Minerals Engineering, Vol. 14, No. 9, pp.

975-985, 2001
Pergamon © 2001 Elsevier Science Ltd
All rights reserved
0892-6875(01)00105--4 0892-6875/01/$ - see front matter

CURRENT STATUS AND LIMITATIONS OF COPPER SX/EW PLANTS

CONTROL*

L.G. BERGH and J.B. YIANATOS

Chemical Engineering Department, Santa Maria University, Valparaiso, Chile

Email: [email protected]
(Received I January 2001; accepted 27 June 2001)

ABSTRACT

A review of the state of the art and trends in automation and control of copper hydrometallurgical
processes is presented. Besides the great expansion of hydrometallurgical processes world-wide,
there is a number of unsolved problems related to lack of instrumentation, lack of process
knowledge, odd operating practices, and general lack of use of data management and processing. In
general, process control of local objectives are frequently achieved, however, application of mature
and new techniques, successfully adopted in other mineral processing plants, are seldom reported.
In the near future it is expected that intelligent techniques will be incorporated to solve a large
variety of problems. To test and evaluate new control strategies, a pilot SX/EW plant has been set
up and instrumented at the Process Control Laboratory, Santa Marfa University, in Chile. The
control system is organized at three levels in an architecture including field controllers, PLC and
PC networks. The main objectives are to develop and test intelligent techniques in an hybrid system,
and to study measurement and supervisory control problems. It is hoped to produce reliable results
that will contribute to improving the efficiency of this emerging and important process. This paper
describes the current status and limitations of control of SX/EW plants and summarises where the
main deficiencies lie and a proposed methodology to remedy them. © 2001 Elsevier Science Ltd. All
rights reserved.

Keywords
Process control; hydrometallurgy; solvent extraction; process instrumentation

INTRODUCTION TO LX-SX-EW PROCESSES

Leaching, solvent extraction and electro-winning (LX-SX-EW) processes have become increasingly
important to concentrate, purify and separate metal ions and inorganic salts. The most common commercial
application of these processes is in the copper industry (Kordosky, 1992). Although oxide ore is the
preferred leaching mineralization, many operations also leach sulfide ores as well. A large number of
commercial SX/EW plants have been successfully commissioned all over the world in the last two decades
(Arbiter and Fletcher, 1994), and over 90% of them are located in North and South America.

The first applications of leaching in piles started in the 80"s in Chile. In the raffinated copper market, the
production of high purity cathodes by lixiviation, solvent extraction (SX) and electro-winning (EW) has
been growing very fast. The annual overall and SX/EW copper productions in the world are shown in

* Presented at Minerals Engineering 2000, Cape Town, South Africa, November 2000

975
 976 L.G. Bergh and Y. B Yianatos

Figure 1 for recent years. Also the annual SX/EW copper produced in Chile is included. In 1994, for
example, slightly over 200 thousands tons of copper were produced by SX/EW in Chile. In 1998, the
copper production by SX/EW plants was increased to over 1,100 thousands tons, representing 47% of the
total electrolytic copper produced in Chile. Moreover, in 1999 the total SX/EW copper produced was 1,400
thousands tons (Pezoa, 1999). Table 1 shows the actual and future (2001) production of SX/EW copper in
Chile.

12000

10000

8000
t-
O
6000 ..................... ; ..................... i ........................................... ~ ....................

o
4000
..................... i..................... 7........................................... i ....................
2000 .....................~!.....................i................. . ~ , "_.~.;"; : i .........
0
1992 1994 1996 1998 2000 2002
Year

- - • - • Chile SX/EW • World SX/EW • Total

Fig. 1 Annual total and SX/EW (Chile and world) copper productions.

TABLE 1 Actual production and new SX/EW Projects in Chile (Pezoa, 1999)

Company name 1999 capacity Future capacity (2001)

Thousands tpy Thousands tpy
Radomiro Tomic 180 250
SCM El Abra 225
Minera Zaldivar 200
Cerro Colorado 100
I Quebrada Blanca 75
Escondida Oxidos 125
Collahuasi 50
Santa B~bara 45
Manto Verde 50
Chuquicamata 160
Michilla 50
Lomas Bayas 60
El Tesoro 75
Exp. Lomas Bayas 90
Fortuna de cobre 60
Chimborazo 60
Spence 40
Others 180
Total 1220 1735

Typically the process starts with the crushing and grinding of ore. Since these processes are classified as
mineral processing operations, a more detailed discussion on the control problems can be found in Hodouin
et al. (2000), and the current state and future trends in automation of mineral and metal processes in Jiims~i-
Jounela (2000). The ground material is collected in large heaps, which are sprayed with an acid solution to
dissolve out the metal from the ore. The metal is transferred to the aqueous phase, called the pregnant leach
 Currentstatus and limitationsof copperSX/EWplants control 977

solution (PLS). After the leaching stage the solution contains impurities, which have to be removed. A
simplified flow diagram of the process is shown in Figure 2.

The solvent extraction stage decreases the proportion of impurities and concentrates the solution. The metal
ions are selectively removed from the aqueous PLS via the organic phase and pass back into the aqueous
phase. The flow is divided between several parallel SX trains. Extraction and stripping, i.e. the transfer of
metal ions into the organic phase and back into the aqueous phase are performed within the mixer-settlers.
In the mixer, the organic and aqueous phases are brought in to close contact with each other by dispersing
the organic or aqueous phases into droplets a few millimetres in size within another phase. After mixing,
the two phases are separated in large settlers by gravity. To improve the overall efficiency of transfer of
metal from the aqueous solution, the solvent extraction process has internal recycling and some material is
returned to the leaching stage.

Ore "1 L~ehing I"

p

Organic stream

I I .,J
i
Loaded organic !
ppmg
s .
Rich electrolyte Poor electrolyte

Copper "q""nEleetrowinning

Fig. 2 Typical flow diagram of a LX/SX/EW plant.

STATE OF T H E ART IN COPPER HYDROMETALLURGY C O N T R O L

Control objectives

The aim of a LX-SX-EW process is to obtain copper cathodes at low cost by treating liberated ore,
previously prepared by comminution and size separation processes. The process to be controlled in LX-
SX-EW can thus be classified into four different categories:

• Minerals liberation processes (crushing, grinding and size classification)

• Metals separation processes by reaction and mass transfer (leaching, solvent extraction)
• Metals separation processes by electrolysis (electro winning)
• Peripherical processes (feeding systems, pumping, filtering, material handling, etc)

The overall control objective of a LX-SX-EW plant is to produce high quality commercial cathodes, which
meet some standards, while maximising the net revenue of the plant (Bergh et al., 2000). Usually these
objectives require some trade-off between the concentration and flow of PLS, the flow and concentration in
the organic phase of copper and contaminants, the degree of entrainment of organic in aqueous, the
aqueous carry over in organic, the consumption of reagents, the electric current stability and distribution,
the time scheduling of cathodes and anodes changing, among others.
978 L.G. Berghand Y. B Yianatos

Control approaches

Ideally real time plant-wide optimisation should be the right approach to LX-SX-EW plant control, as in
other processes (Herbst et al., 1995), i.e. the adjustment of the operating conditions of the various units as a
function of the raw ore properties and feed rate, metal market prices and energy and reagent costs.

However, a first major problem with such an optimal approach is that the major disturbances of the plant,
which are the variations of the raw ore properties, are extremely difficult to predict, as well as their effects
down to the final products. These almost unpredictable and immeasurable properties of the feed ore
continuously change the steady-state and dynamic behaviours of the plant, as well as the location of the
optimal operating targets.

An essential feature of classical control and optimisation strategies is the availability and effectiveness of a
mathematical model that accurately describes the steady-state and dynamic characteristics of the process in
the whole operating range, including its non-linear behaviour. This leads to a second major problem for
LX-SX-EW control, since satisfactory mathematical models are not frequently available for the following
reasons: a) the physics and chemistry of the sub processes involved are poorly understood, and b) the
relevant information about the state of the sub processes are quite difficult to measure and describe. Some
empirical modelling has been reported (Aminiam et al., 1997) but usually the validity of such models is
restricted and specific to a plant, and its use requires information not commonly available on-line in
industry.

LX-SX-EW optimisation and control cannot be performed without a minimum amount of information on
the input disturbances, the process states, and the final product quality. This is in fact the bottleneck of LX-
SX-EW control. The efficiency of the operating strategies is totally dependent upon the quality of the
information which is used, because on one hand it is used to build the knowledge encapsulated in the
models which the control strategy is based upon, and on the other hand it is the input to the real-time
optimisation and control algorithms. Therefore, due to both instrumentation and modelling problems, the
usual approach is to separate the optimisation and control problem for each unit operation. This control
hierarchy is widely accepted in the LX-SX-EW industry and is a mature technology. Usually, only the
regulatory control of the low level is implemented, the set points being manually selected. Decentralized
SISO stabilizing control is thus a mature technology widely used in the LX-SX-EW industry.

However, for time variant and non-linear LX-SX-EW processes that undergo large unknown disturbances,
these multiple SISO control loops sometimes exhibit poor performances. Three types of avenues are used
to cope with these problems: MIMO regulatory and supervisory optimising control based on mathematical
models, AI techniques based on empirical or heuristics process models, and finally the application of
operating rules in expert systems. Studies or application of multivariable regulatory control, adaptive, non-
linear and robust control theory to the LX-SX-EW processes are not reported in the literature. A similar
situation prevails for the promising alternatives of AI techniques and expert systems.

Since comminution and classification are processes discussed in Hodouin et al. (2000), this paper will
focus only on leaching, solvent extraction and electro winning processes. Peripheral processes control
strategies are not discussed here.

Leaching process

To obtain and maintain a close steady-state operation it is necessary to have good control of the solution
flow through the operations. The leaching operation is usually controlled using pumps and through
programmable logic controllers (PLC). In general, the leaching stage is governed by three factors:
recirculated raffinate flow from solvent extraction, PLS dam level, and PLS pump flow from the collector
tank into the solvent extraction feed pool.

The magnetic flow meters, pond level controls, alarm system in the control room, variable speed motors in
the raffinate pumping, and operator visual inspections are some of the main tools utilized to reach a
solution balance within the plant. The pond levels are usually controlled by the number of pumps operating
 Currentstatus and limitationsof copperSX/EWplants control 979

at one time in the heap PLS pond. PLS stacking on the heaps is periodically utilized to eliminate PLS
overflows from the PLS feed pond to the raffinate pond.

Depending on the complexity of the leaching processes, the PLS may be formed from several leaching
outputs. In this case an appropriate managing of the streams mixing is crucial to minimise disturbances
(both copper concentration and pH of PLS) coming down into SX.

Solvent extraction process

The combinations of flows and their control in the solvent extraction stages are important to maintain
satisfactory production levels. The control of every entering extraction and re-extraction flow is performed
by the following flow control loops: pregnant leach solution entering the first extraction stage; organic
solution entering the second extraction stage; poor electrolyte entering the re-extraction stage and
recirculation from the aqueous launder to the re-extraction stage mixer box.

The operator-defined organic/aqueous ratio determines the proportions of the organic flow in the first
extraction stage and the re-extraction stage feed of the poor electrolyte.

The four factors governing the solvent extraction stage are poor electrolyte, PLS, organic and recirculated
rich electrolyte flows. Each of the flows are controlled. An orifice plate measures the organic flow, while in
the other stages a magnetic flow meter is applied.

In general, the most critical problems found in SX plants are related to (Farfas et al., 1994):

- pH of PLS
Copper concentration in PLS
- Crud formation
Aqueous carry over in organic
- Organic entrainment in aqueous
Phase separation times

In many cases the copper concentration and pH are regulated from samples that are taken manually and
spread. Chemical assays are known every 4 or more hours. Thus, for instance the average desired value of
10 g/L of Cu varies from 9.8-12 g/L while pH varies from 1.6-2.8. It must be noticed that important losses
of efficiency and selectivity are observed under deviations of-+0.3 g/L and _+0.2 in pH.

Some problems are associated with entrainment from one phase into the other, leading to plant instability,
low quality of copper cathodes, excess consumption of reagents, increased anodic corrosion, transport of
undesirable elements to the electrolyte, and extra losses of organic. Some problems may be partially solved
with the help of filters, settlers, column flotation, separators, coalescers, flocculants, etc. However, some of
these problems may be avoided with high quality process information and appropriate control strategies
and operating practices. Presently, in most operations the level of the aqueous/organic interface is not
measured directly, and the level is regulated manually. This condition favors the entrainment of organic
phase into the aqueous phase to be processed in the EW plant. The problem here is the loss of organic
(expensive) and the loss of capacity of the EW cells, the need to sacrifice some banks of cells because of
contamination. Furthermore, the usual lack of coordination of the flow rates and mixing of different process
streams cause disturbances and increases the organic losses. Another problem is the chloride contamination
of the electrolyte, which has an important impact on the life of permanent cathodes.

Electro winning process

The control loop for the rectifier is usually designed to maintain the direct current consistent with the set
point. The aim is to protect the rectifier from overheating and to prevent major consequences such as
unbalanced Cu in the plant, accelerated corrosion of anodes, increase in cathodes noduling, generation of
short circuits and decrease of cathode purity.
980 L.G. Bergh and Y. B Yianatos

The electrolysis cells have two important automation applications: operation of the baths, and voltage
monitoring and accounting. The bath operation includes a schedule for anode changing, cathode changing
and cleaning intervals for each cell. All the bath voltages and currents are scanned at predefined intervals.
The lower limit value is used to check for short circuits. Values higher than the upper limit indicate
passivity of the electrodes. Temperature measurement is carried out by means of an infrared camera
installed on the handling crane. An imminent short circuit can be seen from locally increasing temperatures.
The supervision of cell voltage and temperatures allow trace ability of short circuits on the basis of cell
location information (Siemens, 1998; Spitzlei, 1994).
Electrolyte treatment is managed with the goal of maintaining a predetermined constant temperature.
Sensors e.g. for temperature, level, flow, pressure and conductivity, are used. The main task is to control
the flow of hot water or steam to the heat exchanger. All motors, stirrers, pumps and valves are controlled.
The levels in vessels and tanks are kept within acceptable limits.

Anode and cathode preparation machines are equipped with their own PLC controllers. They handle the
internal logic sequences of the preparation processes and monitor the input and output conveyors.

Current misdistribution and short-circuiting in EW cells are the main causes of cathode rejection. The
electrolyte temperature and its relation with the proper current density is also controlled. Low temperature
causes crystallization.

INFORMATION ACQUISITION AND PROCESSING

Instrumentation

In comparison with other mature mineral processes, a rather slow incorporation of instrumentation and
control systems has been observed, pHmeters, conductimetry (to measure phase continuity in mixers) and
flow meters can be considered mature instrumentation.

On-line analysers for monitoring and control purposes have been reported lately. A number of applications
exist with Outokumpu's Courier system (Hughes and Saloheimo, 2000) and Amdel stream analysers,
however these techniques are still active in the sense that their availability is too sensitive to maintenance
and calibration programs.

Even when, in general, mature sensors for detecting interface position are available in the market, the
application of these sensors in SX plants to control the organic/aqueous interface position is rarely found,
and is not reported in the literature.

Data management and communication

Real-time and historical information is useful for global plant optimisation. Smart data management
systems are required for efficient communication between the business staff (information on metal
inventories, costs, production objective, equipment availability), the process engineers (information for
production optimisation and control), the laboratory (quality control), the environment department, and the
operators of the various units. In addition to the data exchange facilities, the format of the information must
be easily adapted to the various objectives of data processing (local control, loop tuning, mass balance
calculation, process modelling, maintenance and trouble-shooting, performance indicator display, real-time
optimisation). Bascur and Kennedy (1999) extensively described the availability of the data management
architectures available and their benefits. Innovative communication systems between remote locations are
emerging. In general, actual data managements and communications facilities are not frequently used in
LX-SX-EW plants.

Data reconciliation

Because of the inherent inaccuracies of measurements, the raw data delivered by sensors, such as flow rates
and chemical assays, contains errors. Data reconciliation procedures are used to correct measurements and
make them coherent with prior knowledge about the process. Frequently, mass conservation equations are
 Currentstatus and limitationsof copperSX/EWplants control 981

used as a basic model to reconcile redundant data with prior knowledge constraints (Crowe, 1996). At the
same time, data reconciliation techniques may be used to infer unmeasured process variables such as flow
rate and composition of internal streams of a complex unit. Applications for LX-SX-EW plants are not
reported, but with the consolidation of on-stream analysers it is expected they will be rapidly adapted.

Pattern recognition

Historical or real-time sets of measurements on multivariable processes contain massive amounts of

information about the behaviour of the operation. However, they are difficult to exploit because of the high
number of available variables, their poor reliability and finally the lack of measurements for the most
important properties as mentioned above. Statistical or AI techniques are, in general, active or emerging to
extract, from these data sets, pieces of information which may be useful for monitoring, predictive
maintenance, diagnosis, control and optimisation. A study to correlate the copper concentration of poor
electrolyte with other measured variables was performed by Katajainen (I 998) using plant data.

Process supervision, fault detection and isolation

In general, some methods are emerging to detect either sensor biases or model inadequacy using
multivariable statistical tests on the residuals of material balance constraints (Berton and Hodouin, 2000;
Hodouin and Berton, 2000). ANN are also active methods to detect and diagnose faults (Aldrich and Van
Deventer, 1995). Supervision of the control strategy for processes such as flotation columns is used to
detect sensor or operating problems using data validation and expert systems (Bergh et al. 1999; Bergh and
Yianatos, 2000). In particular, no applications have been reported in LX-SX-EW plants. It is expected that
they will appear, as soon as the information about the relevant process variables allows the development
and implementation of supervisory control, on top of conventional distributed control systems.

RESEARCH PROPOSAL

Since L X - S X - E W processes for copper production are relatively new, most of the operating problems and
process inefficiencies are partially attributed to plant design and reagents selection. However, the
information about the relevant variables in these processes are poor and inaccurate, such that stabilization
of the plant around a steady state can only be partially achieved by decentralized conventional control of
flows, levels and so on.

To improve the process efficiency and quality of the products, the incorporation of reliable instrumentation,
such as on-stream analysers and automatic titrators, is needed. When this has been achieved, the
development and implementation of supervisory control will be possible. This supervisory control will be
an hybrid system that will incorporate all the new techniques successfully applied in other fields, such as
mineral processing.

The lack of process knowledge and accurate on-line information on relevant process variables has largely
contributed to little development in information and control systems. In most large-scale plants,
experimentation under well informed conditions (measurements, monitoring, control) cannot be performed
due to the risk involved and because of the lack of facilities. Usually, the analysis of operating data is
inconclusive due to the frequent confusion in estimating effects from corrupted data. On the other hand,
pilot scale plants are not fully instrumented and practically they are manually operated.

A successful approach in many other processes is to set up a pilot scale plant with a range of facilities. At
least the most important variables have to be monitored on-line and basic control loops should be able to
operate the plant at a truly steady state. The information and control system should allow the plant to move
safely from one steady state to another, and to keep a record of all the control and manipulated variables.
Only when these facilities have been achieved, can experimentation dealing with finding correlations
between operating variables be carried out. Therefore, a research program should start with the set up and
instrumentation of a pilot plant. A complete control and information system has to be developed in order to
obtain reliable data to model and control the process, and to test the application of new techniques. The
specifications and description of a pilot scale plant installed at Santa Maria University is presented.
982 L.G. Berghand Y. B Yianatos

SX/EW pilot plant at Santa Maria University

The SX/EW plant is schematically described in Figure 3. Two extraction units, 75x33x36 cm 3, with a
capacity of 90 L, were constructed in a special resin, 4 mm thick, and covered with a chemical resistant
film of isoftalic resin. The residence time in mixers was designed to be 3 minutes, with a superficial
velocity of decantation of 2.4 mm/s. The organic and aqueous flows are pumped from tanks to cells at a
maximum rate of 2 L/min. The stripping unit is of similar dimensions. An electrowinning unit, divided in
three cells of 6 SS 316, 21.6x21.6 cm 2, cathodes, working with a current density of 200 A/m 2, and
producing 2.4 kg of copper in one 8-hours shift. The system is completed by 21 Pb/Cd/Sn anodes. In Figure
3, the design copper and acid concentration for the main streams are included.

cu3 #L
Acid 0.5 g/L Cu 0.3 F,/L
Acid 4.7 g/L
Organic

Extraetionl ~ Extraction2 ~'~Raffmat)

"x._./
I Aqueous t
~1 stripping
c.2.7¢L c.4,~ _L' / c.a3~
Acid I$0WL • [ I Acid 174g/L
Cu 300 g/h Copper "~''"lEleetr°winning rl

Fig. 3 Flow diagram and main streams in the SX/EW pilot plant.

The instrumentation and control system is described in Figure 4. The pregnant solution flow rate is
controlled by a variable speed peristaltic pump and the total aqueous flow rate to E-1 can be controlled by

I' ' .I ~/......(~

E-1

III,
Fig. 4 Instrumentation and control system of the pilot plant.
 Currentstatusand limitationsof copperSX/EWplantscontrol 983

manipulation of the control valve opening, installed in the recycle stream, and by selection of the suction
pump speed. Each cell has five compartments. The first has a variable speed suction pump with an impeller
designed to produce small droplets of organic. The drop size distribution depends on the suction pump
speed. The inlet flow rate depends also on the gap between the impeller and the suction hole. This gap can
be manually adjusted, then the main control variable is the impeller speed. The motor speed is controlled by
using frequency controllers. The other three compartments are used to separate the organic and aqueous
phases. The organic flow rate to E-1 is the organic overflow from E-2. The mixer speed is manipulated by
setting the set point of a frequency controller. The speed of all the mixers are the same. In the mixer a
conductivity probe is used to detect phase continuity. The organic/aqueous and the organic/air interfaces in
the decanter are detected by measuring a conductivity/capacity profile. These signals are processed to
estimate the interface positions. Similar instrumentation is used in E-2. The charged organic flow rate from
E-1 to S-1 is controlled by a variable speed peristaltic pump. The electrolyte flow rate to S-I is controlled
by a variable speed peristaltic pump and the total electrolyte flow rate to S-1 can be controlled by
manipulation of the control valve opening, installed in the recycle stream, and by selection of the suction
pump speed.

The charged electrolyte to EW is controlled by a variable speed peristaltic pump. The electrical current is
measured and controlled. Temperatures at different positions are measured in order to detect current
misdistributions.

The control is organized at three levels as is schematically shown in Figure 5. A field level for frequency
variators (mixers and suction pumps), speed controllers of peristaltic pumps, and electrical current at EW.
A second level of control is organized in a PLC (General Electric, Fanuc 90/30, CPU 350). All field
measurements are communicated as inputs to the PLC, while all the local controller set points are outputs
from the PLC. Emergency and shut down procedures are programmed in PLC. All the information acquired
is communicated to a PC server. In here reside the high level functions, such as process monitoring,
processing of information, process supervision (including fault detection, data validation, supervisory
control) and data base generation, All this information is available in a PC Ethernet network for monitoring
and off-line data analysis.

P L C GE-Fanuc 90/30 PC Server PC

Monitoring Monitoring
CPU I/O modules
Supervision Data analysis
350
Data base

Conductivity profile in E-1

II
Off-line:
I/P transducer, control valve reflux E-1
Conductivity profile in E-2 I/P transducer, control valve reflux E-2 Stream essays [
Conductivity profile in S-1 I/P transducer, control valve reflux S-1
Conductivity in mixer E-1 Pregnant solution flowrate setpoint (peristaltic pump)
Conductivity in mixer E-2 Charged organic flowrate setpoint (peristaltic pump)
Conductivity in mixer S-1 Spent flowrate setpoint (peristaltic pump)
Pregnant solution flowrate Mixers E-l, E-2 and R-1 speed setpoint (frequency)
Charged organic flowrate Speed suction pump setpoint in E-1
Spent flowrate Speed suction pump setpoint in E-2
Electrical current in EW Speed suction pump setpoint in S-1
Temperature profile in EW Electrical current setpoint in EW

Fig. 5 Information and control system architecture.

984 L.G. BerghandY. B Yianatos

Techniques and methodologies to be experimented

The plant has been constructed and installed to produce repeatable results. The first objective is centered on
the organic-aqueous phase separation problem. Two main problems are considered: the estimation of
interface position and the supervisory control. The conductivity profile data will be analysed by using
simple empirical models and by artificial neural networks. Fault detection and supervisory control will be
developed as hybrid systems, where conventional techniques (control and statistics) and Artificial
Intelligent techniques (expert systems, fuzzy systems, neural networks) will be combined. In general, the
approach will be that successfully employed in the study of some other processes, as flotation in columns
(Bergh et al., 1999).

CONCLUSIONS

Electrolytic copper production by using LX/SX/EW plants has been growing very fast in the last decade.
Today several projects are under development so that production capacity will continue to increase over the
following years. Chile is the most important producer of SX/EW copper.

In general, the most critical problems found in SX plants are related to: pH control of PLS, copper
concentration control in PLS, crud formation in cells, management of aqueous carry over in organic,
organic entrainment in aqueous and phase separation times.

The main problems in achieving a high quality cathode are related to: operation of the baths, voltage
monitoring and accounting, current misdistribution and short-circuiting in EW cells, electrolyte temperature
control and current density control.

So far improvements in operating and control systems have seldom been reported. Instrumentation
problems and lack of adequate mathematical models have largely contributed to this situation. On the other
hand a number of mature techniques, successfully applied in other mineral processes, are available to be
tested when basic instrumentation and control problems have been solved.

A pilot scale SX/EW plant has been designed and constructed at the Process Control Laboratory in the
Santa Marfa University. This plant has been instrumented and has an information system to develop and
test the most recent supervisory techniques. It is expected that a better management of the organic/aqueous
separation problems will provide significant insight for improving the control and operation of this
emerging process.

ACKNOWLEDGEMENTS

The authors would like to thanks Conicyt (Project Fondecyt 1990859) and University of Santa Maria (Project
992723) for their financial support.

REFERENCES

Aldrich C. and J. Van Deventer, Comparison of different artificial neural nets for the detection and location of
gross errors in process systems, 1995, Ind. Eng. Chem. Res., 34, pp. 216-224.
Aminiam H., Bazin C., Hodouin D. and Jacob C., Modelling and Simulation of a Solvent Extraction Plant,
1997, Proceedings 5 th Southern Hemisphere Meeting on Mineral Technology, May 6-9, Argentina, pp.
337-340.
Arbiter N. and Fletcher, A.W., Copper Hydrometallurgy Evolution and Milestones, 1994, Mining
Engineering, 46(2), pp. 118-123.
Bascur O.A, J.P. Kennedy, Real-time information management for improving productivity in metallurgical
complexes, 1999, Control and Optimisation in minerals, metals, and materials processing. D.Hodouin,
C.Bazin, A.Desbiens Edts, Metallurgical Society of the CIM, pp. 3-16.
 Current statusand limitationsof copperSX/EWplants control 985

Bergh L.and J. Yianatos, Improving controllability of Flotation Columns, 2000, Proceedings XXI
International Mineral Processing Congress, Rome, Italy, July 22-25, pp. C3 24-31.
Bergh L., Yianatos J., Acufia C., Pfrez H. and F. L6pez, Supervisory Control at Salvador Flotation
Columns, 1999, Minerals Engineering, 12(7), pp. 733-744.
Bergh L., J~imsii-Jounela S-L and D. Hodouin, State of art in copper hydrometallurgical processes control,
2000, Proceedings IFAC Symposium Future Trends in Automation in Mineral and Metal Processing,
22-24 August, Espoo, Finland, pp. 74-79.
Berton A. and D. Hodouin, Statistical Detection of Gross Errors in Material Balance Calculation, 2000,
Proceedings CONTROL 2000, SME Annual Meeting, Salt Lake City, February,pp. 157-166.
Crowe C., Data Reconciliation progress and Challenges, 1996, J. Process Control, 6(2/3), pp. 89-98.
Farias L., Reghezza A. and J. Vergara, Continuous improvement of the Solvent Extraction Process, 1994,
Proceedings IV Meeting of the Southern Hemisphere on Mineral Technology, Concepci6n, Chile, pp.
137-162.
J~ims~i-Jounela, S-L, Current status and future trends in the automation of mineral and metal processing,
2000, Proceedings IFAC Symposium Future Trends in Automation in Mineral and Metal Processing,
22-24 August, Espoo, Finland, pp. 96-108.
Herbst J.A., Pate W.T., Flores R.T. Zarate H.A., Plant wide control: the next step in mineral processing
plant optimization, 1995, Proceedings of the XIXth International Mineral Processing Congress, SME,
Colorado, USA., vol 1, pp. 211-216.
Hodouin D., J/ims/i-Jounela, S-L., Carvalho T., Bergh, L., State of the art and challenges in mineral
processing control, 2000, Proceedings Future Trends in Automation in Mineral and Metal Processing,
22-24 August, Espoo, Finland, pp 59-73.
Hodouin D., A. Berton, An algorithm for fault detection and isolation using mass balance constraints, 2000,
Proceedings XXI International Mineral Processing Congress, July 22-25, Rome, pp. A 59-65.
Hughes, D., Saloheimo, K., On-stream analysis in copper SX-EW processes, 2000, Proceedings Future
Trends in Automation in Mineral and Metal Processing, 22-24 August, Espoo, Finland, pp 361-366.
Katajainen, E., Multivariable correlation analysis of a solvent extraction process, 1998, M.Sc thesis,
Helsinki University of Technology, Laboratory of Process Control and Automation, Espoo, 128 p.
Kordosky, G.A., Copper solvent extraction, the state of the art, 1992, JOM, 44, pp. 40--45.
Pezoa Alejandro, Panorama de la Hidrometalurgia del Cobre, Resumen de Proyectos SX/EW en Chile,
1999, Minerla Chilena, 218, August, pp. 43-53.
Siemens S.A., Automatizaci6n de una Planta de Electrolfsis de cobre, 1998, Minerla Chilena, 200, pp. 87-
92.
Spitzlei P., Process instrumentation and automation in copper plants, 1994, Congreso Hidrometalurgia del
cobre, Universidad Santa Marfa, Vifia del Mar, Chile, pp. 42--48.

Correspondence on papers published in Minerals Engineering is invited b y e-mail to

bwills @min-eng.com