Lehigh UniversityLehigh Preserve

Theses and Dissertations

1987

Coal cleaning in an air fluidized bed /Yih-Tun TsengLehigh University

Follow this and additional works at: https://preserve.lehigh.edu/etd

Part of the Mechanical Engineering Commons

This Thesis is brought to you for free and open access by Lehigh Preserve. It has been accepted for inclusion in Theses and Dissertations by anauthorized administrator of Lehigh Preserve. For more information, please contact preserve@lehigh.edu.

Recommended CitationTseng, Yih-Tun, "Coal cleaning in an air fluidized bed /" (1987). Theses and Dissertations. 4788.https://preserve.lehigh.edu/etd/4788

..

....... ~ .. . . . . ... - . . -

COAL CLEANING IN AN AIR FLUIDIZED BED

by

Yih-Tun Tseng

A Thesis

Presented to the Graduate Committee

of Lehigh University

in Candidacy for the Degree of

Master of Science

• in

Mechanical Engineering

Lehigh University

1986

..

. ,,.,

\

'

r

. _,, .. ' .. -··· ........ .. - -·--··--·· ----- ~·---·-·--·· .... , .

•

This -thesis is accepted and approved in partial

fulfilment of the requirements for the degree of Master of

Science.

(date)

Professor Charge

Chairman of Department

I I

•

CONTENT

Abstract • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 1·

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . 2

Equipment and instrumentation • • • • • • • • • • • • • • • • • • • • • 8

Experimental procedure • • • • • • • • • • • • • • • • • • • • • • • • • • • • 20

Results • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 22

Conclusions • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 62

..

..

'

•

ACKNOWLEDGEMENT

The author is grateful to Bulent Kozanoglu. for his

help in carrying out these experiments and analyzing the

data. • • • • • • . It is a pleasure for the author to record his gratitude

to Dr. Edward K. Levy for his valuable guidance and

encouragement during this study.

\

l •

.. •

· I

•

ABSTRACT

The purpose of this study is to develop an efficient "

method of separating coal refuse from coal in an air

flu~dized system of coal and magnetite. Special emphasis is

on processing coal particles with a top size of 20 mesh and

with fines smaller than 100 mesh. -~

By fluidizing crushed raw coal and magnetite near the

minimum fluidization velocity, good segregation can be

obtained. The ferromagnetic magnetite particles can be

separated from coal through use of magnetic separation

techniques after segregation between the coal and its refuse

is achieved.

In this work, experiments involving mixtures of coal

and magnetite were carried out. The effects of particle size

and density, bed depth, fluidizing velocity and multiple

stages of cleaning on sulfur reduction were investigated.

Finally, a rectangular inclined fluidized bed design was

developed for continuous processing of the coal.

l

0

\

..

INTRODUCTION

Background

' I (''

I

Two major benefits can be derived from coal cleaning.

One is to increase the heating value of the coal and the

other is to reduce air pollution from sulfur. In addition,

coal cleaning reduces fuel transportation costs and results

in .lower capital costs for new power plants. The latter

benefit occurs because boilers can be built and maintained

more inexpensively if they are required to burn coal with

less sulfur and ash and a more consistant quality. About 35

percent of the coal in the United States • receives some

cleaning before it is burned [l].

The irnpuri ties in raw coal of concern in this study are

the noncombustible portion -- the ash and the pyritic sulfur.

Sulfur exists in coal in several forms. Some sulfur is part

of the minerals that make up the ash, while other forms of

sulfur are chemically associated with carbon. The sulfur has

three major classifications:

. Organic sulfur -- Physical coal cleaning methods are

•

•

not able to separate I organic

sulfur from coal.

Pyritic Sulfur -- This is

sulfur.

an inorganic form of

'

Sulfate Sulfur -- While hard to remove, this is

usually present in negligible

amounts.

2

\ '

"-, I ' / ' ' /

'

'

Physical coal cleaning can separate free pyrite from

coal, but it· cannot separate pyrite that is finely dispersed

in the coal. \ .

Liberation is the process of breaking the ·raw coal

particles in order to release some of the free ash (which ,,,

includes pyrite) fragments when the coal particles are

fractured. During the coal crushing process, many other

particles, containing solely inherent ash or free ash are

reduced in size. Breaking these types of particles changes

the raw coal's size distribution, but it does not increase

the amount of liberated impurities in the coal [2].

When wet coal cleaning methods are used, the fine coal

particles are the most difficult to clean, because they tend

to stay suspended with pyrite in the cleaning process.

Consequently the coal fines are discarded at many coal

cleaning facilities. Because the fines contain up to 25

percent of the heating value of the raw coal, there is

growing interest in recovering the combustible portion of

these small particles [l].

As an alternative to wet cleaning, dry cleaning ha:s ' .

several advantage:

• Higher thermal efficiency .

• No wastwater treatment problem .

• Reduced handling problems due to coal freezing

and plugging of and bins.

Some types of dry .systems which can be used in coal cleaning

3

J

..,

\ • •

- . : ~~· "l''"r ·-.~ ···-~.. , .. ,,

• I

include pneumatic jigs, pneumatic tables and air·fluidzed

beds. The first two are in commercial use. Air tables are

effective cleaning devices down to 28 mesh (3], but they are

·i,

inadequate when fine fractions of coal are dealt with.

In a fluidized bed where particles of different

densities and sizes exist, there is a tendency at near

minimum fluidzation conditions for the solids to stratify in

the vertical direction according to the density, and to a

'

lesser extent, size. Two fechanisms control the manner in

I

which particles segregate (l] .

. As a bubble rises through the bed, particles are

carried along with the wake of the bubble. This

tends to counteract solid segregation by causing

remixing of settled material. This, however, is also

the mechanism by which the less dense particles find

their way to the top of the bed .

. As bubbles move vertically upward through the bed,

they cause the dense particles to descend. The

dense particles tend to fall through the free space '

of the bubbles and drift downward through the

emulsion phase in the region disturbed by the bubble

as the bubble passes by .

Bubbles play a very strong role in mixing and

stratification, and the degree of segregation which occurs

depends very much on the excess velocity. This is ~llustrated

in Fig.l where Curve A is the case of a strongly segregated

4

...

'· .

. . . . .

...

j

/

•

Top

+> .c: tJl

•ri (l)

=x: re, Q) ~

•

C I I I

I I

I I

I

' I

(A) (B) (C)

I

• ' ' I

..

Uo = Umf. completely stratified Uo > Umf. partially mixed Uo >> Umf. well mixed

, A I

l I

Bottom

.

Concentration of dense component

Figure 1. Effect of fluidizing vel·ocity on ·solids stratification. (Ref. 4 )

5

·,.,, .

..

..

bed very near minimum fluidization, curve Bis the concentration profile for the case of soliids at. an intermediate fluidization velocity and Curve C is the case of a perfectly mixed bed operated at high fluidizing velocities.

Previous Research

Particle segregation has been studied by many researchers in the last twenty years.

One of the first studies was carried out by Rowe and Sutherland. They concluded that no bubbles appear and

therefore no particle mixing occurs at gas velocities up to about 1.2 Umf in binary mixtures [5]. Later, Rowe, et al., studied the mixing effect of a single bubble passage through a two dimensional fluidized bed [6]. Rowe, Nienow and Agbim found that the effects of density ratio and excess velocity are considerable while those of size difference and shape are slight in segregation of binary mixtures (7].

Chen and Keairns [8] performed segregation experiments with dolomite and char and observed sharp segregation near the flow rate of the jetsam particle's minimum fluidization velocity (The component which tends to segregate and settle to the bottom is called jetsam, while the one which tends to float is called flotsam in the technical literature). They found that segregation is much more sensitive to particle density difference than to particle diameter. All of their results supported the conclusions of the earlier study by Rowe, et al .

6

I

L

'

Nienow, Rowe and Cheung [9] studied the segregation and

mixing behavior of binary mixtures of powders differing in

density and developed a mixing index. Their mixing index

agress with the experiments with a standard deviation of+/-

9 percent for mixtures with up to 50 percent jetsam by

volume.

One of the coal cleaning related segregation Btudies

was perf armed by Nienow, Rowe and Chiba [ 1 O]. They carried

out experiments for ternary systems of char, sand and shale

and observed that large particles of low density such as coal '

tend to remain on the surface and can be separated from

denser ones, such as shale in a fluidized bed of intermediate

density such as sand, Fig. 2. Their experiments also showed

that the flotsam particles are drawn deeper into the bed with

increasing gas velocity. But the degree of segregation in

their experiments was far from perfect.

Weintraub, Deurbrouck and Thomas (11] studied

segregation in ternary systems. They found that minerals of

different density, such as coal and coal refuse, can be

separated in a fluidized bed of fine size and high density

material. They successfully separated various fractions of

coal and ash impurities ranging from 9.5mm top size down to

"'._ 3 O mesh bottom size in times ranging from sixty seconds for

the coarse material up to five minutes for the finest

particles in a fluidized bed of small magnetic particles.

Fig. 3 shows the distribution of coal and refuse with

7

"f•·

"

,,

I

location in the bed and with time in one of their

experiments.· They observed that the coal becomes uniformly '

dispersed within the magnetite while the.refuse material '

sinks to the bottom of the bed. They also found that the

degree of segregation is affected by the size distribution of

magnetite.

Objective of Investigation

The objective of this study is to develop an efficient

method of separating coal refuse from coal in an air

fluidized system of coal, coal refuse and magnetite.

Special emphasis is on processing coal particles with a

top size of 20 mesh and with fines smaller than 100 mesh.

Fina1·1y a continuous inclined bed was designed, built

and tested for this application.

EQUIPMENT AND INSTRUMENTATION

Fluidized Beds

Three fluidized beds were used for this investigation.

The first, borrowed from Dr. J. Chen of the Chemical

Engineering Department, was a 10.16cm (4 inch) cylindrical

bed (see Fig. 4). A series of fluidization experiments was

carried out with magnetite, glass and plastic particles.

These were intended to determine the effects of particle

density and size on minimum fluidization velocity ~nd to

develop qualitative guidelines on the fluidization and

stratification behavior of mixtures of particles.

The second bed was a 15.24cm (6 inch) diameter

8

.. .

.. '

..

..

. . .

n • 1·49 t1/cmJ r1a1& ~

Sand JctlGID : 90°/o Sand

3·5°4 Shale

Flotsaa: 6·5°/o Char

d,• O·BBmm

d,• 3·03~

dp • 5 ·18 "'"'

-p . • 2·S0g/cmJ

p • 0·97 vi Chl3

Shale rich

.• rich , H ··'-- , .. .. ·-~- ,. . .';

Bed hc.ght.

•

6

(Ill 4'

(u-u.,l • 13 cm/s

Sand

. --=-~~- ... . . . . . . . . .•' . :

8

6 . (u-u., l • 28 cm/s

4 Sand

0 o 0-2 0·4' 0-6 0·8 1·0 0·2 0·4 0·6 O·B 1·0

Weight fraction.xJ, (1-x,l

Figure 2. Segregation pattern--char and shale in sand. (Ref. 5 )

9

..

..

•

•

. . .

. .. . .

...

•

"

-Figure 3.

.·~ .

'

• ' ' ' ' '

RELATIVE CONCENTRATION

'

% component in increment average% component in bed

' ' ' ' ' ' ' " ' ' ' ' '

-·-

' ' ' .

' ' '

fu ~. t~i .... fi-fl r 1, .;,, r~, ~ _;1 ...

' ' ' \

\ • I

\

I \ C -•

~' ~ C:1 (1

·I ,~1 (~ r] gj ~ lf"' .,.

•:·.\ ,- . . ... _,,__ ...... . . . .

' \ \ \ \

f16

l,s I r' .. ~13

l,2

r'o ~~

ls I -r1 r6 rs r4

I

•

Distribution of coal and refuse with location in bed and with time. (Ref. 5)

10 . .

...

'

cylindrical bed fabricated from plexiglass for easy

obser~ation. of the solids. The grid was of .porous plate

design. The bed was designed to permit easy removal of

layers of the bed ma.terial through use of a vacuum .sampler.

In the batch experiments, solid concentration was measured by

suddenly shutting off the fluidizing air, defluidizing the

bed and measuring t-he composition of each layer (see Fig. 5) .

Based on the results from the above two beds, the 132cm

··(52 inch) long, 10.16cm (4 inch) wide rectangular bed

illustrated in Fig. 6 was designed. With this inclined

rectangular bed, material is fed to the bed at one end and

flows out from the other end while a vacuum sampler is used

N I I

to remove the material in the top layer.

Air Flow Rate Arrangement

Air flow rate was measured using four Schutte &

Koerting rotameters. The tubes and floats used for this

study, listed in Table 1, were selected to give a range of

air flow rates up to 0.694 m3; min (24.5 ft 3;' min) at

room temperature and pressure.

illustrated in Fig. 7.

Vacu11m Sampler

The flow circuit is

J . -

I: - I

The vacuum sampler is illustrated in Fig. 8. This

device consisted of a jet pump (model GL from Penberthy

Oudaille Co.), a gas flow control valve and a catch container -

for collecting the solids. The thicknesses of the bed layers

were measured with a millimeter scale.

11 ,.

•

..

•

•

•

:~ .

~ Distrib~tor

Pressure tap • Air from rotameter •

Figure 4. The first fluidized bed •

-

.. Pressure tap~~~~.

Figure 5.

.. . . .

.. . .

. .. ..

Detachable part for easy access to material in the bed

" '

Distributor

••,---- Air from rotameter

The second fluidized-bed. I

' 12 . '

•

Distributor

•

0 0

•

Air from rotameter

Figure 6. The inclined fluidized bed.

·-

Tube Float Max. flow (m3 I min)· Model Model at 21 ° c, 101.4 kpa

2-B R-22 0.022 ..

.

3-HCF 34-J 0.154

....

4-HCF 44-J 0.321

..

5-HCF 54-J. 0.694 .

.. . TABLE 1. Rotameter tubes and floats.

13

•

-

. . . .

•. • ! ••

4 Air outlets

5-HCF 4-HCF 3-HCF ... .

2-B

Flow control valve Flow control 0 valve

Pressure regulating valve

Figure 7. The flow Circuit.

Inlet

>:l I ~

Con trol valve '

·"'-Air inlet

Discharge

.

Suction Catch con tainer .

I

Measuring gadget •

I •

Figure a. Vacuum sampler.

•

14

t • I t· I

..

•

...

Solid Separation

Where.a sufficiently large size difference existed,

between the different materials used in a fluidization

experiment, the materials were separated after fluidization

by dry sieving.

For the experiments involving magnetite and coa.l, a

magnetic separator, purchased from Eriez Magnetics Company,

was used to separate the coal from the magnetite. This

device, Shown in Fig. 9, consists of a vibratory feeder for

transporting material uniformly to the belt and a belt and

pulley set. One pulley was made of a permanent magnet and

the other was made of steel. The speeds of the feeder and

belt are adjusted manually by the controller. By

appropriately adjusting the speeds of the feeder and the

belt, coal can be efficiently separated from magnetite for

different combinations of coal and magnetite.

A demagnetizing coil, also purchased from the Eriez

Magnetics Company, was used to restore the magnetite to it's '- _>

original unmagnetized state. This was necessary to prevent

agglomeration of the magnetite after passage through the

magnetic separator. This demagnetizing coil was held

vertically and the magnetized magnetite particles were poured

, .. through the coil. In preliminary tests, the magnetized

magnetite were found to be fully demagnetized after passing

through the coil twice.

15

•

... 0\

:

•.

r

' I

I

Controller

\

• '

•

.. ..

•

·-- Vibratory feeder

Belt & pulley set ::

..

Figure 9. Magnetic separator •

•

•

• '

. . ,.

Float and Sink Apparatus

The purpose of these tests is ·to determine the effect

of the size distribution of the crushed coal on its ability

to liberate and release pyrite.

The Float-Sink apparatus is made of Pyrex glass. The

two flasks are joined together and filled with solvent to

just above the top part of the joint. A weighed coal sample

and additional solvent are added to bring the liquid level to .

three quarters full. The· solvent and sample mixture is

stirred very well. This keeps the sample from balling and

insures uniform wetting of the sample. This is needed for

achieving proper segregation. Solvent is again added to

1.27cm (1/2 inch) of the top. The sample is again stirred

well in both the top and bottom flasks but gently to avoid

combining the contents of the two flasks. Then the stopper

is carefully inserted, the flasks are separated, the samples

are filtered and dried, and the corresponding sink and float

solid residues are weighed.

Material Preparation

Each kind of material, illustrated .in TABLE 2, was '. I

divided into narrow size ranges by use of dry sieving.

In addition, TABLE 2 also includes density, theoretical

minimum fluidization velocity and experimental _minimum

fluidization velocity for each size range of the materials.

17

..

..

'·

-

... ' .. .

. .. ..

Material ;_.,

Plastic

Glass

Magnetite

~

•

I

.Particle Particl-e (Umf) Diam. Range exp.

µ.m Mesh No. (Mesh No.) (cm/s)

850 20 d) 20 14.20

600 JO 20) d) JO 12.80

425 40 · JO) d > 40 8.70

.. • JOO 50 d) 50 4.60

212 70 50) d ) 70 4.11

1.50 100 70)d)l00 2.28

125 120 100)d)120 1.49

JOO 50 40) d > 50 13. 47 .

250 60 50) d > 60 7.37

212 70 60) d > 70 5.97 180 80 7q) d > 80 3.99 150 100 80) d >100 3.53 125 120 100)d>l20 1.86

106 140 120) d )140 1.37

90 170 140) d) 170 1.16

75 200 170)d>200 0.70

63 230 200) d )230 O. 67

45 325 2JO)d ;>325 0.52

I

·r·ABLE 2 : · List of Bed i1aterials

18

•

(U ) mfthe.

p (g/cmJJ

(cm/s) 20.47 1.00 "

13.39 8.J8

6.613 2.25

4.139 2.444

1.414

14.37 5.00

10.15

7.37

5.35 3.73 •

2.59 1. 87

1.35 0.94

. o.66

0.34

\ I I I t

••

/

.•

. .

...

-

' i •

. .. .. ..

. .,, . .

Particle Particle (Urnfix • (U ) p Material Diam. Range mfthe. (g/cmJ)

(cm/s) µm Mesh No. (Mesh No.) (cm/s)

Coal #1 850 20 d) 20 J0.20 28.50 1 • .50

600 JO 20) d) 30 22.50 20.10

425 40 JO) d > 40 12.50 10.60

300 50 40) d) · .50 7.00 6.80

250 60 50 > d) 60 2.70 2.50

212 70 60) d > 70 2,JO 2.20

180 80 70) d) 80 1.70 1. 80 · .

. 150 100 \ 80) d )100 a.Bo 0.70

. .

Coal l/2 850 20 d> 20

600 JO 20 > d) JO

425 40 JO) d )_ 40

·300 50 40) d ) 50 250 60 50) d) 60 .. II

212 70 60) d) 70

180 Bo 70 > d > 80

150 100 80) d ) 100

Coal #J 850 20 d) 20 . . . .. .

600 JO 20) d) JO 425 40 JO) d) 40

JOO 50 40)d > 50 250 60 50 > d) 60 " II

212 70 60) d) 70

180 80 . 70) d ) 80 . '

150 100 80) d )100 .

I

. Coal #1 - An· anthracite coal. 'Mineral contents is unknown.

Coal #2 - Raw coal from PP&L Rushton mine. ·, Coal #J - Ravi coal from Minesota Power.

II . . .

- ·rhe same as J.n coal dl. • • • t •

. ·rABLE 2 (continued)

•• •

19

.. . . .. -·

. .

Chemical Analysis (

In order to know the sulfur concertration in each layer •

of coal after steady state fluidization was achieved,

samples of coal from each of the tests and from each bed layer

were sent to the Warner Laboratory in Cresson, PA, for sulfur

analysis.

EXPERIMENTAL PROCEDURE

The experiments were divided into three phases. In the

first phase, the 10.16cm (4 inch) diameter bed was used to

obtain the minimum fluidization velocities of each size of

all the materials mentioned above. This was accomplished by

measuring the bed flow rate -- pressure drop profiles. The

pressure drop between the top and bottom of the bed was

obtained by subtracting the pressure drop across the grid

from the gage pressure measured at the plenum.

In the second phase, the detachable 15.24cm (6 inch)

diameter bed was used. Experiments for each particle

combination were preceded by explorary runs to determine the

time required to reach steady state. These runs consisted of

repeated measurements of the composition at the top of the

bed to observe its change with time.

Each experiment was run with the bed initially well

· mixed. The different particles were added simultaneously to

the bed and fluidized at velocities, much higher than the

minimum bubbling velocity, for 5 minutes to insure that all

the particles were well mixed. Then, the air was shut off

20

\

•

(

suddenly and increased gradually to the desired velocity.

Most experiments were then run at the minimum bubbling

velocity. After the bed was fluidized for 30 minutes, the

air was shut off abruptly. The materials was removed layer

by layer by use of the vacuum sampler. The number of layer

depended on the bed depth, with each layer typically being 1

to 2 cm thick.

Flotsam concentration p·rofiles were obtained by

separating the flotsam from the magnetite in the magnetic

separator.

In the final phase, the 132 cm (52 inch) long 10.16 cm

(4 inch) wide inclined bed was used. Fluidizing experiments

were performed at minimum bubbling velocity with the bed

inclined at different angles. During these tests,material was

fed in at one end and removed at the other. Tests to measure

residence time of particles in the bed as a function of feed

rate of bed material and angle of inclination were carried

out.

21

'

"

RESULTS

Fluidization Behavior of Mixtures

In the first phase of the experiments, each of the

samples listed in TABLE 2 was loaded separately into the

fluidized bed and bed pressure drop was measured as a

function of fluidizing velocity. Shown in Figure 10, 11 and

12 are typical pressure drop -- flow rate curves for three

materials. The point of minimum fluidization is indicated in

each case and is also tabulated in TABLE 2.Figure 13 contains

a summary of minimum fluidization velocity as a function of

particle size for each material. Even though the three coal

types differ from one another, no significant difference

exists in their minimum fluidization curves. Therefore, the

Umf versus particle size characteristics are represented by

the curve marked" Coal" in Figure 13.

Because of its low density, plastic with particle

density of the order 1 g/c.c should be closest in behavior to

the carbonaceous portion of the coal. It was, therefore,

expected that when the experiments were run with mixtures of

coal and magnetite, the behavior would be at least

quantitively similar to that of plastic and magnetite. Both

plastic and glass possess the property of being easily

observed in mixtures with magnetite.

The minimum fluidization velocities of binary mixtures

of magnetite with plastic and glass were measured and some

comments on the results are shown in TABLE 3.

22

'

AP [cm Wate~]

20

15

10

5

•

•

,

I

I

I

I

• I I

I

I

•

•

1 Umf I I I

0.5

·' . ~

• •

'

' •

Material: Glass (50 Mesh)

• - . •

1.0 1.5 2.0

• 3 [m x 10 kg/sec]

Figure 10. Typical fluidization curve for glass. •

•

I I .

. I

.-

•

AP , I

,

(cm Water]

20

15 •

10

5

I I

I I

I I

I

' I

' • I I I I I Umf I I

0.5

,.

• • ,·

Material: Magnetite (140 mesh)

1.0 1.5

Figure 11. Typical fluidization curve for magnetite .

•

'

•

..

•

2.0 • 3

[m x 10 kg/sec]

..

c-

•

6P [cm Water]

20

15

10

5

r·

,

0.5

l • I I I I I

• • I

. •

I

Umf

1.0

.

Material: Plastic (30 Mesh)

• •

1.5 2.0 • 3

[m x 10 kg/sec]

Figure 12. Typical fluidization curve for plastic .

. I

•

• •

..

.

.

I I . . •

..

•

Umf [cm/sec]

20

15

10

5

.. .

.

' I

. -

l

,

100

Figure

•

13.

• •

. ;t

I

. .

•

Magnetite

600 . .. 500 . 200 300 400

Variation of minimum fluidization velocity with particle size .

• •

• ,

Coe.l

-~

Plastic

700 800

d [µ.m]

.. •

·•

Minimum fluidization velocity of mixtures of 50 mesh

magnetite w·i th plastic is plotted as a function of mixture

concentration in Figure 14.

Figure 14 indicates that the minimum fluidization

velocity of the mixture typically lies between the two

individual minimum fluidization velocities but it is a little "

higher than either in the range between 65 and 75 volume

percent of magnetite.

Each of the mixtures listed in TABLE 3 was fluidized at

its minimum fluidization velocity for ten minutes. It was

observed that when particles individually having identical

minimum fluidization velocities were mixed together, they

fluidized uniformly with bubbling occuring over the entire

height of bed and with relatively little segregation of

materials. One example of this situation is a mixture of 50

mesh magnetite with 20 mesh plastic.

In cases where the minimum fluidization velocities of

the individual materials are not close to each other, then

the material with the larger minimum fluidization velocity

tends to segregate and settle to the bottom of the bed with

the other material floating to the top. Two examples of this

situation are mixtures of 50 mesh magnetite with 40 mesh

~ -. plastic and 6 o mesh magnetite with 7 o mesh glass.

Finally, if Urnf 1 >> Umf2 and Umf 1 > Ugas > Umf2,

component 1 settles to the bottom of the bed and forms a

packedlayer while component 2 forms a fluidized layer on I )

27

I

.·

N a,

.

,

-

•

TABLE 3. Minimum fluidization velocities -of different mixed material.

Plastic (Mesh Size) Glass (Mesh Size) . Material

30 40 20 50 \

Umf [cm/sec] 12.80 8.70 14.20 4.60 .

Volume ~ .

50 r 30 10 50 30 10 30 10 0

Magnetite 50 14. 57 + 12.83•

50Mesh - " 15. 64 + 14. 39 + 70 15.64 +. •

Umf: - -. ' . =13.47 cm/s 14.75+ 12.95 ' 90 ' .• +

Magnetite· 50 , f?.O·Mesh •

70 Umf =7.37 cm/s 90 10.79+ 11.13 + 9. 36 +

+ Well mixed with no segregation.

* Good segregation. Bed free surface has almost 100% glass.

** Bottom one third of the ved is packed. The rest of the

bed has a larger flotsam concentration in comparison to

the bottom.

70 100 :

4.11 2.28 !

50 -~, 50 J

-

7.89•~ 4. 72'!- j I

r i l I

·-

I I

' ;

i

I i .

' i ' I ' I -

20

Umf

[cm/sec]

15

10

5

•

30 Mesh Plastic with 50 Mesh Magnetite

40 Mesh Plastic with 50 Mesh Magnetite

25 50 75 100

--(volume% of magnetite)

· Figure 14. Minimum fluidization velocities of binary

• mixtures.

29

.....

• _________ _....,_ .. _A. ... ----.-..-.. -----·-·--~ ............. . ----~ .... • -· ··-· Ao•... ·#-~ ·- -·~ . ·• ....... -.. ~ ......... #----· ·--·· -~··· -~-..... -·---- ........... ·- .... . ....

top. This occured,for example,with a mixture of 60 mesh .

magnetite and lOOmesh glass.

Effects of Gas Velocity and Bed Depth

In the second phase, experiments were carried out in

the 15.24 cm (6 inch) diameter bed to determine the effect of

bed depth on the time required for the bed to achieve steady

state conditions. In these experiments, a mixture of 10

weight percent coal and 90 weight percent magnetite was added

to the bed, fluidized at an air velocity much larger than the

minimum bubbling velocity for 5 minutes to mix the mixture

well and then fluidized at near minimum bubbling velocity

for different periods of time. (The minimum bubbling velocity

is defined as the velocity at which small bubbles begin to

form at the distributor. From experiments, Umb / Umf usually

lies between 1.6 and 1.8,and Umb can be determined by

observation when bubbling begins to occur.) The air supply

was then shut off, the bed material was sucked out and

analyzed layer by layer. The fluidization time was increased

and the experiment was repeated. This procedure was continued

until the concentration distribution remained unchanged with

duration of the experiment.

The results are summarized • in Figure 15 where the

'transient response time of the bed is plotted as a function of

bed depth. It was found that the time to reach steady state

increases sharply with bed depth. Beds of just a few

centimeters in depth reach steady state in 2 to 3 minutes

30

..

..

..

.,

• •

' . .

•

. ··- ..

-... ·,

. .

. . . . .

•

..

...

30

25

\ " 20 ~ .,... a '-"'

cu ~

-rt +l

Q) 15 Ul ~ 0 p. Ul (1)

P::

10

5

•

l 2 3 4 5 6 7

Bed depth (cm)

8

•

. •

9

Figure 15 • The effect of bed depth on the time to reach stea1y state conditions.

31 ':" .

•

•

.• •--

'

I • • • I

10

• ~ ...

•·

•

. .

••

while beds of 10 cm in depth need 25 to 30 minutes to reach

steady state conditions.

In the segregation experiments which follow, each

mixture was fluidized for 30 minutes to insure that steady

state conditions were reached, then the bed material was

removed layer by layer through use of the vacuum sampler to

obtain the concentration profile as a function of distance

from the top of the bed.

The mixtures of 50 mesh magnetite with 30 mesh plastic

and of 100 mesh magnetite with 60 mesh coal were

fluidized individually at different air flow rates to

determine the effect of air flow rate on segregation

conditions. The results for the mixture of 50 mesh magnetite

with 30 mesh plastic, shown in Figure 16, demonstrate that an

air flow rate close to minimum bubbling velocity has the best

effect on segregation. The results for a mixture of 100 mesh

magnetite with 60 mesh coal are shown in Figure 17. No

segregation was achieved at either flow rate because these

materials have almost

velocities.

identical • • minimum fluidization

The effect of bed depth on the shape of the steady

state concentration profile was determined from the

._ experiments with mixtures of 170 mesh magnetite and 100 mesh

coal. The results are shown in Figure 18 as weight fraction

of coal versus the dimensionless distance below the free

surface of the bed.

32

... • •

. . ..

..

,.__ # '-"

u •rt .µ Ul

... ro r-i ~

~ 0

C • 0

•rt .µ u Al ~ ~

(l) J:; ~ r-i 0 >

\

...

•

60

50

40

30

20

10

' 0 I

' ' ' ~\

Uo

• -.... )(· ....

[cm/s] · Uo/Umb

10.7 0.75

14.2 1.00

18.0 1.27

Mesh Volume%

·,. mag 50 80

1

,· .. \ ..

plas 30 20

\ '--J. \ .

• \ .

•

\ : \ ·. \ .

• \ •• >(

X • \ .. • •

• • ~ •• '- .

'o )( '-...... . 0

0 ...... • --- ....__Q . ' . '

''M.. 'o\ '•)( \

• • • • . ')(

• • r

2. 3 4 5 6 7 8

Bed Depth (cm]

0 \ \ \ • . ~ ·. ', 'i( ....

9 10

Figure 16. The effect·of air flow rate on segregation .

... .

33

. .

•

.. I• • I•

•

..

..

•

' - - ---'

' ·' .

•

• •

-

. .. ..

.. . • ..

...

•

50

,.....

"" '"" r-4

40 ,0 0 u

q..,. 0

t: 0

~r-4 Jc .µ 30 , ___ ,,,, ~ u «s Jc ' ~ µ..

Q)

a :;:,

,...,t

0 20 :>

10

. ..

o •. I • •

1 ·2 3

Uo [cm/s]

- -----x--

•

)( ·..-.

3.2

3.9

·mag

plas

)(-- --.. ... ,..._,- t'

'

. -:

Uo/Umb.

0.82

1.00

Mesh Volume%

100 80

60 20

" It ~, "lt' ••

• ·-• • • • •

•

4 5 6 7 8 9 10

· Bed Depth [cm] •

. Figure 17. The effect of air flow rate on segregation.

34

•

•

•

•

...

. .. •·

• •

.. .

. .

The mixing index is a method to judge the efficiency or

quality of segregation. Different • • mixing indexes provide

different ways to evaluate the degree of segregation. For

objectivity, four different mixing indexes are shown in

TABLE 4 to compare the segregation conditions in mixtures.

It was found that the differences between the profiles of

shallow and deep beds are relatively small, thus indicating

that the depth of the bed has very little effect on

efficiency of segregation.

Sink and Float Tests

In the sink and float tests, 50 gram samples of coal

N0.2 .and.coal NO. 3 (coal N0.2 -- raw coal from PP & L,

coal N0.3 -- raw coal from Minnesta Power) were tested with a

liquid solvent of 2.0 specific density. The purpose of these

tests was to determine the effect of the size distribution of

the crushed coal on its ability to liberate and release

pyrite. The coal was separ~ted into narrow fractions. The

number of grams of feed coal which sank in the heavy medium

was then determined as a function of the size of the feed

coal. These results are shown for the two coals in TABLE 5

and Figure 19. In both cases a relatively large increase in

the amount of material which sank occured in the size range

~ between 80 to 100 mesh. This most likely indicates a critical

size below which more extensive liberation of the pyrite

begins to occur. Interestingly, in the case of the Minnesota j Power coal, almost none of the coal sank until the grid size

35

.. '

. . .

•

1·

.

• • f I original·coal weight concentration 7 0

\ - .3.81

' \ \

. .. . .

\ \ o 'I I I

6 ~ \ .. ' . \.\ • • 7 cm ••

• ', \ • -0-0--0 4 cm

' s:: 5 ~ \ --*---*- 3 cm 0

-r4 ' 2 +> .~\

cm - •

«s • M \\ ~+l ~ Q) ~' ... r-4 u

«s ~ 4 ' '- ··o, 0 0

,_ -- ' I (J u -.)( '

' ' ~"- ' '\o ' - ' 4, \ ' ••A .. 3 \ 4 ..

\

' \ • f

X 0

- \ ~---~ ·2

1

o. 2 5 0.5 0. 7 5 1.0 .. .

Z / L.( z: distance from the free surface)

Figu~e 1s. The effect of bed depth on segregation. _.., .

• -....

36 ...

..

... -- - ... .a .... --- .. -·--····-········· .

·- •• ·~- . ·' - • -J- • • f •

r' I I

,_ . . . - ..

e I I""'

. ... . . .. . ' .

• 1 "• r ,. l"-r, '' ' I • ~ !' ... > f •; I

• I:• , .. '

'··,, ~. _: .. _.:, .. _ ... ..: ......... · .. ,

. ~ - .......... --- . - . .. .. ..

• • • I

....

TABLE 4. The mixing index calculations. --· -

•' ·- - -- - . .

Xj 0.90 (average jetsam volume percentage) --H=2cm H=3cm H=4cm H=7cm

% X % % X % % X % % X %

(cm] w V [cm] w V (cm] w V :cm] w V

o.5 0.5 17 0.5 7. 0 1 7

1.0 4~0 10 1.0 4.0 11

1.5 3.4 9 1.5 4.0 11

2.0 2.0 6 2.0 3.8 10

2.5 3.5 9

3.0 3.5 7

Il 0.922 0.966 •

I2 0.0733 0.0677

I3 0.922 0.922

I4 0.865 0.870

Xhdh

Il =

h=H

h

!2 =

X = X H H

n 2 L (V - V ) i=l j i j

~

0.5 6.0 15 1.0 7.2 lS

1.0 4.2 11 2.0 3.5

2.0 3.7 10 3.0 3.3

3.0 2.6 7 4.0 3.3

4.0 2.3 6 5.0 3.1

6.0 3.0

7.0 3.0

0.966 0.911

0.0566 0.0877

0.944 0.911

0.885 0.884 .

X (H- h)

I

H X = X

(1 - -V ) j

H H

-* V , • J

9

9

9

8

8

a

.

Perfect Perfect I •

segre . mixing

0 l·

1 0

0 1

0 1

, (Tho-Ching Ho, Kirkpatrick, Wang, Hopper)

(Daw)

I3 = X / X , (Rowe, Nienow, Agbim)

I4 =

tpp layer bed

the .... % refuse in • 1n the produ~t recovery of the

•

the product at a 50 % mixture.

37

, (Weintraub, Deurbrouck ,Thomas)

. ---·-·------· --~- .. - -·· .

. ··- ··~•· -~ .~.,., .. .,.. ~ ·-- ....... _._... ... . ---.----------------... --····--- .. . '

..

TABLE 5

Effect of size of particles on mass of coal sinking during sink/float test. Based on 50 gram samples.

Size ----# 30

# 40

# 50

# 60

# 70

# 80

#100

#140

Minnesota Power Coal

(g) ----1.0

1.3

1.2

1.5

3.0

3.0

9.0

11.0

38

PP&L

(g) ----

4.1

4.5

3.5

5.3

5.5

4.5

6.0

8.5

....

•

..

'

-

,_, •• _ ........ - ...... • ••-• I '

•

From sink and float test: ....

PP&L coal. ·

---------- Minnesota Power c.oal.

25%

20%

15%

10%

5%

• •

• I I

• I ~ / )c

/ /

-- --~-r /)( )C

X • I I

I )( /

I I

X / I.

I

I I

I

I I I

40 50 60 70 80 100 140 <140

Mesh size

Figure 19. Sink/ float test.

39

. .

.. ,

•

-----------...,....-~---=sc-.,-,. --,"·,

was reduced to below 60 mesh, whereas in the PP & L coal, at

least s percent sank even in the larger fractions. This may •

indicate that the PP & L coal is amenable to physical

cleaning in larger size fractions than is the case for the

Minnesota Power coal.

Fluidized Bed Segregation with Coal and Magnetite

Experiments with mixtures of magnetite and coal were

carried out to measure the coal concentration and sulfur

concentration in the coal as a function of the dimensionless

distance from the top of the bed.

These experiments consist of two stages. A batch of

PP & L coal was separated into narrow size fractions, each

of which was weighed as shown in TABLE 6.

In the first stage of segregation, each of the coal

fractions (200 gram) was fluidized with the suitable size

fraction of magnetite (1800 gram). The minimum fluidization

velocity of the magnetite was chosen to be slightly higher

than that of the coal to achieve good segregation conditions.

Initially the mixture was fluidized at an air flow rate much

higher than minimum bubbling velocity for 5 minutes to mix

the mixture well. It was then fluidized at minimum bubbling

velocity for 30 minutes to achieve steady state conditions.

Bed depths were in the range of 5 to 8 cm. Divided into four

layers, the material in the bed was sucked out layer by layer

through use of the vacuum sampler. Each layer of material was

put into the magnetite separator and the magnetite was

40

•.

.-TABLE 6. Size distribution for PP&L coal

Particle Diam _____ ... ______ _ m Mesh No.

425

300

250

212

180

150

106

40

50

60

70

80

100

140

<140

I

Particle Range (Mesh No.)

30 > d > 40

40 > d > 50

50 > d > 60

60 > d > 70

70 > d > 80

80 > d > 100

100 > d > 140

140 > d

41 •

Weight (g)

1494

1500

1167

451

782

1160

468

1023

'

Weight%

19

19

15

6

10

14

6

13

'

·,

separated magnetically from the coal. Pyrite is not

ferromagne~ic, so the liberated pyrite goes with the coal

fraction. By weighing the coal in each layer, the coal

concentration profiles were obtained and by chemical analysis

of the coal samples from each layer, the sulfur concentration

in the coal was obtained.

In the second stage (top), the top coal layer from the

first cleaning stage was added to clean magnetite to form 10

percent coal mixtures. In the second stage (bottom), the

bottom layers of the beds from the first cleaning stage

was also added to clean magnetite to form 10 percent coal

mixtures. Segregation experiments were then repeated as in

the first stage. This process is illustrated in Figure 20.

The purpose of this second set of experiments was to

determine the effect of multistage operation on coal cleaning

ability. Samples of coal from each of the tests and from

each bed layer were sent to a testing laboratory (Warner

lab.) for sulfur analysis.

The coal concentration profiles in the first stage,

second stage (top) and second stage (bottom) are shown in

Figures 21 to 25. These show that the coal concentration

profiles in the first stage, the second stage (top), and

second stage (bottom) have almost identical profiles for each

combination of coal and magnetite. A summary of the results

is shown in Figures 26, 27 and 28.

From the results of the chemical analyses for the

42

•

•

•

'

'

..

-

-

coal and magnetite ( narrow size fraction)

If'

"

'

--

:,

Magnetite

-

----..

.,_ ___ ... -.....

-- .

---.....

TWO identical mixtures in the 1st stage

... ,

.... •

....._ ,

'

....... ,

I ...... ,.

.... ,

Magnetite

The· second stage (top) The second stage (bottom)

Figure 20. Multistag.e process • • ••

43

I

• f

.. . • f

..

•

•

•

•

•

..

..

•

• •

. .. . .

-••

· ....

201 o

15%

101

51

Original coal weight% -- 101

40 mesh coal & 50 mesh magnetite

4

..

+ -The 1st trial in the 1st stage. 6 -The 2nd trial in the 1st stage. • -The 2nd stage (top). 0 -THe 2nd stage ·c bottom).

+

•

0.25 0.5 o. 7 5 1.0

Z / L

( Z: distance. from the free surface) •

, • I I I

..

Figure 21. Coal concentration profiles (40 mesh coal).

••

44

.. . . ~-

. .

I

•

•

•

. .

.. • I

• .. ' .

..

.. •

•

•

-

. .. . ..

--

i:: 0

-ri +J •, «s .:. . ~

~ .µ s:: Q)

,-t u «s s:: 0 0 uo

251

20%

15\

101

51

• •

• •

..

Original coal weight I -- 10%

60 mesh coal & 70 mesh magnetite

+ -The 1st trial • the 1st stage. in 6 -The 2nd trial in the 1st stage. e -The 2nd stage ( top ) . O -The 2nd stage ( bottom ) .

• •

• 0 +

•

•

I

0.25 0.5 o. 7 5 1.0 z / L

( z: distance from the free surf~~~) .

Figure 22. Coal concentration profiles (60 mesh coal) . ..

45

,,.

.. . . .. .

•·

• •

.. .

I

'

. , I

• • •

' I

\

..

• •

.. .

-.

-

..

J

• 251

.•· 0

20% +

151

101

51

J

Original coal weight% -- 10% 80 mesh coal & 100 mesh magnetite

+ -The 1st 6 -The 2nd e -The 2nd 0-The 2nd

•

0.25 I

trial trial stage stage

•

•

• the 1st stage . in in the 1st stage. ( top). (bottom).

A ........ -----oo- . +

~

•

o.s 0.75 1.0 Z / L

..

( Z: distance from the free surface)

Figure 23. Coal concentration profiles (80 mesh coal) .

46

..

•

•

'

•

.. ,.

• •

-

- .

...

'

..

•

. .

•

'

. . ..

~ 0 ..... .µ &U

• M ~.µ

i:: Cl)

ri u cd h 0 0 uo

...

251

201

., 15%

101

51

,.

I I • I f

, ••

original coal. weight I -- 101 .. . . .

'•

100 mesh coal & 140 mesh magnetite

+ -The 1st trial in the 1st stage • ~ -The 2nd trial in the 1st stage. • -The 2nd stage (top). o -The 2nd stage (bottom).

•

• 0 9 \ -f

. ' 0

A A

A

0. 2 5 0.5 o. 7 5 . 1.0

Z / L

( z: distance from the free surface)

Figure 24. Coal concentration profiles (100 mesh coal) .

47

•

..

•

•

•

• •

-

. .

.. . ~. -

~ ..

. . •,•

...

.. .

. .

. .

r

...

. ,•

~ 0

-ri .µ ftS

• H ~~

i:: cu

.-t u ,a s:: 00 uu

251

20\

15%

10\

5%

'•

Original coal weight I -- 10 I

140 mesh coal & 140 mesh magnetite

+ ·-The 1st trial • the 1st stage. in A -The 2nd trial in the 1st stage. e -The 2nd stage ( top ) . o -The 2nd stage ( bottom ) ·. •

•

• ..

o. 2 5 0.5 0.75 1.0

Z / L

( Z: distance from·the free surface)

Figure 25. Coal concentration profiles (140 mesh coal).

48

•

·... .

•

•

•

...

, .

..

• •

•

original coal weight I -- 101

... <D 40 mesh coal & 50 mesh magnetite • - 60 mesh •

coal & 70 mesh magnetite 80 mesh coal & 100 mesh magnetite

mesh coal & 140 mesh magnetite •

25% - @ l~O mesh coal & 140 mesh magnetite . . . •

i:: 0

•

-r4 .µ

20% ., • M

~+J C Q)

~u ftS i:: 00 00

15%

. -.· ..

,

10\ .•

..

5\

...

.. .

... 0.25 0.5 0.75 1.0

... Z / L

( z: distance from the free surfa~e)

•

" Figure 26. Coal concentration profiles in the 1st stage. Summary fer all size fractions.

I •

. . . ' • 49

. ..

•

. ..

• • -

.. ;

•

.

..

. .

...

- ~

..

..

•

original coal weight% -- ld%

25\ ·mesh magnetite (1) 40 mesh coal & 50

$ 60 mesh coal & 70 mesh magnetite·· 80 mesh coal & 100 mesh magnetite

.. s:: @ 100 mesh coal & 140 mesh magnetite O· @ 140 mesh coal & 140 mesh magnetite •r-1 .µ 20% RS

• H ~.µ

i:: Cl)

r-40 cd i:: 0 0 uo

151

10%

5%

o. 2 5 0.5 0.75 1.0 Z / L

( z: distance from the free surface) .,

Figure 27. Coa~ concentration profiles in the 2nd stage (top). Summary for all size fractions.

50

•

.

• •

.. .

.. .

. ..

•

•

..

- ..

...

• •

• !- I· I •

.- . . . '·

••

•• •

Original coal weight% -- 10%

40 mesh coal & 50 mesh magnetite 60 mesh coal & 70 mesh magnetite··

251 80 mesh coal & 100 mesh magnetite 100 mesh coa·l & 140 mesh magnetite

• 140 mesh coal & 140 mesh magnetite . .

\ ..

,::: 0 .. ..... 20% +l ,u . ~

~.µ ~ Cl)

r-4 0 ctl. ~ 0 0 uo 151

10%

5%

• •

. -.. ., . . o. 25 0.5 0.75 1.0 z I L

.... ( z: distance from the free surface")

Figure 28. Coal concentration profiles in the 2nd stage (bottom). Summary for all size fractions.

51

I

•

•

I

. .

.

'-..

parent coal, the sulfur concentration is plotted as a

function of size fraction of coal in Figure 29. Compared to

Figure 19 in the sink and float test, the sulfur

concentration curve in Figure 29 and the sink portion . ,

concentration curve have the same basic shape.

The sulfur concentration profiles in the first stage,

second stage (top) and second stage (bottom) are shown in

different ways from Figures 30 to 37. In all stages of

cleaning, the sulfur concentration in the coal in the top

layer is always less than that in the other layers. This

occurs because pyrite has almost the same density as

magnetite and tends to be distributed uniformly within the

magnetite. Coal, in the other hand, floats to the top of the

bed resulting in a layer of coal which is deficient in

sulfur. In all cases, the second stage of cleaning resulted

in additional reductions in sulfur concentration for the coal

at the top of the bed.

Inclined Fluidized Bed

In the final phase, the inclined bed was fabricated,

installed in the laboratory, debugged and tests were carried

out with it. Fluidization experiments were performed at

minimum bubbling velocity with the bed inclined at different

angles. During these tests, material was fed in at one end

and removed from the other. The results showed that the bed

was fluidized uniformly over the entire length. Tests to

measure residence time of particles in the bed as a function

52

., .

..

-

:~ .

s:: 0 ~

~ ~ ::s ctS ~ ~ M .µ ::s s:: tJl (1)

0 s:: 0 0

•

4

3

' 2

1'

1

40

'

..

-

50

{

Instrumental sulfur

- - - - - Pyritic Sulfur

•

"""" , / ,,,,

60

Mesh

• I /v

)( /

• / 'C /

~ /

' ' -- ,,

,<

70 80 100 140 >140

• size

Figure 29. Sulfur concentration in

raw coal from PP & L

53

..

•

I

•

...

•

•

'

•

'

• . •

'

. . .

. .. ..

.. .

.. .

..

..

. .

4\

3%

2%

1% .'•

.... .

(

••

Original coal weight I -- 10% •

40 mesh coal & 50 mesh m·agnetite ••

+ -The 1st trial in the 1st stage • A -The 2nd trial in the 1st stage. • -The 2nd stage ( top ) .• o -THe 2nd stage (bottom) .

A --

Z:

I

l .. I I •

I I I • I

o. 2 5

' ) , '

distance

• •

0.5 o. 7 5

z I L .._

from the free surface

I • I ' f

1 • 1 I I

.. ...

1.0

..

)

Figure 30. Sulfur concentration profiles (40 mesh coal).

\_/ 5 4

..

•

•

,

I I

-

• • •

•

I U It

orig~nal coal weight% -- 101

60 mesh coal & 70.mesh magnet~te

+ -The 1st trial in the 1st stage. ~ -The 2nd trial in the 1st stage. e -The 2nd stage (top). o -The 2nd stage (bottom). • •

• . .

.. . _ ..

i: I 0

I - ..... -· -~ 4\ ~ «J

M ..µ 0 .. ~ i::

:l Q) ~o .-i ,::: ::1 0 (I) 0 3%

' I I ..:

I

. . . 2%

I

I • . . w.

I I

,., •

1% I -I I .. I

0.25 0.5 0.75 1.0 - .. .. . . . :, ·' Z / L

( z: distanc& from the free surface) .... )

Figure 31. Sulfur concentration profiles (60 mesh coal).

55

.... .. • •

' I

. . ...

. .. ..

•

.. .

•

..

•

•

-

-..

. .

..

•

... .

~ 0

•rt 4% • .µ ~ RS

H .µ

J.-4 i:: :::i QJ ~o r-i i::: :::i 0 Ul 0 3%

1%

·:- .

Original coal weight t -- 10% ' ' • ' I

· 80 mesh coal & 100 mesh magnetite

+ -The 1st trial in the 1st stage. A -The 2nd trial in the 1st stage. • -The 2nd stage (top) . o -The 2nd stage ( bottom ) •

I l I I I •

I I

I

+ •

• +

I

I

I I I I I

.. ,

~ . . ..

0.25 0.5 0.75 1.0 z / L

( Z: distance from the free surface:

Figure 32. Sulfur concentration profiles (80 mesh coal).

56

..

•

'.,

•

. .

••

-

..

. . .,

• •

-

•

. '

-. .. ..

..

•

....

'j

'"., ···!'-,

•

· ...

4%

3\

2%

1%

.

'

original coal weight% -- 10% 100 mesh coal, 140 mesh magnetite

+ -The A -The • -The o -The

I I

I

I I

I I

I

I

let trial 2nd trial 2nd stage 2nd stage

'·-

-t·

0. 2 5

I the 1st stage. in in the 1st stage. ( top). ( bottom ) .

A

+ +·

•

0.5 0.75

z I L

•

I

I 1.0

.... '

( z: distance from the free surface )

Figure 33. Sulfur concentration profiles (100 mesh coal) •

57

•

·.-.. ..

•

..

•

-:-"'""'

. .. .

I

•

-

, ...

. •

I ·•• ••

•

.. .. .

..

..

..

-

·' .

. .

.... .

.... .

4%

3\

2%

1%

•

Original coal weight% -- 10%

·140 mesh coal & 140 mesh magnetite·

+ -The A -The e -The O -The

I I I I I

I I

1st 2nd 2nd 2nd

I I I I I I I

0.25

trial trial stage stage

A

in the 1st stage. in the 1st stage. ( top). ( bottom ) •

...

0 .. :t

-to

A

·O. 5 o. 7 5

z I L

1.0

( Z: distance from the free surface)

Figure 34. Sulfur concentration profiles (140 mesh coal).

. . ..

. . 58

•

•

•

.

. •

.

'

...

..

-

-

.. .

•

..

original coal weight% -- 10 I · •

(I) 40 mesh coal & 50 mesh magnetite 60 mesh coal & 70 mesh magnetite

mesh coal 100 mesh magnetite 80 &

.... ~ 100 mesh coal & 140 mesh magnetite 140 mesh coal & 140 mesh magnetite

® • ..

.'•

i:: . 0 4\

•rl • .6,J ~ «J

M .µ

M S:: :::J Q) G). 'H 0 r-i J:'.:: :::1 0 3% Ul 0

0.25 0.5 0.75 1.0

Z / L

( Z: distance from the free surface)

Figure 35. Sulfur concentration profiles in the 1st stage. Summary for all size fractions .

•

59

. •

. ..

-

-• ' I '

• •

. . ..

.

•

. .. ..

•

-

-..

•

• •

.

•

Original coal weight% -- 10%

(D 40 mesh coal & 50 mesh magnetite ... •. ® 60 mesh coal & 70 mesh magnetite

® 80 mesh coal & 100 mesh magnetite 100 mesh coal & 140 mesh magnetite

• 140 mesh coal & 140 mesh magnetite

,:: 0

•ri 4% • +l ~ «s

H +l

M s:: ::I Q) CH 0 .-. ,:: ::I 0 Ul 0 3%

~--~@------------2%

1%

0. 2,5 0.5 0. 7 5 1.0

Z / L

( z: distance from the free surface)

Figure 36. Sulfur concentration profiles in the 2nd stage (top). Summary for all size fractions.

. . . ..

•.

60

•

• •

'

-/ ...... ~,

...

-

•· . • •

. '. . .

..

. .

.. .

.. -

...

•

..

. .

•

.. . . ..

....

i:: 4\ 0 -ri

• .µ ~ ro

~ .µ

M i::: ::, (1) ~u ,... i:: 3% :l 0 Ul 0

2%

lt

original coal ~eight% -- lOt

(I) 40 mesh coal & 50 mesh magnetite 60 mesh coal & 70 mesh magnetite 80 mesh coal & 100 mesh magnetite

100 mesh coal & 140 mesh magnetite 140 mesh coal & 140 mesh magnetite

o. 25 0.5 o. 75 .

Z / L

•

I 1.0

( Z: distance from the free surf ace, )

Figure 37. Sulfur concentration profiles in the 2nd stage (bottom). Summary for all size fractions .

61

.'• •

•

•

1. •

..

•

of feed rate of bed material and angle of inclination were

carried out and these results are shown in TABLE 7.

Visual observation of the bed shows that when a

suitable mixture of magnetite and coal is fed to the bed, the

. exit stream is highly segregated, with the top layer rich in

coal. The results are shown in TABLE a.

CONCLUSIONS

A study of the fluidization experiments of coal and

magnetite in a air fluidized system has resulted in the

following conclusions :

. When particles which individually have identical

minimum fluidization velocities are mixed together, they

fluidize uniformly with bubbling occuring over the entire

height of the bed and with relatively little segregation of

materials. • • minimum fluidization velocities of the

• When the

individual materials are not equal, the material with

the larger minimum fluidization velocity tends to segregate

and settle to the bottom of the bed with the other material

floating to the top .

. If Umfl >> Umf2 and Umfl > Ugas > Umf2, component 1

settles to the bottom of the bed and forms a packed layer

while component 2 forms a fluidized layer on top. Therefore,

some amount of component 2 which is trapped in component l

has no chance to rise upward .

• The time to achieve steady state conditions increases

62

·---... . . .. -·------... , ...•. -... _ .......... ------ff .. -.... ~ •• ....... __ ............. ·-••• • '•· ..... .._...;. .. .._ ........ • '• I t ...... I"••-••"·- - ...... , .... , I

"

,.

TABLE· 7. Test results with inclined bed.

....

-

•

Q 8

(c.c/sec) {degree)

0 .14

42 0.28 0.55

. 1. 1

0.14 0.28

25 0.55 1. 1

a .14

10 0.28

-- 0. 55 · · . l . 1

Q - material feeding rate

8 - angle of inclination

.h1 - bed depth at inlet

h2 -_bed depth at exit (

bl

(cm)

9 9 8 6

.

9 .

8 7 6

8 .. 8 7 f,

63

h2 Resident Time

(cm) (min:sec) .

3 2:50 3 2:45

2.5 2: 15· 2.5 1:55

2.5 6:30 2.5 6:25 2.5 5:45 2.5 5:30 ....

2.5 10:30 --· . -- - - .

2.5 10:30 .,

2.5 9;50 .2.4 9:30

....

-·

•

TABLE 8

TEST RESULTS WITH INCLINED BED

. Resident

Q e hl h2 Time Mixture (C.C/sec) (degree) (cm) (cm) {min:sec)

1 10 0.14 8 2.5 10:30 ' •

~

2 10 0.14 8 2.5 10:30

1 -- mixture of 10% Weight mesh 60 PP&L coal

with 90% Weight mesh 70 magnetite.

2 -- mixture of 10% Weight mesh 80 PP&L coal ¢

with 90% Weight mesh 100 magnetite.

64

..

Coal W. Concentration • Top Layer lil

21%

23%

sharply with bed depth. ..

.Air flow rates close to minimum bubbling velocity

(Umb / Umf = 1.6 to 1.8) has the best effect on segregation.

; The depth of the bed has little effect on efficiency

of segregation .

. With both the Minnesota Power and PP&L coals, finer

particles release more pyrite and there is a critical size ..

below which more extensive liberation of the pyrite begins

to occur .

. The coal concentration profiles are qualitatively

similar in shape for all cases studied .

. The sulfur concentration in the coal in the top layer

is always less than that in the coal in other layers of the

bed because pyrite has almost the same density as magnetite

and tends to be distributed uniformly within the magnetite .

. Multistaging is more effective than single stage

cleaning on sulfur segregation in coal from the bottom of the

bed to the top .

. For certain size fractions of coal, sulfur reduction

can always be achieved by fluidizing the coal with the

pertinent size fraction of magnetite .

• Al though a complete study of the continuous process

is beyond the scope of this experiment, the inclined bed

shows a promising way to clean the coal.

The experimental data and results developed in this

study support the final conclusion that an air fluidized bed

65

• r ,· i • ' : '!1 ~· I

is effective in segregating coal-minerals from coal when a

··bed mixture of coal and magnetite is used.

(

66

.. 0 a

l ,l•

• \ ~·:-:;,

•

REFERENCES

'

1. ''New per spec t iv e on co a 1 c 1 ea n in g '' E PR I Jou r n a 1 ,

November 1985,15.

2. "Coal cleaning test facility : 1985 plan", EPRI CS-4071,

Projects 1400-6, 1400-11 Interim Report, July 1985.

3. E.K. Levy, "Coal cleaning in a multistage air fluidized

bed", Research Proposal, January 4, 1985.

4. P. N. Rowe, A. w. Nienow and A. J. Agbim, "A preliminary

Quantitative study of particle segregation • in gas

fluidized beds -- binary systems of near spherical

particles", ~rans. Inst. Ch~m. En~s.,

P. 324.

Vol. 50, 1972,

5. P.N. Rowe and K.S. Sutherland, "Solids mixing studies i·n

gas fluidized beds part II: The behavior of deep beds

of dense materials", Trans. !!!st. Chem. ~ngrs., Vol. 4 2,

1964, T 55.

6. Rowe, Partridge, Cheney, Henwood and Lyall, "The

Mechanisms of Solids Mixing in Fluidized Beds", !ra!}s.

Instn. Chem. Eng., Vol. 43, T 271 (1965).

7. P. N. Rowe, A. w. Nienow and A. J. Agbim, "A preliminary

8.

quantitative study of particle segregation I

in gas

fluidized beds -- Binary systems of near spherical

particles", Trans. Instn. Chem. Engr., Vol. 50, 1972.

J. L. P. Chen and D. Keairns, "Particle segregation I in

a fluidized bed", Canadian J. Chemical Engineering, Vol.

53, August 1975, P. 395.

67

I

9. A. w. • Nienow, P. N. Rowd, L. Y. -L. Cheung, II A

quantitative analysis of the mixing of two segregating

powders of different density in a gas-fluidized bed",

Po~~er Tech_!!2_logy, 20, 1978, P. 89-97.

10. A. w. Nienow, P. N. Rowe and T. Chiba, "Mixing and

segregation of a small prpportion of large particles in

gas fluidized beds of considerably smaller ones", AICHE

§.y!!!Eosiu~ ~erie~ No. 176, Vol. 74, P. 45.

11. W. Weintraub, A Deurbrouck and Thomas, "Dry-cleaning

coal in a flui-dized bed medium", DOE Report RI-PMTC-4

(79) •

I

68

•

h

H

-V

j

·v • • J1

w

X bed

X top

• •

• •

• •

• •

• •

• •

• •

Nomenclature

The depth from the surface

Bed depth

Average jetsam volume percentage

Jetsam volume percentage in the i layer

Weight

Average Jetsam volume percentage

Jetsam volume percentage in the top layer

)

\.

\

•

69

----~-----.-.-··.,,..,.,,.;I~·-,·, .. '

,t

Biography

NAME: Yih-Tun Tseng

DATE OF BIRTH: Jan. 18, 1960

PLACE OF BIRTH: Hsin-Chu, Taiwan, R.O.C.

NAME OF FATHER: Chia-Sheng Tseng

NAME OF MOTHER: Hwan-Chin Sam

INSTITUTION ATTENDED: National Taiwan University

DEGREES : B.S. (Aug. 1, 1978)

'

..

...

4 I