OPTIMUM PROCESSING OF

1 MM BY ZERO COAL

Dennis I. Phillips

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

Doctor of Philosophy
in
Mining and Minerals Engineering

R.-H. Yoon, PhD, Chair
G. H. Luttrell, PhD
G. T. Adel, PhD
A. L. Wicks, PhD
E. R. Palowitch, PhD

April 1998
Blacksburg, Virginia

Keywords: Coal Processing, Coarse Coal Flotation,
Fine Coal, Column Flotation
Copyright 1998, Dennis I. Phillips

OPTIMUM PROCESSING OF
1 MM BY ZERO COAL

Dennis I. Phillips

(ABSTRACT)

Coal in the finer particle size ranges (below 1 mm) has always suffered from poor
cleaning efficiencies. This problem has been exacerbated in recent years with the increased
amount of high ash fines due to continuous mining machines and the mining of dirtier coal
seams. In the present work, it is proposed to improve overall plant efficiencies by processing
coarser coal in column flotation than is now commonly treated by that method. Column
flotation for coarse coal is supported by actual lab and plant test data that result in a full-scale
column plant installation. The fundamentals of coarse particle detachment from bubbles are
reviewed and a new simplified model is developed which better handles cubical and
rectangular coal particles.
Much of the lower efficiency of fine coal cleaning is due to poor size separation of
the fine-sized raw coal which results in misplaced high ash fines reporting to the coarser size
streams. By sending coarser material to column flotation, the finest size separation that takes
place in a plant can be as coarse as 0.5 mm or greater. The proper use of wash water in a
flotation column then becomes the best mechanism for desliming of the high ash clays. This
work quantifies the benefits of removing the high ash fines from the plant product and
increasing overall plant yield by increasing the amount of near-gravity coarse material. The
resulting yield gain is greater than that obtained from only the increased fine coal recovery.
Methods of column operation for improved coarse coal recovery are also evaluated.

ii

ACKNOWLEDGEMENTS

Great appreciation is expressed to my advisor, Dr. Roe-Hoan Yoon, for his
encouragement and guidance throughout the duration of this investigation. Special thanks
are also given to Dr. Gerald Luttrell and Dr. Greg Adel for their continued interest,
assistance, and inspiration. Thanks also to Dr. Al Wicks and Dr. E. R. Palowitch for their
advice and guidance.
The author would like to acknowledge the United States Department of Energy for
the funding of the in-plant testing part of this project through Contract No. DE-AC2292PC92208, Engineering Development of Advanced Physical Fine Coal Cleaning for
Premium Fuel Applications. Thanks also to those at the Lady Dunn Preparation Plant for all
the assistance during and after plant testing.
Special appreciation is expressed to Robert Martin for many valuable years of
guidance and discussions of optimal coal processing that inspired me to pursue this path.
Gratitude is also expressed to my fellow graduate students, Wayne Slusser, and other friends
in the department of Mining and Minerals Engineering.
Thanks to my parents for their continued encouragement over the years. Finally,
abundant thanks are given to my wife Rhonda, daughter Ashley, and son Austin for their
continued love and sacrifices which made this endeavor possible.

iii

................................ 8 2......................................................................................................................................................................................................................................................................................... 1 1...... 3 1....4 SCOPE OF RESEARCH ........................................................................................................ 5 1......................................................................................TABLE OF CONTENTS CHAPTER 1 INTRODUCTION. 7 CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT ...............1 INTRODUCTION ............................................................................................................5 REFERENCES ...................................................................... 8 2................................. 17 Multiple Particle Results................ 8 2............................................ 26 CHAPTER 3 PLANT TEST AND SCALE-UP .............6 REFERENCES .................................................... 1 History of Fine Coal Processing..3 A SIMPLIFIED DETACHMENT MODEL FOR CUBOIDAL PARTICLES ................................................................................................... 19 Single Particle Results ..................... 1 1....4 EXPERIMENTAL ......................................................................................................... 25 2............................................................5 SUMMARY AND CONCLUSIONS ...................................................................................................................................... 27 3......................2 CURRENT DETACHMENT MODELS .....2 CURRENT INDUSTRY PROBLEMS .......... 6 1.................................................3 POTENTIAL IMPROVEMENTS TO FINE COAL PROCESSING ........... 27 3............................1 FINE COAL PROCESSING ................................................................................ 21 2............................................................ 28 iv ..........2 PROJECT BACKGROUND .................... 4 Coarse Coal Flotation Not New............ 13 2.. 17 Sample Preparation and Procedure...................................................1 INTRODUCTION ...........................................................

...............................................................................................56 Second Series Revisited  0......................................................7 REFERENCES ...................................................................59 3......................................5 mm x 0.................................25 mm x 0 .................. 73 4..................................... 49 Second Series Revisited .......................................... 31 Pilot Column Circuit Description and Operation....... 63 Full-Scale Column Results .................25 mm x 0...................................................................First Series.................................................................................................................................................................................... 70 4........ 30 Test Column Feed Sources........................................................................................................................................................................................... 67 3........................................57 Second Series Revisited  0..... 70 4........... 65 3....................4 RESULTS AND DISCUSSION ........................................... 70 4.................5 FULL SIZE FLOTATION COLUMN SCALE-UP ....3 EXPERIMENTAL .......................................... 60 Applicable Pilot-Scale Testing .........................1 INTRODUCTION ...................49 Parametric Tests ............................................................................................................................................................. 54 Second Series Revisited  0.................. 72 4.....................................................................25 mm.......................................46 First Series  0................................ 30 Objectives Of Plant Testing...... 30 Test Plan and Installation ..................................................................................... 59 Lady Dunn Plant Circuit ....................2 ASH FOR MOISTURE SIMULATIONS.....6 SUMMARY AND CONCLUSIONS ... 30 30-Inch Column Chosen..... 60 Full-Scale Column Design ............................ 35 Parametric Testing..............................................................................................5 mm x 0 ........................................................25 mm ................................................. 28 3......................................... 39 First Series  0............................... 74 v ...............................................................................................................................5 REFERENCES ......................................................................................................................................4 CONCLUSIONS .............................. 35 Preliminary Flotation Testing................. 69 CHAPTER 4 YIELD ADVANTAGES OF IMPROVED DEWATERING ........ 38 Parametric Tests ...................................................................................44 First Series  0..........Plant Selection..........................5 x 0.........5 x 0.................................................................................................................................................................................................................. 32 3....................................3 BASE CASE AND MODIFICATIONS...........Second Series ....................................

.................................................................................................................... 93 CHAPTER 6 CONCLUSIONS ........................................................................ 99 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA ................................................................................................................................................ 91 5................. 78 5-Block seam............................................................................................................................................................................................ 80 Gravity Separations – Spirals and Heavy Media Cyclone (HMC) .........................................................................................................1 INTRODUCTION ................ 169 vi ........................................... 75 5.......................................... 87 General ................. 75 5................................................................................ 90 5............................................ 94 CHAPTER 7 RECOMMENDATIONS FOR FUTURE RESEARCH............................2 CASE CIRCUITS AND EVALUATION METHODS ................................. 78 Column Case Values ........................................................................................................ 87 Case C ....... 84 Yield Optimization................................................................................................... 97 APPENDIX A TABULATED PLANT TEST DATA................. 76 Column Flotation .......................................................................................... 86 Case A ............................................................................................................................................................. 83 Spirals.................. 126 APPENDIX C QUANTITATIVE EVALUATION DATA .................................................4 CONCLUSIONS ......................................................................................................................5 REFERENCES ................................................................................................................................................. 163 VITA .... 85 5.................. 86 Case B ............................3 RESULTS OF CASE EVALUATIONS ......................................................................... 80 Sizing Devices (Classifying Cyclones and Sieve Bends)............... 87 Case D .................................. 84 Heavy Media Cyclone (HMC) ..........................................................................CHAPTER 5 QUANTITATIVE EVALUATION OF THE APPLICATION OF COARSE COAL FLOTATION TO PLANT CIRCUITRY ...................................................... 77 Stockton Seam.......

......... 0.......................... A three-phase contact with upward surface tension force.......... as a function of the size of the aggregate........................ 18 Figure 2-7..................... .................. ..... 14 Figure 2-3.. Common coal processing methods and sizes............................. 2 Figure 2-1............................... . 45 vii ..... Common misconception of three-phase contact with cubical particle........................ 11 Figure 2-2.......... N2R (rpm2 x radius of motion).... Three-phase contact point on top of cuboid particle.................. Combustible recovery vs........... Predicted detachment force vs.... b) contact angle and c) surface tension.............................35)........ 15 Figure 2-6....... 33 Figure 3-3.................................................. 42 Figure 3-9.................................. Ash rejection and combustible recovery for 2-inch Laboratory and 30-inch column (Tests 101-110)..................................................... .................. 20 Figure 2-10......................... 38 Figure 3-7................. .........................  = 1................... Spherical particle at interface............................................. ( = 65o...... Adhesion strength of aggregates of spheres..................25 mm x 0........ 29 Figure 3-2. 23 Figure 2-12............. 20 Figure 2-9.............. Frother at 8.. 19 Figure 2-8............. ... Particles in tube during rotation........ ................... Liquid drop on hydrophobic solid.. Effect on detachment force of changes in a) particle shape......... no frother................ .. ..................... Pilot column test circuit arrangement.......... 22 Figure 2-11...............  = 72. 14 Figure 2-4............... 50% detachment point for various frother dosages................... First Series (401-416)................... 40 Figure 3-8............ 15 Figure 2-5...... ash for 2-inch Laboratory and 30-inch columns (Tests 101-110)............................ 37 Figure 3-6.... 10............ .............. Measured and predicted detachment force for each particle weight................. Centrifuge test device for particle detachment measurements.............. 3-D Response Plots................................. A 2 x 1 x ½ shape used for b) and c)......... 24 Figure 3-1..... First series (401-416) combustible recovery by particle size........................................... Size by size detachment force....................................................... 34 Figure 3-4............ 23 Figure 2-13.......... Performance of Microcel pilot column on coarsened cyclone overflow (tests 201-215)......................................................... ............ from Nutt (1960). Grade-Recovery plots by size for tests 401-416............................. and 12 ml/min....................................................... particle size for various shapes and of equal mass................... 10 particles each................................................................. Flotation column schematic showing bubble generation system.... .................... 36 Figure 3-5........ Lady Dunn Plant flowsheet during 30-inch column testing showing the two major feed sources.................LIST OF FIGURES Figure 1-1.........................

................... Base case is $20/ton sales........... First Series (401-416). Product ash by size from 2-inch column tests on high-vol 5-Block seam coal.................. 58 Figure 3-15.......... Effect of collector on combustibles recovery (pilot-scale)..........................25/100 BTU premium..5...........6....... diesel rates 250 and 500 g/Tonne............................... diesel dosage. 53 Figure 3-13......... Full-scale 4-meter diameter column installation at the Lady Dunn Preparation Plant.... 30% fines moisture................. 3-D Response Plots.. ..................... 84 viii ...25 mm........... 7..... Relationship of air fraction. ..................500 BTU...... Full-scale MicrocelTM column and bubble generation system............25 mm.... 0.................... feed rate range 80 and 93 g/min. Second Series (451-465) combustible recovery by particle size................. 66 Figure 3-21............ feed rate range 80 and 93 g/min....................... 48 Figure 3-11.......... Characteristic partition curve for gravity separation.................. Recovery by size from 2-inch column tests on high-vol 5-Block seam coal..... 62 Figure 3-17..... ...... 78 Figure 5-3..... 12............. ................ 79 Figure 5-4.3 ml/min...... Characteristic partition curve for classification................................. Size-by-size recovery in the 4-meter Microcel column........Figure 3-10... 10..... ....... Second Series grade-recovery plots by size for tests 451-465.......................... 77 Figure 5-2................. 63 Figure 3-18.............. Effect of feed rate increase on recovery by particle size for 30-inch pilot-scale column.... 82 Figure 5-5................... ........ . 55 Figure 3-14...................5 x 0.............................. $0. 9............ Second Series Revisited (401-416)..................................5 x 0.......... Simplified flowsheet (new circuit) for Lady Dunn preparation plant....... diesel rates 250 & 500 g/Tonne................... 65 Figure 3-20.................. 71 Figure 5-1........................ and % feed solids for the Second Series (451-465)......... 64 Figure 3-19...... 3-D Response Plots............................................. Particle size and devices for the case circuits... and 12 ml/min................. Results of 4-meter and pilot-scale column tests....... Total effect of moisture reduction on fines portion at dewatering cost of: a) $1 per ton of fines treated...... 0... 52 Figure 3-12............. b) $5 per ton of fines treated........... 67 Figure 4-1........ Frother at 5........................ ............... 61 Figure 3-16.. Frother at 8..... ..............................

5 (or 1.1 INTRODUCTION Fine Coal Processing Although much research regarding improvements in fine coal cleaning has been performed in the last twenty years.150 mm x 0 (plus 3/8 inch) (3/8 inch x 28 mesh) (< 28 mesh) (100 mesh x 0) History of Fine Coal Processing Fine coal processing on even a moderate scale had its beginnings as recently as the 1940’s and 1950’s. the processing of material smaller than 1 mm remains difficult and results in very poor efficiencies when compared to those of coarser sizes. Some of the common methods and sizes for coal processing are illustrated in Figure 1-1.5 mm 0. A major problem that is seldom properly addressed is that of high ash fine particles (slimes) reporting to a coarser size stream from which they are not properly removed or cleaned. This problem of misplaced sizes is often the source of much of the high ash and low yields found in fine coal cleaning. Phillips CHAPTER 1 1. CHAPTER 1 INTRODUCTION Page 1 . but the actual particle size at which one processing method gave way to another was never well defined and continues to vary from one company to another. whether real or perceived. separation efficiency is reduced as the particle size becomes smaller. While some cleaning devices have higher efficiencies than others. they often involve greater economic expense. The nomenclature of coal sizes varies but a general guide and one that will be used in this work is: Coarse Intermediate Fine Ultrafine +10 mm 10 x 0.Dennis I.0) mm x 0 0. there is little agreement as to what particle size ranges are best suited for each of the devices employed. For any coal cleaning device. Not only is there little consensus in the coal industry as to the overall best cleaning methods for particles finer than 1 mm.

Common coal processing methods and sizes. Cyclone 100 mm 6 mm Spirals 1 mm 0.M.M.5 mm Discard 0. Water-Only Cyclone 1990 to Present Conventional Flotation H.Dennis I. Deister Table 1965 to 1990 Conventional Flotation H.M. CHAPTER 1 INTRODUCTION Page 2 .150 mm 0 mm Particle Size Figure 1-1.V. Cyclone H.M.V. Phillips Early to Mid 1900’s Jig Discard or ship raw (dry) Hand Picking 1950 to 1975 Conventional Flotation H.

an additional classifying cyclone circuit. The underflow of the cyclone is then cleaned in spirals. Since heavy media cyclones replaced Deister tables for the intermediate sizes. The second problem is the high separating gravity at which a spiral makes a separation. the overall circuit suffers due to insufficient removal of the ultrafine fraction. In the distant past fine sizes were often discarded and in later years were cleaned on either Deister tables or in water-only cyclones. Even though several devices such as spirals and heavy media cyclones can process these sizes with reasonable efficiencies on the plus 0. The increased ash was due to the presence of minus 0. spiral concentrators (spirals) were chosen even though their efficiency is similar to that of Deister tables on the minus 1 mm coal (Deurbrouck and Palowitch 1963). Nearly all of the ultrafines feeding a spiral report to the clean coal product of the unit. occasionally. It is becoming increasingly popular to simplify circuitry by cleaning the coarse size along with the intermediate size in a heavy media cyclone.25 mm (Martin 1994). One is that the spirals actually clean only a narrow size range well and perform poorly on anything smaller than 0. The resulting higher ash product in effect requires compensation by lowering of the separating gravities on the coarser circuits. usually above a 1. For this reason as well as their low operating costs. The ultrafine sizes were historically shipped raw and dry or were discarded. 1. The author has seen 0. although some steam coal plants and most metallurgical (coking coal) plants use flotation since it remains the only method suitable for cleaning the ultrafine material.150 mm products have a 14% ash where a 10% ash was expected. There are three main problems with this circuit.150 mm material with an ash of over 40%.80 SG cut-point. Phillips Coarse coal was formerly often cleaned in a jig but is now most commonly cleaned in a heavy media vessel. CHAPTER 1 INTRODUCTION Page 3 .150 mm fraction. it was desirable to use a separate cleaning device for just the fine size in order to reduce magnetite consumption in the HMC circuit. This misplacement can be as much as 20% of the ultrafines (minus 0. Usually some attempt is made to further deslime the clean coal stream such as the use of sieve bends or even.5 x 0. resulting in lost yield. there is often sufficient fines of high ash to cause a considerable rise in the product ash. Conventional cell flotation (a trough shaped tank with agitators) is the type of flotation most often employed.2 Current Industry Problems Processing coal smaller than 1 mm in size has always been more difficult than cleaning coarser sizes and remains so today.150 mm) feeding the cyclone.Dennis I. Today the ultrafines are still discarded in many coal plants producing power plant fuel (steam coal). The intermediate sizes were formerly cleaned in jigs or Deister tables but are now almost totally processed in heavy media cyclones. although column flotation has gained a few operations in recent years. A common circuit today consists of sending minus 1 mm material to a classifying cyclone with the overflow (nominal 0.150 mm x 0) going to either flotation or discard. Even with this additional desliming. Ultrafine material that is misplaced to the underflow of the classifying cyclone creates the third problem with this common circuit.

5 mm was 8% ash while the total spiral product (all the minus 0. % Ash % 81. indicating the poor performance of the spirals. Table 1-1.150 mm material had an ash of 30.150 x 0. Truly proper sizing at 0. it contributed enough ash to change the total clean coal from 8. This problem is found to be worse with coal feeds having high clay contents which also cause inconsistent product ashes due to the variations in feed clay from hour to hour.7 8.150 mm. or even 1 mm.3 Individual Wt. Beech Plant total clean coal. if 0. Phillips Table 1-1 shows the clean coal data from an actual plant test (Phillips 1992) on a relatively new plant having heavy media cyclones and spirals with the minus 0. 1995).96 88.150 mm. For instance. 1991). or 0.25 mm.5 mm x 0 material can be processed in flotation columns.045 1.12 17. the plus 0. have been almost exclusively on material finer than 0. % Ash % 100.5 mm is difficult due to poor efficiencies of common devices and due to the high separating gravities for water based devices such as spirals. Without the misplaced minus 0.86 Cumulative Retained Wt.7 16.150 mm material.1 Ash % 7.22 37.2% of the total product.96 8.150 0 x 0. cleaning coal finer than 1 mm and especially finer than 0.8 8. the spiral product would have been 17% ash.06 26.0 20.57 3.97 96.0 7. By applying flotation columns to coarser sizes of 0. Size (mm) +1 0.8% ash to 9.9 9.48 Cumulative Passing Wt.0 9.80 98. Size cuts of 0.150 mm (100 mesh) discarded.150 mm cannot be economically achieved with today’s sizing equipment (Firth et al.Dennis I.8 2.1 1. With these screening devices.2 30.5 x 1 0.5 mm.3 7. unlike classifying CHAPTER 1 INTRODUCTION Page 4 .5 mm) was 20.5 mm can be made very well on vibrating screens and to a slightly lesser extent on sieve bends.045 x 0. The complications added by poor desliming (removing fine sized particles) increase the problems and inconsistencies.6% ash.22% and even though it represented only 3. A new device known as column flotation may provide some or most of the answer to the problems of cleaning fine coal.11 12. The limited locations in which columns have been applied in the coal industry. % 81.5 0. When properly applied.48 18.5%.22 1. The misplaced minus 0. In this sample.1 37. there is no need to make any size cuts near 0. the wash water in a column is the best desliming system available (Luttrell et al.3 6.86 Potential Improvements to Fine Coal Processing As previously discussed.0 9. many of the previously mentioned problems can be overcome.17 100.

Since heavy media cyclones perform well and have no equal on coal greater than 0. then there are two options to consider for improving the processing of coal finer than 1 mm. then a plant can consist of only heavy media cyclones above some size and column flotation on the finer size. As will be shown in later chapters. This circuit should be much simpler and more efficient than current practice due to the ease with which a coarser classification size cut can be made. Option 1 is to use spirals for the coal from 1 mm down to some size such as 0. Option 2 would involve heavy media cyclones down to 0.25 mm and use column flotation for all material finer than the lower size. These devices perform a much more positive size separation. This two circuit plant could conceivably have a very high efficiency on all sizes down to zero. during the 1980’s due to the belief that it was too difficult to float the coarser. Option 1 assumes that the practical upper limit for flotation is 0.5 or 0.5 or 1 mm.25 mm and that spirals will still be utilized above that.5 or 0.150 mm fraction and conventional flotation for the material finer than 0. they will also be more efficient.Dennis I. the current standard is the use of spirals for the 1 x 0. this overload resulted in the coarser particles being lost from the bubble due to the preferential loading of the finer sizes. The author believes that the coarser coal was cleaned well in the earlier years. If the amount of coal above 50 mm (2 inch) in size is insufficient to justify a heavy media vessel. Option 2 is obviously desirable due to its simplicity. The ultrafine ash is entrained in the froth normally and even more so when the cells are “pulled hard” to recover the coarse particles. it was actually relatively common to use conventional flotation cells on coal this coarse in the late 1960’s and 1970’s. but that during the later 1970’s and early 1980’s the increased use of continuous mining machines and the mining of dirtier coal seams brought two changes to the flotation systems: 1.75 or 0. As mentioned previously.5 mm and column flotation on the material finer than that. it remains to optimize and simplify the fines circuits. It lost favor. Coarse Coal Flotation Not New Although very few plants are currently using flotation for particle sizes up to 0.5 mm. there is very little misplacement of fines to the coarse stream. Increased amounts of fines which literally overloaded the carrying capacity of the cells.150 mm. This actually allowed some of the entrained ash to drain from the froth but it also drastically reduced the froth’s ability CHAPTER 1 INTRODUCTION Page 5 . they did understand that by using less frother and drying up the froth. If columns are to be applied for a coarser flotation size. the product ash would drop. however. Phillips cyclones. Increased fine ash resulted in higher flotation feed ashes.25 mm. coal. plus 0. 2. Although most operators did not realize that the entrainment was the main problem. By narrowing the size range feeding the spirals.

150 mm as well or better than current methods. To investigate how a column can fit into the overall plant circuit and should include not just the efficiency of cleaning but also the interrelationships of misplaced sizes. To investigate and recommend techniques for improving the flotation of coarse coal. It is the author’s opinion that coarser coal can be floated well in a properly designed system and that with the use of wash water in a flotation column. requires more understanding of the fundamentals involved. Phillips to carry coarse particles due to froth overload and due to a froth that was too dry and immobile. 3. After a few years of cells operating in this way. these coarse coal flotation improvements and circuit recommendations must be shown to be better than current practice. To be of benefit to industry. 2. this research has the following goals: 1.Dennis I. 1.4 Scope of Research To prove that improvements in coarse coal flotation can be utilized in ways such as those shown as Options 1 and 2 of the previous section. the problems of earlier years can be corrected. To show that flotation columns can process coal above 0. To understand the fundamentals of coarse particle flotation and detachment. This requires actual tests with a column on coarse coal. a reputation was developed that flotation cells could not really handle the coarser flotation sizes. To further the fundamental understanding of potential processing improvements. CHAPTER 1 INTRODUCTION Page 6 . 4.

.. Clarkson. Martin. Bureau of Mines RI 6239. Proc. E.. G.. Denver.O.. R.Dennis I.-H. Seventh Australian Coal Preparation Conference.W. Phillips.. F. and Palowitch. A. Australian Coal Preparation Society. Edward. 1963.I. B.. 1992. Tao. Performance Characteristics of Coal-Washing Equipment: Concentrating Tables. D. M. S-Rock Coal Beech Plant Sample. U. 1995. Personal Communication.S. Yoon..5 References Deurbrouck. 1994. Australia. The Impact of Fine Classification on Coal Preparation Performance.R. Mudgee. 1991. Co. O’Brien. Phillips 1.P. Firth.. SME Annual Mtg. WV. From efficiency test. Luttrell. D. D. H. Beckley. CHAPTER 1 INTRODUCTION Page 7 . Effects of Froth behavior on the Column Flotation of Fine Coal.. R. C.

Many factors such as hydrodynamics of bubble particle collision.150 mm (100 mesh) generally have the highest recovery. A rising bubble has streamlines of liquid flowing past it which exert a stress on the particle.Dennis I. From observation of a transparent laboratory flotation cell or column cell. As the bubble rises through the pulp-froth interface.1 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Introduction In recent years much has been learned about the fundamentals of bubble particle attachment (Jowett 1980). True measurement of all the attachment forces remains a difficult task. film thinning time. but the particle’s difficulty in remaining attached to the bubble throughout the froth zone and thus flow into the flotation cell launder.2 Current Detachment Models There are many dynamic circumstances under which a coarse particle can become detached. This is often not the case since coal particles finer than 0. attached to bubbles. the energy released from expanding and coalescing bubbles may break the bond of attachment. 2. and actual attachment forces are involved with the probability and forces of bubble particle attachment. What force is required to detach a particle from a bubble? What is the difference in magnitude for this force between fine particles and coarse particles? These questions will be investigated in this chapter. be possible to measure or in some way quantify the force required to remove a CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 8 . It should however. rising through the pulp zone and then becoming detached during or shortly after emergence through the interface between the pulp and the froth zone. one can observe coarse particles. From those studies it has been shown that coarse particles have higher probabilities of attachment to bubbles than do finer particles and thus should result in high flotation recovery. The problem with coarse particle flotation is not the ability to attach coarse particles to bubbles. especially near the agitator of a conventional flotation cell. A concentration of particles can be observed near the interface and falling downward into the pulp zone. Turbulence within the flotation system can cause liquid flows that may literally rip the particle from the bubble. Phillips CHAPTER 2 2.

Yoon and Mao (1996) showed that at the level of basic forces.Dennis I. Phillips coarse particle from a bubble surface. one can begin to understand the parameters that cause or prevent detachment. one can see the relative importance of each contributing factor. the probability of detachment (Pd) can be shown as W a E1   P d exp  . By looking at the fundamental forces involved in the detachment of a particle from a bubble or other gas interface. By utilizing models of these detachment mechanisms.

Xu and Yoon 1992). As long as the particle remained attached. The heavier particles produced angles closer to the maximum angle and thus were closer to the point of detachment. The static contact angle was measured. the following equations allowed for the verification of test measurements. By equating upward and downward forces at static equilibrium. The exponential function was used because the multiple small particles on the bubble surface would have distributed E’k values. 2-2 From the Laplace equation for pressure difference between the inside and outside of the bubble. this angle was always less than the maximum contact angle. Morris (1950) was one of the first to attempt to actually measure the particle detachment force on a visual scale. one can substitute and rearrange equation Eq. 2-1 in which a particle can be detached from a bubble if the kinetic energy (E’k ) pulling on the particle is greater than the sum of the work of adhesion (Wa) and the energy barrier (E1) (Laskowski 1989) (Laskowski. P d 2 w d  Sin  ( h R l )  4 L d 2 4 Eq. Small paraffin coated cylinders were attached to large bubbles in water. E   k Eq. 2-2 to yield: Sin  d 4 1 1  R2  R1  w  d   .

(r l)  Ld 4 Eq. 2-3 Where: h = distance from surface of liquid to midpoint of bubble R = radius of bubble l = length of rod d = diameter of rod CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 9 .

and contact angle on the spheres. the surface tension force vector. interface deflection.e. Smaller bubble sizes provide a higher internal gas pressure and thus a greater detachment force. dynes cm-1)  = static contact angle  = surface tension (dynes cm-1) L = density of liquid (grams cm-3) p = density of particle (grams cm-3) Morris determined that internal gas pressure of the bubble and thus bubble size was a critical factor in detachment. Phillips w = weight of rod P = internal gas pressure of bubble (force area-1 . 2-2. provides for a complicated solution of any equation.g. In a mathematical model. Huh and Mason (1974) worked with mathematical models of asymmetric spheroids solving complicated equations to determine the effect of rotation. He did not. The models ignored internal gas pressure of the bubble but did show that the largest particle that can remain attached to an interface is very dependent on contact angle. The use of spheres in modeling requires many assumptions and as shown in Figure 2-1. this change in upward force changes the location of the interface on the particle which again changes the surface force vector direction and also the perimeter distance of the interface. These complicated interrelationships have required the use of extensive equations and heavy computing power for solutions that also involve several assumptions. also changes direction which in turn changes the magnitude of the upward force component. CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 10 . actually measure any detachment forces but only the change in contact angle with changes in particle weight. The solid-liquid-gas interface will move up the edge of the particle until equilibrium is reached. As the interface moves.Dennis I. . Nearly all other researchers in the area of particle detachment have made use of spherical particles in the tests and models. however. as suggested by Eq. Morris verified that the internal gas pressure was greater than the static pressure outside the bubble.

the sum of all forces is zero. one article was discovered describing some previous work involving centrifuge tests. Nutt (1960) used spherical particles (glass beads) coated with silicone oil placed at the water surface in a tube and rotated in a centrifuge at increasing speeds until the particles detached. ro r     Zo Spherical particle at interface. Just before detachment of the particle of radius r. His model also involved complicated interrelated geometry for which he claimed an exact analytical solution is not feasible. Fc < Fs + Fb. After completion of this author’s work using centrifuges for particle detachment measurements and after extensive literature research. For detachment to take place at the semi-static interface.Dennis I. 4 3 r  P a 2r cos  cos( . Phillips  H h Figure 2-1. a. the centrifugal force must be just greater than the sum of the surface tension force and the buoyant force. and at the critical centrifugal acceleration.

 .

 ) .

 L a 3 1 1 4 3  3   2 2 2 2 2 2 /  L a  r cos  1 .

sin 



  3 r

3 r 1 .

2-4 It was left to Schulze (1977) to provide further analytical solutions. showed that the maximum limit of flotation is given by: CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 11 . sin   2 sin   Eq. Scheludko et al. When considering the particle mass as the only detachment force. (1976).

to hCrit calculate the energy of detachment. is given by: d max. g Eq. the largest floatable particle diameter. that the contact angle. 2-6 This equation is for spherical particles and ignores the internal gas pressure of the bubble since it assumes that RP<<RB. g 4 sin  ( P . By solving Eq. heq F 2 3 r  P 3   2 P L    L  1  cos 3   3 3h sin 2   2 2 sin  sin(   ) 2RP a RP   Eq.Dennis I. In a separate equation they also derived the maximum floatable particle size when the removal is determined not by the weight of the particle but by the kinetic energy of its impact with the bubble. which is not the case for coarse particles.  . is less than 40o. Jowett (1980) simplified some of Morris’ analysis. Schulze (1977) was also able. To access the significance of high turbulence zones in impeller flotation cells. That equation is not considered here since this work concerns particle detachment after complete attachment to the bubble. 2-5 where it is estimated that  = /2. Ignoring the capillary pressure and the residual hydrostatic term. E ab  ( F )dh . 2-6. d. Phillips 3 sin  / 2 2 g rmax. for the first time. and that the particle is static.

 L )g Eq. 2-7 where g = u2/R. He had only limited success in verifying this model with actual plant data. Based on CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 12 . u = relative bubble velocity and R = radius of gyration of an eddy. Schulze (1984) worked extensively on defining the shape of the meniscus of the liquid at the three phase interface using spherical particles. He did make the point that actual plant data confirmed the very rapid drop in recovery which is typically observed in flotation recovery above some critical particle size. This very rapid drop is predicted by models that allow for variations in detachment due to current eddies caused by variations in impeller turbulence. He also explored the mathematics of many minute points of bubbles and aggregate particle motion. since some plants were floating copper at much larger sizes than others with little explanation of the difference.

Phillips Figure 2-1.G 2 . and ignoring the hydrostatic and gas pressures: Dmax.Dennis I. he developed an expression to approximate the maximum floatable particle size by making some approximations to solve for the meniscus height Zo.

100 mm are not at all spherical.3 A Simplified Detachment Model for Cuboidal Particles As mentioned previously. Water is the flotation liquid in nearly all coal and minerals processing plants. This angle . This is especially true of coal which has particles that are either cubical or rectangular overall (cuboid) and have edges varying from sharp to rounded. When a liquid and gas interface is in contact with a solid. nearly all detachment models have involved the use of spherical particles and the associated complications. The major force to be overcome if a particle is to be detached from an interface or a bubble is the capillary or surface tension force (and the more fundamental quantity. 3  sin  sin(  ) 2 g Eq. the surface of the liquid bends at some angle  relative to the solid. since other liquids will have different contact angles with the solid. surface free energy). In actual plant practice nearly all particles > 0. the greater it resists being wetted by the liquid and thus greater is the contact angle. 2. CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 13 . Increasing the contact angle of the solid to improve flotation is the principle on which most minerals flotation is based. is a function of the hydrophobicity of the solid. The more hydrophobic is the solid. Through this work it was concluded again that the contact angle and particle size are the major determinants for particle detachment. 2-8 Crawford and Ralston (1988) provided a partial review of work by previous investigators and performed numerous contact angle measurements. For purposes of this work it will be assumed that the liquid is water.

CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 14 . The surface tension is a contraction or tension force at the liquid-vapor surface. an upward component of the force would require a contact angle () greater than 90o. the meniscus must bend upward from the particle as in Figure 2-4.  lv Figure 2-3. Consider cuboid particles located at the surface of the liquid much as would be found with a very large bubble. this cannot be the case since the direction of the surface tension force would actually be downward rather that upward. similar to that of liquid in a graduated cylinder. The surface tension force vector lv (Figure 2-2). If the liquid is against the vertical face of the solid. Phillips lv Liquid  Vapor Solid Figure 2-2.Dennis I. Common misconception of three-phase contact with cubical particle. To have an upward component of the surface tension force. Contact angles above ~70o are unknown in coal and mineral flotation. However. One would expect the liquid to form a meniscus against the side of the particle as in Figure 2-3. to the solid. Liquid drop on hydrophobic solid. acts at the three-phase contact point and in a direction away from the solid and at angle .

but are not limited to. In an actual flotation cell these other forces include. 2-9 If the above conditions are met. the sum of the buoyancy and the vertical portion of surface tension forces must be greater than the gravitational force. At static condition with a particle suspended at the surface of a liquid. additional forces must be applied for a particle to become detached. Three-phase contact point on top of cuboid particle. Buoyancy Surface tension (vert )  gravity force Eq.  Figure 2-5.Dennis I. For the particle to remain attached to the surface. the turbulence from the impeller of the cell agitator. the upward forces involved are buoyancy and the vertical component of the surface tension force. Close inspection of particles suspended at the surface reveal that the three-phase contact is indeed made at the top edges of particles. With the surface tension force properly positioned one can consider all relevant forces to derive an expression for detachment force. The downward force is gravity. eddy currents from the bubble moving CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 15 . The only way remaining for the surface tension to support the particle for  < 90o is with the three-phase contact point on the top of the cuboid as in Figure 2-5. Phillips  Figure 2-4. A three-phase contact with upward surface tension force.

To develop an expression for the point at which detachment occurs requires setting the sum of forces equal to zero. One method of measuring the detachment force on an individual particle basis is to place the particle at the surface of a liquid in a glass tube which is then rotated in a centrifuge. Phillips through the liquid. These detachment forces are difficult to measure in plant practice and it is therefore very difficult to quantify the exact amount of force required to remove a particle from a bubble. Increasing the speed of rotation will eventually dislodge the particle. and the energy release when bubbles break through the pulp/froth interface and coalesce.Dennis I. The increased speed causes an increase in the gravitational force until that force is greater than the sum of the buoyancy and surface tension. F = Buoyancy + Capillary surface tension force – Gravitational force = 0 or F Vl  l g perimeter  sin  .

2-10 where perimeter is simply 2(length + width) and Vl = volume of liquid displaced by particle l = density of liquid (grams cm-3) g = gravitational force  = surface tension of liquid (dynes cm-1) Vp = volume of particle p = particle density (grams cm-3) Since the particle is nearly 100% immersed in the liquid then Vl = Vp and V p g  l . V p  p g 0 Eq.

 p  2l w sin  0 Eq. 2-11 The force required to detach a particle from the interface can be expressed in terms of g and the general model becomes: g .

2l w sin  V p  l .

 p 

CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT

Eq. 2-12

Page 16

Dennis I. Phillips

A model often has more relevance if g is expressed in multiples of earth’s
gravitational force G, where G = g / 981 cm/sec/sec.

2.4

Experimental

In order to validate the model in Eq. 2-12, a set of experiments was undertaken
involving a laboratory centrifuge capable of holding what are essentially large test tubes.

Sample Preparation and Procedure
A high grade coal from the Pocahontas #3 seam in West Virginia was selected
because of it high contact angle and its very slow oxidation rate, which would ensure
consistent data over a several week period. The coal was collected from the clean coal
stream of an operating preparation plant and was later screened into four sizes covering the
range of what is considered coarse coal for flotation.
16 x 20 mesh (1.18 x 0.850 mm)
20 x 30 mesh (0.850 x 0.600 mm)
30 x 40 mesh (0.600 x 0.425 mm)
40 x 50 mesh (0.425 x 0.300 mm)
Initially ten particles were placed in the tube for a given size. Particles were eased
onto the surface of the distilled water. If a particle broke through the surface and fell to the
bottom, it was re-floated. The tubes were placed into the centrifuge and rotated at slowly
increasing speeds in increments of approximately 500 rpm. After each increase in speed the
centrifuge was stopped using an electric brake which eased the centrifuge to a stop with
minimal disturbance to the particles. The number of particles remaining at the surface were
counted and the centrifuge was then restarted to the next higher speed.

CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT

Page 17

Dennis I. Phillips

Tachometer
Probe

Figure 2-6.

Centrifuge test device for particle detachment measurements.

The values for percent remaining at the surface were then plotted versus the G-force
at that speed. G-force was calculated from the speed according to the following formula:

2 

rpm  r
G

91.19 g

Eq. 2-13

where the radius of gyration r = 7.6 cm for the combination of centrifuge and water level in
the tubes (10 ml of distilled water was used for all tests). The value of g was 981 cm/sec2.
Even though the particles were screened over a narrow size range there remained various size
and shapes of particles within a given size range. For this reason the point at which 50% of
the particles remained at the surface was chosen as the G-force to represent the force required
for detachment of that size and condition.

CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT

Page 18

the coarser particle sizes require less force to detach than do the finer particles (Figure 2-8). while one or two would not detach until a very high speed. This linear relationship to surface tension  is predicted from the general detachment force model in Eq. 2-12. Multiple Particle Results Although the ten particles used in each tube were from the same size range. Particles in tube during rotation. As expected. The drop in the detachment force required of approximately 65% from no frother to a high frother dose corresponds to the drop in surface tension of approximately 69% for a drop of from 72 to 50 dynes/cm. they were still of slightly varying sizes and one or two particles would usually drop at a very low speed. Phillips Figure 2-7. The point at which 50% of the particles remained on the surface was thus chosen to represent the average G-force required to detach the particles.Dennis I. CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 19 . The addition of frother to lower the surface tension had a small effect on the force required for detachment as seen in Figure 2-9.

50% detachment point for various frother dosages. Size by size detachment force. 10 particles each.Dennis I. no frother. 800 30 x 40 mesh 700 16 x 20 mesh G-Force 600 500 400 300 200 100 0 0 20 40 60 80 100 Frother (ppm) Figure 2-9. Phillips 100 90 % Remaining on Surface 16x20 M esh 80 20x30 M esh 70 30x40 M esh 60 40x50 M esh 50 40 30 20 10 0 0 4000 8000 12000 G-Force Figure 2-8. CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 20 .

and thickness) and mass was the single attribute chosen to characterize the particle in the 2-D plot. As expressed in the simplified model of Eq. Nutt also found that generally the aggregate of ten particles was detached from the surface as a whole and thus even though the data presented in Figure 2-8 and Figure 2-9 appear as expected on a trending basis. Single Particle Results Selecting typical size single particles from the previously mentioned size classes allowed for more accurate measurements and modeling of the detachment force. This is the most stable position from a buoyancy perspective and once in this position. Phillips As discussed above. the results from tests using ten particles were instructive but they also illustrated the problem with having multiple particles together on the surface. Particle mass was chosen as the single number to represent each particle size which was then plotted versus the G-force in Figure 2-11. width. The maximum surface tension force is then available to resist any reorientation of the particle. By using the actual dimensions of each particle.Dennis I. Several particles of coal were sized. one dimension of each particle is much less than the other two dimensions and thus the particles tend to orient themselves in a horizontal position as they are placed on the surface. A trendline is shown in lieu of a plot of a specific equation since the detachment force is calculated from three characteristics unique to each particle (length. Nutt (1960) found as much as 50% difference in the detachment force for one particle compared to that of ten mono-sized spherical particles (Figure 2-10). Actual particle shape and dimensions affect the perimeter length of the particle at the three-phase contact. The plot shows excellent agreement between the measured and predicted detachment forces. placed at the liquid surface of the centrifuge tubes and then subjected to the centrifugal force at increasing speeds as before. the predicted detachment force was then calculated from the model and plotted for each particle alongside the measured force. Figure 2-8 shows the overlap in the results for the two coarsest sizes while Figure 2-9 displays some variability in results for repeated tests. the absolute values may be in question. This concern for accurate reproducibility led to using only one particle per tube in further testing. Single particles were viewed under a microscope attached to an image analysis system which provided more precise measurements of particle dimensions. 2-12. The actual force required to detach the particles in units of G-force was calculated from the speed at which detachment occurred. CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 21 . the maximum perimeter length is presented to the surface tension force. For coal and most minerals. more than just the particle volume must be known.

6 1. CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 22 . This is seen in Figure 2-12.6 0. Phillips 1.8 1.4 N^2R 1. The detachment force of the sphere was calculated using Eq.2 0 0 1 2 3 4 5 6 7 Number of particles 8 9 10 11 Figure 2-10. (1976). where the flatter particles require a much higher detachment force than the equivalent mass of particle in either a sphere or cube.8 0. N2R (rpm2 x radius of motion). as a function of the size of the aggregate. 2-5 from Scheludko et al. Since the surface tension force functions at the three-phase interface.2 1 0.4 0. the perimeter length relative to total particle mass has a very strong effect on particle detachment. from Nutt (1960). Adhesion strength of aggregates of spheres.Dennis I.

500 2. 5000 G -F o rc e fo r D e ta c h m S phere 1 x 1 x 1 C ube 4000 1.500 1. particle size for various shapes and of equal mass.  = 72. CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 23 . Phillips G-Force for Detachment 1400 Measured 1200 Predicted 1000 Predicted Trendline 800 600 400 200 0 0 0.000 1.500 P a rtic le S iz e (m m ) Figure 2-12.000 0.Dennis I.444 3000 2 x 1 x 1/2 3 x 1 x 1/3 2000 1000 0 0. ( = 65o.000 2. Measured and predicted detachment force for each particle weight. Predicted detachment force vs.5 x 1.5 x 0.0015 0.0005 0.  = 1.002 Particle Mass (gr) Figure 2-11.35).001 0.

Given that detachment is the main reason for poor coarse coal flotation. Phillips Changing the contact angle or the surface tension has a significant effect on the detachment force required for particle removal. b) contact angle and c) surface tension.1 to 1 mm is usually rectangular and somewhat flat shaped. a) Particle Shape Effect b) Contact Angle Effect c) Surface Tension Effect 4000 G-Force 3000 Cube 1 x 1 x 1 65 Degrees 1. When moving from coarse to fine. Effect on detachment force of changes in a) particle shape.000 Particle Size (m m ) 0. Contact angle can be improved and is usually done so for coal by the addition of a hydrocarbon such as kerosene or diesel. Thus any increase in surface tension would be counter-productive.500 Particle Size (m m ) 20.500 1 . from 28 to 42 degrees by the addition of 2500 g/tonne. The 0. this is good news for coarse coal flotation.5x1.Dennis I. A 2 x 1 x ½ shape used for b) and c). but only under excellent conditions and then inconsistently.444 55 Degrees 2x1x1/ 2 45 Degrees 1x3x1/ 3  = 1. The 0. The physics of bubble size generation require the addition of frother to produce the low surface tension which maintains the small bubbles necessary for high flotation capacity.000 0. Previous work at Virginia Tech has shown that kerosene can increase the contact angle on partially oxidized Pittsburgh #8 seam coal.500 1 .000 0.150 x 0. Since coal in the size range of 0. the model indicates that to improve the flotation of coarse coal requires either an increase in surface tension and/or an increase in contact angle.000 1 .5 mm material has been known to float in many locations but is still believed by many to be too difficult to undertake commercially.35  = 72 72 Dynes 65 Dynes 55 Dynes 2000  = 1. but from the CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 24 .000 1 . This follows the results of plant tests discussed later in this document and also matches some of what is known from actual plant experience. but not nearly as much as the effect of flat and elongated particles (Figure 2-13).35  = 65 0 0.35  = 72  = 65 1000  = 1.500 1 .5 x 1 mm coal has been known to float.5x. The shape and location of the sharp bend of the curves is very important in Figure 2-12 and is confirmed in the plots of Figure 2-13.500 Particle Size (mm) Figure 2-13. below 1 mm the detachment force required begins to increase significantly and below 0. The plant test data will show some results of increased diesel. The plant testing chapter will discuss further many of the reasons for the apparent contradictions.000 1 .5 mm it begins a rapid increase.500 2 0.

With the simplified model presented (repeated below for convenience.5 mm will be difficult to float under normal conditions. 2. Phillips rapid change in the slope of these plots it is obvious that particle sizes close to 0.Dennis I. The approach taken here is to consider the particle already attached to the bubble and then to develop a better understanding of the force actually required to remove it from the bubble as well as the factors involved.5 Summary and Conclusions In the past. and liquid surface tension have on particle detachment. since that size represents the vast majority of minerals and all coal that is subjected to flotation. contact angle. g . mass.) one can easily see the direct effect that particle shape. most theoretical efforts have been spent on quantifying and solving equations concerning spherical particles at the air-water interface. Rectangular or cuboid type particles were utilized in both testing and modeling.

2l w sin  V p  l .

5 mm for most typical coal conditions. the model indicates that the only methods remaining to increase the particle detachment force require either: (i) increasing surface tension . The model predicts a sharp drop in the force required for detachment at the coarser particle sizes. should float well since they are flat and somewhat rectangular. at or above 0. This is the size area at which detachment is much more likely.  p  General Model The particle perimeter portion of the model points out a surprising fact of detachment that was never mentioned by researchers considering spheres: a much stronger force is required to detach a flat particle than is required to detach either a sphere or cube of comparable mass. This area also agrees with most plant practice when serious efforts are made to float the coarser particles.5 mm. Since coal particle shape is difficult to control without increasing the amount of ultrafines. which is achievable in a small amount. particle size tends to be around 0. The corner of the “L” shaped curve resulting from plotting detachment force vs. This means that coarse coal particles. CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 25 . which is not practical since flotation recovery generally benefits from small bubbles and a reduced surface tension is needed to maintain the small bubbles or (ii) increasing contact angle .

2815-28.G.. Physico-chemical Elementary Processes in Flotation. 1992. pp.Dennis I.V. Nutt. Journal Chemical Society. 613-626.. Jowett. Schulze. J. S. Journal of Colloid and Interface Science. 95. S.. Yoon. Xu..J. Journal of Colloid and Interface Science. 1989. R.. R. 2.M. Journal of Mineral Processing.-H. pp. S. Int. Huh. J. 187.133-141. and Mason. 348 pp. pp. AIME. 1988. Int. Morris. 5. A. 1. Measurement of Equilibrium Forces Between an Air Bubble and an Attached Solid in Water. The Influence of Particle Size and Contact Angle in Mineral Flotation. 740-754. Chemical Engineering Science. 23. Laskowski. and Yoon. Proceedings of the International Symposium on Fine Particles Processing. 25.. A. and Bojadjiev. Transactions AIME. A Theory on the Upper Particle Size of Floatability. 1-24. 271-289. Scheludko. NV. Attachment of Particles to a Liquid Surface (Capillary Theory of Flotation). and Ralston.. Elsevier. C. 47.. and Mao.. H. D. 1974. Faraday Transactions I.6 References Crawford. 91-95.W. Chap.-H. 1976. Toshev.J.241-259. 37 Formation and Disruption of Particle-Bubble Aggregates in Flotation.. T. B. Mineral Processing Extractive Metallurgy Rev... Laskowski. 1980. 1960... 4.Z. pp. Phillips 2. 1984. 12. pp. 12.. L. pp. J. pp.T. Las Vegas. Journal of Mineral Processing. The Flotation of Axisymmetric Particles at Horizontal Liquid Interfaces. R. Froth Flotaton: The Adhesion of Solid Particles to Flat Interfaces and Bubbles. 181. Application of Extended DLVO Theory IV. H. Schulze.. 1950. LES Techniques. CHAPTER 2 FUNDAMENTALS OF COARSE PARTICLE DETACHMENT Page 26 . 1977. 1996. pp. C...

Dennis I. Phillips

CHAPTER 3

3.1

Plant Test and Scale-Up

Introduction

The greatest inefficiencies in coal preparation are in the minus 1 mm particle size
range. As discussed in Chapter 1, there are considerable problems with the actual cleaning of
the material in this size range as well as problems due to misplaced sizes being sent to a
circuit that cannot treat that size. It is common for plants to either discard or treat the minus
0.150 mm material in conventional flotation while the plus 0.150 mm material goes to
spirals, water-only cyclones or heavy media cyclones. It is the intent of this work to evaluate
the potential for increasing the size range that can be treated in flotation. With today’s high
ash flotation feeds, conventional flotation cells cannot be used to float coarse coal due to the
high amount of entrained ash that comes with the froth product when controlling the cell for
high recovery. Today, the best candidate to be evaluated for use in the flotation of coarse
coal is column flotation with wash water.
When properly operated, columns tend to have better recovery due to the high volume
of consistent sized bubbles which provide better capture of coarse and fine particles and thus
increase the flotation rate. Probably the greatest advantage to flotation columns for coarse
coal is in their ability to handle high amounts of high ash slimes (clays, etc) through the use
of froth washing. This use of wash water not only reduces and stabilizes the product ash
content but it allows the coarse coal to be fed to the column without any prior desliming or
fine size cuts. This means that the intended top size feeding the column will be the smallest
size cut made in the preparation plant. The coarser the size cut, the less material that is
misplaced into the cleaning device for the next coarsest size. There is a tremendous
difference in the misplacement of fines to coarse when the size cut is changed from 0.150
mm (100 mesh) to 0.5 mm.
A small laboratory size column (2-inch), is the simplest method to evaluate column
flotation for a given coal. Since much of the loss of coarse coal is believed to be due to
dropback in the froth zone (Finch and Dobby, 1990), it is important that any scale-up testing
be performed in a column with sufficient froth travel distance to provide for potential loss of
coarse coal. Wall effects in small laboratory columns may actually stabilize the froth zone
and reduce dropback compared to that of full-scale columns. When the quantity of coarse
coal feed that would be required for proper testing was considered, it became apparent that a
lab or pilot plant would be insufficient for testing. Although there are many plants with
operating flotation cells, nearly all have a feed that is too fine in size consist for the testing

CHAPTER 3 PLANT TEST AND SCALE-UP

Page 27

Dennis I. Phillips

considered. At the time of this project there were only two plants operating flotation
columns on coal and they were also utilizing a very fine coal feed. Fortunately, the Center
for Coal and Mineral Processing at Virginia Tech was involved in a project with Amax
Research and Development that was being funded by the U. S. Department of Energy (DOE)
to investigate the production of premium coal fuels (DOE Contract No. DE-AC22-92PC92208). As will be detailed later, a plant site became available in which a sufficiently
large test column, 750 mm (~30-inch) diameter, could be tested on a fine and coarse feed up
to 1 mm in size.
This testing provided an opportunity to study the real potential for floating coarser
sizes than commonly tried. Size-by-size analysis is provided along with the effects of several
parameter changes. Test work also provided scale-up information to be used in designing the
world’s largest full-scale columns then in use on coarse or fine coal.

3.2

Project Background

A small part of the large Premium Fuels project was Task 3 Development of Near-Term
Applications, which was intended to specifically address the use of advanced flotation and
selective agglomeration processes for recovering coal lost in existing coal preparation plants.
The overall goal of Task 3 was to produce a clean coal product which could be sold in
existing markets by one or both of the following strategies: 

Increase the percentage recovery of marketable coal from the ROM coal. 

Improve the quality and value of the marketable coal (heating value, sulfur or ash
content, and handling characteristics) in a cost-effective manner.

Under the above task was Subtask 3.2 Engineering Development, which had a primary
objective of pilot-scale testing and engineering development of the selected applications.
Previous laboratory testing had shown that from an economic and technical point of view,
column flotation was a much more viable process and was thus chosen as the most likely
candidate for meeting the near-term application criteria.

Plant Selection
The Lady Dunn Preparation Plant, near Montgomery, West Virginia, was volunteered
by the Cyprus Amax Company and provided an excellent test site for proving the advanced
flotation column. The flotation feed is typically around 40% ash and has a high percentage
of minus 325 mesh material in the flotation feed. Also, the plant had existing mechanical
conventional flotation cells and therefore results could be directly compared to existing

CHAPTER 3 PLANT TEST AND SCALE-UP

Page 28

Dennis I. Phillips

conventional technology. At the time of the pilot testing, the plant was mid-way through an
expansion program. The plant flowsheet consisted of heavy-media vessel for coarse coal (+6
mm or + ¼ inch), heavy-media cyclones and Deister tables shared the load for the 6mm x
0.150 mm (¼ inch x 100 mesh), and conventional flotation on the minus 0.150 mm overflow
from desliming cyclones (Figure 3-1).
Raw
Feed

Feed to
Column, 100
series

Screen
+ 1/4”

H.M .V.
0.150mm x 0

- 1/4”

Sieve

Deister Tables

Classifying
Cyclones

-1 mm Feed to
Column, 400
series

Conventional
Flotation

H.M .C.
Screen Bowl
Centrifuge

Clean
Coal
-.250 mm used for
200 & 300 series

Figure 3-1.

Refuse

Lady Dunn Plant flowsheet during 30-inch column testing showing
the two major feed sources.

Phase I of the expansion program involved the addition of the heavy media cyclone
circuit while the design of Phase II was not completed but would include a major change to
the flotation circuit. The author considered this an excellent opportunity to prove the benefits
of increasing the size of feed to flotation by utilizing column flotation and proposed that the
pilot-scale test plan be modified to include not only testing of existing flotation feed (-0.150
mm) but to also test up to 1 mm coal. The plan was accepted and the coal company was
receptive to the possibility of including changes in the Phase II design.

CHAPTER 3 PLANT TEST AND SCALE-UP

Page 29

emphasis on coarse coal recovery. which are usually preferred due to the simplicity of installation. Determine the optimum size feed for a column at this plant. Normal in-plant testing could have possibly involved an 8. With the normal variations within a plant process it is sometimes difficult to maintain a proper feed and this was especially true at the Lady Dunn Plant. i.. Phillips 3. 30-Inch Column Chosen Past scale-up of the MicrocelTM column to full-size units has proven successful even from laboratory size units. 1995.or 12inch diameter column. CHAPTER 3 PLANT TEST AND SCALE-UP Page 30 . The critical scale-up parameters that would be required for optimum recovery at a reasonable product ash were investigated. Comparison of the flotation column to the existing conventional cells on current flotation feed (minus 100 mesh).e. Test Plan and Installation To meet the stated objectives required in-plant testing with a column of sufficient size to provide reasonable scale-up information and utilize equipment similar to that used in industry. Performing process testing in an operating plant required that considerable attention be paid to feed source and control. Testing of the 30-inch column at this plant could also provide a general research element by furthering the understanding of the relationship of bubble capacity and collector dosage to recovery of the coarser size fractions of coal. Recovery of various size fractions and product quality (ash) were the major results to be measured. aeration rate. Coarser material and a very stable froth concentrate were added difficulties that made equipment selection and proper test circuit layout very important. frother and collector dosage were varied to determine their effects on the recovery of various coal sizes. Parameters such as feed rate.Dennis I.3 Experimental Objectives Of Plant Testing Testing of a 30-inch diameter MicrocelTM flotation column at the Lady Dunn preparation plant began in June. The objectives were:    Determine the benefit of applying the advanced flotation processes to this operating plant.

150 mm fraction that was mostly clean coal from the HMC circuit. The scale-up indicated that the total height for the 30-inch column would need to be no more than 21 feet. There was also concern that the feed pumps. One was a raw minus 0. the underflow of which was considered minus 60 mesh and was sent to flotation (Figure 3-1). The raw 0.150 mm (-100 mesh) cyclone overflow and the other was a minus 0. Prior to an expansion two years earlier. and 4) a totally raw coal sieve screen undersize stream with some material too coarse for flotation (Figure 3-1). Phillips However. a heavy media cyclone circuit was added to clean the Deister table feed. The concern was that a small column has such a short froth travel distance that the coarse coal would not drop back into the froth zone in equal proportions to that of a larger column. To develop preliminary feed and control parameter information for the 30-inch column.150 mm x 0 coal from the classifying cyclone overflow was the first material tested in the 30-inch column. which suited the height limitations within the plant. Four separate feed streams were examined before a final slurry was selected as the feed for parametric testing. the tables remained in service as well as the HMC circuit. 2) a coarser combined stream of cyclone overflow and partially cleaned sieve screen undersize. pipes and valves associated with a smaller column may have plugging problems due to the misplaced coarser feed sizes found in an operating plant. Deister tables on 1/4 inch x 100 mesh.250 mm (-60 mesh) screen underflow containing a 0. The four streams were 1) a fine coal stream of minus -0.250 x 0. the largest Microcel test column available was chosen. two existing flotation feed streams were available. 3) a similar stream containing an increased amount of the coarser screen undersize. To accommodate these concerns. Test Column Feed Sources At the time of testing. Around 1993. The clean fine coal from the HMC traveled across a large sieve. This valve was adjusted to provide sufficient flow to the column feed sump on the floor CHAPTER 3 PLANT TEST AND SCALE-UP Page 31 . a 30-inch diameter column was loaned from the Virginia Tech pilot plant.15 mm (100 mesh) cyclone overflow. and conventional flotation on the minus 100 mesh material. with increasing demands on the plant for increased production. Scale-up predictions were made from this preliminary test data and a flotation rate was developed. there was concern that a small diameter column would not properly simulate a large column in coarse coal recovery.Dennis I. Thus. the Lady Dunn Plant flowsheet was in a stage of transition. But. To provide a reasonable froth travel distance and to allow for more froth zone dropback. the plant consisted of heavy media vessels for coarse coal (+1/4 inch). the 30-inch column having been previously constructed and tested under an earlier DOE project. A sample thief with a valve was installed in the existing flotation cell distribution box which was fed from the classifying cyclone overflow. which allowed for a higher plant throughput. a drum of flotation feed was collected from the existing plant and tested in a 2-inch diameter laboratory column at Virginia Tech.

While the results of testing the 0. by noting the time required to fill a fixed volume container. The pump and sump were located 2 floors below and over 50 feet horizontally from the column feed area. Pilot Column Circuit Description and Operation Although equivalent to a fully functional full-scale commercial unit. Testing of this raw feed was labeled as the 400 series (Appendix A). a full-stream sample cut could be taken and.5 to 1 tonne per hour (TPH) of clean coal for most coals. the feed for which was taken from the underflow of a temporary fixed sieve receiving raw Deister table feed (6 mm x 0).5-mm fraction was screened from the samples and accounted for separately. testing on a coarser feed was more desirable to all parties since the intent was to prove column flotation on a broader application basis. The plant provided a feed pump with a remotely variable speed controller which was adjusted from the control area (near the top of the column) to maintain or adjust a given volumetric feed flow. This provided a true raw coal feed containing natural minus 1 mm fines. By moving the flexible pipe.25 mm (60 mesh) flotation feed. A general layout of the column testing circuit is shown in Figure 3-2. Screen analyses of the products indicated that there was insufficient coarse material in the blend and the coarse feed valve box was modified to accept nearly all of the material in the screen undersize pipe. The major difference was the limited capacity of the test unit due to its 30-inch diameter compared to the 3-meter or larger in diameter of most commercial units. A valve and collection box were mounted on the side of the sieve underflow pipe which was part of the feed to the flotation cells. The column feed piping discharged into a small head box just prior to entering the column. Diesel fuel oil was added as a collector into the stream feeding the feed sump. The two tests performed on the latter feed were labeled as the 300 series (Appendix A). The test column has a capacity of 0. the fixed sieve was changed to one with a smaller opening (1 mm) and the underflow was fed to the column feed sump. The flow of this combined feed proved to be unstable and of insufficient volume for parametric testing. The column was fed by an 80-gallon feed sump to provide a consistent feed volume. Phillips below.150 mm x 0 material were excellent (see test series 101-110. Since the new column feed was coarser than necessary. After completion of the spiral testing. the plus 0. CHAPTER 3 PLANT TEST AND SCALE-UP Page 32 . The plant had recently installed a test spiral concentrator. the 30-inch test column was considered a pilot-scale column.Dennis I. a positively measured flow could also be taken. Appendix A. The feed sump was allowed to have a slight overflow to ensure a constant flow to the column. Several tests were performed on this coarser feed and are labeled as the 200 series (Appendix A). This coarse material was blended with the classifying cyclone overflow to simulate the raw coal size consist for a minus 0. The results from testing of the fine material are labeled as the 100 series.

Phillips Figure 3-2.Dennis I. Pilot column test circuit arrangement. The slurry recirculation pump was located on a lower floor at the bottom of the column. CHAPTER 3 PLANT TEST AND SCALE-UP Page 33 . small microbubbles were produced (see schematic in Figure 3-3). By recirculating the slurry through the spargers along with the addition of air and frother. A tailings valve was also located in this area of the column and discharged through a section of hose which was maneuvered to provide a full-stream cut for the tailings or refuse sample. The frother was injected into the suction line of this pump and the air was injected just prior to the spargers.

The air flow system was equipped with a pressure regulator to provide a constant pressure to the flowmeter. water. The froth product from the column launder discharged through a 6-inch diameter pipe to the existing flotation cell product launders. This lack of slope caused the column concentrate launder to back-up frequently and necessitated the use of launder water to help move the froth through the pipe. Phillips MICROCEL COLUMN FLOTATION 30 Inch Test Unit Microcel Air Spargers Wash Water Feed Column Concentrate Valv e Air Distribution Ring Slurry Distribution Ring Tailings Valv e Figure 3-3. An orifice plate flowmeter with differential pressure transmitter and digital readout was used to measure the air flow to the spargers. CHAPTER 3 PLANT TEST AND SCALE-UP Page 34 . and pulp level were controlled from the control area at the top of the column. For this reason many of the concentrate percent solids values are slightly lower than actual values since the sample was taken at the pipe discharge. Wash water was measured with a paddlewheel type flowmeter and displayed electronically. Air.Dennis I. Microbubble Generator Air Inlet Frother Inlet (Optional) Circulation Pump Flotation column schematic showing bubble generation system. This allowed for accurate flow measurements and air flow recording on a standard temperature and pressure basis. A minimal height difference between the column and existing cells provided very little slope for the froth concentrate pipe over a long distance.

Air fraction was calculated as: Air Fraction 1 . The signal from a similar transmitter placed at a lower level was also displayed at the control area allowing an air fraction (fraction of air by volume in a given section of pulp) to be calculated from the combination of the two level signals. Pulp level in the column was maintained with a PID loop controller which received a signal from an electronic level transmitter on the column and sent a proportional signal to the tailings discharge valve. Manual gate valves were used to adjust the air and wash water flows. Phillips A pressure regulator was also provided for the wash water but low water pressure required the installation of a small in-line booster pump.Dennis I.

lower level .

This was an excellent method for determining column conditions such as turbulence. A small sight glass near the top of the column provided a means to view the pulp/froth interface area. As the feed rate CHAPTER 3 PLANT TEST AND SCALE-UP Page 35 . but no air. The detailed results and test parameters are provided in order of series number in Appendix A. Normal test procedures for the laboratory column call for maximizing air and washwater input to the column without reaching an air overload or “flooding” condition. Those results are discussed separate from the two series of parametric tests (400 series) which represent the bulk of the testing data. The samples were sent to the laboratory at the end of each day’s testing. and 300. 3. approximate bubble size. or excessive air flow. upper level fixed distance Eq.4 Results and Discussion Preliminary information was gathered from 2-inch lab column tests and pilot unit test series 100. Several full-stream cuts were taken for each sample and were collected in 5-gallon (19 liter) containers with sealed lids. 200. Column tests are then run over a range of feed rates. Testing involved waiting 30 minutes after any change in operating parameters to allow all conditions to stabilize before sampling. in the column (Finch and Dobby. 3-1 where level was in vertical inches of slurry pressure and the fixed distance between level transmitters was measured by the transmitters with slurry. Preliminary Flotation Testing A 2-inch diameter column was utilized for the initial laboratory testing of the classifying cyclone overflow sample. 1990).

This generally produces an ash versus combustibles recovery curve as shown in Figure 3-4. ash for 2-inch Laboratory and 30-inch columns (Tests 101-110). CHAPTER 3 PLANT TEST AND SCALE-UP Page 36 . Both sets of results are plotted in Figure 3-4 and one can see that the same ash/recovery curve was produced in the two-inch lab column as in the 30-inch diameter pilot-scale column. The initial tests with the 30-inch column in the Lady Dunn Preparation Plant were on the same classifying cyclone overflow stream as tested in the 2-inch column and are labeled as the 100 series in Appendix A (test points 101-110). Before this testing was fully developed. it was decided to try for a coarser feed as discussed in a previous section. The purpose of the initial testing was to develop a general “ballpark” for the expected operating parameters. Combustible recovery vs. Phillips is increased the limit of the froth carrying capacity is eventually reached. From the lab results it was determined that a 30-inch column would have a capacity of no more that 100 gpm of feed slurry at the percent solids tested.Dennis I. resulting in lower ash products but also a lower recovery. however. Figure 3-5 shows that the efficiency of the two units is also comparable since the points lie along the same curve. 100 Feed 90 80 Combustible Recovery % 70 60 50 2 inch Lab Column 40 30 30 inch Pilot Column 20 10 0 0 5 10 15 20 25 30 35 40 Product Ash (% ) Figure 3-4. Results were excellent and compared well to the laboratory tests.

CHAPTER 3 PLANT TEST AND SCALE-UP Page 37 .Dennis I. Results from the two tests performed with this attempt are labeled tests 301 and 302 in Appendix A. Testing then began on the coarser feed (200 series) but again this was not fully developed since the stream did not contain as much coarse material as anticipated. Phillips 100 90 Combustible Recovery (% ) 80 70 60 50 40 2 inch Lab Column 30 inch Pilot Column 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100 Ash Rejection (% ) Figure 3-5. An attempt was made to send more coarse material to the column but it was discovered that the available pipe did not carry enough material to provide sufficiently coarse feed to the pilot column. The 200 series feed was predominantly classifying cyclone overflow and thus consisted of well liberated fines. Ash rejection and combustible recovery for 2-inch Laboratory and 30-inch column (Tests 101-110). The 200 Series Tables in Appendix A give the results obtained with the limited testing performed on this feed. Figure 3-6 shows that the performance was generally well below 10% product ash with a good recovery even though the column had not yet been optimized.

and feed rate. Box-Behnken designs are response surface designs. or high. 0. medium. frother and diesel dosage. Box-Behnken designs are available for 3 to 7 factors (or variables) and are formed by combining two-level CHAPTER 3 PLANT TEST AND SCALE-UP Page 38 . To minimize the number of tests while producing a response surface with statistical meaning. Once testing began on the coarse feed there was an urgency to quickly determine the scale-up information for a potential full-scale column installation. and low. but to produce all possible combinations of variables and ranges would be a tremendous task. requires an organized testing method. Tests that produce several combinations and magnitudes of variables are necessary. Performance of Microcel pilot column on coarsened cyclone overflow (tests 201-215). and +1. Phillips 100 90 80 70 60 50 40 30 20 10 0 0 5 10 15 20 25 30 35 40 Concentrate Ash (% ) Figure 3-6. a pipe was installed to the temporary sieve underflow which allowed a true natural sized raw coal stream (1mm x 0) for column feed. the Box-Behnken experimental design was chosen and a computer program Design Expert by Stat Ease was selected for the design and to interpret the data. made specifically to require only 3 levels of each variable. Test numbers 401 to 416 are the initial tests to determine the general range of parameters for operation with the coarser coal and were also later considered as a parametric test series (First Series). coded as -1. Parametric Testing To best determine the effect of various operating parameters such as air rate. At the conclusion of an unrelated test program in the plant.Combustible Recovery % Dennis I.

Changes in specific characteristics of the coal particles (i. when the main parameters (i. specifically the recovery of the coarser fractions of coal. and feed rate solids) were entered into the statistical analysis program as a Box-Behnken experimental design. degree of liberation and hydrophobicity) result in a different grade-recovery curve.Dennis I. Even though this series of tests was not a designed parametric test. Parametric Tests . A major advantage of Box-Behnken designs is in their nature of requiring only 3 levels.e.First Series The first set of results (tests 401 to 416) from the coarse raw coal feed provides the most consistent data set. whereas. The initial set of data for flotation of the raw coarse coal (tests 401 to 416) provided a much more consistent data set. but with only a fraction of the experiments needed for a full three-level factorial. such as low wash water pressure. Phillips factorial designs with incomplete block designs. Because there are only three levels. during the several days required to run the primary designed parametric test series (tests 451 to 465). Other problems.e. column conditions affect CHAPTER 3 PLANT TEST AND SCALE-UP Page 39 . It was very difficult in an operating plant to produce a consistent feed for a small stream such as that required for this parametric test. Although the test parameters covered a wide range of operating conditions. The three factors used in this project required only 15 experiments versus 32 for a full three-level factorial. as shown in the operating parameter list for the 401 series found in Appendix A. nearly all of the results for a given particle size fit along a single grade-recovery curve (Figure 3-7). The results produce a consistent data set and prediction model that is better than that produced with the main parametric design. After further review of the data it was determined that the variations in control parameters for the initial raw coarse coal testing fit a Box-Behnken experimental test design. The intent of this initial testing was to vary key operating variables from low to high to determine likely operating points as well as gain scale-up information. The problems encountered may explain some of the inconsistencies found when evaluating that data set. Most desirable statistical properties are created by the Box-Behnken. frother dosage. Changes in several of the key operating parameters were performed. The intent of the parametric testing was to determine the effect of various operating variables on the performance of the flotation column. collector dosage. Of the 15 tests. For example. For these reasons the results of two separate parametric test series (first and second series parametric tests) will be presented and discussed. the quadratic predictive model is most appropriate. were also experienced. a good correlation was found and several definite trends were realized. The Box-Behnken design provided a measure of the contribution of each parameter to the given response and also allowed the influence of joint interactions between the various test parameters to be estimated. two are repeats of the midpoint levels to determine statistical variance. there were considerable variations in feed solids to the column. and were meant to cover the range from low to medium to high for several of the key parameters..

25 mm x 0 0 5 10 15 20 25 30 35 40 Cumbustible Cumbustible Recovery Recovery % % Concentrate Ash (%) Feed 100 100 90 80 70 60 50 40 30 20 10 00 0.5 0. entrainment of non-floatable material into the froth was not a problem. Feed Cumbustible Recovery % 100 90 80 70 60 50 40 30 20 10 0 0.25 0.Dennis I.5 mm x 0 10 0 0 Figure 3-7.25 mm mm 0 55 10 15 20 25 25 30 30 35 35 40 40 Concentrate Ash (%) Feed 100 90 Cumbustible Recovery % 80 70 60 50 40 30 20 0. 5 10 15 20 25 Concentrate Ash (%) 30 35 40 Grade-Recovery plots by size for tests 401-416. The wash water flow was sufficient to remove the entrained high ash particles.5 xx 0. Conditions that result in a limited carrying capacity provide room on the bubble surface for only the most hydrophobic particles and result in a low ash but low recovery product. Phillips a result’s location on a grade-recovery curve for a given coal. CHAPTER 3 PLANT TEST AND SCALE-UP Page 40 . The close fit to a common graderecovery curve indicates that for most of the tests.

045 mm (100 x 325 mesh) size fraction and the 0. These sizes were chosen because one of the main emphases of this test program was to determine the applicability of column flotation to a coarser coal than commonly practiced. including the effect of fines.25 mm.5 x 0.25 mm x 0 size class was chosen since it is also a relatively coarse top-size for flotation and also includes the effects of fine coal in the feed.5 x 0. high recovery of the coarser coal particles can result in an excessive amount of ash in the finer particle size of the product. The 0.25 mm.25 mm size class was chosen since it was the coarsest size that showed the potential for reasonable recoveries on the Stockton seam coal. and 0. the test results and parameters were subjected to statistical analysis using the Design Expert package for the computations. The variations in recoveries from the 0. Figure 3-8 shows that some tests had a much lower recovery for each size class than other tests. A hump shaped curve is typical for coal. Phillips The three particle size classes scrutinized were 0.25 mm material alone can often be misleading unless the total effect.5 mm x 0 was chosen to show the overall results of floating coarse and fine particles together.045 mm x 0 recovery was always slightly below that.5 mm) relative to the finer sizes based on the various operation parameters as shown in Table 3-1.045 mm to 0. As can be seen from the size-by-size recoveries plotted in Figure 3-8. To determine the relationships of the operating parameters that may be causing variations in recovery. CHAPTER 3 PLANT TEST AND SCALE-UP Page 41 . it illustrates well the inconsistencies of coarse particle recovery (plus 0. the best recovery during each test was for the 0.Dennis I.5 mm x 0. as confirmed by the optimum recovery for the sizes from 0. The combination 0.150 x 0.5 x 0. is also considered. while some dropped off only at the coarser sizes. For example.25 mm x 0 (60 mesh x 0). The combined 0. 0. Although this is a “busy” plot.

and diesel dosage (grams per metric tonne.045 100 0.94 for all three size classes investigated.150 90 . CHAPTER 3 PLANT TEST AND SCALE-UP Page 42 .0 mm 80 1.0 x .70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Figure 3-8. Although several of the parameters were investigated.150 x . +1.5 mm 80 .25 x . The parametric model fits had R-squares of at least 0. Phillips 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 Particle Size in mm First series (401-416) combustible recovery by particle size. In most cases the quadratic model provided the best fit.25 mm 90 . only three variables were considered at one time. those with the most significant effect on combustibles recovery were feed rate (kg/min). frother dosage (ml/min).5 x . Complete statistical results are found in Appendix B which also develop the predictive models.045 mm x 0 100 Cumulative Combustible Recovery % Dennis I. To better grasp the effect of variables. g/T).

61 68 425 1. Rec.5 ? 10 30 520 758 414 80 28.08 Comb.69 14.83 64.97 10.79 38.6 30 518 696 104 28 404 80 42.66 42.39 30.29 24.60 68 425 1.22 50.43 1.60 17.0 14.51 10.71 13.2 1.95 12.6 17 11.80 16.68 44.86 8.41 68 425 1.2 12.40 9.6 ? 11 18 346 445 105 22 410 100 39.6 1.66 55.36 36.46 8.98 10.79 17.96 7.41 37.23 11.67 94.3 12.53 41.52 9.68 12.66 1.91 12.48 51.5 mm lpm lpm (cms) (ml/min) (ml/min) (g/T) (g/T) (gpm) (inch) 401 100 31.38 45.22 Feed % Solids Conc.70 87.99 44.02 43.34 68.70 8.41 54.54 60.7 17.9 1.64 11.6 30 721 961 110 28 403 100 43.44 13.19 91.84 48 544 2.73 7.00 5.30 37.6 ? 11 30 571 745 105 24 411 90 45.38 64 425 1.77 11. Test # Feed 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 Test # First Parametric Tests (401-416) results and operating parameters.63 16.51 21.58 58.47 9.48 8.6 17 10 30 487 707 104 28 408 80 44.82 28.6 ? 8 45 1282 2020 32 CHAPTER 3 PLANT TEST AND SCALE-UP 24 32 Page 43 .6 ? 11 30 405 530 115 26 412 103 42.64 11.10 38. 7.32 55 425 1.67 8.28 37.95 32.5 ? 8 30 831 1249 32 415 80 32.13 13.6 1.6 18 10 15 201 254 105 26 409 100 38.11 72.33 9.4 1.06 45.06 33.6 2.09 10.21 29.87 9.93 14. Sep Eff.76 47.5 30 690 920 110 402 100 29.40 15.84 59.43 37.78 55 425 1.69 34.0 1.6 13 24 48 539 753 104 30 407 60 27.57 52.9 14.79 68 465 1.61 1.3 1.92 40.6 30 423 552 104 24 405 80 22.84 51.47 79.6 20 11.0 1.05 1.62 11. Total Ash % Conc. % 84.93 8.81 32.82 50.2 13. Phillips Table 3-1.4 1.61 45.62 10.49 68 425 1.5 12.51 69.42 18.73 10.62 47.3 12.71 1. Feed Froth Rate Rate Rate Rate Rate Water Rate Rate Fraction (Diesel) -.88 6.38 7.73 92.00 12.65 57.7 13.88 5.07 12.14 42.9 13.84 9.00 37.98 44.80 9.76 38.Dennis I.18 87.21 41.40 41.36 84.63 68 408 1.59 71.% Ash Rej.5 12.29 65 425 1.3 15.38 93.61 1.32 55 425 1.7 10.65 39.71 61.8 12.48 71.75 48.9 1.04 10.74 82 425 1.05 Feed Feed Conc Conc Conc Wash Air Air Air Frother Collect Coll.62 12.65 58.50 86.00 5.62 5.14 90.73 6.02 7.00 68 425 1.37 50.11 17. Tails 34.7 1. Yield 5.2 13.0 17 10.91 39.73 Tails % Wt.6 13 11.99 92.55 52.03 8.7 15 11.70 12.91 34.96 71.85 45.0 2.65 46.73 1.% 60.23 2.1 1.24 57.22 13.07 51.95 15.6 ? 10 40 731 1007 103 Turb 413 80 45.98 91.10 43.18 58.30 66.6 ? 8 45 1091 1521 32 416 95 27.98 57.55 55 425 1.19 42.64 1.35 67 408 1.59 55.93 1.19 Comb.54 1.59 1.20 94.42 55.2 1.09 27.14 50.6 48 1307 1805 104 28 406 60 40.37 64.7 1.07 40.60 4.06 9.5 mm Sump Depth % (gpm) (kg/min) (kg/min) (Tph/m2) -.67 36.64 89.69 1.53 13.85 41.15 70. Coll.8 11.62 89.65 86.76 1.50 39.52 6.

Phillips First Series .Dennis I.

0.030 -0. An advantage of using the coded factor coefficients is that the values are directly comparable rather than being related to the units of measure as in the actual factor values of the above equation.45) for the feed rate coefficient indicates that recovery decreases with increasing feed rate.646E-05 +3. An R-Squared of 0.27 -6. one can see that the feed rate has the major effect on recovery.25 mm x 0 The predictive models in the statistical package help to determine the effect each variable has on the response. In other words.917E-03 = * Feed Rate * Frother * Diesel * Feed Rate2 * Frother2 * Diesel2 * Feed Rate * Frother * Feed Rate * Diesel * Frother * Diesel The coefficients shown in Table 3-2 are equivalent to those in the above equation except that they are based on the coded factors (i.53 -4. CHAPTER 3 PLANT TEST AND SCALE-UP Page 44 .e.48 -15. The size of the coded coefficients relate directly to the observed change in the response and thus from Table 3-2.47 -0. except that air rate can also play a role if varied significantly. which in this case is combustibles recovery.154E-07 +0. small values indicate significant coefficients in the model. On that basis the feed rate is the major determinant for recovery and the frother is a distant second within the range of actual factors tested.942 (indicating a relatively good fit) was calculated for the quadratic model whose equation in terms of actual factors is shown below: 0.25 mm x 0 Recovery +240. The Prob>|t| column gives the probability of getting a coefficient as large as that observed. the top feed rate is +1 and the lower is –1.61 -3. The negative sign (-12. This corresponds to previous flotation column experience. when the true coefficient equals zero.044 -0.

Dennis I.00 1550 38.00 3-D Response Plots.00 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 100 80 .25x0 Rec. Actual Constants: Frother = 8.0 60 40 20 0 2000 45. Frother at 8.33 1100 Diesel 31. and 12 ml/min.67 650 Feed Rate 200 Figure 3-9.67 650 Feed Rate 200 25. 10.00 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 100 80 .25 mm x 0. 25. Actual Constants: Frother = 10. CHAPTER 3 PLANT TEST AND SCALE-UP Page 45 . Phillips DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 100 80 . First Series (401-416).00 1550 38.33 1100 Diesel 31.67 650 Feed Rate 200 25.0 60 40 20 0 2000 45.0 60 40 20 0 2000 45.33 1100 Diesel 31.25x0 Rec.00 1550 38. 0.25x0 Rec. Actual Constants: Frother = 12.

24 11.4067 0. 0. This was as expected since the larger bubbles which occur at a low frother dosage have a limited surface area and are quickly overloaded. Coefficient Estimate 77. Factor Intercept A-Feed Rate B-Frother C-Diesel A2 B2 C2 AB AC BC Response surface model coefficients and significance. This is probably because the smaller bubble size and the increased collector dosage together increased the flotation rate constant and provided extra bubble carrying capacity.45 4.9677 0.4637 0. indicating sufficient bubble surface area to carry the range of coal particles available in the feed slurry.66 3.32 -4. At a high frother dose (12 ml/min) little change in recovery was noticed with changes in feed rate. Phillips Table 3-2.52 0.90 11. A medium frother dose indicated the same performance except that at the higher feed rate.78 -0.60 0. one can see that at a low frother dose (8 ml/min) the increased feed rate reduced the recovery.88 -0.09 -0.1655 0.50 10.59 1.042 0.97 5.70 -0. restricting their carrying capacity for coal.0130 0.49 1.25 mm x 0 coal.Dennis I.80 5.89 -0. except that some improvement in recovery was predicted for higher frother dosages.27 5.25 mm x 0. recovery improved over that with the lower frother dose.25 mm x 0 size model as illustrated in 3D plots (First Series 401-416.38 -1. The increased rate constant and bubble capacity were sufficient to collect middlings particles previously rejected.9658 0.42 7. First Series . First Series. Figure 3-9).58 -0.5684 In the first plot of the 0.69 T for H0 Coeff=0 Prob>|t| -3.4140 0.56 2.1787 0.14 6.05 DF 1 1 1 1 1 1 1 1 1 1 Standard Error 3. Diesel dosage had basically little effect on flotation of the 0.78 9.045 1.95 -12.

5 x 0.5 x 0.25 mm The predictive quadratic model for the coarser size had an R-Squared of 0.76 -0. 1. 0.56 -87.25 mm Recovery +995. The predictive equation in terms of actual factors is: 0.56 = * Feed Rate * Frother * Diesel CHAPTER 3 PLANT TEST AND SCALE-UP Page 46 .0 being ideal).35 -12.969 (indicating an excellent fit.

recovery of the 0.096E-05 +1.Dennis I. predicts that at the highest feed and diesel dosages. however. From the coefficients. At the medium frother dosage of 10 ml/min and a low diesel dosage. Unlike the smaller size. At the lower frother dose (8 ml/min) the combustible recovery is highest at the low feed rate just as for flotation of the smaller size range. which of course is impossible. Response plots in Figure 3-10 are based totally on the model equation and thus indicate recoveries above 100%. At the low feed rate. that is.11 +5. At the highest diesel dosages the recovery increased again due to the increased particle hydrophobicity brought about by the high amounts of collector available in the slurry.5 x 0.745E-03 +0. it can be seen that although feed rate and frother dosage had an effect on recovery. The Prob>|t| column (p values) shows that the most significant factors were frother and diesel and the interactions of frother and diesel. nonetheless.15 -0. This was probably due to the excess diesel. When viewing the 3D plots for the predictive model of 0. the recovery actually dropped with increased collector addition. diesel dosage definitely had the major effect.25 mm material increases. one notices differences from the plot of the smaller particle size range. above that needed to coat the coal. At low frother and low feed rate.25 mm recovery in this test series.033 * Feed Rate2 * Frother2 * Diesel2 * Feed Rate * Frother * Feed Rate * Diesel * Frother * Diesel The coded factor coefficients shown in Table 3-3 are directly comparable rather than being related to the units of measure as in the actual factor values in the above equation.70 +4. Figure 3-10 however. CHAPTER 3 PLANT TEST AND SCALE-UP Page 47 .5 x 0. the relationship between feed rate and recovery was similar to that of low frother dosage. This deviation from other experiences may be explained by the higher p value for feed rate in the presented model which indicates that the feed rate is not a good predictor of 0. The plots are useful. encumbering the frother and causing increased bubble size with less bubble surface area. to show the magnitude of the factor effects.5 x 0.25 mm size range (Figure 3-10). increased feed rate meant lower recovery. Phillips -0. It is well known that the fines preferentially attach to the bubble surface and that coarse particles remain attached only if there is sufficient bubble surface area available after attachment of the fines. At the highest feed rates one would expect increased diesel to have little effect on coarse particle recovery due to insufficient available bubble surface area and the preferential loading of the finer particles. diesel fuel dosages had a major effect on recovery of the coarse particle size. increasing diesel dosage appeared to lower recovery due probably to the decreased effectiveness of the frother as described above.

00 Figure 3-10.25 Rec.00 200 Feed Rate 25.25 mm. 0.00 650 30.25 Rec.00 1550 40. First Series (401-416).25 Rec.00 200 Feed Rate 25.00 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 200 150 .00 650 30. CHAPTER 3 PLANT TEST AND SCALE-UP Page 48 . 10. Phillips DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 200 150 .0 136 64 -8 2000 45.0 100 50 0 2000 45. Actual Constants: Frother = 12.00 1550 40. Actual Constants: Frother = 10. and 12 ml/min.00 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 280 208 .00 200 Feed Rate 25.00 1100 Diesel 35. Frother at 8.Dennis I.5 x 0.00 650 30. Actual Constants: Frother = 8.5x.5x.00 1550 40.00 1100 Diesel 35.5x.0 100 50 0 2000 45.00 1100 Diesel 35. 3-D Response Plots.

5 x 0. probably due to the excess frother “wetting” the surfaces of the coal particles.0268 0.25 mm material can be recovered nearly as well as the finer material.150 mm (60 x 100 mesh) fraction floated as well or better than any size fraction. First Series.02 -7.07 20.20 2.29 9.64 2. Response surface model coefficients and significance. Factor Intercept A-Feed Rate B-Frother C-Diesel A2 B2 C2 AB AC BC Coefficient Estimate 72.83 10.8491 0.91 2. utilized for coal coarser than 0.01 42.68 17.0464 0.0271 At a high frother dosage (12 ml/min) combustible recovery appeared to have been affected only by the diesel dosage.5 mm particle size.50 2.91 Prob>|t| 0.5 mm particle size. if ever.42 41.87 -0.0752 0.0628 0.59 DF 1 1 1 1 1 1 1 1 1 1 Standard Error 6.05 42. this plot does show that for most of the tests.17 -1.26 11.5 x 0.13 19.18 2. The actual size-by-size recoveries shown also illustrate the reason that flotation is seldom. At higher diesel dosages the coal surfaces are not “wetted” by the frother and maximum recovery was projected by the 3D model. 0.21 11. Figure 3-8 indicates that for some conditions the 0.91 -14.54 -0.71 58.28 34.13 4.1738 0. First Series .31 6.25 mm.Dennis I. Phillips Table 3-3. the 0. the combustible recovery began to drop off rapidly above 0.25 x 0. At the low diesel dosage the recovery of the coarse coal was depressed.2814 0. However.15 T for H0 Coeff=0 -1. Even with the best combination of parameters.28 4.0059 0.15 2.

0. and is therefore shown only in Appendix B.5 mm x 0 is essentially a combination of that shown for the 0. between that of the minus and plus 0.25 mm fractions. The intent of the second CHAPTER 3 PLANT TEST AND SCALE-UP Page 49 .Second Series Enough was learned about the operating parameters from the preliminary testing to determine the most likely parameter settings for further parametric testing.5 mm x 0 combination are. Parametric Tests . as would be expected. These settings became the midpoints for a Box-Behnken experimental design.25 mm x 0 and the 0. The statistical data for the total 0.25 mm material.5 mm x 0 Results of the 0.5 x 0.

Although earlier testing had shown that the diesel fuel dosage also affected the coarse coal recovery. feed rate was also intentionally used as a variable. Table 3-4 gives the settings and testing order for the Box-Behnken design. the intent was to remove it as a variable by holding the diesel dosage relatively constant. The percent solids in the feed slurry was to be held constant at 10%.Dennis I. Table 3-4. The diesel fuel tended to first coat the fine coal particles due to the much higher amount of surface area. any variation in the amount of fine coal caused the amount of diesel available for coarse particles to vary considerably. the actual percent solids of feed to the pilot column varied considerably. and 60 gallons per minute (gpm)). which would provide a constant g/tonne diesel dosage even as feed flow was varied. air volume and frother dosage were varied. screen wear on the feed system. 50. Phillips series of parametric testing was to determine the effect of bubble size and air fraction on coarse coal recovery. Although the volumetric amount was held constant for a given feed flow. This was performed by feeding a different amount of diesel for each of three volumetric slurry feed rates (40. Since coarse coal recovery was very sensitive to diesel dosage. Since only the remaining diesel fuel was then available for the coarser particles. and raw coal pumping surges (all of which were unique to this test series). Since any slight variation in feed volume could cause the bubbles to be more or less loaded and therefore affect recovery. the inconsistencies in coarse recovery may be due to this unintended variation in the ratio of diesel to solids. To do this. Test matrix for the Second Series parametric test. The percent solids variations of from 7 to 14% had a major impact on the diesel dosage as well. Run Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Feed Rate Low High Medium Medium Medium High Medium Medium Medium High Low High Low Low Medium CHAPTER 3 PLANT TEST AND SCALE-UP Air Medium High Low High High Medium Medium Medium Low Low Low Medium High Medium Medium Frother Low Medium High High Low High Medium Medium Low High Low Low Low Medium Medium Page 50 . but due to variations in plant feed. the grams per tonne of feed dosage varied with the percent solids changes.

Test results were entered into the Design Expert statistical computer program. CHAPTER 3 PLANT TEST AND SCALE-UP Page 51 . that is. and feed flow. 0. At a low feed rate and high air flow.25 mm x 0 (60 mesh x 0).25 mm x 0 quadratic predictive model had an R-Squared of 0.5 x 0.25 mm x 0 fraction (Second Series. The medium frother dosage showed a similar result although not as pronounced. there was little variation in recovery for the finer material in the Second Test Series. 0. as in the First Series. there are many hydrodynamic situations that can cause detachment of the coarse particles.897.25 mm fraction the p-value is 0. and that was minor. variations in feed. At a low frother dosage the model predicted a higher recovery for the higher feed rate. At these conditions the column would be most turbulent. When the program’s predictive quadratic model was used to develop the 3D response plots for the 0. The initial variables entered into the program were the design parameters. the only significant factor for the recovery response was feed rate. and 0. frother.946.5 x 0. high air corner.11 and the R-Squared is 0. It is obvious that either these results are unique or that something else was happening that would account for the deviation from predictions based on prior experience.25 mm x 0) plots since the majority of the coal is in that size range. Although the quadratic model is the best for the 0. This was contrary to normal flotation results since higher feed rates tend to overload the froth.5 x 0. air rate. Appendix B). The predicted results for the combined 0. Review of the 3D predictive plots for the coarser coal (0. air. causing lost recovery. fine coal is difficult to detach from a bubble under most conditions. there was a high recovery of coal as one would typically expect. One the other hand.25 mm coal at the low feed rate.5 mm x 0.Dennis I. and frother (within the test ranges) were found to have very little affect on the fine coal recovery. The only significant coefficient in the model is frother but it shows a decrease in recovery with increasing frother which is not as would normally be expected.5 mm x 0 particle size range (Second Series) essentially take the shape of the finer size (0. once attached. the lowest recovery is at maximum feed flow and air.25 mm) showed the results to be more erratic.25 mm x 0 material in Figure 3-11 show that. The tight grouping of the test points for the 0. At the high frother dosage the response plot also looked typical with a much higher recovery of the 0. which may explain the poorer recovery.25 mm. This was expected since even though coarse coal can attach easily to a bubble. Phillips The same three major size classes were considered for the second parametric test series as for the First Series. For each of the three frother dosages.5 x 0. Although the 0.

25 mm 20 10 0 0 5 10 15 20 25 30 35 40 Concentrate Ash (%) Feed 100 90 Combustibles Recovery % 80 70 60 50 40 30 0. Phillips Feed 100 90 Cumbustible Recovery % 80 70 60 50 40 30 0.5 mm x 0 20 10 0 0 5 10 15 20 25 30 35 40 Concentrate Ash (%) Figure 3-11.5 x 0. CHAPTER 3 PLANT TEST AND SCALE-UP Page 52 . Second Series grade-recovery plots by size for tests 451-465.Dennis I.25 mm x 0 20 10 0 0 5 10 15 20 Concentrate Ash (%) 25 30 35 40 Feed 100 90 Combustibles Recovery % 80 70 60 50 40 30 0.

It also shows that the 0.25 mm x 0 material in the overall flotation results for the combined 0. CHAPTER 3 PLANT TEST AND SCALE-UP Page 53 .5 90 .5 x 0.250 mm can float as well as up to 0.5 x 0.5 mm x 0 fraction. shows considerable movement up and down the graderecovery curve for the 0. The recovery of the 1 x 0. Even more than in the First Series of coarse flotation.045 to 0. 454 455 456 457 463 464 465 Particle Size in mm Figure 3-12. coal up to 0. Second Series (451-465) combustible recovery by particle size.25 mm coal can be expected to float reasonably well if proper attention is paid to column sizing and operation.5 mm material varied considerably.Dennis I.250 mm material had the best flotation recovery with only a slight drop for the 0.5 x 0. however.150 100 0 x . although some tests had remarkably high recoveries for material this coarse.5 x 1mm 90 . Phillips 80 80 453 70 70 60 60 50 50 40 40 458 30 30 459 20 20 460 10 10 461 0 0 462 +1mm 452 0.25 mm material (Figure 3-12). Figure 3-12 shows that regardless of significant changes in operating parameters.250 x .150 mm.150 x .250 451 .045 x .25 mm material. The lower section of the figure illustrates the heavy participation of 0. this Second Series showed that on a size-by-size basis the 0.045 100 Cumulative Combustible Recovery % Figure 3-11.

Dennis I. Phillips

Table 3-5.

Test #

451
452
453
454
455
456
457
458
459
460
461
462
463
464
465

Second Series Parametric Tests (451-465) results and operation
parameters.
Ash %
Solids %
Feed Conc. Tails Feed Conc. Tails % Wt. Comb.
Yield Rec.%
36.95
38.62
36.29
34.26
33.87
35.63
42.61
34.70
38.86
40.24
42.80
36.31
45.02
42.44
42.30

11.18
11.24
11.23
11.43
10.08
11.12
10.60
9.33
10.66
11.52
11.78
12.63
12.08
12.60
12.68

67.83
69.13
67.32
66.05
66.29
51.39
66.45
56.41
68.06
68.43
72.56
69.73
75.53
74.06
75.10

7.32
7.06
8.27
6.76
7.31
7.35
14.18
13.16
13.43
10.10
9.19
12.24
13.24
10.08
13.90

12.35
15.40
12.73
12.31
15.90
14.55
12.32
15.57
12.01
15.51
14.17
10.47
17.25
11.50
13.77

3.11
5.23
3.40
2.33
2.76
11.38
6.28
7.64
7.44
5.75
3.86
4.81
8.66
4.90
6.32

54.50
52.71
55.31
58.20
57.68
39.14
42.69
46.10
50.88
49.54
48.97
58.53
48.10
51.44
52.55

76.79
76.22
77.08
78.42
78.43
54.04
66.50
64.02
74.34
73.35
75.52
80.29
76.90
78.11
79.52

Comb.
Ash
Sep
Rej.% Eff. %
83.52
84.66
82.89
80.59
82.84
87.78
89.38
87.61
86.05
85.81
86.52
79.64
87.09
84.74
84.25

60.30
60.88
59.97
59.00
61.26
41.82
55.88
51.63
60.39
59.16
62.03
59.93
63.99
62.85
63.77

Table 3-5 cont’d.
Test #

Feed
Rate
(gpm)

451
452
453
454
455
456
457
458
459
460
461
462
463
464
465

41
60
50
50
50
60
50
50
50
60
40
60
40
40
50

Feed
Conc
Conc
Rate
Rate
Rate
(kg/min) (kg/min) Tph/m2
11.7
16.5
16.1
13.2
14.2
17.2
27.6
25.6
26.2
23.6
14.3
28.6
20.6
15.7
27.1

6.4
8.7
8.9
7.7
8.2
6.7
11.8
11.8
13.3
11.7
7.0
16.7
9.9
8.1
14.2

0.84
1.14
1.17
1.01
1.08
0.88
1.55
1.55
1.75
1.54
0.92
2.20
1.31
1.06
1.87

Conc
Rate
-.5 mm
0.73
1.03
1.02
0.87
0.98
0.83
1.38
1.37
1.54
1.37
0.76
1.90
1.04
0.86
1.66

Wash Air
Water Rate
lpm
lpm
54.6
47.8
40.9
43.7
43.7
43.7
51.8
51.8
47.8
47.8
47.8
54.6
50.5
50.5
50.5

323
350
306
350
350
329
329
329
306
306
306
329
350
329
329

Air
Air
Frother Collect Coll. Coll. Froth
Rate Fraction
(Diesel)
-.5 mm Depth
cms
%
(ml/min) (ml/min) (g/T) (g/T) (inch)
1.2
1.3
1.1
1.3
1.3
1.2
1.2
1.2
1.1
1.1
1.1
1.2
1.3
1.2
1.2

5.0
3.5
3.5
2.0
3.3
3.3
11.0
11.9
9.3
10.0
11.0
9.5
12.9
13.6
11.3

4.5
7.3
9.3
9.3
5.5
11.0
8.0
8.0
6.0
9.0
6.0
6.3
6.0
7.5
7.5

16.5
24
20
20
20
24
20
20
20
23
16
23
16
16
19

1107
1141
974
1191
1102
1096
568
612
600
764
876
631
608
799
550

1358
1383
1171
1445
1348
1385
699
765
749
1041
1128
745
834
1136
724

24
20
26
24
28
25
29
30
27
28
25
27
25
25
28

Second Series Revisited
After extensive review and cross plotting of the variables and other operating
parameters, the question of inconsistent results from tests 451-465 was resolved. The major
problem stemmed from the uncontrollable variation in percent solids of the feed. Figure 3-13

CHAPTER 3 PLANT TEST AND SCALE-UP

Page 54

Dennis I. Phillips

is a plot of the effect diesel dosage (on a gram per tonne basis, g/T) had on the air fraction. It
indicated that above a threshold value of diesel (~1200 g/T for this system), the air fraction
dropped rapidly. A decrease in air fraction from the 10 to 13% range to below 4% indicated
much larger air bubbles were being formed and thus a much lower bubble surface area was
available for the attachment of coarse coal. Since fine coal was more strongly attached to the
bubble surfaces than coarse coal, the coarser particles were the first to be lost when particle
loading on the bubbles became high. The larger bubbles may also have caused increased
turbulence that also resulted in detachment of the coarse particles. Feed solids versus diesel
fuel dosage has been plotted on the right axis of Figure 3-13. Since the intent was to hold a
constant diesel dosage on the assumption of a constant percent feed solids, it is no surprise
that there is an unintended linear correlation between feed solids and diesel dosage.

16

16

14

14

12

12

10

10

8

8

6

6

Air Fraction

4

Feed Solids (% )

Air Fraction %

The statistical analysis was re-evaluated using diesel fuel dosage, frother dosage, and
feed rate as variables. In the previous analysis of this series, air flow was found to have a
very small effect and could be dropped to allow room for diesel dosage to be evaluated in the
statistical model. With only 15 test points, 3 variables are the most that can be evaluated
using the Box-Behnken design.

4

% Feed Solids
2

2

Linear (% Feed Solids)

0
0

200

400

600

800

1000

1200

1400

0
1600

Diesel Fuel (g/T for -.5 mm)
Figure 3-13. Relationship of air fraction, diesel dosage, and % feed solids for the
Second Series (451-465).

CHAPTER 3 PLANT TEST AND SCALE-UP

Page 55

Dennis I. Phillips

Second Series Revisited

25 -0.25 mm x 0.32 1. Second Series Revisited. diesel dosage accounted for a slight recovery increase at all but the lowest frother dosage.59 2.501E-05 * Diesel2 -1. differences in recovery for this fine coal fraction are small and thus it is unreliable to predict a recovery response within the range of operating parameters tested.38 1.68 2. From the 0.3379 0.2236 0.19 -0. At a low feed rate.53 * Feed GPM +0.3094 0.8237 0.98 -0.90 0.16 0.5442 0. CHAPTER 3 PLANT TEST AND SCALE-UP Page 56 .70 2. high feed rate.636E-03 * Feed GPM2 -1.25mm x 0 Recovery = +37.75 2.086 * Feed GPM * Frother +4. Response surface model coefficients and significance.25 mm x 0 The quadratic model for the 0.29 -2.85 4. 0.044 * Diesel +13.44 -1.05 * Frother +6. 0.67 2.867 while the predicted equation in terms of actual factors is: 0.466E-04 * Feed GPM * Diesel +0.15 Prob>|t| 0.426E-03 * Diesel * Frother Table 3-6.24 -1.66 -1.66 -1.54 * Frother2 -9.87 DF 1 1 1 1 1 1 1 1 1 1 Standard Error 3.3150 The 3D response plots for this evaluation are relatively flat and thus are shown only in Appendix B.94 2.50 T for H0 Coeff=0 -1. However.2643 0.29 -3.25 mm x 0 plots of the predicted coal recovery.63 -1.84 -5.8180 0.67 1.31 1.30 -0.25 mm x 0 fraction had a R-Squared of 0.09 0. Factor Intercept A-Feed GPM B-Diesel C-Frother A2 B2 C2 AB AC BC Coefficient Estimate 86.7639 0.78 3. and low frother dosage. The best performance was at the medium frother dosage while the lowest recovery was found at the extremes of high diesel dosage. the changes in recovery due to differences in frother dosage were small.

Phillips Second Series Revisited .Dennis I.

762 while the predicted equation in terms of actual factors is: 0. A change in recovery at low diesel dosages was the most significant variation observed. increasing the diesel dosage improved the recovery by overcoming the effects of excess frother. The quadratic model for the 0. 0.25 mm material was frother and it is only weakly so. Increased feed rate normally decreases recovery due to bubble surface overload.026 * Diesel * Frother CHAPTER 3 PLANT TEST AND SCALE-UP Page 57 . Recovery dropped considerably at all feed rates with increasing frother dosage.18 * Feed GPM +0. At the higher frother dosages.5 x 0.14 * Diesel +12. As discussed earlier.5 x 0.25 Recovery = -30. the Second Series results were not nearly as reliable as those of the First Series.25 mm fraction had a R-Squared of 0.5 x 0.5 x 0.25 mm A much broader range of response in the prediction model was seen in the combustible recovery of the 0. but in this case the increased feed flow may have diluted the frother and reduced its negative effect.39 * Feed GPM * Frother +0. An unexpected response was the increase in recovery with increasing feed flow and low diesel. The decrease in recovery at high frother dosage was possibly due to “wetting” of the coarse coal by the excess frother which reduced their hydrophobicity sufficient to allow the particles to drop back into the pulp from the froth.049E-05 * Diesel2 -4. The p values of Table 3-7 indicate that the only significant factor in the model for the 0.579E-03 * Feed GPM * Diesel +0.429E-03 * Feed GPM2 -9.42 +0.58 * Frother2 -2.25 mm fraction 3-D plots (Figure 3-14).5 x 0.71 * Frother +2.

0 1050 Diesel 50.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel .0 1050 Diesel 50. Second Series Revisited (401-416). 0.0 700 Feed GPM 40.5.25Rec Actual Constants: Frother = 9. 3-D Response Plots.5mmx.0 700 Feed GPM 40.0 1225 55.0 1225 55.5mmx.25 mm.0 1225 55.5mmx.6 100 80 60 40 20 0 -20 1400 60.5 100 80 60 40 20 0 -20 1400 60.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel . CHAPTER 3 PLANT TEST AND SCALE-UP Page 58 .0 875 45.0 875 45.3 ml/min.5 x 0. Frother at 5.6.25Rec Actual Constants: Frother = 5. 7.0 875 45.0 700 Feed GPM 40.Dennis I.0 1050 Diesel 50.25Rec Actual Constants: Frother = 7. 9.0 Figure 3-14.3 100 80 60 40 20 0 -20 1400 60. Phillips DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel .

49 0.24 -11.51 16.78 DF 1 1 1 1 1 1 1 1 1 1 Standard Error 15.95 -0.88 12.70 0.32 1.24 11. Response surface model coefficients and significance.11 0.25 mm.31 5.14 16.9850 0. Factor Intercept A-Feed GPM B-Diesel C-Frother A2 B2 C2 AB AC BC Coefficient Estimate 83.63 0. Second Series Revisited.01 7.2099 Second Series Revisited .5 x 0.24 T for H0 Coeff=0 Prob>|t| 0.16 15.020 -0.3973 0.03 7. 0.4255 0.5204 0.1023 0.94 4.77 -16.55 12.24 7.78 0.7642 0.40 -2.59 12.89 -0. Phillips Table 3-7.09 -15.7112 0.4772 0.67 -9.Dennis I.83 22.

and is therefore shown only in Appendix B.5 mm x 0 The predicted flotation response of the 0. Therefore. the 30-inch pilot-scale column was removed from the plant shortly after completion of the last test series. frother. diesel. but it does indicate very well the detrimental effect of excess frother on coarse coal recovery. 3.5 mm x 0 is essentially a combination of that shown for the 0. CHAPTER 3 PLANT TEST AND SCALE-UP Page 59 .25 mm x 0 and the 0. Replication of the major parametric test for this project at the Lady Dunn Preparation Plant was not possible due to the onset of construction for a major plant upgrade. The statistical data for the total 0. and feed rate. Laboratory test work on coarse particles is always difficult due to particle settling.25 mm material.5 x 0. etc.5 mm x 0 composite was obviously a combination of the two previous size ranges. 0. All of the test work presented here showed that coarse coal can be floated successfully when proper attention is paid to control parameters such as air. as discussed previously. differences in samples. Producing a coarse slip stream in an operating plant is also difficult.5 Full Size Flotation Column Scale-Up The successful testing of the 30-inch pilot column coincided with an expansion project at the Lady Dunn Plant which prompted interest in a full-scale column that would clean coal coarser in size than was currently being sent to flotation.

A heavy media cyclone circuit was installed as part of a recent plant upgrade. Plans for the new circuit included a new heavy media vessel for the coarser sizes and an additional heavy media cyclone circuit for the intermediate sizes. 14-16%) and the combustible recovery seldom exceeded 20%. Therefore. Newer continuous mining methods brought increased amounts of fines. Phillips Lady Dunn Plant Circuit The evolution of the Lady Dunn Preparation Plant flowsheet parallels the classic history of coal preparation. For these reasons. A two-stage spiral circuit was tested for the minus 1 mm material. Deister tables. The original plant was designed and built in 1951 for an easily cleaned metallurgical coal (Cedar Grove seam) produced by old conventional mining methods which also provided coarser feed coals.200 tph.25 mm instead of the normal 0.. frother.25 mm (60 mesh).5 mm. and feed rate. but more capacity was needed to handle the planned increase in raw feed tonnage to 1. especially below 0. A change in coal feed to predominantly steam coal (Stockton seam) brought a more difficultto-clean high-middlings coal along with more reject material. The 550 tph flowsheet of heavy media vessel. the selection of an appropriate circuit(s) to treat the fine sizes required more study. Test work with the pilot-scale column showed that a substantially lower clean coal ash (i. Applicable Pilot-Scale Testing Testing of the pilot-scale unit indicated that the column is capable of recovering coal in all sizes up to at least 0. The primary flotation feed was the raw minus 100 mesh classifying cyclone overflow with a secondary feed coming from a clean coal sieve effluent. the ability of the column to treat material from a coarser size-cut of 0. Previous test work conducted at Virginia Tech (Jha et al. 8-10%) and higher recovery (70-80%) was possible by column flotation.25 mm material and all of the minus 0. The testwork also showed that the recovery of the coarser particles is sensitive to operating parameters such as air.15 mm (100 mesh) meant that a much more efficient circuit was possible. In addition. but performance of the spiral deteriorated rapidly on the finer sizes. 1997) showed that a column froth becomes overloaded when the amount of floated solids per unit of available bubble surface area becomes too high. The original flotation circuit at the Lady Dunn Plant consisted of four 5-cell banks of mechanical flotation cells.e.e. including high ash clays. alternative methods for treating this size fraction were considered..25 mm (60 mesh) was sent to column flotation.Dennis I. all of which needed to be addressed in full-scale design. spirals were installed to treat the 1 x 0. The ash content of the flotation product was relatively high (i. and conventional flotation was no longer adequate for the new plant feed material. A simplified version of the new plant flowsheet is shown in Figure 3-15. which overloaded the fine coal circuits. High ash slimes normally entrained in flotation froths were easily rejected by the column wash water system. As usual. This condition can occur CHAPTER 3 PLANT TEST AND SCALE-UP Page 60 .

25 mm sized feeds rather than the minus 0. This plot shows that the coarse particles can be recovered by column flotation. finer particles tend to remain preferentially attached to the bubbles and coarse particles are lost.25 mm fractions. The upper two (overlapping) curves show that recovery drops with increasing feed rate in nearly identical fashion for both the minus 0. Screen Microcel Flotation Columns ½” x 1mm Spirals H. Simplified flowsheet (new circuit) for Lady Dunn preparation plant. Phillips when the amount of floatable feed solids increases or the bubble surface area is reduced (by reducing frother dosage or air rate).25 mm material displays a rapid drop in recovery with increasing feed rate for the coarser material. As the bubble surface area is reduced. CHAPTER 3 PLANT TEST AND SCALE-UP Page 61 . Unfortunately.15 mm (100 mesh) and the minus 0.25mm x 0 Screen + !/2” Classifying Cyclones H.M. As shown in Figure 3-16 feed rate has a major effect on column performance.V. but these particles are lost much faster than the fine particles as the column becomes overloaded. These two curves show that column flotation should be considered for minus 0.M. The curve for the 0.Dennis I.C. the recovery and clean coal ash decreases as middlings particles are rejected back into the flotation pulp.15 mm feed (100 mesh) traditionally used in coal flotation circuits.5 x 0. Raw Feed 0. 1mm x 0 Screen Bowl Centrifuge Figure 3-15.

CHAPTER 3 PLANT TEST AND SCALE-UP Page 62 . i. As shown in Chapter 2. medium (30-40 kg/min) and high (40-50 kg/min).25 mm x 0) and coarse (0. The recovery improved very little as the diesel dosage was increased at the lower feed rates. Effect of feed rate increase on recovery by particle size for 30-inch pilot-scale column. The other variable that has a significant influence on coarse coal recovery is the hydrophobicity of the particles. low (20-30 kg/min). Three separate feed rate groupings are shown. the contact angle has a significant effect on the amount of force required to detach a particle and the contact angle is essentially a measure of the hydrophobicity of a surface. while diesel dosage had a much larger impact on recovery for the higher feed rates..2 5 mm 0 20 25 30 35 40 45 50 Fe e d Rate (k g /m in ) Figure 3-16.5 x 0 . Figure 3-17 shows that recovery increases with increasing diesel dosage (g/tonne) for both the fine (0.5 x 0.e. The increased hydrophobicity imparted by the diesel allows for a much stronger bond between the coarse particles and bubbles. Phillips Recovery vs Feed Rate 100 Re co ve r y % 90 80 70 60 50 40 0 . However. This reduces the probability of coarse particle detachment which would otherwise occur at the higher feed rates where the froth carrying capacity may be exceeded. the rate of increase in recovery is much higher for the coarser material.25 mm) material. The addition of a collector will often increase hydrophobicity and for coal the most economical collector is common diesel fuel.2 5 mm x 0 10 0 .1 5 mm x 0 30 20 0 .Dennis I.

The apparent leveling off of carrying capacity at the larger particle sizes is probably due to the skewedness of the particle size distributions for coal. Full-Scale Column Design The coal company made the decision for the author to proceed with design of fullscale Microcel columns based on the pilot-scale test work. however.5 metric tonnes (1. a considerable amount of material and particle surface area is usually found at the smaller diameters.5 x. Equations developed by others. This production capacity was in good agreement with the test values reported elsewhere (Luttrell et al. have not been validated for particle sizes much above 40 m.25 m m x 0 80 0. are only valid for smaller particle sizes. C a 0. this equation.G.25 m m Re co ve r y % 70 60 0. Although often overused.65 short tons) per hour per square meter. 1993). The first major design concern was the unit capacity since it affected the column diameter and number of columns.5 x 0. The d80 size refers to the 80% passing particle size of the floatable material rather than the feed size.5 x 0.2 5 mm @ 2 0 -3 0 kg / min 40 0. Phillips C ollector vs Recovery 100 90 0. Although the d80 may be relatively coarse.2 5 mm @ 3 0 -4 0 kg / min 20 0 . 3-2 by Finch and Dobby (1990) would indicate an unrealistic value of 23 Tph/m2 for froth carrying capacity (Ca) when using a typical d80 particle size of 400 m and a particle S. (p) of 1.25 m m 30 0 . 3-2 Page 63 .041d 80  p CHAPTER 3 PLANT TEST AND SCALE-UP Eq.Dennis I. Eq. From the pilot-scale testwork.5 x. Effect of collector on combustibles recovery (pilot-scale).5 x 0. clean coal capacity of the froth appeared to be in the range of 1. (1988). as well as the plots by Espinosa-Gomez et al.2 5 mm @ 4 0 -5 0 kg / min 10 0 0 500 1000 1500 2000 2500 Die s e l (g /T ) Figure 3-17.5 x.2 5 mm x 0 A ll F eed s 0 .25 m m 50 .40.

The recirculation of column CHAPTER 3 PLANT TEST AND SCALE-UP Page 64 . Three 4-meter diameter columns would provide 12. the pump motor horsepower is only 100 hp compared to 75 hp for the 3-meter unit. Further savings are realized on structural steel. an internal launder was used to minimize the possibility of immobile froth in the center of such a large column since keeping the froth mobile would help to prevent coarse particle dropback.5 total square meters of surface area. piping.5 square meters each or 37. This area requirement could be achieved with six standard 3-meter diameter columns. and other costs associated with fewer units.5 Tph/m2. the maximum projected clean coal tonnage of 55 tonne/hr (60 tph) would require 37 square meters of surface area. an alternative was sought which led to the design and development of the largest diameter columns known to exist for coal processing. In the interest of reducing costs (as well as space limitations). In addition. which is less than that of successful 3-meter diameter columns currently operating in the field. Wash Water Feed Column Concentrate Slurry Distribution Ring Air Inlet Air Distribution Ring Valve Tailings Valve Microbubble Generator Frother Inlet (Optional) Circulation Pump Figure 3-18. This would be adequate to handle the maximum projected feed to the column circuit.Dennis I. Figure 3-18 depicts the operation of the patented bubble generation system for the Microcel column. Phillips Using the 1. The internal launder made the maximum froth travel distance less than 1 meter. The full-scale system is similar to that of the 30-inch unit except that the flow is downward through the spargers rather than upward. Also. The 4-meter diameter columns provide 77% more area and capacity than 3-meter units at only 25% more cost. Full-scale MicrocelTM column and bubble generation system.

Diesel consumption is CHAPTER 3 PLANT TEST AND SCALE-UP Page 65 .15 mm (60 mesh) material. In fact. Full-scale 4-meter diameter column installation at the Lady Dunn Preparation Plant. Plans were underway to enlarge the froth handling system as well as modify the screen bowl centrifuges with larger inlets.70 S. 1992. 1993. float for the 0. Level is controlled by a proportional tailings valve via a PID loop controller using a signal from a level transmitter. These are excellent results for this relatively high middlings coal. Jha et al. Luttrell et al. Full-Scale Column Results The column circuit (Figure 3-19) was commissioned in August of 1996 and immediately began producing a product within specifications. Figure 3-19. The data points fall on essentially the same grade-recovery curve with a product ash of less than 10% (well below the 12% ash target). The major limitation on column production at the Lady Dunn Plant was froth handling capacity. Phillips slurry through a sparger with the addition of air and frother provides for a very consistent bubble size in any quantity needed. Results from initial testing of the 4-meter units show that the performance is similar to that predicted from the pilot-scale unit (Figure 3-20). Combustible recoveries of up to 80% have been obtained with the expected normal being 75% at a 10-12 % product ash. Phillips 1997).Dennis I. Initially. washability data indicates that a 9% cumulative ash would be obtained at 1. the columns had the capability of producing more froth at a satisfactory ash than could be fed to the dewatering circuit.25 x 0.G. Detailed descriptions of the Microcel technology have been provided elsewhere (Davis et al. 1996.

Phillips 500-800 g/tonne (1. 97 20 10 0 0 5 10 15 20 25 30 35 40 Concentrate Ash (%) Figure 3-20. However.5 lb/ton). Results of 4-meter and pilot-scale column tests. The size-by-size recoveries shown in Figure 3-21 indicate excellent recoveries up to and including particle sizes of 0. Recovery (%) 70 60 50 40 30-inch column tests 30 4-meter Mar. while frother dosage is 270 g/tonne (~0. CHAPTER 3 PLANT TEST AND SCALE-UP Page 66 . the columns are currently heavily loaded and a high bubble surface area must be maintained if a high combustible recovery is to be obtained.6 lb/ton) at peak demand.5 mm.0-1. Feed 100 90 80 Comb.Dennis I. Figure 3-21 illustrates this fact by showing higher recoveries for all sizes at the higher frother dosage.

25 mm cut and sending the minus 0. Phillips 100 Comb.25 mm + . but coal recovery drops off rapidly above this size.045mm .Dennis I. The success of this test work was made tangible by the installation of three MicrocelTM flotation columns.5 mm Particle Size Figure 3-21.6 Summary and Conclusions Coarse coal flotation is alive and well. a flotation column should provide very good recoveries at a low product ash.25 mm to a flotation column should work well in most coal processing plants.25 mm x 0 material at a combustibles recovery of 75%.15x.25x.15mm .5 mm can also perform well in a column. Figure 3-12 shows that particle sizes up to 0. each four meters in diameter. The testwork presented here illustrates well the potential for coarse coal flotation in a properly operated flotation column. As long as the misplaced +0. Since it is difficult to separate fine particles accurately by size. Results from testing the 30-inch diameter column indicated that a product of 10-11% ash could be obtained from the 0. Results matched CHAPTER 3 PLANT TEST AND SCALE-UP Page 67 .5 mm material in the column feed is minimal. Recovery % 90 80 70 60 50 40 High frother Medium Frother 30 20 10 0 Cumulative . The high flotation feed ash containing much clay was a major indicator that a flotation column would perform much better at the Lady Dunn Plant than conventional flotation cells. 3.25 mm can be floated consistently in a column. making a nominal 0.045 mm x 0 . Size-by-size recovery in the 4-meter Microcel column. with little misplaced material.5 x . It has shown that coarse coal up to 0. In fact. The 4-meter MicrocelTM columns at the Lady Dunn Preparation Plant produced 10-12% clean coal ash from a 35-40% feed ash at combustible recoveries of 70-80% (weight yields of 50-60%). the original mechanical flotation cells produced an average of 14-16% ash clean coal at only a 20% combustibles recovery.

to pilot-scale testing. to laboratory tests.15 mm) that makes it to the spiral circuit reports to clean coal at 20-25% ash. Phillips well with those predicted from the 30-inch pilot-scale test unit and illustrate the successful scale-up that is possible with proper testing and design. material of the same size (0.25 mm (60 mesh) material by the columns rather than a traditional minus 0.15 mm material. CHAPTER 3 PLANT TEST AND SCALE-UP Page 68 . This project proceeded successfully from an idea. The pilot-scale test data also suggests that any of the coarser 0.25 x 0.25 x 0. The columns can produce a 6-9% ash product at over 70% combustible recovery for the 0. The decision to treat minus 0. In comparison.15 mm (100 mesh) feed was validated by the performance data obtained using the full-scale columns. and then to full-scale commercial operation.5 x 0.Dennis I. At the time of installation these were the largest known flotation columns processing coal anywhere in the world.25 mm material that may be misplaced to the columns would also be recovered at values close to that of the smaller material.

Luttrell.Dennis I.J. R. Smit. and Luttrell.. DOE Contract DEAC22-92PC92208.A. 1 (1). D. T. Jha.. CHAPTER 3 PLANT TEST AND SCALE-UP Page 69 .-H. 1990.I. M..7 References Davis.. Minerals Engineering. Yianatos. V.. Australia. Elmsford. Lexington. G.S. Installation of 4Meter Diameter MicrocelTM Flotation Columns at the Lady Dunn Preparation Plant. Fish. Phillips. Proceedings.A. Finch. A Comparison of Column and Conventional Flotation Cells for Fine Coal Cleaning. J.. R-H. G. Proceedings. Proceedings.J.J. T.. Lexington. 1997. Espinosa-Gomez. 1992.I..... L. Column Carrying Capacity: Particle Size and Density Effects. 9th International Coal Preparation Exhibition and Conference. R...S. G. and Dobby. G.. C.S. M..H. and Feeley... Smit. Lexington. and Dobby. J. and Yoon. 1993. 14th International Coal Preparation Exhibition and Conference. U. Phillips 3. KY.. Topical Report.A. KY. F. Evaluation of Column Flotation for Fine Coal Cleaning at the Lady Dunn Preparation Plant.J. KY. XVIII International Mineral Processing Congress. Proceedings.. New York. Mankosa. and Yoon. D. Luttrell. Toney. pp 77-79. Column Flotation. 13th International Coal Preparation Exhibition and Conference. 1997. 1996. Design and Scale-up Criteria for Column Flotation.I. R.. Sydney. Mankosa... L.2 Engineering Development of Near-Term Applications. Phillips. Pergamon Press. Yoon. Jha. M. G. Phillips. M.J.H. 1988.L..B..-H. Toney. Subtask 3. D..C. T. Fish. F. H.

which usually brings a premium or an increase in sales price. then the original heating value is maintained but the total tons have increased. In many instances however. The importance of successful dewatering of fine coal has been recognized since water was first used as a medium for coal cleaning. the amount of saleable coal is reduced by the amount of moisture reduction. For this reason. Although the coal industry has long been interested in methods for removing more surface water from fine coal. it appears that often little additional thought is given to it once the mechanical systems are operating properly. In fact. any excess moisture is a reduction in the thermal value of the coal and thus of some importance to the utility.1 YIELD ADVANTAGES OF IMPROVED DEWATERING Introduction With the advent of column flotation and its use of wash water. If the overall plant performance is reconsidered in terms of the potential yield gain. For metallurgical (coking) coal markets there is little incentive to reduce moisture below that required by the contract since the contracts are on a dry ash basis. Removing some of the water from the coal results in an increase in its thermal (BTU or calorie) value. if water is removed from coal. Phillips CHAPTER 4 4. With steam or utility coals. any fine coal amenable to flotation can now be cleaned to a reasonably low ash. If combustible material comes with this ash as in the case of adding middlings (middle gravity material) by increasing media gravities. many advantages to even a small improvement in dewatering are often overlooked. all of the fine coal cannot be tolerated in the product due to its excessively high moisture content. However.Dennis I. Improvements in the fine coal CHAPTER 4 YIELD ADVANTAGES OF IMPROVED DEWATERING Page 70 . Premiums paid are usually only a little more than the value of the coal tonnage reduction and thus there is little revenue remaining to cover the cost of the additional dewatering. Many operations consider the extra effort required to perform better dewatering on the fines to be not worth it. water is just like rock in that it has no heating value. however. 4. it can be replaced with an equivalent amount of pure ash and maintain the same heating value per ton.2 Ash for Moisture Simulations To coal. more effort may be expended on recovering more fine coal and its associated dewatering.

25/100 BTU premium. $ Benefit per Ton of Fine Coal Treated $ Be ne fit pe r Ton of Fine Coa l Tre a te d The two ways of realizing the benefit of fines moisture reduction on steam coal at a typical plant are shown in Figure 4-1. Phillips circuit often benefit the coarse circuit more than the fine circuit. 12. the original heating value is maintained by raising the media gravities on the coarse coal circuits. If the fines moisture drops by at least 5 percentage points. For the 1000 tph plant example this is $200 per hour or approximately $1. For the example used. even after the $5 per ton of dewatering cost. then the value of the additional yield will result in a net benefit after only a 1 point drop in fines moisture (upper line on graph). a drop in flotation product ash from 9. $0. the slope of the benefit line is the same but a greater moisture drop is required to break-even on the dewatering costs (Figure 4-1b) than was required at $1 per treated ton. At Peak Downs Plant in Australia. If. If the cost of dewatering the fines is $5 per ton of fines treated. a 10 point moisture drop in the fines (from 30 to 20%) produces a gain of $3. Any moisture drop on the fines above 1 point then brings considerable revenue.40 per ton of fines treated.Dennis I. the cost of additional dewatering is recovered on the basis of premiums for the higher heating value coal. CHAPTER 4 YIELD ADVANTAGES OF IMPROVED DEWATERING Page 71 .2 million of additional revenue. Base case is $20/ton sales. b) $5 per ton of fines treated.5 to 6% by the addition of new flotation columns allowed the coarse circuit ash to be raised such the total plant saw more than 4 points yield increase (Brake and Eldrige 1996).500 BTU. a dewatering cost of $5 per ton can never be recovered. 11 10 9 8 7 6 5 4 3 2 1 0 -1 -2 0 -3 -4 -5 -6 W ith Inc reas ed Coars e Circ uit Y ield B TU G ain O nly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Pe rce n ta g e Po in t M o istu re Dro p o n F in e s a) Figure 4-1. 11 10 9 8 7 6 5 4 3 2 1 0 -1 -2 0 -3 -4 -5 -6 With Increased Coarse Circuit Yield 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 BTU Gain Only Percentage Point M oisture Drop on Fines b) Total effect of moisture reduction on fines portion at dewatering cost of: a) $1 per ton of fines treated. 30% fines moisture. on the other hand. In Figure 4-1a the cost of additional dewatering is $1 per ton of fines dewatered. If only the BTU premiums are sought. Maintaining a constant BTU and increasing coarse circuit yield will produce a revenue gain for any moisture drop greater than approximately 5 points.

BTU premiums are at $0.85 0.04 485. Only the media gravities on the bath and cyclone circuits are considered truly variable and thus were the only ones changed.15 53. A common mistake made when evaluating moisture reductions is failing to reduce plant product tonnage by the amount of moisture reduction.00 13251 212. Feed Wt% Cut SG Ep Feed Rate D M Bath 36.66 56. Rather than collect the BTU premiums.1 tph for the example in Table 4-3.00 9000 Plant Totals ** ** 1000. one can see the effect of all changes on total plant results.49 0.500 BTU currently being produced by the plant (Table 4-1).88 56. AR AR Ash Yield Rec.65 81.71 78.9 Froth Flotation 10.00 41.0 8.9 Coal Spirals 5.88 77. To optimize yield for each simulation the gravities were adjusted on the principle of maintaining a constant incremental ash plus moisture (Abbot 1981) on the media circuits.57 1. 30% fines moisture.7 Page 72 .84 64.3 Base Case and Modifications The plant example used includes common cleaning devices as shown in the circuit description in Table 4-1. Phillips 4.09 1. Table 4-1.01 12500 604.9 100.02 360.00 ** ** 100.9 D M Cyclone 48.25 per 100 BTU and the initial fines moisture is 30%. Froth flotation is the source of the high moisture fines but the concepts presented here apply to any high moisture portion that has potential for further dewatering.37 4.00 10224 39.65 55.Dennis I. The simulations show that this increased yield produced an additional 15. By showing the results in table form.41 8.80 6.52 0.27 12. it is more profitable to increase the separating gravity as in Table 4-3 while maintaining the same BTU as in the base case of Table 4-1.09 64.0 10.34 1.9 7. The only change for Table 4-2 is a drop in fines moisture from 30 to 25% which results in increased product BTU and a lower product tonnage. Moist BTU Tons CHAPTER 4 YIELD ADVANTAGES OF IMPROVED DEWATERING 58.4 19.00 Clean Clean Comb.7 7.00 12918 293. Circuit Description Base case (original) plant settings.73 85. Assumptions used include a base sales price of $20 per ton for 12.48 30.

52 0. Feed Wt% Cut SG Ep Feed Rate D M Bath 36. Circuit Description Plant settings for 25% fines moisture with BTU gain only. Circuit Description 100.21 80. Phillips Table 4-2.27 83.09 1.15 53.28 7.42 12590 601.37 4. Ash Yield Rec.35 12500 619.27 12.0 D M Cyclone 48.02 360.04 485.23 6.4 100.34 1.8 Conclusions Any improvements in the ash or moisture of the fine coal circuit can have tremendous benefits in the total plant product.0 4.09 64.45 80.72 59.00 13147 219.36 58.00 12918 293.00 12791 305.00 Clean Clean Comb.48 25.80 6.00 13251 213.8 Plant settings for 25% fines moisture with increased coarse circuit yields. AR Clean Clean Comb. It is often overlooked or forgotten that normal plant operations include compensations made in the coarse circuit due to higher ash or moisture of the fines.0 Table 4-3.65 81.06 4.88 56.31 57. Moist BTU Tons 54.84 64.53 0. Moist BTU Feed Wt% Cut SG Ep Feed Rate D M Bath 36.4 19.7 8.15 53.9 8.00 41.66 56.34 1.73 85.00 9750 Plant Totals ** ** 1000.57 1.73 85.27 12.00 ** ** 100.0 10.00 ** ** 100.3 Froth Flotation 10.48 25. CHAPTER 4 YIELD ADVANTAGES OF IMPROVED DEWATERING Page 73 .71 78. AR AR Ash Yield Rec.0 8.0 10.49 0.02 360.56 0.9 Coal Spirals 5. The value of the yield increases possible in the coarse coal circuits due to improvements in the fine circuit can often easily justify these improvements even though the apparent benefit to the fines circuit is small.57 1.0 D M Cyclone 48.0 9.09 1.Dennis I.85 0.7 7.00 AR Tons 54.9 7.4 19.09 64.00 10224 39.00 10224 39.41 7.00 41.65 55.85 0.88 77.04 485.84 64.3 Froth Flotation 10.9 Coal Spirals 5.00 9750 Plant Totals ** ** 1000.

R. and Eldrige.. KY. Newcastle. G. I. 1996. The Optimization of Process Parameters to Maximize the Profitability From a Three-Component Blend. J.Dennis I. 87-106. Conference. pp. 237-251. The Development of New Microcel Column Flotation Circuit for BHP Australia Coal’s Peak Downs Coal Preparation Plant. Phillips 4. Proceedings First Australian Coal Prep.5 References Abbott. CHAPTER 4 YIELD ADVANTAGES OF IMPROVED DEWATERING Page 74 . Brake. Proceedings 13th International Coal Preparation Exhibition and Conference. Lexington.. 1981. pp..

e.150 mm) previously considered too high in feed ash for treatment. The advantages of coarser particle flotation are two-fold: 1) the combustible recovery or cleaning efficiency for column flotation may be greater than other devices. high ash fines are often discarded to refuse with no flotation due to the difficulty in treating with conventional flotation cells.150 mm in size. but those circuits are expensive and difficult to operate. Spirals are often the current choice for material down to 0. relatively coarse coal should be floatable under the proper conditions.5 mm fraction with reasonable efficiencies. Heavy media cyclones can process particles as small as 0. Phillips CHAPTER 5 5.Dennis I.150 mm. However.150 to 0. but it can also be deslimed of high ash fines in the column. 0.150 mm but detailed test data shows that spirals perform very poorly on particles below 0. Anytime that high ash fines can be removed from the product. sufficiently hydrophobic and not oxidized) that column flotation be extended to much coarser sizes than that of current practice. CHAPTER 5 QUANTITATIVE EVALUATION…. 2) the ability to make a much coarser size cut as the smallest size cut made in the plant.25 mm in size.e. they can be replaced with middlings particles from the coarser coal circuit.5 mm remains a difficult area. Not only can it be floated.1 QUANTITATIVE EVALUATION OF THE APPLICATION OF COARSE COAL FLOTATION TO PLANT CIRCUITRY Introduction The model discussed in previous sections showed that fundamentally. the yield can be increased and the overall product will be coarser. The ability to successfully clean fine particle size coal (minus 0. The pilot-scale testing and later full-scale operations at the Lady Dunn Preparation Plant showed that coarse coal can indeed be floated quite successfully in a column. Even though several devices such as spirals and heavy media cyclones can process the 1 x 0. the size range below 0. It is proposed that in all cases where the coal is amenable to flotation (i. Flotation is the only choice for cleaning material finer than 0. Page 75 . By raising the separating cut-point on the coarser or intermediate sized coal. But what are the benefits of treating coarser particle material (i. is of a significant benefit by itself (Davis 1993).5 mm) that is currently already being processed by other devices? Can coarse coal flotation benefit plant operations and what are the real advantages? Processing coal smaller than 1 mm in size has always been more difficult than processing the coarser sizes and remains so today. provides for much less misplacement of high ash fines to clean coal.

5 mm. Case C provides for the column to handle everything minus 0. In case C and D the smallest size cut made in the plant is 0.5 to 1 mm. Column flotation treats all material below 0. In addition. A common circuit today consists of sending minus 1 mm material to a classifying cyclone with the overflow (nominal 0. and D circuits was evaluated and modeled separately.150 to 1 mm and heavy media cyclones (HMC) for the plus 1 mm material (Figure 5-1). Some comparisons were made to actual plant performance where that information was available.5 mm in size and the spirals then see a very narrow size range of 0. therefore utilizes column flotation for the minus 0. Phillips Since most of the advantages of improved performance for a given particle size range may actually be realized in another circuit.150 mm fraction.2 Case Circuits and Evaluation Methods Four different plant configurations were utilized for the evaluations. C. The base case (case A). CHAPTER 5 QUANTITATIVE EVALUATION….Dennis I.150 mm x 0) going to either flotation or discard. Development of the case studies involved the understanding and modeling of several devices. this work evaluated the use of column flotation in four different plant circuit configurations.5 mm. A discussion of the rationale and modeling of each cleaning or sizing device follows. Each device of the A. 5. which was common in the coal industry prior to the utilization of spirals.25 mm. The underflow of the cyclone is then cleaned in spirals. which results in very little misplacement of fines to the oversize portion. Various models for sizing and gravity separation were employed for the predictions of total plant product. The effects of misplaced particle sizes on total performance was also considered. In case D the spirals are eliminated and the heavy media cyclones treat everything above 0. B. the same four case circuits were evaluated on two separate coal seams to determine the influence of coal type on the results. Page 76 .5 mm. Case B is similar to case A except that the top size going to column flotation is 0. spirals for 0.

150 mm material in these case circuits. CHAPTER 5 QUANTITATIVE EVALUATION…. that a column is the best choice and is state of the art. HMC Particle size and devices for the case circuits. Phillips 0 0.25 0. Column Flotation Several authors. Detachment is a major concern for floating coarse particles and the fundamental model presented previously shows that it should be possible to float coal up to at least 0. and a few plant installations have shown the clear advantage of flotation columns over conventional flotation cells when treating a high ash. Page 77 .5 mm.5 1. test work conducted for the USDOE.15 0. minus 0.150 mm feed. The author’s belief that coarser material can be floated well in a column led to the test work at the Lady Dunn Preparation Plant which along with the eventual full-scale installation there showed that coarser particles can be floated successfully in a column. Columns are thought by many in the industry (Fonseca 1995) to be effective only on the ultrafine material and that material coarser than 0. For this reason it was taken as a given that if flotation is to be used on the minus 0.150 mm cannot be recovered well in a column.Dennis I.0 mm + A Column Spirals HMC B Column Spirals HMC C Column Spirals HMC D Column Figure 5-1.

25 mm material. Page 78 . feed rate range 80 and 93 g/min. The small feed rate effect was overcome at the higher diesel dosage however. A 12% ash product was a common result and was near the “elbow” of the grade-recovery curves during testing for both the minus 0. CHAPTER 5 QUANTITATIVE EVALUATION….25 mm feed.Dennis I. 5-Block seam The 5-Block seam predictions were based on laboratory column tests of that seam from Central West Virginia.150 mm feed and Case B with minus 0. Com bus tible Re c ove ry 1 00 90 80 70 60 50 40 30 20 10 Hig h Hig h L ow L ow Die s e l / Lo w Die s e l / High Dies el / Lo w Dies el / Hig h Fe ed Fe ed Fee d Fe ed 0 Pa rtic le Size Figure 5-2. the proper amount of diesel collector can produce high recoveries on particle sizes up to at least 0. Figure 5-2 illustrates that in the case of a high volatile coal. Recovery by size from 2-inch column tests on high-vol 5-Block seam coal.5 mm. The 12% ash product was thus used for both Case A at minus 0. Phillips Stockton Seam The data used to predict column flotation results in the case evaluations came from actual test data of the 30-inch and 4-meter diameter columns as tested on the Stockton seam coal and is presented in detail in Chapter 3. The effect of feed rate is also shown on the two low diesel dosage tests in which a slightly lower feed rate (80 vs. 93 g/min) produced higher recoveries. The recent work in a 2-inch column provides further support for the control of operating parameters necessary to produce high recoveries on the coarser coal particles. diesel rates 250 & 500 g/Tonne. In all four tests air fraction was held constant by controlling frother to a relatively constant ppm basis. The sample had a top size of 0.150 mm and minus 0.5 mm and thus predicted performance on both fine and coarse sizes.

20 51. If anything. The full size-by-size results from the 5-Block seam tests are given in Table 5-1.71 13.17 62.36 71. Phillips Increased recovery through the use of higher diesel or frother dosages or lower feed will bring higher product ashes as expected from a grade-recovery curve.045 .64 8.46 51. With conventional flotation cells any change made to increase coarse recovery will usually produce a major increase in fine particle ash.49 4.150 P roduct Ash % 12 Figure 5-3.39 36.13 100.00 36.75 46.71 100.6mm Total 0 0 x .Dennis I. diesel rates 250 and 500 g/Tonne.67 34. Figure 5-3 shows that a drastic increase in fine particle ash does not happen. BR-5 Size . 14 10 8 6 P a rticle S iz e .045 x .Ash Ash % Flt.07 42.77 6.10 8.31 4.00 100.250 High High Low Low 4 .64 70.36 62.00 19. this is not the case.250 x .Ash 16.12 23. the product ash increases less for the finest sizes than for the coarser.09 23.66 40. feed rate range 80 and 93 g/min.88 11.150 x .32 12. but rather the product ash increases a comparable amount for all sizes. % Wt.61 14. Size-by-size results of 2-inch column flotation tests on 5-Block seam coal.045mm x 0 Total FEED CONC Feed Cum Conc Cum % Wt.69 23.10 70. TAILS Tails Cum Ash % Flt.150 .97 9. % Wt.6 x .42 69.13 9.49 7.50 71.75 Page 79 . Product ash by size from 2-inch column tests on high-vol 5-Block seam coal.045mm 2 Dies el / Low Feed Dies el / High Feed Dies el / Low Feed Dies el / High Feed .00 CHAPTER 5 QUANTITATIVE EVALUATION….250 x .150 x . However.74 8.18 65. with a column using wash water in a deep froth.250mm .Ash 23.00 7.48 31.48 51.79 8. Ash % Flt. Table 5-1.

09 57.70 45.150 and 0.70 59.250 x .150 mm) material due to the lower feed ash on the 0.250mm .05 88. problems.68 48.71 62.5 mm material.70 59. while all circuits require a coarser cut at either 0. often unacknowledged.91 79.68 54. Case C and D look at utilizing a column on material up to 0.150 mm and should result in less misplaced high ash fines to the spiral circuit and 2) the column may perform better than spirals on material smaller than 0.5 mm cut can be made very economically on conventional vibrating screens and 2) a column can make the equivalent of a much lower gravity cut than can spirals. The yields for the coarser material are higher than the finer (-0. Flotation yields for the coarse material vary considerably depending on the operating parameters used. As discussed elsewhere.39 31. All of the raw minus 0. the coarse is always the most difficult to float and thus results in a lower product ash than that of the finer material. Eff.67 47. 57. For these reasons. Sizing Devices (Classifying Cyclones and Sieve Bends) Circuits A and B require a fine size cut at either 0.78 90. while an 8% ash and 61% yield value was used for the 5-Block seam. 48.36 59.25 x 0.25 mm) are the cause of many. Phillips BR-5 Size .5 mm.68 70.09 89. The major advantages of these circuits are 1) much better sizing since a 0. Ash Yield Rec.36 Column Case Values The potential advantages of Case B over Case A are 1) size classification is better at a 0.68 94. no misplacement of sizes is considered for this coarse size cut. 42.16 71.61 70.61 70. Comb. For the Stockton seam a value of 8% ash and 60% yield was used for this coarse material.150 mm or 0.00 53. The inefficiencies of the finer size cuts (0.150 x . From Bottom Wt.60 62. Ash Sep Yield Rec.38 31.58 Sep Eff.Dennis I.16 62.25 mm cut than at 0.5 mm material came from the same test data as for the smaller sizes.045mm x 0 Total Individual by Size Wt.045 . Expected column performance values for the 0. usually report to the clean coal stream and if not removed downstream.37 70.08 83.58 59. There was no split feeding of sizes to the columns.25 mm.5 and 1 mm is common to all circuit configurations and can be performed in actual plant practice at relatively high efficiencies with little misplacement of fines.5 or 1 mm.150 .09 89. Misplaced fine slimes in circuits such as spirals.25 x 0.08 54. The two devices CHAPTER 5 QUANTITATIVE EVALUATION….68 94.5 mm material was fed to the column together in all tests and the products were then screened to determine size-by-size results.74 59. Rej. Rej.08 72. The values chosen here are those easily achievable if sufficient froth capacity and collector is used. The coarser cut of 0. contribute significantly to a higher product ash.55 90.6 x . Page 80 .25 mm.65 75. Comb.67 Cum.

however. To correct for the fines that bypasses to the coarse stream. Very little if any separation is performed on the ultrafine material finer than about 0. will refer to the probability of the given size reporting to the oversize of a sieve bend or the underflow of a classifying cyclone.045 mm (325 mesh) and thus the ultrafines travel with the water. This is illustrated on the partition curve by the lower tail which levels out at some partition coefficient greater than zero (Rf). Eq. Rf . The bypass amount. The partition curve displays the probability of a particle reporting to either the oversize or undersize flowstream. By removing the effect of the bypass (Rf). Water that reports to the oversize stream is often referred to as ‘by-pass’.Dennis I. A factor. is real and varies for different geometrical designs and circumstances and thus must be accounted for at some point in calculations and comparisons. As shown in Figure 5-4 the probability or partition coefficient is plotted on the Yaxis while the particle size is plotted on the X-axis. Pcci 100 Pi . the partition value or coefficient. one can better see the fundamental classification behavior of a device. Pc. A partition curve best describes the performance of a separating device.  Some water always reports to the oversize flowstream.  The sharpness of separation is determined by the slope of the line on both sides of the d50 and is represented by a slope factor. . As used herein. A common partition curve is ‘S’ shaped and is generally characterized by several parameters:  The d50 is the particle size which has a 50% chance of reporting to either the oversize or undersize portion of the separating device and is found at the vertical middle of the plot. represents the percentage of the ultrafine material that reports to the oversize stream. Phillips nearly always used for making the fines size cuts are classifying (hydro) cyclones or fine wire sieve bends. 5-1 is used to calculate the partition coefficients Pc for the corrected efficiency curve.

R f 1.

Rf CHAPTER 5 QUANTITATIVE EVALUATION…. 5-1 Page 81 . Eq.

00 1 10 100 d50c 1000 Size in microns Figure 5-4. d is the particle diameter.70 0.40 0.  d    .90 0.Dennis I. where  is the slope factor.20 0.00 0. Phillips 1. Characteristic partition curve for classification.80 Actual Partition Curve Pc to Oversize 0.60 0. and d50c is the corrected d50.10 Rf d50 0.50 Corrected Partition Curve 0.30 0. One of the better models developed in recent years for defining fine particle classification is the Lynch Rao model (Lynch and Rao 1975) of Eq. 5-2.

1 exp  d 50 c    Pcc  d    exp  .

If actual data is collected on a classifying device this equation can be fitted to the data by variations in the slope factor and the d50c. Page 82 . 2 exp   d 50 c   Eq. 5-3 (Heiskanen 1993). 5-1 and Eq. CHAPTER 5 QUANTITATIVE EVALUATION…. the true partition efficiency curve can be calculated as in Eq. 5-2 By combining Eq. 5-2.

Phillips   d   .Dennis I.

1 exp   d 50 c     R Pc 1 .

R f  f   d   exp  .

25 mm cut in the case B circuit.7. A portion of the gravity partition curve between the partition coefficients (Pc) 25% and 75% is commonly used to determine the sharpness of separation. Ecart Probable Moyen (Ep). This combined with the classifiers on the feed to spirals provide what is essentially a two stage classifying circuit. Gravity Separations – Spirals and Heavy Media Cyclone (HMC) Spiral and heavy media cyclone performance was predicted from the use of partition curve modeling. it provides for the deterioration in performance that takes place as sieve screens wear. Misplaced fines across a sieve will have an ash value similar to that of that size in the plant feed.150mm while a value of 4 to 6 is common for fine wire sieve bends (Firth et al. A computer program jointly developed by Eric Yan and the author was utilized to quickly predict ash and yield results from the various size feed washabilities.Improved). the misplaced fine material kept the ash value of that size in the feed. classifying cyclone underflow fines may actually have a slightly higher ash due to the higher density of the higher ash particles. This value is given by Moorehead as the most common for a 0. is CHAPTER 5 QUANTITATIVE EVALUATION…. For evaluation purposes. In order to have sufficient and consistent detail in the feed washability. A slope index. the program expands the washability to 360 gravity points by the use a modified clamped spline fit. Partition curve modeling for gravity separations is identical to that of size classification except that the particle density parameter replaces that of the size parameter on the X-axis. In order to evaluate the benefit of just adding more classification circuits. For simulation of the 0. 2   exp  d 50c    Eq.0 was taken from actual plant data collected by Moorehead. a 15-inch diameter classifying cyclone was utilized with an  value of 2. A value of 4. 1995) (Moorehead 1998). Page 83 . On the other hand.5 to 4 is common for a classifying cyclone making a cut near 0. The program then has three models available for determining the actual partition coefficients. Although this comes at additional capital and operating costs there may be considerable benefits in the further removal of fines from the coarser streams.0 was chosen for the sieve bends making the 0.150 mm size cut in the case A circuit.150 mm cut in cyclones. The value of 4. two-stage cycloning was considered for Case A (labeled Case A-Improved) and two-stage sieving for Case B (labeled Case B. Many plants that have considerable amounts of high ash material in the ultrafine size have installed additional classifying devices on the clean coal stream of the spirals. and although it is at the low end of the 4 to 6 range given previously. 5-3 An  value of 2. The lower values represent a flatter slope of the partition curve and thus more misplacement and poorer performance.

of the 25% Pc divided by 2 (Eq.Dennis I. of the 75% Pc and the S.G. 5-4). Phillips defined as the difference of the S. Ep d 75 .G.

.6 Specific Gravity Figure 5-5. which is better than the actual 22% ash reported on this size fraction at the Lady Dunn plant.5 by 1 mm size range as in Case C and D. the same Ep of 0.25 mm.15 mm material since this size is known to perform poorly in a spiral. Heavy Media Cyclone (HMC) Predictions for the heavy media cyclone performance assume that all particle sizes of interest are combined into only one size of cyclone.6 1. % to Clean Coal 100 90 Partition Coefficient 80 70 60 50 40 Ep def ining portion 30 20 10 0 d50 1 1.90 S. an Ep of 0.4 2.G. Although changes in operating CHAPTER 5 QUANTITATIVE EVALUATION….8 2 2.10 S. Characteristic partition curve for gravity separation. for the material coarser than 0. For the processing of the 0. Even with this poor performance the spirals produced 15.G.30 and a gravity cut-point of 2.4 1. As recommended by David Chedgy of CLI Corp.25 x 0.54% ash product on Stockton coal. Page 84 .15 and a lower gravity cut-point of 1.G.2 1.95 S.15 and a cut-point of 1. This is considered good performance for the overall average of a large bank of spirals. Spirals The spirals were modeled using an Ep of 0.2 2. d 25 2 Eq. 5-4 A lower value means a steeper slope and a sharper separation. was chosen due to the narrow size range which is optimum for a spiral. was chosen for the 0.

Phillips parameters will affect cyclone performance the same Ep was used for all cases and thus HMC performance was held constant.Dennis I. However. For a perfect separating device. The higher the Ep.06 was chosen as a typical value for all modeling of the plus 1 mm material in heavy media cyclones.0333  50 c d Eq. The effective incremental ash is then calculated by dividing the difference in the two yield ash products by the difference in yields (Eq. 5-5 is from a HMC handbook developed by Chris Wood (1990a) at the JKMRC center in Australia. with a non-perfect separator. Eq. IncAshEff ( yield 2 )(ash2 ) . The cut-point was then varied on the HMC to produce a constant total product ash for all case circuits. Yield Optimization It has been shown by Abbott (1981) that for yield to be maximized in a multi-circuit plant. indicates that both the 0.G. the more this occurs.. all units should be operated at the same incremental ash value. For the 0.09 Ep for the 0.5 x 1 mm to HMC in Case D. An Ep of 0.075 Ep predicted in Eq. similar to laboratory float/sink. the mean particle size. a small increase in separating gravity not only brings the next neargravity particle but also brings material that is several gravity points away from the cut-point.09 was chosen due to the smaller particle size.71 mm.60 S. This is conservative compared to the 0. 5-5 The Ep increases rapidly below a particle size of 1mm and a table of size-by-size Eps for full scale installations found in Wood (1990b). the incremental ash is defined as the ash of the next lump to float if the separating cut-point is raised by a very small amount. an Ep of 0.5 x 1 mm is typical or somewhat conservative. and d. 5-5 where 50c is 1.06 Ep for the coarser sizes and the 0. Ep d 0. is 0. the cutpoint for the coarser particles. For actual plant separating devices the incremental ash can be obtained by samples at two slightly different cut-point gravities or by simulation. 5-6).

( yield1 )(ash1 ) yield 2 .

For different particle sizes the moisture CHAPTER 5 QUANTITATIVE EVALUATION…. for a plant that is limited by the BTU value of the product. incremental inerts must be held constant. Page 85 . The term incremental inerts is defined as the incremental ash plus the incremental moisture. yield1 Eq. 5-6 The incremental ash concept as defined above applies to any plant in which dry ash is the controlling parameter. However.

however.150 mm (100 mesh). Just as in most operating plants. Case A One of the most significant differences was due to improving the classification at 0. the heavy media cyclone was the only unit that had significant control of the separating gravity.9 tph for the Stockton and a similar amount for the 5-Block seam. For the evaluations undertaken in this work. the incremental ash concept could then be applied between those two devices. CHAPTER 5 QUANTITATIVE EVALUATION….3 Results of Case Evaluations After numerous simulation runs the results were compiled into Table 5-2 and Table 5-3 which show the tons. The evaluation results are discussed case by case. The additional substitutions of column or spirals were set up with the most likely performance results. It was then left to the heavy media cyclone circuit to be adjusted to bring the total plant ash (always dry basis ash) back to the same product spec as in the base case. A target ash of 13% was used for the Stockton seam which compares with that of an existing plant processing this seam. many plants that have a high ash feed are starting to realize the importance of effective desliming and have put additional sieves and or cyclones on the spiral product stream to remove more of the high ash slimes from the product. yield. Page 86 . 5. the resulting yield was just that – a result and not an optimization. the plus 1 mm material were separated in a heavy media bath for the coarser portions and in a HMC for the remaining material. If. the constant incremental ash concept could not be followed in Cases A through D. By using a two-stage classifying cyclone circuit the amount of misplaced raw fines in the clean coal product dropped from 19. In practice. Phillips will obviously vary considerably and results in the need for the finer sized coal to have a much lower incremental ash to allow for the higher moisture values. as that is a typical ash for this seam. This points out the importance of good classification especially when processing coal with high ash fines. In this evaluation the conditions of Case A were taken as the base case and represent an existing operation that is just meeting a target ash spec. and ash results by size and circuit for each case.2 tons per hour (tph) to 5. For this reason. Reducing the amount of this high ash material in the product caused a drastic drop in the total product ash which allowed the yield to increase considerably on the HMC circuit from Case A to Case A-Improved. For the 5-Block seam a target ash of 12% was used. as well as to maintain the ability for direct comparisons between cases.Dennis I.

simply sending material up to 0. Page 87 . Case C Allowing the flotation column to treat the 0.5 mm material rather than spirals caused a decrease in the yield on this size fraction but also dropped the ash.25 mm cut rather than 0.Dennis I. CHAPTER 5 QUANTITATIVE EVALUATION….25 mm to flotation rather than only 0.150 mm results in an overall yield increase of 0.7 percentage points on both seams.5 x 1. Apparently the incremental ash was lower on the HMC circuit and thus the lower flotation ash allowed the HMC cut-point to be raised such that the HMC yield increased by roughly 1.25 percentage points for both seams. Due to similar feed sizes and ashes the results were similar for the Stockton and 5-Block seams.5 mm. Since the HMC represents over 80% of the feed.9 to 1.150 mm the amount of misplaced fine material dropped from 19.25 x 0.2 to 9.4 tons for the two-stage sieve (Case B-Improved). Case D Removing the 0.5% on Case C over Case B-Improved. Phillips Case B Case B illustrates the advantages of making a coarser size cut for sizing the feed to flotation. By making a 0. the total yield for both seams increased by approximately 0. The major advantages to Case D circuit is (i) the ability to separate at a lower gravity cut-point (ii) having only two cleaning circuits in the plant and (iii) no fine slimes problem since the smallest size cut made is 0. The main cons of this circuit is the increased difficulty of rinsing the magnetite from this finer size material.5 x 1 mm material is being treated by HMC quite successfully in many plants but it does require more screens and an increased media recovery circuit. this is still 2.0 mm material from spirals and putting it into the HMC circuit provided an overall yield increase of ~0.0% for the 5-Block seam.85% for the Stockton and 1.5 tph of additional clean coal. Even with two-stage classification. Although relatively small compared to the increases for the other Cases. The 0.2 tons for a one-stage sieve (Case B) and from 5.

00 Ash % 12.95 0.6 88.51 0. Stockton Seam Case A Column Spirals Spirals Misplaced HMC Size -0.00 11.00 8.25 x 0.3 19.2 1000 Feed Tons 93.95 0.2 1000.8 19.5 mm HMC +1 mm Total Case D Size Column -0.150 to .3 1.250 to 1 mm Misplaced -0.00 12.7 67.00 62.9 12.150 mm HMC +1 mm Total B-Improved (Less Misplacement) Column -0.250 to 1 mm Misplaced -0.00 13.82 100.54 13.00 71.15 1.82 63.90 13.9 498.4 816.3 5.00 15.23 Yield % 54.250 mm Spirals 0.25mm 0.2 33.250 to 1 mm Misplaced -0.8 Clean Tons 51.150 mm HMC +1 mm Total Case C Size Column -0.1 639.4 598.82 100.250 to 1 mm -0.5 16.00 Ash % 12.8 16.68 13.12 13.150 mm +1 mm Total A-Improved (Less Misplacement) Column -0.5 to 1 mm Misplaced -0.43 13.8 1.536 Ep d50 0.7 0.44 0.6 26.54 13.632 Yield % 52.70 61.30 0.90 42.2 0 816.00 13. Case evaluation results.06 1.6 67.572 Ep d50 0.8 5.09 0.5 mm Spirals 0.5 622.2 Clean Tons 51.95 0.00 0.25 x 0.30 63.2 1000 Feed Tons 95 43.25 mm Spirals 0.2 816.4 511.2 1000 Clean Tons 30.150 to .7 67.2 1000 Feed Tons 95 43.25 mm Column 0.00 Ash % 12.5 x 1 mm HMC +1 mm Total Feed Tons 59.3 88.25 mm Column 0.06 d50 1.32 75.0 CHAPTER 5 QUANTITATIVE EVALUATION….00 Ash % 12.30 0.00 13.8 67.0 Feed Tons 72.90 42.2 816.2 1000.00 8.6 Clean Tons 37.00 74.86 75.90 42.9 816.623 Ep 0.90 Ash % 12.2 816.597 Ep d50 0.9 12.250 mm Spirals 0.10 1.32 75.90 42.15 2.15 1.3 9.150 mm Spirals 0.00 15.86 Yield % 54.00 11.00 12.2 495.8 88.0 Feed Tons 85.00 13.82 100.82 100.00 57.06 1.32 60.00 77. Phillips Table 5-2.150 mm HMC +1 mm Total Case B Size Column -0.10 1.00 64.4 618.0 524.150 mm 0.00 60.08 62.00 Ash % 12.2 0.75 1.3 Clean Tons 46.86 Yield % 52.00 12.6 0 45.Dennis I.2 529.0 32.39 59.46 13.00 77.6 45.06 1.63 63.32 60.06 1.8 636.2 468.15 2.55 13.5 mm HMC 0.00 0. Stockton seam.5 mm Misplaced -0.3 88.08 Yield % 54. Ep d50 0.06 1.574 Ep d50 0.95 0.6 Clean Tons 50.6 26.15 1.00 61.2 630.90 0.8 9.26 64.00 Page 88 .86 75.32 13.62 Yield % 54.

79 Yield % 52.250 mm Spirals 0.150 mm Spirals 0.950 0.5 mm Spirals 0.950 0.0 94 71.9 478.3 0 0.250 to 1 mm Misplaced -0.61 58.00 72.41 8.5 94 71.25 mm 0.47 19.00 59.32 0.06 1.150 mm HMC +1 mm Total B-Improved (Less Misplacement) Column -0.9 CHAPTER 5 QUANTITATIVE EVALUATION….42 56.1 800.150 mm 0.55 100.9 453.00 12.1 19.150 to .0 21.631 Ep 0.0 Feed Tons Clean Tons 96 54. Phillips Table 5-3.9 0 0.00 11.00 14.639 Yield % 52.6 41.3 51.94 12.15 1.16 12. 5-Block seam.40 75.38 12.1 16.7 30.250 to 1 mm Misplaced -0.75 75.9 451.06 d50 1.00 56.06 1.30 0.0 94 71.950 0.586 Ep d50 0.900 0.4 1000 587.50 Yield % 56.250 to 1 mm Misplaced -0.3 42.00 Ash % 11.3 800.00 Ash % 11.43 41.70 11.88 12.00 Ash % 11.9 Feed Tons Clean Tons 78.150 to .100 1.549 Ep d50 0.00 0.0 19.3 31.150 mm +1 mm Total A-Improved (Less Misplacement) Column -0.40 75.8 1000 567.00 57.40 61.43 41.3 51.28 60.9 464.30 0.3 1.2 21.0 Feed Tons Clean Tons 105.25 x 0.70 10.9 34.41 8.95 59.9 427.09 0.750 1.50 Yield % 56.8 58.85 0.00 53.588 Ep d50 0.0 1.0 9.8 1000 601.70 59.9 482.96 Yield % 56.25 mm Spirals 0.15 2.607 Ep d50 0.25 mm Column 0.150 mm HMC +1 mm Total Case B Size Column -0.75 75.86 12.5 mm HMC +1 mm Total Case D Size Column -0.96 60.3 31.4 1000 585.1 59.40 61.7 28.00 Ash % 11.00 56.5 x 1 mm HMC +1 mm Total 5-Block Seam Feed Tons Clean Tons 64.06 1.25 mm Column 0.1 9.6 800.4 5.06 1.0 800.0 42.1 1000 595. Case A Column Spirals Spirals Misplaced HMC Size -0.1 59.1 94 71.15 1.15 12.40 75.1 800.69 15.47 12.70 9.250 to 1 mm -0.43 41.6 Feed Tons Clean Tons 105.950 0.1 16.69 15. Ep d50 0.150 mm HMC +1 mm Total Case C Size Column -0.43 41.00 Ash % 11.47 19.25 x 0.19 Ash % 11.55 100.15 1.5 to 1 mm Misplaced -0.4 800.55 100.40 75.41 15.0 5.250 mm Spirals 0.70 10.41 15.06 1.5 mm HMC 0.5 mm Misplaced -0.77 Yield % 56. Case evaluation results.1 1000 599.55 100.00 0.15 2.100 1.00 66.7 Feed Tons Clean Tons 103.00 Page 89 .Dennis I.36 58.

Dennis I.23 23.000 per year available to support it plus the capital savings of not building the spiral circuit.2 63.844.848. Compared to the gain over the base case.67 million additional.2 3. the yield improved with each case up to a yield gain of 4% for the Case D circuit (Stockton seam).67 % $2.86 - - 622.39 % $1.668.7 -0.000 -3.37% -$444.6 A-Improved Case Add’l Tons Add’l Yield 59. Table 5-4.08 32. Since the first step taken by a plant with considerable high ash fines should be to improve the size classification methods for 0.512. Sending up to 0.37 point drop in yield from Case A-Improved to Case B. Stockton seam.8 63. Although sending the 0.5 0. The exception to this is the 0. As shown in Table 5-4 this change alone is worth a projected $2. Gain Over Case AImproved (Less misplacement) Additional Additional Annual Annual Add’l Add’l Revenue Revenue Tons Yield @ $20/ton @ $20/ton (6000 hrs) (6000 hrs) - Gain Over Case A Clean Tons Per Hour Yield % A 598.0 63.7 1.4 4.5 x 1 mm material to HMC (Case D) will cause some additional capital costs and additional magnetite consumption.000 C 636.000 16.0 2. Compared to Case A-Improved sending a well sized 0. the gains of using coarser size column flotation is also compared to the Case A-Improved.150 mm.90 40. Compared to the poor classification of Case A the circuit changes could produce up to $4.25 mm topsize to columns (Case B-Improved) rather than a 0.000 13. This shows that a good classification (two-stage cyclones) at 0. there is $336.8 million in additional revenue.7 2.25 mm. The additional revenue for the additional tons produced by each circuit change was calculated assuming a $20 per ton net realization and 6000 hours per year of actual production.00 % $2.37 % $2.6 3. Page 90 .85 points or over $1 million.150 mm desliming. Since most plants with serious fine desliming problems would first make an effort to make a better cut at 0.86 20.22 % $3.000 CHAPTER 5 QUANTITATIVE EVALUATION….5 mm to columns rather than spirals creates a total of $1.004.150 mm topsize results in a yield gain of 0. The results of the comparisons are very similar for both seams and thus to reduce confusion only the Stockton seam will be discussed here.85 % $1.76 % $4.150 mm is better than a poor classification at 0.8 million.000 D 639.020.864. Phillips General The gain from changing from the base case (A) to each of the case circuits is shown in Table 5-4 and Table 5-5.000 B-Improved 630.62 37. Case evaluation summary.400. the Case A-Improved is probably the best case to use as a basis for comparison.000 - - - B 618.9 1.3 62.06 % $4.000 8.6 61.

000 B-Improved 595.9 A-Improved Case Add’l Tons Add’l Yield 56.080.8 1. Although construction cost estimates are not given here.1 2.252. 5-Block seam. Page 91 .000 Since the cost of a basic flotation circuit is already in the base case (Case A). Gain Over Case AImproved (Less misplacement) Additional Additional Annual Annual Add’l Add’l Revenue Revenue Tons Yield @ $20/ton @ $20/ton (6000 hrs) (6000 hrs) - Gain Over Case A Clean Tons Per Hour Yield % A 567.4 Conclusions The simulations and the resulting evaluation summary has shown that nearly each level of circuit change can make a significant improvement in coal yield.200. Any revenue realized from the sale of this additional coal is pure profit.50 17.71 % $2.804.50 27.40 % $4.376.96 31. 5.000 - - - B 587. The coarser sizes have a very high carrying capacity in columns and thus only a few additional columns would be needed.0 1. Phillips Table 5-5.77 19.98 % $2.0 59. Case evaluation summary.9 1.0 58.6 59.27 % $324.0 3.19 34. a coarser size cut for separating the fines from the next coarser size cleaning device provides additional yield by eliminating most of the misplaced high ash fines and CHAPTER 5 QUANTITATIVE EVALUATION….1 1.7 58.7 0.9 60.052.000 10.000 D 601.7 3.6 1.69 % $2.00 % $1. the cost of adding coarser material to the flotation circuit would be relatively small.79 - - 585.000 2. The simulation results presented here represent a realistic and somewhat conservative approach to show the advantages of making changes in the traditional plant flowsheet.000 16. The two major advantages are: 1.752.46 % $1. This additional coal comes with no additional cost for the coal since it was already mined.71 % $3. The dewatering capacity for the additional floated coarse material is already there in the spiral circuit product stream.000 14.Dennis I. the author’s experience indicates that the payback should be less than 1-2 years for any of the case circuits.17 % $3.000 C 599.028.

150 mm) allows the column to perform the desliming and ash rejection while simplifying the circuit.Dennis I. Page 92 . the use of a flotation column on coarser sizes than traditionally thought (+0. CHAPTER 5 QUANTITATIVE EVALUATION…. Phillips 2.

of Queensland. Implementation of MicrocelTM Column Flotation for Processing Fine Coal.5 References Abbott. C. CHAPTER 5 QUANTITATIVE EVALUATION…. Challenges of Coal Preparation. Proc.C.. 250-276. Jr. KY. Mudgee. 1981. T. Doctoral Dissertation. A Performance Model for Coal-Washing Dense Medium Cyclones. M. Coal-Washing Dense Medium Cyclones. The Optimization of Process Parameters to Maximize the Profitability From a Three-Component Blend. Van L. pp. R. pp 246-269. Australia. Newcastle. Lexington. 1990a. A. 238-249. pp. Coal Prep 93. J... Handbook for Operators and Plant Designers. pp. Personal correspondence. pp. 1993. 80. Phillips 5. O’Brien. Wood. B. 11th Int. Cagliari. Fonseca. Conference. Lynch. 87-106. Page 93 . and Rao. Univ. 828-834.Dennis I. Chapman & Hall. of Queensland. Seventh Australian Coal Preparation Conference.. A. C. Mining Engineering. Australia.. The Impact of Fine Classification on Coal Preparation Performance. Krebs Engineers.. 1995.. K... Julius Kruttschnitt Mineral Research Centre. pp.. 1990b. Mineral Processing Congress. 74-77. 1998. 1975.J. Firth.. pp. Proc. D. 17. C. Wood.G. Clarkson. Moorehead. Modeling and Scale-Up of Hydrocyclone Classifiers. Univ. 1993. Davis. Australian Coal Preparation Society. Heiskanen. Particle Classification.. Proceedings First Australian Coal Prep. pp. 1995. Australia. Edward.

5 mm in size can be processed successfully in a column flotation cell. The addition of diesel for increased recovery is not new information but when combined with the use of wash water in a column. The size-by-size analyses of the pilot. results in excessive entrainment of high ash fines.045 mm (325 mesh) particles while also indicating a low ash comparable to that of the slightly coarser sizes in the froth concentrate. Together these two results indicate excellent removal of only the high ash clays. Phillips CHAPTER 6 CONCLUSIONS The work presented here has significantly increased the understanding of coal processing. This is especially true at the higher feed ashes due to the need to float coarse coal with high air and frother rates which. 3.  Factors that affect the amount of bubble surface area such as air rate and frother dosage. The major findings of the present investigations and their potential contributions to obtaining optimum processing of 1 mm x 0 coal are: 1. from the froth concentrate. Any truly non-floatable material has no effect on the froth loading or column capacity. and not fine coal. Smaller bubbles have more surface area and also produce a wetter froth which is less likely to collapse from the coarse particles. A flotation column properly utilizing wash water is the best desliming system for removing high ash clays. 2. Detailed pilot-scale work showed that successful coarse coal recovery in a flotation column is sensitive to several parameters:  The foremost factor is the loading of floatable material.Dennis I. one can minimize what is normally a considerable increase in product ash.and full-scale test results show good recovery of the minus 0. With proper attention to operating parameters. The bubble surface area only builds one layer of particles and then the froth is considered overloaded and any additional coal is dropped back into the pulp zone. in conventional cells. CHAPTER 6 CONCLUSIONS Page 94 . coarse coal up to at least 0. A column can float coarse coal as well or better than conventional flotation cells due mainly to the use of wash water in a column.  Proper dosage of collector (diesel) can increase coarse particle recovery but at a slight increase in fine particle ash.

particle size. The top size for coal flotation in any system should be at least 0. The simplified model developed in this work shows that particle shape has more affect on coarse particle detachment than was previously thought. A major benefit of making a coarser cut is less misplaced high ash slimes to affect the other circuits. The size-by-size recovery plots from the pilot-scale and fullscale testwork (Figure 3-12 and Figure 3-21) show that the 0.150 to 0. Through the work of Chapter 2 it was concluded that the contact angle.150 x 0. This undertaking was fortunate in that it was able to proceed from the concept through lab testing to pilot-scale and then to full-scale commercial operation. Contact angle is also important and can be altered to a limited extent.1 mm are not at all spherical.150 mm material. If the flotation feed upper size cut were increased from 0. and particle shape are the major determinants for particle detachment. Scale-up to a large flotation column has proven successful. From the case examples it was shown that an increase of almost 1 percentage point in yield ($1 million additional revenue on a 1000 tph plant) can be realized by increasing the top size of flotation feed from 0. Most models are for spherical particles but in actual plant practice nearly all particles > 0.5 mm.7 percentage points in yield ($2 million) could be realized. It is the author’s belief that it is better to process the fines in column flotation and then set up a screen bowl centrifuge to classify out the amount of ultrafines necessary to meet the moisture specification. Phillips 4. 6. 5. The original concept of this work was to improve fine coal processing by making a coarser top-size cut for the feed to column flotation.25 mm fraction generally floats as well or better than the minus 0.25 mm (60 mesh) rather than 0. A screen bowl CHAPTER 6 CONCLUSIONS Page 95 .150 mm(100 mesh).Dennis I.25 to 1 mm range are relatively flat with considerable perimeter length. It is acknowledged that the ultrafine coal recovered when using flotation for all the fines can result in a high product moisture and therefore many plants prefer to discard the ultrafine material rather than process it.150 to 0. This is especially true for coal which has particles that are either cubical or rectangular overall and have edges varying from sharp to rounded. fine ash reports to the plant product and relatively coarse low ash coal is discarded unnecessarily.25 mm and thereby reducing the amount of high ash fines misplaced to clean coal. due to the difficulties in making an accurate ultrafine size cut. 7. misplacement of high ash fines to clean coal would be nearly eliminated and an increase of up to 1. Considerable economic benefits can be realized by increasing the size at which the finest size cut in a plant is made. However. The force required for detachment varies in direct proportion to the perimeter of the flat projection of the particle but varies only by the sine of the contact angle. The good news is that flats remain attached better than cubes or spheres and most coal particles in the 0.

the centrifuge overflow ports are relatively easy to adjust and thus recovery can be optimized frequently. It is site-specific but the increased cost of more flotation should be offset by the yield gain from reduced misplacement and reduced classifying costs.Dennis I. Phillips centrifuge can classify ultrafines better than a cyclone due to the high G-force as well as there being only particles of similar density. Although not adjustable on-line. CHAPTER 6 CONCLUSIONS Page 96 .

It would also be helpful to study the frothing capabilities of the centrifuge main effluent in terms of residual frother and solids. Although floating material this coarse will. 4. 3. be much more delicate. at best.5 mm but some results showed potential up to 1 mm. An improved understanding of the classifying capabilities of screen bowl centrifuges would be beneficial to those who want the product moisture reduction by discarding ultrafines through the centrifuge and also to those wanting maximum solids recovery. Is it possible to develop a similar technique that also looks at flotation by size and thus allows prediction of full-scale column carrying capacity? 5.Dennis I.5 mm. CHAPTER 7 RECOMMENDATIONS FOR FUTURE RESEARCH Page 97 . and labor intensive than continuous column testing. The ability to more accurately predict carrying capacity and coarse size recoveries from 2-inch column tests would be helpful. The release analysis technique of flotation prediction is much less time. The opportunities to compare 2-inch and full-scale columns are improved as more full-scale columns are installed to process coarser coal. there could be many things learned that could be applied to full-scale flotation of smaller sizes such as minus 0. further research in the following areas is recommended: 1. material. It is known that there is a relationship between diesel and frother in flotation and it was noticed in this study that excess diesel reduced the air fraction in the column due to its affect on frother and thus bubble size. 2. An in-depth in-plant study of the relationship and thresholds of the diesel frother relationship and the resulting effect on coarse coal recovery would be very helpful to operators planning to float coarser coal sizes. It is suggested that the control parameters highlighted in this work be utilized to explore column flotation up to at least 1 or even 2 mm. The coarse coal flotation undertaken in this study concentrated on floating the material up to 0. Phillips CHAPTER 7 RECOMMENDATIONS FOR FUTURE RESEARCH On the basis of the present study.

Phillips 6. Last but not least is the need for affordable means of breaking the strong downstream froths created due to the use of the stronger frothers in columns. The ultrafine coal that is in the main effluent of the fine coal centrifuges only adds to the stability of the froth due to its fine particle size.Dennis I. CHAPTER 7 RECOMMENDATIONS FOR FUTURE RESEARCH Page 98 .

Dennis I. Phillips APPENDIX A TABULATED PLANT TEST DATA APPENDIX A TABULATED PLANT TEST DATA Page 99 .

7 2.0 27.6 1.47 0.56 94.58 54.24 0.01 17. Rec.0 14.72 8.6 11.37 55.13 4.0 42.93 52.96 60.86 41.62 2.0 21.59 42.7 0.0 416 446 432 442 442 481 551 544 544 544 1.6 6.34 0.79 11.0 14.6 9.13 35.20 40.8 18.6 4. Tails % Wt.1 2.00 No Samples Analyzed 39.22 41.70 1.54 4.61 APPENDIX A TABULATED PLANT TEST DATA Comb.06 96.75 11.58 69.55 39.70 64.0 14.73 4.21 33.41 2.8 2.05 95.38 91.31 2.17 Frother Collector (Diesel) (ml/min) Test Series 101-110 (Cyclone Overflow) Operating Parameters Wash Air Air Air Air Water Meter Rate Rate Fraction % in H2O lpm (cms) 14.08 64.79 63.38 47.52 58.4 0.79 78.28 7.2 10 8 Collector Froth Depth (inch) (g/T) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Page 100 .54 10.17 46.4 10.00 0.Dennis I.6 1.83 4.67 0.22 6.6 7.0 14.86 92. Phillips Test # 101 102 103 104 105 106 107 108 109 110 Test # 101 102 103 104 105 106 107 108 109 110 Feed Rate (gpm) 60 60 60 60 44 80 110 100 100 100 Feed Rate (kg/min) 9.82 94.92 2.66 3.67 63.0 14. Sep Eff.4 9 9.48 18.76 13.5 25.35 1.5 1.22 34.84 4.8 19.45 5.45 39.00 21.6 0.59 56.45 Conc Conc Rate Rate (kg/min) (Tph/m2) 5.43 2.72 64.85 56.01 88.73 55.0 24.11 51.29 20.94 49. Yield 41.6 Feed Ash % Conc.49 36.0 2.0 4.5 2.86 71.10 2.07 35.44 18.0 27.87 3.8 27.35 0.51 36.6 1.4 3.15 5.9 16.0 32 19 23 24 22 18 19 15 ? 15 (ml/min) ? ? ? ? ? 8.59 5.80 64.41 26.60 0.10 6.95 3.% 4.0 24.3 7.67 95. % Ash Rej.44 46.0 2.61 44.0 32.0 41.0 12.65 38.89 6.18 4.72 59.00 1.05 91.75 82.64 40.76 5.2 13.0 2.6 1.82 6.% Comb.0 41.68 24. Test Series 101-110 (Cyclone Overflow) Test Results % Solids Tails Feed Conc.0 41.69 31.76 44.0 12.

21 50.78 0.20 61.76 7.045 mm x 0 Cumulative Feed Concentrate Wt.20 100.17 61.64 55.03 31.05 0.15 54.12 37.24 100. Rec.77 22.09 4.045 mm x 0 Cumulative .69 43.93 61.150 .85 31. % Ash % Cum Wt.96 30. Sep Eff.83 17.73 95. Ash Rej.38 94.70 59.56 61.79 5.66 17.44 48.04 5.90 39.46 3.69 APPENDIX A TABULATED PLANT TEST DATA 6.18 47.93 25. % Ash % Cum Wt.67 32.74 100.40 90.19 56. Rec.02 58.86 40.11 12.95 3. 1.93 7.045 0.44 63.44 3.83 71. F.19 47.00 95.95 34.38 71.66 7.11 41.00 61.61 78.70 83.32 24.25 78.15 44.00 3.20 1.300 mm . From Bottom Wt.80 55.30 100.00 0.5 mm . Ash Rej.72 5.09 0.72 5.82 94. Rec.75 6.87 95.61 94.150 x .64 94.16 60.34 17.80 Cum.77 77.150 x .03 0.14 20.14 16.24 2.69 33.31 100.09 55.50 76. Sep Eff.67 41.92 100.18 66. From Bottom Wt.18 63.300 x .16 56.82 Cum.84 6.43 62.73 95.57 Page 101 . Yield C.35 75.75 13.70 5.16 38.87 48. Ash Rej. % Ash % Cum F.10 35.49 40.34 49.86 39.70 6.38 42. Phillips +. Yield C.40 17.Ash Sep Eff.5 x . Ash Rej.17 5.24 54.43 73.38 83.29 63.52 23.00 3.22 17.51 50.Ash Sep Eff.63 94.47 3.62 40.48 60.73 5.05 47. 38.Dennis I.14 62.69 18.49 40. 0.Ash F.48 38. Yield C. % Ash % Cum Wt.90 0.78 40.88 19.14 6.150 .90 48.98 62.57 42.Ash F.81 34.48 60. 30.25 58.18 94.99 13.00 7.Ash Test 110 Tails Individual by Size Wt.44 58.300 mm .36 4.96 47.04 10. Yield C.72 5.300 x .00 1.38 9.16 94. % Ash % Cum Wt.89 66.49 55.5 x .18 54.31 30.53 55.50 26. Rec.35 58.85 9.82 94.17 37.16 3.22 7.69 6.51 40.045 0. % Ash % Cum F.87 98.57 30.82 64. F.Ash Results by Size Test 104 Tails Individual by Size Wt.64 55.61 Feed Concentrate Wt.

% 49.63 47.03 73.35 53.23 76.48 89.92 5.24 72.63 Page 102 .26 Ash Rej.41 68.33 34.71 53.98 78.78 79.65 68.02 8.20 72.09 55.59 5.37 69.04 87.99 69.54 5.44 10.72 1.99 75.61 35.08 87.91 7.31 2.31 10. Phillips Test Series 201 .60 86.69 85.41 42.28 63. 7.31 13.55 7. Sep Eff.75 2.25 88.84 37. Rec.63 9.02 63.79 51.10 49.88 74.31 88.20 3.97 8. % 44.43 39.10 36.03 67.07 82.35 91.74 Comb.55 60.46 56.62 8.21 93.08 1.94 9.58 8.Dennis I.27 2.81 43.45 6.87 2.04 2.47 48.50 3.08 7.05 64.46 62.75 4.48 9.42 2.69 9.14 11.98 9.06 88.05 5.46 33.66 46. Add Push Water 9.50 Ash % Conc.66 3.13 82.70 4.52 9.30 72.67 34. Yield 28.56 4.71 93.22 35.19 67.00 33.27 68.99 4.96 3.64 63.215 Test Results Test # 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 Feed 46.59 78.04 66.14 70.14 Tails 62.65 35.14 88.70 70.76 10.09 7.06 15.96 5.27 63.20 65.96 81.91 68.40 % Wt.% 95.24 55.94 58.37 Comb.55 42.82 87.22 2.49 66.52 33.18 50.40 35.68 APPENDIX A TABULATED PLANT TEST DATA Feed 5.19 5.35 57.15 74.78 4.61 6.95 78.44 5.88 6.97 72.56 54.10 9.37 40.34 57.57 88.71 4.83 7.05 % Solids Conc.51 37.00 4.55 63.49 60.96 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Tails 3.85 7.55 8.34 8.82 8.

6 17.1 15.3 18.3 10.0 20.0 18.2 22.9 14.0 20.3 21.0 18.7 APPENDIX A TABULATED PLANT TEST DATA 6.9 9.1 11.0 7.2 8.6 10.8 22.9 9.5 9.8 7.0 20.6 19.6 11.4 18.4 10.1 11.6 21.2 7. Phillips Test Series 201 .2 10.2 15.2 12.Dennis I.5 21.215 Operating Parameters Test # 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 Feed Rate (gpm) 100 100 120 80 60 80 100 115 100 100 100 100 100 100 100 Feed Rate kg/min Conc Rate kg/min 21.4 21.7 19.2 18.5 18.4 8.4 11.8 10.5 2 2 2 2 2 Air Fraction % 16 23 19 17 23 19 17 18 17 12 17 17 17 17 17 Frother (ml/min) 8.7 11.0 20.3 23.2 9.8 7.0 18.8 10.6 11.0 Air Rate (cms) 2 2 2 2 2 2 2 2 2 1.4 13.6 9.3 8.6 8.2 Collector (Diesel) (ml/min) 0 0 0 0 0 18 18 18 18 18 18 27 36 18 18 Collector (g/T) 0 0 0 0 0 641 405 624 564 553 641 948 1361 545 598 Froth Depth (inch) 25 25 25 27 27 24 22 24 22 25 29 29 28 31 19 Page 103 .0 Wash Water gpm 12.0 12.0 18.0 12.9 10.0 20.

Ash Rej.49 33.35 49.00 Head Ash 10.75 21.150 x . 57.045 mm x 0 50.04 36.90 52.85 46.61 7.82 60.77 Cumulative 100.28 44. 79.99 65. Ash Sep Eff.64 34.57 5.35 Wt.52 Feed Wt.99 7.75 70.29 10. Yield 93.00 30.97 51.64 81.45 94.38 33.93 68.09 78.97 5.72 91. Sep Eff. 36.Dennis I.58 44.57 95.71 82.71 45.72 48.56 67.46 23.045 6. Yield Cum.85 .62 39.24 57.52 36.23 100.61 Wt.90 69.075 x .Ash 32.86 8.00 4.02 60.09 57. Rej.15 54.75 54.91 51. Yield 60.79 43.93 .46 10.92 85.59 60.92 6. Rec.46 90.Ash F. Rej.70 6. Ash Sep Eff.66 28.23 95.43 60.07 32.76 29.96 64.00 79.32 51.67 80.91 .64 Individual by Size C.22 Cumulative 100.37 60.93 94. Yield 75. Rec.09 80.46 59.07 70. % Ash % Cum Wt.30 7.075 10.075 11.39 89.31 .80 52.71 100.62 100. Yield 41.90 32.10 87.14 64.00 Head Ash 17.Ash F.15 53.87 49.99 4.66 34.Ash F.92 6.70 93. Ash Rej. % Ash % +.Ash F.01 42.81 Individual by Size C. Rec.52 37.00 25.23 9.65 45.66 56.150 mm 30.97 60. Ash Rej.33 63.61 32.94 56.47 63.87 9.78 6.19 42.20 27.02 .11 58. Sep Eff.045 mm x 0 52.99 67.69 8.150 mm 6.33 56.80 14.73 44.99 42.33 53.33 63. 58.43 Results by Size Test 203 Concentrate Tails Cum Wt.55 59.38 33.21 Cumulative 100.05 45.46 60.00 3.02 .75 93.15 49. 37.85 55.29 33.93 55.45 6.34 40.19 6.73 Cum.82 9. Rec. Ash Sep Eff.07 .78 5.98 10.00 13.045 mm x 0 76.45 49.18 78.36 63.33 64.48 100.65 95.99 7.49 43.58 47.045 7.81 35.90 72.11 54.02 70.Ash Yield 42.59 56.150 x .25 7.09 54.82 63.86 APPENDIX A TABULATED PLANT TEST DATA 22.73 45.07 9.79 78.96 57. % Ash % Cum F. % Ash % +.55 70.45 93.53 88.95 84.56 Individual by Size C.65 13.63 46.16 4.65 67.09 30.00 66.37 60.04 16.46 79.Ash F.01 100.35 5.00 Head Ash 9.00 6.99 Wt.27 69.13 79. From Bottom C.075 x . From Bottom C.50 48. Rec.73 54.150 mm 32.78 42.12 Wt.92 4. 44.82 Test 204 Concentrate Tails Cum Wt.15 55.96 Page 104 . From Bottom C.28 Cum.62 25. % Ash % Cum Wt.88 6.78 48.11 10.54 14.75 33.36 6.Ash 33.Ash F.28 9.35 42.46 8.075 x .49 48.46 17.80 35.61 36.36 43.19 8.38 33.23 .92 64.58 93.12 37.65 86.78 Test 205 Concentrate Tails Cum Wt.06 18.59 69.46 33.76 5.78 93. Phillips Feed Wt.40 54.33 94.46 Feed Wt.54 85. % Ash % Cum Wt.25 50.00 48.05 11.36 69.00 62.33 78.70 51. % Ash % Cum F.38 81.34 57.37 81. Sep Eff.95 91.67 7.69 28.29 85.84 100.32 69.150 x . Rec.38 6.37 60. % Ash % +.40 44.56 50.045 6.22 67.29 47.79 81.22 Wt. F.81 88.78 93.075 9. Rej.98 6.49 67. % Ash % Cum Wt.04 78.46 .33 94.04 85.69 49.

25 7.14 .11 0.82 83.25 41.75 48. Ash Rej.21 48.Ash 37.71 31.10 74.68 8.43 50.40 74. % Ash % +.Dennis I. % Ash % +.48 40.55 8.00 Head Ash 12. Phillips Feed Wt.90 81.11 8. Yield 54.27 76.25 88.99 Cumulative 100.46 73.64 15.72 89. Rec.150 4.78 14.71 9.68 2.06 88.47 Cum. From Bottom C.Ash 39.62 8.55 39.91 75.Ash 1.91 Wt.Ash 1.300 mm 0.32 6.28 88.65 32.37 Feed Wt.045 mm x 0 74.97 59.68 76.00 Head Ash 11.72 63. Sep Eff.Ash F.26 48.94 37.26 47.41 8.47 47.86 9.96 8.44 62.21 39.19 69.91 62.00 Head Ash 12.25 70.50 63.04 14.62 Wt.06 87.21 0.69 49.24 .61 8.18 74.27 74. Rec.06 56.82 39.16 63.66 46.13 40.05 52.00 21.150 4.77 69.95 60.78 40.63 Wt.75 8.77 40.300 x .59 64. Sep Eff.Ash 39.92 71.34 100.Ash F.10 Feed Wt.Ash F.88 38.59 72.07 46.62 7.01 69. Rec.02 8.98 41. % Ash % Cum F.045 mm x 0 76. Yield 60. % Ash % Cum Wt.32 47. % Ash % Cum Wt.49 Cumulative 100.39 100.66 40.53 43.41 69.93 45.94 Test 206 Concentrate Tails Wt.24 46.150 x 0.11 88.27 91.04 9.045 15.37 67.71 69.45 8.61 78.48 8.48 8.78 7.69 APPENDIX A TABULATED PLANT TEST DATA 0. Rec.99 75.09 Individual by Size C.92 0.70 59.05 Individual by Size C.25 63.82 77.85 75.52 45.91 0.53 8.67 Cumulative 100.76 77.59 90. 81.80 68.97 Cum.78 73.74 70.23 78.20 Wt.82 87.59 9.56 60.85 70.045 mm x 0 78.90 45.41 8.78 89.76 13.93 86.20 88.11 13.64 39.54 71. 64. Yield 75.53 Page 105 .92 7.85 69. Sep Eff.51 62.24 68.00 6.85 72.27 88. 74. Rec. % Ash % Cum Wt.17 33.35 62.23 76.300 x . Sep Eff.60 56.11 84.66 91.86 68.69 8. Yield 48.74 81.69 47. Sep Eff.86 .00 6.61 Cum F.77 91.28 47.63 2.52 54.92 92.47 50.89 89.51 45.045 17.89 1.52 Wt.84 55.11 5.19 Test 208 Concentrate Tails Wt.25 68. Ash Rej.49 89.72 9.80 Cum F.80 87.64 1.17 54. Ash Rej.27 48. Rec.12 40.38 65. 80.150 4. Sep Eff.07 9. Ash Rej.48 56.24 91.37 36.60 34. % Ash % Cum F.00 4.01 69.21 78. Yield 50.12 62.45 64.93 39.80 81.45 .20 81.150 x 0. Ash Rej.97 59.85 .80 39.18 100. % Ash % Cum F.16 64.62 67.39 68.38 9.71 85.92 71.17 54.39 30.77 5.59 89.045 19.66 39.54 77.46 Individual by Size C.88 36.59 69. From Bottom C.Ash 1.150 x 0.50 Cum.02 13.59 8.53 63. % Ash % +.77 88.77 9.00 31.51 55.77 100. Yield 58.00 21.96 75.300 mm 1.35 Wt. From Bottom C.36 74. 73.300 x .84 41.55 91.91 48.87 100.43 53.97 70.96 Test 207 Concentrate Tails Wt.93 Cum F.35 76.72 8. 63.300 mm 1.99 76.94 91.97 69.86 68.16 7.55 39.28 68.80 81. Ash Rej.59 89.03 100.80 88.63 55.59 84.78 90.27 .

045 20.33 13. Sep Eff.86 88.92 73.00 0.81 80.45 55.Ash 2.37 74.37 42.05 89.00 22.150 x 0.47 48.51 51.33 8.24 100.26 54. Rec.58 85.14 .16 57.83 6.81 Cum F.93 39.76 85.59 Wt.75 89.61 8.Ash 1.Ash 1.16 10.63 6.92 64.68 48.12 6.59 7.52 37.77 53.86 92.24 58.150 5.57 88.150 5.20 43.87 73.Ash F.28 8.70 88. 79. Ash Rej.52 63.14 91. % Ash % +.02 42.00 8.87 37.92 89.34 100.99 69.61 Cumulative 100. Ash Rej.21 44.36 . Yield 66.72 73.58 65. 80.90 33.75 52.150 x 0.71 67.96 89.81 100.45 8.59 50.43 34.16 54. Sep Eff. 68.88 60.69 92.84 7.70 Wt.65 100. Ash Rej.10 . Sep Eff.71 63.66 37.00 Feed Wt. % Ash % Cum F. Sep Eff.98 80.22 85.19 69.13 79.17 84.82 8.81 85.04 91.28 33.54 4.85 78.13 1.22 39.Ash 37.88 67. Yield 50.00 7. % Ash % Cum F. % Ash % +.Dennis I.05 85.90 65.00 6.Ash 33.91 65.300 x .09 51.67 Cum F. Rec.07 67. Ash Rej.59 67.62 42.94 9.96 79.90 Test 209 Concentrate Tails Wt.04 7.Ash F.00 27.72 100.045 mm x 0 73.Ash 34.87 41.46 Wt.78 13. % Ash % Cum F.07 68.63 83.22 75.43 84.76 40.69 13.34 88.45 72.27 Test 210 Concentrate Tails Wt.35 42.44 68.14 Cumulative 100.17 48.64 63.85 8.53 45.14 29.20 42.85 80.62 46.83 67.87 43.45 9.33 67. Rec.30 .68 80.54 9. Rec.53 100.92 64.57 74.01 1.63 APPENDIX A TABULATED PLANT TEST DATA 0.43 35. 74.150 x 0.34 56.19 65.12 88.02 86.045 21.03 Individual by Size C. Rec.05 86.27 68.72 1.20 0.42 72.87 38.12 10.150 5.00 Head Ash 13.58 34.300 x .86 7.67 89.30 68.25 8.44 91.87 67.97 56.32 7.96 88.99 70.83 6.79 86.96 79.11 Test 211 Concentrate Tails Wt.97 71.39 65.24 Wt. Yield 64.51 50.92 45.97 35.45 59.13 59.51 7.90 58.81 34.16 55.39 Cum.54 .12 89.045 mm x 0 71.53 24.25 41.52 63.60 61.69 69.67 35.00 13.90 66.83 Page 106 .78 66.82 71.51 73.00 4.36 39.22 78.22 10.27 48.045 mm x 0 71.34 63. Yield 56.02 59.08 7.85 8.11 68.96 88.61 68.71 Cum.47 70.51 Feed Wt. Rec.300 mm 1.00 Head Ash 12. Ash Rej.79 Wt.76 36. From Bottom C.66 89.57 28. Yield 69.300 mm 1.72 56.61 Wt.300 mm 1.27 69.50 10. From Bottom C.87 88.13 8.26 48. Phillips Feed Wt.89 Cumulative 100.54 .045 21.60 73.65 65.56 74.80 75. Ash Rej.48 34.37 88.300 x .87 0.02 Individual by Size C.57 87.00 8.58 45.34 Individual by Size C. Yield 58. 68. Sep Eff. % Ash % +.49 14.31 0.28 8. % Ash % Cum Wt. From Bottom C.14 7. % Ash % Cum Wt.88 49.90 1.52 Cum F.05 54.30 68.72 Cum. Sep Eff.67 33. % Ash % Cum Wt.96 6.65 46.Ash F.54 8. 74.87 9.

05 68.Dennis I.91 .045 mm x 72.45 9.05 7.83 85.00 7.31 66.58 89.33 Feed Wt. Yield Individual by Size C.72 37.74 91.Ash 1.70 46.20 .150 x 0.59 45.00 9.300 x .59 45.Ash 35. 63.87 63.77 1.03 91.42 0.49 46.79 5.62 8. Ash Rej. Ash Rej.24 82.24 6.300 x .045 21. % Ash % Cum Wt.88 6. From Bottom C.00 74.50 Page 107 .045 22.37 7.69 38.87 88.150 x 0.00 68.24 88.150 5. 82.96 Wt. Rec. Sep Eff.10 8.84 69. Ash Rej.33 35.Ash F.40 100.77 44.07 88.64 83.61 100. From Bottom C.34 70.300 mm 1.18 55.08 17. Yield Individual by Size C.06 70.10 5.13 53.98 57.51 78.18 78.65 73.65 34.12 8. % Ash % Cum F.65 Cum F.150 x 0.81 18.15 86. Rec.15 42.06 31. 82.17 Cum.56 7.86 10.55 66.67 12.84 56.79 0.48 73. Ash Rej.34 100.76 65.66 57.77 56.49 56.43 8.57 87. 70.70 Cum F.43 74.27 .90 70.39 11.08 7.34 6. Ash Rej. Rec.43 79. 78.Ash F. Yield Individual by Size C.20 46.14 70.71 49.01 11.34 7.37 87.33 47.99 100. % Ash % +.07 30.43 13.37 88.14 34.07 66.23 57.63 7.88 66.Ash 34.66 72.33 88.60 76.79 72.47 58. Yield 55.90 33.50 55.56 54.36 11. Phillips Feed Wt.33 35.14 46.70 35.33 91.05 76.65 43. % Ash % Cum Wt.89 85.40 Feed Wt.62 40.07 52.92 43.49 91.31 88. Sep Eff. % Ash % +. % Ash % Cum F.08 7.12 68.300 x .06 74.21 9.07 74.43 64.65 74.43 88.03 90.57 59.78 9.Ash F.27 68.47 Cum.40 34.00 72.07 88.00 72.72 36.15 88. Sep Eff.14 Wt.94 82.03 72.03 72.05 47. Rec.76 91.27 68.70 48.51 90.77 7.300 mm 1.23 44. Rec.78 75.55 Wt.75 .08 65.41 5.41 43. 69.99 Wt.14 Test 212 Concentrate Tails Wt.27 33.34 7.66 0.85 7.66 Wt.63 39.17 66.94 57.84 1. Sep Eff.21 0.32 84.51 67.150 5.00 34.21 34. Yield 58. % Ash % Cum F.045 20.78 90.08 13.49 90.08 70.045 mm x 70.63 77.07 82.06 87.23 70. Rec.47 19.95 56.19 87. Sep Eff. From Bottom C.15 4.34 APPENDIX A TABULATED PLANT TEST DATA 1.07 68.27 77.20 81.61 Test 214 Concentrate Tails Wt.04 Wt.13 90.04 70.20 43.07 78.00 Cum F.98 83.46 77.68 0 Cumulative 100.34 47.89 8.96 74.00 73.86 100.02 5.36 0 Cumulative 100.33 70.00 7.33 62.Ash 1.40 44.12 7.045 mm x 71. % Ash % +.00 7.36 48.14 55.40 Test 213 Concentrate Tails Wt. % Ash % Cum Wt.09 75.Ash 1.85 45.47 . Yield 57.66 100.59 77.300 mm 1.66 Cum.28 7.86 0.04 82.Ash 9.05 0 Cumulative 100.98 87.89 .95 36.150 5.85 7.24 78.57 10.10 2.94 47.63 32. Ash Rej. Sep Eff.20 68.

150 4.Ash F. % Ash % Cum Wt.11 78. Phillips Feed Wt.00 4.97 6.42 3.88 63.77 37.04 Cum.43 84.50 35.99 42.50 Cum F.00 18.05 50.98 63.33 9.06 43. % Ash % +.75 90.51 55. Yield 75.33 33.98 .045 mm x 0 73.08 77.09 Page 108 .91 100.16 9.62 60.150 x 0.48 9.14 43.14 0.99 Test 215 Concentrate Tails Wt.07 60.95 58.30 8.47 69.92 51. Sep Eff. Rec. Sep Eff.16 40.54 65.14 8.89 69.91 Individual by Size C. Ash Rej.78 81.56 90.47 88.35 30.Dennis I.045 20.88 64.29 61.13 25.09 60.94 .32 88.04 50.98 Wt. Ash Rej.84 Wt. From Bottom C.14 8.88 69.65 59.46 60.85 4.300 mm 0.51 72.83 61. 72.97 72. 77.00 8. % Ash % Cum F.32 88.59 APPENDIX A TABULATED PLANT TEST DATA 1.99 35.59 8.01 43.Ash 1.27 43.04 83.Ash 35.94 Cumulative 100.38 59. Yield 50.70 100. Rec.300 x .07 32.57 12.32 69.62 88.19 8.24 10.47 72.

% Ash Rej. Phillips Test Series 301-302 Test Results Test # Feed 301 302 Ash % Conc.37 46. Rec.08 1.5 Conc Rate (Tph/m2) Wash Water 1. Sep Eff.23 2.64 APPENDIX A TABULATED PLANT TEST DATA 14.4 14.5 7 Collect (Diesel) (ml/min) 0 25 Coll.0 2.56 Comb.25 12.55 N N Tails % Wt.17 5.43 45.3 30.2 12. Add Push Water Feed 55.47 Tails % Solids Conc.% 35.88 40.Dennis I.39 92.76 9.20 40.% 52. 38.4 Air Rate lpm 544 544 Air Rate (cms) 2. Yield 3.0 Air Fraction % 22 18 Frother (ml/min) 10.50 Comb.8 Conc Rate (kg/min) 8.82 Test Series 301-302 Operating Parameters Test # 301 302 Feed Rate (gpm) 100 100 Feed Rate (kg/min) 23.84 59. (g/T) 0 579 Feed Sump (gpm) 110 110 Froth Depth (inch) 20 26 Page 109 .52 59.91 10.97 7.53 87.23 8.

29 60.50 48.18 Page 110 .44 14.045 0.18 13.64 49.79 49.24 100.84 98.66 47.38 9.15 1.40 61.Dennis I.51 10.66 33. % +. Sep Eff.39 41. % Ash % Cum Wt. From Bottom Ash % Cum Wt.25 100.19 35.35 55.25 4.64 39.Ash 18.150 .Ash F. Ash Rej.55 60.14 70.02 42.22 92.83 4.23 63.36 9.21 38.57 11.34 35.00 17.54 53.34 52.30 89.06 42.09 91.99 Test 302 Wt.93 91.48 54.31 100.66 89.62 56.82 12.06 10.80 23. % +.35 64.70 38. Rec.26 35.045 mm x 0 Cumulative 3.55 50.78 8. Sep Eff.53 6.20 19.00 Results by Size Test 301 Feed Concentrate Tails Individual by Size Cum. Yield C.Ash F.13 72.03 19.26 8.02 63.84 52.01 26.14 55.47 62.13 62. % F.98 66.52 6.49 85.51 89.24 8.21 56.32 100.300 x . Sep Eff.22 52.13 58. From Bottom Cum Wt.35 7.15 20.76 45.46 92.70 42.150 .00 3. F.Ash 16.300 mm . Rec.23 89.36 66.95 47.01 66.36 46.35 89.88 39.35 13.83 48. % Ash % Cum Wt.11 22 63. Wt.91 28. Yield C.06 APPENDIX A TABULATED PLANT TEST DATA 4.90 39.46 33.79 35. Rec.25 32.150 x 0. Rec.45 100.26 89.21 40.31 60.23 37.00 4. Yield C.48 30.61 11.88 70. Ash Rej.01 10.12 4. Phillips Wt. Ash Rej.Ash F.78 30.76 64.28 95.64 8.37 35.00 Tails Ash % Individual by Size Cum. Ash Rej.89 2.53 92.11 24.75 10.150 x 0.38 56.18 47.89 38.300 x .300 mm .54 10.85 7.00 28.045 0.69 39.45 92.44 58.88 38.76 90. Sep Eff.50 35.90 58.02 62.46 92.51 58.58 24.045 mm x 0 Cumulative 3.99 45.85 27.95 72.15 39.21 Concentrate Cum Wt.59 54.12 13.54 10.20 7. F.75 57. % Ash % Cum Wt.42 21.76 92.52 22.47 16.47 48.61 40.60 18.83 54. Yield C.16 29.Ash 39.00 Feed Ash % 16.49 49.01 100. Wt.01 16.

80 9.09 10.41 54.96 71.95 15.10 43.53 13.64 11.96 7.41 37.59 71.85 41.71 13.07 12.88 5.50 39.34 68.68 12.98 57. % 44.14 90.69 14.66 55.% 60.65 86.13 13.06 45.62 47.05 Page 111 .73 Tails 5. 13.73 6.02 7.86 8.95 32.60 4.98 Comb.59 55.14 50.54 60.79 17.37 50.65 39.81 32.07 51.62 5.09 27.47 9.15 70.23 11.48 71.62 89.93 14.84 51.19 42.98 91.83 64.67 94.39 30.20 94.65 46.38 7.70 8.43 37.57 52.95 12.40 41.40 9.93 8.40 15.28 37.38 93.22 50.14 42.21 29. Rec.85 % Wt.29 24.02 43.84 59.30 37.52 9.84 9.99 44.36 84.24 57. Sep Eff.82 50.48 51.53 41.50 86.37 64.00 37.97 10.98 10. Phillips First Parametric Series (401-416) Test Results Test # 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 Feed Total Ash % Conc.33 9.51 69.18 58.44 % Solids Conc.22 13.62 12.% 84.91 34.71 61.19 91.80 16.51 21.73 92.55 52.66 42.60 17.61 45.67 36.08 Ash Rej.19 Comb.Dennis I.88 6.64 89.52 6.64 11.91 12.68 44.63 16.18 87.76 47.42 18.65 57.69 34.62 11.06 33.10 38.62 10.00 12.99 92.73 10.87 9.70 12.07 40.48 8.11 17.51 10.67 8.06 9.77 11.03 8.04 10.79 38.22 APPENDIX A TABULATED PLANT TEST DATA Feed 7.82 28.00 5.21 41.36 36. Tails 34.70 87.38 45.30 66.00 5.75 48.42 55.76 38.73 7.11 72.47 79.91 39.46 8. Yield 45.65 58.92 40.58 58.

6 13.84 1.29 1.2 15.00 1.6 24 10 10 11 11 11 10 10 8 8 8 30 30 30 30 48 48 30 15 18 30 30 40 30 30 45 45 Coll.05 1. Phillips First Parametric Series (401-416) Operating Parameters Test # 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 Feed Rate (gpm) 100 100 100 80 80 60 60 80 100 100 90 103 80 80 80 95 Feed Conc Conc Conc Rate Rate Rate Rate (kg/min) (kg/min) (Tph/m2) -.9 44.0 1.54 1.61 1.6 1.0 12.60 1.5 mm 31.2 10.74 1.6 1.61 1.6 1.3 27.8 22.6 11.6 1.76 1.6 11.59 1.5 42.5 mm (g/T) 920 961 696 552 1805 753 707 254 445 745 530 1007 758 1249 1521 2020 Feed Sump (gpm) 110 110 104 104 104 104 104 105 105 105 115 103 Froth Depth (inch) 24 28 28 24 28 30 28 26 22 24 26 Turb 32 32 32 32 Page 112 . -.6 1.0 29.0 12.9 11.93 1.6 1.3 32.6 11.7 42.61 2. (g/T) 690 721 518 423 1307 539 487 201 346 571 405 731 520 831 1091 1282 Coll.5 14.1 12.3 27.41 1.6 1.9 12.4 12.4 1.69 1.71 1.3 12.5 1.38 APPENDIX A TABULATED PLANT TEST DATA Wash Water lpm 48 68 68 55 55 55 55 68 68 68 68 82 68 67 65 64 Air Rate lpm 544 465 425 425 425 425 425 425 425 425 425 425 408 408 425 425 Air Rate (cms) 2.6 17.6 1.43 1.9 45.7 13.32 1.7 1.7 13.23 1.64 2.2 40.55 1.35 1.79 2.2 28.63 1.32 1.0 13.7 43.Dennis I.73 1.7 38.6 Air Fraction % Frother (ml/min) Collect (Diesel) (ml/min) 17 15 17 20 13 13 17 18 ? ? ? ? ? ? ? ? 10.78 1.6 14.6 1.5 1.2 45.66 1.8 39.6 1.5 11.49 1.6 1.

300 x.02 .5 mm 22.15 61.36 84.250 x .96 39.19 10.67 36.54 Cum F.82 59.98 52. Ash Rej.79 58.59 40.06 60.65 45.300 mm 38.51 38.Ash 0.37 33.82 Test 403 Feed Wt.84 59.59 12.89 62.43 10.20 70.18 54.90 64.94 .59 34.48 24.38 .06 60.32 6.45 44.38 88. From Bottom Cum Wt.28 10.13 6.70 89.26 59.85 87. Sep Eff.67 60.80 76.42 .92 22.32 13.79 64. Ash Rej.49 51.07 9.62 52.06 34.48 61.150 10.03 4.67 9.42 .81 10. Yield C.17 40. % Ash % F.08 54.66 48.65 55. Ash Rej.83 73.87 59.91 10.14 21.11 59.30 48.18 44.06 Cumulative 100.00 Results by Size Test 401 Tails Individual by Size Cum.96 10.58 43.07 80.11 84. Rec.23 32.35 52.85 10.07 Cum F.74 40.47 25. % Ash % Cum Wt.03 80.98 91.96 10.29 70.47 100.82 27.38 93.66 81.78 16.43 30. Ash Rej.20 53.25 1 x .34 65. Sep Eff.88 Concentrate Tails Wt.19 10.93 88.53 62.28 77.36 95.06 100.045 16.27 85. Rec.07 9.63 9.83 58.91 Page 113 .38 43.75 16.53 44.16 72.85 87.38 97.15 63.50 64. Wt.36 Concentrate Cum Wt.87 45. Rec.07 39.67 40.58 100.36 84.Ash 36.07 85.23 52.07 88.62 65. % Ash % +0. Ash Rej.64 85.22 52.17 44.36 48.13 55.62 90.15 39.36 100.30 43.17 10.37 34.40 32.47 47.07 36.45 44.05 44. % F.10 4. Yield C. From Bottom Cum Wt.00 58.79 64.91 43.36 41.66 81.67 95.18 10. % Ash % Cum Wt.67 APPENDIX A TABULATED PLANT TEST DATA Individual by Size Cum.83 25.00 36.250 mm 13.95 53.08 75.84 79.Ash 52.36 84.88 Cumulative 100. From Bottom Ash % Cum Wt.43 14.57 37.44 . Yield C.31 65. % Ash % F.07 80.04 25.75 Individual by Size Cum.27 37. Rec.45 11.54 45.66 74. Wt.05 79.06 60.79 64.150 13.53 62.045 mm x 0 33.07 88.12 31. Rec. Sep Eff.61 64.99 74.71 45.43 25.07 15. F.20 70.04 90.54 57.00 60.88 40.37 84.28 12.08 53.19 60.Ash 9.85 87.26 77.045 16.39 64.54 52.29 77.045 mm x 0 34.90 73.65 45.Ash 34.53 91.77 77.54 47.84 87.38 50.07 36.22 13. Ash Rej.20 52. Wt.47 10.06 60.76 .00 12.82 43.36 12.77 23. Rec. Sep Eff.83 58. % Ash % +0.19 6. Phillips Feed Wt.64 12.150 x .04 25.60 67.47 9.10 28.18 44.150 x .Ash 60.08 75.44 26.00 9.43 91.300 mm 35.24 21.26 .27 97.68 43.98 52.300 x.08 77.54 45.24 77.80 9.98 52.38 50.36 17.78 78.5 x .22 25.36 12.96 42.65 38.30 32.00 30.18 44.54 36.150 x .09 37.68 44.04 90.71 79.99 11.53 91.57 . Yield C.85 74.55 Test 402 Feed Wt. F.150 14.14 30.43 38.21 7.97 6.01 75.Ash F. Yield C.43 28.84 79.91 61.81 73.93 70.00 29.34 65.18 44.26 81.38 38.65 45.54 33.79 64.24 82.44 15.045 16.72 17.045 mm x 0 31.82 59. F.43 25. % Ash % Cum Wt.16 12.01 77.16 66.00 9. Yield C.27 11.67 Concentrate Tails Wt.06 34.61 100.06 51.65 23.60 67.98 48.65 45.50 66.39 38.Dennis I. Sep Eff.Ash 58.97 53.23 10.17 80. % Ash % +1mm 3.43 84.66 45.26 75.98 52.62 65.65 74.Ash 40.54 .03 100. Sep Eff.06 Cumulative 100.13 9.40 46.

27 34.150 x .84 85.00 10. % Ash % Cum Wt.96 38.20 50.86 .11 28.88 38.67 41.98 29.70 12.35 49.Ash 0.02 34.00 10.67 23. Rec.92 78.47 71.95 86.50 Feed Wt.39 79.54 84. % Ash % Cum Wt.84 100. Rec.37 62.46 63.86 .37 Cumulative 100.250 mm 13.34 35.31 55.55 60.41 57.04 66.74 57. Sep Eff.72 41.46 47.72 13.76 99.44 47.76 85.11 12.045 mm x 0 36.64 37.81 47.90 37.13 59.73 100.19 59.53 .17 45.00 48.Ash 0.50 Test 404 Concentrate Tails Individual by Size Cum.93 58.36 55.84 71.045 mm x 0 30.20 .74 8. % Ash % Cum Wt.37 48. 47.79 11.62 60.52 83.34 38.66 100.06 49.Ash F.69 89.18 .35 13.045 15.150 10.81 42.67 49.82 26.51 28.67 73.65 62.92 91.06 83.97 17.56 66.87 32.46 38.54 84.93 57.045 16. Yield C.73 67.48 8.43 8.28 34.07 57.09 3.49 81.50 2.61 .51 5.92 57.17 6.15 78.58 14.150 10.80 45.20 63. F.47 97.48 17. Ash Rej.47 93.88 39.16 90.39 28.00 58.150 10. F.59 8.35 82. Sep Eff.26 .71 10.250 x .91 43.20 .95 29.70 69.5 x . Wt.Ash 39.89 49.250 x .85 43.96 70.04 87.23 45. % Ash % Cum Wt.47 9.42 36.5 x .Ash F.66 41. Yield C.00 72.Ash 38.03 55.30 59.26 89.66 71. Yield C.04 87.68 75. % Ash % +1mm 4.23 82.71 11.51 87.50 39.10 49.30 63.17 Page 114 .09 77.68 .58 42.04 23. From Bottom Wt.99 45.07 11. F. Ash Rej.04 27.28 10.62 7.00 69.51 96.76 Feed Wt.78 81.66 66.81 56.01 13.72 91.75 78. From Bottom Wt.44 37.44 6.99 66.13 36.95 12.00 37.5 mm 22.85 Cum F.05 1 x .43 15.84 39.57 40.51 87.70 75.20 84.75 53.01 11.81 98.87 9.91 12.46 17.90 49.75 9.66 33. Ash Rej.67 55.47 93. % Ash % Cum Wt.89 10. Rec.64 17.16 61.22 84.70 89.86 58.75 54.94 92. Phillips Feed Wt.11 12.32 90.150 x .17 66.80 13.64 76.97 90.09 6.74 58.33 85.95 19.74 80.68 8.11 APPENDIX A TABULATED PLANT TEST DATA Sep Eff. % Ash % +1mm 3.69 55.46 64.94 10.97 36.51 5.77 26.87 41.03 37. Rec.06 60.04 82.58 55.00 8.46 Cum F.36 34.32 82.73 98.45 71.38 10.00 82.75 8.58 32.62 Test 406 Concentrate Tails Individual by Size Cum.37 49.Ash 36.64 69.27 49. Sep Eff. Yield C.29 82.04 89. Rec.04 51.59 5. Ash Rej.95 41.13 12.64 64.045 16.99 71.150 x .12 76.41 51.94 82.89 86. 34.38 91.58 28.86 49.30 100.45 100.46 80.79 4. Yield C.98 39.03 13.85 39.99 79.35 67. Wt.87 49.03 39.250 x .09 57.58 9.50 38.12 22.75 78.77 67. Wt.07 36.98 84.83 Test 405 Concentrate Tails Individual by Size Cum.93 40.03 Sep Eff.51 100.83 36.23 8.30 11.52 31.66 1 x .22 16.250 mm 15.35 72. Ash Rej.40 45.83 13.30 3.54 16.19 88.42 82.25 12.Dennis I.97 48. From Bottom Wt.72 91.49 8. Ash Rej.38 33.98 38.5 x .33 93.41 47.08 12.74 79.73 5.87 68.250 mm 14.53 .98 84.40 99.98 11.18 33.41 83.Ash F. % Ash % +1mm 3.65 83. Yield C.41 13.67 10.28 .22 11.68 .00 29.73 67.58 11. 64.66 42.73 13.5 mm 24.22 7.5 mm 20.46 86.34 7. Rec.13 12. % Ash % Cum Wt.10 79.36 4.07 18.04 87.91 48.49 51.35 28.11 66.73 Cumulative 100.10 79.12 14.85 27.02 41.76 36.49 Sep Eff.56 41.10 43.42 Cum F.045 mm x 0 30.27 59.08 4.21 45.84 6.Ash 0.62 65.21 28.25 13.06 31.63 73.62 Cumulative 100.16 40.72 1 x .96 65.34 4.

08 34.90 54.83 Test 407 Concentrate Tails Individual by Size Cum.18 8.Ash F.99 14.79 29.87 8.59 86. Phillips Feed Wt.16 42.04 38. Wt.35 74.10 51.86 100.71 39.92 100.64 56.56 35.52 43.31 8.60 11. % Ash % Cum Wt.87 92.Ash 41.78 64.83 41. Rec.00 59.150 x .75 31.61 57.42 30. Sep Eff. From Bottom Wt.42 80.57 33.5 x .85 77.30 7.23 50.12 89.46 43.78 16. From Bottom Wt.23 6.01 11.91 83.22 83.32 13.13 98.37 11.06 44.83 37.01 61. % Ash % Cum Wt.46 69.045 .97 58.67 30.14 20.48 48.32 94.150 x .59 68.93 42.37 83.72 32.96 92.00 11.Ash 0.44 32.00 68.44 9.71 34.045 mm x 0 Cumulative 6.21 39.76 3. Ash Rej.39 37.67 10.93 15.00 8.69 71.92 34.95 22.68 57.92 10.04 40.41 37.74 90.5 x .57 100.59 70.69 82.27 14.61 6.99 48.63 82. Yield C.58 79. % Ash % Cum Wt.11 7.81 16.00 11.78 100.90 54.50 33.49 53.73 32.69 94.00 -100. Ash Rej.06 44.20 15.62 14. Ash Rej.40 70.5 mm .45 49.39 49. Sep Eff. Sep Eff.39 37.24 62.25 14.23 49.35 71.68 46.07 47.21 64.61 69.09 37.18 10.10 25. % Ash % +1mm 1 x .34 8.40 92.75 45.38 8.89 81. From Bottom Wt.51 84.00 0.37 24.Ash F.03 8.41 18.92 Cum F. Yield C.97 58.17 9.40 Feed Wt. Yield C.14 92.64 31. F. % Ash % +1mm 1 x .87 24.5 mm .00 41.250 x .67 58.99 86. Wt.86 40.Ash 0.92 75.250 mm .41 Test 409 Concentrate Tails Individual by Size Cum.5 mm .17 78.07 21.56 51. F.99 37.83 65.00 81.83 47.150 x .50 11.98 52.24 71.80 73.52 27.01 61. % Ash % Cum Wt.75 9.68 94. Ash Rej.77 24.29 71. Yield C.24 13.91 41.85 35.58 64.57 76.99 51.99 37.73 27.70 45.21 62.37 11.37 5.46 52. F.04 52.30 29.96 92.58 9. Rec.37 10.01 62.86 4.73 7.80 38.00 32.21 69.150 .86 90.12 40.90 4.40 70.66 43. Rec.Ash F.82 8.5 x .150 .10 30.61 37.98 14.02 50.71 9.74 89.09 11.61 92.64 38.15 42.90 16.32 35.04 17.00 81. % Ash % Cum Wt. Yield C. Rec.35 86.66 26.52 77.34 14.83 65.14 92.00 0. Ash Rej.63 46.29 96.27 89.Ash 37.00 13.40 98.19 59.52 8.73 7.60 56.045 .67 57.68 100.81 13.59 68. % Ash % Cum Wt.00 8.79 36.05 7.06 56.05 94.71 100.81 15.36 75.86 26.84 57.00 50.40 37.045 .51 40.52 45.33 69.60 60.08 52.74 56.93 60.80 99.83 Cum F. Ash Rej. Sep Eff.43 44.92 9.62 10. Wt.44 13.Ash 0.43 15.21 39.05 27.17 50.39 22.34 82.30 29.28 86.73 46.88 38.56 43.84 11.50 29. % Ash % +1mm 1 x .82 92.79 37.78 57.045 mm x 0 Cumulative 3.48 47.84 100.07 31.92 11.045 mm x 0 Cumulative 3. Rec.36 64.91 Cum F.66 25.46 100.37 33.73 15.87 53.91 92.64 42.92 38.65 65.98 6.16 37.250 x .Dennis I.00 Feed Wt.70 93.14 APPENDIX A TABULATED PLANT TEST DATA Page 115 .96 8.67 31.00 51.00 40.95 Test 408 Concentrate Tails Individual by Size Cum.67 9.150 .00 36.46 82.27 86.42 1.45 5.82 92.42 0.73 41.25 59.52 44.28 50.250 mm .250 x .40 41.28 71.13 79.26 38.79 9.60 14.55 34.54 93.65 5.72 75.42 7.04 11.86 29.38 16.18 26.73 49.36 34.08 5.41 56.27 89.84 7. Yield C.95 94.13 2.01 24.31 72.67 100. Sep Eff.62 34. Sep Eff.250 mm .32 52.78 54.70 48.46 37.00 32.12 45.34 56.24 52.06 35.78 16.45 8.00 30.Ash 41.24 66.55 71.96 9.34 54. Rec.

39 73.41 9.06 9.08 9.27 44.69 10.150 x .76 55.93 .36 66.01 46. Sep Eff.35 73.98 Cumulative 100. Sep Eff.81 10.Ash 0.49 78.30 38.09 8. Wt.84 28.62 42. Sep Eff.38 91.20 42.80 99.62 10.79 24.83 9. Ash Rej.20 Cumulative 100. % Ash % Cum Wt.10 92.48 91.53 60.46 94.54 46.15 1 x .Ash 54.23 68.89 65.48 46.28 .26 61.Ash 55.79 14.80 33. Rec.05 32.82 14.49 8.64 62.80 56.20 36.14 12.49 98.42 13.94 59.26 45.34 35.98 32.045 mm x 0 39.71 25.09 62.48 91. Wt.045 14.56 34.01 44.05 39.14 91.72 5.48 100.80 52.37 33.55 93.045 14.04 3.30 31.18 34.5 x .93 14.85 67.06 49.82 38.97 5.12 1 x . F. F.86 .05 11.39 64.00 37.67 51.89 6.27 41.21 49.83 50.38 91.19 37.37 73.66 Test 412 Feed Wt.18 94.00 10.62 60.59 10.89 10.48 28.30 . % Ash % +1mm 5.57 21.40 57.82 30.09 50. Ash Rej. Yield C.20 42.35 8.18 91.96 72.21 36.93 16.03 98.00 55. Rec.87 51.00 33.41 . From Bottom Cum Wt.150 x .88 47.250 mm 18.92 1 x .5 mm 19.19 91.09 41.30 100.79 42.37 56.21 40.37 61.76 10.09 83.03 5.84 52.11 51. Sep Eff.88 13.60 85.89 98.89 41.47 36.30 96.02 14.10 39. Rec.98 27.74 33.71 9.22 70.26 41.63 54.65 APPENDIX A TABULATED PLANT TEST DATA Individual by Size Cum.00 55. % Ash % Cum Wt.95 40.13 99.250 x .40 14.48 94.69 9.97 36.34 10.05 77.74 38.84 47.10 Concentrate Tails Wt.49 32.52 10.01 71. Yield C.5 mm 21.71 100.09 50.53 .250 x .59 Cum F.Ash 39.27 54.67 .Ash 0.69 43. Phillips Test 410 Feed Wt.045 14. Yield C.38 88.00 33.71 9.56 90.16 73.92 14.22 70.41 29.81 Page 116 .95 13.25 47.83 52. Sep Eff.45 43.71 68.89 Concentrate Tails Wt. % Ash % F.05 100.93 2.5 mm 19.15 33. % Ash % Cum Wt.37 51.045 mm x 0 36. Rec.09 77.98 14.150 8.10 92. % Ash % F. % Ash % F.92 89.74 8.49 34. Wt. From Bottom Cum Wt.94 44.09 62.84 51.20 .80 53.11 93.99 34.55 44.07 52.150 x .84 53.85 61.57 73.53 60.34 24.28 51.84 14.35 12.17 10.30 57.Ash 0.72 49.65 39.84 55.84 7.13 10.11 51.23 8.89 10.250 mm 10. Ash Rej.73 9.53 20.80 37.34 9.80 57.95 14.Ash 40. F. Ash Rej.21 94.99 93.80 100.71 8. Yield C.5 6.58 11.16 54.55 51.150 8.26 27. Yield C.86 .22 Individual by Size Cum.68 91.83 90.65 55.01 91.46 31.61 44.21 9.87 61. Rec.15 27.37 56.250 x .52 50.05 95.94 62.57 7.5 x . % Ash % +1mm 4.51 9. % Ash % +1mm 4.13 Cumulative 100.250 mm 16.69 47.75 54.55 37.50 42.95 Concentrate Tails Wt.93 45.90 54.88 27.67 91.Dennis I.70 67.74 6.15 36.42 62. Ash Rej.79 93.95 6.17 33.81 9.62 3.66 6.47 9.63 45.42 36.50 42.60 36.45 46. From Bottom Cum Wt.80 5.60 10.4 39.56 47.63 8.19 37.91 73.59 .79 38.62 88.18 77.18 43.08 39.61 51.78 .25 58.08 6.00 9.25 73.09 31.38 39.79 93.00 9.13 29.08 35.99 8.36 62.61 35.98 10.48 8.13 63.82 91.30 Test 411 Feed Wt.68 17.57 5.85 44.81 45.57 35.09 38.75 47. Rec.98 31.92 62.03 14. Yield C.00 54.47 42.5 x .48 30.10 37.82 91.62 42.92 62.09 .21 51.41 Cum F.91 42.Ash 55.26 9.21 Cum F.Ash 41.62 45.68 43.045 mm x 0 33.21 51.54 14.150 9.75 66.49 33. Ash Rej.22 14.91 Individual by Size Cum.16 89. Sep Eff.26 44.21 36.01 92.

94 8.12 38.82 51.65 55.46 45.90 36. Yield 29.07 73.69 33.75 87. % Ash % Cum Wt.96 7.29 37.74 62.08 46.Ash 0.99 89.08 62.Ash 46.79 86.55 53.87 73.04 10. 61.08 39.15 90.66 11.99 40. Ash Sep Eff.89 58.64 11.89 9. Yield Rej.69 9.03 Cum F.Ash 44.96 63.37 86.48 50.99 Feed Wt.10 36.00 71.59 60.30 11.88 100.08 87.41 36.45 39.61 42.045 mm x 0 Cumulative 7.73 78.28 96.19 23. From Bottom C.19 45.82 53. % Ash % +1mm 1 x .20 48.50 89.27 67.81 90.24 52.65 90.02 46.60 13.05 86.250 mm .93 26. C.93 67.97 35.20 14.250 x .48 89.21 80.150 .40 69. C. Yield 45.61 Test 414 Concentrate Tails Wt.00 13.33 99.Ash F.54 42.33 78.93 71.90 16.10 59.77 72.04 66.68 Cum.66 76.40 100.99 71.15 66.00 66.75 58.Dennis I.22 84.34 60.53 46. Ash Rej.78 27.250 x .75 43.54 53.08 31.98 11.96 68.17 82.16 82.46 45.02 Cum F.30 Individual by Size Wt.14 87.94 43.16 64.03 39. % Ash % Cum F.10 83.04 45.95 53.33 44.09 22.18 100.53 25.61 29.57 40.80 56.250 x .39 84.44 12.18 Cum.92 41.93 65.75 85.07 67. % Ash % +1mm 1 x .Ash F.32 82.77 31.250 mm .05 74.87 6.05 41. Yield 37.51 13.28 94.00 6.00 11.5 mm .04 12. % Ash % Cum F.87 66.23 55.94 18.63 55.53 17.47 54.32 39.44 16.31 13.16 16.36 10.52 22.18 84. % Ash % Cum Wt.81 58.Ash 39.18 83.24 29. From Bottom Wt.31 66.88 23.5 mm .03 53.91 100.62 46.Ash 0.98 12.045 .14 82.90 41.64 100.06 42.04 84.61 42. Rec.95 81.80 29.07 76.04 51. C.02 49.63 8.35 12.97 56.28 71.23 43.64 71. Yield Rej.00 55.88 50.70 10. Sep Eff.72 80.48 89.15 39.15 76.40 3.49 11.43 64.89 97.17 12.39 74.40 81.23 83.38 61.50 90.47 100.83 48.87 36.150 .66 70.33 15.43 41.94 43.70 32. Rec.045 .16 7.01 26.61 86.5 mm .150 x . Ash Sep Eff.35 12.61 13.150 .22 46. Yield Rej.76 14.65 11.96 Individual by Size Wt.00 44.05 16.86 79.84 100.94 63. Ash Sep Eff.16 82.65 37.99 10.86 50.83 31.10 34.82 82.9 50.33 76.50 51.7 56.36 5.42 3.00 50.66 46.79 14.07 Cum.045 .07 22.84 43.150 x .045 mm x 0 Cumulative 5.54 5.74 15.65 8.52 48.66 Cum F. % Ash % Cum Wt.5 x .71 35. Ash Rej.86 28.00 38.32 26. Rec.15 20.46 11.14 82. % Ash % +1mm 1 x .59 71.25 90. Sep Eff.13 100.45 21.73 80.Ash 0. % Ash % Cum F.17 44.86 41.51 Individual by Size Wt.85 98. C.20 9. C.00 13.00 38.51 66.17 7.82 65.11 34.29 34.49 66.25 90.75 36.51 18.77 6.47 78.Ash F.21 13.91 5.91 6.39 88. Rec. 13.93 67.66 22.96 63.65 44.30 19.27 47.00 59.29 Page 117 .14 91.34 23. From Bottom Wt.22 44. Rec.41 15.150 x .43 65.84 8.84 5.95 11.78 86.19 Test 415 Concentrate Tails Wt.66 44.39 39.74 45.61 5.18 31.38 34.90 58.95 30.63 28.12 68.84 97.86 79.16 64.045 mm x 0 Cumulative 5.29 14.62 12.26 87. na na na na 6.70 32.95 13. Sep Eff.250 mm .29 16.52 56.66 54.18 13.88 82.65 Feed Wt.70 80.5 x . 17.04 38.26 42.77 35.88 27.93 73.08 Test 413 Concentrate Tails Wt.87 23. Ash Rej.93 40.73 33. Phillips Feed Wt.58 71.70 100.26 87.51 40.65 APPENDIX A TABULATED PLANT TEST DATA Wt.58 5.00 12.42 28. Rec.49 47.5 x .14 98.68 47.49 5.39 83.50 43.87 46.45 47.86 83.01 15.66 45.13 14.

25 14. Sep Eff.84 70.88 87.64 78.18 81.85 12.77 Test 416 Concentrate Tails Individual by Size Cum.62 36.89 14.76 8.22 45.64 30.34 63.00 69.01 24.57 10.25 42.78 48.96 14.01 54.39 43. Yield C.81 52. % Ash % Cum Wt.96 52.12 13.71 24.68 57.62 53.40 5.32 4. Sep Eff.150 x .63 46.79 88.80 73.79 100.77 43.23 23.21 53.00 12.68 57.18 29.79 5. Phillips Feed Wt.Ash F.27 12.44 39.83 11.61 86.78 84.88 87.5 x .78 82.46 79.99 75.57 100.37 54.04 31.72 100.17 96. Ash Rej. % Ash % Cum Wt.64 83. Yield C.60 27.26 Cum F.44 67.37 15.0 0 42.20 41.250 x .045 .38 15.94 34.97 64.57 84.30 15.10 81.37 86. F.20 15.30 4.62 53.87 43.66 11.26 43.81 77.99 45.72 43.73 66.67 83. Wt.00 78.Ash 43. Ash Rej.30 14.37 66.58 90.88 99.250 mm .72 82.17 12. From Bottom Wt.62 86.150 .22 8.76 82.72 71. Rec.56 40.96 54.37 67.5 mm .44 62.44 13.83 70.83 9.84 66.17 74.77 13.30 APPENDIX A TABULATED PLANT TEST DATA Page 118 .41 83.00 78. Rec.01 60.36 12.Ash 0.84 66.68 74.76 44.045 mm x 0 Cumulative 11.35 72.Dennis I.87 82. % Ash % +1mm 1 x .71 46 69.

73 12.52 Ash Rej.31 45.33 10.81 86.10 51. % 60.09 84.82 55.26 41.10 APPENDIX A TABULATED PLANT TEST DATA Feed 7.43 10.16 13.50 52.14 42.43 54.90 78.32 15.77 Page 119 .97 58.% 83.26 33.80 36.24 11.69 46.08 78.18 11.41 68.89 80.35 15.05 85.79 76.12 10.04 66.76 7.24 42.61 34.86 40.32 7. 11.11 79.38 87.40 2.90 % Solids Conc.47 17.64 87.85 63.32 % Wt.70 38.32 66.03 59.53 74.99 62.71 55.16 62.52 11.33 2.78 89.06 75.81 8.Dennis I.43 72.56 69.74 84.86 4.31 7. 12.29 76.84 87.62 36.68 39. Sep Eff.44 42.30 60.55 Comb.00 61.54 48.90 14.17 10.25 Comb.20 57.60 9.22 77.13 67.10 9.29 34.51 14.50 13.29 51.10 50.40 12.39 66.88 49.35 14.66 11.42 78.52 79.88 51.97 59. Phillips Second Parametric Series (451-465) Test Results Test # 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 Feed 36.88 59.87 35.08 12.25 11.23 3.24 13.11 5.64 7.31 58.63 12.53 48.06 8.38 6.95 38.50 64.52 80.61 86.44 5.31 15. Rec.23 11.83 69.66 82.30 Ash % Conc.93 63.01 15.55 12.39 59.52 84.76 11.34 73.35 75.66 4.06 68.02 74.08 13.60 12.05 66.90 6.57 12.02 42.59 82.63 60.18 13.63 42.43 10.45 56.24 10.77 Tails 3.78 12.08 11.28 7.44 52.73 75.27 6.75 3.68 Tails 67. Yield 54.% 76.19 12.

88 1.3 1.2 14.0 16.7 8.9 7.6 11.6 14.3 9.37 1.87 0.6 47.03 1.3 3.17 1.7 9.84 1.6 25.2 Air Fraction % Frother (ml/min) 5.0 9.0 7.1 6.54 0.8 47.20 1.7 27.7 11.75 1.86 1.0 6.0 3.55 1.3 4.76 1.3 11. Phillips Second Parametric Series (451-465) Operating Parameters Test # 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 Feed Rate (gpm) 41 60 50 50 50 60 50 50 50 60 40 60 40 40 50 Feed Conc Conc Conc Rate Rate Rate Rate (kg/min) (kg/min) Tph/m2 -.3 6.5 16.98 0.5 mm (g/T) 1358 1383 1171 1445 1348 1385 699 765 749 1041 1128 745 834 1136 724 Froth Depth (inch) 24 20 26 24 28 25 29 30 27 28 25 27 25 25 28 Page 120 .5 Collect (Diesel) (ml/min) 16.2 1.6 50.7 43.73 1.1 13.0 8.5 11.0 9.6 26.1 1.0 6.6 20.5 7.90 1.2 0.6 15.2 1.1 1. -.66 APPENDIX A TABULATED PLANT TEST DATA Wash Water lpm 54.3 5.5 7.0 6.5 50.2 1.Dennis I.3 9.2 1.92 2.7 43.9 13.0 11.5 24 20 20 20 24 20 20 20 23 16 23 16 16 19 Coll.37 0.2 17.5 mm 11.01 1.87 0.7 8.2 6.3 1.0 3.38 1.5 Air Rate lpm 323 350 306 350 350 329 329 329 306 306 306 329 350 329 329 Air Rate cms 1.2 23.3 10. (g/T) 1107 1141 974 1191 1102 1096 568 612 600 764 876 631 608 799 550 Coll.5 12.8 13.2 1.06 1.8 47.3 1.08 0.02 0.9 9.7 7.54 1.8 11.2 1.7 16.7 51.3 1.5 50.8 47.2 27.9 8.83 1.9 43.1 1.3 11.3 28.0 8.1 14.4 8.1 1.31 1.55 1.8 51.8 40.5 2.14 1.8 54.04 0.0 11.5 3.

05 80.84 88.42 65.96 47.36 87.44 86.21 3.52 72.07 29.82 13.30 35.16 81.59 12.58 3.86 55.19 64.86 79.59 69.84 .18 100.33 11.23 60.78 .03 65.79 83.250 mm 19.69 11.32 77.66 60.00 Test 453 Tails Ash % Cum F.44 .65 74.38 71.55 20.Ash 38.35 43.88 84.00 20.50 47.54 11.72 67.63 5.79 12.150 x .73 24.Ash 24.86 92.76 77.03 65. % Ash % +1mm 1. Ash Rej. Sep Eff.41 79.86 60.99 94.41 81.50 12.60 80.80 78.78 83.56 77.29 Cum F.80 76.70 23. Sep Eff.47 89.38 1 x . % F.250 mm 20.52 48.31 77.23 60. Rec. C.63 82.70 76.72 100.83 54.81 60.69 .55 63.58 90.15 80.00 24.34 79.84 88.95 37.66 12.05 78.97 71.22 65.24 100.96 11.70 31.09 38.59 88.02 12.90 85.21 42.47 13.85 76.70 23.21 69.91 49.37 55.66 11. From Bottom Wt.32 36. Ash Rej.28 40.60 100.24 11.08 82.00 Results by Size Test 451 Tails Individual by Size Cum.00 11.73 50.08 14.95 9.28 91. % Ash % Cum Wt. Sep Eff.90 85.36 68. C.55 12.00 20.98 1 x .58 80.Ash 36.51 32.58 54.15 54.81 52.045 16.04 76.Ash 0.045 mm x 0 35.67 57.57 29.98 39.92 73.15 80.150 x .25 54.250 x . Sep Eff.53 66.40 36.79 43.30 67. Phillips Feed Wt.92 10.5 x .50 76.14 .48 13.84 10.98 39.31 34. Rec.00 11. Rec.37 43.250 x . Wt.80 9.22 .39 22.25 39.57 40.74 50.150 x .5 x .10 59.78 84.66 84.12 74.75 Cumulative 100.82 63.49 76.76 77.76 37. Rec.75 13.21 26.05 25.80 8.150 9.35 74. Ash Rej.11 36.04 77.12 52. % Ash % +1mm 1.06 58.57 60.78 48.18 1. % F.54 53.52 12.13 77.49 67.23 13.045 mm x 0 35.89 59.54 77.05 9.79 54. Ash Rej.62 63.13 Individual by Size Cum.07 47.96 APPENDIX A TABULATED PLANT TEST DATA Page 121 .13 50.62 52.30 42. Yield Yield 19.83 66.045 16.41 57.250 x .25 62.99 86.96 13.10 12.55 82.83 36.28 23.Ash 40.44 Cum F.9 11.76 12.52 64. % Ash % +1mm 1.00 77.73 45.64 13.14 .66 34. Wt.66 60.16 55.94 65.79 64.35 34.97 75.26 61.52 60.14 66.23 100.69 100.13 48.16 65.31 68.33 80.81 79.97 Cumulative 100.61 Concentrate Wt.34 12.39 73.Ash 0.93 52.62 53. Sep Eff.81 89.59 12.61 36.64 5.250 mm 19.24 2.51 .29 Concentrate Wt.75 4.93 55.43 48.77 54. C.35 13.82 41.08 82.13 12.54 73.70 83.73 74.01 12. From Bottom Ash % Cum Wt.23 1. Yield Yield 43.22 84.69 12.31 71.48 82.48 59.04 41.34 50.29 36.48 .19 73.49 41.67 67.26 22.52 60.62 38.03 81.80 45.4 1 x .87 71.96 17.35 68.62 81.5 mm 16.27 13.50 52.58 13.87 Concentrate Wt.96 Feed Wt.5 mm 16.045 16.Ash 36.62 Feed Wt.21 79.06 19. % Ash % Cum Wt.150 11.Ash Yield Yield 30.21 24. From Bottom Wt.Dennis I.07 75.68 61. C.97 39.53 54. Sep Eff.5 mm 15. F.Ash 0.85 12.67 45.73 69.16 77.06 11.56 38.05 67. Ash Rej.88 59.03 61.49 77.88 54.21 24. C.74 11.150 11.00 11.19 12.20 56. Rec. % F.62 68.55 83. Rec.36 87.99 4. % Ash % Cum Wt.82 96.00 9.00 Test 452 Tails Ash % Cum F.61 58.22 83.5 x .14 66.13 52.71 76.11 Cum F.61 37.97 .21 84.46 Cumulative 100.32 Individual by Size Cum.11 .12 63.11 58.52 38.68 70.92 53.81 79.23 83.83 66.03 58. C.11 83.045 mm x 0 33.49 47.18 43.49 83.61 88.82 38.51 84.06 23.93 48.92 12.38 83.33 38.99 82. Wt.44 38.21 87. Ash Rej.32 11.21 82.04 9.74 .

78 75.31 51.04 22.60 33.150 x .9 .96 24. Ash Rej.22 85.59 51.89 57.24 81.43 80.00 67.90 58.45 47.75 Cumulative 100.56 41. C.28 77. Phillips Feed Wt.99 11.66 .51 23.76 27.87 79.14 92.03 61.57 79.62 13.26 Test 456 Concentrate Tails Wt.83 68. % Ash % +1mm 2.47 12.50 28.14 6.03 8.61 57.92 57. From Bottom Wt.59 11.41 .87 Feed Wt.39 5.Ash 33.82 9. Wt.29 Individual by Size Cum.57 59.84 .59 17.53 79.18 71.62 5.14 83.25 67.77 72.04 80. C.80 14. Sep Eff.69 67.03 38.43 82. % Ash % Cum F. C.71 46.64 48.00 11.98 .38 66.33 13.40 58.18 39.94 33.Dennis I.08 88. Yield Yield 9.88 13.09 17.08 88.08 12.35 11.86 60.14 54.250 x .90 16.5 x .05 1 x .97 .25 10.39 Individual by Size Cum.02 10.07 7.76 12. Ash Rej.75 98.5 mm 16.61 1 x .63 35.78 98.00 23.56 14.05 Individual by Size Cum.48 40. Ash Rej.32 56.04 87.43 82.33 .72 77.150 11.22 47.51 38.98 66.68 78. Wt.045 16.40 61.34 88.09 66.55 35.06 51.Ash 0.70 11.74 12.045 mm x 0 32.59 10.15 39.Ash 0.81 84.Ash F.19 32.72 51.84 44.05 12.45 36.79 56.86 100.5 mm 18.42 80.69 75.12 100.22 85.78 48.150 x .88 81.54 52.33 12.86 83.53 71.41 40.66 41.64 35.20 53. % Ash % +1mm 1.5 x .50 29. % Ash % Cum Wt.95 24. Sep Eff.03 11.73 62.07 45. From Bottom Wt.05 26.39 54.03 58. Rec.27 96.20 9.61 65. % Ash % Cum F.150 x . Sep Eff.68 88.00 25.65 72.50 3.26 32.04 .250 mm 18.77 42.00 22.40 13.62 54.150 9.42 80.20 31.75 4.13 8.250 x .76 51.90 50.83 94.64 17.250 x .37 43. Yield Yield 9.77 65.95 58.250 mm 19.54 53.43 47.09 66.045 mm x 0 37.95 63.84 61. Wt.150 9.88 Cumulative 100.14 54.47 10.55 Test 454 Concentrate Tails Wt.89 71.68 10.93 86. From Bottom Wt.21 7.21 14.00 27.80 72.79 10.08 2.36 84.17 69.5 x .04 87.84 38.79 9.02 82.95 3.63 Cum F.250 mm 17. Rec. Ash Rej.62 90.36 .Ash F.87 20.82 66.93 84.52 67. % Ash % Cum F. Sep Eff.27 11.Ash 0.46 12. Yield Yield 6.55 44.50 26.34 53.25 Feed Wt.11 82.05 .12 79.25 98.79 61.17 67.20 45.00 66.92 48.60 11.47 26.31 66. Rec.09 36.10 66.12 100.51 66.72 84.90 70.49 5.84 68.00 66.27 6.82 APPENDIX A TABULATED PLANT TEST DATA Page 122 .61 53.77 50.48 29. C.56 .43 100.94 50.70 34.78 41.59 59.24 38.00 Test 455 Concentrate Tails Wt.45 76.67 5.20 78.47 54.25 66. % Ash % Cum Wt.91 82.48 73.91 6.17 83. C.42 80.83 56.30 28.045 16.77 79.43 1. Sep Eff.68 57. Ash Rej.31 58.39 6.36 8. % Ash % Cum Wt.57 87.74 84.23 55.90 59.88 23.67 1 x .57 47.29 8.43 9.25 .02 63. Sep Eff.36 29.11 82. Rec.18 12. C.69 Cumulative 100.Ash 35.50 83.92 71.50 83.76 32.59 71.27 43.95 13.Ash F.67 32.93 11.94 Cum F.12 2. Rec.37 39.93 29.17 80.31 83.05 9.43 75.90 10.31 84.85 11.84 52.48 26.11 14.92 100.13 21.84 39.68 78.18 10.045 mm x 0 36.75 60.83 61.48 11.70 Cum F.42 43.67 10. Rec.33 22.00 20. Ash Rej.26 58.25 76.08 100.42 23.68 57.25 30.86 77.00 51.79 98.77 59.01 79.06 55.5 mm 15.81 30.04 39. % Ash % +1mm 1.26 34.00 11.50 59.22 78.14 85.045 17.90 70.11 30.37 64.34 53.24 81.79 41.31 52.92 70.32 58.84 68.65 13.Ash 34.80 13.12 84.87 34.46 67.00 10.48 13.

11 64.75 85.48 9.93 11.06 74.03 87.12 58.22 100. Rec.50 77.19 17.90 77.69 66.98 9.44 51.35 38.25 62.31 10.43 5.92 35.Ash 0.19 88.13 62.06 11.Ash 0.17 19.150 .08 41.31 8.28 56.21 84. Yield Yield 3.90 22.18 44.83 11.56 68.60 64.045 .39 12.14 40.61 Feed Ash % 26.33 86.10 43.66 2.87 46.67 78.14 12.72 64.77 84. Sep Eff.96 77.74 64.5 mm .06 Individual by Size Cum.61 51.74 44.36 37.69 72. Phillips Wt.51 66.5 x .47 8.00 Feed Ash % 33.05 60.48 81.84 9.18 86.34 33.68 87.150 x .50 46.40 65.62 42.250 mm . Wt.00 10.3 100.87 74.63 50.92 43.72 86.05 43.62 11.86 Cum F. Rec.34 3.40 99.17 10.17 38.47 77.18 41. C.250 x .61 42.30 40.68 20.93 46.39 37.39 40.33 33.89 36.Ash F.15 65.72 22.76 11.54 68.75 10.69 66. Ash Rej.38 APPENDIX A TABULATED PLANT TEST DATA Page 123 .08 80.58 35.05 60.045 mm x 0 Cumulative 1. From Bottom Wt.64 97.72 13.78 7.28 48.29 75.00 Wt.46 74.24 64. Wt.62 15.94 50.93 52.70 Feed Ash % 28.8 17.88 77.88 74. % Ash % Cum Wt.06 79.41 11.49 10.66 56.66 Cum F.38 55.23 13.88 51.045 mm x 0 Cumulative 2.54 4.85 85.96 39.09 88.86 11.22 15. % +1mm 1 x . C.93 39.50 6.60 28.36 94.37 30.52 75.70 64.99 24.27 53.5 x .41 58.05 47.43 10.48 65.24 87.00 52.66 100.74 80.23 33.00 10.150 x .27 65.80 97.34 76. Rec.22 77.45 10. Rec.54 51.03 61.87 78.92 79.06 50.44 99.19 30. Yield Yield 28.16 43.99 61.32 89.5 mm .24 87.59 46.41 Individual by Size Cum.58 90.98 10. C.01 65.045 mm x 0 Cumulative 2.35 44. Sep Eff.85 75.17 10.045 .60 100.96 77.02 55.86 39.32 48.94 44.70 34.95 74.00 56.60 51. Ash Rej.89 42. Ash Rej.66 79.20 15.73 57.66 10.Ash 0.75 85.59 77.78 84.12 79.09 17.90 50.06 11.15 73.51 46.41 38.87 15.36 85.01 65.40 4.60 2.95 59.86 46.42 8.16 46.87 9.39 86.10 64.53 93.80 84.88 29.69 63.150 .54 74.92 56.69 41.19 58.52 38.07 9.93 13. Sep Eff.27 86.250 x .00 66.79 10.65 30.43 35.68 15.19 88.41 Test 457 Concentrate Tails Wt. Rec.89 Cum F.02 87.07 8.03 9.85 55.00 Wt. Wt.52 66.85 68.63 5.5 x .93 52.03 42.77 29.38 55. % Ash % Cum F.35 3.250 mm .01 48.32 11.63 76.63 14.12 17.Dennis I.69 3. % Ash % Cum Wt.91 81.28 12.47 73.49 89.20 85.045 .Ash 38.66 100.85 11.12 100.65 50.92 12.06 68. % Ash % Cum Wt. Sep Eff.09 66.53 28.73 13.36 50.10 21.03 35. % Ash % Cum F.02 60.150 x .00 93.13 43.Ash F.59 70.50 77.38 11.59 77.50 57.53 77.88 2.33 100.70 64.48 4.45 9.44 52.45 Individual by Size Cum.05 74.88 87. Yield Yield 12.09 88.27 26.92 100.33 2.23 44.54 46.150 .77 18.59 46.72 45.15 48. C.55 65.Ash 34. Sep Eff.70 66.00 68.63 7.52 52.10 42. Ash Rej.33 12. From Bottom Wt.65 25. Ash Rej.13 67.59 18.25 19.63 Test 459 Concentrate Tails Wt. % +1mm 1 x .34 23.06 57.250 x .20 71.32 79.88 13. % +1mm 1 x .40 65.07 64.39 19.81 31.250 mm .45 50. C.47 23. Sep Eff.57 12.93 29.09 67.Ash 42. From Bottom Wt.34 86.51 17. Rec. % Ash % Cum F.52 46.88 36.5 mm .88 Test 458 Concentrate Tails Wt.14 7.02 88.66 34.31 12.32 73.47 58.50 89.12 100.00 12. Ash Rej.01 57.54 80.40 34.28 28.66 11.70 5.64 79.89 67.Ash F. C.

23 74.38 91.46 45. Ash Rej.52 Cum F.50 21. Ash Rej.53 80.77 59.Ash 40.20 83.68 59.56 48.35 13.54 13.250 x . Ash Rej.Ash 42.94 55.34 85.96 59.66 29.150 x .61 54.63 APPENDIX A TABULATED PLANT TEST DATA Wt.69 45.97 16.16 90.63 14.67 41.94 8.46 78.53 73.51 62.150 .250 mm .5 x .54 73.44 10.92 12.98 87.58 14.55 100.43 6.13 98.78 Cum F.59 49.32 54. C.08 53.80 Feed Wt.85 18.31 36.47 83.12 83. From Bottom Ash % Cum Wt.045 mm x 0 Cumulative 2.00 27.09 12. C.47 60.16 44.13 10.92 100.98 87.05 68.03 Concentrate Wt.045 .14 96.11 61.76 65.82 86.00 Wt.03 47. Ash Rej.91 77.05 78.55 88.82 62.04 Test 462 Tails Individual by Size Cum.33 100.045 .5 x . Ash Rej.62 40.64 59.Ash 36.64 12.06 10.61 80.11 77.61 23.97 6.29 79.01 14.68 62.80 43.29 59. Ash Rej.17 17.02 72.Ash 1.76 79.39 17. Sep Eff. Sep Eff.72 72.62 12. From Bottom Ash % Cum Wt.09 6.97 50.41 40.05 68.93 63.50 100.51 73.5 x .48 12.79 46. From Bottom Ash % Cum Wt.30 86.44 11.01 57. Sep Eff.5 mm .66 13.55 4.22 12.69 23.73 6.16 40.150 x .90 34.11 42.84 30. % 1.10 100. Rec.02 13.93 73.22 69.52 62. F.22 69.23 49.22 20.Ash Yield Yield 38.00 78.04 44.62 80.01 66.84 17.95 7.98 79.20 81. C.03 15.03 72.52 86.52 10.89 44.00 81. Rec.55 88.90 83.80 85.08 44.79 66.07 85.02 5.41 78.99 48.Ash 0.93 87.43 34.67 10.66 80.93 Page 124 .05 68.03 25.28 56.34 46.60 83.82 59.43 49.52 48.94 14.64 77.62 80.09 81.56 20.12 75.41 13.47 38.77 42.07 77.00 Wt.250 mm .59 100.20 76.10 77.49 47.00 35.78 16.74 10.05 100.36 52.53 80.32 13.48 14.51 85.04 21.93 33.Dennis I.66 9.250 x .Ash Yield Yield 34.17 9.30 69.74 55.85 31.00 11. % 2.99 100.150 x .10 49.64 59.00 29.045 mm x 0 Cumulative 3. Wt.49 61.52 86. C.54 52. Rec. C.63 59.72 49.07 58.55 53.51 79.97 30.51 87.65 52.47 28.49 13.93 43.13 35. Rec. Sep Eff.79 68.51 68.07 10.02 26.55 33. F. % Ash % +1mm 1 x .81 40.93 50. Sep Eff. Phillips Feed Wt.045 mm x 0 Cumulative 2. % Ash % Cum F.57 79.52 77.49 81.57 67.97 75.80 69.75 9.45 11. % Ash % Cum F.23 61.66 86.38 8.63 12.30 63.98 43.14 76.10 49.23 61.89 54. % Ash % Cum F.97 9.79 70.64 12.51 73.15 50.71 98.27 12. Sep Eff.96 75.91 81.87 23. Wt.04 72.66 71.94 62.03 42.08 75.24 40.00 12.26 29.89 19.87 36.Ash 0.045 .15 5.150 .32 24.73 73.47 78.30 75.79 4.81 62.60 82. % Ash % +1mm 1 x .82 86.87 Concentrate Wt.35 85.09 79.250 mm .35 66.77 67.70 79.49 42.42 85.75 14.73 58.58 52.40 12.86 11.05 17.93 45.49 9. Wt. Rec.24 45.31 74.250 x .30 83.25 40.88 82.59 39. C.16 48.94 11.00 11.91 63.14 49.06 57.06 42.52 62.17 13.53 86.65 94.24 Feed Wt.70 77.39 22.33 13.00 11.29 4.84 83.81 59.67 49.25 Concentrate Wt.74 46.56 22.95 51.150 .31 Cum F.01 62. Rec.93 69.98 13.11 11.00 Test 460 Tails Individual by Size Cum.75 65.16 Test 461 Tails Individual by Size Cum.59 100.92 99.52 27.03 48.82 42.35 85.5 mm .08 75.12 27.29 79.94 82.46 34.85 15.84 24.11 61.77 68.5 mm . % 2.98 18.20 62.55 13.54 5.74 22.43 61.Ash Yield Yield 31.36 77.02 62.24 74.25 14.92 13.02 65.39 13. % Ash % +1mm 1 x . F.

Ash 52.15 12.80 12.74 82.88 11.06 12.01 74.12 84.79 47.86 4.95 81.150 10.40 13.12 80. Rec.09 42.49 100.56 74.17 93.84 22.08 10.15 75.045 mm x 0 30.00 11.91 64.28 1 x .150 9.40 16.17 83.06 64. Sep Eff.58 98.78 81.74 48.48 67.32 84.17 19.250 x . C.97 94.02 57. C.02 45.55 74.Ash 1.70 44.96 3.61 86.40 52.55 79.61 43.35 71.54 .40 38.70 52.86 94.89 87.60 55.Ash 57.06 16.43 1 x .09 63.94 82.63 67.09 76.59 26.Ash 45.50 15.36 82.30 62.61 86.44 42.Dennis I.33 14.25 79.15 85.30 41.45 45.09 52.44 51.52 84.97 88.42 13.27 60.31 14.250 x .01 51.60 Cum F.10 70.80 78.93 9.75 100.13 25.09 76. Yield Yield 13.85 9.49 11.57 89.10 63.5 mm 25. Wt.99 82.18 Concentrate Wt.25 57. Rec.76 95.14 14.91 75.70 69.71 Page 125 .98 .52 55.90 85.150 x .62 90.99 Test 464 Tails Ash % Cum F.39 62.Ash 66.97 79.73 77.23 67. Yield Yield 35.09 44.06 60.88 85.40 65.19 4.79 78.69 50.26 48.80 15.07 43.45 46.58 67.36 APPENDIX A TABULATED PLANT TEST DATA Wt. Sep Eff. Sep Eff.11 56.50 23.06 30.045 14. Sep Eff. % Ash % Cum F.96 62.20 66.98 67.88 44. C.90 63.96 12.150 x . % Ash % Cum F.21 84.46 81.83 46.55 84. Sep Eff.5 x .99 Cum F.66 44.61 83.55 87.10 61.11 15.78 77. C.09 63.85 47.83 51.79 15. Ash Rej.00 38.90 12.36 79.68 75.Ash 0.11 52.05 11.74 62.70 37. From Bottom Wt.75 45. % Ash % +1mm 4.00 12.12 70.18 45.10 75.90 57. % 5.45 78.72 45.23 66.10 61. Wt.250 mm 18.92 83. Ash Rej.01 83.11 100.32 38.62 39.43 11.23 Concentrate Wt. Yield Yield 42.81 48. Ash Rej.67 67.80 78.045 mm x 0 29.46 42. Rec. Rec.47 75. Rec. % 8.05 14.50 13.12 52.94 82.94 33.045 mm x 0 32. % 4.44 86.82 52.10 76. Sep Eff. Ash Rej.60 18.01 48.09 5.62 5.99 41.80 80.05 75.11 84.93 68.85 31.23 40.36 23.82 23.39 79. Rec.74 82.150 9.00 Wt. Wt.11 11.14 63.250 x .25 71.81 76.47 68.54 45.250 mm 16.54 52.77 43.73 62.65 67.37 41.27 60.57 76.42 81.99 58.33 65.90 87.21 15.5 mm 23.00 45.42 88.77 80.21 83.17 .07 82.02 Feed Wt.33 65.15 15.74 61.07 52.75 67.44 Feed Wt.41 95. % Ash % +1mm 3.23 .27 36.74 15.03 82.63 .52 84.20 66.46 87.15 85.79 51.43 86.79 6.57 77.55 87.87 50.53 Individual by Size Cum.48 .48 11.14 13.97 65.03 12.60 27.39 1 x .22 49. % Ash % Cum F.65 67.84 66.68 21.08 Cum F.07 54.Ash 40. From Bottom Wt.06 Individual by Size Cum.2 .36 10.69 84.81 51.10 54.49 100.41 7.37 83.03 74.82 86.68 71.77 58. From Bottom Wt.88 Cumulative 100.37 81.84 100.16 15.05 21.00 82.72 12.9 Cumulative 100.37 10. C.08 19. % Ash % +1mm 3.11 86.00 Test 463 Tails Ash % Cum F.25 63.150 x .00 12.36 43.00 Wt.11 84.50 41.38 70.95 .85 Test 465 Tails Ash % Cum F.06 32.72 44.64 47.53 70.5 mm 20.75 80.98 5.42 51.19 45.81 45.05 81.85 8.35 .35 86.63 85.77 35.78 .74 100.250 mm 17.06 78.05 70.49 14.81 55.5 x .53 .68 70.92 50.13 6. Phillips Feed Wt.66 Cumulative 100.72 66.75 80.5 x .34 80.99 42.Ash 0. C.Ash 42.10 Individual by Size Cum.00 12.00 40.07 66.46 Concentrate Wt.045 14.88 49.62 . Ash Rej.94 48.91 82.54 4.99 35.09 81.23 12.36 21.97 80.12 87.14 20.045 14.84 53. Ash Rej.44 78.

Phillips APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 126 .Dennis I.

Phillips Design .0804 PRESS 523.28 15.8549 Source R-Squared Linear 0.8407 Quadratic 0. or equivalently maximizing the "PRED R-SQR".06 F Value Prob > F 21.11 1.Dennis I.0001 0.000 88722.0000: PRESS statistic not defined Predicted R-Squared 0.25x0 Rec.00 200.93 157.81 "Model Summary Statistics": Focus on the model minimizing the "PRESS".93 1431.55 3.00 12.14 DF 1 3 6 6 0 16 Mean Square 87165.00 +1 Level 45.00 8. Name Feed Rate Frother Diesel Units kg/min ml/min g/ton -.00 *** WARNING: The Cubic Model is Aliased! *** Sequential Model Sum of Squares Source Mean Linear Quadratic Cubic Residual Total Sum of Squares 87165.9420 Cubic Case(s) with leverage of 1.88 Adjusted R-Squared 0.6635 0.13 "Sequential Model Sum of Squares": Select the highest order polynomial where the additional terms are significant.5mm Type Numeric Numeric Numeric -1 Level 25.18 436.75 < 0.2578 5545.18 1308.36 0.68 90. Model Summary Statistics Root MSE 4.31 26. APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 127 .Expert Analysis Response: Factor A B C First Series (401-416) .00 2000.8009 0.

81 5.02 15.32 -4.81 DF 9 6 15 Coefficient Factor Estimate Intercept 77.78 9.Dennis I.00 200.5mm Type Numeric Numeric Numeric -1 Level 25.9658 0.95 A-Feed Rate -12.42 +4.88 -0.66 3.38 2 B -1.38 = *A *B *C * A2 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 128 .42 BC 7.78 -0.82 Prob > F 0.25x0 Rec.14 6.97 5.96 15.22 12.1655 0.70 AC -0.50 AB 10.0130 0.06 F Value 10.00 +1 Level 45.59 1.8549 0.9677 0.89 -0.36 1556. +77. Phillips Response: .60 7.26 1431.4067 0.045 1.88 73. PRESS 3.90 11.80 5.5684 7.43 23.914 Desire > 4 ANOVA for Response Surface Quadratic Model Source Model Residual Cor Total Sum of Squares 1466.05 DF 1 1 1 1 1 1 1 1 1 1 Standard Error 3.00 2000.0045 R-Squared Adj R-Squared Pred R-Squared Adeq Precision 0.27 C-Diesel 5.42 B-Frother 4.9420 0.0804 9.24 11.00 8.V.1787 0.27 +5.09 2 C -0. Factor A B C Name Feed Rate Frother Diesel Units kg/min ml/min g/ton -.56 2.042 0.96 Root MSE Dep Mean C.95 -12.90 Final Equation in Terms of Coded Factors: .00 Mean Square 162.73 7.00 12.69 t for H0 Coeff=0 Prob > |t| VIF -3.53 5.49 1.60 0.60 90.4637 0.52 0.32 2 A -4.53 13.58 -0.4140 0.22 10.25x0 Rec.

524 0.044 -0.108 3.748 -0.61 -3.906 -0.70 84.717 7 -0.48 -15.51 64.488 6 0.47 -0.17 0.10 0.70 0.14 69.243 0.139 0.454 0.16 79.72 77.065 9 -1.030 -0.718 -0.44 0.05 * B2 * C2 *A*B *A*C *B*C Final Equation in Terms of Actual Factors: .29 80.90 85.000 0.79 62.853 -1.98 58.186 0.05 0.321 13 0.262 12 -1.388 0.19 0.27 -6.867 0.Dennis I.88 68.73 0.008 0.15 0.87 87.09 -0.12 84.399 0.750 0. Phillips -1.742 -5.142 11 Page 129 .78 60.74 72.320 10 0.301 -1.072 -1.012 0.917E-03 = * Feed Rate * Frother * Diesel * Feed Rate2 * Frother2 * Diesel 2 * Feed Rate * Frother * Feed Rate * Diesel * Frother * Diesel Diagnostics Case Statistics Standard Order 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Actual Value 74.05 0.84 59.91 80.14 80.522 0.960 0.646E-05 +3.75 62.50 +10.62 0.67 82.05 0.17 87.199 -0.174 14 0.001 0.42 +7.050 1.258 -0.045 2 0.521 1.411 1.610 -4.62 56.154E-07 +0.347 0.96 0.171 1 0.134 -0.904 0.04 Student Residual Leverage Residual 5.005 0.109 5.01 79.25x0 Rec.099 8 1.001 0.798 0.501 0.53 -4.56 0.096 0.168 15 -0.182 5 -0.83 63.22 0.349 0.65 85.640 -1.91 79.916 4 -0.14 0.90 0.393 0.575 16 -2.70 70.002 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Outlier Run t Order 2.48 Predicted Value 69.33 79.70 -0.155 Cook's Distance 0.634 3 0.995 -0.357 0.540 0.05 82.140 0.594 -0.64 73. +240.57 56.

00 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 100 80 . Phillips 3D Response Plots First Series (401-416) 0. Actual Constants: Frother = 12.00 1550 38.33 1100 Diesel 31.25x0 Rec.00 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 130 .0 60 40 20 0 2000 45.0 60 40 20 0 2000 45.0 60 40 20 0 2000 45.Dennis I.25 mm x 0 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 100 80 .33 1100 Diesel 31.00 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 100 80 .00 1550 38.25x0 Rec.25x0 Rec.67 650 Feed Rate 200 25. Actual Constants: Frother = 10.33 1100 Diesel 31. Actual Constants: Frother = 8.67 650 Feed Rate 200 25.67 650 Feed Rate 200 25.00 1550 38.

69 0.9689 0.8330 0.25 Rec.5mm Type Numeric Numeric Numeric -1 Level 25.5x. or equivalently maximizing the "PRED R-SQR".Expert Analysis Response: Factor A B C First Series (401-416) .59 2394.52 0.00 200. APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 131 .0480 3058. Model Summary Statistics Root Adjusted Predicted Source MSE R-Squared R-Squared R-Squared Linear 10.000 48931.0001 0. Phillips Design .00 12.56 268.00 2000.59 7184. Name Feed Rate Frother Diesel Units kg/min ml/min g/ton -.95 4.00 *** WARNING: The Cubic Model is Aliased! *** Sequential Model Sum of Squares Source Mean Linear Quadratic Cubic Residual Total Sum of Squares 40307.00 +1 Level 45.00 8.83 DF 1 3 6 6 0 16 Mean Square 40307.75 F Value Prob > F 19.26 44.9222 0.6434 Quadratic 6.72 195.0000: PRESS statistic not defined PRESS 3075.16 1171.24 "Sequential Model Sum of Squares": Select the highest order polynomial where the additional terms are significant.81 8614.36 < 0.0011 Cubic Case(s) with leverage of 1.95 0.65 "Model Summary Statistics": Focus on the model minimizing the "PRESS".Dennis I.7913 0.

54 -0.91 -14.0628 0.26 +11.00 Mean Square 928.53 5.0007 R-Squared Adj R-Squared Pred R-Squared Adeq Precision 0.9222 0.19 13.65 Factor Intercept A-Feed Rate B-Frother C-Diesel A2 B2 C2 AB AC BC DF 9 6 15 Coefficient Estimate 72.00 12.29 9.91 2.0011 14.0059 0.21 11.0752 0.87 -0.28 34.13 19.75 Prob > F 0.72 268.71 58.232 Desire > 4 ANOVA for Response Surface Quadratic Model Source Model Residual Cor Total Sum of Squares 8355.60 7.00 2000.V.00 +1 Level 45.91 0.22 10.02 -7.90 Final Equation in Terms of Coded Factors: .25 Rec.18 2.73 7.52 8624.5x.50 2.59 DF 1 1 1 1 1 1 1 1 1 1 Standard Error 6.26 11.28 4.01 +42.15 2.43 23.87 = *A *B *C * A2 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 132 .20 2.02 15.5mm Type Numeric Numeric Numeric -1 Level 25.25 Rec.83 10.8491 0.0464 0.15 t for H0 Coeff=0 Prob > |t| VIF -1.0268 0.31 6.00 200. +72.53 13.64 2.69 50.24 Root MSE Dep Mean C.42 41.2814 0.05 42.1738 0.01 42. Phillips Response: . PRESS 6.0271 7.13 4.9689 0.02 -7.68 17.33 8614. Factor A B C Name Feed Rate Frother Diesel Units kg/min ml/min g/ton -.91 -14.00 8.75 F Value 20.41 44.22 12.07 20.5x.Dennis I.17 -1.

004 0.411 36.214 -0.42 -5.882 15.116 -0.57 50.105 0.723 0.188 0.64 0.992 0.127 0.990 1.24 -1.306 0.80 65.995 22.34 0.234 -1.706 1.409 1.5x.28 +34.108 1.76 -0.016 -0.130 -0.81 60.42 +41.424 1.22 83.70 0.56 8.283 0.53 1.393 8.745E-03 +0.49 6.59 0.544E-03 0.454 Student Cook's Outlier Residual Distance t 1.55 17.521 77.263 0.012 0.000 0.606 1.96 -1.35 0.65 77.04 20.031 -0.02 7.388 -0.11 +5.49 Predicted Value Residual Leverage 38.501 13.941 0.172 -0.145 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Run Order 3 9 14 5 1 7 2 10 8 15 16 12 4 6 13 11 Page 133 .35 -12.00 69.922 -0.89 14.59 * B2 * C2 *A*B *A*C *B*C Final Equation in Terms of Actual Factors: .292 0.36 43.06 0.002 0.159 0.640 27.287 0.388 75.931 -0.31 0. Phillips -0.45 -0.496 0.18 -0.853 83.83 0.258 72.370 -0.77 -2.25 Rec.45 -9.20 60.79 0.59 0.718 49.56 -87.32 78.033 = * Feed Rate * Frother * Diesel * Feed Rate2 * Frother2 * Diesel 2 * Feed Rate * Frother * Feed Rate * Diesel * Frother * Diesel Diagnostics Case Statistics Standard Order 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Actual Value 45.38 0.70 +4.25 0.862 -0.142 1.05 +42.429 0.798 66.904 56.Dennis I.18 7.15 -0. +995.085 -0.009 -0.42 0.99 61.008 -0.84 0.71 +58.128 -2.38 2.46 -0.301 60.00 47.960 65.05 -2.906 49.02 0.42 0.56 -0.463 -0.006 -0.096E-05 +1.004 0.255 0.41 4.260 0.

00 1550 40.00 1550 40.00 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 134 .0 100 50 0 2000 45. Actual Constants: Frother = 10.0 136 64 -8 2000 45.00 200 Feed Rate 25. Phillips 3D Response Plots First Series (401-416) 0.5x.25 mm DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 200 150 .00 200 Feed Rate 25.00 200 Feed Rate 25.0 100 50 0 2000 45.00 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 200 150 .00 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 280 208 .25 Rec. Actual Constants: Frother = 8.00 1100 Diesel 35.25 Rec.00 650 30.00 1100 Diesel 35.5x.00 650 30.5x. Actual Constants: Frother = 12.00 1100 Diesel 35.25 Rec.00 650 30.5 x 0.00 1550 40.Dennis I.

57 "Model Summary Statistics": Focus on the model minimizing the "PRESS".57 0.47 596.47 1788.00 +1 Level 45.9421 0.Expert Analysis Response: Factor A B C First Series (401-416) .00 8.86 F Value Prob > F 19.8275 0.00 *** WARNING: The Cubic Model is Aliased! *** Sequential Model Sum of Squares Source Mean Linear Quadratic Cubic Residual Total Sum of Squares 75948.61 247. or equivalently maximizing the "PRED R-SQR".2135 4881.5mm Type Numeric Numeric Numeric -1 Level 25.81 DF 1 3 6 6 0 16 Mean Square 75948.0000: PRESS statistic not defined Predicted R-Squared 0.00 2000.0001 0. Phillips Design .19 1.98 < 0.9196 PRESS 798.8552 Cubic Case(s) with leverage of 1.00 12.57 0.00 200.5mmx0 Rec.6304 -2. APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 135 . Name Feed Rate Frother Diesel Units kg/min ml/min g/ton -.20 41.86 "Sequential Model Sum of Squares": Select the highest order polynomial where the additional terms are significant.000 78109. Model Summary Statistics Root Adjusted Source MSE R-Squared R-Squared Linear 5.16 0.57 125.7844 Quadratic 4.26 20.83 8471.Dennis I.

6250 0.46 13.65 +4.86 F Value 10.00 +1 Level 45.04 -3.4114 0.9421 0.22 10.87 C-Diesel 14. Factor A B C Name Feed Rate Frother Diesel Units kg/min ml/min g/ton -.00 -0.53 5. PRESS 4.Dennis I.00 2000.53 0.02 6.5mmx0 Rec.02 15.V.73 7.63 8471.0926 0.59 A-Feed Rate -9.5mm Type Numeric Numeric Numeric -1 Level 25.00 8.30 1.98 11.04 2 A -3.27 0.39 2 B -0.43 23.18 125.1901 0.88 1.30 7.76 t for H0 Coeff=0 Prob > |t| -2.276 Desire > 4 ANOVA for Response Surface Quadratic Model Source Model Residual Cor Total Sum of Squares 2036. +74.9196 10.1965 VIF 7.66 AB 18.4475 0.57 68.65 13.39 = *A *B *C * A2 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 136 .78 2 C 10.0045 R-Squared Adj R-Squared Pred R-Squared Adeq Precision 0.8552 -2.65 B-Frother 4.45 0.19 3.34 Root MSE Dep Mean C.24 20.85 Prob > F 0.59 -9.60 7.6128 0.52 -0.00 Mean Square 226.57 DF 9 6 15 Coefficient Factor Estimate Intercept 74. Phillips Response: .53 13.12 7.48 2.00 200.0607 0.16 2161.31 4.58 1.98 DF 1 1 1 1 1 1 1 1 1 1 Standard Error 4.81 2.87 +14.28 BC 19.00 12.5mmx0 Rec.0637 0.90 6.22 12.11 AC 10.90 Final Equation in Terms of Coded Factors: .

53 0.301 -0.049 0.228 2.459 6 0.62 62.252 1 0.01 0.585 0.478 0.082 8 1.807 0.16 48.411 1.16 86.034 -0.550 14 0.68 60.89 53.16 -0.258 1.46 48.960 -4.246 15 -0.98 * B2 * C2 *A*B *A*C *B*C Final Equation in Terms of Actual Factors: .824 0. Phillips -0.70 0.933 0.14 54.43 0.77 0.853 0.904 1.91 +1.015 0.422 -1.380 -1.640 -4.462 0.03 0.156 11 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 137 .440 0.99 74.521 -1.22 72.197 0.07 0.143E-03 +0.69 Residual Leverage 6.582 0.388 -1.198 -0.28 +19.995 0.982 7 0.192 0.547 5 -1.00 80.74 83.050 -1.196 0.70 0.985 0.155 1.798 3.71 0.Dennis I.03 62.21 83.11 +10.72 58.445 10 -0.062 -1.171 0.339 9 -0.338 35.199 13 -0.501 -0.18 0.011 = * Feed Rate * Frother * Diesel * Feed Rate2 * Frother2 * Diesel 2 * Feed Rate * Frother * Feed Rate * Diesel * Frother * Diesel Diagnostics Case Statistics Standard Order 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Actual Value 70.19 85.367 0.88 74. +504.67 55.250 4 -1.454 Student Cook's Outlier Run Residual Distance t Order 1.92 75.57 0.78 +10.906 -4.40 0.408 2 0.00 68.393 -0.5mmx0 Rec.002 -0.44 54.29 61.921 16 -1.837 -0.84 -8.19 +1.33 0.52 70.44 85.31 0.43 84.46 54.93 66.71 78.66 +18.090 0.70 78.11 Predicted Value 64.663 12 -1.90 -37.316E-05 +0.009 0.82 70.003 -0.58 -0.005 0.718 1.31 0.49 75.909 0.443 3 0.58 0.217 0.

Actual Constants: Frother = 10.Dennis I.5 mm x 0 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 150 120 .00 1100 Diesel 35.5mmx0 Rec.00 650 30.0 90 60 30 0 2000 45.0 90 60 30 0 2000 45.0 90 60 30 0 2000 45.00 200 Feed Rate 25.00 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 150 120 . Actual Constants: Frother = 12. Actual Constants: Frother = 8.00 200 Feed Rate 25.00 1550 40.00 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 138 .00 650 30.5mmx0 Rec.00 1550 40.00 1100 Diesel 35.5mmx0 Rec.00 1100 Diesel 35. Phillips 3D Response Plots First Series (401-416) 0.00 1550 40.00 200 Feed Rate 25.00 DESIGN EXPERT Plot Actual Factors: X = Feed Rate Y = Diesel 150 120 .00 650 30.

86 565.9466 0.25mmx0 Rec Name Feed GPM Air Vg Frother Units gpm cm/sec ml/min Type Numeric Numeric Numeric -1 Level 40. APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 139 .8264 -0.30 *** WARNING: The Cubic Model is Aliased! *** Sequential Model Sum of Squares Sum of Source Squares DF Mean 98575.0168 0. or equivalently maximizing the "PRED R-SQR".0000: PRESS statistic not defined PRESS 461.0413 0.78 8.33 Cubic 0.10 5.74 Quadratic 14.67 3 Quadratic 148.03 0.17 292. Phillips Design .0000 0.74 4.22 24.05 15 Mean Square 98575.08 4.89 0.0309 8.19 35.67 F Value Prob > F 2.13 Prob > F 0.1555 0.45 1 Blocks 1.19 1 Linear 105.60 +1 Level 60.0309 "Sequential Model Sum of Squares": Select the highest order polynomial where the additional terms are significant.Expert Analysis Response: Factor A B C Second Series (451-465) .450E-03 6589. Lack of Fit Tests Sum of Source Squares Linear 162.2118 -0.36 "Model Summary Statistics": Focus on the model minimizing the "PRESS".450E-03 DF 9 3 0 1 Mean Square 18.9996 Case(s) with leverage of 1.00 1.41 6 Cubic 14.00 1.13 0.78 F Value 2139.0892 Cubic 0.000 Pure Error 8. Model Summary Statistics Root Adjusted Predicted Source MSE R-Squared R-Squared R-Squared Linear 4.450E-03 1 Total 98845.16 6.33 3 Residual 8.45 1.3937 0.092 1.7181 Quadratic 1.90 565.450E-03 "Lack of Fit Tests": Want the selected model to have insignificant lack-of-fit.30 9.Dennis I.

12 0.30 F Value Prob > F 7.15 -3.54 2.49 -4.201 Desire > 4 ANOVA for Response Surface Quadratic Model Source Block Model Residual Lack of Fit Pure Error Cor Total Sum of Squares 1.60 1.73 1.19 28.95 1.23 3.87 1.14 Mean Square 1.61 1.51 0.0590 0.86 1.86 1.22 -3.48 -1.08 14.55 -2. Phillips Response: .Dennis I.58 2.33 8.60 +1 Level 60.20 -1.58 4.29 0.2097 0.25mmx0 Rec = +83.24 1.79 1.25mmx0 Rec Factor A B C Name Feed GPM Air Vg Frother Units gpm cm/sec ml/min Type Numeric Numeric Numeric -1 Level 40.89 81.59 10.49 2.8264 -0.0892 11.450E-03 DF 1 9 4 3 1 14 R-Squared Adj R-Squared Pred R-Squared Adeq Precision DF 1 1 1 1 1 1 1 1 1 1 1 Standard Error 1.07 2.0309 0.26 -3.00 1.34 292.26 3.25 * A APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 140 .00 1.6867 0.54 -2.0311 565.V.88 0.64 -3.19 254.25 -1.9466 0.1957 0.09 -1.3801 1.13 0.33 14.2963 0.450E-03 269.60 Root MSE Dep Mean C.78 8.97 1.0254 0.64 -3.07 Final Equation in Terms of Coded Factors: .10 5.75 t for H0 Coeff=0 Prob > |t| VIF 0.93 0.0660 0. PRESS 1.43 -1.36 Factor Intercept Block 1 Block 2 A-Feed GPM B-Air Vg C-Frother A2 B2 C2 AB AC BC Coefficient Estimate 83.99 0.31 1.90 6.1009 0.91 0.62 -2.30 9.09 3.

585 0.103 11 0.847 0.15 0.36 81.339 2 -0.650 12 83.05 * Feed GPM +25.57 0.47 -1.66 * Air Vg2 -0.11 0.337 13 84.033 * Feed GPM2 +53.833 7 Page 141 .631 6 82.47 80.54 0.70 82.20 81.086 14 2. Phillips -1.62 0.23 -0.49 -4.321 -0.38 82.784 15 77.026 0.12 86.14 *B *C * A2 * B2 * C2 *A*B *A*C *B*C Final Equation in Terms of Actual Factors: .337 0.77 0.72 79.09 +3.14 0.064 0.35 0.659 -1.55 0.074 2 79.22 -3.63 0.11 0.36 * Air Vg +13.106 0.384 10 80.36 0.61 85.21 79.73 * Frother2 -4.07 -1.07 -1.144 0.53 0.527 7 86.798 -0.005 9 0.909 15 -0.67 0.Dennis I.143 4 -0.898 1.748 3 84.862 -0.102 14 77.54 -2.15 * Air Vg * Frother Diagnostics Case Statistics Standard Actual Predicted Student Order Value Value Residual Leverage Residual 1 81.002 0.625 4 81.569 13 -1.231 0.84 +3.84 0.004 11 84.223 5 82.22 * Feed GPM * Air Vg -0.595 5 -0.867 Note: Predicted values include block corrections.738 8 -0.29 +0.350 1.974 0.576 1 0.172 0.47 -0.97 +1.735 0.872 0.22 0.90 0.24 0.12 78.59 2.93 68.53 -0.82 -0.473 12 -0.296 3 1.17 83.57 83.07 -0.339 6 -1.718 9 67.100 8 82.699 10 -0.032 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Outlier Run t Order 1.848 1.25mmx0 Rec = -41.321 -0.655 -0. Cook's Distance 0.58 * Frother +0.21 * Feed GPM * Frother +6.09 -1.097 2.177 0.984 -0.890 -1.85 79.79 0.58 83.962 -0.140 1.68 0.14 -0.281 0.

3 60.1 Feed GPM 40.0 1.3 55.3 55.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Air Vg .1 Feed GPM 40.0 1.0 1.0 1.3 100 90 80 70 60 50 40 1.0 1.25mmx0 Rec Actual Constants: Frother = 7.2 45.0 1.0 1.0 1.6 100 90 80 70 60 50 40 1.2 45.25mmx0 Rec Actual Constants: Frother = 9.0 1.Dennis I.2 Air Vg 50.25 mm x 0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Air Vg .3 55.0 1.4 100 90 80 70 60 50 40 1.2 Air Vg 50.2 45.3 60.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Air Vg .0 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 142 .0 1.1 Feed GPM 40.25mmx0 Rec Actual Constants: Frother = 5.2 Air Vg 50. Phillips 3D Response Plots Second Series (451-465) 0.3 60.0 1.

46 4729.5mmx.31 1 Linear 977.57 3.83 3 Quadratic 1763.25Rec APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 143 .41 0.2575 0.96 15 Mean Square 67648.11 0.11 Prob > F 0.3201 0.75 42.46 1 Total 70946.30 9.46 DF 9 3 0 1 Mean Square 230.1109 0. Phillips Design .93 1 Blocks 243.6661 Cubic 1.00 1.0801 0.46 "Lack of Fit Tests": Want the selected model to have insignificant lack-of-fit.1161 Quadratic 8.85 103.000 2.8765 4648.Expert Analysis Response: Factor A B C Second Series (451-465) .49 103.0000: PRESS statistic not defined PRESS -0.9895 Case(s) with leverage of 1.9992 0.42 311.09 6 Cubic 311.78 F Value 93. Lack of Fit Tests Source Linear Quadratic Cubic Pure Error Sum of Squares 2074.8973 0.93 243.1127 "Sequential Model Sum of Squares": Select the highest order polynomial where the additional terms are significant.31 325.97 Response: .5mmx.10 5.86 0.54 42.5218 -4.60 +1 Level 60.34 3 Residual 2.1127 2.25Rec Name Feed GPM Air Vg Frother Units gpm cm/sec ml/min Type Numeric Numeric Numeric -1 Level 40.94 293.34 0.57 0.67 17950.Dennis I. Model Summary Statistics Root Adjusted Predicted SourceMSER-Squared R-Squared R-Squared Linear 14.80 F Value Prob > F 1.78 2.00 1.30 *** WARNING: The Cubic Model is Aliased! *** Sequential Model Sum of Squares Sum of Source Squares DF Mean 67648.

11 0.86 67. PRESS 8.5239 0.98 C-Frother -7.88 0.0990 0.90 6.8725 0.55 Block 1 4.V.47 B-Air Vg -2.07 Final Equation in Terms of Coded Factors: .46 3298.6661 -4.8765 7.47 * A -2.10 5.55 -0.56 -0.55 78.1940 0.16 13.78 A-Feed GPM -0.50 7.92 313.42 6.24 -2.92 AC -18.20 0.86 1.79 1.5mmx.37 4. Phillips Factor A B C Name Feed GPM Air Vg Frother Units gpm cm/sec ml/min Type Numeric Numeric Numeric -1 Level 40.18 1 1 1 1 1 1 1 1 1 1 4.78 2.1043 0.13 5.40 t for H0 Coeff=0 Prob > |t| VIF -0.0435 0.28 AB -16.1022 42.30 9.2947 1.1127 R-Squared Adj R-Squared Pred R-Squared Adeq Precision 0.74 5.97 Coefficient Factor Estimate Intercept 61.78 Block 2 -4.27 3.33 A2 17.00 1.09 1.Dennis I.86 1.54 8.86 2 B 9.31 2740.70 -2.46 F Value Prob > F 3.59 10.31 304.02 2 C -1.80 311.03 Root MSE Dep Mean C.14 2.8973 0.0884 0.54 2.11 -0.886 Desire > 4 ANOVA for Response Surface Quadratic Model Source Block Model Residual Lack of Fit Pure Error Cor Total Sum of Squares 243.60 +1 Level 60.17 -2.31 BC 6.79 7.9187 0.25Rec = +61.91 1.98 * B APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 144 .58 2.95 1.30 Mean Square 243.51 DF 1 9 4 3 1 14 Standard DF Error 1 8.45 103.34 2.00 1.19 17950.

12 0.898 -0.40 5.891 12 60.07 0.11 0.236 2 -2.47 -8.350 1.092 14 2.86 +9.862 -0.157 7 Page 145 .27 29.15 0.57 11.99 +35.73 0.802 7 80.39 0.859 10 80.95 75.38 65.872 -0.127 11 0.49 0.142 0.88 -0.25Rec +784.735 0.321 -1.787 1 -0.57 -3.21 78.829 13 70.47 -5.31 +6.478 10 -0.22 1.655 1.270 5 61.659 0.028 0.764 8 -1.90 0.42 83.89 -1. Cook's Distance 0.88 71.92 -0.146 2 77.85 0.02 -1.89 0.235 9 0.013 0.97 56.82 1.59 74. Phillips -7.50 0.269 11 75.321 -0.18 +901.81 4.398 0.823 6 0.47 63.862 5 -0.91 +0.011 0.699 6 87.002 0.5mmx.044 14 48.34 +9.596 9 31.20 -2.196 0.58 54.295 15 0.77 -1610.37 -16.067 8 67.125 1.962 0.746 13 -0.685 0.88 -0.06 89.51 *C * A2 * B2 * C2 *A*B *A*C *B*C Final Equation in Terms of Actual Factors: .984 -1.54 0.07 75.60 -2.111 Note: Predicted values include block corrections.798 -0.847 0.890 -0.225 0.038 4 -0.Dennis I.36 54.389 0.053 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Outlier Run t Order -0.92 -18.80 0.33 +17.14 70.848 -0.791 4 73.40 0.054 16.28 -16.789 3 -0.50 +8.55 0.99 63.758 12 1.155 0.39 -0.200 0.807 15 46.532 3 80.78 1.26 -0.18 = * Feed GPM * Air Vg * Frother * Feed GPM2 * Air Vg2 * Frother2 * Feed GPM * Air Vg * Feed GPM * Frother * Air Vg * Frother Diagnostics Case Statistics Standard Actual Predicted Student Order Value Value Residual Leverage Residual 1 65.

3 60.0 1.5 150 125 100 75 50 25 0 1. Phillips 3D Response Plots Second Series (451-465) 0.5 x .0 1.3 150 125 100 75 50 25 0 1.2 Air Vg 50.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Air Vg .2 Air Vg 50.0 1.0 1.5mmx.3 55.2 45.6 150 125 100 75 50 25 0 1.3 55.0 1.5mmx.25Rec Actual Constants: Frother = 5.Dennis I.25 mm DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Air Vg .3 55.0 1.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Air Vg .1 Feed GPM 40.0 1.0 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 146 .1 Feed GPM 40.0 1.0 1.1 Feed GPM 40.3 60.0 1.25Rec Actual Constants: Frother = 9.5mmx.2 45.3 60.2 45.0 1.25Rec Actual Constants: Frother = 7.2 Air Vg 50.0 1.

Phillips Design .0604 0.08 10.6719 Cubic 0. Lack of Fit Tests Sum of Source Squares Linear 288.26 3 Quadratic 258. APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 147 .1506 -0.Expert Analysis Response: Factor A B C Second Series (451-465) .30 9.77 5.76 0.30 *** WARNING: The Cubic Model is Aliased! *** Sequential Model Sum of Squares Sum of Source Squares DF Mean 93991.09 10.3466 0.47 6 Cubic 30.2165 0.60 +1 Level 60.78 147.9998 0.0000: PRESS statistic not defined PRESS 777.5mmx0 Rec Name Feed GPM Air Vg Frother Units gpm cm/sec ml/min Type Numeric Numeric Numeric -1 Level 40.48 51.67 147.Dennis I.068 1 Total 94446.068 DF 9 3 0 1 Mean Square 32.15 739.79 Quadratic 30.0574 0.7578 Quadratic 2.0358 0.16 "Model Summary Statistics": Focus on the model minimizing the "PRESS". or equivalently maximizing the "PRED R-SQR".37 0.068 6296.01 15 Mean Square 93991.09 43.10 5.00 1.0604 "Sequential Model Sum of Squares": Select the highest order polynomial where the additional terms are significant.7767 -0.000 Pure Error 0.00 1.10 F Value 468.31 3 Residual 0.40 F Value Prob > F 1.9313 0.62 0.42 1 Blocks 12.068 Model Summary Statistics Root Adjusted Predicted Source MSE R-Squared R-Squared R-Squared Linear 5.26 0.48 1 Linear 153.42 12.10 0.31 Cubic 0.62 Prob > F 0.9980 Case(s) with leverage of 1.

58 2.54 2. Phillips Response: .9313 0.86 1.75 7.0919 0.0528 0.12 5.73 30.60 +1 Level 60.Dennis I.0529 0.5mmx0 Rec Factor A B C Name Feed GPM Air Vg Frother Units gpm cm/sec ml/min Type Numeric Numeric Numeric -1 Level 40.86 1.16 Factor Intercept Block 1 Block 2 A-Feed GPM B-Air Vg C-Frother A2 B2 C2 AB AC BC Mean Square 12.2818 1.98 -2.60 10.38 30.71 -6.70 -2.36 1.70 -2.72 1.72 0.79 1.76 79.0497 147.35 2.90 -1.759 Desire > 4 ANOVA for Response Surface Quadratic Model Source Block Model Residual Lack of Fit Pure Error Cor Total Sum of Squares 12.31 0.09 DF 1 1 1 1 1 1 1 1 1 1 1 Standard Error 2.1177 0.59 Root MSE Dep Mean C.V.90 6.07 1.6163 0.33 1.19 = *A *B *C * A2 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 148 .1001 0.00 1.21 1.22 -2.28 -1.24 0.68 -2.16 3.99 2. PRESS 2.30 F Value Prob > F 6.3101 0.0604 0.30 9.13 -1.068 DF 1 9 4 3 1 14 R-Squared Adj R-Squared Pred R-Squared Adeq Precision Coefficient Estimate 81.19 0.068 454.48 411.80 2.59 10.54 -1.6719 9.33 2.5mmx0 Rec +81.22 -4.07 Final Equation in Terms of Coded Factors: .90 -1.10 0.16 -2.39 -6.16 -2.12 +5.95 1.48 45.91 1.62 0.16 4.72 -2.2697 0.02 0.7767 -0.00 2.48 739.00 1.10 5.55 t for H0 Coeff=0 Prob > |t| VIF 1.

91 81. Cook's Outlier Run Distance t Order 0.99 0.520 3 0.890 -1.880 7 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 149 .20 0.57 83.241 -0.32 * Feed GPM * Frother +11.61 * Feed GPM -15.25 1.862 -0.133 2.35 78.939 1.052 * Feed GPM2 +97.10 0.321 -0.33 0.837 4 79.32 84.02 76.655 0.24 0.18 0.937 11 82.92 0.98 -2.14 0.04 0.34 85.09 * B2 * C2 *A*B *A*C *B*C Final Equation in Terms of Actual Factors: .189 0.36 -1.798 -0.93 62.15 0.334 7 85.872 0.114 4 0.069 0.75 81.55 81.920 3 83.153 -1.66 * Air Vg2 -0.082 14 74.00 +2. Phillips +0.035 -0.732 1 0.08 -1.779 6 83.03 -1.39 -6.906 Note: Predicted values include block corrections.39 76.11 -0.154 -0.321 -0.294 12 0.50 +4.69 0.984 -0.71 -6.515 15 0.27 0.647 9 61.743 15 74.425 0.962 -0.350 1.848 1.847 0.29 * Air Vg * Frother Diagnostics Case Statistics Standard Actual Predicted Student Order Value Value Residual Leverage Residual 1 79.88 81.918 9 0.867 12 80.79 * Frother2 -6.33 * Frother +0.84 77.87 81.66 0.003 0.85 0.95 0.028 2 78.798 13 1.253 5 80.55 76.68 1.395 -0.693 8 0.08 -2.898 1.024 -0.536 1.14 -0.06 0.90 * Air Vg +13.89 3.252 -0.575 13 82.23 0.897 10 0.Dennis I.38 -0.285 6 0.39 * Feed GPM * Air Vg -0.659 -0.26 -1.109 14 0.51 0.5mmx0 Rec = -52.038 11 0.11 80.187 -0.59 0.735 0.834 5 0.125 8 80.393 2 3.325 10 79.

0 1.0 1.5mmx0 Rec Actual Constants: Frother = 5. Phillips 3D Response Plots Second Series (451-465) 0.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Air Vg .0 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 150 .Dennis I.1 Feed GPM 40.0 1.3 60.0 1.3 55.0 1.0 1.2 45.1 Feed GPM 40.2 Air Vg 50.6 110 100 90 80 70 60 50 1.3 55.0 1.0 1.3 60.2 Air Vg 50.2 45.0 1.3 55.5 110 100 90 80 70 60 50 1.5 mm x 0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Air Vg .5mmx0 Rec Actual Constants: Frother = 9.2 Air Vg 50.3 110 100 90 80 70 60 50 1.0 1.2 45.0 1.1 Feed GPM 40.5mmx0 Rec Actual Constants: Frother = 7.3 60.0 1.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Air Vg .

32 "Model Summary Statistics": Focus on the model minimizing the "PRESS".00 5.50 0.1971 "Sequential Model Sum of Squares": Select the highest order polynomial where the additional terms are significant.39 3 32.7193 -5.7482 PRESS 461.45 Blocks 1.Dennis I.12 0. or equivalently maximizing the "PRED R-SQR".5mm) ml/min Type Numeric Numeric Numeric -1 Level 40.30 *** WARNING: The Cubic Model is Aliased! *** Sequential Model Sum of Squares Sum of Mean Source Squares DF Square Mean 98575.19 1 1.45 1 98575.00 700.0000: PRESS statistic not defined Predicted R-Squared -0.19 Linear 98.000 0 Total 98845.Expert Analysis Second Series .25mmx0 Rec Factor A B C Name Feed GPM Diesel Frother Units gpm g/T (-.95 Residual 0.1889 0. APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 151 .1765 Quadratic 2.00 9.8667 0.23 6 22.Revisited (451-465) Response: .80 Quadratic 134.60 +1 Level 60.3666 0. Phillips Design .05 15 6589.99 0.00 1400. Model Summary Statistics Root Adjusted Source MSE R-Squared R-Squared Linear 4.5667 Cubic Case(s) with leverage of 1.93 2.37 Cubic 35.79 4 8.67 F Value Prob > F 1.48 1811.

90 +0.2643 0.66 -1.3150 2.99 81.Dennis I.Revisited (451-465) Units gpm g/T (-.30 Mean Square 1.38 t for H0 Coeff=0 Prob > |t| VIF 1.63 35. Phillips Response: .84 2 C -5.44 -1.19 B-Diesel -0.00 1400.68 2.20 3.70 2.8237 0.19 232.3379 0.59 BC 2.31 AC 1.69 4.00 5.29 -2.30 -0.32 DF 1 9 4 14 Coefficient Factor Estimate Intercept 86.19 -0.7482 6.04 1.25 3.5mm) ml/min Type Numeric Numeric Numeric -1 Level 40.29 Block 1 4.89 0.5667 -5.00 700.87 DF 1 1 1 1 1 1 1 1 1 1 1 Standard Error 3.69 1811.7639 0.67 1.1597 R-Squared Adj R-Squared Pred R-Squared Adeq Precision 0.25mmx0 Rec Factor A B C Name Feed GPM Diesel Frother Second Series .42 6.09 0.90 A2 0.2236 0.28 Block 2 -4.75 2.85 4.32 1.86 18.19 25.94 2.3094 0.95 F Value Prob > F 2.8180 0.463 Desire > 4 ANOVA for Response Surface Quadratic Model Source Block Model Residual Cor Total Sum of Squares 1.V.60 Root MSE Dep Mean C. PRESS 2.79 269.07 3.25mmx0 Rec +86.63 C-Frother -1.66 2 B -1.5442 0.28 A-Feed GPM -2.15 0.54 7.00 9.88 22.29 AB -3.78 3.25 -0.60 +1 Level 60.05 Final Equation in Terms of Coded Factors: .85 8.24 -1.63 -1.66 = *A *B *C * A2 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 152 .8667 0.67 2.50 -1.16 0.

37 -1.47 84.581 8 82.299 10 80.31 +1.511 7 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 153 .649 -0.33 0.566 Note: Predicted values include block corrections.611 11 0.086 6 82.12 85.91 -1.877 -1.809 1.01 0.39 -1.58 83.725 3 84.019 -0.638 8 0.73 0.168 -1.426E-03 = * Feed GPM * Diesel * Frother * Feed GPM2 * Diesel2 * Frother2 * Feed GPM * Diesel * Feed GPM * Frother * Diesel * Frother Diagnostics Case Statistics Standard Actual Predicted Student Order Value Value Residual Leverage Residual 1 81.685 -2.87 * B2 * C2 *A*B *A*C *B*C Final Equation in Terms of Actual Factors: . Cook's Outlier Run Distance t Order 0.Dennis I.074 1 0.299 4 0.25mmx0 Rec +37.117 0.558 4 81.88 1.041 3 1.362 2 79.93 68.12 76.919 1.480 6 0.813 2.692 15 77.24 -0.45 -0.010 0.792 0.36 81.54 -9.065 0.526 14 0.55 2.118 -1.25 1.55 1.93 -0.57 83.053 7 86.05 +6.152 13 1.02 0.046 12 0.53 +0.948 -0.595 -1.84 -5.46 2.72 78.636E-03 -1.500 -1.466E-04 +0.506 -0.787 1.472 15 3.515 9 0.358 2 0.526 9 67.84 0.835 1.005 -0.61 82.471 12 83.38 82.15 0.33 0.044 +13.20 82.98 -0.881 5 7.951 10 0.055 -0.200 14 77.570 11 84.17 82.70 83.010 -0.35 -2.19 0.752 1.21 80.29 -3.69 0.59 +2.953 -1.85 79.35 0.987 -1.79 0.492 1.58 -1.036 0.030 -0.030 13 84. Phillips -1.822 2.086 +4.501E-05 -1.612 5 82.283 0.301 -0.92 0.88 0.77 -0.936 -0.57 0.

3 80 70 60 50 1400 60.25mmx0 Rec Actual Constants: Frother = 5.0 700 Feed GPM 40.25 mm x 0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel 100 90 .0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel 100 90 .0 700 Feed GPM 40.0 1050 Diesel 50.0 1225 55.0 875 45.6 80 70 60 50 1400 60.Revisited (451-465) 0.0 1050 Diesel 50.0 700 Feed GPM 40.0 1050 Diesel 50.0 875 45.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel 100 90 .25mmx0 Rec Actual Constants: Frother = 7.25mmx0 Rec Actual Constants: Frother = 9.5 80 70 60 50 1400 60.0 875 45.0 1225 55. Phillips 3D Response Plots Second Series .0 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 154 .Dennis I.0 1225 55.

98 181.00 1400.86 6 Cubic 725.00 9.02 "Model Summary Statistics": Focus on the model minimizing the "PRESS".0000: PRESS statistic not defined Predicted R-Squared -0.5mm) ml/min Type Numeric Numeric Numeric -1 Level 40.74 4 Residual 0.5432 -11.00 700.31 1 Linear 1123.11 0. Model Summary Statistics Root Adjusted Source MSE R-Squared R-Squared Linear 13.000 0 Total 70946.4825 4729.47 0.90 0.25Rec .7624 0.37 200.94 1.80 "Sequential Model Sum of Squares": Select the highest order polynomial where the additional terms are significant.00 5.30 *** WARNING: The Cubic Model is Aliased! *** Sequential Model Sum of Squares Sum of Source Squares DF Mean 67648.5mmx.44 F Value Prob > F 1.Second Series .3677 0.93 1 Blocks 243.96 15 Mean Square 67648.60 +1 Level 60.31 374.3740 PRESS 4714.93 243. Phillips Response: Factor A B C . APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 155 .1875 0.1780 Quadratic 13.2279 Cubic Case(s) with leverage of 1.Revisited (451-465) Name Feed GPM Diesel Frother Units gpm g/T (-.02 37799.Dennis I. or equivalently maximizing the "PRED R-SQR".12 3 Quadratic 1205.

31 258.69 4.40 -2.04 1.25Rec +83.98 Block 2 -4.16 12.06 37799.Dennis I.V.31 Block 1 4.63 A2 0.70 0.89 -0.020 -0.2279 -11.00 1400.60 +1 Level 60.88 22.7642 0.3973 0.47 67.03 AC 7.5mmx.5204 0.74 3298.54 7.01 7.7624 0.63 +0.00 5.05 Final Equation in Terms of Coded Factors: .7112 0.59 12.24 2 B -11.30 Mean Square 243.67 AB -9.32 1.03 Root MSE Dep Mean C.78 0.98 A-Feed GPM 5.4772 0.88 12.2099 2.00 9.31 +5.78 DF 1 1 1 1 1 1 1 1 1 1 1 Standard Error 15. PRESS 13.24 = *A *B *C * A2 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 156 .11 0.02 DF 1 9 4 14 Coefficient Factor Estimate Intercept 83.24 0.24 t for H0 Coeff=0 Prob > |t| VIF 7.16 20.42 6.94 B-Diesel 4.98 725.1023 0.5mm) ml/min Type Numeric Numeric Numeric -1 Level 40.25Rec Factor A B C Name Feed GPM Diesel Frother Second Series .51 16.4255 0.09 2 C -15.31 2328.94 +4.43 0.25 3.78 181.14 BC 16.Revisited (451-465) Units gpm g/T (-.3895 R-Squared Adj R-Squared Pred R-Squared Adeq Precision 0. Phillips Response: .44 F Value Prob > F 1.9850 0.95 -0.83 22.77 C-Frother -16.24 11.3740 4.77 -16.86 18.5mmx.49 0.20 3.644 Desire > 4 ANOVA for Response Surface Quadratic Model Source Block Model Residual Cor Total Sum of Squares 243.55 12.00 700.

500 -1.337 3.67 -9.344 0.470 1.47 62.058 9 11 75.12 0. APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 157 .18 0.049E-05 -4.548 0.90 0.669 8 15 46.339 1 6 87.54 0.611 1.75 -0.07 65.59 79.028 1.195 0.71 +2.500 2 5 61.809 1.23 -15.787 1.76 0.167 10 3 80.09 -15.49 0.39 6.006 0.146 5 12 60.38 4.558 6 10 80.31 11.70 0.03 +7.81 6.39 +0.850 3.877 -1.252 -2.067 0.936 0.85 0.506 -0.953 -1.18 +0.497 0.948 -0.112 -1.14 0.001 -0.502 -0.42 72.58 56.25Rec -30.026 = * Feed GPM * Diesel * Frother * Feed GPM2 * Diesel2 * Frother2 * Feed GPM * Diesel * Feed GPM * Frother * Diesel * Frother Diagnostics Case Statistics Standard Actual Predicted Student Cook's Outlier Run Order Value Value Residual Leverage Residual Distance t Order 1 65.755 0.59 0.095 1.Dennis I.579E-03 +0.70 -8.21 80.283 0.42 +0.301 -0.649 -0.40 0.14 +12.58 -2.38 81.17 -3.31 0.12 -1.02 8.721 0.305 -1.792 0. Phillips -11.664 0.104 0.429E-03 -9.68 1.68 0.14 +16.020 -0.998 14 8 67.95 74.27 35.74 -1.356 15 9 31.65 0.06 -4.77 -6.585 13 4 73.494 12 7 80.444 7 Note: Predicted values include block corrections.06 88.309 1.99 60.998 0.598 11 2 77.096 8.402 0.384 0.5mmx.919 1.97 63.752 1.987 -1.36 51.039 4 14 48.350 0.88 69.20 3.94 0.135 3 13 70.605 -1.023 -0.14 66.78 * B2 * C2 *A*B *A*C *B*C Final Equation in Terms of Actual Factors: .

5mmx. Phillips 3D Response Plots Second Series .5mmx.0 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 158 .0 875 45.25 mm DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel .6 100 80 60 40 20 0 -20 1400 60.0 1050 Diesel 50.0 1225 55.5 x 0.0 1225 55.3 100 80 60 40 20 0 -20 1400 60.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel .5mmx.25Rec Actual Constants: Frother = 5.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel .0 1225 55.5 100 80 60 40 20 0 -20 1400 60.Dennis I.0 700 Feed GPM 40.0 875 45.25Rec Actual Constants: Frother = 9.0 700 Feed GPM 40.0 1050 Diesel 50.0 700 Feed GPM 40.0 1050 Diesel 50.25Rec Actual Constants: Frother = 7.0 875 45.Revisited (451-465) 0.

00 1400.5mmx0 Rec Units gpm g/T (-.50 6 Cubic 74. Phillips Response: Factor A B C Name Feed GPM Diesel Frother . or equivalently maximizing the "PRED R-SQR".2318 0.Revisited (451-465) Type Numeric Numeric Numeric -1 Level 40.8306 0.2707 6296.48 49. APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 159 .00 9.72 F Value Prob > F 1.00 5.96 "Model Summary Statistics": Focus on the model minimizing the "PRESS".71 3 Quadratic 218.7723 -7.8506 PRESS 783.57 36.60 +1 Level 60.000 0 Total 94446.48 1 Linear 148.33 0.57 3912.90 4 Residual 0.00 700.3364 0.42 0.42 18.1373 Quadratic 4.30 *** WARNING: The Cubic Model is Aliased! *** Sequential Model Sum of Squares Sum of Source Squares DF Mean 93991.69 1.42 1 Blocks 12.42 12.5mm) ml/min Second Series .94 0.01 15 Mean Square 93991.0000: PRESS statistic not defined Predicted R-Squared -0.4494 Cubic Case(s) with leverage of 1.Dennis I.40 "Sequential Model Sum of Squares": Select the highest order polynomial where the additional terms are significant. Model Summary Statistics Root Adjusted Source MSE R-Squared R-Squared Linear 5.

33 79.4494 -7.2895 0.18 -0.2797 2.00 700.69 4.2617 0.8306 0.6532 0.00 1400.31 -1.19 -0.32 4.54 +0.5mmx0 Rec +85.66 -1.03 AC 2. PRESS 4.02 5.57 A-Feed GPM -1.88 22.8506 5.25 0.05 Final Equation in Terms of Coded Factors: .2356 Desire > 4 Standard Error 4.72 F Value R-Squared Adj R-Squared Pred R-Squared Adeq Precision 0.90 t for H0 Coeff=0 Prob > |t| VIF 2.16 5.17 -3.91 4.18 Prob > F 0.18 B-Diesel -0.Dennis I.04 1.53 3.86 2.5436 0.75 = *A *B *C * A2 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 160 .44 3.044 -1.40 0.51 DF 1 1 1 1 1 1 1 1 1 1 1 2.86 18.8581 0.22 0.30 ANOVA for Response Surface Quadratic Model Source Block Model Residual Cor Total Sum of Squares 12.79 -1.59 Root MSE Dep Mean C.627 DF 1 9 4 14 Coefficient Factor Estimate Intercept 85.5mm) ml/min Type Numeric Numeric Numeric -1 Level 40.20 3.2345 0.47 3912.54 A2 0.75 2 B -2.7472 0.00 9.90 454.12 7.57 Block 2 -4.79 Block 1 4.96 Mean Square 12.15 3.35 1.60 +1 Level 60.5mmx0 Rec Factor A B C Name Feed GPM Diesel Frother Second Series .54 7.17 C-Frother -3.47 BC 4.48 40.48 -0. Phillips Response: .21 74.Revisited (451-465) Units gpm g/T (-.80 18.V.42 6.61 -0.9673 0.25 3.66 2 C -6.00 5.94 AB -5.48 367.

87 81.596 -2.602 14 0.506 Note: Predicted values include block corrections.439 0. Student Cook's Outlier Run Residual Distance t Order -1.948 7 85.575 6 -0.55 81.546 0.416 9 1.306 4 -0.55 78.500 2 78.498 7 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 161 .75 80.081 3 1.032 -1.05 3.32 80.412 15 -1.06 -0.94 -5.91 80.23 3.30 0.016 0.09 0.467 0.93 63.51 * B2 * C2 *A*B *A*C *B*C Final Equation in Terms of Actual Factors: .39 1.130 12 0.201 10 1.25 2.987 13 82.450E-03 -2.621 8 -0.065 +14.970E-03 = * Feed GPM * Diesel * Frother * Feed GPM2 * Diesel2 * Frother2 * Feed GPM * Diesel * Feed GPM * Frother * Diesel * Frother Diagnostics Case Statistics Standard Actual Predicted Order Value Value Residual Leverage 1 79.20 0.72 -0.44 0.13 +6.57 83.649 11 82.018 -0.5mmx0 Rec +17.70 0.03 -1.953 10 79.40 0.149 0.569 1.26 -2.95 -1.03 +2.656 0.936 6 83. Phillips -2.15 0.283 9 61.008 0.714 5 -1.204 1.Dennis I.112 13 -1.037 -0.88 82.85 1.47 +4.795 11 1.029 -0.04 1.346 3.19 -1.83 -2.79 0.752 3 83.759 0.037 -0.059 8.854 3.787 4 79.014 0.70 -1.463 0.66 -6.75 -4.763 1.553 0.84 75.38 0.000 0.188 -1.31 +7.34 84.801 2.38 -0.301 15 74.175E-05 -2.919 14 74.149 0.26 0.66 0.11 80.02 75.675 0.792 8 80.436E-03 +0.09 0.68 0.35 +0.345 -1.501 1.407 0.35 83.190 2 0.51 0.39 76.012 1 -0.877 5 80.48 0.809 12 80.

0 1225 55.0 875 45.0 700 Feed GPM 40.0 1050 Diesel 50.0 700 Feed GPM 40.5mmx0 Rec Actual Constants: Frother = 5.5mmx0 Rec Actual Constants: Frother = 7.0 1050 Diesel 50.0 875 45.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel 100 90 .0 1225 55.0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel 100 90 .0 1225 55.Dennis I.0 1050 Diesel 50.Revisited (451-465) 0.5 mm x 0 DESIGN EXPERT Plot Actual Factors: X = Feed GPM Y = Diesel 100 90 . Phillips 3D Response Plots Second Series .3 80 70 60 50 1400 60.6 80 70 60 50 1400 60.5mmx0 Rec Actual Constants: Frother = 9.0 APPENDIX B STATISTICAL ANALYSIS OF PLANT TEST DATA Page 162 .0 875 45.5 80 70 60 50 1400 60.0 700 Feed GPM 40.

Dennis I. Phillips APPENDIX C QUANTITATIVE EVALUATION DATA APPENDIX C QUANTITATIVE EVALUATION DATA Page 163 .

APPENDIX C QUANTITATIVE EVALUATION DATA Page 164 .Dennis I. Phillips Figure C-1. +1 mm. taken from Coalpro computer simulation program. Heavy Media Cyclone feed washability. Stockton seam.

partition parameters. Typical simulation results showing Heavy Media Cyclone partition curve. Taken from Coalpro simulation. for Stockton seam.Dennis I. APPENDIX C QUANTITATIVE EVALUATION DATA Page 165 . and product qualities. Phillips Figure C-2.

APPENDIX C QUANTITATIVE EVALUATION DATA Page 166 . Stockton Seam.150 x 0.25 mm.Dennis I. Phillips Figure C-3. Spiral feed washability. 0.

Dennis I. Stockton seam.25 mm size. Phillips Figure C-4.150 x 0. 0. Simulation results showing partition curve and product qualities for the Spirals. APPENDIX C QUANTITATIVE EVALUATION DATA Page 167 .

Phillips Figure C-5. Steep slope indicates a high amount of near-gravity material. +1 mm. 5-Block seam. APPENDIX C QUANTITATIVE EVALUATION DATA Page 168 . Heavy Media Cyclone feed washability.Dennis I.

Preparation Engineer. After spending his summers working in underground coal mines. in Mullens. he worked as Senior Mining Engineer in their corporate Coal Business Development group for three years. from University of West Virginia College of Graduate Studies in 1990. as Preparation Superintendent for Knox Creek Carbon. degree in Civil Engineering from West Virginia University in 1977. He worked as a Mining Engineer for Gates Engineering. operation. and after part of the Massey Company was partitioned to Shell Mining Company. Steel. Phillips VITA .B. He later received an M. He spent a few months providing consulting services in plant design. He is a Registered Professional Engineer and has several publications and presentations at national meetings to his credit. VITA Page 169 .D. he received a B.S. 1954. Phillips spent 15 years in coal preparation plant engineering and operations management. and optimization before joining the Center for Coal and Minerals Processing at Virginia Tech in 1992 as a Sr.T. Mr.A.S. Massey Coal Company as Sr. where he was elected Senior Class President. as a Preparation Engineer for U. He grew up in the community of Corinne and graduated in 1972 from Mullens High School. West Virginia. Dennis Ivan Phillips was born on May 17. and six years for A. Research Associate while working part-time on his Ph. and Assistant Preparation Director.Dennis I.

Sign up to vote on this title
UsefulNot useful