Thedevelopment ofcontinuousfluidized-bed ... · bed ion-exchange system, and anumber ofideas on moving theresinwereproposed andpatented3, 4.These systems aregenerally referred toasmoving-bed

The development of continuous fluidized-bedion exchange in South Africa, and its use inthe recovery of uranium

SYNOPSIS

by A. K. HAINES*. M.Sc. (Chem. Tech.). Ph.D. (Member)

The status of the South African research and development programme on continuous ion exchange, which wasaimed at the technical and economic assessment of that process for the recovery of uranium, is reviewed.

The first phase of this programme, which has now been completed, involved development investigations on variouspilot plants and culminated in the successful operation of a large-scale demonstration plant and the incorporationof the system in a number of new South African uranium plants.

This account highlights the engineering aspects and the design fo the system, its adaptation to process flowsheets,and plans for future development.

SAMEVATTING

Die stand van die Suid-Afrikaanse navorsings- en ontwikkelingsprogram in verband met deurlopende ioonuitruilingwat die tegniese en ekonomiese evaluering van daardie proses vir die herwinning van uraan beoog, word in oenskougeneem.

Die eerste fase van hierdie program, wat nou afgehandel is, het ontwikkelingsondersoeke by verskillende proef-aanlegginge behels en het uitgeloop op die geslaagde bedryf van 'n grootskaalse demonstrasieaanleg en die insluitingvan die stelsel by 'n aantal nuwe uraanaanlegginge in Suid-Afrika.

In hierdie verslag val die klem op die ingenieursaspekte en ontwerp van die stelsel, die aanpassing van die stelselby prosesvloeidiagramme, en die plan ne vir toekomstige ontwikkeling.

Introduction

The unit processes for the recovery of uranium thathave been applied in South Africa by way of the acid-leaching route are well known, and comprise leaching,solid-liquid separation, concentration and purificationof the metal, and precipitation of the product. In thedevelopment of these unit operations, the areas that havereceived most attention up to now are solvent extractionand ion exchange for the concentration and purificationstep. However, as regards the chemistry, more work hasprobably been done on the leaching step.

There have been many developments in theconcentration-purification process, the original approachbeing the precipitation of a uranyl phosphate complexwith copper metal. This approach was found to be un-practical, but it was discovered at about the same timethat uranyl sulphate could be adsorbed onto strong-baseanion-exchange resins. Ion exchange was then intro-duced on a commercial scale using conventional fixed-bedequipment. Since then, the fixed-bed ion-exchangeprocess has been combined with solvent extraction (theBuffiex process!) and the direct solvent extraction ofpregnant solutions2 has been introduced - developmentsthat have led to savings in capital and operating costsand to an improvement in the quality of the product.

The advantages of countercurrent contact wererecognized very soon after the introduction of the fixed-bed ion-exchange system, and a number of ideas onmoving the resin were proposed and patented3, 4. Thesesystems are generally referred to as moving-bed systems,and many variations have been proposed for the treat-ment of water. An important consideration with these

.Formerly National Institute for Metallurgy, Randburg, Trans-vaal; now Sentrachem Limited, Johannesburg.

systems is that, although countercurrent flow can beachieved, the bed of resin remains tightly packed.

The key to savings of real significance in capital andoperating costs lies in the treatment of the concentration-purification process and the solid-liquid separationprocess as an integrated system. In this way, the wholeproblem of uranium concentration and purification canbe considered in a wider context in that modificationscan be introduced that will not only bring about savingsin the concentration-purification step but will allow forsimplification of the solid-liquid separation step,thereby bringing about further savings.

The use of conventional fixed-bed ion exchange andsolvent extraction has placed two major constraints onthe solid-liquid separation system.(1) The clarity of the pregnant solution has to be as

high as possible. In ion exchange the use of apacked bed of resin, whether stationary or moving,leads to the blocking of the interstices if the clarityof the solution is poor. Frequent back-washing willthus be required or, in the case of a moving bed,frequent cycling. In solvent extraction, solventlosses tend to be inversely proportional to theclarity of the solution and represent a significantfraction of the operating costs of the concentration~purification step.

(2) The volume of solution produced per ton of solidstreated is required to be kept to a minimum. Thisis because the cost/scale factors of both fixed ion-exchange plants and solvent-extraction plants aregenerally unfavourable. Furthermore, in the caseof solvent extraction, solvent losses can be regardedas proportional to flow-rate.

The fluidized-bed system is inherently more suitablefor the treatment of solutions of poor clarity in that it

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY JULY 1978 303

allows free passage to fine particles of slime. It thereforeformed the basis of the development programme oncontinuous ion exchange in South Africa, the basic aimof the programme being the relaxation of the previouslymentioned constraints, which would permit the intro-duction and development of alternative routes forsolid-liquid separation.

The obvious answer to the problem of these constraintsis the complete elimination of solid-liquid separation.However, when this was attempted in South Africa, itwas not commercially successful because of majorengineering problems.

Historical Aspects

The magnitude of the potential cost savings resultingfrom the treatment of unclarified solutions has longbeen realized and is widely accepted. There are, ofcourse, cases where the production of solutions of highclarity is sometimes virtually impossible, and in-slimerecovery operations are a necessity as in certain parts ofWyoming in the V.S.A.

The concept of a multiple-stage fluidized-bed columnfor the treatment of unclarified solutions and slimes con-taining uranium was suggested as far back as 1955 byLevin5. His idea, which did not progress beyond labora-tory scale, is basically the same as that which has nowbeen put into operation. It differed in one respect - themethod of resin transport. He suggested shutting off theflow of feed and allowing the resin to drain from tray totray under gravity. This was a good simple idea, but, interms of what is now known, it is unlikely that thismethod of resin transport could have been scaled upsuccessfully.

An alternative fluidized-bed system was investigatedin Australia at about the same time. This was the jigged-bed system described by Arden et al.6 in 1958. However,this technique suffered from a number of practicaloperating problems and was never pursued. A similarsystem was proposed in 1957 in the V.S.A. by Chase7.

In 1961, Cloete and Streat initiated work at ImperialCollege on the basic concepts of their reverse-flow ideafor the transport of resin in fluidized-bed columns. Thisled to the acceptance of a British patent in 1967 with apriority date of October 19628.

Interest in the multiple-stage fluidized-bed contactorwas revived in South Africa in the mid-sixties byLloyd9, who suggested countercurrent cyclone washingsimilar to that described by Arden6, followed by directtreatment of the cyclone overflow by continuousfluidized-bed ion exchange as a simple uranium-recoveryflowsheet. In a subsequent paper10, Lloyd gave indica-tions of the cost savings that could be expected from theintroduction of a slimes process.

In 1968, the National Institute for Metallurgy (NIM)used the work of Cloete and Streat as a starting point inthe development of a continuous ion-exchange systemthat could be applied commercially to the recovery ofmetals in the hydrometallurgical processing of minerals.Improvements in the design of the column and traysand in the control of the resin inventory were patentedby NIM in May 197111 and in June 197212.

The development of fluidized-bed ion exchange with

304 JULY 1978

reverse-flow transport of resin has been pursued by theV.S. Bureau of Mines since the late 1960S13-16, but theirwork has not yet progressed beyond the pilot-plant stage.

Today there is world-wide interest in the applicationof fluidized- bed ion-exchange technology to the recoveryof uranium. Ritcey, Slater, and Lucas17 and Streat andQuassim18 published the results of laboratory studies,and, more recently, workers from Davy Powergasdescribed their development programme19. 2°. PechineyVgine Kuhlman have also been involved in a develop-ment programme, and are understood to have a smallcommercial plant in France that is handling slimes witha solids content of 6 per cent by mass. Alternativefluidized-bed approaches are the Porter system21, whichis being commercially operated by Rossing UraniumLtd, the horizontal tank system suggested by Lloyd22,and the Himsley column23. Although the Himsleysystem is rather ingenious, its operation is complicated.

South African Pro~ramme

As shown in Table I, the overall South African pro-gramme for the development of continuous ion exchangeand its application to the recovery of uranium is, and hasbeen, a collaborative effort between NIM, the ExtractionMetallurgy Division of the Atomic Energy Board (AEB),the uranium industry, and the chemical industry.

This paper discusses the parts of the programmeconcerned with the recovery of uranium under the mainheadings Engineering, Process, and Plant Design.

En~ineerin~The plant that has evolved from the programme of

research is now well known and has been described in anumber of publications24-27. It is essentially a cylindricaltower divided into a number of contacting stages thatare separated by trays for distributing the liquid. Eachstage contains resin, the volume of which is determinedby the requirements of the process.

The feed solution is pumped into the bottom of thetower and flows upwards, fluidizing the resin in eachstage and leaving the tower by a simple overflow launder.During the period of solution flow, the resin beds expandto the height of each stage, the tower is fully flooded withresin, and there is no wastage of volume.

The resin flows countercurrent to the solution. It is fedto the column and moves through it intermittently,control being achieved by the use of synchronous timers.Slugs of resin leaving the bottom stage are collected in aclosed bin, known as the transfer bin, where they areisolated in readiness for transfer to the second column.The transfer of resin from bin to column is done underhydraulic pressure.

At the beginning of the engineering programme, it wasdecided that all the development should be done on arelatively large scale. Vp to that time, only smalllaboratory-scale models had been built, and it was quiteclear that large-scale equipment would be required ifthe engineering aspects of the system were to inspireconfidence. Thus, a hydraulic mock-up with a diameterof 1,8 m was built in 1969.

The objects of the engineering development pro-gramme were as follows.(a) To develop the hydraulic characteristics of the

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MININGAND METALLURGY

Work Responsibility for investi ationWork in

Main topic Detailed projects completed progress Gold- Chemicalor or NIM AEB uranium engng

stopped planned industry industry

Engineering development Hydraulic tests in 1800 mm CIX rig toof the CIX contactor establish parameters for mechamcal design X X

Further simplification of tray design andancillary circuits (valves, pumps, resin- X Xtransfer vessels, etc.)

Instrume:Jr.ation and control circuits forresin movement and process operation X X

Application of CIX togold-recovery flowsheets Details not shown X X X

Application of CIX to therecovery of zinc Details not shown X X X

Application of U nclarified Tests in 600 mm CIX pilot plant on strong-CIX to solutions base resins with nitrate elution X X Xuranium-recovery Tests in 250 mm pilot plant with strong-flowsheets base resin circuit and H.SO. elution X X X

Proving of CIX and strong-base resincircuit in 2500 mm-diameter demonstration X X Xplant

Commissioning of full-scale CIX plant inCCD-CIX-SX circuit* X X X

Slimes Tests by the Chamber of Mines on counter-current cyclone washing followed by direct X Xtreatment of cyclone overflow by CIX

Tests on the Relix process, Involvinguranium recovery by ion-exchange resins X X Xdirectly in the leaching vessels

Trials with slime feed to 600 mm CIX pilotplant X X X

Development work on cyclones and othersand--slime and solid-liquid separation X X Xtechniques

Development work on slime feeds to CIX rigs X X

Proving of integrated circuits for slimefeeds to 2500 mm CIX demonstration plant X X X Xwith heavy resin

General Development of analysers for uranium inresins and solutions for CIX control circuits X X

Laboratory and pilot-plant investigation ofweak-base resin circuit using chloride X Xsolution

Development and evalua- Laboratory and on-plant testing of resinstion of new and improved from manufacturers of commercial resins X Xion-exchange resins forrecovery of uranium Laboratory testing of a high-density ion-

exchange resin for uranium applications X X X

Testing of a high-density resin in the 250mm CIX pilot plant with unclarified solu- X X Xtion as feed

Testing of a high-density resin in CIX rigs !fed with slimes to evaluate hydraulic para-meters, etc. X X X

Proving of a high-density resin in the 2500mm CIX demonstration plant with a slime X X X Xfeed

Studies on the effects of Si, Cl, SO., pH,etc. on the loading and elution of resins X X X X

TABLE I

THE DEVELOPMENT OF OONTINUOUS ION EXO1lANGE IN SOUTH AFRIOA AND ITS APPLIOATION TO THE REOOVERY 0.11' URAN'IUM

*Circuit incorporating countercurrent decantation, continuous ion exchange, and solvent extraction.Continued on next page

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY JULY 1978 30~

Work Responsibility or InvestIgatIOnWork in

Main topic Detailed projects ooIT{>leted progress Gold- Chemicalor or NIM AEB uranium engng

stopped planned industry industry

Development of methods Development of new and improved mathe-for design and costing of matioal models for process design X X X XCIX plants (includingscale-up) Development of techniques for process

design, costing, and flowsheet optimization X X X

Compilation of a general engineering designmanual for CIX X X

Evaluation of scale-up parameters X X X

Radio-active tracer tests in CIX columns X X

Design and costing of ClXplants for specific gold Details not shown X X Xapplications

Design and costing of CIX Laboratory and pilot-plant tests to obtain X X Xplants for specific uranium process design parametersapplications

Process design and costing studies forindividual mining groups and companies m X X XSouth Africa

Compilation of a process design manual anddata source book for uranium applications X X

Eoonomio comparison of CIX circuits withconventional resins and those with high- X X Xdensity resins

Design and costing studies for CIX applica-tions in other countries X X

Development of new and Fundamental work on development of newimproved ion-exchange and improved resins (metal-speoifio resins, X X X Xresins high-density resins, eto.)

Fundamental investigation of kinetics,mechanisms, and poisoning of resins X X X

Review of existing bench-scale proceduresand equipment for testing of resins and X X Xdevelopment of improved techniques

Evaluation of new and im-proved ion-exchange resins Details not shown X Xfor platinum, gold, andco er

TABLE 1':-- (Continued)

pp

£

distribution trays in each stage so that the fluidiza-tion of the resin beds would be stable and that therewould be no bypassing of the resin beds.

(b) To develop a simple control technique for resintransport and so ensure that the stages wouldremain full of resin under all conditions and wouldnever become depleted. Depletion of the stages hadbeen observed in earlier designs and constituted amajor criticism of the column system.

(c) To develop the cycle ofresin flow so that movementof the resin in and out of the column would bestable under all conditions.

(d) To develop control systems that would ensurebalanced conditions when two or more columns werelinked to form a recovery module.

(e) To develop the general dimensions of the systemand so achieve a plant that could readily be adaptedin size to suit any required throughput.

Although it was known at the time that continuous

306 JULY 1978

ion exchange was workable as a process, it was clear thatthese five points were vitally important if what wasessentially a laboratory concept was to form the basis ofa stable operating plant.

Design of the Distribution Tray

It is generally recognized that fluidized-bed systems,because the particles are in constant motion, can bescaled to much larger diameters than packed-bedsystems without the danger of maldistribution of thefluid. This, of course, assumes that the bed is evenlyfluidized, which is largely dependent on the pressuredrop across the flow-distribution system, whether it isa system employing a perforated plate or a pipe mani-fold. An important aspect of the column design offluidized-bed ion-exchange plant is that the fluid isredistributed at each stage so that the flow at the top orbarren end of the column (the most critical) is almostperfectly uniform. Mass transfer is therefore good, and

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY

1 2 4 5 6

95,0 97,5 100,0 100,0 101,095,0 99,5 100,0 100,0 101,093,5 99,5 100,0 100,0 100,097,5 99,5 100,0 100,0 100,0

101,0 101,0 100,0 99,5 99,097,5 99,5 99,0 99,0 99,093,5 99,5 99,5 99,0 99,097,5 97,5 96,5 99,0 99,099,5 99,5 99,5 100,0 99,0

103,0 99,0 99,0 100,0 94,5103,0 99,0 98,5 100,0 94,593,0 99,0 94,5 100,0 94,397,0 103,0 99,5 100,0 93,798,5 100,0 99,5 100,0 95,098,5 100,0 96,5 100,0 97,098,5 100,0 97,5 97,5 98,093,5 99,0 99,5 96,5 99,094,5 98,0 99,5 97,5 99,095,5 98,5 97,5 97,5 97,097,5 99,0 96,5 96,5 98,099,5 100,0 97,5 96,5 97,097,5 100,0 99,5 94,5 99,093,5 99,5 100,0 99,5 99,093,5 103,5 100,0 99,5 99,0930 99,5 100,0 99,5 990

Cycle

2---

1 100,0 100,02 89,2 90,13 88,2 94,04 78,4 97,05 78,4 91,16 70,6 89,17 70,6 87,189 65,7 84,2

10 58,8 85,111 66,7 82,312 67,4 88,113 53,9 65,314 51,0 71,315 57,8 79,2161718 60,8 68,219 60,8 82,22021 68,6 64,422 68,6 64,422232425 63,7 89,1

4 5 6

100,0 100,0 100,090,1 90,0 78,487,1 80,0 83,381,2 84,0 88,287,1 88,0 87,384,2 85,0 86,388,1 85,0 83,3

80,2 88,0 78,475,2 85,0 80,476,2 80,0 84,369,3 86,0 86,371,3 80,0 70,669,3 72,0 78,473,3 77,1 84,3

69,3 70,0 93,167,3 69,0 88,2

67,3 70,0 88,267,3 70,0 88,2

63,4 70,0 88,2

Cycle

123456789

10111213141516171819202122232425

TABLE nEFFECT OF CYCLING ON RESIN INVENTORY - WITH CAPSFlow.rate of solution 1,9 l/minFlow-rate of resin 0,98 l/minFractional volume of resin transferred 0,9

Resin inventory in each stage at end of cycle( %of volume at start of run)

Stages

TABLE III

EFFECT OF CYCLING ON RESIN INVENTORY - WITHOUT CAPSFlow-rate of solution 1,9 l/minFlow-rate of resin 0,9 IjminFractional volume of resin transferred 0,9

Resin inventory in each stage at end of cycle( % of volume at start of run)

Stages

The resin inventory for runs 8,16,17,20,22,23, and 24 was not taken.

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY JULY 1978 317

resin losses due to surges are minimized. In addition toconsiderations of pressure drop, the size of the hole, andits pitch and fabrication must be taken into account. Aworkable design has been developed, as is demonstratedby the successful operation of a column 4,25 m in dia-meter. Indeed, the commercial plants of this designappear to be more stable than the various pilot plantsthat have been operated.

Control of Resin Hold-up in the Stages

Depletion of resin in the stages has been eliminatedby the development of a simple cap that is placed overeach perforation of the bottom tray and is arranged togive a pressure drop similar to that across the rest of thestages (Fig. 1).

The cap works on the principle that, during a reversalof flow, the resin immediately leaves all the stages thatare separated by perforated plates and enters the stagebelow. However, the caps force the resin from the bottomstage to move by a longer path, delaying it before itdrops into the conical bottom. Thus, at the end of thereverse-flow period, less resin has been transferred fromthe stage above the bottom tray than from the otherstages, and the resin in that stage is above the equili-brium volume established during the previous solution-flow period. However, this accumulated resin is re-distributed over the column once the flow of solutionrecommences, and thus the hold-up in each stage ismaintained.

Tables Il and III compare two runs on a laboratory

'" *$-(

------

------

Fig. I-A typical cap used in the bottom stage

308 JULY 1978

v

column (100 mm in diameter) containing seven stages.One run was done with caps on the bottom stage, andthe other with a simple perforated plate, all the otherstages being separated by perforated plates. The effectof the caps is clearly shown. The concept of the cap hasbeen patentedll and has proved very successful incommercial operation.

Resin-transport CycleThe concept of a reversal of flow to achieve interstage

transport of the resin is now generally well known. Inthe early stages of development, it emerged that thetransport of resin through the trays is predictable andconsistent. These two points are particularly importantand have been put to use in the establishment of theoverall cycle of the column. This cycle comprises fivesteps and can be summarized as follows.

(i) Forward-flow period. Solution flows up through thecolumn, thereby fluidizing the beds of resin con-tained in each stage.

(ii) Settling period. The flow of solution is shut-off, andthe resin is allowed to settle to the bottom of eachstage so that the transfer of resin in a dense stateis ensured during the following period, and theback-flow of solution between stages is minimized.

(iii) Reverse-flow period. A known volume of solution isallowed to drain from the column by way of thetransfer bin. This drainage causes a known volumeof resin to transfer between the stages, resin fromthe bottom stage moving into the transfer bin.

(iv) Delay period. Solution continues to circulate throughthe bottom of the column and the transfer bin. Thiscirculation ensures that all the resin remaining in thebottom of the tower after the reverse-flow period ismoved through to the transfer bin.

(v) Resin-transfer period. The transfer bin is isolatedfrom the column, and the slug of resin transferredinto it is lifted to the top of the second columnunder hydraulic pressure. This transfer takes placeduring the forward-flow period.

In addition to ensuring stable movement of resin, thissequence ensures that no valves close on the resin, andthat the movement of resin is gentle, thus keepingattrition losses to a minimum.

General Plant DimensionsThe general dimensions of ion-exchange plants need

to be considered separately for each particular case, andlittle can be discussed here. The most important aspectin establishing the general shape and dimensions of thecolumn is the flared top. This flaring reduces surgingduring the introduction of the resin slugs, and so preventsentrainment of resin in the overflow.

ProcessIt is clear that the introduction of continuous ion

exchange to uranium extraction has made no differenceat this stage to the chemistry of the ion-exchangeprocess. The object of the process-development work thatwas conducted together with the engineering programmewas to establish the feasibility of continuous ion exchangeon a scale larger than that of the laboratory, and toevaluate its long-term operating performance withrespect to various points such as plant reliability,

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY

process efficiency, treatment of unclarified feeds, resinattrition, resin poisoning, and reagent usage.

As mentioned earlier, the main object in the develop-ment of continuous ion exchange was to relax thecomtraints placed on the solid-liquid separation plant.Th.', only significant work has been carried out onso-called unclarified solutiom, which contain about 300to 500 p.p.m. of suspended solids. Only preliminarywork has been conducted on slimes, and investigation,>are still proceeding. Flowsheets for slimes are thereforenot discussed in this paper.

Three basic flowsheets have been investigated atvarious scales of operation:(1) a strong-base resin circuit with nitrate elution,(2) a weak-base circuit with nitrate elution, and(3) the Bufflex or Eluex circuit, which is a combination

of ion exchange and solvent extraction.

A fourth flowsheet - a weak-base resin circuit withcWoride elution - is being studied in the laboratory.

BARRENELUTEDRESIN

,

I II

I

I

IX

LOADING

IX

ELUTION

FEED

I

I I1LOADED RESIN

Strong-base Resin with Nitrate ElutionThis is the flowsheet that was used in South Africa

before the advent of solvent-extraction technology. Itsadaptation to continuous ion exchange is depicted inFig. 2. The ion-exchange section of this flowsheet wastested during the campaign at Hartebeestfontein on apilot plant of 600 mm diameter. The results have beenpublished elsewhere26, but it is worth recording the mainconclusions here. The object of the test run was tocompare the capacity of the continuous plant with thatof the main plant operating under similar conditions.The conditions set were from 0,002 to 0,003 g of U3Osper litre of barren solution, and from 8 to 10 g of U3Osper litre of concentrated eluate. These conditions wereachieved with a flow-rate of pregnant solution of5,9 m3fh per cubic metre of wet-settled resin. Thecomparative figure for the main plant was 2,6 m3fh percubic metre of wet-settled resin, indicating that thecontinuous plant has more than twice the capacity of anequivalent fixed-bed plant.

FRESH ELUATEI

HNO3

,MAKE -UP

LIME AMMONIA

FeREMOVAL

ADUPRECIP.

CONC. ELUATE IRONHYDROXIDE

ADU

Fig.2-Flowsheet using a strong-base resin and nitrate elution

SINGLE COLUMN

CONC. ELUATE

~ ,I II II r , II I II I I

I

BARREN

IXLOADING

IXSCRUBBING

FEED

I IL J

RESIN

AMMONIA AMMONIA

~IX

ELUTIONIX

REGENERATIONADU

PRECIP.

I .I II IL__- -~~S~~--_J

FRESH ELUATE t NITRATEMAKE-UP

ADU

Fig. 3-Flowsheet using a weak-base resin and nitrate elution

JULY 1978 309.)OI,lRNAL OF THE $OI,lTH AFRICAN IN$TITUH OF MINING AND METALLURGY

Main plant Pilot plantConstituent using using Solvent

IRA-400 XE-299 extraction

,°8, % 92,42 94,6 96,03

.' % 3,20 3,400., % 1,52 0,06 0,03l,p.p.m. 1,000 <3 20s, p.p.m. 90 <3 <10,p.p.m. 13,9 6 0,7,p.p.m. 3,600 - 150,p.p.m. <1 <0,3 <1,p.p.m. <3 4 < 10,p.p.m. 20 < 10 3,p.p.m. 30 12 0,3,p.p.m. 2,000 300 140

g, p.p.m. 50 60 10n,p.p.m. 200 30 50, p.p.m. 16 50 4i,p.p.m. <3 5 <1

,p.p.m. <3 6 <3,p.p.m. 2 <3 2,p.p.m. <3 <3 <3

, p.p.m. <1 <1 1, .m. < 10 <30 < 10

NoJ MAK

----- ------------. -----"\

r1,---- ---------1 1I, t 11 1

1IX IX 1 ADU1

ELUTION RESULPHATION I PRECIP.1

t 1 1J~ 1 J1 11 1_I 1.__- ------..

\

ADU

SULPHURICACID

,

SX ADUSTRIPPING PRECIP.

....t

I

TABLE IVADU ANALYSIS - WITHOUT NITRATE RECOVERY

uSOSiAABCaCdCoCrCuFeMMMNPbSnTiVZn pp

BARREN 1I I

II

IXLOADING

IXELUTION

II II I1----------..

FEED RESIN CONC. ELUATE

Fig. 4-Flowsheet for the Bufflex process

BARREN 1-------11

IXLOADING

~ 11L-------RESINFEED

Weak-base Resin with Nitrate ElutionThe development of this flowsheet formed the major

part of the campaign in the 600 mm-diameter pilotplant26. The campaign had as its object the establish-ment of a flowsheet that would give an ammoniumdiuranate (ADD) product of equivalent purity to thatproduced by the solvent-extraction process. The basicflowsheet is given in Fig. 3.

The object was almost achieved, as can be observedfrom Table IV, but only with the introduction of adisplacement scrubbing system26. The use of thistechnique, however, led to an unacceptably high con-sumption of nitrate. Furthermore, when the nitrate wasrecovered by the use of ammonia, the product was foundto contain unacceptably high levels of silica, which wasdue to the partial removal of silica from the resin by theammonia. In view ofthese problems, the use of weak-baseresins is not recommended at this stage.

In other respects, the weak-base resin is most accept-able. Good loading and elution efficiencies were achieved,although at a reduced flow-rate as a result of the slower

SULPHURIC ACIDMAKE-UP

.u...............

t

AMMONIA

!

AMMONIA

SX

LOADING

... u u u. u u

SOLVENT STRIP. LIQUOR ADU

Cl

E-UP

MONIA

Fig. S-Flowsheet using a weak-base resin and chloride elution

~10 JULY me ,",OURNALOFTHE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY

loading kinetics of the macroporous resins that were used.A factor observed during this campaign, and one that

had a significant effect on the testwork, was the un-usually high rate of fouling of the macroporous weak-base resins by silica. There is still some controversy asto whether the high rate of fouling is due to the func-tionality of the resin or due to its structure. Recenttestwork seems to indicate that the more importantparameter in the rate of silica fouling in pregnantsolutions of uranium is structure.

The Bufflex Flowsheet

Owing to the difficulties in the adaptation of weak-baseresins to the extraction of uranium (mainly economic),the development programme reverted to the use ofstrong-base resin but modified to the Buffiex system!, inwhich the ion-exchange plant upgrades the pregnantsolution, making it more acceptable economically forsolvent-extraction processing. The basic flowsheet isgiven in Fig. 4. Most of the development of this circuit,which combines two processes, was done on a demonstra-tion plant of 2500 mm diameter. This work forms thesubject of another paper28 and will therefore not bediscussed here.

Weak-base Resin with Chloride Elution

This flowsheet (Fig. 5), which is used at two D.S.uranium plants29. 3°, is being evaluated on a laboratoryscale. This flowsheet has possibilities, but there are anumber of considerations to be taken into account par-ticularly with respect to the consumption and cost ofreagents, the quality of the ADD product, losses ofchloride to the main circuit, and materials of construc-tion. It does, however, offer the potential of a pureproduct in a single step.

Plant Desi~n

Once a basic flowsheet has been chosen, the next stepis the establishment of the design and size of plant thatwill give the desired recovery and eluate concentration.The basic mechanical components have already beendiscussed, and this section is confined to a discussion ofthe current approach to the establishment of basicprocess conditions in relation to the loading column. Theelution column can be designed similarly, but procedureshave not reached as advanced a stage as for the loadingsection.

In the process design the variables that need to befixed are(a) the diameter of the column,(b) the height of each stage,(c) the resin inventory in each stage,(d) the number of stages, and(e) the flow-rate of the resin.

Diameter of the Column

The determination of the diameter is based onhydraulic considerations, and involves a determinationof the expansion characteristics of the resin underconsideration.

Experience has shown that the upper limit for bedexpansion should not be greater than 3 for resins ofstandard column grade. If the bed expansion is greater

than this, losses of fines become excessive. With resinsof the grade used in resin-in-pulp processes, higherexpansion is possible. Because the expansion of a bedresin depends on the loading of uranium, the expansionwill be greatest in the top stage, and, for this reason,tests on expansion should be done with eluted resin inbarren solution. The diameter of the column is estimatedfrom the superficial velocity, which gives the desiredbed expansion for the top stage. A velocity of 18 mfhwas used in the design of the demonstration plant28.

Height of Stage

Although the height of the stage is considered to be anindependent variable in the design calculations and isfixed at some practical value, it is possible to optimizeits value. For large columns, in which access to thestages could be necessary, a minimum stage height ofI m is recommended.

Number of Stages

The number of stages required for any recovery isdetermined by the following method.(I) The inventory of resin in each stage is estimated

from the height of the stage, the diameter of thecolumn, and the bed expansion of the eluted resin,

(2) The concentration profiles are calculated for solutionand resin.

(3) The inventory of resin in each stage is adjusted toaccount for the dependence of bed expansion onresin loading.

(4) The profile is recalculated until convergence isreached.

The calculation procedure is numerical, and is basedon a knowledge of the kinetics of mass transfer, theequilibrium between solution and resin, and the periodicoperation of the column. Models are used to describe theequilibrium and the kinetics, and all that are requiredfor a design are laboratory-determined data on theequilibrium and kinetics.

It is not the intention to discuss here the modellingof the column because the models are still under develop-ment and details will be published at a later date. How-ever, a comparison between the results of design simula-tions and the operating data is worth while.

Fig. 6 compares measured and predicted concentrationprofiles for the demonstration plant (diameter 2500 mm)and the pilot plant (diameter 250 mm). These two plantswere operated in parallel at the same time and under theoperating conditions shown in Table V. The conditionswere exactly the same for both plants, except that thevolume of resin transferred out of the pilot column percycle was inadvertently set at 18 litres instead of thescaled-down value of 16,5 litres. This naturally led to thepilot unit having a more favourable profile and a lowerresin loading than the demonstration plant, and it thusgave a better extraction. The important point, however,is that the model used was capable of predicting bothprofiles from the same laboratory data and that thepredictions in both cases compared favourably with themeasured profiles.

Flow-rate of Resin

The flow-rate of resin is determined by materialbalance once the required recovery and the maximum

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY JULY 1978 311

PILOT PLANT 40 DEMONSTRATION PLANTz0I-<{ 30a::

I-z-w"couo 20z M0::::>uo>

z 10enwa::

zPILOT PLANT 0 1,0 DEMONSTRATION PLANT

I-<{a::I-z-W, 0 Iu co

'zo0 C')

u::::>

0>z

0,010I-::::>...J

0en

0,001

2 3 4 5 6 BARREN PREGNANT I 2 3 4 5 6 BARREN

STAGE STAGE

loading for the resin in equilibrium with the pregnantsolution are known. Two methods for the control ofresin flow are possible:(a) variation of the volume transferred per cycle for a

fixed cycle time, and(b) variation of the cycle time for a fixed volume trans-

ferred per cycle.In general, it is preferable to tramfer a volume as

close to the stage hold-up as possible. The maximumpossible mass transfer performance is achieved when theratio of resin transferred to stage hold-up is 1. In practice,the resin loading is set at about 75 per cent of the equili-brium value.

The Elution ColumnThe detailed modelling of the elution column hasnot

z0I-<{a::1--z'W cou OC')O,I

z::::>0uo>

z0I-::::>

...J

0en

1,0

0,01

0,001

PREGNANT I

yet been completed, and the present approach to itsdesign is therefore based on simple residence time de-rived from practical experience. The time required forcomplete elution is determined in the laboratory by theuse of a single fluidized-bed stage similar to that used inthe determination of the adsorption kinetics.

In the determination of the size of the elution column,the following factors are taken into account:(1) if possible, the elution column is made the same

height as the adsorption column so that constructionand operation are simplified,

(2) a minimum of five stages is recommended to reduceby passing, and

(3) an expansion of the resin bed of not less than 1,2is used for stability.

z0I-<{a::1--z'~ 0: 20z::::>00>U

Zenwa::

40

30

10

0LOADED I

RESIN

312 JULY 1978

Fig. 6-Cornparison of measured and predicted concentration profiles

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY

Fig. 7-General view of the first full-scale commercial continuous-ion-exchange p:ant in South Africa

TABLE V

OPERATING DATA FOR TEST CAMPAIGN ON THE DETERMINATIONOF SCALE-UP PARAMETERS

DemonstrationplantColumn Pilot plant

Diameter, mStage height, mForward-flow period, minSettling period, minReverse-flow period, minDelay period, minTransfer period, minFlow-rate of solution, l(minV 01. of resin transfer, litreResin

0,250,93

180I45

201418

DuoliteAlOI DU

2,51,0

180I45

2014241650Duolite

AlOI DU

The design models used in the evaluation of thenumber of contacting stages are based on certainassumptions pertaining to the mixing patterns of boththe resin and the solution phases. It is policy to con-tinually update and improve these models, and themeasurement of the mixing patterns by means of radio-active tracers has been undertaken. These tests haveyielded interesting results and will be published at alater stage. In general, however, they indicate that theassumptions used in the models are conservative andthat segregation of the resin is insignificant, which is ofconsiderable importance in the scale-up of a continuousion-exchange system.

Conclusion

The initial aim of the programme - the developmentof a commercially acceptable plant that would relax theconstraints placed on the solid-liquid separation process

- has been achieved. This is demonstrated by therecent successful operation of a full-scale commercialcontinuous ion-exchange plant (Fig. 7). In this applica-tion, the ion-exchange system has been successfullyincorporated in a Buffiex process flowsheet, which is nowregarded, technically and economically, as the mostadvanced process for the recovery of uranium in SouthAfrica.

The Future

Although the initial object has been achieved, theresearch programme has gone only part of the way. Upto the present, relatively clear solutions (about 300p.p.m. of suspended solids) have been handled, and it isobvious that still greater benefits could be obtained ifthe system could be adapted to feed solutions containingmuch higher levels of solids.

Some preliminary work was carried out in 1974 in the600-mm-diameter pilot plant in which slimes with asolids content of up to 10 per cent were treated success-fully. This limit was determined by the denisty of thecommercially available resins. Attempts have been madeto produce heavy ion-exchange materials1°, 31, but the

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY JULY 1978 313

heterogeneous materials that were synthesized have notachieved commercial success owing to problems ofchemical and physical instability. An entirely new con-cept is under development, which involves a homo-geneous organic ion-exchange resin that has beenproduced with sufficient density to allow for the treat-ment of feeds with a solids content of 20 per cent andthat has kinetic and equilibrium properties similar to theresins commercially available. This resin should open upa whole new generation of process flowsheets.

In addition to the development of flowsheets forheavy resins and slimes, work is being undertaken onthe engineering of the system and the design procedures.

Materials of construction have not been discussedbecause the optimum choice of materials has not formedpart of the development programme. With the presentflowsheet, the two sections of the plant can be separatedwith respect to materials of construction: the loadingsection, which has to be designed to treat large volumesof very dilute solutions of sulphuric acid (pH 1,5 to 2)at ambient temperature and, in certain instances (e.g., inapplications in the Orange Free State), solutions con-taining significant quantities of chloride; and the elutionsection, which has to handle, at temperatures of up to50 °C, 10 to 15 per cent solutions of sulphuric acid thatcontain traces of solvent.

The materials of construction recommended at presentare 316L stainless steel for the loading section and fibre-glass-reinforced plastic (GRP) for the elution section.Although successful in operation, these materials have a.number of disadvantages including the following.(a) Both are expensive.(b) Stainless steel cannot be used in the presence of

high levels of chloride.(c) There is a lack of expertise and experience in the

fabrication of very large GRP vessels in SouthAfrica.

It is therefore essential that experience in GRPfabrication should be built up rapidly in South Africa.Alternatively, modifications should be made to thedesign of the distribution trays to allow for the use ofrubber-lined mild steel, including solvent-resistant types.This is probably still the cheapest material of construc-tion for vessels that need to be resistant to sulphuricacid.

The following are further points worthy of considera-tion:(1) simplification of the valving and pumping systems,(2) development of washing techniques for the resin, and(3) development of an improved design of transfer bin.

Acknowled~ementsThe contribution of all those involved in the develop-

ment programme is gratefully acknowledged. The pro-gramme has been in progress for over eight years, andmany members of staff of NIM, AEB, and industry havecontributed to the work. To single out any names wouldtherefore be very difficult. However, particular acknow-ledgement is made of the unfailing assistance and co-operation of the managements and staffs of the mineson which the pilot-plant testwork was carried out.

The permission granted by NIM, AEB, and theuranium industry (Nuclear Fuels Corporation of S.A.

314 JULY 1978

Ltd) to publishacknowledged.

paper is also gratefullythis

References

1. FAURE, A., et al. The prodllction of high-purity uranium ata South African gold mine. J. S. Afr. Inst. Min. Metall.,vo!. 66. 1966. pp. 319-341.

2. MEYBURGH, B. G. The design, erection and operation of aPurlex plant at Buffelsfontein Gold Mining Company,Ltd. J. S. Afr. Inst. Min. Metall., vo!. 70. 1970. pp. 55-66.

3. ATMORE, M. G. M., and PORTER, R. R. Ion-exchangeprocess S.A. Patent Application 55/2526. March 1955.

4. ATMORE, M. G. M. Apparatus for contacting liquids withsolids. S.A. Patent Application 55/2643. September 1955.

5. LEvIN, J. A suggested method for semi-continuous counter-current ion-exchange operation, suitable for application toclear solutions or unfiltered pulps. Government MetallurgicalLaboratory Progress Report, Leaching no. 198. Jan. 1955.

6. ARDEN, T. V., et al. Extraction of uranium from acid leachpulps by jigged-bed ion.exchange. Proceedings of 2nd U.N.International Conference on Peaceful Uses of Atomic Energy,vo!. 3. 1958. pp. 396-414.

7. CHASE, C. K. Continuous countercurrent ion.exchange inhydroemtallurgical separators. Min. Engng, N.Y., vo!. 9,no. 9. 1957. pp. 1001-1003.

8. CLOETE, F. L. D., and STREAT, M. Improvements in andrelating to solid-fluid contacting. Brit. Pat. 1 070 251.Acc. Jun. 1967.

9. LLOYD, P. J. The hydrometallurgical recovery of metalsfrom pulps by countercurrent cyclone washing and resin-in-pulp ion-exchange. Proceedings of a Symposium onChemical Engineering in Mineral Processing, Johannesburg,1969. South African Institution of Chemical Engineers.Apr. 1969. (Unpublished).

10. LLOYD, P. J. Unit operations of solid-liquid separatIOn,product concentration and recovery in hydrometallurgicalprocessing systems. Proceedings of an International Sym-posium on Advances in Extractive Metallurgy and Refining,London, 1972. London, Institution of Mining and Metal-lurgy, 1972.

11. CLOETE,F. L. D., and HAINES, A. K. Contacting liquids anQ.solids in countercurrent. S.A. Pat. 71/3139. Acc. 14th May,1971.

12. HAINES, A. K. Equipment for contacting liquids and solids.S.A. Pat. 72/4328. Acc. 20th Jun., 1972.

13. GEORGE, D. R., Ross, J. R., and PRATER, J. D. By-producturanium recovered with new ion-exchange techniques.Min. Engng, N.Y., vo!. 20, no. 1. 1968. pp. 73-77.

14. ROSENBAUM,J. B., and GEORGE, D. R. Cost reductions 111ion-exchange processing of uranium ores. The recovery oJuranium. Sao Paulo, International Atomic Energy Agency,1970. pp. 297-309.

15. ROSENBAUM, J. B., and Ross, J. R. A counterCUrI'elHcolumn for fluid-bed ion-exchange of uranium ore slurriesProceedings International Symposium on Hydrometallurgy.Chicago, AIME, 1973. pp. 535-551.

16. TRAuT, D. E., NICHOLS, I. L., and SEIDEL, D. C. Designcriteria for uranium ion-exchange in a fluidized system.Trans. Soc. Min. Engrs, AIME, vo!. 260. 1976. pp. 24-28.

17. RITCEY, G. M., SLATER, M. J., and LucAs, B. H. A. Acomparison of the processing and economics of uraniumrecovery from leach slurries by continuous ion-exchange orsolvent extraction. Proceedings International Symposium onHydrometallurgy. Chicago, AIME, 1973. pp. 419.474.

18. STREAT,M., and QUASSIM,R. Y. Recovery of uranium fromunclarified liquors by ion-exchange. Ibid., pp. 475-496.

19. NADEN, D., WILLEY, G., and NEWRICH, G. M. The use offluid bed ion-exchange to reduce uranium recovery costs.1O5th Annual Meeting AIME, Las Vegas. Feb. 1976.

20. NADEN, D., and WILLEY, G. Development of a fluid bedweak base ion-exchange process for the recovery of uranium.The theory and practice of ion exchange. Cambridge, Societyof Chemical Industry, Ju!. 1976.

21. PORTER, R. R., and ROSSING URANIUM LTD. Continuousion-exchange process and apparatus. S. A. Pat. 71/8632.Dec. 1971.

22. LLOYD, P. J., and WHITE, D. A. Contacting liquids withsolids in fluidized beds with countercurrent flow. S.A. Pat.App!. 71/8233, Dec. 1971.

23. HIMSLEY, A., and FARKAS, E. J. Operating and designdetails of a truly continuous ion-exchange system. Thetheory and practice of ion exchange. Cambridge, Society ofChemical Industry, Ju!. 1976.

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY

24. HAINES, A. K., et al. The recovery of zinc from pickleliquors using ion-exchange. J. S. Afr. Inst. Min. Metall.,vol. 74. 1973. pp. 149-157.

25. TuNLEY, T. H., KOHLER, P., and SAMPSON, T. D. TheMetsep Process for the separation and recovery of zinc,iron, and hydrochloric acid from spent pickle liquors.J. S. Afr. Inst. Min. Metall., vol. 76, no. 10. 1976. pp.423-427.

26. HAINES, A. K., et al. The use of weak-base resins in therecovery of uranium from unclarified leach liquors. Pro-ceedings of Eleventh International Mineral ProcessingCongress, 1975. Rome, Aziende Tipografiche Bardi, 1976.Paper 32, pp. 1-28.

27. JENSEN, J. H., and TAYLOR, A. Recent developments inion-exchange engineering. Mining Congress Journal, Apr.1977. pp. 26-28.

28. CRAIG,W. M., et al. The design and operation of a continuousion-exchange demonstration plant for the recovery ofuranium. J. S. Afr. Inst. Min. Metall,. vol. 78, no. 12.Ju1.l978 (this issue). pp. 316.324.

29. RITCHIE, M. 1. Utah Construction and Mining Company'sutlique Shirley Ba.3in uranium mill. The recovery of uranium.Sao Pallio, International Atomic Energy Agency, 1970.pp. 147-154.

30. GARDENER, H. E., KUNIN, R., and KEROLLIS, R. E. Appli.cation of weak base resin for recovery of uranium at Uravan.lO2nd Annual meeting, AIME. Chicago. 1973.

31. STREAT,M., and CALT.AGHAN,P. D. Continuous ion-exchangeThe synthesis of high detlsity ion.exchange materials bythe modification of metal oxide gels. Symposium on NovelSeparation Techniques. London, Institution of ChemicalEngineers, Jan. 1973.

Company AffiliatesThe following members have beenadmitted to the Institute as Com-pany Affiliates.

AECI Limited.AfroxjDowson and Dobson Limited.Amalgamated Collieries ofS.A. Limit-

ed.Apex Mines Limited.Associated Manganese Mines of S.A.

Limited.Billiton Exploration SA (Pty) limit-

ed.Blackwood Hodge (SA) Limited.Blyvooruitzicht G.M. Co. Ltd.Boart International Limited.Bracken Mines Limited.Buffelsfontein G.M. Co. Limited.Cape Asbestos South Africa (Pty) Ltd.Compair SA (Pty) Limited.Consolidated Murchison (Tvl) Gold-

fields & Development Co. Limited.Deelkraal Gold Mining Co. Ltd.Delfos & Atlas Copco (Pty) limited.Doornfontein G.M. Co. Limited.Durban Roodepoort Deep Limited.East Driefontein G.M. Co. Limited.East Rand Prop. Mines Limited.Engineering Management Services

(Pty) Ltd.Envirotech (Pty) Ltd.Free State SaaiplaasG.M. Co. Limited.Fraser & Chalmers S.A. (Pty) Limited.Gardner-Denver Co. Africa (Pty) Ltd.

Goldfields of SA Limited.The Grootvlei (Pty) Mines Limited.Harmony Gold Mining Co. limited.Hartebeesfontein G.M. Co. Limited.Highveld Steel and Vanadium Corpo-

ration Limited.Hubert Davies Heavy Equipment

(Pty) Ltd.Impala Platinum Limited.Ingersoll Rand Co. S.A. (Pty) Ltd.Johannesburg Consolidated Invest-

ment Corp. Ltd.Kinross Mines Limited.Kloof Gold Mining Co. Limited.Lennings Holdings Limited.Leslie G.M. Limited.Libanon G.M. Co. Limited.Lonrho S.A. Limited.Loraine Gold Mines Limited.Marievale Consolidated Mines Limit-

ed.Matte Smelters (Pty) Limited.Natal Cambrian Collieries Limited.Northern Lime Co. Limited.O'okiep Copper Company Limited.Otjihase Mining Co. (Pty) Limited.Palabora Mining Co. Limited.Photometric Sorters.President Stern G.M. Co. Limited.Pretoria Portland Cement Co. Limit-

ed.Prieska Copper Mines (Pty) Limited.Rand Mines Limited.R. J. Spargo Limited.

Rooiberg Minerals Development Co.Limited.

Rustenburg Platinum Mines Limited(Union Section).

Rustenburg Platinum Mines Limited(Rustenburg Section).

St. Helena Gold Mines Limited.Shaft Sinkers (Pty) Limited.S.A. Land Exploration Co. Limited.Stilfontein G.M. Co. limited.The Griqualand Exploration and Fi-

nance Co. limited.The Messina (Transvaal) Develop-

ment Co. Limited.The Randfontein Estates Gold Mining

Co. Witwatersrand Ltd.The Robbins Co. (Africa) (Pty) Ltd.The Steel Engineering Co. Ltd.Trans-Natal Coal Corporation limit-

ed.Tvl Cons. Land & Exploration Co.Tsumeb Corporation Limited.Union Corporation Limited.Vaal Reefs Exploration & Mining Co.

Limited.Venterspost G.M. Co. Limited.Vergenoeg Mining Co. (Pty) Limited.Vlakfontein G.M. Co. Limited.Welkom Gold Mining Co. Limited.West Driefontein G.M. Co. Limited.Western Areas Gold Mining Co. Ltd.Western Deep Levels Limited.Western Holdings Limited.Winkelhaak Mines Limited.

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY JULY 1978 315