Beneficiation of Taconite Ore RI 8572

7/28/2019 Beneficiation of Taconite Ore RI 8572

http://slidepdf.com/reader/full/beneficiation-of-taconite-ore-ri-8572 1/16

Report of Investigations 8572

Beneficiation of a Hematitic Taconite

by Reduction Roasting,

Magnetic Separation, and Flotation

By R. E. Peterson and A. F. Colombo

UNITED STATES DEPARTMENT OF THE INTERIOR

James G. Watt , Secretary

BUREAU OF MINES



This pub I cation has been cataloged as follows:

Peter on, R. E. (Roy Erne t), 1926-

Beneficiation of a hematitic taconite by reduction roasting,

magnetic separation, and flotation.

(Rep ort of investigations / United cares Dep t . of th e Interior, Bu •

reau of Mines ; 8572)

Inc ludes bibliographies .

l. Ore-dressing . 2. Taconite, J, Co lombo, A. F . (Arthur F . ). II.

Tide. III. Series: Report of investigations (United Scares . Bureau of

Mines) ; 85 72 .

TN23.U43 [TN538.17) 622s (622' .341) 81 -3883 AACR2



CONTENTS

Abstract •••••••

Introduction • • • • • • • • • • • • • • • • • • • • • • • • • •

Raw materials • • • • • • • • • • • • • • • • • • • • • •

Oxidized taconite •••••••

Lignite ........................•.........•..•.•...•..•.•..•.•..•..•.

Reduction roasting • • • • • • • • • • • • • • • • •

Equipment and procedures ••••••

Test results • • • • • • • • • • • • • • • • • • •

Beneficiation • • • • • • • • • • • • . • • • • • • • . ................................Equipment and procedures ••••

Beneficiation resu l ts •••••••

Summary and conclusions • • • • • • • • • • . .1.2.

3.

4.s.

1.

2.

3.4.s.6.

7.8.

9.

1o.11.

ILLUSTRATIONS

Western Mesabi Range sample areas • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •Reduction-roasting flowsheet • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

Temperature profi le of taconite in re la t ion to i t s posi t ion and

retention time in the ki ln • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

Beneficiation flowsheet • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

Size analyses of grinding-circuit products • • • • • • • • • • • • • • • • • • • • • • • • • •

TABLES

Size and i ron analysis of taconite sample • • • • • • • • • • • • • • • • • • • • • • • • • • •

Chemical analysis of taconite sample • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

Proximate analysis and heating value of l ign i te • • • • • • • • • • • • • • • • • • • • •Description of roasting equipment • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

Processing rates and dist r ibut ion of mater ials • • • • • • • • • • • • • • • • • • • • • •

Size analyses of cyclone dust , ki ln-c lass i f ie r overflow, and scrub-

ber sludge •..................•...............................••..•

Analyses of products from reduction roasting • • • • • • • • • • • • • • • • • • • • • • • •

Size and iron analysis of roasted taconite • • • • • • • • • • • • • • • • • • • • • • • • • •

Description of beneficiat ion equipment• • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

Grinding-mill power requirements • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

Analyses of products from beneficiat ion pi lo t plant • • • • • • • • • • • • • • • • •

1

1223446

99

12

13

25

710

11

3

3

448

8

9

99

12

13



BENEFICIATION OF A HEMATITIC TACONITE BY REDUCTION

ROASTING, MAGNETIC SEPARATION, AND FLOTATION

by

R. E. Peterson 1 and A. F. Colombo 2

ABSTRACT

The Bureau of Mines has undertaken beneficiat ion invest igat ions on sam

ples representing large tonnages of oxidized taconi tes in order to develop and

evaluate applicable technologies for producing iron oxide concentrates from

these materials. Information derived from the invest igat ions can be used to

expand th e domestic iron raw-materials base and to es tabl ish th e impact thesetaconite deposits may have on future national iron ore requirements.

This report describes a process that includes reduction roast ing, mag

netic separat ion, and f lo ta t ion, applied to a western Mesabi Range sample

taken near Coleraine, Minn. The sample contained 36.8 pet iron and 46.0 pet

s i l i ca , and represented 0.5 to 1 bi l l ion tons of crude mater ial . An average

of 962 lb /hr of oxidized taconite was roasted with 33 lb /hr of l igni te in a35-foot-long rotary ki ln . Pulverized l igni te and natural gas maintained the

kiln operating temperature a t 850° C. The roasted taconi te was beneficiatedby magnetic separation and cat ionic f lo ta t ion in a 900-lb/hr pi lo t plant and

yielded a product analyzing 68.1 pet iron and 4.6 pet s i l i ca , and containing

82.4 pet of th e iron in th e kiln feed.

INTRODUCTION

To provide technology that can assure th e U.S. iron and stee l industry acontinuing supply of domestic iron ore, the Bureau of Mines Twin Cities

Research Center has conducted research on processing domestic iron ore

resources. In Bureau invest igat ions , samples of large resources of i ronbearing material were beneficiated to evaluate the i r potent ial .

The deposits chosen for this invest igat ion are located on th e western

Mesabi Range in Minnesota, which contains an estimated 10 bil l ion tons3 of

1Metallurgist , Twin Cities Research Center, Bureau of Mines, Minneapolis, Minn.2Research supervisor, Twin Cities Research Center, Bureau of Mines, Minneapo

l i s , Minn . (now Research Director , Reno Research Center, Bureau of Mines ,Reno, Nev.).

3Throughout this report , tons refer to long tons (2,240 pounds).



2

GRANITE, GREENSTONE , QUART Z I T E

Scal e, miles0 I 2 3

0123 4 5

Scale, kilomelers

FIGURE 1. - Western Mesabi Range sample areas. Samples were taken from the Biwabik iron

formation (darkly shaded area), which is bounded by deposits of granite, green-

stone, and quartzite to the north, and Virginia argillite to the south.

oxidized taconite4 assaying 30 to 40 pet i ron. Samples were taken from three

sections of th is range. The sect ions, as shown on the map in f igure 1, were

ident if ied as WMR-1, WMR-2, AND WMR-3. Each sample was subjected to two

al ternat ive technologies: one that involves select ive f locculat ion and ca t i -onic f lotat ion,5 and the other a process tha t includes reduction roast ing,magnetic separat ion, and f lo ta t ion. The t es t objectives were to optimize

both technologies and to obtain operating data on concentrating the taconi tes

to a product containing less than 5 pet s i l i ca . Results have already been

reported on the select ive f locculat ion and cat ionic f lo ta t ion of the samples

from the WMR-1, and WMR-2 sect ions,6 and on the reduction roasting and bene

f ic ia t ion of the WMR-1 sample. 7 This current report summarizes the resul ts

of reduction roasting, magnetic separat ion, and f lo ta t ion tes ts on the WMR-2

sample, which represents an area west of the f i r s t block of oxidized taconite.

Results of tes ts on the WMR-3 sample wil l be given in subsequent Reports ofInvestigations.

RAW MATERIALS

Oxidized Taconite

A 600-ton sample of a 9-mile-long block of material located between Nash

wauk and Coleraine, Minn., was obtained and crushed to minus f ive-eights inch

40xidized taconite is synonymous with "nonmagnetic taconite" and refers to

taconite in which the iron oxides cannot be recovered by a drum-type sepa

ra tor having a maximum f ield in tens i ty of 3,000 gauss a t i t s surface.5Frommer, D. W., and A. F. Colombo (assigned to U.S. Department of the Inte-

r io r ) . Process for Improved Flotat ion Treatment of Iron Ores by Select iveFlocculation. U.S. Pat. 3,292,780, Dec. 20, 1966.6Jacobs, H. D., and A. F. Colombo. Select ive Flocculat ion and Flotat ion of a

Mes a bi Range Hematit ic-Goethit ic Taconite. BuMines RI 8482, 1980, 11 pp.

Cationic Flotat ion of a Hemat i t ic Oxidized Taconite. BuMines RI 8505,

1981, 11 pp.

7Peterson, R. E., and A. F. Colombo. Reduction Roasting and Benef ic iat ion of

a Hematit ic-Goethit ic Taconite. BuMines RI 8549, 1981.



3

(table 1) and then blended to make i t homogeneous. The portion taken for theroasting tes ts was crushed again to minus three-eighths inch to shorten the

time required to roast the part ic les to the center . An analysis of the sample

shows that i t contained 36.8 pet Fe, 0.7 pet Fe 2+, and 46.0 pet Si02 ( table 2).

Previous geologic studiesB indicated that th e or iginal minerals were magnetite,

chert, iron carbonate, and iron s i l i ca te . Weathering of this mineral assem

blage caused oxidation of the magnetite to hematite and some goethi te, and

oxidation of iron carbonates and s i l icates to goethite . Petrographic analysesof th e sample for this tes t indicated that i t consisted of hematite and

goethite in about equal amounts, quartz as the major gangue mineral, and magnet i te , iron carbonate, and iron s i l icates in trace amounts.

TABLE 1. - Size and iron analysis of taconite sample, percent

Size f raction, meshl Weight Iron Iron distr ibut ionPlus 3• • • • • • • • • • • • • • • • • • • • • • • 10.0 43.9 11.9

Minus 3 plus 4••••••••••••••• 9.5 43.9 11.3

Minus 4 plus 1o . . . . . . . . . . . . . 14.7 47.3 18.9

Minus 10 plus 20 • • . . . . • • . • • . • 14.6 47.9 19.0

Minus 20 plus 100• • • • • • • • • • • • 17.1 50.0 23.3

Minus 100 plus 200 . • . • • . . • • . • 8.2 24.2 5.4

Minus 200 plus 325 • • • • • • • • • • • 7.9 12.3 2.6Minus 325 plus 500 ••••••• •••• 4.5 9.6 1.2Minus 500•••••••••••••••••••• 13.5 17.3 6.41Standard Tyler sieve size.

TABLE 2. - Chemical analysis of taconite sample

Percent Percent

Components: Components--continued

Fe • • • • • • • • • • • • • • • •

Fe2+••••••••••••••Si02 • • • • • • • • • • • • • •

cao .••••. . . .• . .•• .Al 2o3•••••••••••••

MgO • • • • • • • • • • • • • • •

Ti0 2 • • • • • • • • • • • • • •

co 2 •••••••••••••••

LOI Loss on ignit ion.

36.8

.746.0

.8 1

.6 1

.27

.22

• 2

s . . . . . . . . . . . . . . . . . . . .

K20• • • • • • • • • • • • • • • • • •

c. . . . . . . . . . . . . . . . . . . .Mn • ••••••••••••••••••

Na2o . . . . . . . . . . . . . . . .p • • • • • • • • • • • • • • • • • • • •

LOI a t 400° C••••••••LOI a t 1,000° C••••••

Lignite

0.11

.10• 1• 1• 1.01

1. 92.3

North Dakota l igni te was used as the burner fuel and as th e reductant

added with th e material fed to the ki ln . Table 3 gives the proximate analyses

and heating values of the "as received" l igni te . A portion of the minus2-1/2-inch "as received" l igni te was screened at one-fourth inch, and the plus

1/4-inch fraction was used as the reductant. Lignite pulverized to 60 petminus 200 mesh was the burner fuel .

8Grunner, J . W. The Mineralogy and Geology of the Taconites and Iron Ores of

the Mesabi Range, Minnesota. Office of the Commissioner of the Iron RangeResources and Rehabilitation. St. Paul, Minn., 1946, 127 pp.



4

TABLE 3. - Proximate analysis and heating

value of l igni te

Moisture • • • • • • • • • • • • • • • • • • • • • pct ••

Volati le matter • • • • • • • • • • • • • • pct ••

Fixed carbon• • • • • • • • • • • • • • • • • pct ••

Ash• • • • • • • • • • • • • • • • • • • • • • • • • • pct ••

Heating value • • • • • • • • • • • • • Btu/lb ••

REDUCTION ROASTING

Equipment and Procedures

33.4

29.1

31. 1

6.4

7,260

The object ive of the reduction roasting operat ion was to convert th e goe

th i te (Fe 2o3 • H20) and hematite (Fe 20 3 ) to magnetite (Fe 3 0 4 ) . This wasaccomplished by heating th e iron oxides in the presence of a reductant to

remove one-ninth of the oxygen and the water combined in goethi te. The roast

in g step was conducted in a 35-foot-long rotary kiln having a 4.5-inch annular

dam a t the discharge end. The ki ln was l ined from the feed end to the mid

point with f ire-clay bricks and from the midpoint to the discharge or hot endwith 70-pct-alumina bricks. The flowsheet for the process is shown in f ig

ure 2 and the equipment is described in table 4.

TABLE 4. - Description of roasting equipment

Balling disk • • • • • • • • • • • • • • • •

Cyclone dust col lector ••••••Rotary kiln • • • • • • • • • • • • • • • • •

Wet scrubber • • • • • • • • • • • • • • • •

Stack fan •..................

Classif ier • • • . • • • • • • • • • • • • • •

5-foot diameter.

3-foot diameter.

34 inches by 35 feet .

1 foot 6 inches by 21 feet 9 inches.

20 hp.

2 by 11 fee t .

Oxidized-taconite and l igni te feed rates to the ki ln averaged 962 and

33 lb /hr , respect ively. The retent ion time of the taconite in the ki ln , ro ta

t ing a t 1.3 rpm, was 120 min. A coal burner 9 and a premix natural gas burner,both mounted in the discharge end of the ki ln , provided the process heat. Thegas burner preheated th e ki ln , igni ted and maintained igni t ion of the coal

burner, and regulated the kiln temperature. An automatic temperature control

system regulated the gas burner to hold the temperature near 800° C a t the

control thermocouple. This thermocouple was located 4.5 feet in from the

burner end of the kiln because the temperature was s tead ier a t this point than

i t was nearer to th e burner. At the thermocouple nearest the burner, the tem

perature averaged about 850° c.

9Nigro, J. C., and R. K. Zahl. Pulverized Coal-Firing Systems for InduratingIron Oxide Pel le ts . Ch. in Agglomeration '77. American Inst i tu te of Min

ing, Metallurgical , an d Petroleum Engineers, Inc . , New York, v. 1, 1977,

pp. 437-464.

Work ci ted in footnote 7.



4

I Feed

2

Analysis

Solids rate--- lb/hr--

I r on - - - - - - -pet--

Fue 1-- - MM Btu/hr--

NA Not available.

NA p Not applicable.

I

Ore

962

36.8

NAp

2

Lignite

33

NA

0.24

Exhaust

Scrubbermakeupwater

Exhaust

Rotary kiln

Stream No.

3 4 5 6 ?

Stream name

Recycle Excess Natural Pulverized Scrubberdust dust gas lignite slurry

251 15 NAp 102 10

22.3 22.3 NAp NA 29.0

NAp NAp 0.44 0.74 NAp

FIGURE 2. - Reduction-roasting flowsheet.

?

Recycle water

8

9

8 9 10

Scrubber Roasted Classifier

sludge ore slimes

10 839 70

29.0 40.6 11.4

NAp NAp NAp



6

Most of the entrained dust was removed from the kiln exhaust as i t passed

through a cyclone dust collector . (The dust accumulated in the cyclone hopper

is designated "cyclone dust" in this repor t . ) Additional par t icula tes and

water-soluble gases were removed from the cyclone exhaust by a we t scrubber.

The cyclone dust was pellet ized in a balling disk, using only water as th e

binder, and recycled to the kiln. Once these pellets entered the kiln bed,there was l i t t l e further loss of th e dust, even though the pel le ts d is in te-

grated before they were discharged.

Since the hot magnetite discharged from the kiln would oxidize to hema

t i t e i f exposed to a i r , the ki ln product was water-quenched in the pool of ascrew c lass i f ie r . A low-grade slime, the class i f ier overflow, was routed to

the ta i l ings pond. The c lass i f ie r sands, a t 18 pet moisture, were stockpiledin a covered concrete bin for subsequent beneficiat ion tes ts .

Test Results

A 5-day continuous-roasting campaign was conducted to provide roasted

taconite for beneficiat ion tes ts and operating data for subsequent reports .Data were averaged and samples were composited over an 86-hour period of

steady-state operation.

The energy equivalent of the l igni te for the burner and the reductant was0.98 MM Btu/hr. In addit ion, 0.44 MM Btu/hr was contributed by the natural

gas burner. The combined energy consumption was 1.42 MM Btu/hr, or 3.3 MMBtu/ton of taconite. In a commercial operation, using a larger kiln l inedwith insulating brick, the fuel consumption should be much less. Figure 3shows the profile of the temperatures in the ki ln , with respect to both theposit ion of th e taconite in the kiln and the corresponding retent ion time.

The time scale indicates the average length of time that the charge had been

in the kiln. For example, a t a point 25 feet into the ki ln , the charge hadbeen in for 85 min and reached a temperature of 580° C.

Part iculate entrained in the ki ln exhaust was removed by the cyclone dust

collector a t an average rate of 266 lb /hr . The pellet iz ing and recycling rateaveraged 251 lb /hr , leaving an average excess of 15 lb/hr of dust. The largecirculating dust load resul ted part ia l ly from the pel le t breakage on the

vibrating conveyor used to feed the ki ln .

The material balance in table 5 shows that 7.2 pet of the feed weight wasrejected from the kiln discharge product in the class i f ier overflow. This

overflow had a pH of 9.6, indicat ing dissolut ion of lime from the taconite and

basic oxides from the l igni te ash. Size analyses presented in table 6 showthat th e solids in this stream contained only 9.4 pet plus 325 mesh material .Since the fines in the feed were low grade, i t was advantageous to reject themfrom the roasted product in a desliming operation before they were mixed with

higher grade fines formed during comminution. As expected, the slimes removedby th e class i f ier were low in iron, 11.4 pet, and did not respond well to mag-netic separation. The magnetic f raction recovered from a Davis tube tes t



u0

DISTANCE FROM FEED END OF KILN, feet

0 5 10 15 20 25 30 359 0 0 r - - - - - r - - - - - , - - - - - - . - - - - ~ r - - - - - ~ - - - - - r - - - - ~ ~

800

700

600

400

300

200

100 ~ - - - - - - - - ~ - - - - - - ~ - - - - - - - - ~ - - - - - - - - ~ - - - - - - - - _ J0 25 50 75 100 125

TIME, min

FIGURE 3 .- Temperature profile of taconite in relation to it s position and

retention time in the kiln.

7



8

contained excessive s i l ica , 12.9 pet, indicating the the presence of the

slimes would increase the difficulty of making an acceptable grade concentratein th e beneficiation circui t .

TABLE 5. -Processing rates1 and distr ibut ion of materials

Material

Taconite ••••••••••••••

Ash from l ignite4 •••••Total • • • • • • • • • • • •

Roasted taconite ••••••

Classif ier overflow •••Scrubber sludge •••••••

Cyclone dust ••••••••••

LOI from ore ••••••••••

Oxygen loss fromreduction• • • • • • • • • • • •

Total • • • • • • • • • • • •

-'9629

971

83970101522

15971

Iron analysis ,pet

36.8

11.6

NAp

OUTPUT

40.6

11.4

29.0

22.3

NAp

NAp

NAp

LOI Loss on ignit ion in roasting.NAp Not applicable.

-'99.1 99.7

.9 .3100.0 100.0

86.4 96.0

7.2 2.3

1. 0 .81.5 .9

2.3 .o

1.6 .o100.0 100.0

1Natural gas consumption was 441 f t3 /hr . Combustion a ir ratesfor the gas and coal burners were 3,171 and 5,050 f t3 /hr ,respect ively.

M o i s t u r e ~ f r e e basis .3Weight obtained from difference between the to ta l product weight

and the calculated ash weight.4Consumptions of l igni te for reductant and fuel were 33 and

102 lb /hr , respectively.

TABLE 6. - Size analyses of cyclone dust, ki ln -c lass i f ie r overflow,

and scrubber sludge, pet

Size f raction, mesh1 Cyclone dust Kiln-classif ier overflow Scrubber sludge

Plus 100 • • • • • • • • • • • • • 20.4 o.o 0.3Minus 100 plus 200 ••• 15.8 .o .2

Minus 200 plus 325 ••• 13.0 9.4 .9

Minus 325 plus 500 ••• 17.6 32.6 .9Minus 500 • • • • • • • • • • • • 33.3 58.0 97.7

1standard Tyler sieve size.

The deslimed roasted taconite contained 40.6 pet Fe and 13.1 pet Fe 2+, asshown in the product analyses reported in table 7, and 96 pet of th e iron fed

to the kiln. Although the ferrous-to-total- i ron rat io (Fe 2+/Fe) was only 0.21

in the plus 3-mesh fract ion (table 8), the overal l rat io , 0.32, was near the

natural magnetite rat io of 0.33. An iron recovery of 94.6 pet was obtained inthe magnetic fract ion from a Davis · tube tes t and indicated a good conversion

of the ferric oxides to magnetite. This magnetic f raction contained 67.1 pet

iron and 6.7 pet s i l ica .



9

TABLE 7. -Analyses of products from reduction roasting, percent

Component Cyclone dust Classif ier slimes Scrubber sludge Roasted orel

Fe • • • • • • • • • • 22.3 11.4 29.0 40.6

Fe 2+•••••••• .6 2. 1 1. 2 13.1

Si0 2•••••••• 65.7 78.5 48.7 41.2

cao . . . . . . . . . .1 0 1. 0 .75 .7 8Al203 ••••••• .8 1. 1 3.2 • 61MgO • • • • • • • • • (.1 • 51 • 30 • 27s. . . . . . . . . . . .06 .10 .1 5 .0 4

K20••••••••• ( .1 ( .1 ( .1 ( .1c. . . . .••. .•• .5 .9 1. 8 .3Mn • • • • • • • • • • ( .1 ( .1 (.1 ( .1

Na 2o••..•••• • 1 • 19 • 28 ( .1

p• • • • • • • • • • • (.01 (.01 .0 3 .01

1Feed to beneficiation pi lo t plant.

TABLE 8. - Size and iron analysis of roasted taconite , percent

Size f raction, meshl Weight Iron Iron distr ibut ion Fe2+jFe

rat ioPlus 3• • • • • • • • • • • • • • • • • • 3.5 44.9 4.0 0.21

Minus 3 plus 4• •..•...•• 8.7 49.5 11.0 .27

Minus 4 plus 1o . . . . . . . . 14.9 51.7 19.8 .3 2Minus 10 plus 28• • . • • • • . 12.7 51.7 16.8 .3 5

Minus 28 plus 65 • • • • • • • • 16.2 53.2 22.1 • 37Minus 65 plus 100••••••• 5.5 47.4 6.7 .38

Minus 100 plus 200 •••••• 19.2 23.3 11.4 .38

Minus 200 plus 325 •••••• 6.3 15.1 2.4 • 38Minus 325 plus 500 •••••• 6.2 16.2 2.6 • 36Minus 500• • • • • • • • • • • • • • • 6.8 18.6 3.2 • 36

lstandard Tyler sieve size.

BENEFICIATION

Equipment and Procedures

The beneficiation flowsheet, which includes three stages of magnetic

separation and two stages of f lotat ion, i s shown in figure 4. A br ie f

description of the equipment is given in table 9.

TABLE 9. - Description of beneficiat ion equipment

Rodmill • • • • • • • • • • • • • • • • • • • •

Cobber magnetic separator ••

Ballmill •••••••••••••••••••

Rougher magnetic separator.

Hydrocyclone• • • • • • • • • • • • • • •

Finisher magnetic separator

Rougher f lotat ion cel ls ••••Scavenger f lotat ion cells ••

2- by 4-foot mill , 690-pound rod charge.

30-inch diameter by 18-inch wide concurrent drum.Maximum surface field strength of 1,400 gauss.

3- by 3-foot mil l , overflow type. Ball charge

was 2,000 pounds.


3-inch diameter.


Bank of 2 cel ls , 4.8 ft3 .Bank of 2 cel ls , 3.0 ft3 .



Romi

8

Sem No

I

2

3

4

5

6

?

8

9

/0

Ays

Sem nme

p

RmiC

C

BmiR

R

Co

Co

Fns

Fns

fe

tonc

aeds

gon

c

aeu

owo

owton

c

ae

Wegds10

35

65

34

36

38

29

59

33

56

F-

46

84

60

67

10

64

64

60

15

69

Fl+-

11

26

14

26

26

28

22

24

39

24

S02-

4.2

89

18

N

83

63

NA

13

71

80

Fds-10

78

92

55

.8

57

43

94

11

93

NNaae.

FG

4B

caoow

9

II

12

II

1

13

R

S

R

foaooao

foao

ton

c

aec

ae

52

07

51

45

53

61

19

11

21

N

27

I46

55

10

88

14

S foao

ton 45 4

313

40 45

1!4

15

15

Ta

tan

-

49

18 36 N 1

2

. 0



80

70

60uc.

ciw

50

<X1-wa::

1- 40r~w

30w~1-<X

_J20>~

=>u

10

STANDARD TYLER SIEVE SIZE, mesh

28 48 100 20 0 32 5 40 0

KEY• Rodmill discharge

o Cyclone underflow

• Ballmill discharge

0 Cyclone overflow

11

The roasted taconitewas fed to a rodmill a t an

average rate of 977 lb/hr( dry basis) with enough

water to produce a slurrycontaining 62 pet solids .

After being ground to ab out93 pet minus 48 mesh(fig . 5) , the slurry wasconcentrated with a cobber

magnetic separator . Thenonmagnetic product wasdewatered in a screw c lass i -

f ier t o yield a coarse t a i l ing1 0 and an overflow thatwas r outed to the water

reclamation system . Con centrate from the cobber

was ground in a ballmill to53 pet minus 400 mesh and

upgraded in a rougher mag netic separator . Therougher concentrate wasclass i f ied in a hydrocyclone.

o L _ - - - - ~ ~ ~ - - - L - - - - ~ ~ ~ ~ ~ ~ L - L - L - L - - L - - ~ The cyclone underflow was1,000 60 0 500 400 300 200 100 80 60 40 30 returned to the ballmill for

PARTICLE 01 AMETER, ,._m

FIGURE 5. - Size analyses of grinding-circuit products.

regrinding; th e overflow, a t

78 pet minus 400 mesh , waspassed through a f inishermagnetic separator . After

the f inisher concentrate was treated with 0.10 lb/ ton of cationic collectorArosurf MG-98A11 , a s i l ica - bearing f roth was removed in a rougher f lotat ionoperation performed a t a natural pulp pH of 8. 5. This rougher f roth wasrefloated in a scavenger step . The scavenger underflow was returned to th e

rougher f lota t io n step ; the scavenger f roth was combined with th e ta i l ings

products from the three magnetic separators in the water reclamation system .The rougher f lotat ion underflow , which was the f inal concentrate, was thickened and f i l tered .

The ta i l ings products, except the class i f ier sands, were flocculated with

0. 06 lb/ton of a cationic flocculant ( Superfloc 330) , and pumped to the water

reclamation system, where they were dewatered by two thickeners arranged inseries . The f i r s t thickener produced an underflow that was pumped to a t a i l -

ings pond; th e second yielded an overflow that was recycled to th e pi lo t plantas reclaimed water, and an underflow tha t was routed back to the f i r s t thickener feed well .

10 In this report , "tai l ing" is used to refer to the residue from one step inth e beneficiation process , and "tai l ings" to the combined residues fromth e various steps .

11 Reference to specif ic brand names is made for ident if icat ion only and does

not imply endorsement by the Bureau of Mines .



12

Beneficiation Results

The roasted taconite was processed during a 5-day around-the-clock tes t

in the beneficiation pi lo t plant. The data reported here are averages from a32-hour period of steady-state operation when the col lec tor addition was main

tained at 0.10 lb/ ton.

Of the 43 gal/min of water used by the beneficiat ion plant , 83 pet wasrecycled. Because magnetic flocculation promoted rapid set t l ing of suspended

fines, the reclaimed water was relat ively clear , having a turbidity of 38 ppms i l ica equivalents, as determined with a Hellige turbidimeter, and containing

19 ppm Ca 2+ and 4.2 ppm soluble s i l ica . The increase in calcium concentrationfrom the 12 ppm in the Minneapolis ci ty water, was at tr ibuted to calcium inthe ore and th e basic constituents of the l igni te ash.

The to tal power requirement for grinding was 18.9 kwhr/ton ( table 10).

Since 37.5 pet of the feed weight was discarded a t the cobber before ballmillgrinding, and th e grindabil i ty of the ore was improved by roasting, 12 the

power consumption was below that obtained in tes ts processing unroasted ore(25.7 kwhr/ton). 13

TABLE 10 . - Grinding-mill power requirements

Mill

Rodmill ••••••••••••••

Ballmill •••••••••••••

Total • • • • • • • • • • •

Kwhr/ton

4.6

14.3

18.9

Hp-hr/ton

6.2

19.3

25.5

The circulating load rate in the ballmill grinding circui t ( f ig . 4) was

high, about 400 pet. The high-density iron minerals were preferential ly segregated with the coarse part ic les, entered the underflow product of th e

cyclone, and caused the ballmill circulating load to have a high iron content.The presence of the i ron-r ich circulating load increased the iron content of

th e rougher feed, causing the unusual circumstance of having the rougher con

centrate higher grade than the f inisher concentrate. A ta i l ing containing

10.0 pet iron was obtained from the rougher magnetic separator.

The f inisher concentrate contained 8.0 pet s i l i ca . In the f lotat ion c i r -

cuit , substantial f roth product was removed to at ta in a concentrate with lessthan 5 pet s i l ica . This froth contained 40.3 pet iron and 4.5 pet of the ironin th e rodmill feed. The iron recovery for th e beneficiat ion plant was

85.8 pet. Adjusting for the 4 pet loss in th e roasting system, the overal lrecovery was 82.4 pet. As shown in table 11, the f inal concentrate contained

68.1 pet iron and 4.6 pet s i l ica .

12Vik, R. A., C. B. Daellenbach, and W. M. Mahan. Influence of Thermal Treatment Upon Grindabil i ty of Low-Grade Nonmagnetic Iron Ores. BuMinesRI 7567, 1971, 30 pp .

13second work cited in footnote 6.



13

TABLE 11. - Analyses of products from beneficiat ion pi lot plant , percent

Magnetic separation Flotat ion

Component Tail ing Scavenger Rougher

Cobber Rougher Finisher ta i l ing concentrate

Fe •• • • • • • • • • • • • • 8.4 10.0 13.5 40.3 68.1

Fe 2+• • • • • • • • • • • •

2.6 2.6 3.9 13.3 22.1Si0 2 • • • ••••••••• 83.9 83.3 77.1 42.0 4.6CaO • • • • • • • • • • • • • • 60 .47 .66 .73 .56

Al 203••••••••••• .8 .7 1. 0 1. 0 .sMgO • • • • • • • • • • • • • .18 .14 .19 .21 .11

s . . . . . . . . . . . . . . . .06 .OS .OS .OS .03K20 • • • • • • • • • • • • • <.1 <.1 <.1 (.1 (.1

c. . . . . . . . . . . . . . . 2.0 .9 .7 .3 .2Mn • • • • • • • • • • • • • • • 04 .06 .06 .11 .09Na 2o•••••••••••• • 15 .18 .38 .43 .29

p • • • • • • • • • • • • • • • <.01 <.01 .02 .03 .04

SUMMARY ANDCONCLUSIONS

Pilot-plant roasting tes ts demonstrated that the oxidized taconi te sam

ple, representative of 0.5 to 1 bi l l ion tons of iron resource, could be con

verted to th e magnetic form by reduction roasting in a rotary ki ln . A Davis

tube separation of the roasted taconi te resul ted in a recovery of 94.6 pet of

the iron in th e magnetic fract ion. Since the Fe2+/Fe rat io in the par t ic les

coarser than 3 mesh was signif icantly lower than the rat io in the f iner par

t ic les , the taconite should have been crushed to minus one-fourth inch before

roasting.

I t was demonstrated that desliming the kiln product was an effect ivebeneficiation step. The slimes f ract ion , containing 11.4 pet iron as compared

with 11.8 pet iron in the combined beneficiat ion plant ta i l ing, const i tuted aweight reject ion of 7.2 pet of the feed weight.

An acceptable grade concentrate was obtained from the deslimed product by

two addi t ional benef iciat ion s teps: magnetic separation and cat ionic f lo ta-

t ion. By magnetically beneficiat ing the roasted taconi te before f lo ta t ion, alarge part of the gangue was eliminated without being f inely ground or con

suming reagents. Because grinding and reagent costs are major expenditures in

an iron ore f lotat ion process, advantages gained in these areas would be a s ig -

nif icant part of th e just i f ica t ion for beneficiat ing by roasting and magnetic

separation before f lotat ion. Through f lo ta t ion, the s i l ica content in the

concentrate from the f inisher magnetic separator was lowered from 8.0 to

4.6 pet. This f inal concentrate analyzed 68.1 pet iron and contained 82.4 petof th e iron in the kiln feed.

GOVERNMENT PRINTING OFFICE : 1981 -703-002/77 INT. · BU .OF MINES,PGH. ,PA. 25566

Beneficiation of Taconite Ore RI 8572

Documents