ITEM VIEW

The high-temperature pressure oxidation of a Witwatersrand pyrite concentrate

dc.contributor.advisorSteyl, J. D. T.en_ZA
dc.contributor.advisorDorfling, C.en_ZA
dc.contributor.authorStrauss, Jakolien Adrianaen_ZA
dc.contributor.otherStellenbosch University. Faculty of Engineering. Dept. of Process Engineering.en_ZA
dc.date.accessioned2019-03-04T10:36:34Z
dc.date.accessioned2019-04-17T08:34:50Z
dc.date.available2019-03-04T10:36:34Z
dc.date.available2019-04-17T08:34:50Z
dc.date.issued2019-04
dc.identifier.urihttp://hdl.handle.net/10019.1/106210
dc.descriptionThesis (MEng)--Stellenbosch University, 2019.en_ZA
dc.description.abstractENGLISH ABSTRACT: The oxidation of the pyrite present in gold and uranium ores is often desirable, to either liberate the gold present in the pyrite, or to generate iron(III) and acid for uranium extraction. The current study aimed to define a rate equation to describe the pressure oxidation (POX) kinetics of a Witwatersrand pyrite concentrate. Iron(III) and sulphuric acid, which is produced during POX of pyrite, may be used to leach uranium, which is also present in Witwatersrand ores. In addition, gold recovery from the POX residue may be improved significantly, especially if the gold ore is partially refractory, as is the case in many Witwatersrand tailings dumps. A rate equation to describe the pyrite oxidation kinetics will, thus, be important for reactor design purposes, as well as for operating and capital expenditure estimations to evaluate the feasibility of flow sheets incorporating POX. The pyrite POX kinetics was investigated in terms of temperature (180 to 210°C), oxygen partial pressure (480 to 1100 kPa), acid concentration (10 to 50 g/L) and particle size (+38 to 150 μm) by using a batch 2 gallon Parr autoclave. A batch model was subsequently developed in MATLAB and employed to confirm the observed rate dependencies. The oxygen gas-liquid mass transfer coefficient was also measured independently during the study, to enable a quantitative description of the dissolved oxygen concentration during modelling of the batch oxidation process. The activation energy was calculated as ~120 kJ/mol in relation to dissolved oxygen concentration, which indicated that the reaction was controlled by a chemical reaction at the surface of pyrite particles and that no diffusional limitations applied. The oxidation rate decreased with increasing acid concentration with a reaction rate order in acid concentration, ranging between -0.2 and -0.3. The oxidation kinetics was found to be relatively insensitive to particle size at oxygen partial pressures lower than 1000 kPa. For all practical purposes, the pyrite oxidation rate was found to be first order in dissolved oxygen concentration; however, this assumption led to poor prediction of the iron(II) and iron(III) solution concentrations during modelling of the batch oxidation tests. Accurate quantification of the iron(II) and iron(III) solution concentrations would also be important to consider for reactor design purposes, as it will dictate the maximum quantities of iron(III) and acid that can be produced during POX. Simulations showed that improved predictions of the iron(II) and iron(III) concentrations are obtained when a direct reaction between pyrite and iron(III) was allowed for. This means that both dissolved oxygen and iron(III) are responsible for the oxidation of pyrite at typical POX conditions, i.e., the pyrite could have a dual rate dependency on the oxygen and iron(III) concentrations in two additive rate-determining steps. Regression indicated that the experimental data may be represented by two additive rate equations with orders of ~0.6-1.0 in dissolved oxygen concentration and half-order in iron(III) concentration. The relative contribution of the two reactions to the overall rate appears to be influenced by the slurry density, particle size, and oxygen partial pressure. It is proposed that a follow-up study should be conducted to quantify the rate dependency of pyrite in solutions of varying iron(III) concentrations, at the temperatures employed during this study and in the absence of dissolved oxygen, that is, to provide an independently measured rate equation to the batch POX model. The homogenous iron(II) to iron(III) oxidation rate should also be measured independently to confirm whether the employed rate equation was correct. Furthermore, the possible effect of secondary minerals, in this case, pyrophyllite, should be clarified by conducting experimental work at higher slurry densities.en_ZA
dc.description.abstractAFRIKAANSE OPSOMMING: Die oksidasie van die piriet teenwoordig in goud- en uraniumerts is dikwels verkieslik om óf die goud teenwoordig in die piriet vry te stel, óf om yster(III) en suur vir uranium ontginning te genereer. Die huidige studie het beoog om ’n tempovergelyking te definieer om die drukoksidasie (DOX) kinetika van ’n Witwatersrand pirietkonsentraat te beskryf. Yster(III) en swaelsuur, wat gegenereer word gedurende die DOX van piriet, kan gebruik word om uranium uit te loog, wat ook teenwoordig is in Witwatersrand-erts. Daarby kan goudontginning van die DOX residu beduidend verbeter word, veral as die gouderts gedeeltelik refraktêr is, soos in die geval van baie Witwatersrand uitskothope. ’n Tempovergelyking om die piriet oksidasiekinetika te beskryf sal dus belangrik wees vir reaktorontwerp doeleindes, sowel as vir bedryfs- en kapitaaluitgawe beramings, om die uitvoerbaarheid van vleoidiagramme wat DOX insluit, te evalueer. Die piriet DOX kinetika is ondersoek in terme van temperatuur (180 tot 210 °C), suurstof parsiële druk (480 tot 1100 kPa), suurkonsentrasie (10 tot 50 g/L) en partikelgrootte (+38 tot 150 μm) deur gebruik te maak van ’n lot 2 gallon Parr outoklaaf. ’n Lotmodel is daaropvolgend in MATLAB ontwikkel en gebruik om die tempo afhanklikhede wat waargeneem is, te bevestig. Die suurstof gas-vloeistof massa-oordrag-koëffisiënt is ook onafhanklik gemeet gedurende die studie om ’n kwantitatiewe beskrywing van die opgeloste suurstofkonsentrasie gedurende modellering van die lot oksidasieproses, moontlik te maak. Die aktiveringsenergie is bereken as ~120 kJ/mol in verhouding met opgeloste suurstofkonsentrasie, wat aandui dat die reaksie deur ’n chemiese reaksie beheer is by die oppervlak van pirietpartikels en dat geen diffusie beperkinge van toepassing is nie. Die oksidasietempo het afgeneem met verhoogde suurkonsentrasie met ’n reaksietempo orde in suurkonsentrasie, met ’n bestek van tussen -0.2 en -0.3. Dis gevind dat die oksidasiekinetika relatief onsensitief is tot partikelgrootte by suurstof parsiële druk laer as 1000 kPa. Vir alle praktiese doeleindes is die pirietoksidasie as eerste orde in opgeloste suurstofkonsentrasie gevind, alhoewel hierdie aanname tot swak voorspelling van die yster(II)- en yster(III)-oplossingkonsentrasies gedurende modellering van die lot oksidasie toets gelei het. Akkurate kwantifisering van die yster(II)- en yster(III)-oplossingkonsentrasies sal ook belangrik wees om in ag te neem vir reaktorontwerp doeleindes, aangesien dit die maksimum hoeveelhede yster(III) en suur wat geproduseer kan word tydens DOX, sal dikteer. Simulasies het gewys dat verbeterde voorspellings van die yster(II)- en yster(III)-konsentrasies verkry word wanneer ’n direkte reaksie tussen piriet en yster(II) toegelaat word. Dit beteken dat beide opgeloste suurstof en yster(III) verantwoordelik is vir die oksidasie van piriet by tipiese DOX kondisies, dit is, die piriet kon ’n duale tempo afhanklikheid op die suurstof- en yster(III)-konsentrasies in twee additiewe tempo bepalende stappe hê. Regressie het gewys dat die eksperimentele data verteenwoordig word deur twee additiewe tempovergelykings met ordes van ~0.6-1.1 in opgeloste suurstofkonsentrasies en half-orde in yster(III)-konsentrasie. Die relatiewe bydra van die twee reaksies tot die algehele tempo lyk of dit beïnvloed word deur die pulpdigtheid, partikelgrootte, en suurstof parsiële druk. Dit word voorgestel dat ’n opvolg studie gedoen word om die tempo afhanklikheid van piriet in oplossings van wisselende yster(III)-konsentrasie te kwantifiseer, by temperature gebruik in hierdie studie en in die afwesigheid van opgeloste suurstof, sodat ’n gemete tempovergelyking, onafhanklik van die lot DOX model, verskaf kan word. Die homogene yster(II) na yster(III) oksidasietempo moet ook onafhanklik gemeet word om te bevestig dat die tempovergelyking wat gebruik is, korrek is. Verder moet die moontlike effek van sekondêre minerale, in hierdie geval pirofilliet, duidelik gemaak word deur eksperimentele werk by hoër pulpdigtheid te doen.af_ZA
dc.format.extent313 pages : illustrationsen_ZA
dc.language.isoen_ZAen_ZA
dc.publisherStellenbosch : Stellenbosch Universityen_ZA
dc.subjectUCTD
dc.subjectPyriteen_ZA
dc.subjectPressure -- Oxidationen_ZA
dc.subjectKineticsen_ZA
dc.titleThe high-temperature pressure oxidation of a Witwatersrand pyrite concentrateen_ZA
dc.typeThesisen_ZA
dc.rights.holderStellenbosch Universityen_ZA


Files in this item

Thumbnail
Thumbnail

This item appears in the following Collection(s)

ITEM VIEW