adec preview generated nozzles of the irrigation equipment used to delivery leachate to the heap....

HEAP LEACHINGBy

DAVID MENNE

ABSTRACT

The preferred alternatives in heapdesign and their motivation are identified as

- .normal heap leaching, to begenerally applied wherever possible,flooded vat leaching, to be appl iedwhere wetted ore 1acks adequatecohesion or if 1imited spacenecessitates it,thin layer leaching, to be appliedwhere deleterious vat chemistrynecessitates it.

A number of the following features wouldprobably be incorporated in the optimizationof normal local heap leach operations:

- pre-cyanidation- pre-agglomeration- retaining walls- staged (counter-current) heap

treatment .- control of calcium levels- reduced pH and trickler irrigation

The following technologies associatedwith heap leaching are expected to becomepreferred methods:

adsorption using high pKa weak-baseresin columns or alternatively;staged heap leaching and directelectrowinning (where scal ing iscontrollable and chloride levels arerelatively low).direct smelting of bullion ingas-fired kilns (for largerinstallations), or alternatively;Micron-type production of bull ionflake by electrodepositon ontoaluminium and subsequent treatmentwith NaOH (for smallerinstall ations).

To overcome many of the problemsresulting from chemistry inherent in WestAustralian gold ores and groundwaters, thereis considerable motivation to move away fromalkali cyanidation to acid leach technologies.

The Aus.I.M.M. Perth and Kalgoorlie Branches, Regional Conference on "Gold-Mining, Metallurgy and Geology",October 1984.

229

230 D MENNE

INTRODUCTION

The occurrence of a smaller oxi de cap,accessible by cheap open cut mining andoverlying 1arger sul phide reserves is acommon feature of Australian gold deposits.

Furthermore relatively small tailingsdumps resulting from early operations aboundin each of the Australian goldfields.

Heap leaching provides a low-costoption for economically winning gold fromsuch oxide caps and tail ings heaps where theore grade and/or tonnage or initialavailability of water is inadequate tosupport slurry processing.

Such an approach is particularlyattract i ve to enterpreneurs who of necess ityhave to 1imit their initi al developmentexpenditures until cash flow con be generatedto better del i neate an ore body or prov i dethe more substantial infrastructure andfacilities required to mine and beneficiate,by slurry processes, gold from the underlyingmore competent and oxygen-consuming sulphides.

Even where the size and richness of adeposit exploited by open cut miningnecessitates slurry processing, substantialquantities of undergrade material aregenerally also produced, and heap leachingprovides a means for economically exploitingmuch of this material.

However, many heap 1each operati onshave demonstrated a variety of technicalpitfalls, which generally reduce cashflow tounacceptable levels even if reasonable totalgold recoveries are indicated. In othercases the short-cuts needed to generateprofitable cashflows have left fortunes ingold in the residue. Apart from the recordprovided by Kappes (1979}J the pitfalls arenot generally recorded (Gel ard, 1983).

This paper aims to address the majorprocess options for alkaline cyanide heapleaching and related technology for bullionrecovery, and to develop the rationale fordeci ding between these alternatives.Particular emphasis is placed on avoiding orcoping with common technical problems ofinterest to all existing and potential localheap leach operators.

However, to further cater for thevariety of interests represented in thisdistinguished gathering I have attempted toallocate a quarter of the contents tomaterial which cognoscenti will consider tobe old hat, and another quarter tostate-of-the-art developments which will beof limited interest to as yet uncommittedentrepreneurs who wi sh to be provi ded withhard, testable information only. I ask theindul gence of the one group for the subjectmatter included for the other.

The specific aspects of heap leachingtechnology which are considered in this paperinclude

- selection of heap design- feed preparation- gold precipitation, and- bullion production.Mining, comminution and heap/vat

construction also constitute importantaspects. However these aspects cover a vastsubject matter which has already beenaddressed elsewhere (Chamberlain, 1980; Clem,1982; Geldard, 1983; Heinen, personalcommunication; Potter, 1981) and will thusnot be discussed in this paper.

SELECTION OF HEAP DESIGN

The basic options include:- Normal heap design, where a

free-standing heap is built to ahei ght typi cally di ctated by theavailable material handling system.

- Flooded heap (vat) des i gn, where aheap is contained within impermeablebunds or walls allowing fullyflooded operation.

- Thin layer design, where only arelatively thin advancing ore faceis leached.

Normal heap design is generally appl iedbecause of simplicity and low cost.Re-usable pads are indicated where:

- pad sealing is costly- a wi de ore size range or risk of

compaction necessitatessophisticated placement techniques,and wherethe advantages of such pl acementtechniques justify the cost ofdouble handling the ore.

Vat leaching is generally necessary toprevent unplanned heap runaway should the orelack cohesiveness on wetting. (It should benoted that even should testwork indicate thatrunaway will not occur, it is still desirableto pl ace heaps so that if the experts werewrong, the heap will run to a more innocuoustarget than the gold recovery facilities).

Consideration of plant layout orclimate may also necessitate the applicationof vat leaching.Thin layer leaching is indicated if oreconstituents rapi dly deteri orate thecapability of leachate to execute itsdissolution function. This conclusion is incontrast to the original patented aim ofJohnson (1980) to apply thin layer leachingto contain within practical dimensions theproblem of fines migration leading to areduction in heap permeability.Selection tests to rationally decidebetween the above alternative heap leachdesigns include assessment of

- minimum cohesiveness to determinewhether flooded operation isnecessary, and


o MENNE 231

FEED PREPARATION

managementdes i gn

involve\Od 11 beto heap

- maximum kinetics of base, oxygen orcyanide consumption to determine thenecessity of operating a thin layerleach.

A description of the tests advised for theseassessments is beyond the scope of thisart i c1e, but can be executed by the Edward LBateman Pty Ltd (ELB) 1aboratory inKalgoorlie.

Further testwork related to minimumacceptable ore comminution, optimal leachateflux and consequent fines management alsoneed to be executed. Thi s data will affectdecisions concerning the preparation of feedprior to heap building and leaching, and isdiscussed in the next section.Retaining walls e.g. the prefabricatedL-shaped storage wall units shown on fi gure 1(Reid, personal communication) allows morematerial to be stacked on a given pad, inaddition to removing the risk of toe shear orpad runaway which may exist for some oresfo 11 owi ng suffi ci ent rain. The use of suchwalls also provides good sprinkler coveragewhile avoiding sidewall erosion.

Evaporation not only leads to excessivewater consumption but also concentrates saltsand increases the occurrence of harmfulpreci pi tates such as those block i ng thenozzles of the irrigation equipment used todelivery leachate to the heap.Water management to handl e preci pitat ionproblems is very much a 'horses for courses'issue, which draws on the followingfundamental approaches:

get and keep calcium levels lowuse equipment resisting and/orinsensitive to the build-up ofprecipitates [see figures 14a and b(Geldard and Houtgraaf, personalcommunications)]

- ensure that precipitates which doform are not coherent.

Although the chosen wateralternatives have facil ityimplications, the major issuesoperations. Thus this subjectdiscussed in the section devotedoperation.

.'.'

Comminution is generally a major issue infeed preparation. Comminution is generallyexecuted to liberation size, and this may bebest estimated in a firkin roll test:

Typically 100 kg of -32 mm material isleached in a rolling device (e.g. a 80/120litre concrete mixer) for an extended periodaround 48 hours. The residue after leachingis screened into a number of size fractions(typically 32 x 16; 16 x 8; 8 x 4; 4 x 2 and2 x 1 mm). Each size fraction is thenreduced to a size typically applied to slurryprocessing (e.g. 100% under 200 microns) andfurther leached for 24 hours.

An analysis of incremental goldrecovery will i ndi cate whether the increasedrecovery of gold by slurry processingexcludes heap leaching and if not, what thefeed ore size for heap leaching should be.

This approach provides less scatterthan attempts to derive pre- and post-leachassays of coarse size fractions from separatesampl es.Agglomeration, which can be more preciselydescribed as binding is desirable toaccommodate even modes t quant it i es of fi nesin ore intended for heap leaching.

Substances which bind mineral particlesare generally described as pozzolanas, andusually form, when dampened, calciuma1umi nate and s il i cate matr ices wh ich cementfines together, over a period of 10 to 102hours.

fFUTURE HEAP

PROFILE

Storage facilities should be provided toprevent loss of gold if excess leachate flowoccurs following rain or outage of theirrigation equipment.

In general merely a small surge sumpshould separate a normal heap from thebullion recovery facilities.

To provide for drainage during outageor rain, a suitably sized pond should be laidout to take any overflow from this sump.

...'.'

.'.'

.'

PRE ·FABRI (ATEO L·SHAPEO STOR AGE UNITS

FIGURE 1


---------------

232 o MENNE

•It • U to 10.

, • 10 JCl)l W •• _

). , ..

II&Ililll!IJlA(j(jLOl1ERATW O!!f !

I UNAGGlOMERATEO ORE!

lJr.uBU.

lMSTOHE I SlW.f

COlUMN TEST AND !'UHT OPERATION

IUNAuuLOl1fRATW ORE)

tlWJIBU

CHfRT I JASPEROID

COLUMN TEST AND PLANT OPERATION

It

10

""

"

1,00

to.."

I.-,.. ..Cl:

"..)(

I~

I:"'""

f'OURE 1

THE EFFECT OF CLAY ON SUPERFICIALLEACHATE FLUX

1'1to

",., \,., \,., \\

" \ .,,,,.,

\~... ","'".\:

0'• ~,

\\..: ..,.. .".. ."-)C ~l~... ~.... ..,~

""is \; ." \\

~ \;;: \'" HI \It :Jr.~

'" ..,--40,0-1 " H ~ I-S H

ClAY fUCTION Ifd_

Agglomeration generally allows thepermeability of clay-rich ores to bedramatically increased to provide asuperficial leachate flux around 1m3leachate/m2 heap surface per day.

Figure 2, drawn from informationdeve loped by Stroganov (1979), illustrateshow the fraction of unagglomerated clay in anore affects the superficial leachate flux.In practice clay fines are washed into lowerlevels, where they then constitute a greaterfraction. This can lead to seriouspermeation problems, as is well illustratedin the paper by Chamberlin. (1980).

In addition to the increased rate ofleachate flux and thus rate of gold recoveryresulting from agglomeration, agglomerationovercomes the dramatic (typically five-foldrate) loss in cash flow on scale-updemonstrated by Kappes (1981) on virtuallyall ore types as shown on figures 3 to 5.

This removal of the risk of scale-up byapplying agglomeration is borne out by theexcellent correspondence between small andfullscale operations which may be expected ifagglomeration is applied, as shown in figures6 and 7 (Houtgraaf, personal communication).(The corrections applied for leachate flux inthese fi gures are moti vated in the nextsection on Heap leaching).

UAOiATf APPLICATION TIll( OA'I'S •


233

"MABEL MILL:'


I AGGLOMERATED ORE)

100 ~--T"""-r---,--r--- ......--...-r-...-r-""

r :70

'I-60

SO

30

[

LIMITED INDICATED BYBOTTLE ROLL TESTS C"'D 3 • RATE CHANGE

- - - - - - -- - ~ - - ~~~~~ELB\~~~nLUX0"" •.,'" \..tt'

..- PLANTLEACH

20

10

] 4 • 10 20 )(I 40 60 10 110

LEACHATE APPliCATION TIME, OAYS ----••

'OLD MAN"

COLUMN TEST AND PlANT OPERATION

I AGGUHRATED ORE)

3 4 • 10

C"'D] • RATE HAN

IlIl1lTED INDICATED BY EXPlICABl£ BY LOWERIlOTTLE ROLL TESTS fW.S(AL£ LEAOiATE FLUX

- ------------- . ---'}" PLANT

, LEACH

20

lot•

lOO r-----r---....,.--,-......,....,.--.,.-.,-,.-",

1:70

LfACHI'TE APPUCATIOII TIME, ~YS •


---------------_._-- -

234 o t~ENNE

20 30 40 6C 80 100e 10

SIDE VIEW

'WOVEN SOCK

~._:YANIDE SOlUTGN

SLOTS

3 4

~AIN COIL FI TTED WITH FILTER SOCK

FIGURE 10

.. WILe IRISH GIRL'


[CYANDE AGGLOMERATED ORE]

CROSS-SECTION

[

LIMIT INDICATED BYBOTTLE ROLL TESTS- ----- -------------------

l!) ~ANT./' L ACH /'

LEACHATE APPUCATION TIME, OAYS ----••

10C

I'0

le

70

::J?0

li:6C

~ 50

UJa: 4(

~30

20

10

It should be noted that a patent forthe agglomeration process as originallydeveloped by the US Bureau of Mines has beenapplied for locally by Heinen et al. (1980).However this patent app1 ication is beingopposed by some of the heavyweights in thelocal industry.

As the activities of many pozzo1anasare known or easily assessed (e.g. it isknown that 1kg cement or 2kg fine f1yash willbind approximately 30kg clay), it isgenerally possible to estimate how muchpozzo1an is needed to reduce the amount offree clay to a reasonable level. Howeversince any remaining free clay may migrate tolower levels and form relatively impermeablepans which divert 1eachate, it may provedesirable to bind practically all the clay intaller heaps.

Many ores are inherently pozzo1anic,and extremem1y cheap pozzo1anas such asf1yash wastes may be available. Thus theamount of fines which may be economicallybound can vary considerably for individualcases.

Where the cost of reagents foragglomeration is excessive, either the finerfraction or the entire ore needs to betreated by a slurry process. Where a coarserfraction of ore is removed, it will often beamenable to treatment by normal heap1each ing, wi thout further commi nut ion, asgold mineral ization is often concentratednear fracture planes. These options in factcover the historic sand/s1 imes and a11-s1 imesprocesses of gold recovery.Pre-cyanidation of ore is generallydesirable so that when solution is passedthrough the ore the major task of 1eachate iswashing out dissolved gold rather than tryingto get it into solution.

The effect of pre-cyanidation oncash-flow is dramatic, typically increasingit three-fold as shown on figure 8(Houtgraaf, personal communication).

Although equipment was alreadydeve loped to execute such so-called drycyanidation by 1921 [see figure 9, reproducedfrom Ju1ian and Smart (1921)J this processhas been recently patented for ores finerthan 1 mm by Balakrishnan and Skinner (1978).

PLAN

FIGlIRf Q

BLAISDELL MIXERE.QB ':!!lliTENING PULVLROJ lJ~t

WITH CYANIDE SOLUTION-

a

ELEVATION

The Aus.I.M.M. Perth and Kalgoorlie Branches, Regional Confe't"vv vd "Gold-Mining, Metallurgy and Geology",October 1984.

235

,-- ....

, • 10 10 JO It4 60 1(1100

~8

LI I1fS TONE

EfffCT OF L!A£HATE FLUX

IlJWiGl.a1ERATfD ORE J

FIGURE '11.

ALl! GATOR RIDGECOLUMN TEST

CYANIDE AGGLOMERATED ORE

lEACHATE APPLICATION TII1E, OAY$ -

""

In agglomeration cyanidation, theeffect of alkalinity and cyanide strengthmust be carefully considered and establishedby testwork to avoid excessive loss of costlycyanide and/or dissolution of undesiredcations.

The exceptions where pre-cyanidation ofore is undes irab1 e, occur if substancesdeleterious to the leach require theirconversion or removal prior to the leach.Such deleterious substances are generallybase, oxygen or cyanide consumers, such as

unstable sulphides (e.g. pyrrhotite,realgar and orpiment) which mayappear as the su1 phi de zone of anore body is approached.ferric mineral s common to ore zonesnear the watertab1e, andmany copper minerals as listed inTable 1. The benefiication ofcopper-rich gold ores is however avast subject and for heap leachingcan become very comp1 ex as thecopper cannot always be washed outprior to cyanidation. This subjectwill thus not be discussed here.

In conclusion I wish to note thatparti cu1 ar1y coherent base and top layers aregenerally desirable to respectively providefor stable drainage and to allow traffic formaintenance of irrigation equipment withoutexcessive fines generation. Thus thedevelopment of practical heap buildingprocedures to ach i eve such strat ifi ed heapswould be very useful. Higher cement dosagesmay thus be desirable for the critical topand bottom 1ayers. Furthermore dra in co il sfi tted wi th a fi Her sock as shown in fi gure10 (Trichel, private communication) may beincorporated in the base of the heap.

10

10

LfACHATE AfIfIUCATtOtl TIH(, DA~ _

TABLE 1: Cyanide reactivity of copper .inerals (arranged in order ofincreasing reactivity)

Unreactive minerals Reactive .inerals(10 + 101 dissolves) (85 + 101 dissolves)

Cha1copyrite CuFeS Enargite cUlSS4Chrysoco11a CUSiO~~2H20 Bornite CU 5 eS 4Tetrahedrite Cu 3S S3 Cuprite Cu 0

Native Cop~er CuCha1coci te Cu SMalachite cucg3.CU(OHl

2Cove" ite CuS


o

236 o MENNE

" •

I 4 4 I 10

LEACHATE APPLICATION TIME, DAYS-

10

110

10

•

lli.Y..!1..D•Q.l..Q..!W['

COI.UHN TEST

IAGUlOHERATEO ORE J

WILD IRISH GIRLCWJHN TEST

(AGGLOI1£IlATED ORE )

l(A(HAT[ APPlI(ATIOH 1IH(, OA~

>-

i .•<>~ 10I:>

~>- 6llIX

~ sou....IX .0

o;5 10l:)

90

A further motivation for applying dripirri gation is that there is greaterflexibil ity in the selection of pH andcyanide levels. The reduced exposure ofleachate to air when applying drip ratherthan sprinkler irrigation allows more extremelimits of pH levels and cyanide concentrationto be applied without necessarily increasingthe consumption of cyanide excessively. Thisprovides considerable operating flexibilityas well as room for effectively reducingprocessing costs.

HEAP OPERATION

-Recovery of gold proportional to cyanideexposure- is concluded to form a basic lawof heap performance.

A considerable aroount of test andoperating data (Houtgraaf, personalcommunications; Kappes,1981; McClelland andEisele, 1981) indicates that for a heap thatoperates satisfactorily the gold extractionis virtually proportional to the total amountof cyanide per unit of ore passed through theheap. Figures 11 to 14 are typical of thedata that support this view:

Fi gures 11 a and 11 b demonstrate thatchanges in leachate fl ux change therate of gold recovery in about thesame proport ion.Figure 12 demonstrates that leachingpractically ceases on cessation ofirrigation and proceeds at a normalrate on recommencement ofapplication of leachate.Figure 13 demonstrates that at afixed leachate flux, the rate ofleaching is closely related to thecyanide concentration.

Naturally other, more varied andobscure conclusions may be drawn from thetreatment of ores suffering from

excessive bypass because of afailure to agglomerate excess fines,orexcessive levels of constituentswhich interfere with the cyanidationreaction or the stability of thedissolved gold complex aftersolubilization, ora high risk of cyanide loss e.g.from using very high cyanideconcentrations and inadequate pHlevels.

The results shown on fig lla in factsuggest that the presumably fresh(non-leaching)leachate was depositing a slowprecipitating substance which affected theextent of bypassing occurring wtth~ the heap.leachate distribution cons i tutes one of themost critical aspects of heap operation.

Except for vat leaching where floodeddams are created, irrigation equipment suchas sprinkl ers are generally used todistribute leachate.

There is a strong case for using dripirrigation equipment as illustrated in figure14c (Trichel, personal communication) toreduce evaporation and the erosion(particularly of side walls) resulting fromdroplet impact. The problem of attempting toachieve droplet size distributions whichadequately balance such erosion withevaporation and mist losses can then also beavoided.

The Aus.I.M.M. Perth and Kalgoorlie Branches, Regional Conference on "Gold·Mining, Metallurgy and Geology",October 1984.

DMENNE

Water quality management in heap leachingconstitutes one of the most critical areas ofheap operation. Interesting conflicts ofchemistry, requiring innovative solutions formany heap leaching operations occur in thisarea.

In particular, cyanide associates withprotons to form prussic acid at pH levelsbelow 9,5. (Association is practicallycomplete below pH 7,5). In contrast tocyanide, prussic acid is not very effectivein 1eaching gold, and is furthermore easilylost by volatilization. However high calciumand carbonate levels are common in localwater, furthermore bicarbonate commenses todissociate to carbonate above pH 7,3, and ispractically completely dissociated tocarbonate above pH 9.3, and calcium carbonateis not very soluble.

Thus if one wi shes to avoi d theproblems of reduced activity of cyanide.fOrgold dissolution and also the potential forits loss as prussic acid vapours, one mustaccept the risk of calcium carbonateprecipitation. (The precipitation of calciumsalts particul arly affects the effectivecontinued operation of the irrigationequipment used to spread leachate over theheap, and also fouls the carbon commonly usedto collect dissolved gold.)

A further complication in watermanagement that arises in the treatment ofmany local weathered rocks is that magnesiumreleased by weathering has not been washedout of the oxide ore because of the very aridconditions, and/or when such magnesium hasbeen washed out, it has concentrated in thegroundwaters which are most easily availablefor beneficiation of the ore.

Magnes i urn sets up a buffer in the pHrange 8,8 + 0,3 depending on magnesiumconcentration: If cyanide leaching isattempted in this pH range, increased cyanidedosages need to be applied to allow for lossof cyanide by vapourization. Thus either themagnes i urn must be preci pitated pri or to heapleaching, or drip rather than sprayirrigation of leachate must be applied toreduce cyanide consumption.

Because of inter alia the lowsolubility of lime and the potential ofinsoluble precipitates on the surface of limepreventing the full util ization thereof,magnes i urn needs to be preci pitated by NaOHrather than 1ime. Essentially all themagnes i urn can be preci pitated with merestochiometric quantities of caustic since thesol ubil ity product of Mg(OH)2 at ambient isso low (about 10- 11 ).

Whether spray or drip irrigation isapplied, the risk of precipitation of calciumsalts (sulphates as well as carbonates) must

237

be el iminated. If no sul phates are present,the risk of calcium carbonate precipitationmay be avoi ded by a modes t reduction in pHafter the bulk of the calcium has beenprecipitated by sodium carbonate. Sodiumcarbonate can remove calciumstochiometrically, since the solubilityproduct of calcium carbonate is relativelylow (around 10-8 to 10-9 depending onambient temperature). However one mustconsider the influence of pH as the protonassociati~n constant for carbonate is fairlyhigh (100 to 1011 depending on ambienttemperature) .

FIGURE 14

IRRIGATION DEVICES

oI , I, I, ,

III

NUTATlNG HEA/)

FIXED IIASE

lal THE SEHNINGER WOBBLER ®

1111 STATIC FULL - CIRCLE SPRINkLER.

le I EXPOSED VIEW OF LEACHATE FLOW PATH

IN A LABYRINTH ORIPP£R


238 o MENNE

FIGURE 15

5 ~ 15 20I ONI C STRENGTH OF LEACHATE •

t•• q /l I

-5 10 15IONIC STRENGTH OF LEAOiATE

Imeq/I)

20

10

RATIO OfHONO·TOBIVALENTCATIONSFflOMSATURATIONEXTRACTOF HEAP

Ma·

Pi H9 2· J)

1e

140

30

EXCH1.NGEABLISOO!U'1PERCENTAGE 2e

arguments that the effect is due to a changein habit of existing seed crystals. Howeverdespite the great dispute over howmagnetohydrodynamic water treatment works, itdoes seem to work.

A recent survey by Edward L Bateman(Barnea, personal communication), revealedabout a dozen firms offering this technologyand little basis for knowing how to selectthe equi pment of one rather than the other.The only common feature appeared to be that.all appeared to wish to price themselves outof the market.

HEAP PERMEABILITY

IN RELATION TO ADSORPTION CHARACTERISTICS

FIGURE 16

HEAP PERMEABILITY OF NON-SODIC HEAPS

IN RELATION TO IONIC CHARACTERISTICS

Controlled precipitation must be executedwhere precipitation cannot be practicallyel iminated.

Sulphates, which are common in localores derived from sulphide oxidation, presenta particular water management problem asgypsum readily supersaturates, and canunexpectant1y precipitate at a very high rateat favourable sites (e.g. those experiencingmechanical shock). Gypsum fouling is alsonot as easily removed (e.g. by a dilute acidwash) as calcium carbonate.

A successful approach to the control ofcalcium fouling (of activated carbon) whichhas been adopted by the Pinson Mine, Nevada,i nvo1ves feed i ng ca1ci um saturated waters tota il i ngs, where many nuc1eat ion sites on thesolids promote stripping off supersaturatedfractions (Thorndycraft, personalcommunication). Duncan and Smo1ic (1977)record how Cortez Gold Mines successfullyexecuted a similar conditioning step in a1arge barren pond at their Gold Acresoperation. A variation of this techniqueapplied in the Florida phosphate industry atthe Bartow plant relies on the recycle ofgypsum fines to seed and precipitate CaS04as a fine powder in solution rather than as atroublesome solid on equipment.

Alternatively, after the bulk of thecalcium has been removed e.g. by CaC03precipitation, a significant quantity of theresidual calcium may be removed either bysequestering or precipitation. Phosphatesare generally applied for either of theseprocesses. However po1yphosphates used forsequestering generally have to be applied atthree times stochiometric levels to removethe bulk of the calcium as the complexesformed are not particularly strong. Thisexcess po1yphosphate can disperse clayswithin the heap, causing permeabilityproblems.

Simp1 e phosphates are a1 so dispersants,but since they can be practicallystochiometrica11y reacted with calcium (thesolubility product of Ca3(P04)2 is verylow - about 10-32 ), the problem of claydispersion need not arise.

A final, and the most promising futuretechnique for the control of precipitates(which is however at present more at thelevel of a Black art than a science), ismagnetohydrodynamic water treatment.

By pass i ng water at a si gnifi cantvelocity (around 10 m/s) through a strongmagnetic field (over 3000 Gauss as nowavailable from state-of-the-art permanentmagnets) subsequent preci pitat ion of ca1ci umsalts occurs as a slush rather than acohesive build-up. I believe this is due toextensive seeding of crystals occurring inthe magnetic field although there are


D MENNE 239

" 1', ' ,,......

,[...... , 1 •

...... ,0.... '.. ~,o' ...... "

1,I ....... I ...... II

I ......

IAu ([Nl; ,

•• I ..........I·,

II

~ I.... 01 I .,-I~I.

" ~I~ "1... .... Auxl.... I....., ..... I ·11

...... I,I....,. ...... ~'O I 'H

~'< I", ........ ..... II, ,......

." ..... .• I

Breakdown of colloid structure occurs ifthe ratio of dissolved alkali earth to alkalimetals becomes too low e.g. as indicated byfigures 15 and 16 (Quirk and Schofie1d, 1955;Quirk, 1971). If too much calcium andmagnesi urn is preci pitated colloid dispersionoccurs and consequent heap permeabilityproblems become inevitable.

Thi s fact prov i des a good argument forthe removal of essentially all the calciumand the retention of reasonable magnesiumlevels (and application of drip irrigation tocontain losses of cyanide as prussic acidvapours), unless extensive agglomeration isapplied.

Such an approach makes further sense ifone considered the relatively low cost ofNa2C03 (needed to precipitate Ca) incomparison with 2NaOH (needed to precipitateMg), and particularly so if one appreciatesthe feature of local groundwaters that levelsof soluble calcium are typically an order ofmagnitude lower than soluble magnesium levels.

Most of the above problems of watermanagement, preci pitation contro1 and break downof colloid structure are eliminated byapplying an acid leach, and this provides asignificant motivation to move away fromalkali cyanidation to acid thiourea orchloride processing.

to 1 I) ~ to l • \0 h 11 11 13 '",.

fiLcH 17

POURBAIX DIAGRAM'J..u· CN'· H2011e 'n. (N'J

I.

GOLD PRECIPITATIONFigure 17 demonstrates that gold can be

precipitated out of aurocyanide solution tolow tenor levels (as metal) by practicallyall commercial reductants at all pH levels,but more easily in acid solution.

Table 2 provides a brief summary ofsome gold precipitation processes which havebeen applied to date (Gmelin, 1954; Rose,1906; Robine and Lenglen, 1906) and why someare no longer applied.

In general, the use of solublereagents, and any need to manipulate pH todifferent levels for leaching andprecipitation are not desirable in heap1each i ng because of the inherent bu il dup ofsuch reagents wh i ch occurs fo 11 owi ng recyc1eof solution. In fact, the buildup of zinc(unless alkali conditioning be applied), isone of the main reasons why Merri1precipitation is not generally desirable in aheap leach circuit. (The complexitiesintroduced by

- dilution of precipitate by suspendedsolids which occurs if the feedstockis not clarified

- the desirability of deoxygenation,lead addition and absence of variousbase metals and

- the maintenance of a fair calcium1evel

to obtain good results, constitute furtherreasons) •

Some very neat processes have beendeveloped to handle the reagent build-upinherent in recycle operations, which couldoccur in heap leaching. As an example thecyanide recovery process of Crits (1981) usesType II resins and simple e1utions toseparately recover zinc and recycle cyanide.However this general subject area of cyaniderecovery is vast and cannot be adequatelyaddressed in this paper.

The main gold precipitation processesthat thus remain contenders for appl icationin a heap leach circuit are e1ectrowinning,either direct or following priorconcentration by either carbon column orresin column adsorption and elution ofaurocyanide.

Generally at the low solution tenorsinvolved in heap leaching, film diffusion1imits the rates at which gold can beelectroplated. Any excess current is merelywasted on hydrogen production. Furthermorethe increase in overvo1tage as the gold tenordrops, and what appears to be some sort ofback reaction (redisso1ution of dislodgedgold flakes) appears to create a lowerasymptotic gold recovery 1imit in the rangeof typical heap 1eachate tenors.

o 1 1 ) 4 !I , ., • • 10 11 12 U u

pH-_


240

TAIl£ 2: Pred1l1t:uts "'1c~ ~lYe Mea t11sclosed or usedfor ~ recovery of gold fro. cyAn1de solut10ns

PRECI PlTANT RECOROED OR MOST LIKELYDEFECTS

Z1 ne Efficiency is sensitiveto the presenceof base metals anddissolved oxygen.

Carbon Low concentrations ofdeposited gold general 1insufficient to avoidnecessity to attemptrecycling of carbon.

Cathodic Current Efficiency and raterapidly drops atlower tenors (except1n cells which areyet not well provenconaercially).

Cathodic Currents Low capacity. highonto Amalgams cost. requires

clarified solutionto avoid sI ime-1ayer formi ng onamal gam

Cu-Zn Amalgam; Zinc-Amalgam Cost. high residuetenors

Alullinium Requires de-aerationand low Ca-levels

S11icon; Silicon/AlkaliEarth Alloys Cost

CuCI. H~SOt/CuCl/oxidant;

H6S0~/C (J ;S 2/ uSD4 Cost

H2S04/FeS04 Cost

Alkal i ne hydrogen phosphide Toxicity

Acetylene or carbide Cost

Pb-salts Cost

Hg plus Cu plus Zn-salts Toxicity

D MENNE

The abovementioned characteristicsprevent the application of conventionalelectrowinning cell designs, and one or moreof the following concepts must beincorporated to allow practical applicationof direct electrowinning of heap leachate:

electrode scrubbing, wheremicroturbulance greatly reduces thediffusion gradient/gold depositionbarrier,staged (counter-current) heap leachoperation to provide far greaterleachate (electrolyte) tenors, and/orapplication of electrowinningres i dues to pre-treat ore in apre-cyanidation and/orpre-agglomeration step. Anyinefficiency in the cell performanceis then of no major consequence asunrecovered dissolvecl sold.. isrecycled right up to the lastcampaign. The liquors of the lastcampaign can then be stripped ofgold in a single chemicalprecipitation or adsorption step.

A further desirable electrowinningfeature is a large cathode surface to providefor a sufficient rate of gold deposition atthe diffusion limit.

However before laying out money on ace11 us i ng poreous cathodes, one mus t beconvinced that any given cell will in factoperate well despite the occasional presenceof significant quantities of suspended solidsin heap leaches, which can clog poreouscathodes.

Another point worth checking is if, atthe rel atively low leachate conductivity ofsome heap leachates, adequate cathodepenetration will occur. In some cases therapid drop in potential over thesolution/cathode boundary deeper within aporeous cathode can 1i mit go 1d P1at in g to aninadequate outer zone nearest the anode (Paulet al. 1983).

The USBM has been developing stagedheap 1each i ng technology to increase thetenors of heap 1eachates and so reduce thedemands on the electrowinning equipment, andare also addressing the development ofimproved electrowinning techniques (Eisele etal. 1984). ELB has also been addressing thedetail design of cells to achieve practicalelectrowinning of the typical large flowratesand low precious metal tenors, conductivitiesand levels of anode-protective alkalinityinvolved in heap leaching.

However, the high chloride levelsinherent to many local ground waters andcontained within soluble salts in local oreswill probably necessitate further developmentof anode construction even though theexisting approach may be satisfactory forapplications where the water quality isbetter.

The Aus.I.M.M. Perth and Kalgoorlie Branches, Regional Conference on "Gold-Mining, Metallurgy and Geology".October 1984.

D MENNE

Activated carbon allows simple andefficient capture of gold out of leachate,whereafter it may be eluted and electrowonfrom a more concentrated solution.

The first commercial process by whichthis was achieved was developed by Zadra, andhis process has subsequently been modified ina number of ways including the use'of organicadditives and the application of highertemperatures.

AARL elution involves separation of thethree steps of carbon pretreatment, carbonelution and eluate electrowinning • Thisallows considerable flexibility in optimisingeach step and provides a number of otheradvantages over Zadra elution as outl ined intable 3. A more complete paper on thissubject is available on request from ELB.

The Micron Research elution technologyhas been locally developed and is thus fairlywell known here. At the time of writing itappears that early problems apparentlyrel ated to engineering rather than anyfundamental flaw in the process have beenovercome and two commercial units are said tobe behaving satisfactorily in commissioningtrials (at Tower Hill and Edjujina). Howeverinsufficient information is yet available atthe time of writing this paper on the longerterm performance of carbon eluted by thisprocess and the extent to which regenerationmay be dispensed with, to allow a fairassessment of this technology to be made atthis stage.

Fleming and Cromberge (1984a, 1984b,1984c) have provided information whichindicates that high pKa weak-base resins maysupercede carbon as an adsorbent to recovercyanide-leached gold.

241

In brief, resins are far less prone tofoul ing, and a small quantity of very strongelectrolyte ideally suited for electrowinningis easily and rapidly generated from loadedresin by a simple caustic wash.

A high pKa is desirable to avoid theneed for acidulation prior to adsorption andsubsequent neutralization prior to recycle ofleachate.

Although electrowinning of therelatively strong electrolytes produced bythis process can be processed in cells ofvirtually any design, the use of a designwhich incorporates a parallel fluid andcurrent fl ux as ill ustrated by Surfl eet andCrowle (1972) and relatively thin cathodes isdesirable because of the high gold loadingsper unit of steel wool which can be achievedin this type of cell provided conditions areselected to avoid inactive zone in thecathode.

Such high loadings allow practicaldirect smelting of· cathodes without thenecessity of attempting prior acid washing ofthe steel wool, which is after all goldplated and thus protected against the veryattack being attempted.

The electrowinning of gold ontoaluminium foil and the alkali dissolution ofthe foil to produce bullion flake, asdeveloped by Arnold Griffin is an attractivealternative, particularly where smallerquantities of electrolyte are involved.BULLION PRODUCTION

Smelting constitutes one of the moreobscure areas re1ated to heap 1eachoperation. A great variety of practices areapplied, few rationally developed andoptimized and most reminding one of medievalalchemy.

TABLE 3: The relatiye ~rits of AARL and Zadra elution.

The AARL system

is cheaper with respect to operating costs

is more flexible to handl ing changes in feedstock or increasesin contaminants in feedstock

is less sensiti ve to temperature fl uctuati on occurring duri ngelution, and contaminants returned by recycling of eluate

is faster

limits kiln degradation

The Zadra system

can more readily accommodate poor quality of water used to makeup eluate

is tolerant of unoptimized electrowinning call design

is simpler where no sequence control may be applied.


242

Because of the practi ca1 importance ofsmelting in heap leaching operations, thissubject has been addressed fairly extensivelyin a paper which is available on request fromELB. This paper also discusses the bestcrucible, kiln and fuel for smeltingsteelwool cathodes to recover dore bullionunder typical local conditions. Table 4summarizes a set of good flux recipes derivedin th i s paper. Potass i urn rathe.r- than sodiUM haslate~ been specified for the nitrate because ofits superior thermal stability.

ACKNOWLEDGEMENTSThe permission of the Northern

Queensland Company to prepare figures frominformation provided by Marc J Houtgraaf,their Chief Geologist, is gratefullyacknowl edged, as are the personalcommunications of the following contributorswhich have been recorded in the text:

El i Barnea, Managi ng Di rector, Edward LBateman, Haifa, Israel; Denis Geldard,Operations Manager, Whim Creek ConsolidatedNIL, Perth; Arnold Griffin, Principal,Mi cron Research, Perth; Davi d Mu i r, Seni orLecturer, Murdoch University, Perth; IanReid, Estimator, Concrete division, HumesLtd, Perth; Bruce Thorndycraft, Millsuperintendent, Pinson Mining Co, Winnemucca,Nevada; Bernie Trichel, Sales Representative,Plastics division, Humes Ltd, Perth.

The assistance provided by David JCostello, Lecturer, TAFE in regard toaddressing water qual ity management is al sogratefully acknowledged as are the valuablecomments on pi oneering USA operati onsprovided by Harold J Heinen and Gene EMcClelland respectively ConsultingMetall urgist and Manager of Heinen-LindstromConsultants (a Division of Bateman ProcessServices, Inc.), Reno, Nevada, and Judith AEisele, Research Supervisor, Chemicalprocesses, USBM Reno Research Centre, Nevada.

TABLE 4: Recipes useful in direct smeltingof steel wool cathodes

a) Flux per kg Fe.

b) Flux per mole of base metal(generally Cu) to beremoved from dore

c) Flux for additional in-situoxidation (dull or stickyslag; low fineness;brassy, hard ormagnetic dore).

2 kg NaN031,4 ~g Si021,3 kg Na2B402.10H20

1 mole Si020,2 moles Na2B40

7.10H20

55% NaN0345% SiO:;>


o MENNELiterature References

243

Julian, H. Forbes and Smart, Edgar, 1921.Cyani ding gold and silver ores. CharlesGriffin &Co. London (3rd edition, revised byA. W. Allan) pp 168-170.

Quirk, J. P. and Schofield, P. K., 1955.The effect of electrolyte concentrations onsoil permeabil ity. Journal of Soil Science 6(2): 163-178.

Quirk, J. P., 1971. Chemistry of salinesoils and their physical properties. Chapterin Talsma, T. and Philip, J. R. (Editors).Salinity and water use. Macmillan, London.pp. 79-91.

McClelland, Gene E. and Eisele, Judith A.,1981. Improvements in heap 1each ing torecover si 1ver and go1d from low graderesources. USBM report of investigations8612 pp. 9-11.

1981. Des i gn factorsMining Engineering.

Paul, Roger L.; Filmer, Antony O. and Nicol,Michael J., 1983. The recovery of gold fromconcentrated aurocyanide solutions. Out ofOssea Asare K. and Miller, J. D.Hydrometallurgy - research, development andplant practi ce. (Proceedi ngs from the 112thMeeting of the Metallurgical Society of theAIME, Atlanta, Georgia).

Robine, R. and Lenglen, M., 1906. TheCyanide Industry. (Translated by J. ArthurLe Clerc). John Wiley & Sons, New York (1stedit'; on).

Rose, T. K., 1906. The metallurgy ofgold. Charles Griffin &Sons, 5th edition.

Schulze, Reinhold, 1984. New aspects ofthiourea leaching of precious metals.Presented at the International PreciousMetals Symposium, Los Angeles, (27-29 Feb).

Stroganov, G. A. et al., 1979. Optimizinggold ore heap leaching procedures. TsvetnyeMetally No. 11 (Oct) pp. 99-102. Reprinted inWorld Mining, May 1981, pp. 80-81.

Potter, George M.,for heap leaching.(March) pp. 277-281.

Kappes, Daniel W., 1979. Precious metalsheap leaching simple - why not successful?Presented to the Northwest Mining Assocation,Nevada (7 Dec.).

Kappe~, Daniel W., 1981. Scale-upexpenence in gold-silver heap leaching.Chapter 4 out of Sch 1i tt W. J., Larson W. C.and Hiskey J. B. Gold and silver leaching,recovery and economics. [Proceedings fromthe 110th AIME meeting Chicago, Illinois(February 22-26)J.

Gmelin, 1954. Handbuch der AnorganischenChemie: System-nummer 62 (Gold) VerlagChemie, GMBH, Weinheim. pp. 305-311.

Heinen, Harold J.; McClel1and Gene E. andLindstrom Roald E., 1980. Leachingagglomerated gold-silver ores. Australianpatent application 57386/80.

Johnson, Paul Howard, 1980. Thin layerleaching method and apparatus. Australianpatent 508221 (1 Feb.).

Eisele, Judith A., Wroblewski, M.D., Elgers,Carl H. and McClelland, Gene L, 1984.Staged heap leaching and directelectrowinning. [Proceedings from the Firstinternational symposium on precious metalsrecovery, Reno, Nevada (June 10-14)J.

Fleming, C. A. and Cromberge, G., 1984a.The extraction of gold from cyanide solutionsby strong- and weak-base anion-exchangeresins. Submitted to the Journal of SouthAfrican Institute of Mining and Metallurgy.

Fleming, C. A. and Cromberge, G., 1984b.The e1ut i on of aurocyani de from strong- andweak-base res ins. Submitted to the Journalof the South African Institute of Mining andMetall urgy.

Fleming, C. A. and Cromberge, G., 1984c.Small-scale pilot-plant tests on theresin-in-pulp extraction of gold from cyanidemedia. Submitted to the Journal of the SouthAfrican Institute of Mining and Metallurgy.

Geldard, Denis, 1983. The evolution ofgold ore heap leaching and carbon-in-pulptreatment. Austral. Inst. Min. Metall.Bulletin No. 476/477 (July/August) pp 23-25.

Crits, George J., 1981. Cyanide recovery.United States patent 4267159 (12 May).

Duncan, D. M. and Smolik T. J., 1977. HowCortez Gold Mines heap-leached low grade goldores at two Nevada properties. E & MJ (July)pp 65-69.

Clem, B. M., 1982. Heap leaching gold andsilver ores. E &MJ. (April) pp 68-76.

Balakrishnan, Ramachandran and Skinner,Geoffrey Frederi ck, 1978. Extracti on ofgold and silver. Australian patent 33760/78.

Chamberlin, P. D., 1980. Heap leaching andpilot testing of gold and silver ores.Proceedings of the American Mining CongressMining Convention, San Francisco, California(21 to 24 Sept.).


adec preview generated nozzles of the irrigation equipment used to delivery leachate to the heap....

Documents