Knelson's concentrator: extraction evaluation

Vladimir V. Knauf, NATI Jsc, Saint-Petersburg

 

Introduction

Knelson's concentrators are widely used in industry for the extraction of precious minerals (PM) from different ores, river sands and tailing-dump of the integral mine-metallurgical plants. But the results, which are obtained for different raw materials or even on same type of raw materials (extraction percentage of precious elements (PE) and PE-quality of concentrate in terms of the extracted component) can differ considerably.

To elucidate which ore parameters influence the results of enrichment and to be able to predict from what kind of ores there will be maximum extraction, the mineralogical investigations were done. Both the minor-sulphide Cu-Ni ore (the ore feeding for Knelson`s concentrator) and the tailings after the concentrate extraction were studied.

 

Mineralogical study

Using the Knelson's concentrator from the primary ore (40kg) with a grain-size class of 0.5mm, 80g of concentrate with the reduction coefficient equal to 530 were obtained. The total content of Pt, Pd and Au in the primary ore was about 5 ppm.

The quarters of the primary ore and the tailings after the concentrate's extraction were investigated using the "ppm- mineralogy" technology: primary ore and tailings were separated by granulometric fractions (GF) of 250 - 90, 90 - 37 and -37µm. (the fraction GF of 500-250µm was not studied, because when the concentration is 5 ppm, the possibility of the appearance of noble metal grains in this class of grain-size equals to zero practically, see "Home->Technics" section). Using hydroseparator NATI with the reduction coefficient of 18000 - 72000 for primary ore and 8000 - 13000 for tailings in each of 6 GF available the heavy concentrates (HC) were extracted. The weights of GF and HC of the primary ore and tailings are shown in the table 1.

Table 1
Sample
250-90µm
90-37µm
-37µm
Primary ore
GF=244г
HC=13мг
GF=146г
HC=8мг
GF=72г
HC=1мг
Tailings
GF=56г
HC=7мг
GF=37г
HC=6мг
GF=13г
HC=0.4мг

Microprobe samples GF were prepared from the heavy concentrates. Using SEM ABT-55 with ED spectrometer the minerals of noble metals were identified and their volumetric relations were measured.

Mineralogical investigation has shown that there are the same minerals of noble metals both in the primary ore and in the tailings: tetraferroplatinum, sperrylite, moncheite, paolovite, atokite, taimyrite, sobolevskite, kotulskite, mertieite, stillwaterite, guanglinite, palladoarsenide, palarstanide, arsenopalladinite, Pd-electrum, and unnamed phases (Pt,Pd)2Fe, (Pd,Pt)5(Sn,As,Sb,Te)2. The two latter and theirs analogous were mentioned by different authors, but as we know, these minerals have not been registered yet as the new ones.

The relationships between main minerals in the sample are as follows (volume %): tetraferroplatinum - 6.6, phase (Pt,Pd)2(Fe,Cu,Ni) - 1.0, sperrylite - 7.7, moncheite - 0.1, paolovite - 11.5, atokite - 37.5, taimyrite - 0.8, sobolevskite - 1.3, kotulskite - 0.3, phase (Pt,Pd)5(Sn,As,Sb,Te)2 - 14.4, mertieite I + palladoarsenide + Pd5(As,Sn)2 - 14.2, stillwaterite + guanglinite - 0.7, (Au,Ag,Pd) - 3.9.

Besides the diagnostics of minerals under the special programs of image processing the areas of all grains of minerals of noble metals were measured, and their linear sizes and volume (cubic approximation) were calculated. It is important to emphasize, that the distribution of total volume of minerals of noble metals by granulometric classes is directly proportional to distribution of weights (and concentration) of noble metals, because for the transition from volume to weights it is enough to take into account only the density of minerals. However, for the sufficiently exact estimation of a degree of extraction, it is quite justified to operate with volumes without transition to weights of elements, as soon as, in this case, for the transition to weights it is necessary to multiply the obtained volumes on some constant multiplier, identical both for primary ore, and for tailings, because the PM proportions in primary ore and in tailings are the same. (See in more details in section: " Home - > Technics - > Metrological... ").

Intervals of the linear sizes with a step of 10µm, number of grains in an interval (Ngr), total volume of grains in an interval (V mm3), the percent of grains in the given interval (%N) and percent of volume of grains in the given intervals (%V) for primary ore and tailings are given in table 2.

Table 2
Primary ore                                                                                                                 Tailings
Interval (µm)
N gr.
V µm3
% N
% V
 
Interval (µm)
N gr.
V µm3
% N
% V
Total:
656
64796168.9
 
 
 
Total:
412
14784397.1
 
 
0 - 10
118
35352.2
18.0
0.1
 
0 - 10
93
107069.1
22.6
0.7
10 - 20
185
717626.5
28.2
1.1
 
10 - 20
216
2060190.4
52.4
13.9
20 - 30
117
1700286.7
17.8
2.6
 
20 - 30
73
2884875.9
17.7
19.5
30 - 40
71
3042151.2
10.8
4.7
 
30 - 40
17
2629592.0
4.1
17.8
40 - 50
65
5825172.5
9.9
9.0
 
40 - 50
8
3364764.0
1.9
22.8
50 - 60
44
7477958.4
6.7
11.5
 
50 - 60
4
2774605.8
1.0
18.8
60 - 70
21
5370155.1
3.2
8.3
 
60 - 70
1
963299.9
0.2
6.5
70 - 80
20
8300378.8
3.0
12.8
 
70 - 80
0
0.0
0.0
0.0
80 - 90
4
2393110.8
0.6
3.7
 
80 - 90
0
0.0
0.0
0.0
90 - 100
3
2493491.3
0.5
3.8
 
90 - 100
0
0.0
0.0
0.0
100 - 110
0
0.0
0.0
0.0
 
100 - 110
0
0.0
0.0
0.0
110 - 120
1
1528073.9
0.2
2.4
 
110 - 120
0
0.0
0.0
0.0
120 - 130
1
1792656.6
0.2
2.8
 
120 - 130
0
0.0
0.0
0.0
130 - 140
1
2341752.7
0.2
3.6
 
130 - 140
0
0.0
0.0
0.0
140 - 150
1
3290999.9
0.2
5.1
 
140 - 150
0
0.0
0.0
0.0
150 - 160
1
3481118.5
0.2
5.4
 
150 - 160
0
0.0
0.0
0.0
160 - 170
1
4501904.8
0.2
6.9
 
160 - 170
0
0.0
0.0
0.0
170 - 180
2
10503979.1
0.3
16.2
 
170 - 180
0
0.0
0.0
0.0
180 - 190
0
0.0
0.0
0.0
 
180 - 190
0
0.0
0.0
0.0
190 - 200
0
0.0
0.0
0.0
 
190 - 200
0
0.0
0.0
0.0

Histograms 1.1 and 1.2 show the data concerning the primary ore and tailings after extraction of the concentrate. The analyses of histogram 1.1 shows, that the Knelson's concentrator extracts all grains with the size more than 70µm. In an interval 70 - 30µm the share of the extracted grains decreases. The grains with a size less than 30µm are completely lost practically. (The amount of grains less than 30µm in tailings is higher than in primary rock that is impossible actually. This contradiction is explained by losses of the thin fraction of tailings during sampling in and after enrichment. This is the cause of disagreement in the values of the percentage of extraction equal to 4%, which is shown on histogram 1.2. The amount of weight losses for a thin fraction is easy to estimate on the base of data of table 1.

Histograms 1.1 and 1.2

The share of extraction of minerals in each granulometric fraction is shown at the histogram 1.2, the total extraction being 77.2%. So high percent of extraction is explained by properties of ore: the significant part of noble metals (more than 80 %) is contained in minerals, which size is more than 50µm, in spite of the fact that the total amount of grains with the size > 50µm is only 15.5 % (see table 2).

If to consider the histogram of primary ore (PO) as PM-ore passport, and the histogram of tailings (TK) as PM-tailings passport of Knelson`s concentrator looses with a reduction coefficient of 530 (see1.1), it is easy to define a total share of extracted minerals of noble metals from the given ore taking into account the given parameters of enrichment process. It makes it possible to predict percentage of noble metals extraction from other kinds of ores without using labour-consuming researches.

At histograms 2.1 and 2.2 (analogues of 1.1 and 1.2) the data for one of the types of minor-sulphide Cu-Ni ores are given. On the contrary to the previous case, it is characteristic for primary ore, where all of the noble metals are contained in fine grains, which size does not exceed 50µm. That means that the previous and yet mentioned ores considerably differ from each other. If to compare the passport of examined ore to the passport of Knelson`s concentrator losses, it is easy to predict, that the total extraction of noble metals will be about 23 % (see 2.2) at the similar parameters of process (see 2.1).

Histograms 2.1 and 2.2

 

Discussion

The obtained data allow making the following conclusions:

  1. The mineralogical investigations which have been carried out with primary ore and tailings after concentrate's extraction from the concentrator, have demonstrated, that the rather large grains of minerals of noble metals are effectively extracted; the grains more than 70µm size are extracted completely, the grains with a size of 70-30µm are extracted partially and the grains with a size less than 30µm are in the field of 100 % of losses (taking into account the error of sampling).
  2. The distribution of minerals of noble metals by the size of grains can be considered as PM-ore passport of the given ore, which contains all necessary information for the choice of technology of industrial enrichment (extraction) of noble metals from ores.
  3. The reception of the PM-passport of tailings after extraction of a concentrate allows predicting the percentage of extraction of noble metals from ores, which PM-passport is determined. The reliable prediction of extraction of minerals of noble metals is a very important element of an economic estimation of the quality of complex Cu-Ni ores, in spite of the fact that it requires the realization of volumetric mineralogical mapping of a deposit (the ore body) for the reception of the PM-passports for all types of ores.
  4. The reception of the PM-passports of various ores can appreciably influence a choice of technology of industrial enrichment (extraction) of noble metals. For example, for the enrichment of Cu-Ni ores the flotation technique is widely used. It requires thin crushing of primary ore up to -74µm or - 44µm. It is easy to understand while analyzing the PM-passport of ores given on histogram 1.1 that during crushing ore up to - 44µm the significant amount of grains containing 82,5% of noble metals, will be transferred into thin granulometric fractions, in which percentage of extraction on the Knelson`s concentrator is extremely low: about 15 %! It means, that the choice of technology of complex extraction of metals from ores (Cu, Ni, PE) should provide extraction of PM up to a stage of thin crushing (milling) of primary ore, and the way of primary crushing should not result in the material overgrinding (see section " Home- > Technics ").
  5. It is also important to pay one's attention to economic aspect: hardly there will be required the serious calculations to prove, that the production costs for the reception of the PM-passports of ores, are much lower, than alternative way of reception of the similar information - realization of experiments using the industrial equipment with industrial volumes of primary ores. Besides having received the information by the second way, there remains not clear for what reason just such percentage of extraction PE is obtained and whether it is possible to increase it.
  6. PM passportization of ores in combination with a definition of PE contents in ores also gives the answer on the question, which is insoluble by other ways: what is the relationship between the PE located in ores as PM, and, PE located as an isomorphous impurity in the lattices of other minerals dissipated form. As the technologies of extraction for the specified forms of a PE in primary ores essentially differ, then the definition of the ratio PM:PE appears to be a very important parameter for the development of industrial technologies of enrichment of ores and for the production cost analysis (valuation of the outlay) for PE production. The fact, that in some kinds of Cu-Ni ores the dissipated Pd reaches up to 90 %, only emphasizes the importance of definition of PM-ore passports and PM:PE parameters of ores.