This article was downloaded by: [Universal Subs Agency] On: 23 April 2011 Access details: Access Details:

[subscription number 731773863] Publisher Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK

International Journal of Coal Preparation and Utilization

Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t713455898

Magnetite Heavy Media: Standards and Testing Procedures

M. W. Mikhaila; D. G. Osborneb a Coal Preparation Section. Fuel Processing Laboratory, Coal Research Laboratories, Canada Centre for Mineral and Energy Technology (CANMET), Devon, Alberta b Coal Preparation and Handling. PT Kaltim Coal of Indonesia,

To cite this Article Mikhail, M. W. and Osborne, D. G.(1990) 'Magnetite Heavy Media: Standards and Testing Procedures',

International Journal of Coal Preparation and Utilization, 8: 3, 111 121 To link to this Article: DOI: 10.1080/07349349008905179 URL: http://dx.doi.org/10.1080/07349349008905179

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Coal Prepururion. 1990 Vol. 8. pp. 'I I 1-12! Photocopying permitted by license only Minister of Supply and Servim Canada. 1990. Printed in the United Kingdom

Magnetite Heavy Media: Standards and Testing Procedures

M. W. MIKHAIL* and 0. G. OSBORNEt
(Received Murch 20. 1990, infinul form April 6. 1990)
Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011

Magnetite heavy media prare armrate. efficient and e y y to control separators. capable of producing high quality clean coal. ,Magnetite recovery circuit operation and magnetite co~umption arc the main reasons for high capital and'oprating costs. Ovcr the last two decades. several countria have k e n actively involved in establishing magnetite specifications for coal heavy media circuits and in developing working standards for t a t magnetite. This paper reviews media recovery circuits. operating parameten for magnetic separators. magnetite consumption and specifications for heavy medium magnetite. It diuua methcds and instntmmts used in the determination of those physical and chemical properties of magnetite. and rhcological propenin of magnetite suspensions which arc relevant to the application of heavy media separators. The p a w also outlines the development of the international standards for magnetite and magnetire suspensions fating.
Key w r d r Magnetite. heavy medium. huvy medium control, heavy medium recovery. international standards.

INTRODUCTION The importance of heavy medium circuits in comparison with other cleaning processes has increased sicnificantlv in the last twodecades The reason is the need for sham. .. efficient and easy-to-control systems to meet the demand for high quality clean coal. The heavy medium bath handles coarse coal, usually in the + lOmm size range, and the heavy medium cyclone clean coal in the SM.5mrn range. Most of the recently designed preparation plants include the heavy medium cyclone circuit to clean the + 0.5 mm coal because of its accuracy in the separation of fines. Also, recent attempts to clean the -0.5 mm coal by heavy medium cyclone have shown promising results.' The use of heavy medium for fine coal processing increases the consumption of magnetite and demands certain specifications of fine magnetite to achieve optimum

*Coal Preparation Section. Fuel Proeating Labontory. Coal Rcsearch Laboratories. Canada Centre for Mineral and Energy Technology (CANMET). Dcvon. Alberta '*Coal Preparation and Handling. PT Kaltim C d of Indonesia.

M. W. MlKHAlL and D.G.OSBORNE

Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011

separation. More efficient magnetic separators have been introduced along with improved rinsing capacity and elaborate media recovery circuits to minimize magnetite losses. Magnetite resources are abundant as primary deposits, particularly in Canada. However, the market for magnetite for heavy medium use is limited and cannot support the operation of a magnetite mine. As a result, magnetite presently used for heavy medium is a by-product of existing metal mines or is reclaimed from tailings. Several countries have been'actively involved, in the last two decades, in establishing magnetite specifications for heavy medium applications. These countries include Australia. Britian, Canada, Germany, Netherland. South Africa and U.S.A. Many of the tests used were common but some were specific to certain countries. The recently available I S 0 standard on the subject is based on experience gained over many years by several countries. The objective of this paper is to review the standards and standardized testing procedures utilized to characterize the magnetite and the heavy media prepared with the use of such a megnetite. Since the characteristics of the magnetite recirculated in a coal preparation plant depends on the media recovery circuits? such circuits are also reviewed in this paper. This paper is based on a detailed study carried out by Kilborn Engineering (Vancouver) on behalf of CANMET entitled "Characteristics of Canadian Magnetite Resources for Optimum Utilization in Coal Heavy Media Processes"."
M E D I U M CIRCULATION, CONTROL AND RECOVERY

Efficiency of separation in heavy media circuits depends upon magnetite media rheology and stability. This subject is discussed in another paper.' Following the separation in the cyclone or vessel, the medium adhering to the product is rinsed away with water. The diluted medium is then directed to the recovery system which incbdes the following: 1. Rinsing water (recirculated within the pliant). 2. Magnetite rinsed from the products. 3. An overRow or bleed from the heavy media slurry feed. 4. Fresh magnetite to compensate for loss. The discharge of the recovery system circuits of concentrated medium (magnetite plus water) and non-magnetic tailings is a very dilute slurry, which is bled off as effluent and utilized for rinsing water or sent to tailings thickener. There are three alternative media recovery circuits (recommended by the Dutch State Mines, now Stamicarbon) used universally in heavy medium plants: I. Single-stage magnetic separation and magnetite thickening (Figure 1) which is ~sed when the feed to classifying cyclone contains a magnetite concentration of less than I00 g/l (1.08 g/cm3). 2. Two-stage magnetic separation with cyclone thickening prior to the secondary magnetic separation (Figure 2) which is applicable when the dilute medium concentration i; in the range of 250 g/l (I. 19 glcm').

MAGNETITE HEAVY MEDIA

FIGURE I
Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011

Media rccovcq system employing magnctitc thickener

3. Three-stage magnetic separation with cyclone thickening before the second and third magnetic separators (Figure 3) which is used when a medium with a relative density too low to effect the required separation is fed to a magnetic separator.
Recommended (by Stamicarbon) operating parameters for magnetic separators are: Concurrent drum separator with permanent ceramic magnets. 250 g/l maximum feed concentration to the separator. Maximum solids content (magnetic and nonmagnetic) less than 16% by volume. Minimum flux density of 750 Gauss (75 milli tesla) in air at a distance of 50mm from the drum. The relative density of recovered magnetite of 2.1 g/cmJ. Losses of magnetite not to exceed 0.3 g/l. The density of the medium has to be kept constant within certain limits by automatic means in order to obtain the required optimum separation. The density of the medium entering the separating vessel, or cyclone, is usually measured by a

-Q II

-d

FIGURE 2 Media recovery system employing magnetic xpardton.

M . W.MIKHAIL and D.G. OSBORNE

Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011

L I .

L I .

. . I .

FIGURE 3 Media recovery system for low field concentration

nuclear density gauge. There are a number of design alternatives for locating the sensor which measures the correct (set) medium density in the recovery circuit. In any case, it is essential to have a control circuit that responds quickly to changes in medium density.
MAGNETITE CONSUMPTION

The consumption (loss) of magnetite is influenced by several factors, of which the most important are:

a) The fineness of the magnetite. Finely ground magnetite exhibit better stability but
may also show higher viscosity; generally magnetite fineness improves heavy medium separation efficiency but since finer panicles are more difficult to recover, it may lead to higher losses. The fineness of the treated coal. Increased surface area of fine coal panicles increases the amount of magnetite attached to coal particles which is difficult to rinse off. The efficiency of the magnetic separator. Factors such as Gauss rating, position of permanent magnets, feed volume, percentage of magnetics in the feed influence the efficiency. The design of the magnetite recovery circuit single, two or three-stage separation and use of dilute medium cyclones versus thickeners. General poor conditions of operation and maintenance problems.

b)

c)

d) e)

A generally accepted figure of magnetite losses for European. eastern U.S. and South African coal preparation operations would be 0.15 kgltonne of fine coal ( 1 0 4 . 6 mm) and 0.05 kgltonne for coarse coal (+ 10 mm). However, for Canadian coals reported losses are much higher (0.75 to 3.0 kgltonne). This is largely attributable to the presence of a high proportion of finer panicles (due to coal friability) where the average particle size of the 10-0.6mm fraction is 1.8 mm. It is wonh noting that magnetite consumption should be quoted on raw coal fed to the heavy medium

MAGNETITE HEAVY

MEDIA

l IS

circuit. The amount of medium adhering to the products on a sieve bend plus drain screen can be calculated from the following empirical equation:'

where Q = medium in [/tonne product d, =average particle size of product in mm (weighted average size from screen analysis) p , = average relative density of product Mass of solid ( M )in gllitre of pulp (medium) is calculated as follows:

Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011

where p = density of magnetite p, = density of pulp. SPECIFICATIONS FOR H E A W MEDIUM MAGNETITE The following brief and general specifications, based on the British coal mining industry standards and accepted world-wide are:J.6.'" a) Particle-size distribution b) Relative density Magnetic content Maximum 5% by mass larger than 45 pm and 30% by mass smaller than 10 pm. 4.9 to 5 2 g/cml. Not less than 95% by mass.

C)

It is also usually required that magnetite moisture content be less than 10%. METHODS USED IN THE DETERMINATION O F MAGNETITE PHYSICAL PROPERTIES Particle sizes 1. Dry Screening: An automatic sieve shaker is operated with 203 of magnetite for 10 minutes using three screens with apertures of 53, 45 and 3Spm. 1. Wet Screening: A wet vibrating sieve apparatus (NCB design) can be used to separate the same amount of magnetite as dry screening ( 2 0 9 . A 15pm screen can be included to cover a wider range than dry screening. The sample is wetted before screening and gently sprayed during the process. The minus Z5pm is screened over lOum using a Fritsch Analysette sieving apparatus with alcohol as the suspended medium because of its low surface tension. Cyclosizer: A Warman Cyclosizer with five cyclones in series is used with water 3. as the medium with magnetite sample. The precision of separation improves with the increase in elutriation time, but 10 minutes is often adequate. 4. Sedimentometer Size Analyzers:

116

M. W.MIKHAIL and D.G. OSBORNE

Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011

a) Sedigraph: It is based on gravitational x-ray panicle size analysis. The magnetite suspension is pumped upward in a sample cell for detection by x-ray. The x-ray detector gives signals as a function of the Stokes diameter of magnetite suspension in the sedimentation (sample) cell. The range of the application is 1 to 50pm. b) Photosedimentometer Size Analyzer: It is based on concentration measurements of a suspension settling under gravity, panicle size is calculated using Stokes equation. Magnetite has to be demagnetized by heating to above 600C prior to testing. The size range of application is from 2 to 100pm. 5. Microscopic Size Analytical Techniques a) Optical Microscopy: It measures the diameters of particles using a calibrated fibre micrometer eyepiece. The range of panicle~size measurement is 1 to 150pm but the accuracy is dependent on operator's skill. More recent automatic microscopes with image analyzers can be used to accurately determine the size. b) Laser Size Analyzer: The size distribution of particles is determined in the suspension flowing through a sample cell which is illuminated by a continuous laser beam. The suspended particles scatter the light which is collected by a series of lenses and a rotary optical filter. The measured panicle size range is usually flexible and dependent on using certain detector head. M a g n e t i c c o n t e n t analyzer:" 1. Magnetic Chute: It includes a sloping and hinged non-magnetic tray containing a permanent magnet which holds the magnetite in a flowing slurry. The hinged tray is lifted away from the magnet and the magnetic particles are washed off for .analysis. 2. Davis Tube:" It is a laboratory unit used specifically for determining the ferromagnetic and non-magnetic fractions with field intensity ranging from 400-800 Gauss that would give consistent results. It includes a sloping non-magnetic frame which contains a permanent magnet unit above which a hinged tray is fitted. The magnetic panicles are held by the magnet when slurry is poured into the tray. The magnetics are then removed from the tray.for analysis. 3. I S 0 Method:" The apparatus consists of a vertical glass tube (8 mm in diameter) with two permanent magnets at the centre of the tube and both the tube and magnets (IS0 8833). The apparatus is filled with the magnetic suspension. The contents are then drained into a lOOml beaker while magnets are moved up and down to capture magnetite panicles. This magnetite is later on washed into another beaker after the tube is drained. Density m e a s u r e m e n t I . . Bechman Pycnometer: The apparatus functions on the basis that a sample displaces an equal volume of air. Two cylinders of the same volume are used, one as a reference and the other for measurement, each having an identiel movable piston. The measured volume of a sample of known mass is then used to calculate density. 2. Density Bottle: A 50ml capacity density bottle is used to measure the volume of distilled water displaced by a 20 g magnetite sample (IS0 8822). A summary of the above procedures and equipment is shown in Table 1."

'MAGNETITE HEAVY MEDIA TABLE l Methods used in the determination of the physical propenia of magnetite Method or Equipment Pmpefly Panicle size Dry screening Wet screening Cyclosizcr Coulter counter Sedignph or other sedimentation analyen Bahco or Haultain: Pneumatic classification Optical sizing Davis tube IS0 Apparatw Magnetic chute Hand Magnet Demity Only for panicles larger than 53 pm Many variables introduced. for panicles above 25 pm Time<onsuming: only limited poinu obtaind Sensitive to auto coagulating pmpnies High density of panicles c a w rapid vttling Time-consuming. potmlial high loss of very fine material Sophisticated equipment needed Many variables Simple method with good reproducibility Variable magnetic field Accuracy depends on operator. Only approximate Slightly operatordependent Ofien difficult with line powden Limitations

Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011

Magnetic content

Beckman pycnometer Density bottle

METHODS USED I N THE DETERMINATION OF MAGNETITE RHEOLOGICAL PROPERTIES

Viscosity:"

Rotational Viscometen: Conventional rotational viscometen do not take into account the unstable nature of magnetite media nor the centrifuging of panicles at high shear. A modified Stormer viscometer applies a constant shear stress to the medium in the annular gap and measures the speed of rotation (modified by placing an impeller in the bottom of water-jacket of container to maintain a uniform suspension). This viscometer consists of a multi-bladed rotor, rotating in the opposite direction to and also beneath the inner cylinder to keep the medium in a homogeneous state by turbulence. Advances have been made since Starmer viscometer had been introduced. Viscosity measurements are discussed in another paper in this issue.'' . .
.

testing:" F5 [ndex Stability Method: The method was developed by the National Institute for Coal Research (NCIR). South Africa. The apparatus used is made of a copper tube with an overall length of 1.225m. It includes three tapping points. After 5 minutes of settling, the medium between tapping points I and 2 (A) is run off and weighed. After discharging the medium between tapping points 2 and 3, the remaining medium in the tube is weighed (B). The ratio of A to B gives the F5 index. Jar Settling Test: The demagnetized sample is mixed with distilled water (with added wetting agent) in a 500ml calibrated cylinder. The relationship between height
stability

M. W. MIKHAIL and D. G:OSBORNE TABLE 2 Method used in the determination of the rheological propenia of magnnite Propeny Viscosity Method or Equipment Stormer virometcr DeVaney-Shelton consistometer Capillary virometer Rotational viswmetn Stability
Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011

Limitations Reliability only achieved by custom modification Undefined shear raw does not allow precise rhmlogical measuremenu Non-homogmeous movement of panicla affens flour in capillary Commercial units am not suitable to work with wdimenting systems Operator influcncr very significant Not developed to commercial N o standardized method available
u=ee

FS Index Tube
Settling Tests

'

DRL 'U' rube-apparatus

and time produces the settling rate. The slope ob&ed during the free-settling zone represents the settling rate in mm/min with typical rates of 10-20mmlmin.
Magnetic properties testing:"

Magnetite's magnetic property is not a rheological property, however, it influences the rheology and stability of the magnetite slurry. Satmagan Balance: It measures the deflection in the mass of a magnetic material upon .being placed in a magnetic field. The susceptibility of a certain amount of magnetite in a glass tube is measured by a deflector indicator when the sample is exposed to magnets located at certain distance from the sample. Vibrating Sample Mangetometer: It measures magnetic susceptibility! saturation moment, remnance and coercive force.
Demagnetization test

Simon-Carves Demagnetization Method:' A slurry of 1.6 relative density is placed in a 100cc graduated measuring cylinder and thoroughly apitated. After the settling rate is measured. the cylinder is magnetized ip a field of 65 kA/m. and the settling rate remeasured. After that, the cylinder is demagnetized and the process repeated. A summary of the rheolopic procedures and equipment" is shown in Tables 2 and 3.
MAGNETITE CHEMICAL ANALYSIS

Chemical analysis is used to distinguish magnetite from other sources of iron and includes: a. Determination of the total iron content using I S 0 2597 volumetric method. b. Determination of the ferrous iron content by dissolving the sample in hydrochloric acid under nitrogen and titraring ferrous iron from the resultant solution using potassium dichromate.

MAGNETITE HEAVY MEDIA TABLE 3 Methods wd to characterite mapetic properties Method or Equip ment Magnetic SurepdbditylCoercive Fora Satmagan balance Vibrating magnetometer Thennogravimetric Eaton apparatus Demagnetization
Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011

Limitations Single-point measurement of susceptibility Not developed to commercial Mesures hysteresis curve only and requires calibration Reference is HgCo(SCN) (SO414 which must be carefully prepared Operator influence. but reproducibility is reasonable
W g e

Simon-Carve method

More detailed chemical analysis would only be required for new sources of media to determine the following: SiO,, AI,O,, TiO,. P, S, Ba, Cd, Zn. Ca. Mg, Ni, CO, CU, Au and Ag usually by x-ray fluorescence analysis. DEVELOPMENT O F INTERNATIONAL STANDARDS FOR MAGNETITE TESTING Industrial experience and practices by several coal producing countries have been used to develop standard procedures to test magnetite for coal preparation purposes. The International Standard ISO/DIS 8833 (submitted in November 1986) specifies the following properties for testing:" a. b. c. d. e. f. g. Moisture content. Particle size distribution. Magnetic content. Fundamental magnetic properties. Relative density. Total iron content. Ferrous iron content.

All of the above properties may need to be determined for a new deposit but for a known supply of magnetite a. b. c. and e are usually sufficient." The following is a summary of tests performed on samples from several maznetite deposits in Canada (a, b. c, and e) most of which are in compliance with suggested IS0 standards: a. Moisture Content: Total moisture content can be determined either directly (100g sample) or in two stages (I kg sample) in compliance with the 6 0 . 8 8 3 3 method, where the free moisture and air-dried moisture contents are determined separately. The latter method is, however, mainly applicable to large quantities ( I kg sample). Samples are dried in a laboratory circulated air-oven at temperatures of 105-1 IOC.

120

M. W.MlKHAlL and D. G. OSBORNE

b. Particle Size Distribution: The size analysis should comply with I S 0 2591, where the plus 38 pm fractions are wet screened and the minus 38 pm fraction is screened using a sub-sieve centrifugal classifier or a Warman Cyclosizer. c. Magnetic Content: The magnetic chute with a fixed magnetic field strength (permanent ceramic magnets) is recommended (i.e. similar to draft IS0 8833 test method). The multiple pass procedure (usually two passes) ensures total recovery of magnetic material. d. Relative Density: The set pycnometer method on which the volume of liquid is displaced by a known mass of sample (draft I S 0 8833). PROPOSED R&D WORK
Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011

Determination of the rheological properties of a magnetite suspension is the principal area that requires further research as indicated in most major coal producing muntries. Two main subjects are investigated: a. Medium rheology in relation to panicle size distribution, fundamental magnetite properties, the presence of non-magnetic materials such as clay minerals and viscosity reducing chemical additives. b. Medium stability, the methods for its determination and effect of chemical aids on the stability. While (a) is being specifically directed towards sophisticated scientific investigations requiring advanced viscometric methods. (b) is mainly concerned with obtaining a reliable method for quantifying the effect.'There are several methods to determine medium stabiliry (F5Index, Simon Carves settling method and the Czechoslovakian J-Index); these tests, however, are very empirical and therefore prone to operator bias As a consequence, all methods currently employed should be comprehensively evaluated before a suitable method is adapted for standard use. References
I.

2.
3.
4.

5. 6. 7. 8. 9. 10.

D. G. Osborne. "Fine Coal Cleaning by Gravity Methods: A Review of Current Pramice", Cwl Prtpararion 2.207-24 1 (1986). "Characteristics of Canadian Magnetite Resource for Optimum Utilization in Coal Heavy Mcdium Proeessts, Pan 1. Survey of Mametic Medium Coal Cleaning", CANMFT wntran repon . prepared . . by Kilborn ~nginecrine(1986)"Characteristics of Canadian Magnaitc Resources for Optimum Utilization in Coal Heavy Medium Testing and Analysis Stages". CANMET wntran repon prepared by Kilbom Engineering (1986). T. J. Napier-Munn, "The ERffl of Dense Media Viscosity on Separation EBcirncy", Coal P r r p a r q rion. this issue. "Densc Medium Coal Preparation". Sunonam Ltd. Publications (1968). T. E. Abrey, "A Note on the Use of Magnetite in Coal Preparation in Great Britain"; Mining and Marrrials Engineering. 25-29 (1972). "Specification for Magnetite", The NCB Code of Prania for Coal Preparat~on (Draft)(1985). C. C. Graham. "The Application of Magnetite for Coal Rrparation in Australia". F ~ nIntmnadonal t SME-AIME Fall Meeting. Honolulu. Hawaii (1982). i z e Measurement". 2nd Edition. Chapman and Hall. London (1978). T. Allen. "Panicle S Standard Produre". Pcnekc. R. Bosold. "Determination of Magnetics Content of Magnaite Homer City Coal Laboratory (1983).

MAGNETITE HEAVY MEDIA

II.

121

12. 13. 14.

IS.

16.
17.

N. F. Schulz. "Determination of the Magnetic Separation Characteristic with the Davis Magnetic Tubc". Truns. Soc. .Winins &n#ineering. pp. 2 1 I-? 16 ( 1964). "The Testing of Magnetite for Ux in Coal Preparation". Draft Inrernntional S t a n d a h ISOIDIS 8833 (1986). L. Jonker. "The Development of Standard Rocedum for the Evaluation of Magnetite or Uw in Heavy Medium Separation". Mintek Repon. No. MIU. Republic of South Africa (1984). H. F. Yancy. M. R. Cmr and M. Sukaski. "Viscosity. IU Mcasumenr and Impoflance in Dmw Medium Cleaning of the Fine Sizes of Coal". 3rd International Coal Prcparation Congress. Liege (1958). Paper FI, pp. 584. 8. Klein. S. J. Partridge and J. S. Laskowski. "Rheolonv -- of Unstable Mineral Susansions". C a l Prepararion. this issueK. Kamorniczyk. D. Gleisbcrg and M. Reia. "Test Methods to Determine Viscosity. Stability and Corrosion Behavior of Hnvy Medium Suwnsions Consisting of Ferrosilicon, Ferrochmme. and Magnetite". Knapsack A.-G. Cologne, &netall (1973). C. C. Graham and R. Lamb. "Coal Preparation in Denx Media Rheology: .4 Review of Measurement and Control". ACIRL Research Report (Australia). P.R. 82-3 (1981).

Downloaded By: [Universal Subs Agency] At: 04:18 23 April 2011