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

Minerals Engineering 24 (2011) 98–101

Contents lists available at ScienceDirect

Minerals Engineering
journal homepage: www.elsevier.com/locate/mineng

Higher kaolin recovery with a water-injection cyclone


V. Golyk a,*, S. Huber b, M.G. Farghaly c, G. Prölss c, E. Endres c, T. Neesse c, M.A. Hararah d
a
AKW A+V GmbH, Dienhof 26, 92242 Hirschau, Germany
b
Amberger Kaolinwerke Eduard Kick GmbH & Co. KG, Georg-Schiffer-Str. 70, 92242 Hirschau, Germany
c
Department of Environmental Process Engineering and Recycling, University Erlangen-Nuremberg, Paul-Gordan-Str. 3, D-91052 Erlangen, Germany
d
Department of Environmental Engineering, Faculty of Engineering, Al-Hussein Bin Talal University, Maan, Jordan

a r t i c l e i n f o a b s t r a c t

Article history: In kaolin classification, the final product is isolated from the overflow of a 50 mm hydrocyclone with an
Available online 17 August 2010 underflow containing fine sand and residual kaolin. For a higher kaolin recovery, the underflow is usually
further treated using a washing cyclone. This paper presents a new development in fine sand washing
Keywords: using water injection in the conical section of the classifying cyclone. This washing results in a 45–50%
Hydrocyclone reduction of the kaolin fraction (<10 lm) in the underflow. Thus, water injection into the classifying
Classification cyclone leads to a separation comparable to downstream washing.
Fine particle processing
Ó 2010 Elsevier Ltd. All rights reserved.
Control
Water injection

1. Introduction compared their results with conventional cyclones. A classification


study was performed by Udaya et al. (2005) to recover free miner-
In kaolin processing, hydrocyclones are the favored devices for als from a lead–zinc ore. These investigations were carried out
the selective separation of fine particles (kaolin) and coarse parti- using a 100-mm water-injection cyclone. The authors found that
cles (sand). The advantages of the hydrocyclone include its simplic- the cut size d50 gradually increases with the water injection rate
ity, ease of operation, and high volumetric throughput as well as until a value of 100% of the underflow water is reached. Further in-
relatively low capital investment, operating and maintenance creases beyond 100% in the water injection rate lead to sharp in-
costs. However, the physical principle of the hydrocyclone includes creases in cut sizes. The results also indicate that the sharpness
two disadvantages: (1) varying feed conditions that lead to consid- of the separation decreases gently with increased water injection
erable fluctuations in cut size and separation sharpness and (2) fine rates. The study concluded that the optimum water injection rate
particles that are inappropriately discharged with the liquid of the is approximately 100% of the underflow water rate.
underflow. To compensate for these difficulties, a special hydrocy-
clone control has been developed by Neesse et al. (2003a,b,
2004a–c, 2007). 2. Principle of the water injection into the hydrocyclone
Further attempts have focused on the washing of the sediment
in the conical section of the hydrocyclone (Honaker et al., 2001; A schematic of the new concept for water injection is shown in
Udaya et al., 2004, 2005). Tangential water injection into the cy- Fig. 1. The injection flow, which is introduced tangentially to the
clone cone has been applied to displace feed pulp water in the main flow, meets the three-dimensional hydrocyclone flow with
underflow stream and to increase the sharpness of the separation. the tangential, axial and radial components. Simulations of the
Dahlstrom (1954) successfully adopted this technique for deslim- water injection into the hydrocyclone have been carried out by
ing mined products. Kelsall and Holmes (1960) developed the first Dueck et al. (2010). First, the injection flow is transported down-
unit for the classification of sand, in which 48% of the material ward in the outer swirl. The radial component of the injection flow
remaining in the feed solution that was <10 lm was transferred washes the sediment and transports the extracted fine particles to-
to the underflow and was reduced to 11.5% after water injection. ward the inner vortex. A washing effect is achieved if the radial
Experiments with coal were performed by Firth et al. (1995) on a component of the injection flow can overcome the zero-line radius
hydrocyclone with water injection. of the axial velocity before the injected water is discharged
Mohanty et al. (2002) investigated a cyclone with water injec- through the underflow. Finally, the washed fine particles are trans-
tion in the cleaning process of fine coal particles (<45 lm) and ported in the inner swirl upwards to the overflow.
The wash effect depends on the flow rate of the injected water,
* Corresponding author. Tel.: +49 96227039416. the number of injection points and the location of the injection.
E-mail address: vgolyk@akwauv.com (V. Golyk). The distance from the injection point to the inner vortex becomes

0892-6875/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.mineng.2010.07.008
V. Golyk et al. / Minerals Engineering 24 (2011) 98–101 99

Fig. 2. Tangential water injection with three injection points.

Table 2
Injection parameters.

Parameter Value

Fig. 1. Principle of underflow washing by water injection. Number of injections, n 3


Injection diameter, din (mm) 2.5
Distance from apex end (mm) 40
Supply pressure of the water valve 3
shorter as the injection approaches the end of the conical section. (bar)
Water injection rate, Qin (L/min) 1–7
The new technique developed for small cyclones should result in a
minimum fine particles recovery in the underflow with low con-
sumption of injection water. This has been achieved by operating
with the injection at the cone end. 0.25
Dc= 50 mm
0.20 Do= 20 mm
3. Water only experiments Du= 6 mm
Pi = varied
OF

0.15
0

The water injection was implemented in a 50-mm hydrocy- P=0.6 bar


Δ QOF /Q

clone. The cyclone geometry is listed in Table 1. 0.10 P=1 bar


Fig. 2 shows the tangential water injection with three injec- P=1.6 bar
0.05
tion points. Table 2 lists the injection parameters used in the
experiments. 0.00
Before experimenting with kaolin suspension a test series was
conducted with water only to determine the distribution of in- -0.05
jected water in the overflow and underflow. Fig. 3 shows the influ- 0 1 2 3 4 5 6 7 8
ence of wash water flow on the change in overflow rate (DQOF) as a Injection rate (IR), [L/min]
proportion of overflow rate at zero wash water (Q 0OF ).
From Fig. 3, it can be seen that the flow rate of overflow due to Fig. 3. Related overflow flow rate versus water injection rate.

water injection increases gradually as the injection rate increases.


This flow component is significant for the wash effect because it
transports the fine particles from the sediment in the cone to the Fig. 4 indicates a maximum in the related underflow rate at
overflow. The effect is stronger at lower feed pressures. Therefore, medium injection rates. The injection rate at this maximum
one may conclude that water injection functions better at lower reaches 5% of the cyclone throughput. At low injection rates, the
pressure values. underflow is only weakly influenced, with the injected water pref-
Fig. 4 shows the influence of the injection rate (IR) in L/min on erably reaching the inner swirl traveling to the overflow. This effect
the change in underflow (DQUF) as a proportion of underflow rate is desirable because the underflow suspension should not be di-
at zero wash water (Q 0UF ). luted, and only a water exchange will occur. Again, a greater effect
at low pressure was found. At higher injection rates, the tangential
flow in the conical part of the cyclone is intensified, which results
Table 1 in an increase of the air core diameter. This in turn leads to a re-
Hydrocyclone dimensions. duced underflow rate, which influences the separation negatively
and increases the cut size.
Parameter Value (mm)
Fig. 5 shows the change of the ratio of the volume fluxes to
Nominal diameter, DC 50
overflow and underflow (the volume split S) with the injection re-
Inlet 8  18
Overflow diameter, DO 16
lated to the volume split S0 without injection, which is dependent
Underflow diameter, DU 6 on the injection rate (L/min) at different feed pressures. It can be
Length of the cylindrical part 130 seen that at low injection rates, only a low increase of S is observed.
Length of the conical part 350 At high injection rates, the split parameter increases markedly
Length of the apex 135
with the increasing injection rate.
100 V. Golyk et al. / Minerals Engineering 24 (2011) 98–101

0.15 stage, where residual kaolin can be extracted and added to the final
Dc= 50 mm product of the overflow from the desliming stage.
0.1 Do= 20 mm The new technological concept involves the installation of a
QUF /Q0UF , %

Du= 6 mm
0.05 Pi = varied water injector in the desliming cyclone stage. The goal is to replace
0 the washing stage currently used with this water injection. The 50-
mm hydrocyclone has been operated at 1.0 bar and an injection
-0.05 P=0.6 bar pressure of 3 bar. Fig. 7 shows the portion T of every size fraction
-0.1 P=1 bar traveling to the cyclone underflow (separation efficiency T) of the
-0.15 P=1.6 bar 50-mm hydrocyclone for different water additions. The T0-value
(T-value for d ? 0) can be used for the evaluation of the washing
-0.2 effect.
0 1 2 3 4 5 6 7 8
The most important conclusion derived from Fig. 7 is the exis-
Injection rate (IR), [L/min] tence of the minimum T0-value at an injection rate of 3 L/min. This
finding can also be concluded from Fig. 8, which shows that the T0
Fig. 4. Related underflow flow rate versus water injection rate.
curve has a minimum (a maximum wash effect). This minimum
could be expected from the water only experiments. At moderate
2 injection rates up to 3 L/min, a water exchange in the underflow
1,8 takes place, and the solids content in the underflow is reduced only
Related Split S/S0

1,6 by the replaced fines. At low injection rates, a decrease of the cut
1,4 size is observed, which is caused by higher settling rates in the di-
1,2 luted suspension. Higher injection rates lead to a remixing of the
1 fines by intensive turbulence and an increase of the cut size.
0,8 p = 0.6 bar Table 3 presents the results using the novel wash water injec-
0,6 p = 1 bar tion method, as compared to the conventional method. The sieve
0,4 p = 1.6 bar retention (>25 lm) in the overflow, the fraction (<10 lm) in the
0,2 underflow and the solid recovery were measured.
0
0 2 4 6 8
Injection Rate (IR), [L/min]
1
Fig. 5. Related volume split S/S0 versus water injection rate.
Separation efficiency, T0

Td (0.5) Dc = 50mm
Do /Du = 4.8
In summary, the water only experiments indicate that water pi = 1bar

injection gives optimal results at moderate injection rates. In these


conditions, a water exchange of the underflow exists, which results
in the desired high solids content in the underflow. 0,1
no injection
IR=1 L/min
4. Experiments with kaolin suspension IR=2 L/min
IR=3 L/min
T 0,min(IR = 3L/min) IR=4 L/min
Fig. 6 presents a schematic of the last stages of kaolin classifica- IR=5 L/min
tion. There are two stages, both equipped with 50-mm hydrocy- 0,01
clones. The desliming stage produces the final kaolin quality in 0,1 1 10 100
the overflow. After dilution, the underflow is directed to a washing Particle size, µm

Fig. 7. Partition curves of the 50-mm hydrocyclone for different water injection
rates (DO = 20 mm, DU = 4.2 mm).

0,3 32
Separation
Separation Efficency T

0,25 Efficiency 30
Cut Size d50 [µm]

Cut Size d50

0,2 28

0,15 26

0,1 24

0,05 22

0 20
0 1 2 3 4 5 6
Injection Rate IR [L/min]

Fig. 8. Separation efficiency and cut size of the 50-mm hydrocyclone versus water
Fig. 6. Schematic of the kaolin classification. injection rate.
V. Golyk et al. / Minerals Engineering 24 (2011) 98–101 101

Table 3 low and is less than 10% of the feed flow. The investment for the
Comparison of the 50-mm cyclone with and without wash water injection. implementation of a controlled wash water injection system is
Desliming without Desliming with low. Only one cyclone of a hydrocyclone battery is monitored,
injection injection and it controls the other apparatuses via a single control valve.
% >25 lm in the overflow 0.085 0.115
% <10 lm in the underflow 14.6 7.6
Solid recovery (%) in the 41 26 References
underflow
Dahlstrom, D.A., 1954. Fundamentals and application of the liquid cyclone. In:
Chemical Engineering Progress Symposium Series, vol. 15. American Institute of
The sieve retention (>25 lm) in the overflow represents the Chemical Engineers, p. 41.
Dueck, J., Pikushchak, E., Minkov, L., Farghaly, M., Neesse, Th., 2010. Mechanism of
quality requirements for the kaolin. In the case of desliming with- hydrocyclone separation with water injection. Minerals Engineering 23, 289–
out injection, the value of the sieve retention is slightly less than 294.
the result with injection, but both cases meet the demanded con- Firth, B., Edwards, D., Clarkson, C., O’Brien, M., 1995. The impact of fine classification
on coal preparation performance, In: Proceedings of the 7th Australian Coal
dition of <1.5%.
Preparation Conference, Paper E2, Australian Coal Preparation Society, pp. 250–
The fraction (<10 lm) in the underflow represents the kaolin 276.
loss and should be minimized. The water injection reduces these Honaker, R.Q., Ozsever, A.V., Singh, N., Parekh, B.K., 2001. Apex water injection for
losses by about 50%, which corresponds to the value of the under- improved hydrocyclone classification efficiency. Minerals Engineering 14 (11),
1445–1457.
flow in the washing stage. With wash water injection, the kaolin Kelsall, D.F., Holmes, J.A., 1960. Improvement in classification efficiency in hydraulic
recovery in the last cyclone stage increases by 15%. Thus, the cy- cyclones by water injection. In: Proceedings of the 5th Mineral Processing
clone with water injection may replace the washing stage. Congress, Paper 9, Inst. Mining and Metallurgy, London, pp. 159–170.
Mohanty, M.K., Palit, A., Dube, B., 2002. A comparative evaluation of new fine
particle size separation technologies. Minerals Engineering 15, 227–236.
5. Conclusions Neesse, T., Schneider, M., Dueck, J., Donhauser, F., Regler, J., Tiefel, H., 2003a.
Computer controlled hydrocyclone battery with maximum solids recovery. In:
Proceedings: XXII International Mineral Processing Congress, Cape Town, South
The addition of water injection to small hydrocyclones with Africa, pp. 1552–1560.
diameters of 50 mm has been introduced to kaolin classification. Neesse, T., Schneider, M., Regler, H., Donhauser, F., Tiefel, H., 2003b. Technical
processes in the regeneration of bentonite suspensions for hydro-shield drives.
The present investigation confirmed that an optimum injection
Tunnel 3, 29–35.
rate exists when a high washing effect of the fine particles is com- Neesse, T., Schneider, M., Dueck, J., Golyk, V., Buntenbach, S., Tiefel, H., 2004a.
bined with relatively stable cut sizes. This development improves Hydrocyclone operation at the transition point rope/spray discharge. Minerals
the separation markedly, such that an underflow washing stage Engineering 17, 733–737.
Neesse, T., Golyk, V., Kaniut, P., Reinsch, V., 2004b. Hydrocyclone control in grinding
can be replaced by the water injection and can result in the same circuits. Minerals Engineering 17, 1237–1240.
level of kaolin recovery. An alternative is to increase the kaolin Neesse, T., Schneider, M., Golyk, V., Tiefel, H., 2004c. Measuring the operating state
recovery by wash water addition in the desliming stage and to of the hydrocyclone. Minerals Engineering 17, 733–737.
Neesse, T., Tiefel, H., Kaniut, P., 2007. Volume split control of a hydrocyclone group.
maintain the downstream washing stage. However, the water Minerals Engineering 20, 355–360.
injection at the apex is sensitive to variations in the feed. There- Udaya, B.K., Govindarajan, B., Barnwal, J.P., Rao, K.K., Rao, T.C., 2004. Modelling
fore, a controlled wash water injection is unavoidable. The hydro- studies on a 100 mm water-injection cyclone. Physical Separation in Science
and Engineering 13 (3–4), 89–99.
cyclone control system is based on optical monitoring of the Udaya, B.K., Govindarajan, B., Barnwal, J.P., Rao, K.K., Gupta, B.K., Rao, T.C., 2005.
underflow. This monitoring makes the entire hydrocyclone classi- Classification studies of lead–zinc ores fines using water-injection cyclone.
fication more transparent. The water consumption is relatively International Journal of Mineral Processing 77, 80–94.

You might also like