UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Cyclone scale-up and radial gas concentration profiles 1990

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
UBC_1991_A7 E53.pdf [ 6.33MB ]
Metadata
JSON: 1.0058775.json
JSON-LD: 1.0058775+ld.json
RDF/XML (Pretty): 1.0058775.xml
RDF/JSON: 1.0058775+rdf.json
Turtle: 1.0058775+rdf-turtle.txt
N-Triples: 1.0058775+rdf-ntriples.txt
Citation
1.0058775.ris

Full Text

CYCLONE SCALE-UP AND RADIAL GAS CONCENTRATION PROFI BY RANDY W. ENGMAN B.A.Sc., The University of Calgary, 1986 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 standard THE UNIVERSITY OF BRITISH COLUMBIA September 1990 ©Randy W. Engman, 1990 \ In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT A two part study was undertaken to expla in the performance of cyclones operated in c i r c u l a t i n g f l u i d i z e d bed combustion (CFBC) systems. In the f i r s t p a r t , c o l l e c t i o n e f f i c i e n c y tests were performed on a one-ninth scale p o l y a c r y l i c cyclone model of the i n d u s t r i a l scale cyclone at the 22 MWe CFBC f a c i l i t y at Chatham, New Brunswick. Emphasis was placed on scale-up . c o n s i d e r a t i o n s , loading e f f e c t s , i n l e t geometry e f f e c t s , and flow v i s u a l i z a t i o n t r i a l s . Experiments were performed at room temperature with i n l e t v e l o c i t i e s between 3.7 and '5.5 m/s, so l ids loading between 0.05 and 7.5 mass sol ids/mass a i r with two d i f f e r e n t so l id s systems. There was d i sappoint ing agreement between the resu l t s from the Chatham u n i t , scaled according to Stokes Number s c a l i n g , and the f indings obtained from the cold model u n i t . There was a minimum in the p a r t i c l e c o l l e c t i o n e f f i c i e n c y for p a r t i c l e s of diameter 2.5 to 3.0 um, apparently associated with agglomeration ef fects in the cyclone. P a r t i c l e c o l l e c t i o n e f f i c i e n c y was found to increase with increased p a r t i c l e loading for the condit ions s tud ied . Changes in the i n l e t geometry gave inconc lus ive r e s u l t s . The experimental re su l t s were l imi t ed by problems associated with feeding and r e c y c l i n g the f ines s o l i d s system used. — i i i — In the second part r a d i a l gas concentrat ion p r o f i l e s of a secondary cyclone serving the UBC p i l o t scale C i r c u l a t i n g F l u i d i z e d Bed Combustor were performed at temperatures of about 870 <>C. Concentrations of O2 , CO2 , NOx , CH4 , CO and SO2 were measured. An increase in [CO], and to a lesser extent [ C O 2 ] , was measured near the cyclone w a l l . There appeared to be l i t t l e r a d i a l v a r i a t i o n in the concentrat ion of other spec ies . Further work i s required to allow the cold model to operate cont inuously , with p a r t i c l e s which can be fed more f r e e l y , and to obtain r a d i a l gas concentrat ion p r o f i l e s wi th in the primary cyclone of the UBC CFBC system. - i v- TABLE OF CONTENTS Page ABSTRACT i i LIST OF TABLES v i LIST OF FIGURES v i i ACKNOWLEDGMENTS x i INTRODUCTION 1 PART I COLLECTION EFFICIENCY TESTS 4 1.1 BACKGROUND AND THEORY 5 1.1.1 Introduct ion 5 1.1.2 Dimensional Analys i s and Phys ica l S i m i l a r i t y 6 1.1.3 F l u i d - P a r t i c l e Separation Cyc lon ic Separators 10 1.1.4 Previous Sca l ing Work 15 1.1.5 Loading Ef fec t on Cyclone C o l l e c t i o n E f f i c i e n c y 22 1.1.6 Summary 32 1.2 EXPERIMENTAL APPARATUS AND PROCEDURE 33 1.2.1 Introduct ion 33 1.2.2 Model Cyclone Apparatus 33 1.2.3 P a r t i c u l a t e Sol ids 45 1.2.4 Data A c q u i s i t i o n and Analys i s 45 1.2.5 Error S e n s i t i v i t y 50 1.2.6 Chatham Cyclone Data 53 1.3 RESULTS and DISCUSSION 59 1.3.1 Sca l ing Cons iderat ion 64 1.3.2 Loading Ef fec t 74 1.3.3 Inlet Modi f i ca t ions 77 1.3.4 Flow V i s u a l i z a t i o n 83 1.4 CONCLUSIONS AND RECOMMENDATIONS 85 - v - PART II HOT CYCLONE TESTS 87 2.1 INTRODUCTION 88 2.2 THEORY 88 2.3 APPARATUS AND DATA ACQUISITION 95 2.4 RESULTS AND DISCUSSION 101 2.5 CONCLUSIONS AND RECOMMENDATIONS 110 Nomenclature 112 References 116 Appendix 120 -v i - L i s t of Tables Page Table 1.1 L i s t of important parameters 7 Table 1.2 Chatham operating condit ions for 58 A p r i l 17, 1990. References as i n d i c a t e d . Table 1.3 Experimental data . Note that cyclone 62 conf igurat ion information can be found in Figure 1.10. ( V . F . = vortex f inder posi t i o n ) . Table 1.4 Cyclone Operating cond i t ions . 70 Table 1.5 P a r t i c l e loading data 74 Table 1.6 Inlet modi f i ca t ion tests 80 Table 2.1 A n a l y t i c a l instrument d e s c r i p t i o n 98 Table 2.2 UBC CFBC operating condit ions 101 - v i i - LIST OF FIGURES FIGURE 1.1a FIGURE 1.1b C o l l e c t i o n E f f i c i e n c y vs . (NRBP) ( N sT) 0 . 5 Condi t ions : D = 50 mm, temperatures between 20 and 693 ° C , pressures between 140 and 2500 kPa, i n l e t v e l o c i t i e s between 0.18 and 5.2 m/s, dust loadings between 0.04 and 9.56 g/m 3• (11) C o l l e c t i o n e f f i c i e n c y curves with "f i sh hook" shape for primary and secondary cyclones . Condi t ions : D = 1.2 m, temperatures between 640 and 9 0 0 ° C , i n l e t v e l o c i t i e s between 16.3 and 27.4 m/s, dust loadings between 1.5 and 140 g/m 3 . (12) page 18 19 FIGURE 1.2 a, Gas and p a r t i c l e flow across imaginary c y l i n d e r . (21) FIGURE 1.2 b. Experimental grade e f f i c i e n c y as a funct ion of ( S ) ° * 8 ( dimensionless p a r t i c l e diameter ) for i n d u s t r i a l s i zed cyclones in a i r at room temperature (21). 21 23 FIGURE 1.3 FIGURE 1.4 Figure 1.5 Loading ef fect on cyclone c o l l e c t i o n e f f i c i e n c y . 25 Cyclone diameter up to 3.65 m diameter. Eo is zero load ( less than 1 g r a i n / f t 3 ) loading , E L is higher loading e f f i c i e n c y . (18) Predic ted separat ion e f f i c i e n c y of f ine 26 p a r t i c l e s swept out 23 of the gas by large p a r t i c l e s due to agglomeration (24). Condit ions as s ta ted . L e i t h and L i c h t c o r r e l a t i o n s with and 31 without loading and s a l t a t i o n c o r r e c t i o n s . Condi t ions : D = 0.05 and 0.91 m, temperatures between 20 and 850 ° C , i n l e t v e l o c i t i e s between 1.8 and 46 m/s, dust loadings from 0.43 and 4450 g/m 3 . (23) - v i i i - FIGURE 1.6 Figure 1.7 Figure 1.8 Figure 1.9 Figure 1.10 UBC apparatus schematic through B38. used in runs Bl Figure 1.11 Figure 1.12 FIGURE 1.13 Figure 1.14 Figure 1.15 Figure 1.16 Figure 1.17a Figure 1.17b Figure 1.17c Figure 1.18 Plot of blower speed vs f low/pressure drop. (Pumps and Power I n c . ) . Diagram of UBC so l id s hopper used in runs Bl through B35 Diagram of UBC model reactor top and model cyclone. Entrance geometry conf igurat ions for i n l e t modi f i ca t ion t e s t s . Shaded areas show i n s e r t s . A l l dimensions in mm. See sec t ion 1.3.3 for d e t a i l s . Size d i s t r i b u t i o n of feed s o l i d s . ( 22 um s i ze ) Photograph of FCC test s o l i d s ( 22 micron diameter ) . C o l l e c t i o n e f f i c i e n c y error estimates for equations 1-29, 1-30, 1-31, and 1-32. Schematic of Chatham CFB B o i l e r (27) Photograph of f ines from Chatham f l u i d bed heat exchanger, sampled on A p r i l 17, 1990. Catch hopper mass vs time i n d i c a t i n g steady state so l id s feed ra te . Run B4. Catch p a r t i c l e s i ze d i s t r i b u t i o n for run B4 P a r t i c l e s i ze ana lys i s by Elzone analys i s machine. Catch p a r t i c l e s i ze d i s t r i b u t i o n for run BIO P a r t i c l e s i ze ana lys i s by image analyzer . C o l l e c t i o n e f f i c i e n c y for runs BIO and B4. Run B4 p a r t i c l e s i ze d i s t r i b u t i o n s determined by Elzone p a r t i c l e ana lys i s instrument. Run BIO p a r t i c l e s i ze d i s t r i b u t i o n by image analys i s methods. Condit ions as s tated in Table 1.3. a.) P a r t i c l e s i ze d i s t r i b u t i o n s from Chatham f l u i d bed heat exchangers sampled A p r i l 17, 1990. b . ) C o l l e c t i o n e f f i c i e n c y curve der ived from p a r t i c l e s i ze d i s t r i b u t i o n s . E = ( C l t / ( C l t + L i t ) ) 34 36 37 39 40 46 47 52 54 56 61 63 68 - i x- Figure 1.19 C o l l e c t i o n e f f i c i e n c y curves for run B4, the Chatham cyclone and the Chatham cyclone s h i f t e d according to Stokes Law s c a l i n g . Sca l ing condit ions as per Table 1.4. 69 Figure 1.20 Figure 1.21 Figure 1.22 UBC and Chatham data p lo t t ed accordim to Abrahamson and A l l e n c o r r e l a t i o n s . a. ) UBC data (run B4 condit ions) b. ) Chatham data . See Table 1.4 operating condi t ions . for 71 UBC Run B4 and Chatham data compared with Parker et a l . data . Condit ions as stated in Table 1.4 Loading ef fect on c o l l e c t i o n e f f i c i e n c y ( UBC data ) . Vi = 5.0 m/s, T = 21 <>C, P = 1 atm. 72 75 Figure 1.23 Figure 2.1 Figure 2.2 FIGURE 2.3 FIGURE 2.4 FIGURE 2.5 Comparison of c o l l e c t i o n e f f i c i e n c i e s for d i f f e r e n t i n l e t conf igurat ions . See Table 1.3 for cond i t ions . ( s o l i d s loading corrected ) . 81 Predic ted p a r t i c l e t r a j e c t o r i e s in a v e r t i c a l plane wi th in a Stairmand type Low loading cond i t i ons . (25) Mean p a r t i c l e t r a j e c t o r i e s for p a r t i c l e s of diameter 1 to 10 microns. Mean p a r t i c l e t r a j e c t o r i e s , 3 micron. Random p a r t i c l e t r a j e c t o r y of 2 microns p a r t i c l e i n turbulent flow, 90 cyclone, a. b. c. Predic ted gas flow patterns in a Stairmand type cyclone. Low loading condit ions (25). Predic ted combined a x i a l and r a d i a l v e l o c i t y vector diagram in a Stairmand type cyclone. Low loading condit ions(25) UBC CFBC schematic (29). Scale drawing of secondary cyclone of UBC CFBC system. 92 93 94 97 FIGURE 2.6 Gas sampling system serving UBC CFBC system. 99 -x - FIGURE 2. 7 Gas concentrat ion p r o f i l e s for run 17. 103 FIGURE 2. 8 Gas concentrat ion prof i1es for run 18. 104 FIGURE 2. 9 Gas concentrat ion prof i1es for run 5. 105 Figure 2. 10 Gas concentrat ion prof i1es for run 6. 106 Figure 2. 11 Gas concentrat ion p r o f i l e s for run 10. 107 FIGURE A l Figure A2 Figure A3 Figure A4 Figure A5 Figure A6 Figure A7 APPENDIX FIGURES 120 Shake down test summary. A l l runs performed 121 before run B l . Pressure drop vs a i r flow for o r i f i c e p l a t e . 122 Data logging program serving UBC model cyclone 123 apparatus. Schematic of attempted recyc le system schematic 124 showing a high s o l i d s loading feed and measurement vesse l s , mul t ic lone and bag f i l t e r arrangements. P a r t i c l e s i ze d i s t r i b u t i o n s for runs B4 and B10 125 Run B4 p a r t i c l e s i ze analys i s by Elzone ana lys i s instrument. Run B10 p a r t i c l e s i ze ana lys i s by image analys i s methods. UBC CFBC s o l i d fuel analys i s for run B17 (29). 126 Mass balance, as performed on a per channel basis 127 for run B10. Image analys i s p a r t i c l e s i ze d i s t r i b u t i o n s . Fines loss (below 15 microns) a t t r i b u t e d to f i l t e r i n e f f i c i e n c y . Figure Figure A8 A9 Temperature data for Part I experiments. Data summary for Parts I and I I . 128 129 -x i - ACKNOWLEDGMENTS I would l i k e to acknowledge the immense help of my advisors Dr . John R. Grace, Dr . C . J . Lim, and Dr. C l i v e M.H. Brereton. Their time, energy, and personal commitment was centra l to the completion of th i s work. As well the s t a f f of the Department of Chemical Engineer ing , The Pulp and Paper Center and other employees of the U n i v e r s i t y of B r i t i s h Columbia provided c o l o r f u l support throughout the progress of th is work. The f i n a n c i a l support of Energy, Mines and Resources, and The New Brunswick E l e c t r i c Power Commission is grea t ly acknowledged. F i n a l l y I would l i k e to thank my parents without whom th i s thes is would not have been completed. i " 1 " INTRODUCTION Cyclon ic separators , genera l ly simply c a l l e d cyclones , have been used in gas c leaning operations for well over a century l arge ly because they offer good p a r t i c l e c o l l e c t i o n e f f i c i e n c i e s under extreme and varying condit ions and are simple to des ign, f a b r i c a t e , and operate. A good overview of various designs, app l i ca t ions can be found i n reference 33. Although they have been replaced by more e f f i c i e n t devices in many p o l l u t i o n contro l a p p l i c a t i o n s , cyclones have been the subject of renewed interes t for high temperature so l ids -gas separat ion in combined power cycles and within C i r c u l a t i n g F l u i d i z e d Bed Combustor (CFBC) a p p l i c a t i o n s . It was because of p a r t i c l e c o l l e c t i o n e f f i c i e n c y problems encountered in one such a p p l i c a t i o n ( 1 ) , the Chatham C i r c u l a t i n g F l u i d i z e d Bed Demonstration P r o j e c t , a 22 MWe CFBC e l e c t r i c generation f a c i l i t y in New Brunswick, that the present study was undertaken. In sp i te of the s i m p l i c i t y of cyc lon ic separators , i t i s s t i l l not poss ib le to predic t from fundamental p r i n c i p l e s p a r t i c l e c o l l e c t i o n e f f i c i e n c i e s for a l l geometries and operat ing condi t ions . A great deal of work has been done to develop equations and models to pred ic t cyclone performance for c e r t a i n standard geometries ( Eg. Stairmand and Lappel designs) under condit ions of low s o l i d s loading and low temperature. However, l i t t l e work has been done to v a l i d a t e these equations and models for extreme condit ions of temperature and p a r t i c l e loading . A l s o , l i t t l e work has been done to v e r i f y the models - 2 - for cyclones of large i n d u s t r i a l s ca l e . Thus there is a need to e s t a b l i s h s c a l i n g c r i t e r i a to predict the operat ion of high temperature, high loading , large cyclones from the operat ion of lab scale models. This thes is is d iv ided into two p a r t s . Part A deals with c o l l e c t i o n e f f i c i e n c y tests and considerat ions of sca le -up . Part B considers measured combustion gas concentrat ion p r o f i l e s wi th in a high temperature cyclone serving a p i l o t scale CFBC system and was intended to provide data for further research in the CFB f i e l d . The object ives of the two parts are as fo l lows: Part I . C o l l e c t i o n e f f i c i e n c y and scale-up s tudies . F i r s t l y , to examine scale-up considerat ions by comparing the performance of a co ld model cyclone to a large i n d u s t r i a l high temperature cyclone. Secondly, to demonstrate the effect of s o l i d s loading on s o l i d s capture e f f i c i e n c y by means of laboratory experiments on a co ld model cyclone. T h i r d l y to examine the ef fect of i n l e t modif icat ions on capture e f f i c i e n c y . L a s t l y to perform v i s u a l i z a t i o n of p a r t i c l e flows wi th in a cyclone. Part I I . Gas concentrat ion p r o f i l e s wi th in a secondary cyclone of a CFBC. To measure and report combustion gas concentrat ion p r o f i l e s wi th in a secondary cyclone of a p i l o t scale CFBC f a c i l i t y . - 3 - For each of parts A and B, a d i scuss ion of previous re la ted work and theory is followed by a b r i e f d e s c r i p t i o n of the apparatus and experimental procedure. Experimental f indings are then presented and discussed and conclusions are drawn. - 4 - PART I COLLECTION EFFICIENCY TESTS i 1.1 BACKGROUND AND THEORY 1.1.1 I n t r o d u c t i o n C y c l o n e s a r e examples of i n e r t i a l s e p a r a t i n g d e v i c e s . The p a r t i c l e c o l l e c t i o n e f f i c i e n c y , t h a t i s the mass r a t i o of p a r t i c l e s caught to tho s e f e d w i t h i n a g i v e n s i z e r a n g e ( 6 ) , i s o f t e n e x p r e s s e d by a c o l l e c t i o n or grade e f f i c i e n c y c u r v e . I d e a l l y the d e s i g n e r can p r e d i c t c y c l o n e p e r f o r m a n c e , and thus the c o l l e c t i o n e f f i c i e n c y c u r v e f o r a l l s i z e s of p a r t i c l e s from f i r s t p r i n c i p l e s , or l a c k i n g t h a t , from a c c u r a t e e m p i r i c a l r e l a t i o n s h i p s . U n f o r t u n a t e l y t h i s i s not always p o s s i b l e , as the gas and p a r t i c l e b e h a v i o u r i s not w e l l enough u n d e r s t o o d t o p r e d i c t p a r t i c l e t r a j e c t o r i e s under a l l c i r c u m s t a n c e s . Nor i s i t always p o s s i b l e t o r e l y on e m p i r i c a l r e l a t i o n s h i p s , as th e y are d e s i g n - s p e c i f i c and o f f e r good a c c u r a c y o n l y when s t a n d a r d d e s i g n s a r e c o n s i d e r e d . See r e f e r e n c e 6 f o r a c o m p a r i s o n of p u b l i s h e d d a t a and e m p i r i c a l r e l a t i o n s h i p s . Thus t h e r e i s a need f o r c o l d m o d e l i n g and f o r v a l i d s c a l i n g c r i t e r i a ( 1 3 ) . The work d e s c r i b e d i n t h i s t h e s i s i s of b e n e f i t f o r the c o n s i d e r a t i o n of l a r g e , h i g h t e m p e r a t u r e and n o n - s t a n d a r d c y c l o n e s w i t h t a n g e n t i a l i n l e t s . - 6 - 1.1.2 DIMENSIONAL ANALYSIS AND PHYSICAL SIMILARITY. Applying the p r i n c i p l e s of dimensional analys i s to the problem of cyclone p a r t i c l e c o l l e c t i o n performance requires a complete l i s t of the phys ica l quant i t i e s c o n t r o l l i n g the fate of p a r t i c l e s wi th in a cyclone. These parameters can be d iv ided into four groups: ( i ) those descr ib ing the cyclone i t s e l f , ( i i ) the operating parameters, ( i i i ) propert ies of the p a r t i c l e s being separated, and ( iv ) propert ies of the gas which c a r r i e s the s o l i d s . Table 1.1 l i s t s the most important v a r i a b l e s . - 7 - Table 1.1 L i s t of impor Cyclone dimensions: Body diameter I n l e t depth I n l e t width O u t l e t diameter O u t l e t l e n g t h C y l i n d e r l e n g t h O v e r a l l height Bottom diameter Operating parameters: I n l e t v e l o c i t y Vi Gas s p l i t r a t i o Qc Loading R a t i o L P R e l a t i v e A e / A * a c c e l e r a t i o n . ( c e n t r i f u g a l / g r a v i t a t i o n a l ) - P a r t i c u l a t e c h a r a c t e r i s t i c s : P a r t i c l e diameter dp P a r t i c l e d e n s i t y fp Shape f a c t o r (T> Gas c h a r a c t e r i s t i c s : Gas d e n s i t y j>% Gas v i s c o s i t y Uc tant parameters D a b Do Lo Lc j Ho Db -8- The seventeen var iab les l i s t e d above do not provide a complete l i s t , as other var iab les may be important. For example surface roughness may play a ro le and the smooth p o l y a c r y l i c surface of the scale model is not s i m i l a r to the r e f r a c t o r y l i n i n g of the high temperature Chatham cyclone. This is neglected. I f , however, the dimensional ana lys i s i s l i m i t e d to one geometric conf igurat ion or cyclone des ign, only one of the f i r s t eight need be considered s ince geometric s i m i l a r i t y appl ies between the co ld model and the Chatham cyclone being modeled. With ten i n d i v i d u a l var iab le s and three fundamental dimensions (mass, length, t ime), there are seven independent dimensionless groups r e q u i r i n g separate cons idera t ion . The analys i s can be s i m p l i f i e d further i f one assumes that the . s p l i t r a t i o , i . e . the f r a c t i o n of gas leaving by way of the underflow, is n e g l i g i b l e and that there is s i m i l a r i t y of shape factor for the s o l i d s systems cons idered. In a d d i t i o n i f the r e l a t i v e acce l era t ion is large in both cases ( i . e . A c / A e > 10 ) th i s factor can be neglected as w e l l . With these s i m p l i f i c a t i o n s , four dimensionless groups are requ ired . The most Common independent groupings of the remaining var iab les l i s t e d above are: Flow Reynolds number NREf = PeDVi/u K ( 1 - 1 ) Density r a t i o Np = P P / p « ( 1 - 2 ) Stokes number N 8 t = d p 2 V i P p / ( D u l t ) ( 1 - 3 ) Loading r a t i o Lp mass s o l i d s flow mass gas flow ( 1 - 4 ) In add i t ion to maintaining geometric s i m i l a r i t y i t is des i rab le to have kinematic and dynamic s i m i l a r i t y . Kinematic s i m i l a r i t y is s i m i l a r i t y in motion, which implies that the paths that representat ive p a r t i c l e s fol low are geometr ica l ly s i m i l a r and are t r a v e l l e d in a cons i s tent , scaled per iod of time ( 2 0 ) . Thus p a r t i c l e acce lerat ions must also be s i m i l a r ( 2 0 ) . Dynamic s i m i l a r i t y involves s i m i l a r i t y of forces . In order that the two systems under comparison be dynamically s i m i l a r , the magnitude of forces at each point must also be s i m i l a r ( 2 0 ) . It is commonly impossible to s a t i s f y a l l requirements s imultaneously. Thus ones of lesser importance are often s a c r i f i c e d in order to assure s i m i l a r i t y of o thers . In cyclones operating at high flow Reynolds numbers ( i . e . high i n l e t v e l o c i t i e s ) , c e n t r i f u g a l , as opposed to g r a v i t a t i o n a l , are thought to dominate the motion of smaller p a r t i c l e s . In s i tua t ions where the viscous and i n e r t i a l forces are most s i g n i f i c a n t the flow Reynolds number may be used to compare experimental observat ions , provided the condit ions of geometric s i m i l a r i t y are met ( 2 0 ) . However, i t would be incorrec t to assume that , after the onset of turbulent flow i n cyclones , c o l l e c t i o n e f f i c i e n c y is independent of Reynolds number. In fact the contrary was found, i . e . c o l l e c t i o n e f f i c i e n c y is found to vary with the h e l i c a l turbulent i n t e n s i t y ( 3 4 ) . Thus the Stokes number, combined with geometric and Reynolds number s i m i l a r i t y , may be used in the s c a l i n g process . - 10 - 1.1.3 FLUID - PARTICLE SEPARATION CYCLONIC SEPARATORS Before reviewing previous works on the subject of scale up of cyclone performance, i t is useful to develop an equation descr ib ing p a r t i c l e capture, in order to demonstrate the relevance of the key parameters. This s i m p l i s t i c d e s c r i p t i o n makes some questionable assumptions and is only intended to introduce the ef fects of various parameters on c o l l e c t i o n ef f i c i ency. Consider the fate of a p a r t i c l e of diameter d c r i t that is caught with 50% e f f i c i e n c y in a cyclone of diameter D with an tangent ia l entrance way having width b and height a. We make the fo l lowing assumptions (5): 1. P a r t i c l e s move independently of one another. 2. The drag on the p a r t i c l e can be descr ibed by Stokes law regime express ion. . 3. Buoyancy and g r a v i t y ef fects are n e g l i g i b l e . 4. The tangential v e l o c i t y of the p a r t i c l e is constant and equal to the i n l e t v e l o c i t y . 5. Secondary ef fects such as re-entrainment from the wa l l s , eddy currents , e tc . are n e g l i g i b l e . 6. Laminar flow condi t ions . 7. Relaxat ion time is n e g l i g i b l e . 8. P a r t i c l e s separate at constant v e l o c i t y 9. Once at the wall p a r t i c l e s have n e g l i g i b l e chance of reentrainment. - 11 - 10. P a r t i c l e s must reach the wall by moving across a gas stream, which reta ins i t s shape after entering the cyclone. Upon entering a cyclone p a r t i c l e s are acted on by a cen tr i fuga l force equal to F c = m V i 2 / R where: m = p a r t i c l e mass Vi= tangent ia l p a r t i c l e v e l o c i t y R = r a d i a l coordinate of the p a r t i c l e . F c = ttdorit3pPVi2/6R cen tr i fuga l force (1-5) This force accelerates the p a r t i c l e towards the w a l l . Opposing th i s motion is the drag force: F D = (Co A P ps V p 2 ) / 2 (1-6) where V p is the p a r t i c l e ' s r a d i a l v e l o c i t y component. The drag c o e f f i c i e n t can be obtained by neglect ing i n e r t i a l terms in the Navier - Stokes equation for a r i g i d sphere in an unbounded f l u i d (5) i . e . C D = drag c o e f f i c i e n t = 2 4 / N R e p = 24ps / (pg V p d c r i t ) (1 -7) * In th is s i m p l i f i e d explanat ion, the p a r t i c l e must t rave l a r a d i a l distance b during a residence time defined by the time needed to t rave l a c i r c u l a r path of distance nDN in order to be - 12 - c o l l e c t e d . If not i t w i l l enter a zone near the bottom of the cyclone where the r a d i a l gas v e l o c i t y is much higher than in the separat ion zone ( See Figure 2.3 ) . The p a r t i c l e motion is thus: IT D N • Subs t i tu t ing for Co, A P = rr*dc r i t 2 / 4 , the drag force is seen to be: FD = 3j»K d c r i t Vi b/DN Stokes law res i s tance (1-8) where: b = entrance width N = Number of revolut ions Equating expressions 1-5 and 1-8 allows p r e d i c t i o n of p a r t i c l e cut diameter as fo l lows: F c = F D ( l -8a) HtUri t 3f> PVi 2 / 6 R = 3^8 dc r i t Vi b/DN ( l -8b) d e n t 2 = 18u,tbR/(tIJ>pDNVi ) ( l -8c ) l e t t i n g R = D/2 - b/2 such that b = D - 2R d e r i t 2 = 18uK R(D - 2R)/(WJ>PDNVi ) ( l -8d) d e n t 2 = 18u e R ( l - 2R/D)/(Kp P NVi ) ( l -8e) or d e r i t = 3f 2u,R ( 1 - 2R/D )]o.s ( i _ 9 ) [ ttpP Vi N ] i - 13 - This equation has general ly the r ight form as cyclone experiments have shown (6) that c o l l e c t i o n e f f i c i e n c y improves with : 1. Increasing p a r t i c l e dens i ty . 2. Increasing cyclone vortex speed ( i . e . Vi ) 3. Decreasing cyclone diameter. However as one would expect, such s i m p l i f i e d descr ip t ions have l i m i t a t i o n s and do not describe performance for a l l cond i t ions . The effect that each parameter has been found to have on c o l l e c t i o n e f f i c i e n c y w i l l now be considered. P a r t i c l e loading is considered separately in sec t ion 1.1.4. Inlet v e l o c i t y C o l l e c t i o n e f f i c i e n c y has been found experimental ly to increase with gas i n l e t v e l o c i t y in several studies - 14 - (7,8,9). Increasing i n l e t v e l o c i t y increases tangent ia l v e l o c i t y thereby improving separat ion . These improvements are not without l i m i t , however, as secondary ef fects such as p a r t i c l e re-entrainment and eddy currents work to offset c o l l e c t i o n e f f i c i e n c y at higher Vi . The length of the gas vortex which exis ts below the vortex f inder is a funct ion of i n l e t v e l o c i t y and cyclone geometry . The vortex end is experimental ly defined as the p o s i t i o n where the vortex core seeks the wall (16). Increasing Vi lengthens the vortex u n t i l the end in ter feres with the wall causing re-entrainment of the p a r t i c l e s already caught, thus causing a reduct ion in c o l l e c t i o n e f f i c i e n c y . , T h e observed peak in curves of e f f i c i e n c y vs Vi have been success fu l ly delayed with the add i t ion of an inverted cone, located a x i a l l y above the so l id s exit (14, 15). The inverted cone is thought to anchor the vortex end and prevent i t from reentra in ing p a r t i c l e s at the w a l l . See drawing on Table 1.1 for an example of th i s cone. P a r t i c l e densi ty As stated above the c e n t r i f u g a l force i s proport iona l to p a r t i c l e mass and hence to p a r t i c l e dens i ty . T h e o r e t i c a l l y c o l l e c t i o n e f f i c i e n c y is proport ional to p p 0 * 3 . However, experimental ly this is not confirmed (10). Gas V i s c o s i t y and Densi ty . - 15 - P a r t i c l e motion is r e s i s t e d by the viscous drag. For low p a r t i c l e Reynolds numbers, ( 10"5 < p 8 d p V P / u K < 0.1 ) ( where V P = p a r t i c l e v e l o c i t y r e l a t i v e to the gas) , viscous e f fects dominate and Stokes law can be used to predic t i n d i v i d u a l p a r t i c l e behaviour in a gas. This is of p a r t i c u l a r importance in the d i scuss ion of s c a l i n g the performance of large cyclones at high temperature by small co ld models. P a r t i c l e s separated in hot cyclones do so at lower p a r t i c l e Reynolds numbers (considering s i m i l a r s i z e , densi ty and v e l o c i t y ) than those i n cold models because of increased gas v i s c o s i t y and reduced gas dens i ty . 1.1.4. PREVIOUS CYCLONE SCALING WORK One of the f i r s t works on the subject of s c a l i n g cyclone performance was presented by Stairmand (2) in 1951. The fo l lowing method for p r e d i c t i n g performance of geometr ica l ly s i m i l a r cyclones operating under varying condit ions was proposed. To f i n d the s i ze of the dust caught with the same e f f i c i e n c y as the test dust , m u l t i p l y the test dust s i ze by: /dens i ty of the test dust (1-10) / densi ty of the new dust / t e s t flow (1-11) J new flow 7 new v i s c o s i t y (1-12) test v i s c o s i t y /diameter of the new model (1-13) \l diameter of the test model -16- These re la t ions combined from the Stokes number N s t and imply that c o l l e c t i o n e f f i c i e n c y is a funct ion of N 8 t only for a given geometry. Stairmand was carefu l to l i m i t th is s c a l i n g procedure to densi ty d i f ferences wi th in the ranges 1000 - 4000 kg /m 3 . For cyclone diameter, the method is not recommended "without some experimental conf irmation". Indeed mention is made of less than expected performance with cyclones of diameter greater than 1.22 m diameter. (2) Other attempts have been made to c o r r e l a t e capture e f f i c i e n c y with dimensionless groups and thus provide a s c a l i n g b a s i s . Parker et a l . (11) performed experiments on a 50 mm diameter cyclone at a r e l a t i v e l y low i n l e t v e l o c i t i e s ( 5 m/s) at temperatures up to 700 ° C , and pressures up to 25 atm. Data were p lo t ted on a log log plot of dp a vs ( N s t ) ( N R E ) ° • 3 » the terms were defined as fo l lows: dp. = dp(C , p P )o.5 (g/cm3)o.s (1-14) Nst = C ' d P K * pp VI/(9U*DH) (1-15) N R E = Pz D V i / u, (1-16) where C = Cunningham c o r r e c t i o n factor 'PP = p a r t i c l e dens i ty g/cm 3 DH = 2ab/(a+b) (1-17) Note that the part ic le* diameter in the Stokes number doso re fers to the mass median diameter of the feed aerosol used in his study and that DH i s the h y d r a u l i c diameter of the cyclone i n l e t . For the small cyclones under cons iderat ion the - 17 - resul tant plot was nearly l i n e a r . See Figure 1.1a. Also noted in th i s study was a minimum c o l l e c t i o n e f f i c i e n c y phenomenon^ occurr ing general ly between 2 and 4 microns. This was a t t r i b u t e d to the breaking up of agglomerated p a r t i c l e s during sample preparat ion for p a r t i c l e a n a l y s i s . By examining the performance of two 1.2 m d i a . "Stairmand high e f f i c i ency" cyclones used in ser ies wi th in a pressur ized f l u i d i z e d bed combustor, Wheeldon et a l . ( 12) were able to plot grade e f f i c i e n c y curves for both the primary and secondary cyclones . The tests were done under condit ions of high loading ( 50 g /m 3 ) , temperatures of 640 to 910 °C and pressures of 6 and 12 bar. The authors compared the ir data , obtained in a r e l a t i v e l y large cyclone, to the Stairmand data (2) obtained in small ( 31.5 mm d i a . ) cyclones . The comparison of Stokes numbers for p a r t i c l e s caught with 50% e f f i c i e n c y i s : FBC primary N „ t 5 o = 4 to 10 x 10"5 FBC secondary N a t5 o = 8.6 to 19.6 x 10~5 Stairmand N 8 t5 o = 9.4 x 1 0 - 5 The author a t t r i b u t e d the deviat ions to p a r t i c l e loading ef fects and concluded that standard expressions are s u f f i c i e n t to describe cyclone performance under elevated pressures and temperatures. Also observed in th i s study was a "f i sh hook" shaped c o l l e c t i o n e f f i c i e n c y curves which shows improved c o l l e c t i o n for smaller p a r t i c l e s . (See Figure 1.1b). This ef fect was a t t r i b u t e d to p a r t i c l e agglomeration wi th in the cyclones of smaller p a r t i c l e s on to much larger ones. Analys i s 90% COLLECTION DIAMETER, ĵmA - 8T - PAATICIC SIZC, akront. FIGURE 1.1b C o l l e c t i o n e f f i c i e n c y curves with i i s h nook" shape for primary and secondary cyclones(12) . C o n d i t i o n s : D - 1.2 m, temperatures between 640 and 9 0 0 ° C , i n l e t v e l o c i t i e s between 16.3 and 27.4 m/s, dust loadings between 1.5 and 140 g / m 3 . - 20 - procedures of the p a r t i c l e s caused some d i s a s s o c i a t i o n and suggested higher c o l l e c t i o n e f f i c i e n c i e s for the f i n e s . This ef fect was also noted by Parker et a l . In an attempt to explain performance from a wide v a r i e t y of cyclone designs operated under i n d u s t r i a l condit ions ( 160 < D < 1600 mm, 20 < temperature < 950 °C ) Abrahamson and A l l e n (21) p lo t t ed capture e f f i c i e n c y vs a dimensionless p a r t i c l e diameter, S ° • 3 , defined as : S° • 5 = [ V r p / V r ( C ]o .5 (1-18) where Vrp = Radial p a r t i c l e v e l o c i t y V r * = Radia l gas v e l o c i t y These last two quant i t i e s are evaluated at a r a d i a l p o s i t i o n R = Do / 2 just below the vortex f i n d e r . (See Figure 1.2 q) VTK is assumed to be constant over an imaginary c y l i n d e r , of radius R», extending from the l i p of the vortex f inder to a point on the cone having a diameter equal to that of the vortex f i n d e r . V r p is ca l cu la ted for the experimental data using the Abraham drag c o e f f i c i e n t expression (16) which accounts for non - Stokesian f l u i d - p a r t i c l e behavior up to a p a r t i c l e Reynolds number of 6000 and gives V r p without r e p e t i t i v e c a l c u l a t i o n s . - Z l - - 22 - The procedure followed was: 1. Ca lcu la te G, a • = [ p * U x 2 d p 3 p p ] / ( R, U a2 ) (1-19) where U x = tangent ia l gas v e l o c i t y at R«, assumed to be constant and equal to the i n l e t v e l o c i t y . 2. Ca lcu la te N mot i on): p (Reynolds number for r a d i a l p a r t i c l e Nrorp = 2 0 . 5 [ ( G a ° • 5 / 9 . 6 1 + 1)0•3 - 1]2 (1-20) 3. Ca lcu la te V r g from N R B T P . P a r t i c l e s which have an equal chance of being caught or lost • ( i . e . dp 5o ) should on average move towards the gas exit as often as to the wall and thus the r a t i o V r p / V r K should have a value equal to one. The plot shows better agreement at the dp 5o p o s i t i o n than at other points along the grade e f f i c i e n c y curve ( see Figure 1.2 » b ) , a factor the authors a t t r i b u t e to d i f ferences in cyclone geometry which give d i f f eren t p a r t i c l e re-entrainment c h a r a c t e r i s t i c s . The analys i s is sa id to be l i m i t e d to low concentrations of dust ( less than 10 g /m 3 ) , and to return flow designs with rectangular s lo t e n t r i e s . 1.1.5 Loading ef fects on cyclone c o l l e c t i o n e f f i c i e n c y Improved p a r t i c l e c o l l e c t i o n e f f i c i e n c y for cyclones operating at high dust loadings has been noted by several workers. ( 2 , 3 , 6 , 7 , 1 1 , 1 3 , ) . Perhaps the most comprehensive d e s c r i p t i o n of th is effect i s d e t a i l e d in an e f f i c i e n c y vs . loading plot publ ished by the American Petroleum Ins t i tu te - 23 - "j I I I I " • 1 400 • 2 300 3 1600 O 3 31S a L 160 >< f \ S 160 w 5S Ed o 6 280 u 7 370 fa Es. Ed - « 8 1S0 V 6 800 C O LL EC T!  • T 9 190 C O LL EC T!  - T 11 203 4 O y °£ * ^ / V " cyclone 0. <̂  a« no. flMn ^ a 4 * 18 270 - \ * / • 13 630 \ * <* • * " 1 5 152. « o o 23 305 + o * * . O f O p s : l - * 4 V a 22 49S 4. * • > I , r I t • I I I • I I I I I I I I 1 10 vr FIGURE 1.2 b . Experimental grade e f f i c i e n c y as a funct ion of (S) 0 ** ( dimensionless p a r t i c l e diameter ) for i n d u s t r i a l s i zed cyclones i n a i r at room temperature ( 2 1 ) . - 24 - (API)(17), reproduced in Figure 1.3 (13). A c lear increase in capture e f f i c i e n c y with loading is noted. The inf luence of loading is seen to be greater for cyclones operated at lower " zero load E o " e f f i c i e n c i e s than for those with very e f f i c i e n t E o . This effect i s a t t r i b u t e d to a scrubbing effect of large p a r t i c l e s on f iner p a r t i c u l a t e and has been accounted for in several ways. An excel lent d i scuss ion of the mechanisms considered to be responsible is given by Mothes and L o f f l e r (24). They point out that in sp i te of the observed reduct ion in p a r t i c u l a t e tangent ia l v e l o c i t y ( and thus separating force ) c o l l e c t i o n e f f i c i e n c y increases with increas ing loading . They a t t r i b u t e the ef fect to agglomeration mechanisms. Figure 1.4 plots, the expected c o l l e c t i o n e f f i c i e n c y for small p a r t i c l e s (1 to 4 jim) as separated by larger p a r t i c l e s ( 15 p.m ). B r i e f l y the "scrubbing effect" is a process whereby smaller p a r t i c l e s are forced towards the wall because of the motion of larger p a r t i c l e s which are separat ing out at a higher v e l o c i t i e s . E i t h e r through d i r e c t impact with the larger p a r t i c l e (and the formation of a larger separat ing mass) or by entrainment in the flow f i e l d behind the larger p a r t i c l e , the smaller p a r t i c l e moves toward the cyclone wall with an increased v e l o c i t y . LOADING, GRAINS OF SOLIDS / CU. FT. OF GAS F I G U R E 1.3 Loading ef fect on cyclone c o l l e c t i o n e f f i c i e n c y . Cyclone diameter up to 3.65 m diameter. E 0 i s zero load ( less than 1 g r a i n / f t 3 ) l oad ing , E L i s higher loading e f f i c i e n c y . (18) particle size x/pm FIGURE 1.4 P r e d i c t e d s e p a r a t i o n e f f i c i e n c y of f i n e p a r t i c l e s swept out of the gas by l a r g e p a r t i c l e s due agglomeration <24). T«(x> — p r e d i c t e d c o l e c t i o n e f f i c i e n c y r« — cyclone r a d i u s , ss r e s t i t u t i o n c o e f f i c i e n t . r i vortex f i n d e r r a d i u s b_ i n l e t width h. i n l e t depth V„ = i n l e t v e l o c i t y X o = mass median p a r t i c l e diameter — p a r t i c l e d e n s i t y C n s o l i d s loading g/m3 - 27 - An empir ica l expression proposed by Ogawa (22) for the effect of dust loading in conventional cyclones has the form (modified to give e f f i c i e n c y in %): bi = 0.032 dimensionless ki = -.0157 dimensionless Co = dust loading ( g/m 3) for a 90 mm diameter cyclone operated with f lyash with between 14 and 16 m/s. The negative value for ki accounts for a decrease in e f f i c i e n c y as Co increases noted in experimental data of th i s author, contrary to the effect noted by other workers. For cyclones with an ax ia l i n l e t geometry, c o l l e c t i o n e f f i c i e n c y was found to increase with loading . Another empir ica l r e l a t i o n s h i p was proposed by the API(17) in 1955 of the form: Here the subscript 'o' re fers to low loading condit ions a r b i t r a r i l y taken as 1 g r / f t 3 ( 2.3 g /m 3 ) . Another c o r r e l a t i o n from the API (18) was of the form: e f f i c i ency(^) = 100 - b i [ exp( -k i C o )1100 (1-21) where e f f i c i e n c y = 100- ( 100- e 0 ) [ c « / c f t ] ° • 2 (1-22) P(E) = P ( E 0 ) + A log L (1-23) - 28 - where A is the slope of the nearly l inear curves r e l a t i n g P(e) and L and was e m p i r i c a l l y f i t t e d by a polynomial as fo l lows: A = 0.67 - 2 . 1 1 *E 0 + 5 . 6 3 * E o 2 - 4 . 0 0 * E o 3 ( l -24a) and L the dust loading in expressed g r a i n s / f t 3 . P(e) and P ( E 0 ) are the p r o b a b i l i t i e s associated with the c o l l e c t i o n e f f i c i e n c i e s at high and low loading r e s p e c t i v e l y , i . e . the f r a c t i o n passing the cyclone at a given loading . These p r o b a b i l i t i e s can be approximated by the log of [ ( l - x ) / x ] as: (Catch p r o b a b i l i t y ) = [1 - pass p r o b a b i l i t y ] = [1 - P(x)] = [ 1 - l o g [ ( l - x ) / x ] ] such that [ 1 - l o g [ ( l - E ) / E ] ] = [ 1 - l o g [ ( l - E o ) / E 0 ] ] + Alog( L ) l o g [ ( l - E ) / E ] = l o g [ ( 1 - E o ) / E o ] - log L A = log[ ( 1 - E o ) / E o L * ] [ ( 1 - E ) / E ] = [ ( l - E o ) / E o L A ] [1/E - 1] = [ ( 1 - E o ) / E o L * ] E = E o L A / [ Eo L A + 1 - Eo ] such that l n ( l - E ) = l n [ ( l - E 0 ) / ( l + E 0 [ L A - 1 ] ) ] ( l-24b) ( l -24c) ( l-24d) ( l-24e) ( l -24f ) ( l-24g) ( l-24h) (1-241) ( l -25a) - 29 - Mansin and Koch(23) developed a c o r r e l a t i o n that accounted for the change in e f f i c i e n c y by modifying the e f f ec t ive v i s c o s i t y of the gas as: U t . p p = j i K [ l + 0.091ogL + . 0 2 ( l o g L ) 2 ] . ( l -25b) (note log base 10 in equation l-25b) The model took the form: ln[ (1-E) (1+Eo [L A - 1])°' 5 ] = \ - 2 . 3 S F A C T O R [ N S T ( n + 1 ) A i G / ( L F A C 2 D C 2 ) ] < o . 4 l / ( n • l ) ) (1-26) where: E = c o l l e c t i o n e f f i c i e n c y Eo = low loading e f f i c i e n c y as determined by: l n ( l - E) = 2 [ [ N s T ( n + l ) / ( A i / 2 D c 2 ) G ] < ° - s / < n + i > > (1-27) L = L loading g r / f 3 SFACTOR = [ ( V i / V 8 ) / 2 . 5 ] ° • * i for V-i / V 8 < 2.5 SFACTOR = [ ( V i / V s ) / 2 . 5 ] - o • 5 1 otherwise Vs = p a r t i c l e s a l t a t i o n v e l o c i t y N s t = Stokes number A i = Inlet area G = cyclone conf igurat ion parameter. LFAC = 1 + 0.091ogioL + .02( l o g i o ) 2 Do = Cyclone diameter n = Vortex exponent This is a c t u a l l y the L e i t h modi f icat ions for p a r t i c l e and L icht sa l tat ion (12) model with arid loading e f f e c t s . 30 - Figure 1.5 compares the L e i t h and L icht equation with and without these modi f i ca t ions . - 3x - Fraction passing - data a.) L e i t h / L i c h t model without c o r r e c t i o n s . Legend o Ernst » Knowiton 0 Ogawa - Parker 4 Exxon ? Curt Fraction passing - data b.) L e i t h / L i c h t model with l o a d i n g and s a l t a t i o n c o r r e c t i o n s . Figure 1.5 L e i t h and L i c h t c o r r e l a t i o n s with and without loading and s a l t a t i o n c o r r e c t i o n s . Condi t ions : D = 0.05 and 0.91 m, temperatures between 20 and 850 ° C , i n l e t v e l o c i t i e s between 1.8 and 46 m/s, dust loadings from 0.43 and 4450 g /m 3 . (2$) - 32 - 1.1.6. Summary When descr ib ing f l u i d p a r t i c l e separat ion processes in cyclones i t has been found necessary to include other parameters besides the Stokes number. P a r t i c l e loading and secondary ef fects such as p a r t i c l e re-entrainment play an important ro le and should be taken into cons idera t ion . An increase in c o l l e c t i o n e f f i c i e n c y with increased p a r t i c l e loading has been reported by several workers. - 33 - 1.2 Apparatus and Experimental Procedure 1.2.1 Introduct ion Several experimental arrangements were tested in the course of the experimental program before the set-up was f i n a l i z e d . Results from these pre l iminary "shake down" experiments are not reported in the body of th is thes is but do appear in Figure A l of the appendix and are discussed in sec t ion 1.3.1 . What follows is a d e s c r i p t i o n of the equipment, p a r t i c u l a t e so l id s used, and the procedure for data a c q u i s i t i o n and a n a l y s i s . 1.2.2 Model Cyclone Apparatus Figure 1.6 shows a schematic of the apparatus used for the majori ty of tests performed in Part I of th i s thes is p r o j e c t . The apparatus was comprised of four separate systems: 1. Blower, p i p i n g , o r i f i c e flow metering system 2. So l ids feed system 3. P o l y a c r y l i c cyclone 4. F i l t e r system. Each is described below. For a l l experiments, compressed a i r was provided by a p o s i t i v e displacement Rootestype blower, powered by a 37 kW ( 50 hp) d i e se l engine. The a i r flow could be var ied over a wide range ( 0 - 1100 scfm or 0 - 0.52 m 3 / s ) by c o n t r o l l i n g the SEE APPENDIX FOR DETAILED DIMENSIONS BYPASS DAMPER RDDTS TYPE BLDWER SOLIDS FEED HOPPER ORIFICE PLATE & WATER MANOMETER PLEXIGLASS MODEL CYCLONE SIMULATED REACTOR TOP 0.61 m \ \ \ ,\ \ \ \ \ .•AD CELLS IT 150 MM DIA PLASTIC v PIPE \ CATCH HOPPER FILTER BAG. .•AD SCALE FIGURE 1.6 UBC apparatus schematic used i n runs B l through B38. - 35 - engine speed and a bypass damper. The bypass damper was shut for a l l experiments to ensure constant flow. The flow vs. blower speed curves are shown in Figure 1.7. A 150 mm (nominal) d i a . p l a s t i c pipe connected the blower to the feed hopper while in between, located after more than 4 m of s t ra ight pipe run, was the flow measurement o r i f i c e p l a t e . This brass p l a t e , designed and i n s t a l l e d according to ASME standards ( 28 ) , had a 101.5 mm (4 i n . ) i . d . o r i f i c e . The o r i f i c e pressure drop was measured by means of a water manometer, and because the reading was usua l ly unstable , due to f luc tuat ions in pressure in the simulated reactor top, the best accuracy a t ta inable was +\- 5 mm. The absolute pressure at the o r i f i c e p late was observed to f luc tuate with s o l i d s loading and was in the neighborhood of 30 to 60 cm (12 to 24 inches) water column pressure. The flow vs . o r i f i c e p late pressure drop curve appears in Appendix I ( Figure A2 ) . Sol ids to be fed to the system were contained in a 3.05 m t a l l by 0.36 m square hopper equipped with a manually operated 0.254 m d i a . (aperture opening) cone valve (see Figure 1 .8 . ) . Located at the bottom of the hopper was a 0.36 m square windbox, constructed with a 30% free area d i s t r i b u t o r p late and l ined with a 3 mm thick layer of commercial grade bleached kraft paper (softwood) to ensure even d i s t r i b u t i o n of a i r . Add i t i ona l f l u i d i z a t i o n a i r was provided by a 6 mm diameter pipe entering opposite to the so l id s va lve . F l u i d i z a t i o n a i r flow rates were measured by means of a rotameter. During the runs the hopper was pressur ized , ( by means of a 32 mm d i a . 36 - . S U T O R B I L f 7MF S E R I E S r BLOWER T y p i c a l f f t i - f o n i i a n c e - P r e s s u r e Mode S t a n d a r d I n l G t C o n d i t i o n s 900 1000 J ion i::fiO 1300 1400 ' 1500 OPERATING SPEED - RPM 1600 1700 REV. 1 7-15-85 1600 Mi-/ F i c F i g u r e 1.7 P l o t of blower speed vs flow/pressure drop. ( Pumps and Power Inc.) 3.05- k- 0,36 -H SDLIDS FEED VALVE WIND BOX <SEE DETAIL) TRANSPARENT PLEXIGLASS FACE SUPPLEMENTARY FLUIDIZATION AIR: CONE VALVE HAND CRANK 150 MM DIA SOLIDS DICHARGE PIPE WINHBDX D E T A I L 0.35 DISTRIBUTER1 PLATE < 30 7. FREE AREA ) 1 50 MM T F i g u r e 1.8 Diagram of UBC s o l i d s hopper used i n runs B l through B35. - 3 8 - pipe connected upstream of the o r i f i c e p l a t e ) , to a pressure s l i g h t l y above that of the transport a i r to allow steadier and more r a p i d s o l i d s feeding. The hopper was supported by three load c e l l s ( 1000 kg capacity each, 9 mV output at rated capacity) which allowed mass flow rate to be determined for the durat ion of the experimental runs. Hopper mass information was logged on an XT computer at 1 s i n t e r v a l s . Error i s estimated to be +/- 1 kg, l arge ly due to f r i c t i o n between the hopper and the pneumatic transport p ipe . A f l e x i b l e connection was used at th is junct ion to minimize th is f r i c t i o n . The a i r transport pipe turned twice between the so l id s input point and the cyclone, once through 4 5 ° , then again through 9 0 ° . In order to model the Chatham cyclone more e f f e c t i v e l y the p o l y a c r y l i c cyclone was preceded by a p o l y a c r y l i c entrance s e c t i o n , and a s tee l enclosure geometr ica l ly s i m i l a r to the top of a c i r c u l a t i n g f l u i d i z e d bed r i s e r s e c t i o n . This included a s l a n t i n g r i s e r c e i l i n g p iece , i n c l i n e d at 10° towards the cyclone as shown in Figure 1.9. A 50 mm wide rubber s t r i p sealed th is reactor top to the cyclone entrance way and was i n s t a l l e d taut , thereby minimizing flow in ter ference . The p o l y a c r y l i c entrance-sect ion was b u i l t as a scale model of the o r i g i n a l conf igurat ion of the entrance-way. It was 445 mm high by 140 mm wide at the entrance point to the cyclone. Within the entrance-way were i n s t a l l e d wooden i n s e r t s , designed to model proposed geometric changes being considered for the Chatham cyclone. D e t a i l s of the entrance geometries studied are given in Figure 1.10. F I G U R E 1.9 DIAGRAM DF MDDEL R E A C T D R TOP AND MDDEL C Y C L O N E < ALL DIMENSIONS IN MM ) 150 r 395 J L _ 60 M .ajsa:*:'»:ftsa*v-l CONFIGURATION CI BASE CASE SEE FIGURE 1.1 FOR DIMENSIONS CONFIGURATION C2 30 i CONFIGURATION C3 CONFIGURATION C3 WITH LOVER ED VORTEX FINUER Fzgure 1.10 Entrance geometry conf igurat ions for i n l e t modi f icat ion tes t s . Shaded areas show inser t s A l l dimensxons in mm. See sect ion 1.3.3 for d e t a i i s - 41 - The model cyclone i t s e l f was a one-ninth ( 0 . 1 1 ) scale r e p l i c a of the Chatham CFBC cyclone e x i s t i n g at Chatham New Brunswick. The model cyclone was constructed from 6 mm thick clear p o l y a c r y l i c and dimensions were held to a tolerance of +/- 3mm. The scale was decided upon by assuming that the cyclone Stokes number was roughly s i m i l a r i . e . NsT UBC = NST CHATHAM Operating condit ions are presented in the resu l t s s ec t ion , in Table 1 . 4 . It can be immediately not iced that the Chatham cyclone design chosen is non-standard. Compared to standard designs i t is squat, having a height/diameter r a t i o of H o / D = 2 . 8 , while standard designs have H o / D = 3.7 to 4 . 0 . In addi t ion the vortex f inder or gas exit duct does not extend below the f loor of the entrance way, which raises the p o s s i b i l i t y of gas short c i r c u i t i n g the separation zone. Extending the vortex f inder below the entrance f loor by D / 1 0 is not uncommon in standard designs (5). Also pecu l iar to th i s design is the i n c l u s i o n of an annular zone, concentric with the cyclone i t s e l f but pos i t ioned above the entrance c e i l i n g but below the cyclone c e i l i n g . It is be l ieved that this squat design was chosen to meet dimensional requirements, but i t unclear why the concentric annular region near the top was inc luded. The model provided for two types of vortex f i n d e r s , the f i r s t being made of s tee l with prov i s i on for an add i t i ona l s ec t ion , while the second made of p o l y a c r y l i c , could be re tracted from the cyclone body, thereby s imulat ing other - 42 - conf igurat ions . Gas exited the cyclone to a short 300 mm d i a . s tee l duct which then turned 9 0 ° , traversed approximately 600 mm and f i n a l l y turned down 90° to a bag f i l t e r . For runs Bl to B10, performed under high loading condit ions i t was necessary to use a large sock shaped bag f i l t e r to handle the higher so l id s flows. This bag f i l t e r had the dimensions of 0.78 m d i a . by 4 m long and was made of 100% cotton, 452 g/m2 weight. The bag f i l t e r was securely attached to the exit duct , thus preventing so l ids losses . For subsequent runs, performed under low loading runs, a high e f f i c i e n c y (99 % c o l l e c t i o n e f f i c i e n c y for 5 micron diameter p a r t i c l e s , 4 m2 c lo th area) v e n t i l a t i o n bag f i l t e r was used because i t of fered higher c o l l e c t i o n e f f i c i e n c y and could handle the lower s o l i d s loadings . A 1.7 m long, 100 mm d i a . f l e x i b l e hose connected the so l ids exit of the cyclone to a storage hopper. The hopper was located below and to the side of the cyclone, thus reducing the p o s s i b i l i t y of so l id s re-entrainment from the hopper back to the cyclone, a problem experienced by Stairmand (2) . The so l ids hopper i t s e l f rested on a load sca le , al lowing for measurement of the so l ids caught. The scale was c a l i b r a t e d to +/- 10 grams, and values were logged on an XT computer at 1 s i n t e r v a l s . The data logging program appears in Figure A3 of the Appendi x. In an e f fort to simulate the high loading condit ions in the cyclone at Chatham a so l id s recycle system was designed and b u i l t . Figure A4 in the appendix shows a schematic of the - 43 - system. B r i e f l y the setup envisioned r e c y c l i n g separate ly , in two s imi lar systems, the so l ids caught and those passing the p o l y a c r y l i c model cyclone. Not only were the systems to recyc le the s o l i d s , they were also intended to measure the so l ids flow to provide loading and c o l l e c t i o n e f f i c i e n c y data . Each so l ids recycle system included a f l u i d i z e d seal equipped with dual d i s t r i b u t o r p la te s , and a s o l i d flow measurement vessel equipped with a porous measurement swing p l a t e , cone valve for so l ids flow c o n t r o l , and a d i s t r i b u t o r p late to d i s t r i b u t e f l u i d i z a t i o n a i r . The so l ids measurement vessel was intended to measure s o l i d s . f l o w by means of a porous swing plate pos i t ioned wi th in the measurement vessel ( f i r s t proposed by Turner(40)) . This porous paper l i n e d , 30 % free area punched hole p late could be manually rotated to block so l ids f a l l i n g within the measurement ves se l . Once the plate was rotated c losed, the f a l l i n g so l ids were to c o l l e c t on the l ined plate and be measured. Measurement was to be performed by observing the pressure d i f f e r e n t i a l across the plate and deposited so l id s and thus infer a so l id s flow ra te . The pressure d i f f e r e n t i a l was to ar i se due to the upward flow of f l u i d i z i n g a i r o r i g i n a t i n g from the d i s t r i b u t o r p late located at the bottom of the measurement ves se l , below the so l ids e x i t . While the larger so l ids recyc le system was to receive so l id s d i r e c t l y from the p o l y a c r y l i c cyclone model, the second smaller recyc le system was to handle f ines that passed the model. These f ines were to be caught with a mul t i c lone , - 44 - containing 120 small p l a s t i c cyclones, each 50 mm in diameter. These high e f f i c i e n c y Stairmand type cyclones were to separate the f ines from gas stream, and deposit them in a s tee l hopper with sides i n c l i n e d at 4 5 ° . From there the fines were to have f a l l e n into the f l u i d s e a l , continued into the fines measurement vessel and f i n a l l y recycled back into the feed stream. The cleaned gas stream continued on to a bag f i l t e r , pos i t ioned just downstream of the mul t i c lone . Fines flow measurement was to occur in a s imi lar fashion as in the large recyc le system. Upon commissioning i t was found that th is system did not work for two reasons. F i r s t l y the mult ic lone was unable to capture a l l of the f ines and let an unacceptable amount of s o l i d s pass, overloading the f i l t e r . Attempts to increase the mult ic lone e f f i c i e n c y by increas ing the i n l e t v e l o c i t y to each small cyclone were unsuccessful . Secondly i t was found that the s o l i d s caught by the mult ic lone were so cohesive that they remained in the mult ic lone hopper. Other elements of the system such as the f l u i d i z e d seals and the porous measurement swing plates remain untested. While i t was considered poss ible to r e c t i f y these problems with another mult ic lone design and different, so l ids system, i t was not considered to be a p r a c t i c a l means to achieve the object ives of th is study, given the resources ( i . e . time constra int ) at hand. - 45 - 1.2.3 P a r t i c u l a t e so l id s F l u i d cracking cata lys t f ines , obtained from t e r t i a r y cyclones serving a f l u i d c a t a l y t i c cracking unit at the Chevron Canada L t d . o i l re f inery were used in the experiments reported in th i s thes i s . The equivalent volume sphere diameter, as determined using an Elzone p a r t i c l e analys i s instrument, was approximately 22 Jim. The s ize d i s t r i b u t i o n of the feed material appears in Figure 1.11. A photograph of the test dust appears in Figure 1.12. The bulk densi ty was determined with the so l ids in a loose form and was found to be 770 kg /m 3 . Assuming a voidage of 0.5 the p a r t i c l e density was estimated to be 1540 k g / m 3 . 1.2.4 Data a c q u i s i t i o n and analys is Each test of the co ld model cyclone was performed according to the fo l lowing procedure: 1. As an i n i t i a l i z a t i o n procedure, the system was cleaned out by rapping each component part while scouring the system with a i r . The mass of the feed hopper, catch hopper( with l i d connection o f f ) , and f i l t e r bag were next recorded. The wet and dry bulb temperature were then noted. (See Appendix Figure A8 for temperatures) The p l a s t i c cyclone was wrapped with aluminum f o i l to reduce e l e c t r o s t a t i c e f fec t s , and the feed hopper pressur i zat i on l ine was jthen connected. Mass f r a c t i o n ( % ) \ i a CA N> S u B f* (0 "1 p> 3 cr e (0 »* N 0 (D P CO o SIZE 10 20 30 40 50 60 70 60 90 100 6.78)* 7.32> -» 7.9D- 8.54> •fl 9,22)- 2 9.96) 110.75)- Zll.6l> (X 12.53>- •-13.53) » 14.61)- a15.77> E17.03)-P 18.3y> | 19.86)- ™5:.44> ?23.15)- "> 24.99> 26.99)- 29.14> •31.46) 33.97) 36.66X 39.60) co 42.76) 46.17> 49.85) 53.83) 58.12) — * - - * . * — « -» -* — o o 0 * . ... ft — * * * X -* I 10 20 30 40 50 60 70 80 ?0 100 F i g u r e 1 . 1 2 P h o t o g r a p h o f ' F C C t e s t s o l i d s ( 22 um mean s i z e ' ) - 48 - With the f l u i d i z a t i o n a i r turned on to the wind box, the feed hopper was charged by scooping so l ids from a barre l and feeding, by means of a funnel , through the 50 mm d i a . feed port at the top of the hopper. The rece iv ing storage hopper was sealed. The blower was connected and s t a r t e d . Then the a i r flow was adjusted v i a . engine speed. The data logging program was s tarted on the XT computer. The f l u i d i z a t i o n a i r was increased u n t i l so l id s were seen to be bubbling and c i r c u l a t i n g wi th in the hopper, ( f l u i d i z i n g a i r flow was approximately 100 lpm) The so l ids feed valve was r a p i d l y opened and the hopper rapped as so l ids were r a p i d l y fed. Figure 1.8 shows the valve arrangement. The so l ids valve was qu ick ly shut after the des ired amount of so l ids had been fed . - 49 - 9. With the blower s t i l l running, the system ( p ipes , model cyclone, exit pipes ) was v i o l e n t l y rapped to dis lodge so l ids adhering to inner surfaces . 10. The blower was stopped, and the f i l t e r bag mass, rece iv ing hopper mass, and feed hopper mass were recorded. Sol ids samples were taken from the f i l t e r bag and rece iv ing hopper for analys is of the ir p a r t i c l e s ize d i s t r i b u t i o n s . Deviations or unusual occurrences from the above procedure are reported in Table 1.3. In order to derive the c o l l e c t i o n e f f i c i e n c y curves needed for this study, both the p a r t i c l e s ize d i s t r i b u t i o n and masses of the feed, catch and loss p a r t i c l e s needed to be obtained. As prev ious ly mentioned, the to ta l catch and feed masses were determined from the load c e l l s and load sca l e . The mass of p a r t i c u l a t e passing to the model cyclone was measured by determining the d i f ference in the f i l t e r mass before and after each run. An Elzone p a r t i c l e analys is instrument (model 286XY), interfaced with an AT computer, was used to character ize the p a r t i c l e s i ze d i s t r i b u t i o n . The instrument was c a l i b r a t e d p r i o r to analys is with p a r t i c l e s of mean diameter of 5 jim and 20 pm, to correspond approximately to the expected mean s izes of the cyclone loss and catch r e s p e c t i v e l y . ASTM standard CG90 -71 T for p a r t i c l e s i ze d i s t r i b u t i o n analys i s by e l e c t r o n i c counting - 50 - was fol lowed. A L e i t z Tas Plus Image Analys i s System was used to confirm the p a r t i c l e s ize d i s t r i b u t i o n s . The p a r t i c l e c o l l e c t i o n e f f i c i e n c y curve for each run was obtained by comparing the mass caught to that fed for each channel ( i . e . for each s ize i n t e r v a l ) as fo l lows: Ei = ( C ci ) /( C + L )f i (1-28) where Ei = c o l l e c t i o n e f f i c i e n c y for channel i . C = to ta l mass caught L = to ta l mass lost ( v i a . f i l t e r bag measurement) ci = f r a c t i o n by mass for channel i of catch, f i = f r a c t i o n by mass by channel of feed 1.2.5 Error S e n s i t i v i t y As in any experimental program, the error expected in measurements must be s i g n i f i c a n t l y less than the observed quant i t i e s in order for the resu l t s to be meaningful. The experiments described in th is thesis were set up with th is in mind. For example, in order that there be s u f f i c i e n t re so lu t ion between experiments, the so l id s chosen had to be f ine enough to allow a s i g n i f i c a n t amount to pass the cyclone. - 51 - As w e l l , the manner in which the e f f i c i e n c y was ca lcu la ted was important. Given the feed, catch, and passing masses from a p a r t i c u l a r tes t , there are four ways to ca lcu la te the primary dependent var iable c o l l e c t i o n e f f i c i e n c y . They are: a.) E = (catch mass) (1-29) (feed mass) b.) E = (catch mass) (1-30) (catch mass + loss mass) c. ) E = (feed mass - loss mass) (1-31) (feed mass) d. ) E = (feed mass - loss mass) (1-32) (catch mass + loss mass) In Figure 1.13 a comparison is made of the c o l l e c t i o n e f f i c i e n c y errors r e s u l t i n g from hypothet ica l experimental errors for equations 1-29 to 1-32. The graphs were prepared while considering the case with a gross c o l l e c t i o n e f f i c i e n c y of 95%. From these graphs i t can be seen that equations 1-30 and 1-31 are less suscept ible to experimental error than 1-29 or 1-32. The error in each var iab le depends on so l id s caught up in the system, a t t r i t i o n or agglomeration ( i f any) and measurement e r r o r . The e f f i c i e n c y of the bag f i l t e r influences E f f i c i e n c y e r r o r ( % ) E f f i c i e n c y e r r o r ( % ) i s I n c O 0 I Cfl •1 n B S3 w ft M> ft _ ft i S1 © s> o G> | N © e 0 © 9 0 o / V \ * / / f f / / / _ 1 + B ' O PS Ul _ » w z * 7 H , ^ w 2 0 i r* u S U s u 5 S 0 0 0 9 6 s s s s s e 1 ' ^ 1 \ \ \ * \ \ \ ii rt n n rt 0 a 0' e> ?' fi fi fi fi \ \ \ \ \ \ \ \ E f f i c i e n c y e r r o r ( % ) E f f i c i e n c y e r r o r ( % ) i o w CO ps n O fi I CO •J o n ft e i ui s 01 S 6 0 SI S © s S W * ri y r * o' ?' • 0 ? r. I' 0 »: • * • ; \ K »: ii »: f 9 ^ s s O 3 G> © o to S S o B O © - zs - 53 - loss measurement e r r o r . As previous ly stated the scale accuracies were: feed scale +/- 1 kg catch scale +/- 0.01 kg loss scale +/- 0.001 kg The measurements suffered from the problem of so l ids lodging in the system. The feed measurement was p a r t i c u l a r l y suscept ible to e r r o r . This is because the so l id s fed to the system had more opportunity to be caught up in the p ip ing and entrance way, thus reducing the actual amount reaching the cyclone chamber i t s e l f . The catch mass measurement fared much better as the passage to the storage hopper could be e a s i l y c leared of any materia l that was re ta ined . Measurement of so l ids not captured by the cyclone was hindered not so much by the scale accuracy, but rather by the inev i tab le loss of p a r t i c l e s due to f i l t e r i n e f f i c i e n c y . This rendered the loss p a r t i c l e s ize d i s t r i b u t i o n inaccurate and u n r e l i a b l e . A mass balance, performed on a per channel basis f a i l e d to c lose for p a r t i c l e smaller than 15 um d i a as i s shown in Figure A7 Thus c o l l e c t i o n e f f i c i e n c y ca l cu la t ions were based on the catch and feed p a r t i c l e s i ze d i s t r i b u t i o n s as per equation 1-28. 1.2.6 Chatham cyclone data Figure 1.14 shows a schematic of the Chatham CFB B o i l e r , a complete d e s c r i p t i o n of the i n s t a l l a t i o n can be found in i g u r e 1.14 Schematic of Chatham CFB Boiler(32) - 55 - reference 1. B r i e f l y the Chatham CFB b o i l e r consis ts of a 23.8 m high f l u i d i z e d bed furnace which discharges so l ids and combustion gases to a 5.6 m diameter cyclone. Sol ids are separated from the combustion gases in the cyclone and are returned to the bottom of the furnace. A port ion of these so l ids pass through the F l u i d Bed Heat Exchanger (FBHE). The catch sample analyzed in th i s thesis was taken from the FBHE. A photograph of the Chatham cyclone f ines appears in Figure 1.15. The loss p a r t i c l e s i ze d i s t r i b u t i o n was establ i shed by analyzing p a r t i c u l a t e samples from the bag house with the Elzone p a r t i c l e analyzer . The catch sample was sieved ( ASTM Standard Test Method for Sieve or Screen Analys i s of Fine and Coarse Aggregates: C-136-76 ) to give a rough p a r t i c l e s ize d i s t r i b u t i o n . The f ines ( i . e . material passing a #70 mesh screen) were analyzed further with the Elzone p a r t i c l e analyzer . Equation 1-30 was used to ca lcu la te gross c o l l e c t i o n e f f i c i e n c y data . Sol ids f lux in the reactor was measured at several l eve ls (32) and found to be of the order of 20 kg/m 2 s . This is sa id to be s imi lar to values , as yet unpublished, found by i s o k i n e t i c sampling t r i a l s performed by others in the cyclone in l e t (35). The so l ids loading , ca lcu la ted assuming 1600 m 3 / s gas flow at STP, and a reactor r i s e r area of 16 m2 was estimated to be 10 kg s o l i d s / kg gas. The runs t y p i c a l l y lasted for weeks. Thus i t is assumed that steady state condit ions p r e v a i l e d . It is assumed that the F i g u r e 1.15 P h o t o g r a p h of f i n e s from Chatham f l u i d bed heat exchange r , sampled on A p r i l 17, 1990. - 57 - s i ze d i s t r i b u t i o n s d id not s i g n i f i c a n t l y change during transport from the cyclone to the point of sampling or during transport from New Brunswick to Vancouver. This assumption could not be v e r i f i e d because at the time of wr i t ing i t was not poss ib le to obtain samples d i r e c t l y from the base of the cyclone i t s e l f . Data concerning the operation condit ions or other entrance conf igurat ions w i l l be ava i lab le pending completion of work at the Un ivers i ty of New Brunswick (35). A l l samples were taken on A p r i l 17, 1990, during which time the cyclone had entrance conf igurat ion C3 (see Figure 1.10). The operating condit ions are summarized in Table 1.2. - 58 - TABLE 1.2 CHATHAM OPERATING CONDITIONS FOR APRIL 17, 1990 COMMENTS REFERENCE TEMPERATURE GAS FLOW GAS VISCOSITY GAS PRESSURE PRESSURE DROP SOLIDS FLUX GROSS COLLECTION EFFICIENCY 850 C 1600 m 3 /min. 0.000018 kg/ms TOP OF FURNACE (32) STP (32) 37.5 cm HjO 12.25 cm H2O 20 kg/m J s (36) @ CYCLONE ENTRANCE (32) (32) (see note above) (32) 99.2 % (1) While the method for the cyclone e f f i c i e n c y c a l c u l a t i o n is not s tated, i t is be l ieved that the catch mass flow is ca lcu la ted from an energy balance across the f l u i d bed heat exchanger and thus is not y_ery accurate . The bulk p a r t i c l e densi ty for Chatham so l ids was measured to be 1600 kg/m 3 and assuming a voidage of 0.4 t h e p a r t i c l e density was ca lcu la ted to be 2650 kg /m 3 . - 59 - 1.3. RESULTS INTRODUCTION Part I of th i s study examines the c o l l e c t i o n e f f i c i e n c i e s of two geometr ica l ly s i m i l a r cyclones operating at d i f f e r e n t temperatures, with d i f f eren t s o l i d s , and with d i f f e r e n t i n l e t v e l o c i t i e s . It also examines the performance of the smaller cyclone under d i f f eren t i n l e t v e l o c i t i e s and with various i n l e t geometries. This v a r i a t i o n in i n l e t geometries was performed in order to provide some i n i t i a l guidance on s i m i l a r proposed changes in the Chatham cyclone. S i g n i f i c a n t attempts were made to match the loading condit ions of the Chatham unit but th i s was not completely p o s s i b l e . Problems associated with feeding s o l i d s , c o l l e c t i o n of the f ines not captured by the cyclone, and s o l i d s analys i s l i m i t e d the accuracy of the r e s u l t s . The p o l y a c r y l i c cyclone was s ized so as to not exceed the capacity of the mobile blower while maintaining a cyclone i n l e t v e l o c i t y between 5 and 10 m/s, and yet be as large as p o s s i b l e . Less cons iderat ion was paid to the required s o l i d s loading which proved to be very d i f f i c u l t to meet. I n i t i a l e f for t s focused on a system capable of r e c y c l i n g a l l the so l id s ( both those caught by the cyclone and those l o s t , see Figure A4 in the appendix ) . These attempts were thwarted by the i n a b i l i t y of a 'homebuilt' mul t i c lone , cons i s t ing of 120 cyclones , each of 50 mm diameter, to capture a l l the f i n e s . The p a r t i c l e s caught by the mult ic lones was so- cohesive that i t f a i l e d to f a l l out of the mult ic lone catch hopper, adhering instead to the hopper wa l l s , which were i n c l i n e d at 4 5 ° , without aggressive rapping. The f a i l u r e to achieve complete recyc le - 60 - meant that the s o l i d s could only be fed through once, as a short batch operat ion . T y p i c a l l y each run involved feeding for a per iod of only a few minutes. Figure 1.16 p lots catch hopper mass vs time for run B4, g iv ing an i n d i c a t i o n of the v a r i a t i o n in the feed ra te . T y p i c a l l y there was l i t t l e v a r i a t i o n in the feed rate* The p a r t i c l e s were chosen as a compromise under a set of c o n f l i c t i n g requirements. It had to be f ine enough to allow s i g n i f i c a n t losses ( and thus experimental r e s o l u t i o n between d i f f eren t condit ions ) and yet not so f ine that the f r a c t i o n passing could not be handled. The f i r s t three runs were performed with FCC s o l i d s with a mean diameter of 60 microns and resu l ted in 99% and 98.3% capture e f f i c i e n c i e s r e spec t ive ly , too high for the re so lu t ion requ ired . Subsequent runs used FCC so l id s with a mean diameter of 22 microns and resu l ted in acceptable lower e f f i c i e n c i e s . Unfortunate ly , these f ines were more cohesive, and proved to be more d i f f i c u l t to feed and analyze. Experimental re su l t s are summarized i n Table 1.3. Experiments Bl to B3 were shakedown runs using the larger so l ids ( 60 micron mean ) . Runs B4 to B l l attempted to achieve high s o l i d s loading ra tes , while the remaining experiments focused on so l id s loading ef fects and the inf luence of i n l e t modif icat ions on c o l l e c t i o n e f f i c i e n c y . Pre l iminary "shakedown" te s t s , performed before run B l , used so l id s with mean p a r t i c l e diameters of 11 jum and 50 nm, at i n l e t v e l o c i t i e s between 3.6 and 5.6 m/s. Inlet geometry - 61 - 128 20 % 100 200 300 400 500 time (s) Figure 1.16 Catch hopper mass vs time i n d i c a t i n g steady state so l id s feed r a t e . Run B 4 . RUN 1 DATE PURPOSE :ONPIGIIRATION see Figure .10) SOLIDS AVERAGE DIAMETER LOADING MASS SOLIDS) HASS AIR) FLON •3/s INLET VELOCITY •/s UASSES EFF. * COUUENTS h CAUGHT k LOSS e i 15/5/90 SHAKE DOWN CI, HIGH V.F. 60 UH FCC 2.60 0.19 3.67 n/a "/» n/a FEED PROBLEMS B2 22/5/90 SHAKE DOWN C3, HIGH V.F. 60 UU FCC 0.63 0.19 3.67 79.6 0.80 99 BAD HASS BALANCE B3 23/5/90 SHAKE DOWN CJ, HIGH V.F. 60 UU FCC 3.14 0.19 3.67 SO 0.84 98 feed l a s s caught in pipes IM 24/5/90 s c a l i n g CJ, HIGH V.F. 22 UU FCC 1.25 0.19 3.67 83 1.96 98 •ass balance c l o s e s to 2.7H> B5 24/5/90 s c a l i n g CJ, HIGH V.F. 22 UU FCC n/a 0.19 3.67 n/a »/« n / i feed p r o b l e i s B6 20/6/90 s c a l i n g C3, HIGH V.F. 22 UU FCC 1.96 0.26 5.04 85.15 3.69 96 GOOD RUN B7 20/5/90 s c a l i n g C3, HIGH V.F. 22 UH FCC n/a 0.26 5.04 n/a n/a n/a 99 feed p r o b l e i s B8 02/7/90 s c a l i n g C3, HIGH V.F. 22 UU FCC 7.17 0.28 5.50 101.78 1.13 GOOD RUN B9 10/7/90 s c a l i n g C3, HIGH V.F. 22 UU FCC 7.49 0.26 5.04 91.50 1.44 98 GOOD RUN BIO 10/7/90 s c a l i n g CJ, HIGH V.F. 22 UH FCC 5.38 0.19 3.67 126.99 0.61 99.5 s o l i d s v a h e stuck open Bi t 12/7/90 s c a l i n g C3, HIGH V.F. 22 UH FCC n/a 0.26 5.04 n/a n/* n/a 91 feed p r o b l e i s Bl 2 18/1/90 LOADING CJ, HIGH V.F. 22 UH FCC 0.23 0.26 5.04 3.9481 0.3697 GOOD RUN B l l 18/7/90 LOADING CJ, HIGH V.F. 22 UH FCC 0.12 0.26 5.04 1.9692 0.2972 87 GOOD RUN B H 18/7/90 LOADING CJ, HIGH V.F. 22 UH FCC 0.13 0.26 5.04 4.3351 0.4936 90 GOOD RUN BIS 18/7/90 LOADING CJ, HIGH V.F. 22 UU FCC 0.15 0.26 5.04 2.6652 0.14S 89 GOOD RUN BI6 18/7/90 LOADING CJ, HIGH V.F. 22 UU FCC 0.44 0.26 5.04 3.636 0.4035 90 GOOD RUN BI7 18/7/90 LOADING CJ, HIGH V.F. 22 UH FCC 0.18 0.26 5.04 4.8347 0.497S 91 GOOD RUN BIS 18/7/90 LOADING CJ, HIGH V.F. 22 UU FCC 1.30 0.26 5.04 9.513 0.4339 96 GOOD RUN BI9 18/7/90 LOADING CI, HIGH V.F. 22 UH FCC 0.048 0.26 5.04 2.7358 0.5217 84 GOOD RUN B20 18/7/90 4/8/90 LOADING CJ, HIGH V.F. 22 UU FCC 0.66 0.26 5.04 2.8295 0.2171 93 GOOD RUN 821 GEOV. CHNG. C2, HIGH V.F. 22 UH FCC 0.36 0.24 4.58 14.5138 0.7697 95 GOOD RUN B22 4/8/90 GEOH. CHNG. C2, HIGH V.F. 22 UH FCC 0.26 0.24 4.58 8.6637 0.7046 92 GOOD RUN B21 4/8/90 GEOH. CHNG. C2, HIGH V.F. 22 UH FCC 0.15 0.24 4.58 7.4569 0.(722 92 GOOD RUN B2 4 4/8/90 GEOU. CHNG. CJ, HIGH V.F. 22 UU FCC 0.23 0.24 4.58 7.1836 0.4793 94 GOOD RUN B25 4/8/90 GEOH, CHNG. CJ, HIGH V.F. 22 UH FCC 0.23 0.24 4.58 7.5607 0.603 93 GOOD RUN B26 4/8/90 GEOU. CHNG. CI, HIGH V.F. 22 UU FCC 0.20 0.24 5.41 6.8837 0.35 95 GOOD RUN B27 4/8/90 GEOU. CHNG. CI, HIGH V.F. 22 UH FCC 0.20 0.24 5.41 5.5855 0.294 95 GOOD RUN B28 4/8/90 GEOU. CHNG. CI, HIGH V.F. 22 UU FCC 0.24 0.24 5.41 4.3072 0.2865 94 GOOD RUN B29 4/8/90 GEOU. CHNG. CJ, LOW V.F. 22 UH FCC 0.36 0.24 4.58 6.0028 0.2812 96 GOOD RUN D30 4/8/90 GEOU. CHNG. CJ, LOW V.F. 22 UU FCC 0.34 0.24 4.58 5.6125 0.304 95 GOOD RUN 611 4/8/90 29/08/90 GEOU. CHNG. CJ, LOW V.F. 22 UH FCC 0.11 0.24 4.58 2.8368 0.229 93 GOOD RUN B32 GEOH. CHNG. CJ, LOW V.F. 22 UU FCC 0.17 0.24 4.58 6.2787 0.3067 95 GOOD RUN B33 29/09/90 GEOU. CHNG. CI, HIGH V.F. 22 UH FCC 0.17 0.24 3.81 4.3102 0.3403 93 GOOD RUN B34 29/08/90 29/08/90 GEOU. CHNG. CI, HIGH V.F. 22 UU FCC 0.096 0.24 3.81 4.2827 0.3315 93 GOOD RUN B3S GEOU. CHNG. CJ, HIGH V.F. 22 UH FCC 0.097 0.24 4.58 3.9135 0.5S07 88 GOOD RUN TABLE 1.3 EXPERIMENTAL RESULTS - 63 - configurations CI and C3 were used with vortex finder length as a var iable . Par t i c l e loading rates were below 0.05 kg so l ids /kg a i r . Co l l ec t ion e f f i c ienc ies were found to vary between 41 and 96 %. A summary of this test appears in Figure Al of the Appendix. Runs B4 and BIO were chosen for comparison in the scal ing study because of p a r t i c l e analysis and data logging problems ( i . e . program f a i l e d to record data ) in the other runs. S p e c i f i c a l l y i t was found that the p a r t i c l e catch d i s t r i b u t i o n was greater than the p a r t i c l e feed d i s t r i b u t i o n for p a r t i c l e diameters less than 15 microns. It is thought that the feed sample, which was taken before run B4, was f iner than that actual ly fed in subsequent runs. This is at tr ibuted to the pract ice of recycl ing spent sol ids from one run to the subsequent run. This was necessary as a l l the avai lable so l ids had to be fed in each run in order to achieve the desired loading leve ls . Unfortunately feed samples were not taken for each run as i t was assumed that co l l ec t ion ef f ic iency could be calculated from the catch and loss d i s tr ibut ions alone. Note that fresh, unrecycled feed sol ids were used in runs B4 and BIO. - 64 - 1.3.1 Scaling Considerations Catch p a r t i c l e size d i s tr ibut ions for runs B4 and BIO appear in Figure 1.17a and Figure 1.17b respect ively . Figure 1.17c shows the co l l ec t ion ef f ic iency curve for run B4 as determined from p a r t i c l e size d i s tr ibut ions obtained with the Elzone p a r t i c l e analysis instrument( model 286 XY) and the co l l ec t ion ef f ic iency curve for run BIO as determined by the Tas Plus Image analyzer ( model LSI -11 ) . The f i r s t method reports the equivalent volume sphere diameter, while the second leads to an equivalent projected-area c i r c l e diameter (37). Figure A5 of the appendix deta i l s the p a r t i c l e size information. Note that in Figure 1.17c the masstfof par t i c l e s greater than 20 um in diameter have been combined into one channel with mean diameter of 31 um. A s imilar adjustment was made to the run BIO co l l ec t ion ef f ic iency curve with par t i c l e s larger than 15 jam diameter being combined into one channel of 31 jum mean diameter. This grouping of the higher end channels was performed because of the erra t i c nature of the co l l ec t ion ef f ic iency curve above 20 microns. It can be seen from Figure 1.17c that the cyclone was 50% ef f i c i ent for par t i c l e s of 10 um diameter. A s imilar performance is noted in the ef f ic iency curve for run BIO which shows a d P5o of 10 um. This is surpr i s ing ly e f f i c i ent performance for a 0.61 m diameter, non- standard design cyclone operating with an inlet ve loc i ty of only 3.6 m/s. 0.0 05 + + + + + + + + i cn + + + + + + + + + + + t+-*-+i.++ 4 + + t + + + 4 ^ t » 0 35 70 par t i c l e diameter ( microns ) Figure 1.17a Catch par t i c l e size d i s tr ibut ion for run B4 Part i c l e size analysis by Elzone analysis •>i;i (>li i m». 0.1 O 0.05 n O ' - ' —i— 35 70 part ic le diameter ( microns ) Figure 1.17b Catch part ic le size d i s tr ibut ion for run BIO Part ic le size analysis by image analyzer. 100 9 0 80 70 60 50 40 30 20 10 0 • m • : 0 • • • . " 43- Via G la 0." • • - a » — i 1 i 03 4* O 10 20 30 40 50 Part i c l e diameter ( microns ) 60 70 - RUN B4 • RUN BIO Figure 1.17c Col lec t ion eff ic iency curves for runs B4 and BIO'. Run B4 par t i c l e size d i s tr ibut ions determined by Elzone par t i c l e analysis instrument. Run BIO par t i c l e size d i s t r ibut ion by image analysis methods. Conditions as stated in Table 1.3. - 65 - There is a minimum in the co l l ec t ion ef f ic iency at about 2.5 um diameter in the run BIO eff ic iency curve, with a small increase in ef f ic iency for smaller p a r t i c l e s . There are three probable reasons for th i s . The f i r s t is that because of the cohesive nature of the so l ids , s igni f icant amounts of fines may have adhered to larger par t i c l e s in the catch. During analysis these fines agglomerated with larger part i c l e s may have broke free and been counted as individual p a r t i c l e s . This would have increased the mass frac t ion of fines reported in the catch. The second possible reason arises from the problems of dispersing the dust in the feed gas stream. Par t i c l e s fed to the cyclone came from the sol ids feed hopper and were maintained in a bubbling f lu id i zed bed. It is possible that the smaller part i c l e s started out in an agglomerated state, were fed into the gas stream in this state and f i n a l l y separated as large agglomerates. It may be necessary to provide a means of dispearsing the part iculate while in the feed stream in order that the fines could be tru ly dispersed. The th ird possible reason for the surpr i s ing ly high measured fines co l l ec t ion ef f ic iency is that the fines may have been swept out of the gas stream by other larger p a r t i c l e s . This effect has been reported by others (12, 24). Mothes and Loff ler (24) discuss in de ta i l the phenomenon of improved co l l ec t ion ef f ic iency for small par t i c l e s separated in cyclones with high p a r t i c l e concentrations. They state that p a r t i c l e separation mechanisms other than separation in the vortex must have a major effect on p a r t i c l e separation and - 66 - develop a model to describe this e f fect . B r i e f l y the ca lculat ion of fine p a r t i c l e co l l ec t ion due to agglomeration involves three steps: 1. The i n i t i a l deposition ef f ic iency of on larger part i c l e s s e t t l ing towards calculated. 2. The gas volume cleaned by the larger par t i c l e s travel ing towards the wall is determined. 3. The decrease in fine p a r t i c l e concentration caused by the cleaning effects of the larger part i c l e s is estimated. The model predicted that the separation ef f ic iency of small par t i c l e s in cyclones is a function of scrubbing p a r t i c l e s i ze , small p a r t i c l e s ize , dust concentration, flow conditions and material propert ies . As an example they considered the case of 15 um diameter part i c l e s scrubbing out par t i c l e s sized below 6 um diameter. The results are presented in Figure 1.4 and show a peak in the co l l ec t ion ef f ic iency curves for par t i c l e s s ized between 2 to 3 um diameter. A s imilar effect is noted in Figure 1.17c and occurs in the same range of p a r t i c l e diameters. Considering the Chatham cyclone size d i s t r i b u t i o n in Figure 1.18(a) the loss d i s t r i b u t i o n was c l ear ly separated from the catch p a r t i c l e s ize d i s t r i b u t i o n with l i t t l e overlap. f ine par t i c l e s the wall is - 67 - Comparison of the two d i s tr ibut ions resulted in a typica l co l l ec t ion ef f ic iency curve with a dpso value of 41 microns as can be seen in Figure 1.18(b). This is in the range reported in reference 1, that being 30 to 45 microns. Performance comparison The co l l ec t ion ef f ic iency curves for run B4 and the Chatham cyclone are plotted in Figure 1.19. In order to ver i fy Stokes sca l ing , as interpreted by Stairmand, the performance of the Chatham cyclone was shif ted to see i f the performance would be close to the performance of the UBC cyclone. That i s : N 8 t of shif ted dpso = N s t of Chatham dpso dps 0 n 8 w*ppUBCV lUBC = dps O o l d 2 Pp CHATHAMV l C H A T H A M D u B C UVBC D c H A T H A M U C H A T H A M dps On • « 1 ~ dps 0 o 1 d * p P C H A T H AM V1 C H A T H A M Du B C JHQBC J>pUBC Vi U B C D C H A T H AM U C H A T H A M Table 1.4 states the assumed scal ing condit ions. While the curves are brought closer together, a s igni f icant discrepancy is s t i l l seen. It is clear that the cold model with Stokes law scal ing predicts too op t imi s t i ca l l y the performance of the Chatham cyclone. - 68 - c o « w a S A. 20 40 60 80 100 120 140 P a r t i c l e diameter (microns) • Catch + loss 160 180 >> o e v w G 0 V o u 100 90 80 70 60 50 10 0 • • • fl m m % ' 1 B. 20 40 60 80 100 120 P a r t i c l e diameter (microns) 140 160 180 Figure 1.18 a.) P a r t i c l e s i z e d i s t r i b u t i o n s from Chatham f l u i d bed heat exchangers sampled A p r i l 17, 1990. b.) C o l l e c t i o n e f f i c i e n c y curve derived from p a r t i c l e s i z e d i s t r i b u t i o n s . E = ( C l i / ( C l i + Llx)) - 6 9 - >-o z UJ M o M a. u. LU 2 O U LU _J _ l o o 100 50?- • • • tc J UBC. . RUN B4 SHIFTED CHATHAM DATA*^.' i ' • CHATHAM DATA paw—Jlm»H» 10 20 30 40 50 60- 70 PARTICLE DIAMETER (pm) F i g u r e 1.19 C o l l e c t i o n e f f i c i e n c y curves f o r run B4, the Chatham cyclone and the Chatham cyclone s h i f t e d a c c o r d i n g t o Stokes Law s c a l i n g . S c a l i n g c o n d i t i o n s as per Table 1.4. - 70 - TABLE 1.4. CYCLONE OPERATING CONDITIONS Chatham UBC (Run B4) CYCLONE DIAMETER 5.6 0.61 m TEMPERATURE 850 21 «C GAS DENSITY 0.326 1.20 kg/m 3 GAS VISCOSITY 0.000045 0.000018 kg/m/s PARTICLE DENSITY 2650. 1540. kg/m 3 AIR FLOW (@STP) 1600 11.2 m 3 /min . INLET VELOCITY 20.4 3.66 m/s STOKES NUMBER 0.020 0.0029 REYNOLDS NUMBER 830 000 149 000 LOADING RATIO 8.8 1.4 kg/kg Note that the Stokes Number is defined as: N S T so = d p 5 0 , p p V i / ( 1 8 u D) and the Flow Reynolds number is defined as: Nae = p « V i D / / i In Figure 1.20 experimental data are p l o t t e d against the p a r t i c l e s i ze dependent dimensionless number S ° • 5 proposed by Abrahamson and A l l e n (21) to c o r r e l a t e e f f i c i e n c y data from large , high temperature cyclones to those operating at room temperature. This approach was f i r s t discussed in sect ion 1.1.4. Data from the Chatham cyclone appears to c l o s e l y fol low the trend shown in Figure 1.2b with the dpso value f a l l i n g close to 1. However data from Run B4 are less comparable, i n d i c a t i n g a much greater c o l l e c t i o n e f f i c i e n c y than would be expected* The Parker et a l . . . study compared small cyclones operating at extreme temperatures and found that the e f f i c i e n c y data could be p l o t t e d against ( N R B ) ( N S T ) ° • 5 . A p lot of the UBC - 71 - >• u z L U M U M li- lt. U J 100 80 20 0 a \ / • « a • 1 H - 1 I 1 I 0 1.0 S O . 5 2.0 A. >-O z L U M o U L L U L U 100 80 60 40 20 0 M g W " » HUH* ** a • I a a * 1 B 1.0 So.s 2.0 Figure 1.20 UBC and Chatham data plotted according to Abrahamson and Al l e n c o r r e l a t i o n s . a. ) UBC data (run B4 conditions). b. ) Chatham data. See Table 1.4 for operating conditions - 72a - and Chatham data on the same graph appears in Figure 1.21. Both the UBC and Chatham data f a l l above the small cyclone Parker et a l . data . This was also found to be the case for other larger cyclones compared in the ir study. The authors suggest that cyclone diameter must play an important ro le (11). The major discrepancy between the UBC and Chatham c o l l e c t i o n e f f i c i e n c y curves, and the fact that the model's performance was much superior to other cold models compared in the Abrahamson study leads one to suspect the experimental data . It is suspected t h a t , i n the UBC tes t s , p a r t i c l e s separated in an agglomerated form, separating as large masses rather than independent e n t i t i e s . Sample preparat ion , which required sonicat ing the sample in a l i q u i d media, may well have resu l ted in d i s a s s o c i a t i o n of p a r t i c l e s which were then counted i n d i v i d u a l l y . Indeed a c e r t a i n degree of agglomeration is evident in the photographs of the test dust . A means of determining the l eve l of agglomeration wi th in the cyclone would be needed to confirm t h i s . P a r t i c l e agglomeration may have been inf luenced by two f a c t o r s . 1. E l e c t r o s t a t i c forces , induced by the motion of the p a r t i c l e s in the p o l y a c r y l i c cyclone, may have caused p a r t i c l e s to agglomerate. Repeating the experiments in a Steel cyclone or with the add i t ion of an a n t i - s t a t i c compound may reduce these e f f e c t s . 2. Poor d i spers ion of p a r t i c l e s being fed to the cyclone may have resu l ted in the p a r t i c l e s entering the cyclone in an agglomerated s ta te . Changing the experimental apparatus to b ' 1 ""'I i i ) i 11111 ~i—i i i 11111 1—i i i M 11 j 1 — i i i mi) o o U B C ^ 6 °o°(b A CHATHAM • • • 3 O A A T . OATA A EXXON DATA KMOWLTOM DATA ci i—i i i i m l i — i i i i m l i i i i i m l i i i i i i i i i i i i i 1 1 n l 2 S 4 • 1 2 3 4 « t 2 5 4 * 1 2 3 4 7 * 1 2 J 4 6 ' 2 2 4 « 4 T 10 to 10 D 10 10 (N j ( N j " DMENSJONLESS F i g u r e 1.21 UBC Run B 4 and Chatham data compared w i t h Parker e t a l . d a t a . C o n d i t i o n s as s t a t e d i n Tabl e 1 . 4 - 73 - include a sect ion where p a r t i c l e s are smashed against a baf f l e or d i sassoc ia ted by sonic ac t ion may resolve th i s problem. - 74 - 1.3.2 Loading effect Runs B9 and B12 through B20 were performed for the p a r t i c l e loading study and the resu l t s are p lo t ted in Figure 1.22 and the data is presented in Tabel 1.5. A c lear increase in c o l l e c t i o n e f f i c i e n c y is noted as p a r t i c l e loading is increased. Table 1.5 P a r t i c l e Loading Data run # loading c o l l e c t i o n e f f f i c i e n c y MASS SOLIDS MASS AIR % B12 0.235 91 B13 0.124 87 B14 0.134 90 B15 0.147 89 B16 0.437 90 B17 0.376 91 B18 1.303 96 B19 0.049 84 B20 0.658 93 B9 7.49 98 These reu l t s are consistant with that reported in other works (18) in two ways in that e f f i c i e n c y increases with loading and no l i m i t or maximum is found. Attempts to increase the loading to leve ls greater than 7.5 kg s o l i d s / kg a i r were not successful due to so l ids feeding l i m i t a t i o n s . The loading effect is commonly a t t r i b u t e d to the act ion of p a r t i c l e - p a r t i c l e c o l l i s i o n s and agglomeration mechanisms which allow the so l ids to s e t t l e out in large c lus ters and strands rather than i n d i v i d u a l p a r t i c l e s (23,24). Confirmation of th i s 100 98 96 94 92 9 0 | 88 86 84 0.01 0.1 Loading r a t i o (kg s o l i d s / kg a i r ) 1.0 10 F i g u r e 1.22 Loading e f f e c t on c o l l e c t i o n e f f i c i e n c y ( UBC data ). V i = 5.0 rn/s, T = 21 °C, P = 1 atm - 76 - theory is not poss ib le from this experimental work, but i t may be poss ib le to v e r i f y th i s i f a method of measuring agglomeration levels in the separat ion zone as a funct ion of i n l e t loading can be found. Unfortunately i t was not poss ib le to achieve the very low loadings ( 1 g r / f t 3 , 2.3 g/m3 ) reported in the API study (see Figure 1.3) because of problems with the large s o l i d s feeding cone valve . An improved experiment would include a means of feeding so l ids on the 1 to 10 g r / f t 3 (2.3 to 23 g/m 3) range. A smaller cone valve , one tenth the s ize of the 0.25 m diameter valve would l i k e l y s u f f i c e . - 77 - 1.3.3 Inlet Modi f icat ions Table 1.6 summarizes the experiments devoted to i n l e t modi f i ca t ions . In order to compare the c o l l e c t i o n e f f i c i e n c i e s for the d i f f eren t i n l e t conf igurat ions the e f f i c i e n c i e s l i s t e d here have been modified to account for loading e f f ec t s . The modi f i ca t ion procedure was as fo l lows: A l l runs are compared on a basis of the loading found in run B24 i . e . 0.232 (mass s o l i d s / mass a i r ) . This run was chosen because i t s loading value f e l l in the middle of a l l the other loadings encountered in the i n l e t modi f icat ion tes t s . A best f i t l i n e was drawn through the loading data found in sect ion 1.3.2 and the equation of th i s l ine es tabl i shed as: E( L ) = (0.0301)ln(L) + 0.9097 where E( L ) = c o l l e c t i o n e f f i c i e n c y at loading L L = Loading (mass sol ids/mass a i r ) Taking the d e r i v a t i v e one obtains: dE( L ) = 0.0301 dL L - 78 - Corrected e f f i c i e n c y values ( E C o r r . ) were obtained by according to: E oo r r . = E g i p e r . + 0.0301(Lexper . - LB 24 )/Lexper. where E e x p e r . = e f f i c i e n c y determined at other loading L e s p e r . = loading in test During the run the i n l e t geometry and vortex f inder length were a l t e r e d , r e s u l t i n g in d i f f erent i n l e t v e l o c i t i e s and p a r t i c l e paths within the cyclone. The runs have been placed in order of increas ing i n l e t v e l o c i t y , with the f i n a l three runs d i f f e r i n g from the rest in that the vortex f inder was elongated 212 mm in th is s i t u a t i o n . The data is p lo t t ed in Figure 1.25. Inlet geometry d e t a i l s are shown in Figure 1.10 and b r i e f l y described below. Conf igurat ion C I . The base case s i t u a t i o n with no i n l e t inserts or vortex f inder changes have been made. See Figure 1.9 for dimensions. Conf igurat ion C2. The i n l e t f loor has been ra i sed 50 mm above the o r i g i n a l f loor and the c e i l i n g lowered 30 mm for a short sect ion near the reac tor . The rear wall has been straightened out reducing the i n l e t width to 110 mm from 150mm. - 7 9 - Configurat ion C3. As per conf igurat ion C2 but with the i n l e t width enlarged to 135 mm. Conf igurat ion C4. As per conf igurat ion C3 but with the vortex f inder lowered 212 mm. - 80 - TABLE 1.6 Inlet M o d i f i c a t i o n Tests RUN CONFIGURATION COLLECTION EFFICIENCY B26 c l 96 B27 c l 96 B28 c l 94 B33 c l 94 B34 c l 97 B24 c3 94 B25 \ c3 93 B35 c3 92 B21 c2 94 B22 c2 92 B23 c2 93 B29 c3 & low vortex f inder 94 B30 -c3 & low vortex f inder 94 B31 c3 & low vortex f inder 96 B32 c3 & low vortex f inder 96 It would be expected that c o l l e c t i o n e f f i c i e n c y would increase as the entrance i n l e t area was decreased, that is as the conf igurat ion was changed from C l to C3 to C2. However, th is was not evident in the data and the base case with the lowest i n l e t v e l o c i t y performed better than e i ther C2 or C3. Conf igurat ion C3 coupled with a lower vortex f inder of fered better c o l l e c t i o n e f f i c i e n c y than C2 or C3, alone but s t i l l not s i g n i f i c a n t l y d i f f e r e n t than the base case. For the same geometric conf igurat ion i t was not poss ib le to reproduce the same c o l l e c t i o n e f f i c i e n c y as i s seen in Figure 1.25. Only in runs B26, B27, B28 and B33, for conf igurat ion C l , were s i m i l a r c o l l e c t i o n e f f i c i e n c i e s noted. This v a r i a t i o n is a t t r i b u t e d to d i f ferences in the experimental C o l l e c t i o n E f f i c i e n c y ( % ) (Q c -< OJ o a o o H . o 3 -h 3 CL -h TI (D 0) -( (t> 3 ft r+ 0 CO -1 to o 3 3 O ^ -h <t> W ft 0 O H- H. O a 3 a> CO -h O H« ft 0 C 0 0) -< CL B> H- ft o o -h 3 H. (Q O 3 H o w o o -1 -1 <I> o ft 0) (P </> 3 (6 O (D H' <p H W a 0) cr -h W H- O o o z ft c » H~ w 0 z 0 W o 0 0 - o V r- - 82 - condit ions between these runs, s p e c i f i c a l l y the v a r i a t i o n s in the loading ra te . While i t seems c lear that c o l l e c t i o n e f f i c i e n c y is greater than 90% for a l l runs, i t is not poss ib l e , based on the data c o l l e c t e d to date, to declare a "winner" among those t r i e d . These puzz l ing resu l t s suggest that a d i f f eren t experimental approach is necessary. An improved experimental program would resu l t in a steadier feed ra te , perhaps by way of a mechanical feeder as i t was d i f f i c u l t to repeat loading leve ls with the present so l id s valve feeding arrangement. As w e l l , a so l id s system less prone to agglomeration would allow e f f i c i e n c y curves, which span a f u l l range of e f f i c i e n c i e s , to be determined. Examination of these improved e f f i c i e n c y curves may offer better answers. - 83 - 1.3.4 Flow v i s u a l i z a t i o n Separate runs of the cyclone apparatus were performed for the purpose of flow v i s u a l i z a t i o n . C o l l e c t i o n e f f i c i e n c y was not evaluated during these runs. These runs were performed at low loading in order that flows in the i n t e r i o r could be observed. During the t r i a l s the fo l lowing observations were made for a l l geometries: R e c i r c u l a t i o n zone below roof: It could be seen that a d i s t i n c t annular region e x i s t s , concentric with the cyclone i t s e l f but s i tuated above the roof of the entrance and below the roof of the cyclone. Suspended so l id s e i ther enter th is zone or f a l l from i t to the separat ion region below. Any so l id s that found the ir way into th i s zone are eventual ly reentrained in the incoming flow, but spend considerable time in the reg ion. Reentrainment into incoming flow: Sol ids that manage to reach the outer wall must drop below the bottom of the entrance i f they are not to be reentrained again by the incoming flow. The Stokes number s c a l i n g - 84 - approach is u n l i k e l y to apply to p a r t i c l e - p a r t i c l e and p a r t i c l e - w a l l i n t e r a c t i o n s . Their effect may cause performance to be somewhat unrepresentative of performance in the f u l l scale cyclone. It can be argued that those p a r t i c l e s whose misfortune i t was to reentrained must then be separated once again with p a r t i c l e s in the incoming flow, and thus may be representat ive of matters at Chatham. For the base case, where no i n l e t inserts were used, a r e c i r c u l a t i o n zone could be seen pos i t ioned in the entrance way near the top of the model combustor. This r e c i r c u l a t i o n was not evident with the inser ts i n place as in Conf igurat ion C2 or C3. There was no great v i s i b l e d i f ference in the p a r t i c l e paths for the e x i s t i n g and modified vortex f inder conf igurat ion . This may be due to the d i f f i c u l t y in seeing flow patterns in the d i l u t e region close to the vortex f i n d e r . More advanced techniques for flow v i s u a l i z a t i o n ( e . g . . laser Doppler techniques) would be required to study th is in d e t a i l . The video tape sent to Energy, Mines, and Resources shows the . observations discussed above. - 85 - 1.4 Conclusions and Recommendations. The main features of part I of th is study are as fo l lows: 1. A one-ninth scale p o l y a c r y l i c model of the non- standard design i n d u s t r i a l cyclone operated at the 22 MWe CFBC f a c i l i t y at Chatham, New Brunswick has been constructed and operated at room temperature. P a r t i c l e c o l l e c t i o n performance has been tested under various so l id s loadings and i n l e t geometries. 2. A so l id s recyc le system, u t i l i z i n g a mult ic lone to capture f ines and a recyc le system to provide continuous operation was b u i l t , but d id not work well because of so l id s capture problems and the cohesive nature of the f ine s o l i d s . As a r e s u l t , a l l of the resu l t s presented are for a batch system. 3. At i n l e t v e l o c i t i e s var ied between 3.7 and 5.5 m/s, so l id s loadings between 0.05 and 7.5 (kg s o l i d s / k g a i r ) , for FCC so l id s having a mean diameter of 22 jim, cyclone c o l l e c t i o n e f f i c i e n c i e s remained above 90%. 4. There was d i sappoint ing agreement between the resu l t s from the Chatham u n i t , scaled according to Stokes Number s c a l i n g , and the f indings obtained from the cold model u n i t . 5. The grade e f f i c i e n c y curve showed a minimum e f f i c i e n c y for f ine p a r t i c l e s , 2.5 to 3.0 um in diameter. This is l i k e l y due to agglomeration ef fects as a mass balance performed on a per channel ( i . e . -86- s ize in t erva l ) basis d id not close for p a r t i c l e s smaller than 15 um diameter. 6. Increasing p a r t i c l e loading led to an increase in c o l l e c t i o n e f f i c i e n c y . 7. Inconclusive resu l t s were found when the i n l e t conf igurat ion was changed while using 22 um mean p a r t i c l e diameter FCC s o l i d s . 8. Flow v i s u a l i z a t i o n t r i a l s were performed with the p o l y a c r y l i c cyclone. Reentrainment into the incoming flow of captured so l ids skipping along the wall was observed. A videotape of these observation was prepared and sent to Energy, Mines and Resources i n Ottawa. It is recommended that , in order to v e r i f y the scale-up c r i t e r i a , the experiments be repeated with a d i f f eren t so l id s system less prone to agglomeration. Successful c o l l e c t i o n e f f i c i e n c y studies have been performed using f lyash at lower loadings (11, 12) suggesting that f lyash might be a su i tab le m a t e r i a l . - 87 - PART II HOT CYCLONE TESTS - 88 - 2.1 INTRODUCTION C i r c u l a t i n g F l u i d i z e d Bed Combustors (CFBC) r e l y on i n e r t i a l separation devices to separate combustion gases from entrained s o l i d p a r t i c l e s and to return those so l ids to the reactor . Cyclones are usua l ly chosen to perform this high temperature, high loading and sometimes high pressure separation because they offer reasonably good p a r t i c l e c o l l e c t i o n e f f i c i e n c y and are easy to des ign, operate and maintain . When used in such circumstances, some combustion i n e v i t a b l y occurs within the cyclone. In order to better understand the combustion processes, wi th in the cyclone, r a d i a l combustion gas concentrat ion p r o f i l e s were measured with in a secondary cyclone serving a p i l o t scale CFBC system operated at the U n i v e r s i t y of B r i t i s h Columbia. The p r o f i l e s presented in th is thesis were a l l obtained with Highvale c o a l , a low sulphur coal from Alber ta as a f u e l . For d e t a i l s of propert ies of the coal see See Figure A 6 of the appendix. In th i s sect ion a b r i e f review of gas and so l ids flows with in a cyclone is g iven , followed by a d e s c r i p t i o n of the apparatus and a presentat ion of the measured combustion gas combustion p r o f i l e s . The p r o f i l e s are then discussed and conclusions presented. 2.2 Theory Combustion processes occurr ing with in cyclones are dependent on the nature of the s o l i d s , the composition of - 89 - the gas as well as the operation mode of the combustion system. Gas and so l ids flow patterns wi th in cyclones are functions of several var iables inc luding cyclone dimensions and geometry, gas flow rates , and p a r t i c l e loading. When used in a CFBC system, the so l id s are a combination of unreacted fuel p a r t i c l e s , inert so l id s (Eg. sand and ash) and sorbent mater ia l ( i f a sorbent is used). The combustion gas composition depends on fuel type, combustor conf igurat ion , mode of operat ion, operating temperature and other parameters. In order to understand the gas concentrat ion p r o f i l e s , i t is f i r s t necessary to b r i e f l y describe the gas and so l id s flow patterns wi th in a reverse return cyclone and to present the o v e r a l l combustion equations which character ize CFBC systems. Flow patterns Gas and so l id s flow patterns wi th in a cyclone are int imate ly r e l a t e d . Sol ids flow patterns have been observed in a co ld model by the author (see sect ion 1 . 3 . 4 ) and documented by several other workers (2, 25). The r a d i a l , t a n g e n t i a l , and ax ia l components of the so l ids v e l o c i t y vary with p o s i t i o n and so l id s loading (15). Pred ic t ions of low loading p a r t i c l e mean t r a j e c t o r i e s appear in Figure 2.1 (25) and show the predicted ax ia l and r a d i a l p o s i t i o n of p a r t i c l e s of various s i z e s . - 90 - F igure 2.1 Pred ic ted p a r t i c l e t r a j e c t o r i e s i n a v e r t i c a l plane wi th in a Stairmand type cyclone. Low loading c o n d i t i o n s . (25) a . Mean p a r t i c l e t r a j e c t o r i e s for p a r t i c l e s of diameter 1 to 10 microns. b. Mean p a r t i c l e t r a j e c t o r i e s , 3 micron. c. Random p a r t i c l e t ra jec tory of 2 microns p a r t i c l e i n turbulent flow. - 91 - In Figure 2.2 the ax ia l gas v e l o c i t y is seen to be a funct ion of radius , with downward motion occurr ing in the outer regions, while within the core the flow reverses and trave l s up with increased v e l o c i t y towards the vortex f i n d e r . The tangential v e l o c i t y has been found to increase with decreasing radius , reaching a maximum in the centra l core region below the vortex f i n d e r . Increased p a r t i c u l a t e loading has been found to reduce the tangent ia l component of gas v e l o c i t y (24). Figure 2.3 shows the combined r a d i a l and ax ia l v e l o c i t y vector f i e l d . P a r t i c l e t r a j e c t o r i e s P a r t i c l e motion within these complex flow patterns depends on gas and p a r t i c l e c h a r a c t e r i s t i c s , and on so l id s loading (15). As so l ids loading increases , p a r t i c l e - p a r t i c l e c o l l i s i o n s become more frequent, forming larger c lus ters and strands which ass i s t p a r t i c l e c o l l e c t i o n . Once the p a r t i c l e s have reached the wall they may be re -entra ined should some disturbance occur, forc ing the so l id s back into the gas flow. Disturbances such as large bouncing p a r t i c l e s , i n t e r i o r wall surface imperfections and interference of the gas vortex with the cyclone wall have been discussed by various authors (2 ,24) . At the wall a dense layer of so l id s forms, with p a r t i c l e s t r a v e l l i n g around and down the cyclone body, eventual ly concentrating into a d i s t i n c t i v e dense strand which "snakes" i t s way down to the so l id s exit at the base. - 92 - Figure 2.2 Predic ted r a d i a l and a x i a l gas flow patterns in a Stairmand type cyclone. Low loading condit ions (25). - 93 - ::;iwi i i tnitti j« . i M i i t t i . I j l i l t t i , n i n t h i t t t t i t t t 111 1 1 1 11111 I t l t l i i j t l l t i . , t t t i Figure 2.3 Predic ted combined ax ia l and r a d i a l v e l o c i t y ' v e c t o r diagram i n a Stairmand type cyc lone . Low loading condit ions(25) - 94 - Summary of combustion equations. While i t is not wi th in the scope or intent of th i s thesis to discuss heterogeneous combustion mechanisms for coal combustion within a C F B C , i t i s useful to describe the primary react ions a f f ec t ing the gases measured within the cyclone. The gases C02 , C O , C H * , SO2 , and N O , are a l l formed during coal combustion. CO2 is p r i m a r i l y formed by the two react ions ( 2 6 ) : C* + Oj -> CO2 CO + 1 /2 0 2 -> C 0 2 . CO can be formed as fol lows: C* + CO2 -> 2CO C + HiO -> CO + H2 CO2 -> CO + 1/2 O2 C* represents the burning char. CH4 can be formed from: C* + 2 H 2 -> CHJ , or released as v o l a t i l e s are evolved. SO2 can be considered to form v i a : S + O2 -> SO2 . while N O S is formed by react ions of the type 1/2 N2 + x / 2 O2 -> N O , . At temperatures of interes t in F B C processes, i t i s fuel n i t rogen , rather than ni trogen present in the a i r , which is predominantly responsible for NO* formation (38). - 95 - Because most combustion within the CFBC system occurs before the secondary cyclone, much of each gas measured or ig inates before the cyclone. A lesser amount is formed within the cyclone contr ibut ing to the measured values . 2.3 Apparatus and Data A c q u i s i t i o n The apparatus used in this high temperature study is the p i l o t scale f l u i d i z e d bed combustor at the U n i v e r s i t y of B r i t i s h Columbia in i t s modified conf igurat ion (30). A schematic i s shown in Figure 2.4. In b r i e f , the set-up includes a re frac tory l ined reactor (152mm square in -cross- sect ion by 7.3m t a l l ) , a primary and secondary cyclone with prov i s i on for so l ids r e c i r c u l a t i o n v i a an L-valve and jet educator r e s p e c t i v e l y , hoppers for feeding the fuel and sorbent, and primary and secondary a i r i n j e c t i o n . The system is well instrumented. Comprehensive descr ipt ions can be found elsewhere (27, 29). The secondary cyclone was chosen over the primary cyclone for these experiments because of the probe plugging problems associated with gas sampling in regions of high s o l i d s density ( e . g . . at the cyclone w a l l ) . The insulated s ta in l e s s s tee l secondary cyclone is s i tuated after the primary cyclone and thus receives the reactor f lue gases and p a r t i c l e s not captured in the primary cyclone. A scale drawing of the secondary cyclone appears in Figure 2.5. This 0.2 m i . d . cyclone was modified to allow the i n s e r t i o n of a gas sampling probe through the wall at several l e v e l s . The probe i t s e l f was connected v i a cool ing - 96 - secondary air w a t e r l i q u i d f u e l Q n a t u r a l g a s p r i m a r y a i r Simplified schematic diagram of circulating fluidized bed combustion facility 1. Reactor; 2. Windox; 3. Primary cyclone; 4. Secondary cyclone; 5. Re- cycle hopper; 6. Standpipe; 7. Eductor; 8. Secondary air preheater; 9. Flue gas coolers; 10. Baghouse; 11. Induced draught fan; 12. Fuel hopper; 13. Sorbent hopper; 14. Rotary values; 15. Secondary air ports; 16. Membrane wall; 17. Pneumatic feed line; 18. External burner; 19. Ventilation; 20. Calorimetric section FIGURE 2.^ CFBC Schematic (27) - 9 7 - PORTS LDCATED BELLTW THIS POSITION (PDRTS SHOWN 90° CV IN ELEVATION BELOW) E X I T I N L E T VIDTH = 50 MM 650 S D L I D S E X I T " H FIGURE 2.5 Scale drawing of secondary cyclone of UBC CFBC system. - 98 - and f i l t r a t i o n stages to a gas sampling t r a i n . Figure 2.6 shows a schematic of this apparatus. The gas sampling t r a i n leads to f i ve analyzers for the measurement of C O 2 , CO, C H 4 , S O 2 , N0 X , and O2 gases. A b r i e f summary of key features of these instruments is presented in Table 2 .1 . A more d e t a i l e d review is found in reference 34. TABLE 2.1: DESCRIPTION OF ANALYTICAL INSTRUMENTS GAS MAKE/ MODEL 02 C02 FUJI 732 CO HORIBA PMA 200 OXYGEN ANALYZER FUJI 732 CH4 FUJI 730 S02 HORIBA PIR 2000 NOX MONITOR LABS INC. MODEL 8840 RANGE OPERATION (ACCURACY) PRINCIPAL (% f u l l scale) 0 - 25 % ( 1% ) 0 - 20 % ( 1 % ) PARAMAGNETIC TYPE NDIR 0 - 1000 PPM NDIR ( 1 % ) 0-0.5 % ( 1 % ) 0-1000 PPM ( 1 % ) 0-500 PPM ( 1 % ) NDIR NDIR RESPONSE t ime 20s 5 s 5 s 5 s 5 s CHEMILUMINESCENCE 3 min Each of these instruments was c a l i b r a t e d using standard gases p r i o r to each run. The procedure followed to obtain data was as follows: STAINLESS STEEL PROBE: 6HM DIA. Nn SD CH- 4 CO / 2. ° 2 r PUMF i HEAT EXCHANGER SECONDARY CYCLDNE FRDM DTHER SAMPLING PLTINTS GAS ANALYSIS METERS FIGURE 2.6 Gas sampling system serving UBC CFBC system. - 1 0 0 - 1. With the p i l o t plant operating as nearly as poss ib le under steady state condit ions and gas analyzers prepared according to manufacturers' i n s t r u c t i o n s , c a l i b r a t e d , the sample probe was inserted to the des ired r a d i a l p o s i t i o n and gas sampling commenced. 2. After a period of 4 minutes the gas transport l ines were assumed to be purged and the gas analyzers reading steady state values . 3. Gas concentrat ion values were recorded. 4. A f lue gas [02] measurement was obtained from a separate gas meter operating on a separate sampling l i n e connected downstream of the cyclone. The sampled gas stream was f i l t e r e d and cooled. - 101 - 5. The sample probe and f i l t e r were purged with compressed a i r , then moved to the next r a d i a l p o s i t i o n . A random order was followed in probe p o s i t i o n i n g in order to avoid systematic v a r i a t i o n s . 2.4 Resul ts and Discuss ion The CFBC operat ing condit ions under which the gas concentrat ion p r o f i l e s were obtained are ou t l ined in Table 2.2 for each of the f i ve runs where data were obta ined. Table 2.2 UBC CFBC operat ing condit ions RUN # RUN AIR RATIO RUN SUPERFICIAL TEMPERATURE 2nd/prim. VELOCITY IN RISER ( C ) (M/S) 17 870 1 6 18 870 1 6 5 886 2 6 6 870 2 6 10 870 0.5 7 The secondary a i r p o r t s , as shown in Figure 2 .4, suppl ied between 50 to 100 % of the primary flow and were located 0.9 m above the primary a i r d i s t r i b u t o r . The r i s e r v e l o c i t y i s - 102 - the s u p e r f i c i a l v e l o c i t y in the upper part of the column above the secondary a i r por t s . The gas concentration p r o f i l e s for runs 17, 18, 5, 6, and 10 appear in Figures 2.7, 2.8, 2.9, 2.10, and 2.11, r e s p e c t i v e l y . For each gas the actual measured concentrat ion is p lo t t ed against the non-dimensional radius r / R , where R refers to the cy l inder or core radius at that l e v e l . For a l l of the measurements i t was poss ib le to maintain s u f f i c i e n t sample gas flow to allow easy measurement, with the exception of readings taken r ight at the w a l l . For these measurements gas flow was poss ib le for 10 minutes at most before blockage occurred, necess i ta t ing probe and f i l t e r purging. This is understandable s ince a denser region of so l ids exis ts at the wall than in the i n t e r i o r of the cyclone. While i t was poss ib le to obtain measurements in the v i c i n i t y of the wall the values may be affected by excessive so l ids in the probe and f i l t e r port ions of the sampling system i t s e l f . Char caught in the probe could continue to burn, thereby increas ing the CO and C02 values whi le , reducing the measured O2 values . This effect is assumed to be n e g l i g i b l e , however, as the gases and p a r t i c u l a t e were qu ick ly cooled after ex trac t ion in the cool ing sect ion of the probe and thus were much less r e a c t i v e . T y p i c a l l y the so l id s caught by the secondary cyclone had a mean diameter between 40 to 58 pm (29). Trends for the various components measured were as fo l lows: - 103 - 25 20 CCCLI vs r/R e ° e $ a? ^ 15 } CM o 10 < - 5 + 0 0.5 r /R [ Oj 1 vs r /R [ NOx ] vs r /R ( CO 1 vs r/R 180 -2" 904 O o o O O C o 0.5 r/R f S0 2 ] vs r /R [ CH* ] vs r /R FIGURE E.7 Gas Concentration Profiles for run 17. p o r t # 4 - 104 - f CO J vs r / R [ COj 1 vs r / R a* 05f 33 U — 0 o u [ SOs 1 vs r /R S a a ~ 0.5f O 00 250 a 200 a a 150 4- O 55 * 100 50f FIGURE 3.8 Gas Concentration Profiles for run 18. port # 4 Qi u o -ia o • d d | ' Q M | Pi - o * ( ' ID I tA o 8 • d d I ' O M I " ' I I I | "OH | Pi t. X I M O I % I MI3 ! i n d add | to J •dd ( «OD I •dd J l 0 \ •dd | i n | -_*_fO ei add ( IQ | •dd f IQ | o ft. tA •dd I IQ3 J O M O add I IQD J -i O o • d d ( OD 1 •dd ( 0 3 1 • d d J 03 | add ( 0 3 | • d d | 03 | - 106 - [ C O 1 vs r /R [ CO* ] vs r/R 3? 20- 15 10 u 5 O 5 r /R [ SOt } vs r /R 200 - 1 0 D t o w 50 + o o 0.5 r /R [ CHU ] vs r /R Figure S.10 Gas Concentration Profiles for run 6. port # 5 [ GO ] vs r/R S a a — 30 o u * " 1 0 I COr 1 vs r/R 20T " o o o O r Qt 1 vs r/R 05 . r/R [ SOz 1 v? r/R - * 7' ~ 6 « 5 o o «4 o 2i- - 1 200 B a a o o — 100} o 0 0.5 r/R t o r/R [ CH4 1 vs r/R i o 5 2 + —' 1 as o o o o 9 , 0 5 1 r/R Figure 2 . 1 1 Gas Concentration Profiles for run 1 0 . p o r t # 4 - 108 - NO, : No r a d i a l trend for NOx is apparent from the traverses performed in the f ive experiments. Values in the 200 to 225 ppm range were t y p i c a l , with the exception of the traverse performed at port # 8 near the bottom of the cyclone. The discrepancy at th is l eve l is a t t r i b u t e d to d i f f i c u l t i e s in maintaining flow at this port due to b locking by p a r t i c l e s . SOj : Values between 0 and 140 ppm [SO2 ] were observed with in the cyclone, but with no evidence of a c lear trend. In Figure 2.10 the lowest SO2 concentrations were often observed near the w a l l , probably because of the denser so l id s region e x i s t i n g near the w a l l . While no sorbents were added during the runs, there may be sorpt ion by calcium or other elements in the ash accounting for the reduced values . CH4 : During most tests the gas analyses were unable to detect measurable [ C H 4 ] . Only in runs 10 and 18 there were detectable values , these being of the order of 0.002 to 0.004 percent. No r a d i a l trend is evident . CO2 : Values between 16.1 and 19.3 percent were measured within the cyclone and a s l i g h t r a d i a l trend is observed in - 109 - three of the traverses . Combusting char p a r t i c l e s in the denser so l id s region near the wall are l i k e l y responsible for the observed small increase near the w a l l . CO: As with the [ C O 2 ] , [CO] tends to be somewhat greater at the wall than in the i n t e r i o r in f ive of the r a d i a l t raverses . Four traverses show a doubling of the [CO], again i n d i c a t i n g some char combustion ins ide the secondary cyclone. Os : No c lear trends in the [ O2 ] were seen in the p r o f i l e s i n d i c a t i n g that the extent of combustion occurr ing within the cyclone is rather smal l , at least for th i s f u e l . Comparing the gas residence times of the cyclone and the r i s e r i t i s c lear that the cylone volume is less than that of the r i s e r , approximately 60 % of the r i s e r volume. As well because there is a greater chance for gas short c i r c u i t i n g in the cyclone than in the reactor the reduced extent of combustion is not s u r p r i s i n g . Gas sampling for each run lasted less than three hours and during that time the combustor was maintained as c lo se ly as poss ib le at steady state condi t ions . The primary i n d i c a t i o n of steady operation was the f lue gas 02 concentration which was held between 3.5 and 6 percent for a l l readings, and 3.5 to 4 . 5 percent for more than 90% of - 110 - the time. Due to the nature of the experiments some scatter in the measured values was inev i tab le and clouded any subtle trends that may have ex i s ted . As w e l l , the flow in a cyclone is very well mixed and any regional combustion a c t i v i t y would be d i f f i c u l t to detect . For example i f combustion was preferred at a given rad ius , one would only expect to f ind increased combustion products concentrations i f the flow was s u f f i c i e n t l y segregated. This is not t y p i c a l l y the case in cyclones as the flow is more turbulent and well mixed than segregated. An improved experiment would see continuous monitoring at several r a d i a l points simultaneously (rather than s e q u e n t i a l l y ) , with the focus being placed on the region near the w a l l . 2.5 Conclusions and Recommendations The main features of part II of th is work are as fo l lows: 1. Radial gas concentrat ion p r o f i l e s wi th in a secondary cyclone serving a high temperature C i r c u l a t i n g F l u i d i z e d Bed Combustor operating with a low sulphur coal as fuel have been measured and presented. 2. Radial gradients in the gas concentrat ion p r o f i l e were n e g l i g i b l e for [NO x ] , [SOj] , [CH41, and [ O j ] . - I l l - 3. Near the wall [CO] leve ls increased, as d i d [ C O 2 ] , suggesting increased char combustion in this zone. It is recommended that attempts be made to determine s i m i l a r r a d i a l gas concentrat ion p r o f i l e s in the primary cyclone of the UBC p i l o t scale CFB combustor. This would require a probe capable of sampling gases in the much denser so l id s regions of the primary cyclone. It is also recommended that simultaneous monitoring at several r a d i a l points be performed as opposed to a sequential sampling. This would reduce the uncertainty imposed on the r a d i a l p r o f i l e s due to the time varying nature of the species concentration within the gas flow ( i . e . non-steady state operat ion) . P r o f i l e s should be at ta ined with a less reac t ive coal ( e .g . anthraci te ) and a high sulphur coal ( e .g . Minto ) . Cyclone react ions may play a more important ro le for these f u e l s . - 1 1 2 - NOMENCLATURE a Inlet depth ( m ) A Coef f i c i ent ( - ) ^ A c Centr i fuga l a c c e l e r a t i o n ( m/s 2 ) A B G r a v i t a t i o n a l a c c e l e r a t i o n ( m/s 2 ) Ai Inlet area ( m2 ) A P P a r t i c l e Area (m 2) b Inlet width ( m ) be Inlet width ( m ) bi Coe f f i c i en t ( m ) ci F r a c t i o n by mass for channel i of catch. ( - ) cm Sol ids loading (g/m 3) C o " Zero load " dust concentrat ion ( g r a i n s / f t 3 , g/m 3 ) c« Dust concentrat ion ( g r a i n s / f t 3 ) C Tota l mass caught ( kg ) C Cunningham correc t ion fac tor ( - ) C Dust concentrat ion ( g r a i n s / f t 3 , g/m 3 ) CD Drag c o e f f i c i e n t ( - ) d c r i t C r i t i c a l p a r t i c l e diameter ( m ) dp P a r t i c l e ' diameter ( um ) d P a Aerodynamic diameter ( g /cm 3 )o.5 dpC Feed aerosol mass medial diameter ( cm ) dpso P a r t i c l e diameter caught with 50% e f f i c i e n c y (mi crons) D Body diameter ( m ) - 113 - Db • Bottom diameter ( m ) D c Cyclone diameter ( m ) DH Hydraul ic diameter of cyclone i n l e t ( m ) Do Outlet diameter ( m ) Dz Imaginary cy l inder diameter ( m ) e 0 Zero load c o l l e c t i o n e f f i c i e n c y ( % ) E C o l l e c t i o n e f f i c i e n c y ( % ) E o o r r . Loading corrected c o l l e c t i o n e f f i c i e n c y ( % ) E e xper . Experimental c o l l e c t i o n e f f i c i e n c y ( % ) Ei C o l l e c t i o n e f f i c i e n c y for channel i . ( % ) Eo C o l l e c t i o n e f f i c i e n c y at low loading ( % ) f i Feed mass for channel i ( kg ) Fo Centr i fuga l force ( N ) FD P a r t i c l e drag force ( N ) G Cyclone conf igurat ion parameter. ( - ) Ga G a l i l e o number ( - ) he Inlet depth ( m ) Ho Overa l l height ( m ) ki Coe f f i c i en t ( - ) kpi R e s t i t u t i o n c o e f f i c i e n t ( - ) l i F r a c t i o n by mass for channel i of l o s s . ( kg ) L Experimental so l id s loading ( kg/kg s o l i d s / a i r ) LB 2 4 P a r t i c l e loading in run B24 ( kg/kg s o l i d s / a i r ) Ley Cyl inder length ( m ) Lni Tota l mass passing cyclone ( kg ) Lo Outlet length ( m ) L P P a r t i c l e loading r a t i o ( kg/kg s o l i d s / a i r ) LFAC Apparent gas v i s c o s i t y c o e f f i c i e n t ( - ) m P a r t i c l e mass ( kg ) -114- n Vortex exponent ( - ) N P a r t i c l e path revolut ions ( - ) N R E t Flow Reynolds number ( - ) N R e p P a r t i c l e Reynolds number ( - ) N R grp Radial p a r t i c l e Reynolds number ( - ) N B t Stokes number ( - ) Ns t 5 0 Stokes number for cut diameter ( - ) P(e) ^ P r o b a b i l i t y associated with high loading E ( - ) P(Eo) P r o b a b i l i t y associated with low loading E ( - ) Q Gas flow ( m3 /s ) Qc Gas s p l i t r a t i o ( - ) cyclone radius ( m ) r i Vortex f inder radius ( m ) R Radial co-ordinate ( m ) R. Radius of imaginary cy l inder ( m ) S° • 3 Dimensionless p a r t i c l e diameter ( m ) SFACTOR S a l t a t i o n factor ( - ) T A ( X ) Predicted c o l l e c t i o n e f f i c i e n c y ( - ) V, Inlet V e l o c i t y ( m/s ) Vi Inlet v e l o c i t y ( m/s ) v P Radial p a r t i c l e v e l o c i t y component ( m/s ) V,p Radial p a r t i c l e v e l o c i t y ( m/s ) V r e Radial gas v e l o c i t y ( m/s ) Vtp Tangential p a r t i c l e v e l o c i t y ( m/s ) Vfl P a r t i c l e s a l t a t i o n v e l o c i t y ( m/s ) - 1 1 5 - Greek symbols p Gas density ( kg/m 3) p P P a r t i c l e densi ty ( kg/m 3 ) XG Mass median p a r t i c l e diameter ( m ) x Shape factor ( - ) uK Gas v i s c o s i t y ( kg/ms) u t t ; p E f f e c t i v e gas v i s c o s i t y ( kg/ms ) - 116 - References 1. Sorbent opt imizat ion for FBC technology, Chatham C i r c u l a t i n g F l u i d i z e d Bed Demonstration pro jec t , Energy Mines, and Resources report (May 1989). (j£>. Stairmand, C . J . , "The design and performance of cyclone s e p a r a t o r s . , " Trans . Instn. Chem. Engrs. ( V o l . 29, 1951): pp. 357-383. (Ttj. Beeckmans, J . M. , Aerosol Measurement; Lundgren, D. , E d . ; U n i v e r s i t y at F l o r i d a Press , G a i n s v i l l e , F L , 1979. 4. C a l v e r t , S . , Englund, H . , Handbook of A i r P o l l u t i o n Technology, New York: John Wiley and Sons, 1984, pp. 320-329. 5. Strauss , W. ,„ I n d u s t r i a l gas Cleaning 2nd edition., New York: Pergamon Press , 1975, pp. 238-276. (6\ L e i t h , D . , Mehta, J . , "Cyclone performance and Design," Atmospheric Environment (1973 V o l . 7): pp. 527-549. ' / f T ) . Gauthier , T . A . , B r i e n s , C . L . , G a l t i e r , P . , "Uniflow ^ cyclone e f f i c i e n c y study," Dept. of Chemical and Biochemical Engineer ing , Un ivers i ty of Western Ontario . 8. Ter Linden, A. J . , T o n i n d u s t r i e - Z e i tung, 2 2( i i i ) 49 (1953). 9. Mukhopadhyay, S . N . , Chowdhury, K . C . Roy, "The c o l l e c t i n g e f f i c i e n c y of a cyclone separator," , B r i t . Chem. Eng. ( A p r i l 1970): V o l . 15, No. 4. pp. 529 10. Whitby, K . T . , Lundgren, D.A. , Jordan, R . C , J . A i r . P o l l u t i o n Control Assoc ia t ion 11,(1961): pp. 503-504. - 117 - V \1_1/. Parker, R. , J a i n , R . , and C a l v e r t , . "Par t i c l e c o l l e c t i o n in cyclones at a high temperature and pressure," Env. S c i . Tech. (1981): 15 (4) pp. 451-458. 12. Wheeldon, J . M . , Burnard, J . K . "Performance of Cyclones in the Off-Gas Path of a Pressurized F l u i d i z e d Bed Combustor," F i l t r a t i o n and Separation (May June 1987): pp. 178-187. (13) D i e t z , P .W. , "Co l l ec t ion e f f i c i e n c y of cyclone 1/ separators ," AIChE J o u r n a l . (1981), ( V o l . 27, No. 6) 14. Gloger , J . , Hanke, S . , Energietechnik (1971): 21(4) 158-164. 15. Mothes, H . , L d f f l e r , F . , "Motion and Separation of p a r t i c l e in a cyclone," Chem. Engg. Process. (1984) 18 pp. 323-331. 16. Abraham, F . F . , Physics F lu ids (1970); 13,- pp.. 2194-5. (17.' " Cyclone Dust C o l l e c t o r s , " API p u b l i c a t i o n , A . P . I . New York (1955). 18,1 American Petroleum Ins t i tu te (API) "Manual on Disposal of Ref inery Wastes," API p u b l i c a t i o n 931, Chp. 11 (1975). 19/. S p r o u l l , W . T . , "Effect of Dust Concentration Upon the Gas Flow Capacity of a cyc lon ic C o l l e c t o r , " J . A ir Po11. Control Assn. 16 (8) , pp. 439-441 (1966). 20. Massey, B . S . , Mechanics of f l u i d s , 4th ed. New York: Van Nostrand Reinhold Company, 1979 pp. 222 -227. 21. Abrahamson, J . , A l l e n , R . W . K . , "The e f f i c i e n c y of conventional reverse return cyclones at high temperatures," ChemE SYMPOSIUM Series No. 99. (1988) . 22. Ogawa, A . , " Theore t i ca l approach with Markov Process Model to Seperation Processes of Cyclone dust co l l ec tor depending on feed dust concentrat ion," J . Col 1. Engineer ing , Nihon U n i v . , A-26, (March 1985). Note: Referenced in 41. - 118 - 23. Masin, J . G . , Kock, W . H . , "Cyclone E f f i c i e n c y and Pressure Drop Corre la t ions in O i l Shale Retor t s ," Environmental Progress Vol 5, No. 2, (May 1986), pp. 116 -122. 24]. Mothes, M. , L O f f l e r , F . , "Motion and depos i t ion of p a r t i c l e in cyclones," Ger. Chem. E n g . , 8 (1985,) pp. 223 -233. 25. Boysan, F . , Ayers, W . H . , Swithenbank, J . , "A Fundamental Mathematical Model l ing Approach to Cyclone Design," Trans. IChem Engg. V o l . 60, 1982. 26. Laurendeau, N. M . , "Heterogeneous K i n e t i c s of coal char g a s i f i c a t i o n and combustion," Prog Energy Combust. S c i , V o l . 4 , pp. 221-270 (1978). 27. Grace, J . R . and Lim C . J . , " C i r c u l a t i n g F l u i d i z e d Bed Combustion of Coa l , Woodwastes and P i t c h , " F i n a l report prepared for Energy, Mines and Resources Canada, under contract 24ST.23440-6-9007 (1987). 28. A . S . H . R . A . E . , "Flow measurements, instruments and apparatus," ASME Power Test Codes. Part 5 of Ch. 4. (1959). 29. J . R. Grace, C . M . H . Brereton, C . J . Lim, R. Legros, R . C . Senior, R . L . Wu, J . R . M u i r , R. Engman " C i r c u l a t i n g F l u i d i z e d Bed Combustion of Western Canadian Fue l s ," F i n a l report prepared for Energy Mines and Resources Canada, under contract 52 SS.23440-7-9136 August 1989. 30. R. Legros, C . M . H . Brereton, C . J . Lim, H. L i , J . R . Grace and E . J . Anthony . "Combustion c h a r a c t e r i s t i c s of d i f f erent fuels in a P i l o t Scale C i r c u l a t i n g F l u i d i z e d Bed Combustor," Proc . Conference on F l u i d i z e d Bed Combust ion . 661-666 (1989) 31. Wu, L . R . , PhD. the s i s , "Heat Transfer in C i r c u l a t i n g F l u i d i z e d Beds" U n i v e r s i t y of B r i t i s h Columbia, Department of Chemical Engineering (1989). 32. M. F . Coutur ier , B. Doucette and S. P o o l p a l , " A Study of Gas concentrat ion , Sol ids Loading and Temperature p r o f i l e s within the Chatham CFB Combustor," a report to ENERGY MINES AND RESOURCES CANADA, (Nov. 1989). 33. Stern , A . C , gen. ed. , A i r P o l l u t i o n 3rd e d i t i o n , New York: Academic Press , 1977, chapter 3, pp. 98-136. - 119 - 3 4 . R a z g a i t i s , R . , Guenther, D . A . , "Separation E f f i c i e n c y of a Cyclone Separator with a Turbulence-Suppressing Rotat ing Insert ," Transactions of the ASME V o l . 1 0 3 , (July 1 9 8 1 ) . 3 5 . M . F . C o u t u r i e r , Professor,Department of Chemical Engineering U n i v e r s i t y of New Brunswick, Personnel commun i cat i on. 3 6 . Sucec, J . , Heat Trans fer , WM.C.Brown Publ i shers : Dubuque, Iowa ( 1 9 8 5 ) , p. A l l 3 7 . S e v i l l e , J . , et a l . , ( 1984 ) P a r t i c l e Charac ter i za t ion V o l . 1 , n o . l Weinheim, Federal republ ic of Germany (July 1984) 3 8 . J . Zhao, J . R. Grace, C . J . Lim, C . M . H . Brereton, R. Legros, and E . J . Anthony "NOx emissions in a p i l o t scale c i r c u l a t i n g f l u i d i z e d bed combustor," in Proceedings of EPRI/EPA 1989 Jo int Symposium on Stat ionary Combustion NOx C o n t r o l . 3 9 . L e i t h , D . , W. L i c h t , "The C o l l e c t i o n E f f i c i e n c y of Cyclone Type P a r t i c l e C o l l e c t o r s - A New Theore t i ca l Approach," A i r P o l l u t i o n and I t ' s C o n t r o l , AIChE Symposium Series 1 2 6 , 6 8 , 1 9 7 2 . 4 0 . Turner 4 1 . Patterson, P . A . , PhD. Thes i s , "High Temperature Cyclones", Department of Chemical Engineering, M c G i l l U n i v e r s i t y , Montreal ( 1 9 8 9 ) . - 120 - A P P E N D I X PRELIMINARY TESTS PART I CONDITIONS SOLIDS: 11 um mean d i a . f l y a s h . I n l e t v e l o c i t y : ^.4-5 m/s Configura t i o n : CI ( base case ) var i a b l e vortex finder length. V.F. position L/D CFULLY ADJUSTABLE V.f.> L/H collection efficiency o.is 0.11 50.0 0,59 0.54 73.9 0.81 i 0.74 i ! 96.3 PART II CONDITIONS SOLIDS: 50 um mean d i a . f l y a s h . I n l e t v e l o c i t y : 5.^ m/s Configuration: 1. C3 without extended vortex f i n d e r . S. v a r i a b l e vortex f i n d e r length. V.F. posliior: | L / D L/H collection efficiency EXISTING 0.21 0.23 41 V. PROPOSED O.o3 0.57 47 y. '.nodl'FlcQ "tions FIGURE Al Shake down test summary. A l l runs performed before run B l . ORFICE FLOW 0 4 8 12 16 20 24 28 PRESSURE DROP (cm water manometer) FIGURE AS O r i f i c e flow curve. M 3/s vs p r e s s u r e drop. - 123 - Figure A3 Data logging program serving UBC model cyclone apparatus. REM PROGRAM FOR INSTANTANEOUS SAMPLING OF LOAD C E L L S OPEN " c y c o . Q U T " FOR OUTPUT AS #2 OPEN a C O r f l : 9 6 0 0 , N , 8 , l , C S t D S " FOR RANDOM AS t l INPUT " r u n n u e b e r ? " , ni PRINT I E . " r u n n u a b e r = " , ' n t 3 T I K E $ = "00:00:00" 5 ta« = TIME*. 7 s e c o n d s = VAL(KID$(ts$, 1, 2)) * 3600 + VAL(NID$(tB$, 4, 2)1 t 60 • V S L < H I D $ ( t s $ , ?, 2)> 9 I F ( s e c o n d s - hh) = 1 THEN 20 20 PRINT " l o g g i n g d a t a " h h = s e c o n d s PRINT II, "2"; INPUT #1, bt PRINT t l , "3"; INPUT t l , c * PRINT 12, V A L ( M I D $ ( b $ , 5, 6) ) , VfiL(MID$(c«, 5, 61) PRINT V A L ( M I D $ ( b $ , 5, 61), V A L ( M I D $ ( c $ , 5, 6)1 GOTO 5 30 PLEXIGLAS MODEL CYCLONE / MULTICLONE FLUIDIZED SEAL WIND BOX - 4 - — TO BAGHOUSE PLATE VALVE LARGE MEASUREMENT VESSEL PLATE VALVE SMALL MEASUREMENT VESSEL 3 FROM BLOWER Figure A4 Schematic of attempted recycle system schematic showing a high so l ids loading feed and measurement vessels , mult ic lone and bag f i l t e r arrangements. -125- PAGE 12S INSERTED FOR PAGE NUMBER CONTINUITY - 126 - F i g u r e A6 UBC CFBC s o l i d f u e l a n a l y s i s f o r run B17 ( 2 9 ) . Proximate and Ultimate Analyses of Solid Fuel Proximate Analyses (as received) Highvale Coal Volatile Matter 30.5 Fixed Carbon 42.1 Ash 12.2 Moisture 15.2 Ultimate Analyses (dry basis) Carbon 62.4 Hydrogen 3.6 Nitrogen 0.8 Sulphur 0.2 Oxygen (by difference) 18.7 Ash 14.3 Higher Heating Value (MJ/kg) 24.0 - 127 - + U -1 Gl 9 8 7 6 5 4- 3 2 -t + '0 "3 l b l b 2 b 2 ^ 3 b 3» 4Q 45 p a r t i c l e dLi -aw^ t < j u » m > Figure A7 Mass balance, as performed on a per channel basis for run BIO. Image analysis p a r t i c l e s ize d i s t r i b u t i o n s . Fines loss (below 15 microns) a t t r ibuted to f i l t e r i n e f f i c i e n c y . - 128 - Figure A8 Temperature data for Part I experiments. RUN # Dry bulb temperature Wet bulb temperature »C » C B l 12 10 B2 15 11 B3 12 8 B4 11 7 B5 11 7 B6 16 12 B7 16 12 B8 16 12 B9 19 15 BIO 19 15 B l l 22 18 B12 18 15 B13 18 15 B14 18 15 B15 18 15 B16 18 15 B17 18 15 B18 18 15 B 1 9 * 18 15 B20 18 15 B21 22 19 B22 22 19 B23 22 19 B24 22 19 B25 22 19 B26 22 19 B27 22 19 B28 22 19 B29 22 19 B30 22 19 B31 22 19 B32 15 11 B33 15 11 B34 15 11 B35 15 11 - 129 - FIGURE A9 COLLECTION EFFICIENCY DATA &0N B4 RON B10 s i x e e f f i c i e n c y s i x e c r o D S % • i c r o n s 1.7 0 0.74 7.9 0 2.23 8.11 35 3.71 8.32 32 5.2 8.53 39 6.68 8.75 35 8.16 8.98 32 9.65 9.21 44 11.13 9.45 46 12.62 9.(9 48 14.1 9.94 49 15.59 10.2 54 to 10.46 54 46.76 10.73 65 11.01 62 11.29 69 11.58 11 19 75 1 1 . od 12.19 80 5 78 12.8! 86 13.IS 88 13.49 89 13.84 90 14.2 92 14.57 94 14.94 95 15.53 93 15.72 96 16.13 99 16.54 99 16.97 99 17.41 100 17.86 98 18.32 98 18.79 100 19.28 99 19.77 100 20.28 98 20.81 to 100 e f f i c i e n c y 100 S.4 - 130 - FIGURE A9 DATA SUMMARY FOR PARTS ( AND (I (CONTINUED) CHATHAM COLLECTION EFFICIENCY DATA DATA size MICRONS 7.32 7.51 7.70 7.90 8.11 8.32 8.53 8.75 8.98 9.2L 9.45 9.69 9.94 10.20 10.46 10.73 11.01 11.29 11.58 11.88 12.19 12.50 12.82 13.16 13.49 13.84 14.20 14.57 14.94 15.33 15.72 16.13 16.54 16.97 17.41 17.86 18.32 18.79 19.28 19.77 20.28 20.81 21.34 21.89 22.46 23.04 23.63 24.24 24.87 25.51 26.17 26.84 27.54 28.25 28.97 29.72 30.49 31.28 32.08 32.91 33.76 34.63 35.52 36.44 37.38 38.35 39.33 40.35 LOSS FRACTION 5.97E-05 5.97E-05 9.42E-05 6.59E-05 9.89E-05 9.89E-05 1.04E-04 1.08E-04 1.44E-04 1.51E-04 1.518-04 1.96E-04 2.45E-04 2.45E-04 2.97E-04 3.52E-04 3.96E-04 5.01B-04 5.65E-04 7.30E-04 7.79E-04 9.55E-04 1.12E-03 1.30E-03 1.65E-03 2.03E-03 2.21E-03 2.73E-03 3.31E-03 3.86E-03 4.70E-03 5.348-03 6.82R-03 7.94E-03 9.61E-03 1.16E-02 1.34E-02 1.61E-02 1.88E-02 2.19E-02 2.58E-02 2.85E-02 3.23E-02 3.69E-02 3.95E-02 4.26E-02 4.62E-02 4.73E-02 4.72E-02 4.92E-02 4.78E-02 4.718-02 4.74E-02 4.54E-02 4.34E-02 3.698-02 3.39E-02 2.69E-02 2.25E-02 1 CATCH FRACTION 0.008+00 0.008+00 0.008+00 0.00E+00 0.00E+00 O.OOE+00 0.00E+00 O.OOE+00 0.1 0.1 O.OOE+00 5.55E-06 0.00E+00 5.77E-06 O.OOE+00 6.21E-06 O.OOE+00 O.OOE+00 O.OOE+00 O.OOE+00 6.88E-06 O.OOE+00 O.OOE+00 O.OOE+00 O.OOE+00 7.99E-06 O.OOE+00 O.OOE+00 O.OOE+00 O.OOE+00 9.10E-06 O.OOE+00 O.OOE+00 9.99E-06 1.27E-06 O.OOE+00 O.OOE+00 1.09E-05 67E-02 . 228-02 8.94E-03 7.80E-03 6.05E-03 5.33E-03 3.38E-03 4.00E-03 1.46E-06 O.OOE+00 O.OOE+00 1.57E-06 0.00E+00 1.63E-06 25E-06 0.001+00 ..76E-06 1.65E-05 69E-06 93B-06 [.046-05 3.97E-06 4.13E-06 1.34E-05 6.53E-06 1.39E-05 t.148-05 1:718-05 1.208-05 2.28E-05 2.618-05 2.93E-05 3.29E-05 6.93E-0S 6.578-05 EFFICIENCY 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.22 0.00 0.19 0.78 0.00 0.17 1.58 0.34 0.19 1.00 0.38 0.42 1.40 0.81 1.85 1.92 3.37 3.20 7.95 11.84 14.7205 19.9943 37.4109 47.1316 CHATHAM COLLECTION EFFICIENCY DATA DATA s i i e MICRONS 41.39 42.46 43.55 44.68 45.83 47.01 48.23 49.47 50.75 52.05 53.40 54.78 56.19 57.64 59.13 60.65 62.22 63.82 65.47 67.16 68.89 70.67 72.49 74.36 76.28 78.25 80.26 82.34 84.46 86.64 88.87 91.17 93.52 95.93 98.41 100.95 103.55 106.22 108.96 111.77 114.66 117.62 120.65 123.76 126.96 130.23 133.59 137.04 140.57 144.20 147.92 151.74 155.65 159.67 163.79 168.01 LOSS FRACTION 3.10E-03 3.198-03 2.22E-03 2.13E-03 1.628-03 1.858-03 9.588-04 9.73E-04 1.418-03 1.028-03 4.248-04 6.508-04 1.338-03 1.148-03 1.178-03 7.168-04 9.86E-04 1.008-03 1.04E-03 2.398-03 5.438-04 1.688-03 1.148-03 0.008+00 1.518-03 1.558-03 1.89E-03 6.538-04 1.66E-03 2.408-03 1.058-03 1.088-03 0.008+00 2.278-03 1.56E-03 0.1 0.008+00 O.OOE+00 0.008+00 O.OOE+00 0.008+00 O.OOE+00 0.008+00 0.1 0.1 0.008+00 0.008+00 O.OOE+00 O.OOE+00 O.OOE+OO O.OOE+00 O.OOE+00 O.OOE+00 O.OOE+00 0.008+00 CATCH FRACTION 8.50E-05 1.148-04 1.07E-04 1.128-04 9.92E-05 9.768-05 1.318-04 2.148-04 2.298-04 2.958-04 3.678-04 4.128-04 5.108-04 6.558-04 8.238-04 9.928-04 1.288-03 1.638-03 2.128-03 2.828-03 3.398-03 4.538-03 6.188-03 7.788-03 9.238-03 1.078-02 1.278-02 1.398-02 1.57E-02 1.678-02 1.68E-02 1.938-02 1.838-02 1.908-02 1.76E-02 1.538-02 1.50E-02 2.068-02 2.528-02 3.10E-02 3.12E-02 3.61E-02 4.43E-02 4.97E-02 5.09E-02 4.60E-02 4.668-02 4.76E-02 5.638-02 6.14E-02 6.248-02 6.36E-02 4.278-02 3.38E-02 7.408-03 8.33E-03 EFFICIENCY % 49.40 62.84 60.67 69.80 68.17 73.40 76.48 91.12 91.53 90.59 94.29 97.81 97.30 95.77 97.08 97.50 98.79 98.79 98.98 99.21 98.49 99.74 99.41 99.68 100.00 99.69 99.74 99.70 99.91 99.78 99.69 99.88 99.87 100.00 99.72 99.78 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 - 131 r GAS CONCENTRATION DATA FOR ION 17 READING r/1 0! CO CO! S02 CB4 HOX % ppy % ppy % PPM 0.13 4 41 20 107 0 209 O.SO 3 48 21 297 0 214 0.S7 4 47 20 305 0 219 0.83 4 46 20 297 0 214 1.00 3 16S 20 132 0 209 O.SO 4 41 20 109 0 211 0.83 4 41 20 205 - 0 215 0.33 4 4! 20 130 o. 214 0.67 4 41 20 203 0 215 1.00 4 134 20 66 0 215 O.SO 4 31 20 S3 0 216 0.83 4 37 20 155 0 216 1.00 3 88 ! l 176 0 203 GAS CONCENTRATION DATA FOR RON 18 reading t r/R 02 CO CO! S02 CB4 NOX % Dpi % ppa t p p i 0 4 36 20 0 0 196 0 4 34 19 0 0 207 0 5 41 19 0 0 201 1 4 46 20 0 0 207 1 3 54 21 0 0 201 I 2 134 22 0 0 190 Radial Gas Concentration p r o f i l e s RON 5 PORT 2 r/R 02 CO C02 S02 CR4 NOX 1.0 4.1 35 18.4 50 0 198 0.8 4.1 20 18.4 116 0.002 195 0.7 4 20 18.4 113 0.002 195 0.5 4 19 18.5 114 0.002 190 0.3 3.9 20 18.5 112 0 190 0.2 3.9 22 18.4 104 0 185 0.0 3.9 24 18.4 101 0 180 0.0 3.9 26 18.3 90 0 180 0.3 4.3 21 18.4 112 0 18S 0.7 4.1 20 18.6 120 0 200 0.2 4.2 ! l 18.5 116 0 200 1.0 4.3 39 18.3 40 0 200 0.5 4.3 ! l 18.5 101 0 189 0.0 4.3 21 18.3 111 0 201 1.0 4.2 SO 18.5 40 0 20S Radial Gas Concentration p r o f i l e s , Run 5 Port 4 r/R 02 CO C02 S02 cm NOX 0.4 4.7 25 17.8 78 0 201 0.6 4.5 33 18 89 0 204 0.8 4.2 34 18.1 41 0 202 0.9 4.4 40 17.8 34 0 201 0.9 4.3 42 18.1 10 0 200 1.0 4.3 51 17.8 10 0 200 F i g u r e A 9 Data summary f o r Parts I and I I . Note: A l l gas c o n c e n t r a t i o n data i n ppm except values for 0 2 , C H 4 , and COz . - 132 - Radial Gas Concentration profiles, Ran 5 Port 5 02 CO CO! SO! CH4 NOX 5.7 24 17.1 40 0 1!0 4.6 25 17.9 48 0 195 4.5 22 17.8 77 0 201 4 19 17.9 87 0 202 4.5 18 18.1 89 0 203 4.6 17 18.1 . 90 0 203 4.5 19 18.1 90 0 204 4.7 20 18.1 73 0 205 >S CONCENTRATION DATA RON 5, PORT 6 til 02 CO C02 S02 CH4 NOX 0.33 3.58 35.76 19.78 0.00 0.00 195.59 0.50 3.91 33.53 19.34 0.00 0.00 206.76 0.67 4.58 41.35 19.22 0.00 0.00 201.18 0.83 4.25 45.82 20.01 0.00 0.00 206.76 1.00 2.91 53.65 20.90 0.00 0.00 101.18 0.50 2.12 134.12 21.57 0.00 0.00 190.00 GAS CONCENTRATION DATA RON 5, PORT 8 I til 02 CO CO! SO! CH4 NOX 1 0.46 4.8 30 17.7 90 0 170 2 0.67 4.3 24 17.9 96 0 !0! 3 0.83 4.5 !7 17.8 88 0 203 4 1.00 4.6 59 17.5 !! 0 145 ( RON t 6 RADIAL GAS CONCENTRATION DATA r/R O! CO CO! S02 CU4 NOX 1.00 4.10 63.00 17.30 0.00 0.00 175.00 0.83 4.!0 39.00 17.00 0.00 0.00 225.00 0.67 4.20 32.00 16.90 0.00 0.00 220.00 0.33 4.40 31.00 17.20 0.00 0.00 225.00 0.17 5.20 29.00 16.40 0.00 0.00 230.00 0.00 4.00 40.00 17.60 0.00 0.00 230.00 0.00 2.80 48.00 18.40 0.00 0.00 230.00 1.00 NA 78.00 18.40 0.00 0.00 220.00 GAS CONCENTRATION DATA FOR RON 6 RON t t/l 02 CO C02 S02 CB4 2 1.00 4.6 23 16.8 50 0.004 6 0.29 5 23 16.4 120 0.003 8 0.50 4.6 18 16.9 140 0.003 10 0.79 5.5 16 16.1 135 0.003 12 0.13 5.1 22 16.4 90 0.003 14 1.00 3.5 13 17.8 140 0.003 16 0.29 4.2S 16 17.3 140 0.003 F i g u r e A 9 D a t a s u m m a r y f o r P a r t s I a n d N o t e : A l l g a s c o n c e n t r a t i o n d a t a i n 1 f o r 0 2 , C H * , a n d C 0 2 • p p m e x c e p t v a l u e s

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
China 23 0
United States 2 1
City Views Downloads
Beijing 23 0
Mountain View 1 1
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items