving results r e l i a b l e within a narrow range. With some modification further improvements could be made. D. Coal Loading In this study coal loading refers to the mass of coal fed to the reactor per unit volume of expanded gas. Loading i s obviously of c r u c i a l importance i n the commercial v i a b i l i t y of any coal conversion system. By varying the feed rate to the reactor under a constant regime of temperature and residence time, the e f f e c t of loading was studied. Using the short residence time reactor (long, narrow secondary zone) at 700°C and a gas residence time of 0.5 s., the tar y i e l d r e s u l t s in Figure 19 were obtained (see also Appendix D, runs 12 to 16, 22). - 104 -Figure 19 EFFECTS OF COAL SOLIDS LORDING Short RTD Reactor SOLIDS LORDING (KG/M3) x 102 - 105 -Dilute phase pyrolysis results in tar yields that are some 50% higher than dense phase pyrolysis. From this, for desirable tar yields a reactor should operate at a very low loading of coal to hot gas volume (or mass). Clearly this is a d i f f i c u l t condition to achieve while at the same time operating the reactor economically. Similar results were obtained by other researchers with this and other coals (104). The decline in tar yields observed parallels an increase in char and gas yields. While some have suggested this is the result of the presumed auto catalytic effect of char or coke, earlier discussion of gas phase reactions would indicate that other effects may be responsible. Cracking, chain reactions and bimolecular reactions are very dependent on the concentration of neutral parent hydrocarbons. An increase in loading would consequently improve the rates of these reactions. As these are very fast processes, the hydrocarbon substrate could become involved in numerous reactions over the residence time considered. The greater turnover possible with more hydrocarbon present should lead to enhanced secondary degradation•producing more gases and char to the detriment of tar yield as loading is increased. Figure 20 shows the product distribution of the tars obtained from the various loading experiments. Over most of the loading range the distribution remains nearly constant. The sharply higher yields observed in the dilute region are attributable almost entirely to polar and hexane insoluble components. Contrasting this is a slight reduction in aromatics. By combining the yield and distribution results, the data in Figure 21 are derived. The yield of less desirable polar materials - 106 -Figure 20 EFFECT OF COAL SOLIDS LORDING -i—i i i i—i—i i — i — i — i — i — i — i — i — i — i — i — i — i — i i 2.8 SJo ZA 112 U.O 16.8- 19J6 22.4 2S2 2%.Q 30.8 33.6 LORDING (KG/M3)X102 f - 107 -makes up f o r the s l i g h t l y lower or f l a t y i e l d of d e s i r a b l e aromatics. This could i n d i c a t e that polar components, possessing some a r o m a t i c i t y , can be degraded to some extent by the secondary processes so as to l i b e r a t e m a r g i n a l l y more aromatics. As the secondary r e a c t i o n r a t e s improve no greater y i e l d of aromatics occurs. Instead, c r a c k i n g to char and gases predominates from the polar m a t e r i a l s . The r e a l optimum l o a d i n g occurs where the y i e l d of d e s i r a b l e products i s highest or, i n other words, where maximum aromatics s e l e c t i v i t y occurs. Figure 22 de p i c t s the change i n aromatics s e l e c t i v i t y (aromatic y i e l d / p o l a r s + i n s o l u b l e s y i e l d ) with changes i n co a l l o a d i n g . This i n d i c a t e s that w h i l e tar y i e l d s may be highest at low l o a d i n g , the product i s of very poor q u a l i t y w i t h low aromatics s e l e c t i v i t y ( f o r comparison to other petroleum products see Appendix F ) . As the lo a d i n g increases the aromatic f r a c t i o n improves r e l a t i v e to the undesirable products. From t h i s , optimum loa d i n g would appear to be a dense phase feed greater — 2 — 3 than 5 x 10 kg m . By using such dense loa d i n g s , r e a c t o r throughputs can be high w i t h no s i g n i f i c a n t l o s s i n the more d e s i r a b l e products. This r e s u l t i s contrary to much of the current design work f o r p y r o l y s i s r e a c t o r s where low loading i s g e n e r a l l y used. The reasons f o r t h i s are easy to see: by foc u s i n g an o v e r a l l tar y i e l d s alone the incremental improvement i n y i e l d w i t h lower l o a d i n g i s very l a r g e . This work confirms t h i s f i n d i n g , but when the q u a l i t y c o n s i d e r a t i o n s are f a c t o r e d i n the improvement i n y i e l d i s outweighed by u p g r a d e a b i l i t y and value of the products obtained. While po l a r s and hexane i n s o l u b l e components are not e n t i r e l y worthless m a t e r i a l s , i n a market where heavy crude o i l s , shale o i l and various pitches are becoming i n c r e a s i n g l y important as - 108 -Figure 21 EFFECT OF COAL SOLIDS LORDING LOADING (Kg/M 3 ) x 102 - 109 -Figure 22 EFFECT OF COAL SOLIDS LORDING . . y i e l d aromatics--s e l e c t i v i t y • — — • , , • -. • , , , • v y i e l d (polars + hexane insolubles; - 110 -conventional l i g h t crude supplies dwindle, the value of these polar components w i l l be discounted f u l l y . Premium f u e l and l i g h t e r crude f r a c t i o n s w i l l continue to be the most valuable hydrocarbons. To this end, s a c r i f i c i n g t o t a l tar yi e l d s for improved q u a l i t y and greater economy may well be sensible. E. Residence Time and Temperature Effects Both the pyro l y s i s temperature and duration of reaction are expected to be of prime importance. To study the e f f e c t of these variables a series of experiments were performed to cover a matrix of temperature from 500 to 800°C i n 100°C increments for residence times of 0.5, 1.0 and 2.0 s. In a l l cases the long residence time reactor was — 2 3 used. The loading chosen was 5.4 x 10 kg m- and as feed rates were adjusted accordingly for each temperature and residence time. For this _ o group of experiments the actual mean loading obtained was 4.93 x 10 kg _ a m with a percent deviation of 12.8% from this mean. The re s u l t s for a l l twelve of the above runs are shown gra p h i c a l l y i n Figures 23 to 34 as well as i n Tabular form i n Appendix D. The figures are organized i n blocks showing the tar y i e l d s for each time followed by product d i s t r i b u t i o n and the net product y i e l d s . Tar y i e l d s at various temperature and residence times are depicted i n Figures 23 to 26. At the short residence time (0.5 s) the maximum tar y i e l d of 11.65% m.a.f. occurs at approximately 800°C. As the residence time i s increased the y i e l d maximum of 11.33% m.a.f. (1.0 s) and 11.53% m.a.f. (2.0 s) occurs at 700°C. These y i e l d r e s u l t s Figure 23 TEMPERATURE VS RESIDENCE TIME MATRIX Residence Time =0.5 Seconds 12.8-12.0-500 600 700 800 TEMPERATURE (C) TEflPERATURE (C) - 113 -- 114 -are lower than the values obtained by Jarallah (105) using Balmer coal. The temperature at which the tar yield maxima occurs is in agreement with those found in that study. Figure 26 shows a composite three-dimensional overlay graph of the results in the previous three figures. This drawing allows an easy comparison of the relative changes in tar yields. As shown, the yields generally increase with temperature until a maximum value near 700°C is reached. The slope of this line (-^  = 0.0239%/°C) is nearly independent of time. However, after the maxima is attained, a sharp decrease in yield is evident as the vapour residence time is increased. At 800°C the rate of decrease in tar yield is roughly 2.34 wt. % per second of residence time. Why is there such an abrupt change in tar yields with residence times only at the highest temperatures? To deal with this question one need reconsider the chemistry and kinetic changes that occur at higher temperature. A l l rates of reaction increase with temperature. The degree of reaction, the reaction coordinate, is determined by the time the reactions are given to run. Secondary reactions, with typically low activation energies may be approaching a diffusion-limited regime at high temperatures. Reaction rates would be expected to increase in 1 75 proportion to T • as does the diffusivity of molecules in the gas phase (106). This rate of increase would be less than that predicted by the exponential Arhenius expression. For example reaction rates for a typical activation energy of 50 KJ/mol increase by a factor of 3.1 for a 200°C rise in temperature. Diffusivity in the gas phase would increase - 115 -Figure 26 TENPERRTURE VS RESIDENCE TIHE HflTRIX Tar Y i e l d TEflPERflTURE (C) - 116 -only 1.4 times and only 1.2 times i n porous s o l i d s . I f bimolecular secondary routes become limited by d i f f u s i v i t y , other routes such as i n t e r n a l disproportionation, rearrangement and fragmentation w i l l become I of greater s i g n i f i c a n c e . This should lead to increased l e v e l s of smaller fragments, gases and o l e f i n s to the detriment of tar. As well as these p o t e n t i a l k i n e t i c considerations the dramatic tar y i e l d decrease may be due to some other e f f e c t s . Recall that at higher temperatures, the rates of such energy-intensive primary reactions as carbon-hydrogen and carbon-carbon cleavage become s i g n i f i c a n t . This should open up new reaction horizons i n the secondary processes due to the r a d i c a l s and o l e f i n s formed. These reactions should lead to increases i n both smaller (gaseous) and larger fragments (chars) at the expense of tar constituents. The combined e f f e c t of the increase i n the primary reaction inventory and the p o s s i b i l i t y of increased unimolecular secondary reactions could account for the rapid loss of tars at the higher temperatures. The product d i s t r i b u t i o n s and net product y i e l d s obtained i n these experiments are shown i n Figures 27 to 32. From the product d i s t r i b u t i o n s , i t can be seen that the s e l e c t i v i t y of aromatics i s highest at the low temperatures and f a l l s s t e a d i l y or remains e s s e n t i a l l y constant as the temperature i s increased. Subsequent processing and upgrading considerations would favour these low temperature tars as a substitute crude or blending ingredients. The net product y i e l d s for each of these runs shows the same phenomenon as observed e a r l i e r : increased tar y i e l d s are la r g e l y due to - 1 1 7 -F i g u r e 2 7 TEMPERATURE VS RESIDENCE TIME MATRIX Residence Time = 0.5 Seconds 5 0 0 8 0 0 TEMPERATURE (C) - 118 -Figure 28 T E M P E R A T U R E VS R E S I D E N C E T I M E M A T R I X Residence Time = 1 Second HEXANE I N S O L U B L E POLAR HROnflTIC POLAR t INSOLUBLE TEMPERATURE (C) - 1 1 9 -F i g u r e 2 9 TEMPERATURE VS RESIDENCE TIME MATRIX Residence Time = 2 Seconds T E M P E R A T U R E (C) - 120 -Figure 30 TEMPERATURE VS RESIDENCE TIME MATRIX Residence Time = 0.5 Seconds TEMPERATURE (C) - 1 2 1 -F i g u r e 3 1 T E N P E R R T U R E VS R E S I D E N C E T I M E MRTR IX R e s i d e n c e T i m e = 1 s e c o n d 0.8 H 1 1 1 r 1 1 — 1 — " " - " i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — I 500 600 700 800 TEflPERflTURE (C) - 122 -Figure 32 T E f l P E R f l T U R E VS R E S I D E N C E T I M E M A T R I X Residence Time = 2 Seconds TEMPERATURE (C) - 123 -increased polar and hexane insoluble y i e l d s . Figure 33 gives a composite overlay of the asphaltene y i e l d . This figure shows nearly the same o v e r a l l appearance as that given e a r l i e r (Figure 26) for the t o t a l tar y i e l d s . At higher temperatures and longer residence times, the decrease i n tar y i e l d i s due to the cracking of asphaltenes and polars. Cert a i n l y i t would be b e n e f i c i a l i f these materials degraded to give aromatics, but the observed r e s u l t indicate otherwise. The loss of polars does not coincide with any increase in aromatics y i e l d but rather must contribute to the gaseous and char y i e l d s . Figure 34 shows a composite overlay of the net product y i e l d of aromatics. Considering the range of experimental error present, the surface i s f l a t or nearly so over the entire temperature-time matrix. This figure and the preceding one indicate that no q u a l i t y or s e l e c t i v i t y improvement i s gained by operating at higher temperatures. Though tar yields vary considerably, the net value i n terms of premium product produced i s e s s e n t i a l l y the same. Operationally, this would indicate that lower temperatures are preferable. E a r l i e r suspicions that asphaltenes and polars may indeed be nearly the same i n structure are strengthened by these r e s u l t s . The cracking of hexane insolubles to some intermediate form of polar material has been postulated. The results here indicate otherwise i n that no increase in the y i e l d of polar components i s seen; instead both asphaltenes and polars roughly p a r a l l e l each other. F. Extended Temperature Range P y r o l y s i s To extend the examination beyond the range discussed in the above section and to test the l i m i t s of the apparatus a run was carried out at - 124 -Figure 33 TEMPERATURE VS RESIDENCE TIDE nflTRIX Net Product Y i e l d : Hexane Insolubles 0.5 SEC. 1.0 SEC. 2.0 SEC. 500 600 700 800 TEflPERflTURE (C) - 125 -F i g u r e 34 TEHPERflTURE VS RESIDENCE T l t lE • HRTRIX N e t P r o d u c t Y i e l d : A r o m a t i c s TErtPERRTURE (C) - 126 -900°C, a residence time of 1 s. and at the same loading as above. These re s u l t s together with other 1 s. runs are shown in Figures 35, 36 and 37. No operating d i f f i c u l t i e s were observed at the higher temperatures. No agglomeration of coal in the upper feeder flange was observed. Trapping and column scrubbing conditions remained within the desired range. No physical d e t e r i o r a t i o n or thermal stresses were observed i n either of the steel or glass components of the reactor system. Time did not permit an expansion of the temperature-time matrix to the higher l i m i t s . As shown in Figure 35, the total tar y i e l d decrease at an accelerating rate after the maximum observed at 700°C. Again the total tar y i e l d p a r a l l e l s the y i e l d in the polar and asphaltic components. Of i n t e r e s t though i s the increase in aromatics s e l e c t i v i t y as the higher temperature i s reached. Both the lowest and the highest temperature tars have nearly the same aromatics s e l e c t i v i t y . Either of these products would be the best available for upgrading to fue l s . While there i s no s i g n i f i c a n t change i n the net y i e l d of aromatics (Figure 37) the tota l product slate from these two extremes i s very d i f f e r e n t . The low temperature pyrolysis produces very l i t t l e hydrocarbon off-gas (CH 4, C2H4 . . .) but leaves a char with s i g n i f i c a n t residual v o l a t i l e s (see Section 10.J.). The higher temperature pyrolysis produces s i g n i f i c a n t quantities of gaseous hydrocarbons (see' Figure 2) but leaves a char with very l i t t l e r esidual v o l a t i l e s . From an economic point of view i t w i l l be necessary to know the r e l a t i v e values of these chars and gases to determine i f the added operating expense needed to achieve the higher temperatures i s j u s t i f i e d . - 127 -F i g u r e 35 EXTENDED T E M P E R A T U R E RANGE P Y R O L Y S I S TEMPERATURE (• - 128 -Figure 36 EXTENDED TEMPERATURE RANGE PYROLYSIS o HEXANE INSOLUBLE • POLAR « flROflRTIC x POLAR t INSOLUBLE TEflPERflTURE (C) - 129 -Figure 37 E X T E N D E D T E M P E R A T U R E RANGE P Y R O L Y S I S . o o HEXANE INSOLUBLE • + POLAR ' » • PROHPTIC . x x POLAR + INSOLUBLE TEMPERATURE (C) - 130 -G. Extended Residence Time Range P y r o l y s i s Two experiments were conducted to examine an extended range of residence times. Both were carried out at 700°C and the same coal loading as was used i n the previously discussed matrix. The residence times were extended to 0.25 s and 4.0 s. The re s u l t s of these runs are included i n Figures 38 to 40. Operationally, no d i f f i c u l t i e s were encountered i n extending the residence time range either way. The low residence time was achieved by an increase i n the purge gas flow. To accommodate this volume of gas an extension column was added to the f i n a l l i q u i d extraction column and the number of c e l l u l o s e f i l t e r s used i n the cyrogenic trap was reduced. This was found necessary due to tar carryover as evidenced by a s l i g h t d i s c o l o r a t i o n of the f i l t e r i n the f i n a l gas manifold when operating under normal circumstances (see Operations, Section 8). While not conclusively evident from the previous matrix r e s u l t s , the decrease i n tar y i e l d with increased residence time i s observed when the range i s extended further. As well, an increase i n tar y i e l d i s seen for the shortened time. With experimental error, the o v e r a l l trend i s a decrease i n tar y i e l d with the vapor residence time, both at 700 and 800°C. The product d i s t r i b u t i o n s show a f l a t s e l e c t i v i t y for aromatics. The net product y i e l d s show again that polar and a s p h a l t i c materials are almost e n t i r e l y responsible for the changes i n o v e r a l l y i e l d . The net y i e l d of aromatics remains e s s e n t i a l l y unchanged over the en t i r e time range. What these re s u l t s seem to indicate i s that the i n i t i a l stage of - 131 -Figure 38 EXTENDED RESIDENCE TIME RANGE PYROLYSIS i i i i i — i — i — i — i — i — i — — i — \ — i — i — i — 10 2.0 3.0 4.0 RESIDENCE TIME (SEC.) - 132 -F i g u r e 39 EXTENDED RESIDENCE TIME RANGE PYROLYSIS RESIDENCE TIME (SEC.) - 133 -Figure 40 EXTENDED R E S I D E N C E TIME, RANGE P Y R O L Y S I S —i 1 1 1 1 1 1 1 1 1 r — i — i — i 1 1 1 1 1 1 1 1 1—1 0.4 1.0 2.0 3.0 4-0 RESIDENCE T l f l E (SEC) - 134 -p y r o l y s i s , d e v o l a t i l i z a t i o n and the primary reactions generate a mixture of tar hydrocarbons which upon further reaction (with time) do not degrade to more desirable products but rather produce less valuable gases and char. Reactions that could generate aromatics, such as eli m i n a t i o n (neutral or r a d i c a l ) or o l e f i n dimerization do not occur to an extent s i g n i f i c a n t enough to increase the product y i e l d . On the other hand, those aromatics that are l i b e r a t e d appear to be r e s i l i e n t to change as would be thermodynamically predicted. The r e s u l t s then indicate that tar yields can be maximized by going to lower residence times. Aromatic s e l e c t i v i t y i s r e l a t i v e l y unchanged and so, residence time considerations are dictated by the r e l a t i v e values of the less desirable product gases, asphaltenes or char and the operating expenses associated with the reactor and residence time chosen. H. Reactor Comparisons As discussed i n Section 7, two p y r o l y s i s reactors of the same design, both made of 304 s t a i n l e s s s t e e l and with approximately the same 9 i n t e r n a l area (142 cm ) were constructed. The difference between the 3 reactors was th e i r i n t e r n a l volume, one being approximately 45 cm to give a r a t i o of area to volume of 3.16 cm - 1, the other reactor volume being approximately 55 cm to give an (A^/y) r a t i o of 2.58 cm- . Experiments were performed to see i f this difference i n (A^/y) would have any e f f e c t . Using a residence time of 0.5 s, and a mean loading of 4.67 x 10" 2 kg m~3 (± 3.6%) pyrolyses at 500, 600 and 700°C were performed using each reactor. - 135 -The reactor with the highest (A-jyy) r a t i o (low residence time reactor) w i l l be c a l l e d reactor #1. Figure 41 shows the comparative tar y i e l d s obtained. While each curve shows a s i m i l a r change i n y i e l d with temperature, the values obtained for. reactor #1 are, consistently lower. J a r a l l a h (107), using the same Balmer coal at a temperature of 600°C i n a s t e e l walled spouted bed reactor with (A^/y) = 0.20, obtained a tar y i e l d of approximately 18 wt. % m.a.f. By p l o t t i n g tar y i e l d at 600°C against the reactor (A-^/y) r a t i o for these r e s u l t s and the ones obtained i n this study a l i n e a r r e l a t i o n s h i p (within a 1% deviation) i s observed. While the r e s u l t s may not be d i r e c t l y comparable and three points are too few to prove any r e l a t i o n s h i p , the observation i s nonetheless thought-provoking. What these r e s u l t s seem to indicate i s that increasing the (A^/y) r a t i o of s t a i n l e s s s t e e l pyrolysis reactors can have a dramatic e f f e c t on the t o t a l tar y i e l d observed. The decrease amounts to a 3.5 wt % m.a.f. loss i n y i e l d per unit increase i n the s t e e l (A-^/y) r a t i o . Accounting f o r this observation i s rather d i f f i c u l t . For a l l of the runs examined the gas flow throughout the pyrolyser i s well within the laminar regime. The c a t a l y t i c e f f e c t of s t a i n l e s s s t e e l on coking and cracking reaction has been discussed (108). Since the the surface areas of each reactor i s approximately the same there w i l l be an equivalent quantity of pyro l y s i s vapor i n contact with this area i n the nearby streamlines. Detrimental cracking of these proximate hydrocarbons may then occur, producing gases and coke at the metal surface. In a laminar regime the replenishment of reactants at this - 1 3 6 -F i g u r e 4 1 REACTOR COMPAR ISON TEMPERATURE (D - 137 -surface i s dependent on the d i f f u s i o n of these reactants perpendicular to the streamlines. The larger bore downcomer i n the long residence time reactor w i l l have a longer path than the smaller bore short residence time reactor. Surface replenishment w i l l then depend on the reactant d i f f u s i o n path length (or the inverse of the tube diameter). The narrower the tube diameter the greater the opportunity for such a surface reaction, a l l else being equal (surface area a v a i l a b l e , temperature and vapor residence time). These predictions suggest that turbulent flow regimes and high (Aj^/y) r a t i o s be avoided for s t e e l p y r o l y s i s reactors. The product d i s t r i b u t i o n s from the above experiments are given i n Figures 42 and 43. The same general r e s u l t s are obtained for each reactor i n terms of product s e l e c t i v i t i e s . The net product y i e l d s are given i n Figures 44 and 45. The recurrent r e l a t i o n s h i p between increased tar y i e l d and increased polar and asphaltene y i e l d s i s again demonstrated i n these f i g u r e s . Of some i n t e r e s t though are the s l i g h t l y higher y i e l d s of aromatics i n the case of the high (A^/v) reactor. While not conclusive, this observed increase i n aromatics y i e l d when o v e r a l l y i e l d i s decreasing indicate that some b e n e f i c i a l reactions may be occurring at the m e t a l l i c surfaces. From e a r l i e r discussions, aromatics and some aromatic precursors ( o l e f i n s , unsaturated compounds) are known to react v i a me t a l l i c association complexes (109) ( e s p e c i a l l y D iels-Alder r e a c t i o n s ) . If these preliminary r e s u l t s are any i n d i c a t i o n , the decreased y i e l d (and increased char and gas y i e l d s ) may be a harbinger of a more desirable increase' i n aromatic y i e l d wih higher - 1 3 8 -Figure 42 RERCTOR COMPARISON Short RTD Reactor 80-72 64 -56-48-40-32-24 16-8-© O HEXANE INSOLUBLE + + AROMATIC « * POLAR x x POLAR t HROnflTIC r i 500 600 i i 700 TEMPERATURE (C) - 139 -Figure 43 REACTOR COHP f lR I SON Long RTD Reactor . o o HEXANE INSOLUBLE + + AROflflTIC " ® • POLAR - * x POLAR t INSOLUBLE X - V X X o + e * • + 500 600 700 TEflPERflTURE (C) - 140 -Figure 44 REACTOR COMPAR ISON Short RTD Reactor o o HEXANE INSOLUBLE • + flROnflTIC t\ • • » POLAR . x x POLAR £ INSOLUBLE - i — i — i i i — i i i — i i i i i i i i 500 600 700 TEMPERATURE (C) - 141 -Figure 45 REACTOR COMPAR I SON Long RTD Reactor i l _ — 1 1 i — r — i — i — i 1—r— i— i—i 1—i—i r—i 1— i 1 r 50° 600 700 TEMPERATURE (C) - 142 -(A-i/y) r a t i o s . Again, examination of the q u a l i t a t i v e results would favour a d i f f e r e n t reactor design than would the quantitative analysis of tar y i e l d alone. I. Oxygen A single experiment involving a low l e v e l of oxygen i n the pyrolysis gas was performed to see i f any changes i n the product s l a t e would r e s u l t . The basis for this experiment was the theoretical predictions made e a r l i e r i n connection with hydroxyl r a d i c a l chain c a r r i e r s (reactions 17, 25 and 30 i n Appendix A). The conditions chosen were 700°C and a 1 s residence time using the long residence time reactor (#1). Pure oxygen was bled into the pyrolysis gas stream to give a concentration of approximately 1500 ppm (V/V). This represents only 0.024% of the required stoichiometric combustion oxygen. The results from this t r i a l are given as run 45 i n the summary Appendix D. The results are i n the same range as those obtained i n an oxygen free environment, within the range of observed experimental error. The presence of small quantities of oxygen i n this system does not seem to have an appreciable influence on tar y i e l d or product d i s t r i b u t i o n or y i e l d s . This r e s u l t confirms the observations made by other researchers i n the area (44,45). Metal reactor surfaces are known to react readily as chain terminators or r a d i c a l scavengers, p a r t i c u l a r l y for hydroxyl or i t s analog sulfur r a d i c a l s . In order to observe the hypothesized effects of oxygen, i t would be necessary to use a quartz (or l i k e ) reactor. - 143 -J . Thermal Analysis of P y r o l y s i s Chars The char produced i n each of the above experiments was recovered, treated as described previously (Figures 14, 15, 16) and then subjected to thermal gravimetric analysis. By measuring the weight loss of coals or char as a small sample of the material i s heated according to a programmed temperature p r o f i l e , a measure of the residual v o l a t i l e s can be obtained. The f i r s t r e l a t i o n s h i p to be explored was that of the p y r o l y s i s tar y i e l d and the residual v o l a t i l e s remaining i n the char. If off-gasing and coking reactions were of much less s i g n i f i c a n c e than simple d e v o l a t i l i z a t i o n , one would expect a simple inverse r e l a t i o n s h i p between the tar y i e l d and the residual v o l a t i l e s remaining i n the char. As tar y i e l d increases, residual v o l a t i l e s should decrease. Figure 46 i l l u s t r a t e s the r e l a t i o n s h i p between these two variables from a l l of the experiments performed ( a l l temperatures, times and r e a c t o r s ) . The pattern demonstrates no s i g n i f i c a n t c o r r e l a t i o n between residuals and tar y i e l d . While not unexpected this r e s u l t i l l u s t r a t e s the s i g n i f i c a n c e of coal reactions that produce gases and coke. Simple d e v o l a t i l i z a t i o n does not dominate the pyrolysis process. The next r e l a t i o n s h i p explored was that of residual v o l a t i l e s and their dependence on the p y r o l y s i s temperature. As noted, temperature i s a very s i g n i f i c a n t factor a f f e c t i n g the d i f f u s i v i t y ( s o l i d and gas phase), p e r m i s s a b i l i t y and rate of pyrolysis reactions. Figure 47 indicates that a roughly l i n e a r r e l a t i o n s h i p exists between the char r e s i d u a l v o l a t i l e s and the p y r o l y s i s temperature. Within the range of - 144 -Figure 46 RESIDUAL VOLATILES VERSUS PYROLYSIS TAR YIELD o o © CD CD O CD CD 0 O © CD CD CD o 4J8 5.6 6.4 7.2 8 .0 & 8 9 .6 10.4 1 1 2 1 2 . 0 1 2 . 8 13 .6 PYROLYSIS TflR YIELD (UT.%) U) by TGA method (see Figure 15) (.2) residual v o l a t i l e s as weight percent i n chars, - 145 -Figure 47 RESIDUAL VOLATILES VERSUS PYROLYSIS TEMPERATURE 1 1 i i i i i i i i 500 600 "T T 700' T 1 P 800 " i — i — i i i 900 TEMPERATURE (C) (1) residual v o l a t i l e s by TGA method (see Figure 15). (2) temperature of pyro l y s i s run from which these chars were derived. - 146 -error this l i n e can be extrapolated back to room temperature and the previously determined v o l a t i l e s present i n the coal i t s e l f . These r e s u l t s give the only r e a l measure available i n this study of the q u a l i t y of the chars produced. The ultimate fate of such char i s l i k e l y to be for use as f u e l i n a pulverized coal combustor as i t would have l i t t l e value i n the more l u c r a t i v e coke markets as an adsorbent or for use i n blast furnaces. For e f f i c i e n t coal combustion i n a s e l f -propagating coal furnace, the f u e l needs to have 9 to 14 wt % v o l a t i l e matter. If the feed has less than this an a u x i l i a r y f u e l ( o i l , gas) or blended feeding i s required (110). From the r e s u l t s here, only those chars produced at the low temperature end appear to be useful as i s . K. Infrared Spectra of Hexane Insolubles A simple study of the hexane insoluble f r a c t i o n (asphaltenes) of the coal tars produced was undertaken to examine some features of their chemical structure (111). By infrared absorbance spectroscopy some measure of their usefulness as a substitute petroleum feedstock component can be obtained. As i l l u s t r a t e d i n Figure 48, the spectrum of the asphaltenes i s r e l a t i v e l y featureless with s i g n i f i c a n t amounts of background absorbance. Generally a l l of the spectra display very l i t t l e hydrogen (1380 cm"1 -CH 3, 2900 cm - 1 - CH, 730-860 cm - 1 0-H) on a predominately aromatic molecule (1450-1600 cm - 1). Oxygen functional groups are quite evident (quinones 1675 cm - 1, hydrogen bonded phenols 3200-3400, 1200 cm - 1, carboxylic acids 1710 cm - 1, esters 1735 cm - 1, ethers 1200, 1070 - 1 4 7 -1 ^ 1 1 1 1 1 I 3600 3200 1800 1600 1400 1200 1000 800 C M - 1 7 0 0 ° C - 148 -cm- ). Also evident i s the f a c t that asphaltenes do not comprise a unique chemical group. Their i n s o l u b i l i t y i n hexane does not suggest a s i m i l a r i t y i n functional groups. As shown, asphaltenes possess decreasing quantities of phenolic and carbonyl f u n c t i o n a l i t y as either the residence time or temperature i s increased (Figure 49). With this comes an increase i n aromatic!ty and background absorbance. This reduction i n oxygen and other heteroatom f u n c t i o n a l i t y i s as expected from the e a r l i e r discussions on p y r o l y s i s reactions. O v e r a l l , asphaltenes appear to be poor candidates for subsequent upgrading. The high degree of aromaticity and large amorphous size (as evidence by the background dispersion) of these structures makes the addition of hydrogen very d i f f i c u l t . Such structures are well on t h e i r way to becoming char or c h a r - l i k e . While polars were not i s o l a t e d and examined independently, i t i s l i k e l y that they would demonstrate s i m i l a r s t r u c t u r a l features. Materials such as these would most l i k e l y end up i n the r e s i d or tower bottom f r a c t i o n s of a t y p i c a l r e f i n e r y operation. Their ultimate destiny would be for use i n low value products l i k e binder p i t c h , asphalt or heavy combustion f u e l s . - 149 -Figure 49 INFRARED SPECTRA OF HEXANE INSOLUBLES pyrol y s i s temperature °C 1900 1700 1500 1300 1100 900 700 CM -1 Residence time = 1 s - 150 -11. CONCLUSIONS AND RECOMMENDATIONS The preceding theoretical and experimental discussions point to a number of important considerations for coal conversion reseachers. In pyr o l y s i s , as i n any other conversion processes, the entire package of input, operating and output values must be taken into account. Following the course of this study, the recommendation that follow for construction of a commercially oriented pyrolysis system would include the following: • Low temperatures (< 550°C) High temperatures may provide greater tar y i e l d s , but the s e l e c t i v i t y of premium products i s not improved. For a lower operating expense a more valuable tar i s obtained. The chars produced would contain s u f f i c i e n t residual v o l a t i l e s to enable their use d i r e c t l y i n the most probable end use of combustion. The quantity of gaseous products i s also much reduced, simp l i f y i n g downstream processing. • Minimize the vapor residence times Low residence times improves the throughput of the system while s t i l l allowing optimum tar recovery of more valuable product. • Use a heavily loaded system The coal feed rate to pyrolysis gas r a t i o can be r e l a t i v e l y high without impacting on the y i e l d of important tar constituents. So long as p a r t i c l e heat transfer rates are kept s u f f i c i e n t , a very high reactor throughput can be achieved. • Minimize the reactor (Aj_/y) r a t i o - 151 -The reactor should be designed to minimize the area of exposed metal per unit reactor volume. This can be achieved by using such configurations as a short large diameter tubular or moving bed reactor. A l t e r n a t i v e l y non-metal materials of construction could be used (e.g. quartz or ceramic). The pot e n t i a l a v a i l a b i l i t y of addit i o n a l chain c a r r i e r s would also suggest a reduction i n (A-^/y) r a t i o s or a l t e r n a t i v e materials of construction. • Consider the s e l e c t i v e addition of catalysts or other reagents. High temperature i r o n surface reactions may improve product s e l e c t i v i t i e s . With some i n v e s t i g a t i o n , low temperature metal c a t a l y s t s or modifiers could be found to achieve a si m i l a r or improved r e s u l t without reverting to higher (A^/y) reactors. • Consider the chemical implications involved i n the heat-up, pyro l y s i s and quenching of coal tars. The recommendation i n terms of system design that flow from this thesis d i f f e r somewhat from those proposed bv most others i n the coal p y r o l y s i s area. The reasons for this are evident. Coal chemistry and k i n e t i c s are oftentimes oversimplified or ignored because of their inherent complexity. While a complete understanding may not be av a i l a b l e , i t i s this writer's b e l i e f that many useful predictions and design considerations flow n a t u r a l l y from even a cursory examination of th e o r e t i c a l aspects. On the empirical side, much attention i s often focused on a few simple variables or fact o r s , such as tar y i e l d maxima or temperature. The tact taken i n this study was purposefully more broadly based and, while f ar from complete, was aimed at providing a more balanced view. - 152 -I t i s t h e h o p e o f t h i s a u t h o r t h a t t h e t h e o r y , d i s c u s s i o n a n d r e s u l t s p r e s e n t e d h e r e w i l l b e o f a s s i s t a n c e t o t h o s e i n t h e f i e l d o f c o a l c o n v e r s i o n a n d r e l a t e d d i s c i p l i n e s a n d t h a t t h i s u n d e r s t a n d i n g w i l l h e l p f o s t e r p r o d u c t i v e s o l u t i o n s t o t h e w o r l d ' s e n e r g y p r o b l e m s . - 153 -NOMENCLATURE A constant A^ absorbance A. i n t e r n a l surface area 1 a constant B cons tant BDE bond d i s s o c i a t i o n energy b constant C constant Cp heat capacity c concentration D tube hole diameter dp p a r t i c l e diameter dp mean p a r t i c l e diameter Ea a c t i v a t i o n energy F flow rate r a d i a t i o n view factor G coal s o l i d s loading AG Gibbs free energy change h o v e r a l l heat transfer c o e f f i c i e n t h Plancks constant P AH reaction enthalpy k rate constant k^ Boltzmann constant k thermal conductivity g L c e l l path length - 1 5 4 -M molecular weight M mass feed rate t i . f r a c t i o n of reactant i i n mixture 1 N Nussett Number u N R e Reynolds number (Re) P pressure SAT P saturated vapor pressure Pr P r a n d t l number p c o l l i s i o n p r o b a b i l i t y R gas constant r p a r t i c l e radius r mean pore diameter AS entropy of r e a c t i o n U g s l i p v e l o c i t y T absolute temperature Tp p y r o l y s i s temperature t time At i n f i n i t e l y small increment i n time U g s l i p v e l o c i t y V volume Vp p y r o l y s i s v o l a t i l e matter, wt % 00 V p y r o l y s i s v o l a t i l e matter, wt % at t = 0 0 X co a l conversion - 155 -D d i f f u s i v i t y D d i f f u s i v i t y of gas from porous s o l i d pore J O r E em i s s i v i t y e e x t i n c t i o n c o e f f i c e n t 9 ^ temperature gradient p density a standard deviation T vapor residence time T t o r t u o s i t y u v i s c o s i t y v atomic d i f f u s i o n volume X i n t e r n a l porosity of s o l i d p a r t i c l e - 156 -REFERENCES 1. Opportunities for Coal Liquefaction Research i n B.C. S c i . Council of B.C., Project #3904, May 1980. 2. Montano, P., "Characterization of i r o n bearing minerals i n c o a l , " i n Coal Structure Gobarty and Ouchi (Ed.) A.C.S. Symp. Series. #192, Washinton, D.C. (1982). 3. Organic Chemistry of Coal, J.W. Larsen (Ed.). A.C.S. Symposium Series #71, Washington, D.C. (1978), "A Primer on the Chemistry and Constitution of c o a l , " Whitehurst, D.D., p. 3. 4. Zubovic, P. "Minor Element D i s t r i b u t i o n i n Coal," in Coal  Science, Robt Gould (Ed.), A.C.S. #55, Washington, D.C. (1966), p. 232. 5. F l a i g , W., "Chemistry of Humic Substances in Relation to C o a l i f i c a t i o n , " i n Coal Science, i b i d . , p. 58. 6. Haggin, J . , "Interest i n Coal Chemistry I n t e n s i f i e s , " C & E.N., Vol. 60, No. 23 (1982), p. 21-22. 7. Mazunder, B.K., et a l . "A l i p h a t i c Structures i n Coal," i n Coal  Science, Supra Note 4. 8. Whitehurst, D.O., Farcasiu, M., M i t c h e l l , T.O. and Orchest, J . J . , E.P.R.I. Rep. AF-480, July 1977, p. 11, and Supra Note 3. 9. Messenger, L., Attar, A. Fuel, Vol. 58 (1979), p. 655. 10. ASTM Standards, Methods D 3172-3179. 11. Sprangler, R.J., Beckmann, B.G. and Kim, J.H. J . Org. Chem. 42, 2989 (1977). 12. Ratchford, W.P. i n Organic Synthesis, E.C. Horning (Ed.), V ol. 3, J . Wiley Publ., New York (1955), p. 30. 13. T y l e r , R.J. Fuel, 58 (1979), p. 680. 14. Levy, R.L., Fanter, D.L., Wolf, C.U., Anal. Chem. 44_, 1 (1972), p. 38. 15. Manufactured by Vacumetries, Ventura, C a l i f o r n i a , 93003. 16. Meuzelaar, H.L.C., Harink, H.C., Ficke, H.G., J . Chromat. S c i . , _13_ (1975), p. 12. 17. Meuzelaar, H.L.C., Posthumus, M.A., Kistemaker, P.G. and J . Anal. Chem. 45, 8 (1973), p. 1546. - 157 -18. Meuzelaar, Posthumus, Kistemaker, Biomed. Mass Spec. _1, 312-319 (1974). 19. J a r a l l a h , A., Ph.D. Thesis, University of B.C. (1983). 20. Tyler, R.J., Fuel, 59 (1980), p. 218. 21. Kaylhan, F., Ind. Eng. Chem. Process Des. Dev., Vol. 19, No. 1 (1980), p. 15. 22. Coles, E.T. Coalcon Hydrocarbonization Technology, Amer. Inst. Chem. Eng., 82nd Meeting (1976). 23. Chakrabartty, S.K., Brown, P.A.S. "Coal P y r o l y s i s : A State of the Art Review," Al t a . Res. Council (unpublished). 24. Maxwell, P. Hydrocarbon Processing, 55(9), p. 269-271. 25. Steinberg, M., Dept. of Energy and Environment, BNL. 26. Oberg, C.L., Falk, A.V., Silvermans, "Coal Liquefaction by Flash P y r o l y s i s , " Rockwell International, P.O. 5, # EX-76-C-01-2044. 27. Brown, R.F.C., Eastwood, F.W., Harrington, K.J., Aust. J . Chem. 27_ (1974), 23 73. 28. Seybold, G., Jersak, U. Chem. Ber. 1_10 , 1239-1245 (1977). 29. Seybold, G., Heibl, Ch., Chem. Ber. 110, 1225-1238 (1977). 30. King et a l . , Can. J . Chem. 48_, 3710 (1970). 31. Hedaya, E. Angew. Chemie, 367 (1969). 32. deMayo, P. Endeavar, 31_ (1972), p. 135. 33. P y r o l y t i c Methods i n Organic Chemistry, Brown, R.F.C. (Ed.), Monograms i n Organic Chem., #41 (1980). 34. Sharkey, A.G., Shulz, J.L., F r i e d e l , R.A., i n Coal Science, ACS # 55, 1966, p. 643 . 35. St. John, G.A., B u t t r i l l , S.E., Anbar, M. , "Field Ionization and F i e l d Desorption Mass Spectrometry Applied to Coal Research," i n Organic Chemistry of Coal, p. 233. 36. Taylor, J.E. and K u l i c h , D.M. i n Ind u s t r i a l and Laboratory  P y r o l y s i s , L.F. Albright, B.L. Crynes (Eds.) A.C.S. Syrap. Series #32 (1976), Washington, D.C. 37. A l b r i g h t , L.F., Yu, Y.H.C. i n Thermal Hydrocarbon Chemistry, Oblad, A.G., Davis, H.G., Eddinger, R.T. (Eds.), Adv. in Chem. Series #183, A.C.S. Publ., Washington, D.C. (1979), p. 198. - 158 -38. Tyler, R.J. Fuel 1980, V o l . 59, p. 218. 39. Williamson, K.D., Davis, H.G. "Mechanistic Studies on Ethane P y r o l y s i s at Low Pressure," i n Ind. and Lab. P y r o l y s i s , Supra Note 36. 40. Chen, C.J., Back, M.H. "Mechanism of the Thermal decomposition of methane," i n Ind. and Lab. P y r o l y s i s , Supra Note 36. 41. Sakai, T.., Nohara, D. "A K i n e t i c Study on the Formation of Aromatics during P y r o l y s i s of Petroleum Hydrocarbons," i n Ind. and Lab. P y r o l y s i s , Supra Note 36. 42. F i e l d s and Meyerson (1968), Vol. 6, p. 1, Acad. Press, N.Y. 43. Chemical Rubber Co., Handbook of Chemistry, 59th ed. p. F238. 44. Martin, R., Niclause, M., Scacchi, G. "Influences of small quantities of oxygen on thermal cracking of alkanes at 500°C" i n Ind. and Lab. P y r o l y s i s , Supra 36. 45. Taylor, J.E., Supra Note 36. 46. Al b r i g h t , L.F. and Yu, Y-H, "Production of coke and other • p y r o l y s i s products from acetylene, butadiene and benzene i n various tubular reactors," i n Thermal Hydrocarbon Chemistry, p. 193, Supra Note 37. 47. Advanced Organic Chemistry, Part A, F.A. Carey and R.J. Sundberg, Plennum Press, N.Y. 1984, p. 681. 48. Carey and Sundberg, p. 683, Supra Note 47. 49. Ladner, W.R., Newman, J.O., Sage, P.W., "Review of vapor phase cracking of coal derived materials and some model compounds to BTX," J . Inst. Energy [76], June 1980. 50. Louw, J . Chem. Soc-Perkin Trans., 2, p. 1327 (1973). 51. Rock, Mohmand, Angew Chem. Int. Ed. (Engl.), 16_, 104 (1977). 52. Bigley, D.B., Thurman, J.O., J . Chem. Soc. p. 436 (1968). 53. Hageman, Weirsum, Chem. Ber. 9_, 207 (1973). 54. Hageman, Weirsum, Chem. Ber. 9_, 206 (1973). 55. Schaden, Angew Chem. Int. Ed. Engl., 1_6, 50 (1977). 56. Schaden, 51, i b i d . 57. Griitzmacher, Hiibner J . Lieb i g s , Ann. Chem., 748, p. 793 (1973). - 159 -58. Bock, Mohmand, J. Am. Chem. S o c , 99, 1663 (1977). 59. Loudon, A.G., Macoll, A., Wong, S.K., J . Chem. S o c , B, 1933 (1970). 60. Corey, E.J., Block, E., J . Org. Chem. 34_, 1233 (1969). 61. Ladner, Newman, Sage, Supra Note 49. 62. Supra Note 67, p. 63. 63. F i e l d s , Meyerson, Vol. 6, P. 1, Acad. Press, N.Y. (1968). 64. Kelland, A.S., Pu r n e l l , J.H., Shurlock, B.C., "The Pyrolysi s of Propylene," Proc. Royal Soc. (1967), A, 300 (1460). 65. Kistiakowsky, G.B., Ruhoff, J.R,., Smith, H.A., vaughan, W.E. J . Am. Chem. S o c , 58, 146 (1936). 66. Warrener, R.W., K n e l l , K.I., Paddan-Row, M.W., Tetrah. Letters (1977), p. 53. 67. D i e l s Alder Reactions, A. Wasserman, E l s e v i e r Publ. (N.Y.), 1965, p. 50. 68. Sakai, Nohara, Supra Note 41, and Wasserman, Ibid., p. 49. 69. W i l l c o t t , M.R., C a r g i l l , R.L., Sears, A.B., "Thermal Unimolecular Reactions," i n Progress i n Physical Chemistry, Bradford Series (1982), p. 25. 70. Chakrabartty, Supra Note 23. 71. Haggin, J . , Supra Note 6. 72. Dutta, S., Wen, C.Y., Belt, R.J., Ind. Eng. Chem. Process Des. Devel., _16_, 20 (1977). 73. P i t t , G.J., "The Ki n e t i c s of the Evolution of V o l a t i l e Products from Coal," Fuel, 4_1_, 267 (1962). 74. Anthony, D.B., Howard, J.B., A.I.Ch.E.J., 22(4), 625 (1976). 75. Chang, P.W., Durai-Swamy, K., K n e l l , E.W. "Kinetics of Coal P y r o l y s i s Reactions i n a Flash Pyro l y s i s Reactor," i n Coal  Processing Technology, V o l . 6, A.I.Ch.E., N.Y. (1980). 76. J a r a l l a h , A., Supra Note 19. 77. Seybold, G., "Thermolysis of Organic Compounds," Angew. Chem. Int. Ed. Engl., 16, 365-373 (1977). - 160 -78. Scott, D.S., Pickorz, J . "Low Rate Entrainment Feeder for Fine S o l i d s , " Ind. Eng. Chem. Fundam., 1982, 21_, 319-322 (1982). 79. Report from coal analysis, General Testing Laboratories, Vancouver, B.C., Report #8209-15540. 80. For a general description see: Basic Liquid Chromatography, E.L. Johnson, Varian Assoc. (1978). 81. Galya, L.G., Suatoni, J.C., J . L i q . Chromat. 3(2), 229-242 (1980). 82. Alfredson, T.V., J . Chromat. 218, 715-728 (1981). 83. Alfredson (1981), Supra Note 82. 84. Alfredson (1981), Supra Note 82. 85. Dark, W.A., J . Chromat. S c i . , 1_6, 289-293 (1978). 86. Dark, W.A., McGough, R.R., J . Chromat. S c i . , 16_, 610-615 (1978). 87. Liphard, K.G., Chromatographia, 13_, No. 10, 603-606 (1980). 88. Sutoni, J.C., Garber, H.R., J . Chromat. S c i . , 14_, 546-548 (1976). 89. Suatoni, J.C., Swab, R.E., J . Chromat. S c i . , 13_, 361-366 (1975). 90. Jewell, D.M., Albough, E.W., Davis, B.E., Ruberto, R.G. Ind. Eng. Chem. Fundam. 13_(3), 279 (1974). 91. Dark, W.A., Bradford, D.L., J. Chromat. S c i . , _L5, 455 (1977). 92. "Hydrocarbon Types in Liquid Petroleum Fractions by Fluorescent Indicator Adsorption," A.S.T.M. Standards 1319, part 17 (1979). 93. Jewell, D.M., Albough, E.W., Davis, B.E., Ruberto, R.G. Ind. Eng. Chem. Fundam., 13_, 278 (1974). 94. Galya, Supra Note 81. 95. Tar courtesy of A. J a r a l l a h from spouted bed pyrolyser, 500°C, Forestburg Coal, Supra note 19. 96. Alfredson, T.F. Courtesy of Varian Associates, J . Chromat. S c i . , 218, 725 (1981). 97. Dark, W.A., McFadden, J . Chromat. S c i . , 15, 455 (1977). 98. Dark, W.A. , McFadden, J. Chromat. S c i . , Jj>_, 289 (1978). 99. Atlas of Spectral Data and Physical Contents, J.G. G r a s s e l l i (Ed.), CRC Press (1973). 100. Later, D.W., Lee, M.L., Bartle, K.D., Kong, R.C, Vassilor e s , D.L., Anal. Chem. 53, 1612-1620 (1981). 101. Borwitzky, H., Schomberg, G. J . Chromatography, 170 (1979), 99-124. 102. Novotny, M., Strand, J.W., Smith, S.L., Wiesler, D., Schwende, F.J., Fuel (1981), Vol. 60, 213-220. 103. Hirsch, D.E., Hopkins, R.L., Coleman, H.J., Cotton, F.O., Thompson, C.J., Anal. Chem., Vol. 44, No. 6 (1972), 915-918. 104. J a r a l l a h , Supra Note 19, p. 78. 105. J a r a l l a h , Supra Note 19, p. 93. 106. Bennett, CO., Myers, J.E., Momentum, Heat and Mass Transfer, 2nd ed., McGraw H i l l Inc., 1974, p. 488. 107. J a r a l l a h , Supra note 19, p. 93. 108. Supra Note 46. 109. Supra Notes 62-68. 110. Chemical Engineers Handbook, 5th Ed., R.H. Perry, CH. Chilton (ed.), McGraw H i l l Co., N.Y., 1973 , p. 9-20. 111. Organic Structure Determination, D.J. Pasto, C.R. Johnson, Prentice H a l l Inc., Englewood C l i f f s , N.J. (1969), Ch. 4 ) . 112. Gavalas, G.R., Cheong, P.H-K., Jain, R., Ind. Eng. Chem. fundam. 20 (1981), 113-122. 113. Ross, L.L., Shu, W.R. i n Thermal Hydrocarbon Chemistry, ACS #183 (1979), p. 129. 114. Anthony, D.B., Howard, J.B., A.I.Ch.E.J., 22_, No. 4 (1976), p. 625. 115. Physical Chemistry Part I I I : Physical and Chemical K i n e t i c s , R.S. Berry, S.A. Rice, J . Ross ( J . Wiley and Sons, Toronto: 1980), 1047-1080. 116. Chemistry 350 Lecture Mateial 1980, University of Western Ontario. 117. Supra Note 43, F-321. 118. Supra Note 43, F-240. 119. Supra Note 69. 120. Supra Note 115, Ch. 30, 31. 121. Nandi, S.P., Walker, P.L., i n Coal Science, p. 379, Supra Note 4. - 162 -APPENDIX A MECHANISTIC PYROLYSIS CHEMISTRY AND KINETICS A number of authors have attempted to develop k i n e t i c s models for coal and hydrocarbon pyrolysis (112, 113). Anthony and Howard (114) give an excellent review of a number of these studies. In most of the models outlined, an attempt i s made to f i t the d e v o l a t a l i z a t i o n y i e l d ( i n various forms) to an Arrhenius exponential equation: Rate = k[A] " [ B ] 3 where k = A exp(-E /RT) cL The e f f e c t s of reaction order, sequence, r e v e r s i b i l i t y and d i f f u s i o n control may be included as a perturbation of the general rate expression. This empirical approach yields valuable parameters useful f o r the conditions under which they were obtained, but not usually applicable elsewhere. An a l t e r n a t i v e approach i s a p r i o r i k i n e t i c modelling. This involves the determination of the possible reaction pathways that a given material may taken when pyrolysed. Knowledge i s required of the chemical structure and functional groups present and how they may be expected to behave when thermally activated. Obtaining a f i n a l reaction pathway i s i n i t s e l f a very complicated matter. When a workable scheme i s determined the expected ranges of a c t i v a t i o n energy and hence i n d i v i d u a l rate constants may be c a l c u l a t e d . Such values may be obtained from model compound studies, bond energies or a v a r i e t y of - 1 6 3 " -thermodynamic means. The reaction pathway and i t s energetics form an extremely useful q u a l i t a t i v e l y predictive tool. With refinement and a great deal of work, quantitative simulation may also be possible. The advantges of working out an a_ p r i o r i model are numerous. F i r s t l y i t i s a most useful predictive tool to examine the e f f e c t s of c a t a l y s t s , time, reactive additives and coal type. Secondly, such a model would be universal, applicable not only to a l l coal types but to other pyrolysates as well ( t a r s , asphalts, o i l s , gases . . . ) . Moreover, pertubations such as heat and mass transfer may be more e a s i l y superimposed on top of the chemistry. The obvious detraction from modelling pyrolysis in this way i s i t s inherent complexity. For even simple materials, the number of reactions possible can be quite high. As the chemical complexity of the material increases the number and type of reactions Increase exponentially. In this study a generalized a p r i o r i k i n e t i c model i s presented. The reaction pathways derived provide an excellent means of r a t i o n a l i z i n g various trends seen i n the pyrolysis of coal organics as well as the other hydrocarbons. A short summary i s given here, o u t l i n i n g the major points of i n t e r e s t to coal p y r o l y s i s . I t must f i r s t be pointed out that chemical k i n e t i c s are but an arm of the o v e r a l l chemical picture i n p y r o l y s i s . Thermodynamic considerations must f i r s t be met before a discussion of k i n e t i c s i s pertinent. Thermodynamically allowed reactions are those which show an ov e r a l l negative free energy change, as given by the Gibbs expression: AG = AH - T AS - 1 6 4 -From thi s , one derives that high temperatures favour entropy (AS) driven reactions. Large positive entropic changes coincide with reactions that produce small fragments having more degrees of freedom than their precursor molecules. This i s the case for many of the reactions considered, p a r t i c u l a r l y those y i e l d i n g gases CO, CO2, H20, H 2 and to a lesser extent the lower hydrocarbons. Conversely, reactions that reduce the o v e r a l l degrees of freedom, notably building aromatics from small unsaturated fragments and charring reactions, may make the reaction not allowable. In these cases the balance between enthalpy and entropy w i l l determine the permissability of the reaction. For other reactions not involving s i g n i f i c a n t entropy changes, the ove r a l l reaction enthalpy dominates. For bond cleavage, rearrangement and most bimolecular reaction this i s the case. While this may seem a useful tool to predicting the course of coal reactions, problems arise i n obtaining useful values of AH and AS for any p a r t i c u l a r reaction at a given temperature. P y r o l y s i s , as with other high energy s i t u a t i o n s , makes many new el e c t r o n i c energy l e v e l s available to a molecule from which i t may react. Enthalpy and entropy changes then take on a very complex appearance. In this study, the assumption i s made that somehow the thermodynamics work themselves out so that a l l reactions are a v a i l a b l e . Kinetics then enters to dictate how appreciable the rate of that reaction w i l l be. A general Arrhenius expression Is used for s i m p l i c i t y . The pyrolysis reaction pathway flow diagram presented e a r l i e r appear on the next page as Figure A - l . Each of the reactions has been HYDROCARBON m , k C/l u H Id Z * - i 14 c/> cn >-- J I- o < as >» u 3 a z •H Cu cn >» < 3 <: CL, z o * - i H O - 1 6 6 -l a b e l l e d with a p a r t i c u l a r reaction number ( i ) . From these w i l l be derived the rate of that p a r t i c u l a r reaction (R^) from the rate constant (k^) and the concentration of each reactant i n the matrix, [RXHO]. As a l l of these reactions are elementary steps, the conversion to the k i n e t i c expression flows from the stoichiometry of the reaction. The following derivations are then used to derive the o v e r a l l rate expressions: 1. Unimolecular reactions: ( f i r s t order) where the rate i s dependent on the concentration of one reactant only: N 2 ° 5 — ^ 2 N 0 2 + !/2 ° 2 R - ^5^51 - k[N 2 0 5 ] 2. Bimolecular reactions: (second order) where the rate i s dependent on the concentration of two species involved i n a given elementary step k f H 2 + I 2 — ^ 2 HI R = -d[H 2] . k [ H 2 ] [ I 2 ] 3. Reversible reactions: where the reaction occurs by a single bimolecular elementary step ir H 2 + 1 2 f^= 2 HI R = ~ d [ * ? ] = k e [ H 2 ] [ I 2 ] " k [HI] 2 dt f 1 - 167 -4. Simultaneous reactions: where a number of reactions may occur from the same species at the same time A — < T p - -d[HCHOl _ , ., r . d t — M i A j + k 2 l A J 5. Consecutive reactions: where an i n i t i a l reaction produces a second reactant that i t s e l f may react further 1 2 A—*- B C Rl = "kit" A] R 2 = -k^A] + R 2[B] This s i t u a t i o n results i n concentration p r o f i l e s of a, B, C that are strongly dependent on the ratios of kj and k 2. 6. Chain reactions: where some eleraentry steps may be repeated many times: 1 CH4 + F 2 ^"CH 3 + HF + F* 2 •CH3 + F 2 CH 3F + F• 3 F« + GAh ^'CHg + HF 4 CH 3 + F* ^••CH3F - 168 -then: I L ^ M L = k^CH,,] [ F 2 ] " k 2[«CH 3][F 2] + k3[F.][CH„] - k„[ .CH 3] [F •] 7. Hetero c a t a l y t i c r e a c t i o n s : where a heterogenous c a t a l y s t lowers the a c t i v a t i o n energy. The mechanism i n v o l v e s adsorption and desorption of the reactants onto the c a t a l y s t surface C 2Hit + H 2 N.r. o Ni C2Htf + H2 C2H6 then the rate i s given by d^cf^ = k [ C 2 H H ] [ H 2 ] d t (1 + ctCzH^] 2) and "c" i s a constant r e l a t e d to the degree of adsorption on that s u r f a c e . A l l of these r e a c t i o n types are found i n the r e a c t i o n pathway diagram given e a r l i e r . These expressions or s i m p l i f i c a t i o n s of them are used to derive the o v e r a l l k i n e t i c expressions f o r the major products i n a t y p i c a l p y r o l y s i s . Before doing t h i s some a d d i t i o n a l f a c t o r s must be in c l u d e d . For a heterogereous mixture such as a coal hydrocarbon or a mixture of compounds i n the gas phase, the a v a i l a b i l i t y of any p a r t i c u l a r primary r e a c t i o n depended.on the a v a i l a b i l i t y of the pe r t i n e n t f u n c t i o n a l group. For example, e l i m i n a t i o n r e a c t i o n s g e n e r a l l y i n v o l v e heteroatomic species (quinones, c a r b o x y l i c a c i d s , ketones). I f there are no such f u n c t i o n a l groups present i n the - 169 -s t a r t i n g material then this primary reaction i s not availabl e . And so, for any compound or mixture at any point i n time a d i s t r i b u t i o n of fra c t i o n s of each class of reactive functional group or bond type can be given. The fractions (nj_) must to t a l unity at a l l times. These f r a c t i o n s are represented as n D A (Diels-Alder retrogression), n^C-H cleavage), n c (C-C cleavage), (C-X cleavage), njj T ( h e t e r o c a t a l y t i c reactions, n g^x (elimination) and n ^ (chain reactions). This i s a very d i f f i c u l t concept to formulate. As the pyrolysis progresses, each of these values must change and so some system must be used to keep track of the degree of reaction of each functional e n t i t y . These fractions may be thought of as mole fr a c t i o n s or a l t e r n a t i v e l y as fractions of some ove r a l l bond inventory. To date this has been the stumbling block i n completing a f i n a l i z e d working simulation. Assuming that this can be r e c t i f i e d the remaining methods are less taxing. For a given reaction type then, the rate expression may appear as: * - • " i k i [ B 1 If a f i n i t e approximation approach i s used then this s i m p l i f i e s to: AA = N. k. [B] At i l where At i s some small increment i n time [At < K j - 1 ] . From an i n i t i a l concentration of A 0 and B Q after time At the new concentrations are - 170 -Ai = AA + A D Bi = AB + B Q At time t = 0 the i n i t i a l bourday conditions for concentration of each species may be inserted. By employing the above methods using the combined rate expressions given below, the concentration p r o f i l e s of a l l I v products can be found as a- function of tot a l time elapsed at the chosen temperature. Perturbations of time p r o f i l e s , added reactants and quenching p r o f i l e s could then be imposed on this model once a working system has been derived. From the reaction pathway diagram the rate expressions for the major products can be derived by summing over a l l reactions that involve that p a r t i c u l a r species. These expressions appear on the following .1 page. Some equations include the increment counter ( i ) . The rate constants (k^) correspond to the reaction number given in the diagram. Reversible and heterocatalytic reactions have been s i m p l i f i e d . K i n e t i c Expressions for K i n e t i c Simulation: ( 1 ) k6[RHX0] • N X Q d[H ] (2) — = k 1 4[H'] + K 2 2[H'] [RHXO] N f t - kgtRj] (3) i lML- k 1 3[H-][X-] + k 2 4[X-][RHXO] N H - 171 -d [ X 2 J , r ,2 (4) — ~ - k u [ X - r + k 2 J-0 2H][RHXO] N R - 2 k ^ l l X ^ (5) iLcharl = ^[olefin]2 + ^ [ R - ] [olefin] + k ^ R - ] 2 ( 6 ) dUHMj. m _ [ R H X Q ] _ + ^ + ^ + ^ + ^ + ^ XO + K20 [ 6' ] NH + W 1 NH + K22 V * > ] + K 2 3 [ R * ] NH . + % N H[X«] (7) k 6[RHXO] • N X Q + k 1 5[H«][R«] + k i g [ R ' ] 2 + k 2 3 [R •] [RHXO] N R (8) = k 1 2fR.][x.] then: [RHXO] N = [RHXO] N_ 1 + [RH] + [RX] (9) d [ * ' ] = k 4[RHXO] N x + 2 k 1 Q [ X 2 ] + k 2 Q [0'] (RHXO] + k 1 ? [ H ' ] [ 0 2 ] - [k 2 4[X«] [RHXO] N H + k n [ X - ] 2 + k 1 2[X'][R»] + k 1 3[X»][H»]] - 172 -k2[RHXO] N R + 2k g [H 2] + ^ [ R * ] [k 2 2[H«] [RHXO] N R + k [H'][0 2] + k 2 8 [ H • ] [ o l e f i n ] + k 1 4 [ H - ] 2 + k 1 5[H-][R-] + k 1 3[H-][X-]] [k 2 [RHXO] + 2k 3 Nc[RHXO] + k^fRHXO] + k^RHXO] NR^ , + K2 8[H*][olef] + k 1 8[R*] + k^tOHRHXO] N R + k 2 1[0 2H][RHXO] N R + K 2 2[H'][RHXO] N H + k 2 3[R'][RHXO] N R 4+ k 2 4 [ X •][RHXO] N R] - [ k 2 5 [ R - ] [ 0 2 ] + 2 k i g [ R - ] 2 + k 1 6[R.] + k l g[R.] + k ^ [ olef ] [R •] + k 2 9 [ R - ] 2 + k 1 2[R«][X«] + k 1 5[R-][H-]] The rate constants (k^) are p a r t i c u l a r l y d i f f i c u l t to derive. To do these i t i s assumed that there are no mass transfer r e s t r i c t i o n s so that a hard s h e l l k i n e t i c gas model i s assumed. Accordingly, the pre-exporential frequency factor becomes: do) I t y . . ( I D - 173 -where k D i s the Boltzraann constant, h i s Plancks constant and the c o l l i s i o n p r o b a b i l i t y , p, i s assigned a value of 0.2. The variables remaining i n the Arrhenius expression are then the temperature (or temperature - time p r o f i l e ) and the a c t i v a t i o n energy f o r the respective r e a c t i o n (Ea^). These energy values are of course not fixed but represent t y p i c a l values within the range generally observed for a reaction of the one considered. Owing to the advances i n k i n e t i c methods, vacuum technology, p y r o l y s i s and k i n e t i c techniques, many of these reaction types have been examined as model compound studies. This provides the source for the range of a c t i v a t i o n energies required in the above expression. The following table i l l u s t r a t e s a preliminary survey of energy values for each reaction type and their source. These values do not represent an exhaustive examination of the analogous reaction, energies: Reaction Number Description E (KJ mole -*) 1 retro - D.A. 75 - 300 2 C-H cleavage 350 - 420 3 C-C cleavage 350 - 440 4 C-X cleavage 300 - 430 5 het e r o c a t a l y t i c 100 - 350 6 elimination 100 - 300 8 H 2 cleavage 431 - 174- -Reaction Number Description E a ^ K ^ m o^- e ^ 10 X 2 cleavage 160 11 X« combination 2 0 - 6 0 12 r a d i c a l combinations 30 - 80 13 ra d i c a l , combinations 20 - 60 14 r a d i c a l combinations 20 - 40 15 r a d i c a l combinations 30 - 80 16 fragmentation 50 - 100 17 oxygen 10 - 50 18 fragmentation 50 - 100 19 disproportionation 30 - 90 20 chain reaction 10 - 50 21 chain reaction 10 - 50 22 chain reaction 30 - 80 23 chain reaction 30 - 100 24 chain reaction 10 - 50 25 oxygen abstraction 10 - 50 26 dimerization 175 - 250 27 addition 2 0 - 8 0 28 hydrogenation 10 - 50 29 r a d i c a l terminations 20 - 80 30 hydrogen-oxygen 0 - 2 5 - 175 -The above values were obtained l a r g e l y from bond energy tabulations, enthalpy tabulations (118), Wilcott, C a r g i l l and Sears (119), Berry Rice and Ross(120) and Nandi, Walker (121). After considerable work the above project was defered u n t i l time and energy allowed i t s completion. As can be seen, the bulk of the task has been completed i n putting together an o r i g i n a l reaction pathway diagram and derivations of the pertinent mechanistic k i n e t i c and energetic r e l a t i o n s . By examining the r e s u l t s at this stage many q u a l i t a t i v e and semi-quantitative predictions can be made, as referred to i n the body of this work. There i s a considerable amount of information contained i n these sections. I t i s recommended that the reader take the time to digest the ove r a l l picture and those areas that are of p a r t i c u l a r i n t e r e s t and relevance to their own research and in terests. - 176 -APPENDIX B: COAL FEEDER DEVELOPMENT The design of the coal feeder used i n this study i s out l ined i n s ec t i on F . This appendix deals with the c h a r a c t e r i z a t i o n and propert ies of this feeder . The coal used throughout the study of the feeder was the same Balmer coal used in the actual p y r o l y s i s experiments. The coal was sieved to a mean p a r t i c l e diameter of 100 ym with the range of p a r t i c l e s izes being 76 to 124 ym. E a r l y i n the development i t was found that inadequate s i z i n g or high shear rates on s ized p a r t i c l e s allowed f ines (< 76 ym) to enter the coal sample and led to subsequent plugging problems. Presumably this i s due to the greater cohesiveness of these small p a r t i c u l a t e s . For this reason, coal was sized twice before use and again i f i t was recyc led during the course of feeder c a l i b r a t i o n . The operation of the feeder i s quite s imple. The rotor blades create a q u a s i - f l u i d i z e d state ins ide the feeder body. Flow of n i trogen through a r e s t r i c t i o n creates a s l i g h t l y lower pressure ins ide the feeder tube compared to that wi th in the feeder body. This creates a low flow through the hole i n the side of the feed tube which, together with g r a v i t y , draws a steady amount of coal into the tube. This i s then entrained from the feeder and c a r r i e d into the coal conveying tube. For c a l i b r a t i o n of the mass rate of coal feeding the conveying tube was connected to a f o u r - l a y e r bag f i l t e r . This set-up was adopted to match a l l of the condit ions up to the p y r o l y s i s reactor as c l o s e l y as pos s ib l e . By e s t a b l i s h i n g a c e r t a i n feed condi t ion over a s u i t a b l e -'177 -length of time, the feed rate can be determined by measuring the weight change of both the bag f i l t e r and the feeder body i t s e l f . The feeder proved to be a very r e l i a b l e constant rate p a r t i c l e feeder over an extended range of flow rates, tube hole sizes and times. Attainment of steady flow occurs instantaneously once flow i s switched from the bypass l i n e to the feed tube. The use of quick-connect f i t t i n g s allowed for easy changing of feed tubes of a variety of materials ( s t e e l , copper, p l a s t i c ) i n minutes. Performance of a given tube remained the same despite repeated insertions and removals from the feeder. The feeder performance did however turn out to be dependent on a large number of variables: 1 . Rotor Three s t y l e s of rotors were t r i e d , as depicted i n Figure B - l . This figure also demonstrates the r e l a t i o n s h i p between feed rate and the mass of coal i n the feeder for each p a r t i c u l a r rotor configuration. The unusual behavior presumably indicates that feed rates are dependent on the i n t e r n a l hydrodynamic state of the feeder created by the action of the rotors. For r e l i a b l e constant feeding during a given pyrolysis experiment, i t i s evident from this figure that only the horizontal figure - 8 s t y l e rotor i s useful and again only when the coal mass i n the feeder i s greater than about 150 grams. 2. Rotor Revolution Speed As the rotor speed i s increased, the feed rate i s observed to increase l i n e a r l y up to approximately 500 RPM and then remain constant. - 178 -Figure B-1 EFFECT OF ROTOR ON FEEDING i — T — r — i — r — i — i — i i i — i i — i I I — i — i — i — r — i — i — i — i — i 1—1 45.0 80.2 115.4 150.6 185.8 221.0 256.0 MEAN UEIGHT OF CORL IN FEEDER (KGX10** ) (300 RPM, 148 cm3 m i n - 1 , tube hole 0.076 cm). - 179 -This i s shown i n Figure B-2. Similar r e s u l t s were obtained by Scott et a l . (78) though their speed did not exceed 450 RPM so that they did not observe this plateau e f f e c t . 3. Tube Angle The rotor blades were positioned i n each case to be approximately 3-4 mm. from the feed tube. The hole was placed at the mid-point of the feeder body cross-sectional chord. The hole could be oriented facing v e r t i c a l l y upward (0°) or downward (180°) or h o r i z o n t a l l y toward the shaft (270°). The feed rate was observed to deviate s l i g h t l y i n a smooth parabolic curve reaching a minimum at 180° and a maximum at 0°. This presumably r e f l e c t s the e f f e c t of gravity on the flux of p a r t i c l e s into the tube hole. Plugging of the hole was less of a problem at 0° and so a l l future work was done at that angle. 4. Nitrogen Flow Rates and Coal Mass For a given rotor, the feed rate increases as the flow rate i s increased, as shown in Figure B-3. For each rotor a similar family of curves i s generated showing the same shape, but curves are shifted v e r t i c a l l y with a change in flow. For the horizontal-8 rotor, i n the linear portion of feed rate versus coal mass region, the r e l a t i o n s h i p between flow rate and feed rate can be extracted for each tube hole s i z e . This i s shown in Figure B-3. I t was possible to f i t each of these curves to a logarithmic equation of the form - 180 -Figure B-2 ROTOR RPH V E R S U S F E E D RATE RPfl (1) horizontal-8 rotor, Balmer 100 ym, flow 120 cm3/min, tube hole s i z e : 0.051 cm. - 181 -Figure B-3 F E E D CURVES 2.64 + X o X 2.16 1.6ft-o J^1.2 OH cr or e u. 0.7 2 0.2 A-o 53 cn3/niN + "74 CH3/niN « 96 cn3/niN * 120 cri3/niN • i*e cn3/niN • 178 CH3/niN T 1 1 1 1 r — T 1 1 1 1 1 r-r-i 1 [ — T 1 1 1 1 1— 50 70 90 110 130 150 170 190 210 230 250 l r MASS OF COAL IN FEEDER (G) Notes: (1) eggbeater r o t o r , Balmer 100 pm, tube hole s i ze 0.076 cm. M = a log F + B where a, b are constants for each tube hole size, M i s the mass feed rate (g/min) and F the nitrogen flow rate. These feed c o r r e l a t i o n equations were used i n the selecti o n of tubes i n the pyrolysis i •', experiments according to the c r i t e r i a given e a r l i e r . The valve of a "a" did not correlate with the tube hole size or the cross-sectional area of the tube hole. This indicates that at any given flow rate the flux of sol i d s across the tube holes i s not constant. This may be due to the fa c t that the tube hole sizes were of the same order of magnitude as the mean p a r t i c l e diameter and so the flux may be dependent on the permissable orientations of p a r t i c l e s i n the hole opening. Visualization- of the coal feeding by using transluscent conveying tubes showed that s a l t a t i o n of coal occured up to a flow rate of v 3 — 1 approximated 60-cm min . For the tube bore size used, this corresponds to a lin e a r flow v e l o c i t y of aproximately 1.12 m/s. When sa l t a t i o n occurs, the coal p a r t i c l e s s e t t l e along the bottom of horizontal tubes creating a channel along the top of the tube bore along which p a r t i c l e s are conveyed. Plugging due to s a l t a t i o n was never observed. Coal feeding was generally uniform i n density across the tube and over the tube length above the s a l t a t i o n v e l o c i t y . When using very small tube hole sizes (< 0.041 cm), plugging of the hole was a problem, p a r t i c u l a r l y i f fines were present. From the above characterization adequate feeder performance was obtained using the horizontal 8 rotor at 300 RPM with a tube angle of 0°. Under these conditions the res u l t s i n - 183 -Table B-l are determined. For a l l of the pyrolysis runs these correlations and results were r e l i e d on i n c a l c u l a t i n g the desired feed rate, flow rates and choosing the appropriate feed tube. Table B-l Feeder Chara c t e r i s t i c s c o a l : dp = 75 -• 124 um, dp - 100 um rotor: horizontal-8, 300 RPM, 0° hole orientation tube: 0.125 cm s tainless ! s t e e l , holes at mid-point of chord Tube Hole Size D/d^ Feed Range C a l i b r a t i o n C o e f f i c i e n t s (cm) (g min - 1) b a 0.037 3.7 0.06 - 0.16 -0.164 0.055 0.041 4.1 0.06 - 0.16 -0.355 0.115 0.046 4.6 0.13 - 0.47 -0.582 0.172 0.051 5.1 0.48 - 1.71 -2.10 0.633 0.064 6.4 0.50 - 2.01 -2.53 0.764 0.076 7.6 0.60 - 3.0 - -- 184 -Figure B-4 CORL FEEDER C A L I B R A T I O N Tube Hole Size CD o 0.0145" 0 . 0 3 7 c m + + 0.0160" 0 . 0 4 1 c m - « » 0.0180" 0 . 0 4 6 c m . x x 0.0200" 0 . 0 5 1 c m 1 i i i i i — I i i i i 1 1 I r — i p — i 1 i 1 1 1 r — r 4 0 8 0 120 160 200 240 280 320 360 4 0 0 440 NITROGEN FLOU RATE (Cf13/niN) (horizontal-8 rotor, l i n e a r region of Figure B-l) v- 185 -Appendix C T r i v i a l Names and Structures of Polynuclear Aromatics and Other Coal Tar Components. benzene chrysene toluene xylenes or CLU dibenzoanthracene te trahydronap thaiene napthaiene fluorene 00 perylene benzoperylene anthracene phenanthrene coronene 1, l-,j-bin-ap:fehyl pyrene triphenylene p - b e n z o q u i n o n e 2 , 4 - d i m e t h y l p h e n o l ( 3 - n a p t h o l f l u o r e n - 9 - o n e a n t h r a q u i n o n e d i b e n z o f u r a n c a r b a z o l e acridine Appendix D Summary of Data from Pyrolysis T r i a l s and Analysis Run No. Temperature Residence Reactor Feed Rate Loading* Mass /°C Tirae/S Vol./cm 3 (kg/min) x 10 3 (kg/m3) x 10 2 FED/kg x 10 3 12 700 0.5 45 0.136 5.04 8.17 13 700 0.5 45 0.079 1.46 4.43 14 700 0.5 45 0.299 6.50 6.58 15 700 0.5 45 0.736 13.6 7.36 16 700 0.5 45 1.73 31.9 4.40 20 500 0.5 45 0.257 4.75 4.89 Run No. M.A.F. Yi e l d Residual V o l a t i l e s Product D i s t r i b u t i o n wt. % Wt. % Insolubles Polar Aromatic 12 10.04 10.53 - - -10.45 10.88 49.2 21.2 29.6 Mean 10.25 - - ' -13 13.27 8.19 45.6 34.4 19.9 13.95 8.20 48.0 27.1 25.0 Mean 13.61 46.8 30.8 22.5 14 11.07 9.02 43.9 29.2 26.9 10.53 8.83 41.4 29.8 28.8 Mean 10.79 42.7 . 29.5 27.9 15 8.95 10.75 38.0 30.6 31.4 8.93 10.79 41.5 30.5 28.0 Mean 8.94 ; 39.8 30.6 29.7 16 8.98 8.11 39.2 31.3 29.5 — 7.94 39.2 31.4 29.4 Mean 8.98 39.2 31.4 29.5 20 4.99 12.22 19.0 37.5 43.5 4.88 12.12 24.6 34.0 41.5 Mean 4.93 21.8 35.8 42.4 Run No. Net Product Yields % m.a.f. Insolubles Polar Aromatic 12 - - — 5.14 2.22 3.09 13 6.06 4.57 2.64 6.69 3.77 3.48 6.37 4.19 3.06 14 4.86 3.23 2.98 4.36 3.14 3.03 4.60 3.18 3.01 15 3.40 2.74 2.81 3.71 2.72 2.50 3.55 2.73 2.65 16 3.52 2.81 2.64 3.52 2.81 2.64 20 0.95 1.87 2.17 1.20 1.66 2.02 1.07 1.76 2.09 Run No. Temperature Residence Reactor Feed Rate Loading* Mass /°C Time/S Vol./cm 3 (kg/min) x 10 3 (kg/m3) x 10 2 FED/kg x 10 3 21 600 0.5 45 0.262 4.85 5.50 22 700 0.5 45 0.245 4.85 ' 5.15 2 3 500 1.0 54 0.189 5 . 7 3 7 . 57 24 600 1.0 54 0.144 4.37 5.18 25 700 1.0 54 0.122 3.70 4.40 28 500 2.0 54 0.073 4.43 4.70 Run No. M.A.F. Y i e l d Residual V o l a t i l e s Product D i s t r i b u t i o n wt. % Wt. % Insolubles Polar Aromatic 21 7.55 9.67 26.1 37.6 36.3 7.52 9.55 26.8 38.4 34.8 Mean 7.54 26.5 38.0 35.6 22 9.17 7.89 42.1 33.1 24.9 9.44 8.42 38.8 33.6 27.6 Mean 9.30 40.5 33.3 26.3 23 6.22 9.64 20.2 40.0 39.9 6.42 9.65 19.5 - -Mean 6.33 19.8 40.2 40.0 24 10.24 8.84 42.7 30.2 27.0 10.19 8.44 38.6 32.7 28.7 Mean 10.22 40.7 31.4 27.9 25 12.51 6.64 55.6 22.9 21.5 12.30 6.23 58.0 23.2 18.8 Mean 12.41 56.8 23.1 20.2 28 7.08 11.69 29.7 33.4 36.9 6.75 11.58 28.5 38.6 32.9 Mean 6.92 29.1 36.0 34.9 Run No. Net Product Yields % m.a.f. Insolubles Polar Aromatic 21 1.97 2.01 2.84 2.89 2.74 2.62 1.99 2.87 2.68 22 3.86 3.66 3.03 3.17 2.28 2.61 3.76 3.10 2.44 23 1.25 2.49 2.48 1.25 2.54 2.53 24 4.39 3.93 > 3.09 3.33 2.77 2.93 4.16 3.21 2.85 25 6.95 7.13 2.86 2.85 2.69 2.32 7.05 2.86 2.50 28 2.10 1.92 2.37 2.61 2.61 2.22 2.01 2.49 2.41 Run No. Temperature Residence Reactor Feed Rate Loading* Mass /°C Time/S Vol./cm 3 (kg/min) x 10 3 (kg/m3) x 10 2 FED/kg x 10 3 2 9 600 2.0 54 0.08 1 4 . 9 1 6.47 30 700 2.0 54 0.102 6.19 5.10 31 500 0.5 54 0.299 4..54 5.97 33 600 0.5 54 0.295 4.47 5.60 34 700 0.5 54 0.307 4.54 5.22 3 6 700 0 . 2 5 54 0 . 89 2 6 . 7 7 5.35 Run No. M.A.F. Yi e l d Residual V o l a t i l e s Product D i s t r i b u t i o n wt. % Wt. % Insolubles Polar Aromatic 29 9.84 43.0 30.7 26.3 9.69 46.9 28.2 24.9 Mean 9.76 44.9 29.5 25.6 30 11.62 59.0 19.7 21.3 11.43 52.9 23.2 23.9 Mean 11.53 56.0 2f.4 22.6 31 6.47 10.46 27.6 29.9 42.5 5.62 10.38 26.1 36.5 37.4 -Mean 6.04 26.9 33.2 40.0 33 9.58 10.22 41.4 32.5 26.1 9.52 — 37.1 35.2 27.7 Mean 39.3 10.22 39.3 33.9 26.9 34 11.20 9.06 62.3 21.0 16.7 11.14 — 47.1 26.4 26.6 Mean 11.17 54.6 23.9 21.4 36 12.96 7.48 51.9 26.1 22.0 12.99 7.71 49.6 28.7 21.7 Mean 12.97 50.8 27.4 21.8 Run No. Net Product Yields % m.a.f. Insolubles Polar Aromatic 29 4.23 3.02 2.59 4.54 2.73 2.41 4.39 2.88 2.50 30 6.86 2.28 2.48 6.05 2.65 2.73 6.45 2.47 2.60 31 1.78 1.94 2.75 1.47 2.05 2.10 1.62 2.00 2.41 33 3.97 3.12 2.50 3.53 3.35 2.64 3.75 3.23 2.57 34 6.97 2.36 1.87 5.24 2.94 2.96 6.11 2.67 2.39 36 6.72 3.38 2.86 6.45 3.73 2.81 6.58 3.56 2.83 Run No. Temperature Residence l°C Time/S 37 800 0.5 38 800 1.0 39 800 2.0 40 700 1.0 41 700 4.0 43 700 1.0 44 700 1.0 45 (oxygen 700 1.0 46 900 1.0 Reactor Feed Rate Loading* Mass Vol./cm 3 (kg/min) x 10 3 (kg/m3) x 10 2 FED/kg x 10 3 54 0.248 4.59 4.47 54 0.191 5.79 4.00 54 0.074 4.4 9 2.65 54 0.072 2.18 2.66 54 0.035 3.12 2.43 54 0.159 4.82 5.25 54 0.178 5.40 4.80 54 0.174 5.28 4.70 54 0.173 5.79 4.00 Run No. M.A.F. Yi e l d Residual V o l a t i l e s Product D i s t r i b u t i o n wt, % Wt. % Insolubles Polar Aromatic 37 11.48 5.42 56.0 21.6 22.4 11.84 6.12 55^7 22_i2l 22.0 Mean 11.65 55.9 21.9 22.2 38 10.19 5.22 53.3 22.3 24.4 10.69 5.24 52^0 23_;9 24.1 Mean 10.43 5.23 52.6 23.1 24.3 39 8.34 6.50 41.6 31.0 27.4 8.03 5.95 52.5 22.5 25.0 . Mean 8.19 47.1 26.8 26.2 40 12.93 7.02 48.0 36.3 15.7 11.30 6.90 59.3 27.2 13.6 Mean 12.11 53.6 31.7 14.6 41 9.93 5.82 45.7 30.5 23.8 9.80 6.14 34 .4 42.8 22.8 Mean 9.87 40.1 36.7 23.3 43 11.56 8.53 50.3 29.8 19.9 11.11 8.62 50.2 27jA 22.7 Mean 11.33 50.3 28.4 21.3 44 11.79 6.01 55.2 25.7 19.1 11 .89 5.42 58^5 23^5 18.0 Mean 11.84 56.8 24.6 18.6 45 11.49 8.18 54.8 25.6 19.6 11 .68 7.56 5K3 26_i7 22.0 Mean 11.58 53.1 26.2 20.8 46 7.79 1.51 49.0 20.1 30.9 7.52 2.07 49.2 1^3 31.5 Mean 7.65 49.1 19.7 31.2 Run No. Net Product Y i e l d s % m.a.f. Insolubles P o l a r Aromatic 37 6.43 2.48 2.57 6.59 2^64 2.60 Mean 6.51 2.56 2.59 38 3.43 2.27 2.49 5.56 2.56 2.58 Mean 5.49 2.41 2.53 39 3.47 2.59 2.28 4.22 K 8 J 2.01 Mean 3.85 2.19 2.15 40 6.21 4.69 2.03 6.70 3.07 1.53 Mean 6.50 3.84 1.77 41 4.54 3.03 2.37 2 3^37 4.19 2.24 Mean 3.95 3.62 2.30 43 5.82 3.44 2.30 5.58 3.01 2.52 Mean 5.70 3.22 2.41 44 6.50 3.03 2.26 6.95 2.80 2.14 Mean 6.73 2.91 2.20 45 6.30 2.94 2.25 5^99 3A2 2.57 Mean 6.14 3.03 2.41 46 3.82 1.57 2.40 3.70 K45 2.37 Mean 3.76 1.51 2.39 - 199 -APPENDIX E: HEAT TRANSFER AND COAL PARTICLE TEMPERATURE PROFILES An o v e r a l l heat balance on a coal p a r t i c l e can be performed using the following expression (for spherical p a r t i c l e s ) : , T 3h(T - T) 3a F 1 0 e , , J v p (1 - X) c % = i + i i - I (T - T ) - AH p 4* c p dt r r p c dt c y This includes the combined e f f e c t s of convection and ra d i a t i o n . If the heat of pyrolysis i s assumed to be n e g l i g i b l e (AH ->• 0) and the o v e r a l l conversion i s assumed to be zero (X •>• 0) the slowest temperature r i s e time p r o f i l e can be determined. The o v e r a l l heat transfer c o e f f i c i e n t (h) i s calculated from a general c o r r e l a t i o n for laminar flow past immersed spheres as: Nu = a + b R e 1 / 2 P r 1 / 3 hd U p d 1/2 Cp P 1/3 where T — E . = 2.0 + 0.60 ( — — g — ( — S ) k y x k g g The gas density ( p g ) , heat capacity ( C p ) a n d thermal conductivity (kg) are a l l functions of temperature for which appropriate c o r r e l a t i o n s may be found. By solving the above equations by a f i n i t e approximation sequence using i t e r a t i o n s of small time increments (At) the temperature r i s e times may be obtained. Some t y p i c a l r e s u l t s are given i n Table E - l . - 200 -Table E - l P a r t i c l e temperature r i s e times* gas H 2 N2 dp (pm) 100 14 100 200 55 320 500 300 1500 *time calculated to reach within 5% of pyro l y s i s temperature (Tp) where T p = 900K. The above res u l t s are i n s t r u c t i v e when considering residence time/temperature p r o f i l e s and the e f f e c t of hydropyrolysis on p a r t i c heating. - 201 -APPENDIX F Table F - l : P.O.S.A. Di s t r i b u t i o n s i n some Petroleum Fuels wt % Asphaltenes Polar Aromatic Saturate gasoline < 1 l i g h t gas o i l < 2 heavy gas o i l 0 - 1 0 asphalt 20 - 30 < 1 0 - 1 5 0 - 1 0 1 5 - 3 0 10 - 20 15 - 30 20 - 30 20 - 30 85 - 100 60 - 85 40 - 75 20 - 30 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0058901/manifest

Comment

Related Items