OPERATING THE PROCESS

INVENTION SUMMARY

The invention relates to the electrowinning of a metal from an electrolyte. The invention covers two embodiments of the electrowinning (EW) of copper, though the invention is not limited to the electrowinning of copper. The embodiments are:

• Electrowinning of copper as part of a solvent extraction-electrowinning (SX-EW)

process.

• Electrowinning of copper in a liberator section of a copper electrorefinery.

Prior art 1 discloses the use of point-of-use power converters to drive current through the anode-cathode gaps of EW tanks locally rather than relying on a central rectifier to produce current flow through a parallel-series arrangement of anodes and cathodes in a multiplicity of tanks.

In conventional practice, there is only one row of anode cathode pairs in an individual tank with electrolyte flow in a direction orthogonal to the plane of the electrodes - a direction of flow which is not conducive to good electrolyte distribution in between anodes and cathodes.

The inventors have realised that by employing point-of-use converters to drive the inter- electrode gaps, it becomes possible to have many rows of anode-cathode pairs in a single enlarged tank or "Jumbo" cell. This was not possible in the conventional arrangement due to parasitic current flowing between rows of electrodes, which leads to power loss and poor process efficiency. In turn, the use of many rows of electrodes in a single tank permits the electrolyte to be circulated through the enlarged tank in a direction in-line with the orientation of the electrodes. Electrolyte flow through the tank is no longer impeded by the electrodes and the refreshing of electrolyte in the inter-electrode gaps no longer depends on turbulence or other mixing techniques e.g. air sparging.

Advantageously, if mixed metal oxide coated titanium mesh anodes (MMOA) are used in place of the conventional lead anodes, very few pillars are required to support the insulated cross members or beams from which the electrodes hang in the electrolysis tanks. The use of MMOAs facilitates the use of very large tanks with all the cost benefits and process improvements that are thereby enabled.

As will be understood, the term "electrolyte" as used herein is intended to refer to an aqueous electrolyte. The invention thus relates to the electrowinning of a metal from an aqueous electrolyte which contains the metal in ionic form.

As will also be understood, the terms "insulating" or "insulated" as herein described are used in the context of an electrical insulator.

BACKGROUND OF THE INVENTION

In a standard electrowinning (EW) tankhouse containing electrolytic cells (number 100 in Figure 1a and 100a, 100b, 100c, 100d, 100e in Figures 1 b and 1c) for the electrolytic recovery of metals including, but not limited to: copper, nickel, gold, silver, cobalt, zinc, chromium and manganese, from an electrolyte 104 containing the metal of interest in an ionic form, by passing an electrical current between electrically positive electrodes (or anodes) 102 and electrically negative electrodes (or cathodes) 103 immersed in the electrolyte 104. A rich electrolyte (RE) is fed into the cell, for example, from the main electrolyte supply manifold 106, through a valve 107 and into an electrolyte feed manifold 101 in the base of the cell. The feed manifold 101 is typically a loop of 75 mm (3") diameter pipe positioned below the electrodes, the pipe has nozzles angled upwards towards the bottom corners of the electrodes. After passing through the cell, the lean electrolyte (LE) exits the cell through an electrolyte overflow 105.

The valves 107 can be closed to stop the electrolyte feed to the individual cell during cell maintenance or cell cleaning.

The applied electrical current causes the metal of interest - here using the example of copper - to be deposited on the permanent cathode surface. In the following description the term tankhouse means an arrangement, wherein at least one cell (or tank) 100 and at least one power source are present in a building or enclosed structure, that is, a house. In a typical configuration an electrolytic tankhouse comprises a plurality of cells.

In the electrowinning of metals a single cell usually contains a single row of n cathodes and (n+1) anodes arranged vertically, in parallel, in the order anode, cathode, anode, cathode, anode etc., Figures 1b and 1c shows an example "cell section" with five such cells 100a-e.

In an EW plant the number of cells in a section will usually be more than five. We use a low number of cells here for illustrative purposes only.

Figure 1 b shows the cells without any bus-bars or electrodes to give a clear representation of the position of the electrolyte feed manifolds 101. The electrolyte feed manifolds are obscured in Figure 1c. Figure 1c shows sixteen anodes 102 and fifteen cathodes 103 in each cell. The low number of electrodes is again for illustrative purposes. In a modern EW plant between sixty and eighty-four cathodes (and 61 to 85 anodes) are typical numbers in a single copper EW cell.

In standard tankhouse practice, power is fed to the cell through an arrangement of busbars. These typically sit on insulating capping boards, which in turn rest on top of the side-walls of the cells. The state-of-the-art is the well-known "double contact" arrangement with electrical connections at both ends of each electrode's hanger bar. For clarity the sketch Figure 1c shows a single contact system with electrical connections at one end of the hanger bars of the electrodes.

In standard tankhouse practice, the electrical current is fed into a busbar at one end of the section of cells, i.e. at the anodic busbar 108a. The current passes through the anodes 102 into the electrolyte 104 of cell 100a. It then passes into the cathodes 103 of the first cell 100a and then into the inter-cell busbar 109a. From the inter cell bus-bar 109a the current passes into the anodes of cell 100b. After passing through cells 100b-e (and through the inter-cell busbars between them) the current is collected in the cathode busbar 108b at the far end of the section of cells.

In the case of copper EW, the EW section is often combined with a solvent extraction (SX) process giving a technology called solvent extraction - electrowinning (SX-EW).

For electrode materials, the active part of the EW anodes 102 are usually either lead-based alloys e.g. a rolled lead-calcium-tin alloy or mixed metal oxide coated titanium anodes (MMOA) as described in prior art 3 and prior art 9.

The permanent cathodes 103 in copper EW are usually stainless steel blades although a few older refineries may still use copper starter-sheet technology. Typically in copper EW the electrolyte 104 contains copper as copper sulfate with sulfuric acid as a supporting electrolyte.

The RE (Rich Electrolyte) solution with a high concentration of copper is sent from the SX section of the plant to the EW section, entering the cells through the feed manifolds 101. After some of the copper has been recovered in EW, the LE (Lean Electrolyte) solution - now with a lower copper concentration - exits the cells at the overflows 105 and is returned to the SX process.

The decrease in copper concentration between the RE and the LE (i.e. the removal of metal at the cathodes in the EW process), is known as "delta copper" or "ACu". For a copper SX-EW plant, ACu is usually in the range of 1 to 5 g/dm ³ (1 to 5 kg/m ³ ). ACu may however be higher or lower than those values, depending on the operating conditions of the individual copper EW plant.

The composition of electrolyte should be homogeneous throughout the cell. This is not always the case, especially in quiescent areas of the cell, e.g. below the electrodes.

In a typical copper EW process the RE enters the cell with a copper concentration of approximately 40 g/dm ³ and the LE exits from the cell with a copper concentration of approximately 38 g/dm ³ . The ACu value is then 2 g/dm ³ (2 kg/m ³ ).

In industrial copper electrowinning, the cathodic current density (/, units A/m ² ) is usually in the range of 200 to 400 A/m ² , with most plants operating in the middle of that range at ca. 300 A/m ² .

The metal is deposited on the active part of each cathode. The active area of each cathode face is usually in the range of between 1 and 1.2 m ² , giving a total active area of 2 to 2.4 m ² per cathode, though the active area may be smaller or greater than those values. The volumetric flow rate V of electrolyte through the cell _T (units of m ³ /hour), is adjusted according to the desired values of ACu and of the current density. In SX-EW, V is set by the requirements of the solvent extraction plant for the required value of ACu.

When using a plurality of cells in conventional electrowinning, it is usual to have several cells in electrical series powered by a common power source or "rectifier" which converts mains ac electricity to dc electricity for use in the EW cell. Using a common rectifier means that the total current passing through each cell is the same.

Figures 1 a-c show a state-of-the-art copper EW cell 100 which contains a plurality of anodes 102, where all anodes are connected electrically in parallel. The cell also contains a plurality of cathodes 103 where all cathodes are connected electrically in parallel.

The voltage across a cell is therefore approximately equal to the voltage that would be experienced between a single anode and a single cathode. The cell voltage of copper EW is usually in the range of 1.7 to 2.3 Volts and depends on:

• the applied current density

• the electrolyte composition (copper and sulfuric acid concentrations)

• the electrolyte temperature

• the anodes used (lead based or mixed metal oxide coated titanium)

• the concentration of cobalt added to the electrolyte (when using lead anodes).

It is difficult to efficiently convert electrical power from a mains ac voltage to a dc voltage of this magnitude. For this reason it is common practice to connect the cells in series so that they all conduct the same current. The voltage across the series chain of cells 100a-e is equal to the sum of all the individual cell voltages. By this means the voltage rating of the rectifier, is elevated and high efficiency can be obtained. Figure 2 shows the general shape of the polarization curve for copper EW. The copper deposition reaction occurs at potentials negative of 0.34 Volts, oxygen occurs at potentials positive of 1.23 Volts, giving a theoretical voltage for the Cu EW reaction of approximately 0.9 Volts. The remainder of the cell voltage is due to the cathodic overpotential for copper deposition, the anodic overpotential for oxygen evolution, the voltage drop across the electrolyte and voltage drops in the hardware, comprising the electrodes, the electrical contacts and the power electronics.

The industrial EW process is operated in the region at the centre of Figure 2. The anodic and cathodic current densities are identified as j _a and j _c respectively. The current includes the Faradaic current for the electrode reactions (copper deposition and oxygen evolution), current used by any side or parasitic reactions, and current inefficiencies such as those attributable to short circuits, and stray or leakage currents.

The maximum current which can be used practically for depositing metal in an EW tankhouse is around one third of the limiting cathodic current y ^' | _im (units A/m ² ).

In the following equations, it is noted that all units in an equation must be consistent, e.g. if the units of area are m ² , then the concentration should be in mol/m ³ . If the units of current density are mA/cm ² then concentrations of mol/cm ³ are appropriate.

Under mass transport controlled conditions, Eqn. 1 shows the diffusion layer thickness δ is inversely proportional to the limiting current density ( | _im ). The smaller the value of δ, the thinner the diffusion layer - the higher limiting current density. y _lim = (z Co) / 5 (Eqn. 1)

Where

• z is the number of electrons involved in the reduction of the metal ions (for Cu in

sulfate based electrolytes, z = 2),

• F is Faraday's constant 96 485 Coulombs/mole (1 Coulomb = 1 ampere second),

• D is the diffusion coefficient of copper ions, Cu ²⁺ (usually expressed in cm ² /s),

• Co is the bulk concentration of the electroactive species of interest, in this case Cu

(units mol/cm ³ ).

Classically, the rotating disc electrode (RDE) method is used in the laboratory to measure the limiting current density, and the Levich equation (Eqn. 2) shows the relationship between the limiting current _im and the angular frequency ω (a rotation rate which is analogous to electrolyte velocity in an industrial cell), v is the kinematic viscosity of the electrolyte (units of stokes where 1 stoke = 1 cm ² /s).

Vlirn (Eqn. 2)

When the RDE's rotation rate ω is increased, the diffusion layer thickness δ decreases (Eqn. 1.) and y ^' _l im increases proportional to the square root of ω (Eqn. 2).

The current density employed in copper electrowinning is usually in the order of one third (1/3) of the limiting current density In copper electrowinning and electrorefining y ^' i _im is usually considered to be in the order of 1 kA/m ² . The value of yi _im depends on the mass transport of copper ions to the cathode surface, which depends on the specific operating conditions of the tankhouse including cell and electrode geometry, electrolyte composition (primarily metal ion concentration), electrolyte temperature and electrolyte circulation (pumping / mixing / convective regime). Modern copper EW tankhouses operate at current densities in the range of 200 to 400 A/m ² . At the upper end of that current density range, the copper deposit tends to be of a lower grade (higher roughness, and more electrolyte inclusions). Copper quality is important when the aim of the EW plant operator is the production of LME A- grade copper cathode for the market. If the target is high current density and good copper cathode quality, then attention must be paid to the transport of electrolyte to the cathode surface.

Krishna and Das (prior art 13) reported that a fourfold improvement in the mass transport coefficient, K, given in Eqn. 3 resulted in a doubling of the limiting current density ( | _im , units A/m ² ). The units of K are usually expressed as cm/s but for consistency we use m/s,

K = vi _im / (z F Co) (Eqn. 3)

As with the Levich equation, the greater the mass transport coefficient, the higher the limiting current density.

The mass in grammes of a metal (m) deposited during electrolysis can be written as Eqn. 4.

m = (I t M _r η) / (z F) (Eqn. 4)

Where

• I is the current in Amps,

• t is time in seconds,

• M _r is the molar mass in g/mol, for copper 63.55 g/mol,

• η is current efficiency (C. E.) in the range 0 to 1 , though usually discussed as a

percentage value.

Charge is current multiplied by time with units of Coulombs, where Coulombs = ampere seconds, though in the EW industry charge is usually expressed in units of ampere hours (Ah), kiloampere hours (kAh) or megaampere hours (MAh).

In standard EW there are two main ways which the electrolyte circulates in an electrowinning cell • the volumetric flux of electrolyte pumped into the cell through the electrolyte feed manifold

• a mixing effect due to bubble lift from oxygen evolved at the anode

From Eqn. 4 the "electrochemical equivalent" of copper can then be calculated as 1.1855 g/Ah (or 1.1855 kg/kAh or 1.1855 tonnes/MAh), giving Equation 5, where time t is measured in hours.

m = 1.1855 I t η (Eqn. 5)

Delta copper (ACu) with units of kg/m ³ (equivalent to g/dm ³ ) is a value of mass per unit volume; when substituted into Eqn. 5 and using the electrochemical equivalent 1.1855 kg/kAh we obtain Eqn. 6a, where V is the electrolyte volume in m ³ needed for a given ACu, current and time.

ACu V = 1.1855 I t η (Eqn. 6a)

Dividing both sides of Eqn. 6a by time gives Eqn. 6b, which shows that current is directly proportional to ACu and the volumetric flow rate V with units of m ³ /hour.

ACu V = 1.1855 I η (Eqn. 6b)

Considering a given volumetric flow rate through the cross section area of the cell (units of m ² ), a value equivalent to a volumetric flux is obtained (q, units m ³ /s m ^~2 or m/s). The direction of the volumetric flux q can be considered as being from the end of the cell where the rich electrolyte enters the feed-manifold, to the end of the cell where the overflow is located. Volumetric flux differs from the mass transfer coefficient K in Eqn. 3, which is a measure of the electrolyte transport at the cathode surface, extracted from the limiting current y ^' | _im .

Equations 1 to 6 show the importance of the supply of electrolyte to the cathode surface. By improving the flow of electrolyte to the cathode, the diffusion layer thickness δ will decrease and y _Nm will be increased, allowing a quality deposit to be formed at a higher current density. In a standard EW cell the promoter of electrolyte circulation in the inter-electrode gap (I EG) are the oxygen bubbles evolved at the anode which rise up along the surface of the anode, lifting the surrounding electrolyte - thereby decreasing somewhat the diffusion layer thickness at the cathode surface. Prior art 14 (Guerra and Shepherd) tells that the bubble dispersion at the anode appears as an inverted wedge along the anode extending all the way to the cathode at the top of the cell and current distribution could therefore be made more uniform by employing tapered anodes in preference to conventional plate anodes. Tapering of lead anodes was found to be impractical due to the mass of metal required to get a significant taper. Tapering of an MMOA structure is more feasible and may be considered in the future.

Prior art 15 tells that at higher current densities the bubble wedge "breaks down higher in the cell, where motion of bubbles towards the centre of the interelectrode spacing is observed and circulating streams of bubbles are established'.

A number of inventors have attempted to improve the flow of electrolyte within conventional electrolysis cells using a single row of electrodes. In prior art 11 Kuhn summarised a number of innovations for metal recovery including air sparging with an example plot of diffusion layer thickness, δ at different heights along an EW cathode surface. At the top of the cathode the value of δ was 0.15mm as oxygen bubbles evolved at the anode enhanced the mixing of the electrolyte. At the lower part of the cathode δ was as great as 0.3 mm where there is less mixing of the electrolyte. With air sparging δ was decreased to as little as 30 μηι (0.03 mm) along the whole height of the cathode.

Prior art 8 teaches how to sparge a mixture of fresh electrolyte and air in between the electrodes, but air-sparging introduces additional equipment into the tankhouse (requiring more maintenance and manpower) and new technical challenges for the operation of the cell; e.g. an increase in the generation of acid mist above the cell which must be controlled. Another approach described in prior art 16 is the use of a parallel flow device inside an electrorefining cell. Fresh electrolyte is introduced directly in front of the cathode. The use of a parallel flow is claimed to allow the cathodic current density to be >400 A/m ² i.e. increased by as much as 50%.

By using a MMO coated mesh anode instead of a lead -based anode, the overpotential for the oxygen evolution reaction can be decreased by 200 to 300 mV, or 10 to 15% of the cell voltage. Additionally, when using mesh anodes, e.g. MMOAs, the electrolyte can pass through the anode giving enhanced mixing compared to solid plate lead anodes.

In an ideal EW cell the same current density will be experienced at each cathode, with the exception of the first and last cathodes in the cell which are located adjacent to the end-anodes. End anodes which have only one active face typically draw less current than other anodes in the cell. Plant operators will aim to keep all electrodes optimally positioned and spaced so that the resistances of each electrode are similar and that current is distributed evenly. Since current always follows the path of least resistance, it is challenging to control the distribution of current within the cell.

Prior art 1 introduced a new approach to the supply of current to the interelectrode gap between the individual anode and cathode faces. A point-of-use (POU) converter is a power supply which in general is a power supply which provides the desired current at the desired voltage close to the point of usage. Typically this type of converter is used when power is required at a relatively low voltage. In a dc system, power is the product of voltage and current. Hence, for a given power, if the voltage is low the current is large. If large currents are supplied from a distant power supply, the voltage drop in the supplying cables can be large or the cables have to be thick (and hence expensive) so as to have a low resistance to the current flow. Hence there are cost and efficiency advantages in using a POU converter with the input power to the converter being supplied at a relatively high voltage and low current. The relatively high current supplied by the POU converter to the load has only a short distance to travel through the cables connecting the output of the POU converter to the load. Typically a POU converter has a DC output and a DC or AC input. Typically a POU converter will employ switched mode technology. The input and output may or may not be galvanically isolated from each other. Cost and application requirements will determine which arrangement is used. As described in the prior art 1 and prior art 3 there are a number of advantages in using POU converters compared with the conventional practice of using a single central rectifier. This includes the ability to limit or entirely switch off the current which flows between anodes and cathodes when a short circuit occurs between them. Short circuits are typically formed when a metal dendrite is formed between the cathode and the anode. Such short circuits can reduce the current efficiency of the ER or EW process and result in burning of the anode or cathode. Operators attempt to identify and remove such short circuits as early as possible in their formation. The use of POU converters facilitates early detection and through current control prevents damage to the electrodes. All the anticipated benefits of using POU converters are applicable to the present invention.

Prior art 4 described a new structure for an anode using ideas from prior art 1 , eliminating the conducting copper hanger bar - a part of the MMOA structure especially susceptible to anodic corrosion and attack by the acid mist in an electrowinning cell.

Electrorefining (ER) is another electrolytic process for purifying a metal - most commonly for the refining of copper - where an impure cast anode is dissolved and a more pure metal is deposited at the cathode. The cast anode is usually around 99% copper.

As the anode dissolves anodically copper (II) ions, Cu ²⁺ _(aq) , historically referred to as "cupric" ions, enter the electrolyte.

At the cathode surface, copper ions are reduced and high purity copper metal (>99.995% copper) is deposited. As copper anodes are refined in copper ER, the current efficiency of copper deposition in the cathode reaction is typically a little lower than the current efficiency of the copper dissolution at the anode. This results in a continuous increase in copper concentration in the ER electrolyte over time.

Those copper ER impurity elements originating from the anode which are more noble than copper remain as solids and collect in the anode slimes at the base of the cell (e.g. silver, gold, selenium and the platinum group metals). Those impurity elements which are less noble than copper dissolve into the electrolyte (e.g. nickel, arsenic, antimony, bismuth and iron). As with copper, the concentration of dissolved impurity elements in the ER tankhouse electrolyte increases with time.

To control the concentration of copper and the concentration of impurity elements in the ER electrolyte, a bleed stream of ER electrolyte is diverted to a purification circuit for decopperising and impurity control. Decopperised electrolyte can be further processed, for example for the recovery nickel, the purified electrolyte is recycled back to the ER cells.

To remove excess copper from the bled ER electrolyte, the electrolyte purification section often has a smaller, secondary electrolytic circuit (see prior art 2), with cells known as "liberator" cells. Copper is liberated from the electrolyte by electrowinning with insoluble anodes that evolve oxygen in sulfate based electrolytes.

The cathodes used in the liberators are either permanent cathodes with stainless steel blades, or copper starter-sheets, or spent / scrap anodes from the ER tankhouse. The anode materials for copper liberator cells are similar to those used in standard copper electrowinning cells, insoluble anodes are employed. These are usually lead-based alloys (e.g. rolled lead-calcium-tin alloy). Alternatively, MMOAs may also be used.

In a liberator EW section, the target is to remove copper from the electrolyte solution as a solid copper cathode deposit. Some copper cathodes produced in a liberator EW circuit may satisfy the LME purity grade and so be sold directly to the market. More commonly the liberator EW cells produce copper cathodes deposited under non-optimal conditions. These contain unacceptably high levels of impurities such as arsenic, and such rejected cathodes are cycled back to the smelter to be melted and re-cast into new anodes for the copper ER process.

In a liberator section the EW cells can be arranged so that the electrolyte passes through one or more cells in cascade. In the case of a copper ER electrolyte bleed stream, the initial copper concentration can be, for example, in the range of 40 to 60 g/dm ³ . As the electrolyte passes through the liberator cells copper is removed from the electrolyte. Copper concentration decreases and sulfuric acid concentration in the electrolyte increases, giving a final copper concentration typically below 10 g/dm ³ .

In an ER liberator section ACu can then be as much as 30 to 50 g/dm ³ , far higher in comparison to the ACu value in the copper EW section in a SX-EW plant, which as stated earlier, is typically in the order of 1 to 5 g/dm ³ .

PRIOR ART

The following prior art is disclosed:

(1) Grant D: Apparatus for use in electrorefining and electrowinning. WO 2012020243 (A1), published Feb. 16, 2012. (2) Barker MH, Virtanen HK, Grant D: System for power control in cells for electrolytic recovery of a metal: World application: WO2013117805 (A1), published Aug. 15, 2013.

(3) Grant D, Nordlund L, Rantala A, Barker MH, Virtanen HK and Schmachtel S: Self

protected anodes and cathodes in electrolytic cell arrangements. Fl 124587 (B), published Oct. 31 , 2014.

(4) Barker MH and Grant D: Anode for a metal electrowinning process. UK patent application GB1518048.2, filed Oct 12, 2015; and PCT patent application No. PCT7GB2016/053164.

(5) Anastasijevic N, Nepper JP and Barker MH: JuCaTec® - Jumbo electrode technology for copper electrowinning: HydroCopper 2005. Proceedings of the III international copper hydrometallurgy workshop: November 23-25, 2005, Santiago, Chile: Eds. Menacho JM, Casas de Prada JM: ISBN 956-19-0492-6: pgs 467-477.

(6) Nieminen V, Barker MH, Virtanen HK: Anode and method of operating an electrolysis cell:

WO2013132157 (A1), published Sep 12, 2013

(7) Ikeda H and Matsubara Y: No3. Tankhouse at Onahama smelter and Refinery: Extractive Metallurgy of Copper. TMS AIME conference 1976: Volume 1 , Chapter 30, pgs. 588-608.

(8) Rigby GD, Stuart AD, Grazier PE: Electrolytic process and apparatus, US Patent

US200401 1664 (A1 ), published Jan 22, 2004.

(9) Sandoval SP, Clayton CJ, Robinson TG, Zanotti CJ, Zanotti MK: Structure For Copper Electrowinning: US2013126341 (A1), Published May 23, 2013.

(10) Siegmund A, Gadia S, Leuprecht G and Stantke P: Replacement of liberator cells - pilot test study at Aurubis Hamburg using the EMEW technology: Proceedings of Copper 2013: Santiago, Chile, 1-4 Dec. 2013: Volume V, Pgs. 275-283.

(11) Kuhn A: Novel electrode systems for dilute metal bearing liquors: Chemistry and Industry:

1978, 1 July, 447-453.

(12) Nyman B, Hutholm SE, Ekman E, et al.: Method and equipment to control separation of a dispersion liquid-liquid extraction: WO03097205 (A1), Published Nov. 27, 2003.

(13) Krishna P.G and Das SC: Enhancement of operating current density in a copper

electrowinning cell: Hydrometallurgy: November 1992, 31(3), 243-255. (14) Guerra E and Shepherd JL: Tapered anodes for copper electrowinning: Proceedings of the Copper-Cobre 2007 International Conference, Toronto, Canada, August 25-30, 2007. Vol. 5 Electrowinning and Electrorefining, Pgs. 13-22.

(15) Papachristodoulou A, Foulkes FR and Smith JW: Bubble characteristics and aerosol formation in electrowinning cells: J. App. Electrochem.: 1985, 15(4), 581-590.

(16) Filzwieser A and Filzwieser I: Method for operating copper electrolysis cells:

WO2009026598 (A2), published Mar. 5, 2009.

PROBLEMS WITH CONVENTIONAL PRACTICE

Problem 1 - delivery of fresh electrolyte to the surface of EW cathodes

The present design of tankhouses for the electrowinning of metals and the layout of the cells contained within those tankhouse have been determined by several requirements:

• The delivery of rich electrolyte to multiple electrolytic cells, said electrolyte containing metal ions to be deposited at multiple cathodes.

• The supply of current to multiple cathodes in multiple cells to carry out the electrochemical reduction of metal ions to deposit solid metal at the cathode surface.

• That the surface of each electrode should be in electrochemical communication only with the nearest (adjacent) neighbour anode surfaces.

Equation 6b shows the simple relationship between current used in a cell (I, in kA), the volumetric flow rate of electrolyte through the cell (V, in m ³ /hour) and the concentration change of copper (ACu, in kg/m ³ ).

The technology used in most EW tankhouses today is a compromise design which gives delivery of fresh electrolyte and power to a large number of cathodes. Delivery of fresh electrolyte to the surfaces of the cathodes is far from optimal. As seen in figures 1a-c, in state-of-the-art electrowinning cells, fresh electrolyte is fed into the cell at one end, for example through electrolyte feed-manifolds 101 which run along the base of the cell 101 underneath the electrodes 102 and 103.

Electrolyte typically exits the cells though overflows 105 located at the other end of the cell from the manifold inlet.

The net-flow of electrolyte through the cell is perpendicular to the orientation of the anodes and cathodes - a non-optimum orientation for delivering fresh electrolyte to the surface of the cathodes.

Attempting to improve the delivery of fresh electrolyte to the cathode surface by increasing the electrolyte flow rate through the cell is not a solution to this problem as it also means a decrease in the ACu which, is not helpful for the SX process.

If it is desirable to increase the current density at which the tankhouse operates, this can be done by improving the flow of electrolyte between the anodes and cathodes. This has been attempted as described for example in prior art 8, 16 and 17.

Since the first Cu EW tankhouses were introduced over 50 years ago, attempts to increase the productivity of electrowinning tankhouses have usually focused on increasing the length of tanks and the number of cathodes contained in each cell, and on increasing the surface area of the cathodes, (see for example, prior art 5 and 6). The fundamental cell design and

arrangement of electrodes in those cells using a single row of electrodes remains, however, unchanged.

Prior art 7 is a rare example from copper electrorefining of a very large cell containing multiple rows of electrodes - and an electrolyte volume equivalent to approximately 20 standard cells of that plant type. The cell-top furniture was suspended from pillars inside the cell instead of from the cell walls, with several parallel rows of electrodes located inside the same cell.

The enhanced electrolyte circulation in large (multiple row) electrolysis cells was not an advantage in prior art 7 because in Cu ER the metal ion concentration is maintained by dissolution of copper anodes replacing the copper removed from the electrolyte at the cathode.

A large "Jumbo" or multi-row cell design has not to our knowledge been tested for copper EW or copper liberator cells, where the additional benefit of improved electrolyte circulation is obtained.

Other cell designs have tackled the issue of optimizing the electrolyte flow in between the anode and cathode such as the use of cylindrical electrodes (see for example prior art 10), but such cell designs are manpower intensive, difficult to automate and challenging to deploy at large scale.

Problem 2 - Non-optimal delivery of electrolyte to the cathodes in liberator cells

With liberator cells, as for standard EW cells, there is a single pass of electrolyte. Volumetric flux is perpendicular to the plane of the cathodes, which is not ideal for the delivery of copper to the cathode surface.

For liberators, there is a different way of feeding electrolyte into the cells. Electrolyte feed can be in cascade. The electrolyte is fed to the first cell in a series of liberator cells and the electrolyte cascades down through subsequent cells. In a liberator section the electrolyte composition changes through the cascade. Copper concentration decreases and sulfuric acid concentration increases. This is in contrast with standard Cu EW cells where manifolds are typically used in every cell so that the electrolyte is more-or-less of similar composition in all cells.

Problem 3 - non-optimal current density in liberator cells

Liberator EW cells are usually operated at a current density which can be tolerated by the cells with the lowest copper concentration for the following reasons:

• In the later stages of the removal of copper in liberator EW cells, when the copper

concentration in the electrolyte is low, if the current density is too high, then an undesired result can be production of a copper powder deposit (or copper sludge) that adheres poorly to the cathode, necessitating periodic cell cleaning to remove the sludge from the base of the cell.

• At lower concentrations of copper in the liberator cell electrolyte, if the current density is too high in the presence of arsenic in the electrolyte, undesired copper arsenate can be formed.

• At the lowest concentrations of copper, if current density is too high, toxic arsine gas (AsH ₃ ) may be formed at the cathode surface which is a health and safety risk for the plant workers.

In prior art solutions for liberator EW circuits - with the commonly used arrangement of cells and a common rectifier - since all the cells are connected electrically in series and carry the same current, the maximum current density which can be used is set by the copper concentration in the lean electrolyte and so the current densities used in liberator EW cells are typically lower than current densities used in standard copper EW cells. By operating at a low current density in all cells (due to the limitation of the current density that can be tolerated by the cells with the lowest copper concentration) the liberator process is rarely operated at optimum conditions. Some liberator cells may be operating at a much lower current density than the copper concentration in the electrolyte would permit. Liberator cells typically have a lower space-time yield than a similarly dimensioned EW cell. A liberator tankhouse with a common rectifier uses more electrical power than would be used in an optimum case and also requires more cells and more electrodes than would be needed than if the process would be run using optimum current densities.

Problem 4 - stray current in large (multi electrode row) cells

Stray current or leakage current is another issue in tankhouses where cells are powered by a common rectifier. A portion of the current provided by the rectifier can escape to a circuit in the earth so that for a given cell this can decrease the current efficiency of the electrolysis in that cell by several percentage points.

If the conventional practice of using parallel-series connected electrodes powered from a central rectifier is used in the context of the Jumbo tank design then there is a risk of current from one row of electrodes passing into the electrodes in neighbouring rows. In this situation the impact of leakage current can be significant, see prior art 7, page 599.

Problem 5 - stripping of thin cathodes in liberator cells

In liberator cells using permanent stainless steel cathodes, there is a minimum cathode deposit thickness required to allow the deposited copper plate to be amenable to automatic stripping by a "full deposit cathode stripping machine" or FDSM. This is important for the cathodes operating at a low current density, which must be given sufficient electrolysis time to allow a copper deposit of sufficient thickness to grow, thus avoiding problems in the cathode stripping machine.

In practice a copper deposit of less than 3mm thickness may cause issues in the FDSM, including bending, warping or snagging of the copper plate, and may result also in damage to the permanent cathode. OBJECTIVE OF THE INVENTION

The invention describes new layouts for electrical connection of electrodes in electrowinning cells, focusing on, but not limited to, the electrowinning of copper. The recent implementation of MMOAs for sulfate based electrowinning applications as described in prior art 9, and the application of point of use power conversion to such anodes as described in prior art 1 & 2, allows every electrode in the EW tankhouse to be powered individually and independently.

In prior art 2 an improvement was described for powering electrowinning cells by splitting the current in a single cell. Here we combine the electrodes of several cells into rows in a single tank and benefit from enhanced flow and distribution of the electrolyte solution.

In a traditional tankhouse design, the anodes in a cell are powered through an electrical contact on the underside of the hanger bar which rests on a common anodic bus-bar.

Prior art 4 described a new design of MMO anode with an integrated point of use power convertor.

The new design of anodes may be connected electrically by means of a cable and connector arrangement which removes in part, or completely the need for inter-cell bus-bars to carry current from one cell to the adjacent cell. This alternative layout of electrical connections of the electrodes gives a new degree of freedom in orienting the electrodes in a cell.

With recent advances in the design of electrowinning anodes (prior art 4) and advances in the way that individual electrodes, anode-cathode pairs, anode-cathode surfaces or individual inter- electrode gaps (lEGs) can be electrically powered (prior art 1-3), we remove the need for much of the usual EW cell top furniture - eliminating the heavy inter-cell busbars and the heavy copper anode hanger bars - making a new arrangement advantageous in terms of a smaller amount of copper required to construct the EW plant.

The new approach to the control of power at the level of the individual electrode opens new possibilities in the layout, arrangement and size of the cells, electrolyte feed manifolds, cell-top furniture and electrodes in an EW tankhouse, gaining additional advantages including optimization of current density by improved electrolyte distribution between the anode and cathode surfaces.

Improved flow of electrolyte gives the possibility to use higher current densities to produce the required quality of copper deposit, i.e. copper which satisfies the industry standards (BS EN 1978: 1998 Copper Cathodes, Cathode grade designation Cu-CATH-1).

By elimination of the heavy inter-cell bus-bars which usually carry power from one cell in a section to the next cell, the electrodes no longer need to be arranged in cells containing a single row of electrodes, instead the electrodes can be arranged in multiple rows in a larger tank as described for copper electrorefining in prior art 6, in positions which optimize the electrolyte flow between the anode-cathode surfaces.

By using a large EW cell with multiple rows of electrodes as seen in Figures 3a-c, the electrolyte feed to the cell and the electrolyte removal - or electrolyte overflow - from the cell can be arranged such that the net flow of electrolyte through a cell is in the direction running between the electrodes.

By orienting the electrodes along the direction of net electrolyte flow through the cell as (instead of perpendicular to the net-flow as is the current practice), the distribution of electrolyte through the cell is enhanced, as is the mass transport of copper to the cathode surface. Enhanced flow of electrolyte to the surfaces of the cathodes thins the diffusion layer δ, which raises the limiting current density for copper deposition (J _Km ). This gives the advantage of raising the maximum current density (j _iim / 3) at which the cell can be operated whilst maintaining an optimum quality copper cathode product.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings included to provide further understanding of the invention and constitute a part of this specification, illustrate embodiments of the invention and together with the description help to explain the principles of the invention. In the drawings:

Figure 1 shows three views of standard electrowinning cells 100.

• Fig. 1a a 3D image of a single cell 100 with sample electrodes in position.

• Fig. 1 b a plan view sketch of five empty cells 100a - 100e with electrolyte feed manifold 101 visible.

• Fig. 1c is a sketch the same layout of cells 100a - 100e as for Fig. 1 b with single- contact bus-bar configuration and a single row of alternating anodes 102 and cathodes 103 in each of the cells (feed manifold 101 now obscured).

Figure 2 shows a sketch of the polarisation curves for the processes occurring in copper electrowinning in sulfate based electrolytes - copper deposition and oxygen evolution. Figure 3 shows three views of a Jumbo electrowinning cell 110.

• Fig. 3a a 3D image of a Jumbo EW cell 110 for five rows of electrodes with sample electrodes in position on beams 111 supported on pillars 112.

• Fig. 3b a plan view sketch of an empty Jumbo EW cell 110 with the electrolyte feed manifolds 101 visible.

• Fig. 3c a plan view sketch as for figure 3b with single-contact bus-bar configuration and five rows of alternating anodes 102 and cathodes 103 in the Jumbo cell (feed manifolds 101 now obscured).

Figure 4 shows the electrode system within the Jumbo EW cell 110. Figure 5 shows the arrangement (in side view) for connecting converters to the cathode current collector bars.

Figure 6 shows a plan view sketch of an alternative connection arrangement which permits independent control of the current in the inter-electrode gaps.

Figure 7 shows how power may be supplied to the point of use convertors.

Figure 8 is a view of a Jumbo EW cell 110 with overflows at both ends of the cell for use with an alternative flow of electrolyte.

Figure 9 shows two views of a Jumbo cell 110 for use as a copper liberator.

• Fig. 9a is a 3D image of a Jumbo liberator cell 110 with five rows of electrodes, with sample electrodes in position. The main electrolyte feed manifold 101 is visible.

• Fig. 9b is a plan view sketch of an empty Jumbo liberator cell 110 with main

electrolyte feed manifold visible.

Figure 10 shows a further arrangement for a Jumbo liberator cell

• Fig. 10a is a 3D image of a Jumbo liberator cell for five rows of electrodes with separator fences 113 positioned underneath the support beams 111.

• Fig. 10b is a 3D sketch of a single separator fence 132.

DETAILED DESCRIPTION OF THE INVENTION

The first embodiment is shown in figures 3a-c It illustrates the use of a Jumbo electrolytic cell 110 for the electrowinning of metals with multiple rows of electrodes (anodes 102 and cathodes 103) powered by point-of-use power convertors. By way of example, the configuration shown is that typically used for the electrowinning of copper in a copper solvent extraction - electrowinning (SX-EW) operation.

The advantages obtained by using a Jumbo electrowinning cell containing multiple rows of electrodes powered by point-of-use power convertors are: • An improved electrolyte circulation between the electrodes (by a changing the direction of net flow of electrolyte by 90° from the standard practice).

• An increased current density in the cell.

• An improved space-time yield (higher productivity per unit volume) of the tankhouse.

• By powering each cathode surface independently, the process will benefit from all cathodes running at optimum current density during normal operation.

In the SX-EW application, the electrolyte composition throughout the cell should be kept uniform. A continuous flow of rich electrolyte (RE) delivered through the electrolyte feed manifolds into the cell means that cathode current density can be kept more or less uniform throughout the cell.

In Figure 3a support beams 111 rest on top of pillars 112 in the cell. At their ends the beams join with the walls of the cell 110.

Figure 3c shows the arrangement of five rows of alternating anodes 102 and cathodes 103 in a single cell 110.

Figure 3b shows electrolyte feed manifolds 101 positioned under each row of electrodes in the cell. The feed manifolds introduce fresh electrolyte into the cell.

Figures 3a-c show arrangements of electrolyte feed manifolds 101 and overflows 105 where the net flow of electrolyte is from underneath the first row of electrodes (on the left side of figures 3a-c) to the overflows 105 at the opposite end of the cell (on the right side of figures 3a-c).

The rate of electrolyte flow into each individual feed manifold 101 is controlled by a valve 1 13 associated with each feed manifold 101. The valve 1 13 may be a butterfly valve, diaphragm valve, needle valve or other suitable valve which allows a flow rate to be set.

In the arrangement shown in Figures 3 a-c the rate of flow of electrolyte through each valve 113 will be determined mainly by the pressure in the supply manifold 106 and the setting of the butterfly valve 113. Alternatively each delivery manifold may be supplied by its own pump which draws electrolyte from the supply manifold 106 and forces it into the delivery manifold 101 in which case the electrolyte flow will depend on pump speed. Whichever arrangement is employed, the rate at which electrolyte flows through each delivery manifold 101 may be controlled in an open-loop or closed-loop manner.

In the open-loop manner, the valves 113 are set at a predetermined opening and the pressure in the supply manifold 106 is maintained at a constant value. Similarly if the individual pump arrangement is used, each pump is set to operate at a speed which delivers the required flow rate of electrolyte to each distribution manifold 101. Typically in such an open-loop system a human operator will monitor the process by various means (for example by examination of the thickness, and quality of copper cathodes produced at each harvest) and if necessary alter the settings of the butterfly valves or pump speeds.

Alternatively the electrolyte flow may be operated in a closed-loop manner. In this arrangement the valves 113 will be fitted with an actuator and the valve will be controlled from a local or central controller. In the case in which individual pumps are used, variable speed drives will be used to drive the pumps. The electrolyte flow to each delivery manifold 101 will be measured by a sensor. One control option is for the output of this sensor to be compared with a demand signal and the error signal used to modify the setting of valve 1 13 or the individual pump speed. A central controller (for example a computer system) will supply the demand signal. Alternatively the sensor signal can be sent to the central controller (for example a computer) and the central controller uses an algorithm to determine the valve setting or the individual pump speed to achieve the desired electrolyte flow.

The concept of closed loop control can be extended beyond the control of electrolyte flow. Sensor of various kinds may be used to provide the central controller with values of such variables as electrolyte composition, cathode currents, anode currents and electrolyte temperature and flow at various points within the tank (not just in the manifolds) and manifold pressures. Thus the control algorithms may adapt in real time to conditions in the plant. The electrolyte flow, electrolyte concentration and electrode currents can be coordinated to produce optimal copper deposition on the cathodes.

The control algorithms may be modified by human intervention, resulting for example, from analysis of the copper cathode at harvest and acceptability of the copper product.

The information available from the sensors may be used to provide a visual display of the state of the process in real time and historically, permitting operators to make adjustments based on their experience.

The operation of the plant is based around several theoretical guiding principles. The relationship between current density, cathode deposit quality and electrolyte flow was described in Eqns. 1 to 6.

For a given cathode 103 at a given location in the Jumbo cell 1 10, the current density that may be applied will depend on the supply of electrolyte to that cathode surface. By controlling the flow rate of electrolyte through each individual manifold 101 , an electrolyte flow pattern is established in the Jumbo cell 110 which is designed to achieve optimum electrolyte delivery to the cathodes giving the benefit of all cathode surfaces operating at their optimum current densities. The tankhouse may be controlled by a computerised control system, as previously explained, which interfaces with the valves 113, flow sensors and the point of use power convertors 108. The control system may use an algorithm that sets the current density applied by each point of use convertor 108 at each individual cathode surface to the optimum value to yield an acceptable cathode (acceptable in terms which include chemical purity of the copper deposit, copper surface finish and adherence of the copper deposit to the stainless steel of the permanent cathode).

The control system may be self-optimising, requiring a learning phase during plant

commissioning for the algorithm to learn the optimum flow conditions and current densities which may be applied. The learning phase can be an iterative process with initial parameters based on previous experience from other plants and analysis of the cathode product produced in the early harvests.

The current density at individual cathode surfaces is maintained by point-of-use power convertors (described in Prior art 2) positioned for example on the anodes (described in Prior Art 4) or adjacent to the anodes - to control cathode currents, anode currents or current in the inter-electrode gaps.

Figure 4 shows the electrode system within the electrolyte containment tank or Jumbo cell 110 containing electrolyte 104. Five rows of anodes and cathodes are shown though it will be understood that the use of any number of rows is possible.

The arrangement in Figure 4 shows current feed from both sides of the anode hanger bars, though current feed from just one side is also possible.

Cathode current collector bars 114 rest on the cell walls 110. Cathode current collector bars 1 15 rest on the beams 111. The hanger bars 116 of the anodes 102 rest on the cathode current collector bars 114 and 115. The anode hanger bars 116 are non-conducting.

The hanger bars 117 of cathodes 103 rest on the cathode current collector bars 114 and 115 and make electrical contact with the collector bars.

As shown in Figure 5, pads under the anode hanger bars 1 16 make contact with the cathode current collector bars 114 and 115 as will be described further on. As described in prior art 4, point of use convertors 118 are located on the anode hanger bars and supply current to the anodes as indicated by arrows 120 in Figure 4.

As shown in Figure 4 the converters 118 draw current from the cathode current collector bars 114 and 115 via the conductive pads under the anode hanger bars 116. Current flow between the anodes and cathodes through the electrolyte is indicated by arrows 121. Current entering the cathode electrode flows to the cathode hanger bars 1 17 and from them to the current collector bars 114 and 115. Current flows from the current collector bars 114 and 1 15 to the point of use convertors 118. Electrolyte flow direction is indicated by arrow 122. In this arrangement the converters 118 supply a controlled amount of current to the anodes. This arrangement does not permit independent control of the current density in the inter-electrode gaps. The connection arrangement must be modified according to Prior art 1 if independent control of every inter-electrode gap current is required. If we designate the voltage of all the cathodes as 0 V then all the cathodes will be at approximately 2 V. At least two control methods are possible. In the first case the converters 118 can be operated so as to keep all anodes at the same voltage this precludes current flow between anodes due to a voltage difference.

A 2nd case is that in which anode currents are defined and the voltage of each anode is allowed to vary slightly. In this 2nd case a flow of current between anodes may result. However the spacing between the anodes will generally be larger than the spacing between anodes 102 and cathodes 103 and hence this current should be small. Insulating strips on the edges of the cathodes 102 will also help to reduce the magnitude of this current.

The arrangement of anodes 102 and cathodes 103 in this figure is conventional - that is all cathodes in a line and all anodes in a line - so that conventional harvesting practices may be employed. The invention permits the anodes and cathodes to be alternated in a line if this is advantageous.

Figure 5 shows the arrangement (in side view) for connecting converters 118 to the cathode current collector bars 115. Contact pads 119 bear on the cathode current collector bars 115. Conductive bolts 123 screw into the conductive pads 119. Cables 124 carry the current from the converters 1 18 to the conductive bolts 123. Anode hanger bars 1 16 are of insulating material. Current is delivered to the anodes through cable or cables 125.

Figure 6 shows the alternative connection arrangement (in plan view) which permits independent control of the current in the inter-electrode gaps between the anodes and cathodes and which is fully explained in Prior art 1. Anode current collector bars 1 15 in the previous diagram (Figure 5) are replaced by insulating support bars 126. Cathode current collector pads 128 rest on this insulating bar 126. The cathode hanger bars 117 rest on the current collector pads 128. The current collector pads 128 are connected to the point-of-use converters 118 by cables 127. Current is delivered from the point of use convertors 118 to the anodes by cable or cables 125.

Figure 7 shows how power may be supplied to the point of use convertors. This example is based on the situation in which the anode current is 600 A for an anode which has a cathode on both sides. In the diagram anodes 102 and cathodes 103 are designated by lines marked with A and C respectively. The arrows 120 and numbers 129 show the direction and magnitude of the currents supplied by the point of use convertors 118. The location of the converter nodes (i.e. where one or two converters 8 are located) is indicated by the black circles 130.

Five rows of electrodes are, the combined output power of the two converters is approximately 1.2kW if the output voltage is approximately 2 V. The input power required by the two converters will be approximately 1.33 kW if the converter efficiency is 90%. Let us suppose that the input power is supplied at 300 V DC. Approximately 45 A of input power will be required at this node. Consider a tank containing 84 cathodes. If the converters were fed from the ends of the tank then the total current supply at 300 V would be approximately 186 A. This would require somewhat thick cables to be laid across the tank. A solution to this problem is to use intermediate feed points whereby the 300 V supply also rises through the pillars 112 and support beams 111 at intermediate points across the tank. Three intermediate points would reduce the maximum cable current required in this example to 47 A.

Figure 8 shows an alternative arrangement for a Jumbo EW cell 110 with two changes from Figure 3b, which may be applied alone or in combination:

• Instead of the standard practice of using a loop shaped manifold 101 , other manifold designs can be used to give optimum electrolyte distribution in this cell. Here single pipe feed manifolds 131 are shown positioned so that they will run directly underneath the rows of electrodes.

• The net flow of electrolyte in the Jumbo cell in figure 8 - as indicated by the arrows 122 - is outwards from the centre manifold 131 underneath the middle row of electrodes towards the electrolyte overflows 105 positioned at both ends of the cell, on the left and right of figure 8.

The second embodiment is shown in Figures 9a and 9b, the use of a Jumbo EW cell in a copper liberator circuit. The Jumbo liberator cell 110 contains two or more rows of alternating anodes 102 and cathodes 103 for decreasing the concentration of metal in the electrolyte 104 bled from a copper electrorefining plant.

In the Jumbo liberator example, the advantages of using a Jumbo cell are the same as for the EW case.

All cathode surfaces will operate at their optimum current densities during normal operation as described in embodiment 1 , the key differences in the liberator circuit configuration (Figure 9a and 9b) are as follows.

• There will be feed of fresh electrolyte (from the ER tankhouse) supplied to the first row of electrodes in the cell. Figure 9a, 9b all show two electrolyte overflows positioned after the final row of electrodes. In the example given just two electrolyte overflows 105 are shown for illustrative purposes. We are not limited to this number of overflows 105, it may be more, or fewer, nor do we limit the positioning of overflows to the corners of the cell.

• As the net flow of electrolyte is from the first row of electrodes to the second row of electrodes to the third row and so forth, it is no longer necessary to position the electrolyte feed manifold 101 underneath the electrodes as in the conventional practice (shown in Fig. 1a).

Instead the feed manifold 101 is positioned alongside the wall of the cell, before the first row of electrodes.

• As electrolyte 104 passes from the first row of electrodes to the last row of electrodes (from left to right in Figures 9a and 9b) the copper concentration will decrease.

The net flow of electrolyte is indicated by the arrows 122.

• In contrast with the SX-EW process, the optimum current density whilst maintaining optimum plating conditions at each individual cathode surface in each row of the cell will not be the same (or similar) throughout the cell.

• Instead, the optimum current density will be highest in the first row of cathodes and anodes and current density will decrease through subsequent rows of electrodes through the length of the cell as the electrolyte is decopperised (copper concentration decreases) in passing each row of electrodes.

Figure 10a shows the option to guide the electrolyte 104 so that it passes from one row of electrodes to the next row of electrodes in the Jumbo liberator cell 110, by placing physical barriers or fences 132 (shown in Figure 10b) in between the rows of electrodes. The barriers 132 are positioned in the cell beneath the electrode support beams 111 , between the support pillars 112 and the floor of the cell.

The direction of electrolyte flow can arbitrarily be called the x-axis (figure 10b). The function of the fences 132 is to change electrolyte flow in the y-axis (along the row of electrodes) and also in the z-axis (from the top of the cell to the bottom of the cell, or vice versa) in a manner analogous to the separation fences used in solvent extraction technology (see for example Prior art 12).

The barriers 132 are pressed up against the support pillars 112 and the base of the cell so that the major part to the electrolyte 104 passing from one row of electrodes to the next row must pass through the barrier arrangement 132 in the desired manner.

As illustrated in Figure 10b, in the case of the barrier 132, the electrolyte 104 flows over the first panel, down between the panels and out from beneath the second panel. This directs the electrolyte exiting from the top of the first row of electrodes down in the cell towards the bottom of the second row of electrodes.

Other configurations of barriers 132 may be used. The barriers 132 may also be partial barriers with slits, blades or louvres to adjust the direction of the flow of electrolyte 104. In a liberator EW circuit with each cathode operating at optimum current density, those cathodes electrolyzed at higher current densities will need to be harvested at shorter intervals than those cathodes operating at lower current densities. This differs from the conventional liberator practice where all cathodes operate at similar current densities and are harvested with the same time intervals.

In the case of liberator circuits using permanent stainless steel cathodes, a minimum thickness of copper deposit will be required for the deposited copper plate to be amenable for automatic stripping by a "full deposit cathode stripping machine" (FDSM) as described earlier.

The anodes are able to communicate with the process control system as per advances described in prior art 1 , 3 and 4.

Advantageously, the information available from measuring systems on the plant will be recorded so that an accurate real-time record of the process which produced each copper cathode deposit (particularly the current density and total charge used) can be available to the operators and any computer control system employed to optimise the plant performance in producing subsequent harvests of copper cathodes.

Knowing the charge passed on each cathode since the previous harvesting allows an estimate of the mass of copper on all individual cathode surfaces to be made in real time, thus allowing the control system to plan the harvesting cycle and avoid the occurrence of any stripping issues at harvesting.