Nothing Special   »   [go: up one dir, main page]

The Effect of Grinding Conditions On The Flotation of A Sulphide Copper Ore

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

Minerals Engineering 16 (2003) 1213–1216

This article is also available online at:


www.elsevier.com/locate/mineng
Technical Note

The effect of grinding conditions on the


flotation of a sulphide copper ore
K.L.C. Goncßalves a, V.L.L. Andrade a, A.E.C. Peres b,*

a
CVRD, Diretoria de Desenvolvimento de Projetos Minerais, Br 262 km 296, Santa Luzia, MG 33030-970, Brazil
b
UFMG, Department of Metallurgical and Material Engineering, Rua Espirito Santo, 35/206, Belo Horizonte, MG 30160-030, Brazil
Received 10 April 2003; accepted 23 May 2003

Abstract
The challenge of process development for the beneficiation of Salobo’s copper ore started in 1978 and the studies still go on.
Copper is predominantly present as secondary minerals, such as chalcocite, bornite, and digenite, and liberation requires very fine
grinding. These minerals may undergo rapid oxidation at the alkaline pH range and under mildly oxidising conditions. The oxi-
dation products may adsorb onto the minerals altering their surface characteristics, flotation behaviour, and may also significantly
modify the mechanisms of interaction between the minerals and the collector. These facts impair the flotation process performance
and increase the reagents consumption, the required flotation cells volume, and the overall processing costs. This paper describes the
effect of grinding conditions on the flotation performance. Different media and mill construction materials were tested at bench scale
aiming at evaluating the effects of the pulp electrochemical potential and the availability of iron oxide and hydroxide compounds on
the flotation response. The results indicated that the grinding process affects significantly the flotation metallurgical performance of
Salobo’s ore. The conditions that yielded the highest levels of copper recovery and the fastest flotation kinetics were rubber lined
steel mill and stainless steel media.
 2003 Elsevier Ltd. All rights reserved.

Keywords: Froth flotation; Grinding; Non-ferrous metallic ores; Sulphide ores

1. Introduction cathodes, while the iron grinding media act as anode,


being galvanically oxidised, while oxygen reduction oc-
Grinding precedes flotation in most concentrators curs on the sulphide mineral surface.
(outstanding exceptions are some iron ore flotation The oxygen consumption due to grinding media re-
plants in Brazil). The understanding of fundamental sults in a reducing environment that affects the sulphide
aspects of the two operations is crucial for improving and prevents xanthate oxidation and adsorption. At the
the concentrator performance. same time the sulphide particles surfaces are coated with
The floatability of ores and their separation selectiv- oxides layers. The sulphide minerals ground under these
ity are essentially determined by the surface properties. conditions do not present self-induced floatability and
The surface properties of sulphide minerals are mostly their flotation response in the presence of xanthate is
controlled by the grinding processes and conditions usually poor (Heyes and Trahar, 1979).
(Xiang and Yen, 1998). When two sulphides presenting large differences be-
During grinding, a galvanic contact occurs among the tween their rest potentials are in contact or are ground
sulphide minerals themselves and also among the sul- in a porcelain mill, the sulphide with lower rest potential
phide minerals and the grinding media, resulting in a will act as anode while the other will act as cathode. The
galvanic current due to rest potential differences. The anodic mineral will exhibit enhanced flotation perfor-
rest potentials for sulphide minerals are much higher mance due to the fact that a surface under oxidised state
than that for iron (Rao et al., 1976), so the former act as favours xanthate oxidation and adsorption. On the
other hand, oxygen reduction occurs on the most noble
*
Corresponding author. Tel.: +55-31-3238-1717; fax: +55-31-3238-
sulphide (usually pyrite). Its surface presents, then, very
1815. low affinity towards xanthate due to its reduced condi-
E-mail address: aecperes@demet.ufmg.br (A.E.C. Peres). tion (Xiang and Yen, 1998). This behaviour explains the
0892-6875/$ - see front matter  2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.mineng.2003.05.006
1214 K.L.C. Goncßalves et al. / Minerals Engineering 16 (2003) 1213–1216

enhanced flotation selectivity in the separation between and sodium dithiophosphate (25 g/t) were utilised as
chalcopyrite, galena, and sphalerite from pyrite, when collectors and a polyglycol alcohol (60 g/t) was em-
the ore is ground in porcelain or stainless steel mills as ployed as frother.
compared with iron or carbon steel mills. Flotation kinetics was evaluated by means of froth
Salobo’s deposit, located in the Caraj as area, is the collection at the following time conditions: 1.5, 3, 5, 10,
largest Brazilian copper reserve (geological reserves of 15 and 20 min. Froths collected in the first three flota-
approximately 1 billion tonnes of ore, copper grade tion stages were designated as rougher #1 and the other
0.86%, and open pit mineable reserve of 255 million three froths constituted rougher #2, #3, and #4. Re-
tonnes of ore averaging 1.11% Cu). Results of extensive agents were dosed in all four steps, prior to each flota-
on site pilot plant scale testwork, performed in the 80s, tion stage, 40% of each reagent being dosed prior to
were reported by Pereira et al. (1991). This stage in- rougher #1 and 20% prior to each other stage, namely
cluded a demonstration run lasting 210 h and treating rougher #2, #3, and #4.
380 tonnes of ore. The major conclusions were that A combined platinum electrode was utilised for rest
Salobo’s copper ore is hard to treat mainly due to its potential determinations. An increment of 197 mV was
high Bond work index, its liberation in a fine size range, added to the figure read in the equipment in order to
its complex mineralogy and uncommon mineralogical convert it to the normal hydrogen electrode standard.
associations, and its need of high flotation retention Slurry potentials were recorded after grinding and also
times and high collector consumption. The importance after each reagents addition stage and rougher flotation
of controlling the electrochemical conditions of the pulp stage.
was stressed in the report, despite the fact that this
control was not performed due to technical difficulties
related to time and geographical constraints. 3. Results
CVRD decided to concentrate the efforts on process
development of smaller deposits presenting higher grade The chemical analysis showed 1.35% Cu, 0.55% S,
and easier concentration conditions. Serra do Sossego and 0.99 g/t Au. The copper sulphides present in the ore
project start up is predicted for 2004. In the mean are bornite (4%), chalcopyrite, covellite, and chalcocite/
time, Salobo’s ore should be submitted to further digenite (1% each). Pyrite content is 0.5%. The pre-
investigation, especially concerning the correlation dominance of bornite is confirmed by the Cu/S ratio
between grinding conditions and flotation performance approximately 2.5. The sulphide particles are present in
(Goncßalves, 2002). the fine size range. Associations with magnetite and with
The effects of sodium sulphide additions and of using silicates, as inclusions, are common.
nitrogen as gas phase in flotation will be presented in Fig. 1 illustrates the size distribution of the sample
other publications. after grinding under different conditions. The coinci-
dence of the curves was achieved by using different
grinding times for each grinding condition:
2. Materials and methods
(i) rubber lined jar mill, stainless steel rods: 16 min;
The ROM copper ore sample was crushed in a roll (ii) rubber lined jar mill, carbon steel balls: 15 min;
crusher, in closed circuit with screening, to produce (iii) unlined jar mill, carbon steel rods: 09 min;
100% passing 1 mm with minimal ultrafines production. (iv) ceramic mill, ceramic balls: 29 min.
This sample was homogenised and then cone and
quartered to produce 1100 g fractions for the flotation
experiments.
100
Grinding was performed in the absence of reagents, lined; stainless steel rods
90
at 60% solids: unlined; carbon steel rods
cumulative percentage passing

80
ceramic; ceramic balls
70
(i) rubber lined jar mill, stainless steel rods; 60
lined; carbon steel balls

(ii) rubber lined jar mill, carbon steel balls; 50


(iii) unlined jar mill, carbon steel rods; 40
(iv) ceramic mill, ceramic balls. 30

20
The grinding time was determined based on the re- 10
quirement of achieving 90% < 100# (150 lm). 0

Rougher flotation experiments were performed with 1 10 100 1000


particle size (µm)
1100 g ore samples, at 35% solids, in a 2.5 L Denver
laboratory machine. Potassium amyl xanthate (100 g/t) Fig. 1. Size distributions after grinding under different conditions.
K.L.C. Goncßalves et al. / Minerals Engineering 16 (2003) 1213–1216 1215

300 100

90
200
80

70

copper recovery %
100
60
Eh (mV)

0 50

40
lined; stainless steel rods lined; stainless steel rods
-100 30
unlined; carbon steel rods ceramic; ceramic balls
ceramic; ceramic balls 20
lined; carbon steel balls
-200
lined; carbon steel balls 10 unlined; carbon steel rods
0
-300
0 5 10 15 20 25 30 35 40 45
1 2 3 4 5 6 7 8 9
stages copper grade (%)

1 - after grinding 4 - after reagents 7 - after rougher 3


Fig. 3. Copper recovery as a function of copper grade for different
2 - after reagents 5 - after rougher 2 8 - after reagents
grinding conditions.
3 - after rougher 1 6 - after reagents 9 - after rougher 4

Fig. 2. Slurry potential variation along rougher flotation of the copper


(bornite, chalcocite/digenite, and chalcopyrite). A layer
ore.
of metal sulphide deficient in the metal is formed at the
surface of the minerals, favouring the adsorption of
The condition of eliminating the flotation feed size xanthate oxidised species and also enhancing the self-
distribution as a variable was attained, as illustrated in induced floatability of the copper minerals.
Fig. 1. Grinding in lined steel mill (carbon steel balls) and
Fig. 2 illustrates the slurry potential variation along unlined steel mill (carbon steel rods) yielded negative
rougher flotation, after grinding under different condi- values of the pulp rest potential, approximately )50 and
tions. Table 1 and Fig. 3 show copper flotation recovery )150 mV, respectively, the reducing environment being
for grinding under different conditions. caused by the availability of ferrous ions in the system.
The co-existence of ferrous ions and sulphide minerals in
the system causes a galvanic current that induces the
4. Discussion oxidation of species presenting lower redox potential
(ferrous ions) and oxygen reduction on the surface of the
The similarity among the curves plotted in Fig. 1 sulphide minerals, resulting in a strongly reducing en-
shows that the different grinding conditions utilised did vironment. Iron oxides and hydroxides tend to precipi-
not produce samples with significantly different size tate on the surface of the sulphide particles. Sulphide
distributions. minerals ground under this condition do not present
The results of rest potential variation along rougher self-induced floatability and their flotation response with
flotation are discussed at the light of Fig. 2. xanthate is normally poor.
Grinding in ceramic mill (ceramic balls) and lined The high copper recovery in the first minute of flo-
steel mill (stainless steel rods) yielded positive values of tation after grinding in a lined mill and stainless steel
the pulp rest potential, approximately +200 mV, the grinding media, shown in Table 1, confirms the hy-
oxidising environment being caused dissolved oxygen. pothesis that the existence of an oxidising environment
This condition is ideal to promote the oxidation of during grinding enhances the floatability of copper
xanthate to dixanthogen as well as to oxidise moderately minerals. Grinding conditions that generated a reducing
the surface of copper sulphides present in the slurry environment yielded low copper recoveries in the first

Table 1
Copper flotation recovery after grinding under different conditions
Time (min) Lined-stainless steel rods Ceramic–ceramic balls Lined-carbon steel balls Unlined-carbon steel rods
Cu (%) Cu recovery Cu (%) Cu recovery Cu (%) Cu recovery Cu (%) Cu recovery
(%) (%) (%) (%)
1.5 26.4 58.8 34.4 41.0 34.5 36.5 38.4 21.1
3 18.4 73.7 21.9 64.0 26.0 52.5 26.3 42.1
5 15.4 79.9 17.6 73.0 21.5 61.8 20.9 53.2
10 10.0 89.5 10.5 87.5 9.6 87.7 8.8 86.8
15 8.5 92.1 8.6 91.0 8.0 91.8 7.2 91.9
20 7.6 93.5 7.8 92.5 7.1 93.8 6.5 93.7
1216 K.L.C. Goncßalves et al. / Minerals Engineering 16 (2003) 1213–1216

minute of flotation. A gradual decrease in the float- quate for flotation. Very high collector dosages are
ability of the sulphide minerals is observed for pulp rest necessary to provide the adhesion of the oxidised copper
potentials reaching less positive and then negative val- sulphide particles to air bubbles. Nevertheless, improved
ues. More positive potential values represent a larger flotation selectivity is achieved only for grinding condi-
availability of iron ions in the system, and consequently tions providing an oxidising environment.
a larger amount of oxidised iron species (oxides and
hydroxides) are present on the surface of the sulphide
minerals. 5. Conclusions
The expectation that grinding in a ceramic mill would
yield flotation results similar to those produced by The grinding conditions affect significantly the sub-
grinding in a lined mill (with stainless steel rods) was not sequent flotation stage of this sulphide copper ore. The
confirmed, despite the fact that the pulp rest potential presence of iron ions in the slurry is deleterious to the
was practically identical for both conditions (see Fig. 1). flotation of copper minerals and may be avoided by
The impaired flotation performance of the ore ground in the utilisation of lined mills and non-ferrous or corro-
the ceramic mill may be attributed to the morphology of sion resistant grinding media, such as stainless steel or
the particles, submitted to longer abrasion action (29 pebbles.
min in the ceramic mill · 16 min in the lined mill with Monitoring the slurry rest potential provides an in-
high chromium rods). Another explanation could be the dication of the flotation performance. Mild oxidising
precipitation of oxidised copper and iron species on the potentials (positive rest potentials) are adequate for
sulphides surfaces. The extension of this phenomenon enhanced copper recoveries for favouring xanthate oxi-
was larger for grinding in the ceramic mill due to the dation and adsorption onto the minerals surface.
longer residence time.
Data presented in Table 1 and Fig. 3 show that the
highest selectivity (higher grade for the same copper References
recovery) was achieved for grinding in a lined mill with
stainless steel rods, following ceramic mill with ceramic Goncßalves, K.L.C., 2002. Effect of surface oxidation on the flotation of
Salobo’s copper and gold ore, M.Sc. thesis, CPGEM-UFMG,
balls, then lined mill with carbon steel balls, and finally p. 138 (in Portuguese).
unlined mill with carbon steel rods. The direct contact Heyes, G.W., Trahar, W.J., 1979. Oxidation–reduction effects in the
between the carbon steel rods and the mill shell causes flotation of chalcocite and cuprite. International Journal of
enhanced wear of the shell and grinding media, liber- Mineral Processing 6, 229–252.
Pereira, C.E., Peres, A.E.C., Bandeira, R.L., 1991. Salobo copper ore
ating ultrafine iron particles into the pulp. The rate of
process development. In: Proceedings of COPPER 91, Ottawa,
the oxidation of iron particles reaction is accelerated, Canada, pp. 133–144.
with consequent precipitation of larger amounts of iron Rao, S.R., Moon, K.S., Leja, J., 1976. Effect of grinding media on the
oxides and hydroxides on the surface of the sulphide surface reactions and flotation of heavy metal sulphides. In:
minerals, explaining the impaired selectivity and the Fuerstenau, M.C. (Ed.), Flotation A.M. Gaudin Memorial Vol-
slower flotation rate up to the 10th minute of the test. ume, vol. 1. AIME, New York, pp. 509–527.
Xiang, H.W., Yen, X., 1998. The effect of grinding media and
The high dosage of reagents necessary for the flota- environment on the surface properties and flotation behaviour
tion of Salobo’s ore is responsible for the high final of sulfide minerals. International Journal of Mineral Processing 7,
copper recovery even for grinding conditions less ade- 49–79.

You might also like