ENGLISH ABSTRACT: The sugar industry over the years has been producing sugarcane bagasse as part of the sugar milling process. Currently this sugar mill biomass is incinerated inefficiently as a means of their disposal to produce steam and electricity, which in most cases are only just enough to supply the energy required to run the mills, thereby leaving very little or no extra energy for sale to bring in extra income in addition to sales revenue from sugar. However, the recent instability and uncertainties in the price of sugar and the global call for a green and sustainable environment have necessitated the search for ways of making effective use of this biomass to supply sugar mill energy demands, while producing extra energy in the form of electricity and other energy products for sale and at the same time contributing towards environmental sustainability. The main objective of this work was to develop process models for the processing of sugar mill biomass into energy and energy products. Based on this, biomass to energy conversion process (BMECP) models have been developed for various process configurations of two thermochemical processes; Combustion and Fast Pyrolysis using the Aspen Plus® simulation software. The aim of process modelling was to utilizing sugar cane bagasse as an input energy source to supply the energy requirements of two sugar mill configurations (efficient and less efficient mills), while generating extra electricity and high valued energy products for sale. Four BMECP configurations; 30bar BPST, 40bar CEST, 63bar CEST and 82bar CEST systems were modelled for the combustion thermochemical process. For the fast pyrolysis thermochemical process, two process configurations: Pure Fast Pyrolysis BMECP and Partial Fast Pyrolysis BMECP were modelled. The former BMECP utilizes all available bagasse through fast pyrolysis to produce bio-oil and biochar alongside generating electricity as well as energy to run the sugar mill operations. In the latter BMECP model, only surplus bagasse after separation of the quantity needed to supply the sugar mill energy requirement and electricity production is used to produce bio-oil and biochar. The technical performance of the BMECP models have been analysed and compared based on steam and electricity production rates, process efficiencies and environmental impacts (based on CO2 savings). The effects of boiler operating pressure and bagasse moisture content on the performance of the combustion based BMECP models have also been investigated. Finally, detailed economic models have been developed using the Aspen Process Economic Analyzer (Icarus®) to assess the economic viability of the BMECP models and sensitivity analysis performed to study the response of the BMECP models to variations in economic parameters. Technical performance analysis shows the combustion based BMECP models perform better than the Pure Fast Pyrolysis and Partial Fast Pyrolysis BMECP models with regards to steam and electricity production, thereby giving them higher electrical efficiencies. The electricity generation rate has been shown to increase with increasing boiler operating pressure and decreasing bagasse moisture content while steam production rate has been shown to increase with decreasing bagasse moisture content and decreasing boiler operating pressure. Despite the lower electrical efficiencies of the fast pyrolysis based BMECP models, the analysis shows that their overall process efficiencies compare very well with those of the combustion based BMECP models due to the production of high energy value pyrolysis products. Based on common operating pressure and 50% bagasse moisture content, the Pure Fast Pyrolysis and the Partial Fast Pyrolysis models have proved to be more environmental friendly with hourly CO2 savings of 40.44 and 41.30 tons for the Partial Fast Pyrolysis BMECP and the Pure Fast Pyrolysis BMECP respectively based on a 300 ton of sugarcane/h (81 ton bagasse/h) plant size. From an economic point of view, biomass combustion based on the 63bar CEST BMECP model has proved to be the most economically viable option under current economic conditions. First order total capital investment estimate for this BMECP is about $116 million, producing NPV of $390 million at the end of a 20 year plant life and IRR of 34.51%. The Pure Fast Pyrolysis BMECP model is the least economic viable option. Sensitivity analysis shows this BMECP model is the most sensitive to changes in bagasse and electricity prices; recording -191.61/+446.86% change in NPV for a ±30% change in bagasse price and -91.5/+338.60% for a ±30% change in electricity price. / AFRIKAANSE OPSOMMING: Die afgelope jare het suikerriet-afval (bagasse) by suikermeule ‘n belangrik byproduk van die suiker-industie geraak. Tans word hierdie afval of biomasse verbrand in die suikermeule se poging om stoom en elektrisiteit op te wek; maar die die proses is oneffektief. Die hoeveelheid energie wat opgewek word, is skaars genoeg om die suikermeule self aan die gang te hou; daar is feilik geen sprake ‘n surplus energie waaruit ekstra inkomste verkry kan word toevoegend tot inkomste uit die suiker verkope self. Die huidige onstabiele suikerprys en gepaardgaande onsekerhede sowel as die werêldwye oproep vir ‘n groen- en volhoubare omgewing, noodsaak ‘n nuwe soeke na effektiewe manier om die afvalmateriaal sinvol te verwerk. Die tipe effektiwiteit van verwerking waarna gesoek word moet die volgende uitkomste hê: verskaffing van genoeg energie tydens produksie aan die suikermeuele self; vervaardiging van ekstra energie in die vorm van eletrisieteit en ander energie produkte. Terselfder moet die ook bydra tot die volhoubaarheid van die omgewing. Die grootste gedeelte van hierdie navorsing is gewy aan die ontwikkeling van “proses modelle” om suikemeule afval (bagasse) te omskep in energie en energie-produkte. Om hierdie doel te bereik, is biomassa-tot-energie omskeppingsproses- modelle (BMECP) ontwikkel om verskeie proses konfigurasies van twee termo-chemiese prosesse, naamlik Verbranding (Combustion), en Vinnige Pirolise (Fast Pyrolysis) deur die gebruik van die ‘Aspen Plus®’- simulasie sagteware. Die doel van die proses modelering was om suikerriet biomassa as ‘n bron van energie te gebruik om weer die energie benodighehede van twee denkbeeldige suikermeule vas te stel; een meul is voorgestel as effektief, die ander as minder effektief. Terselfdertyd is gekyk na die hoeveelheid ekstra energie wat elkeen sou opwek en ander hoogs waardevolle energie produkte om te verkoop (bv. ‘bio-olies en bio-char’). Vier “BMECP” konfigurasies (voorstellings) 30bar BPST, 40bar CEST, 63bar CEST en 82bar CEST sisteme is gemodelleer vir die Verbranding termo-chemiese proses. In die geval van die Pirolise (Pyrolysis) termo-chemiese proses, is twee proses konfigurasies gemodelleer: 1. Suiwer Vinnige Pyrolyise BMECP en 2. Gedeeltelik Vinnige Pirolise BMECP. In die geval van eersgenoemde, word alle beskikbare ‘bagasse’ deur vinnige pirolise omskep om ‘bio-olie’ en ‘bio-char’ te vervaardig.Verder wek dit ook elektrisiteit op so wel as die nodige energie om die suikermeule te laat opereer. In die geval van die Gedeeltlike Vinnige Pirolise BMECP , moet daar eers genoegsame ‘bagasse’ opsy gesit word om die suikermeule van genoegsame energie te voorsien vir die volle funskionering daarvan en elektrisiteit-opwekking. Van die surplus of oorblywende ‘bagasse’ kan dan gebruik word om ‘bio-olie’ en ‘biochar’ te produseer. Die tegniese prestasie van al die BMECP modelle is geanaliseer en vergelyk ten opsigte van stoom en elektrisiteits-opwekking; proses effektiewiteit asook die impak op die omgewing ( gebaseer op CO2 –besparings). Die effek van stoomkettel-druk tydens operering asook die bagasse se vog-inhoud. Op die prestasie van die verbrandingsgebaseerde modelle is ook ondersoek. Laastens, uitgebreide ekonomeidese modelle is ook ontwikkel deur die gebruik van die ‘Aspen Process Economic Analyser (Icarus®)’. Sodoende is die ekonomiese vatbaarheid van die BMECP modelle ondersoek. Hierdie sagteware help ook met. Sensitiwiteits-analise in die bestudering van die terugvoer van die BMECP modelle tot veranderlikes in ekonomiese parameters. Rakende effektiwiteit, toon die uitslae dat die verbrandings-gebaseerde BMECP modelle beter vaar as die met betrekking tot stoom- en elektrisiteits-opwekking. Verbrandings-gebaseerde-modelle toon hoër elektriese effektiwiteit. Indien die vog-inhoud van die bagasse laag was en die tempo van stoomketel operasie druk verhoog is, het die tempo van elektriesiteits-opwekking ook gestyg. Ten opsigte van stoom daarenteen, het die stoom-opwekking tempo verhoog in die die vogl inhou van diebagasse laag was asook verminderde stoomketel operering druk. Ten spyte van die laer elektriese effektiewiteit van die Suiwer Vinnig- en Gedeeltelik Vinnig BMECP modelle, dui die analise aan dat hul proses effektiewiteit in die geheel Goed vergelyk met die van die verbrandings-gebaseerde BMECP modelle. Dit is toe te skryf aan die produksie van die hoë-energie draende pirolise produkte. Gebaseer op algemene operering druk van 50% ‘bagasse’ vog-inhoud, het die bogenoemde twee modelle bewys om meer omgewings-vriendelik te wees met uurlikse CO2-besparings. In die geval van Gedeeltelike Vinnige Pirolise BMECP, 40.44 en vir die Suiwer Vinnige Pirolise BMECP 41.30 gebaseer op ‘n 300 ton suikerriet/h (81 ton bagasse/h) plantasie-grote. Ten slotte, vanuit ‘n ekonomiese oogpunt, blyk ‘n biomassa verbranding gebaseer op die 63 bar CEST BMECP model die mees ekonomies-vatbare opsie onder huidige ekonomiese omstandighede. Eerste orde totale kapitale belegging beraming vir hierdie BMECP is ongeveer $116 miljoen, produksie NPV is $390 miljoen aan die einde van ‘n 20 jaar tydperk vir ‘n suikerriet-aanleg. IRP is 34.51%. Die Suiwer Vinnige Pirolise BMECP is die mins-ekonomiese vatbare model. Sensitiewiteits-analises het getoon dat hierdie BMECP model baie sensitief is ten opsigte van verandering in die pryse van bagasse en elektrisieteit; in die geval van NPV is veranderinge van -191.61/+446.86% aangedui op ‘n ±30% verandering in bagasse pryse. In die geval van elektrisieteitspryse, is ‘n sensitiewiteit van van -91.5/+338.60% op ‘n ±30% prysverandering getoon.

Thesis (MScEng)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: In this study, tool forces from rock cutting tests were numerically simulated through a discrete element method (DEM) in association with PFC3D™. Tribological interactions such as contact, shearing, fracturing, friction and wear were presented during these cutting simulations. Particle assemblies, representing Paarl granite and Sandstone-2, were created in PFC3D™ through a material-genesis procedure. The macro-properties of these particle assemblies, namely Young’s modulus, Poisson’s ratio, uniaxial and triaxial compressive strength and Brazilian tensile strength, were calibrated by modelling the uniaxial and triaxial compressive strength test and the Brazilian tensile strength test. The calibration was done through adjustment of the micro-properties of the assembly, namely the stiffness and strength parameters of the particles and bonds. The influence of particle size on the calibration was also investigated. These assemblies were used in the rock cutting tests. Results suggested that DEM can reproduce the damage formation during calibration tests successfully. From the results obtained from the calibration tests, it was also concluded that particle size is not a free parameter but influences the macro-properties greatly. Different rock cutting tools were simulated, namely point-attack (conical) picks, chisel-shaped tools and button-shaped tools. The numerical cutting tools were treated as rigid walls to simplify the simulation and the tool forces were not influenced by wear. In each simulation the cutting tools advanced at a constant velocity. The tool forces acting on the cutting tool, in three orthogonal directions, were recorded during the numerical simulations and the peak cutting forces were predicted by theoretical equations. The damage to the Paarl granite and Sandstone-2 assemblies was revealed as broken bonds, which merge into microscopic fractures. The mean peak cutting forces of sharp cutting tools obtained from numerical, theoretical and experimental models (from the literature) were compared. Finally the influence of factors, including wear on the tool and depth of cut, on the value of tool forces was also investigated. The results from the rock cutting tests revealed that the correlation between the numerical and the experimental models as well as the theoretical and experimental models was not strong when using sharp point-attack and chisel-shaped picks. It was concluded that the influence of wear plays a substantial part in the cutting process and it has to be included during the numerical simulation for the results to be accurate and verifiable. This study also found that there is a non-linear increase in tool forces with an increase in depth of cut, since the contact area increases. At larger cutting depths, chip formation also generally increased and therefore damage to the sample as well as wear on the cutting tool will be minimized at shallow cutting depths. Overall this study concludes that DEM are capable of simulating calibration methods and rock cutting processes with different cutting tools and producing results which are verifiable with experimental data. Therefore numerical prediction of tool forces will allow the design of efficient cutting systems and the operational parameters as well as the performance prediction of excavation machines. / AFRIKAANSE OPSOMMING: In hierdie studie is die kragte wat tydens rotssny-toetse op die sny gereedskap inwerk, numeries gesimuleer met behulp van ‘n diskrete element metode (DEM) in samewerking met PFC3D™. Tribologiese interaksies soos kontak, skeer, breking, wrywing en slytasie is gedurende hiersie snytoetse voorgestel. Partikel versamelings, wat Paarl graniet en Sandsteen-2 verteenwoordig, is in PFC3D™ geskep deur middel van ‘n materiaal-skeppings prosedure. Die makro-eienskappe van die partikel versamelings, naamlik Young se modulus, Poisson se verhouding, eenassige en drie-assige druksterkte en Brasiliaanse treksterkte, is gekalibreer deur modellering van die eenassige en drie-assige druksterkte toets en die Brasiliaanse treksterkte toets. Die kalibrasie is gedoen deur aanpassing van die mikro-eienskappe, naamlik die styfheid en die sterkte parameters van die partikels en bindings. Die invloed van partikelgrootte is ook ondersoek. Daarna is hierdie versamelings in die rotssny-toetse gebruik. Resultate het daarop gedui dat DEM die kraakvorming gedurende kalibrasie toetse suksesvol kan reproduseer. Vanuit die kalibrasie is ook gevind dat die partikelgrootte nie ‘n vrye parameter is nie, maar die makro-eienskappe grotendeels beïnvloed. Verskillende rotssny gereedskap is gesimuleer, naamlik koniese, beitel-vormige en knopie-vormige instrumente. Die numeriese sny gereedskap is gesimuleer as rigiede mure om simulasies te vereenvoudig en die gereedskap-kragte is dus nie deur slytasie beïnvloed nie. Tydens elke simulasie is die sny gereedskap vorentoe beweeg teen ‘n konstante snelheid. Die gereedskap-kragte, in drie ortogonale rigtings, is aangeteken gedurende die numeriese simulasies en die piek snykragte is ook voorspel deur teoretiese vergelykings. Die skade aan die Paarl graniet en Sandsteen-2 versamelings, is voorgestel as gebreekte bindings, wat saamsmelt tot mikroskopiese frakture. Die gemiddelde piek snykragte van skerp sny gereedskap van numeriese, teoretiese en eksperimentele modelle (uit die literatuur) is vergelyk. Ten slotte is die invloed wat faktore, onder andere die slytasie van gereedskap en die snydiepte, op die grootte van die kragte het ondersoek. Die resultate van die rotssny-toetse het aan die lig gebring dat die korrelasie tussen die numeriese en eksperimentale modelle sowel as die teoretiese en eksperimentele modelle nie sterk is tydens die gebruik van skerp koniese en beitel-vormige instrumente nie. Die gevolgtrekking is gemaak dat die invloed van slytasie van sny gereedskap ‘n wesenlike rol speel in die snyproses en dat dit in die numeriese simulasie ingesluit moet word sodat die resultate akkuraat en virifieerbaar is. Hierdie studie het ook gevind dat daar ‘n nie-lineêre toename in die gereedskap-kragte is met ‘n toename in snydiepte aangesien die kontak-area toeneem met ‘n toename in die snydiepte. By groter snydieptes, het die formasie van afsplinterings verhoog en dus sal skade aan die partikel versamelings en die slytasie van die gereedskap geminimeer word by vlakker snydieptes. Algeheel het die studie tot die gevolgtrekking gekom dat DEM in staat is om kalibrasie metodes en rotssny-toetse met verskillende sny gereedskap te simuleer asook om resultate te produseer wat verifieerbaar is met eksperimentele data. Numeriese voorspellings van die gereedskap-kragte sal dus toelaat om doeltreffende sny prosesse en operasionele parameters te ontwerp sowel as om die werkverrigting van uitgrawings masjiene te voorspel.

Thesis (MScEng)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: In recent times, governments around the world have placed increasing focus on cleaner technologies and sustainable methods of power generation in an attempt to move away from fossil fuel derived power, which is deemed unsustainable and unfriendly to the environment. This trend has also been supported by the South African government, with clear intentions to diversify the country’s power generation by including, among others, biomass as a renewable resource for electricity generation. Woody biomass and associated forestry residues in particular, could potentially be used as such a renewable resource when considering the large amount of fast growing hardwood species cultivated in South Africa. Approximately 6.3 million ton of Eucalyptus grandis is sold annually for pulp production while a further 7 million ton of Eucalyptus species are sold as round wood. With these tree species reaching commercial maturity within 7 – 9 years in the South African climate, there is real potential in harnessing woody biomass as a renewable energy source. In this study, pyrolysis was investigated as a method to condense and upgrade E.grandis into energy and chemical rich products. The pyrolysis of E.grandis is the study of the thermal degradation of the biomass, in the absence of oxygen, to produce char and bio-oil. The thermal degradation behaviour of E.grandis was studied using thermo-gravimetric analysis (TGA) at the Karlsruhe Institute of Technology (KIT) in Germany and subsequently used to determine the isoconversional kinetic constants for E.grandis and its main lignocellulosic components. Slow, Vacuum and Fast Pyrolysis were investigated and optimised to maximise product yields and to identify the key process variables affecting product quality. The Fast Pyrolysis of E.grandis was investigated and compared on bench (KIT0.1 kg/h), laboratory (SU1 kg/h) and pilot plant scale (KIT10 kg/h), using Fast Pyrolysis reactors at Stellenbosch University (SU) in South Africa and at KIT in Germany. The Slow and Vacuum Pyrolysis of E.grandis was investigated and compared using a packed bed reactor at Stellenbosch University. The TGA revealed that biomass particle size had a negligible effect on the thermal degradation behaviour of E.grandis at a heating rate set point of 50 °C/min. It was also shown that increasing the furnace heating rates shifted the thermo-gravimetric (TG) and differential thermo-gravimetric (DTG) curves towards higher temperatures while also increasing the maximum rate of volatilisation. Lignin resulted in the largest specific char yield and also reacted across the widest temperature range of all the samples investigated. The average activation energies found for the samples investigated were 177.8, 141.0, 106.2 and 170.4 kJ/mol for holocellulose, alpha-cellulose, Klason lignin and raw E.grandis, respectively. Bio-oil yield was optimised at 76 wt. % (daf) for the SU1 kg/h Fast Pyrolysis plant using an average biomass particle size of 570 μm and a reactor temperature of 470 °C. Differences in the respective condensation chains of the various Fast Pyrolysis reactor configurations investigated resulted in higher gas and char yields for the KIT reactor configurations compared to the SU1 kg/h Fast Pyrolysis plant. Differences in the vapour residence time between Slow (>400 s) and Vacuum Pyrolysis (< 2 s) resulted in a higher liquid and lower char yield for Vacuum Pyrolysis. Local liquid yield maxima of 41.1 and 64.4 wt. % daf were found for Slow and Vacuum Pyrolysis, respectively (achieved at a reactor temperature of 450 °C and a heating rate of 17 °C/min). Even though char yields were favoured at low reactor temperatures (269 – 300 °C), the higher heating values of the char were favoured at high reactor temperatures (29 – 34 MJ/kg for 375 – 481 °C). Reactor temperature had the most significant effects on product yield and quality for the respective Slow and Vacuum Pyrolysis experimental runs. The bio-oils yielded for SP and VP were found to be rich in furfural and acetic acid. / AFRIKAANSE OPSOMMING: Regerings regoor die wêreld het in die afgelope tyd toenemende fokus geplaas op skoner tegnologie en volhoubare metodes van kragopwekking in 'n poging om weg te beweeg van fossielbrandstof gebasseerde energie, wat geag word as nie volhoubaar nie en skadelik vir die omgewing. Hierdie tendens is ook ondersteun deur die Suid-Afrikaanse regering, met 'n duidelike bedoeling om die land se kragopwekking te diversifiseer deur, onder andere, biomassa as 'n hernubare bron vir die opwekking van elektrisiteit te gebruik. Houtagtige biomassa en verwante bosbou afval in die besonder, kan potensieel gebruik word as so 'n hernubare hulpbron, veral aangesien ‘n groot aantal vinnig groeiende hardehout spesies tans in Suid-Afrika verbou word. Ongeveer 6,3 miljoen ton Eucalyptus grandis word jaarliks verkoop vir pulp produksie, terwyl 'n verdere 7 miljoen ton van Eucalyptus spesies verkoop word as paal hout. Met hierdie boom spesies wat kommersiële volwassenheid bereik binne 7 - 9 jaar in die Suid-Afrikaanse klimaat, is daar werklike potensiaal vir die benutting van houtagtige biomassa as 'n hernubare energiebron. In hierdie studie is pirolise ondersoek as 'n metode om E.grandis te kondenseer en op te gradeer na energie en chemikalie ryke produkte. Die pirolise van E.grandis is die proses van termiese afbreking van die biomassa, in die afwesigheid van suurstof, om houtskool en bio-olie te produseer. Die termiese afbrekingsgedrag van E.grandis is bestudeer deur gebruik te maak van termo-gravimetriese analise (TGA) by die Karlsruhe Instituut van Tegnologie in Duitsland en daarna gebruik om die kinetiese konstantes vir die iso-omskakeling van E.grandis en sy hoof komponente te bepaal. Stadige, Vakuum en Snel pirolise is ondersoek en geoptimiseer om produk opbrengste te maksimeer en die sleutel proses veranderlikes wat die kwaliteit van die produk beïnvloed te identifiseer. Die Snel Pirolise van E.grandis is ondersoek en vergelyk op bank- (KIT0.1 kg / h), laboratorium- (SU1 kg / h) en proefaanlegskaal (KIT10 kg / h) deur gebruik te maak van Snel pirolise reaktore by die Universiteit van Stellenbosch (US) in Suid-Afrika en die Karlsruhe Instituut van Tegnologie (KIT) in Duitsland. Die Stadige en Vakuum Pirolise van E.grandis is ondersoek en vergelyk met behulp van 'n gepakte bed reaktor aan die Universiteit van Stellenbosch. Die TGA studie het openbaar dat biomassa deeltjiegrootte 'n onbeduidende uitwerking op die termiese afbrekingsgedrag van E.grandis het by 'n verhittings tempo van 50 ° C / min. Dit is ook bewys dat die verhoging van die oond verwarming tempo die termo-gravimetriese (TG) en differensiële termo-gravimetriese (DTG) kurwes na hoër temperature verskuif, terwyl dit ook die maksimum tempo van vervlugtiging laat toeneem het. Lignien het gelei tot die grootste spesifieke houtskool opbrengs en het ook oor die wydste temperatuur interval gereageer van al die monsters wat ondersoek is. Die gemiddelde aktiveringsenergieë vir die monsters wat ondersoek is, was 177,8, 141,0, 106,2 en 170,4 kJ / mol, onderskeidelik vir holosellulose, alpha-sellulose, Klason lignien en rou E.grandis. Bio-olie opbrengs is geoptimeer teen 76 wt. % (DAF) vir die SU1 kg / h Snel Pirolise aanleg met behulp van 'n gemiddelde biomassa deeltjiegrootte van 570 μm en 'n reaktor temperatuur van 470 ° C. Verskille in die onderskeie kondensasie kettings van die verskillende Snel Pirolise aanlegte wat ondersoek is, het gelei tot hoër gas- en houtskool opbrengste vir die KIT reaktor konfigurasies in vergelyking met die SU1kg/h FP plant. Verskille in die damp retensie tyd tussen Stadige (> 400 s) en Vakuum pirolise (<2 s) het gelei tot 'n hoër vloeistof en laer houtskool opbrengs vir Vakuum Pirolise. Plaaslike vloeistof opbrengs maksima van 41,1 en 64,4 wt. % (daf) is gevind vir Stadig en Vakuum pirolise onderskeidelik, bereik by 'n reaktor temperatuur van 450 ° C en 'n verhittingstempo van 17 ° C / min. Selfs al is houtskool opbrengste bevoordeel by lae reaktor temperature (269 - 300 ° C), is die hoër warmte waardes van die houtskool bevoordeel deur hoë reaktor temperature (29 - 34 MJ / kg vir 375 - 481 ° C). Reaktor temperatuur het die mees beduidende effek op die produk opbrengs en kwaliteit vir die onderskeie Stadige Pirolise en Vakuum Pirolise eksperimentele lopies gehad. Die bio-olies geproduseer tydens Stadige en Vakuum Pirolise was ryk aan furfuraal en asynsuur.

Thesis (MScEng)--Stellenbosch University, 2014. / ENGLISH ABSTRACT: Alkanes can be used as an inexpensive feedstock to produce more valuable alcohols. The biotransformation of alkanes to alcohols provides an alternative to conventional chemical procedures. The scope of this research was to develop a process utilising a biocatalyst to catalyse the oxidation of an alkane to its corresponding alcohol on a larger scale than had been reported on in previous research. The research utilised a recombinant E. coli BL21(DE3) cell, containing the CYP153A6 operon in pET 28 vector, as the biocatalyst. The CYP153A6 enzyme catalyses the oxidation of octane to 1-octanol. The principle objective of the research was to determine the amount of 1-octanol that can be produced by a system utilising this strain of recombinant E. coli as a biocatalyst on a three orders of magnitude larger scale than what had previously been reported on for this reaction system. An additional objective was to model the 1-octanol production performance in the bioreactor. Bioconversion batch reactions, with excess octane used as a substrate, were conducted in 30ml McCartney bottles and in a 7.5L BioFlo 110 Modular Benchtop Fermentor (New Brunswick). The McCartney bottles were not equipped to actively control process conditions.The bioreactor was equipped to control process conditions such as temperature, pH and dissolved oxygen concentration. Experiments in the bioreactor were therefore described as being performed under controlled conditions. The procedures used to grow, maintain and harvest the biocatalyst cells were based on those developed by the Department of Microbial, Biochemical and Food Biotechnology at the University of the Free State. The product and substrate concentrations were determined through gas chromatography (GC) analysis. The McCartney bottle bioconversion reactions, with a 1.33ml reaction volume, produced 1.88 mg 1-octanol per gram of dry cell weight per hour. The bioreactor under controlled conditions, with a 2L reaction volume, produced 14.89 mg 1-octanol per gram of dry cell weight per hour. The formation of a secondary product, octanoic acid, was observed for the bioreactor under controlled conditions experiment at a production of 1.12 mg per gram of dry cell weight per hour. The McCartney bottle experiments did not produce any by-products. The 1-octanol production performance in the bioreactor experiments was empirically modelled. The empirical rate law was based on the form of the Monod equation, with the addition of a product inhibition term. The model achieved an average Root Mean Square Error of less than 5% when compared to experimental data, and was therefore concluded to be accurate within the range of experimental data and conditions tested for. The principal finding of the research is that the cells produced an order of magnitude more product in the bioreactor than in the McCartney bottles. The literature on this reaction system, however, reports only on smaller scale research than that performed in the bioreactor. The improved production results in the bioreactor therefore give the first insight into the potential that this technology has for being scaled up. Of equal significance is the finding that a secondary product developed during the biotransformations performed in the bioreactor. This refutes the assumption that the biocatalyst cells are unable to catalyse any secondary reactions. This aspect of the cells’ performance must be addressed before the biocatalyst cell strain can be considered to be a viable option for utilisation in large-scale processes. / AFRIKAANSE OPSOMMING: Alkane kan gebruik word as ‘n bekostigbare bron om meer waardevolle alkohol te produseer. Die biotransformasie van alkane na alkohol bied dus ‘n alternatief vir konvensionele chemiese prosedure. Die oogmerk en omvang van hierdie navorsing was om ‘n proses te ontwikkel waarin ‘n biokatalisator gebruik word om die oksidasie van ‘n alkaan tot sy ooreenstemmende alkohol te kataliseer, en om vas te stel hoeveel 1-oktanol vervaardig kan word deur ‘n herverenigde E. coli as katalisator gebruik. ‘n Rekombinante E. coli BL21(DE3) sel, wat die CYP153A6 operon in pET 28 vector bevat, is as biokatalisator gebruik. Die CYP153A6 ensiem kataliseer die oksidasie van oktaan na 1-oktanol. Biokonversie lot-reaksies, met oormatige oktaan wat as substraat gebruik word, is in 30ml McCartney bottels en in 7.5L BioFlo 110 Modular Benchtop Fermentor (New Brunswick) uitgevoer. Die bioreaktor was toegerus om kondisies van die proses soos temperatuur, pH and opgeloste suurstof-konsentrasie te kontroleer. Die prosedures wat gebruik is om die groei, onderhoud en oes van die biokatalisator selle te bewerkstellig, is gebaseer op prosedures wat ontwikkel is deur the Department van Microbiese, Biochemiese and Voedsel Biotegnologie van die Universiteit van die Vrystaat. Die produk- en substraat-konsentrasies is vasgestel deur gaschromatografie (GC) ontleding. Die McCartney bottel biokonversie-reaksie met ‘n 1.33ml reaksie-volume het 1.88 mg 1-oktanol per gram droeë-sel gewig opgelewer. Die bioreaktor, wat onder beheerde toestande ‘n 2L reaksie-volume het, het 14.89 mg 1-octanol per gram droeë-sel gewig gelewer. Onder beheerde eksperimentele kondisies is die vorming van ‘n sekondere produk, oktanol-suur, by die bioreaktor waargeneem teen 1.23 mg per gram droeë-sel gewig per uur. Die McCartney bottel eksperimente egter het geen newe-produkte opgelewer nie. Die ontwikkeling van die 1-oktanol in die bioreaktor-ekperimente is empiries gemodelleer. Die empiriese ‘rate law’ is gebaseer op ‘n vorm van die Monod- vergelyking, met byvoeging van ‘n produk-inhiberingsterm. Die model het ‘n gemiddelde vierkantswortel foutvariansie van minder as 5% opgelewer, vergeleke met die eksperimentele data, en word dus binne die rykwydte van die eksperimentele data, en die kondisies waarvoor getoets is, as akkuraat beskou. Die belangrikste bevinding is dat die selle in die bioreaktor ‘n orde van grootte meer produk gelewer het as die selle in die McCartney bottels. Die literatuur oor hierdie reaksie-sisteem berig egter slegs oor kleiner skaalse navorsing as wat in die bioreaktor gedoen is. Die verbeterde opbrengsresultate van die bioreaktor dui daarop dat laasgenoemde tegnologie die potensiaal inhou om opgegradeer te word. Die bevinding dat ‘n sekondere produk in die biotransformasie in die bioreaktor gevorm het, is beduidend. Dit weerspreek die aanname dat die biokatalisator-selle nie sekondere reaksies kataliseer nie. Hierdie aspek moet aangespreek word alvorens die biokataliseer-selle oorweeg kan word as ‘n lewensvatbare alternatief vir gebruik in grootskaalse prosesse.

Thesis (MEng)--Stellenbosch University, 2014. / ENGLISH ABSTRACT: In Base Metal Refineries (BMRs), the copper sulphate leach solution produced during the final pressure leaching stage contains impurities such as selenium and tellurium, as well as other precious metals (OPMs, namely Rh, Ru, and Ir). Se and Te are removed by precipitation with sulphur dioxide prior to electrowinning of Cu. While a small percentage of the dissolved OPMs precipitate with the Se and Te, the largest portion remains in solution and is recycled to the first stage leach. If a larger portion of the OPMs in solution can be recovered in the Se/Te precipitation stage, OPM losses and the OPM inventory of the plant can be reduced. The aim of this project was to determine operating conditions that would allow maximum OPM recovery with minimal Cu and Ni co-precipitation in the Se/Te removal section of a BMR. The effects that the operating temperature, pressure, stirring rate, reagent type, and reagent quantity have on the metal precipitation behavior and precipitate characteristics were determined experimentally. Thio-urea and sulphurous acid were evaluated as precipitation reagents for temperatures of 80°C and 160°C, stirring rates of 250 rpm and 500 rpm, and pressures equal to ambient pressure and 7 bar. 200 % and 320 % excess thio-urea and 720 % and 960 % excess sulphur dioxide were used. The precipitation of OPMs with sulphur dioxide was generally poor; the maximum percentage Rh, Ru, and Ir precipitated were 35 %, 18 %, and 20 %, respectively. It was, however, found that the OPM precipitation increased as the reagent amount was increased. Increasing the temperature further increased Rh and Ir precipitation but affected the Ru precipitation negatively. Thio-urea precipitated virtually all of the Rh contained in the solution irrespective of the values of the process variables studied. As was the case with sulphur dioxide, increasing the amount of thio-urea added resulted in increased Ru and Ir precipitation, while higher temperatures favored Ir precipitation but not Ru precipitation. The maximum percentage Ru and Ir precipitation achieved with thio-urea were 87 % and 60 %, respectively. Complete Se precipitation was observed at all process conditions, while Te precipitation increased as the operating temperature and the reagent quantity were increased. Maximum Te precipitation of 98 % and 90 % were achieved when using 320 % excess thio-urea and 960 % sulphurous acid quantities, respectively, at 160°C and a stirring rate of 250 rpm. Increasing the reagent quantity and temperature did, however, also result in increased copper and nickel precipitation. The statistical analysis of the results allowed regression models to be fitted to predict the percentage metal precipitation as a function of the investigated process variables. These models were used to define an objective function to determine the optimal operating conditions. A temperature of 80oC, a pressure of 7 bar, and 200 % excess thio-urea were proposed as the optimum operating conditions that would yield 98 % Rh, 75 % Ru, and 48 % Ir precipitation with less than 5 % Cu and Ni co-precipitation. Experimental validation tests confirmed the model predicted values and proved repeatability of the experimental data. / AFRIKAANSE OPSOMMING: Die kopersulfaat logingsoplossing wat tydens die finale drukloging stadium in Basis Metaal Raffinaderye (BMRe) produseer word, bevat onsuiwerhede soos selenium en tellurium sowel as ander edelmetale (AEMe, naamlik Rh, Ru, en Ir). Se en Te word voor Cu elektrowinning verwyder deur middel van presipitasie met swaweldioksied. Alhoewel ʼn klein persentasie van die opgeloste AEMe saam met die Se en Te presipiteer, bly die grootste gedeelte in oplossing en word gevolglik na die eerste loging stadium hersirkuleer. AEM verliese en die AEM inventaris van die aanleg kan verminder word indien ʼn groter gedeelte van die AEMe in die Se/Te presipitasie stadium herwin kan word. Die doel van hierdie projek was om bedryfstoestande te bepaal om maksimum AEM herwinning met minimale Cu en Ni kopresipitasie in die Se/Te verwyderingseksie van ʼn BMR te behaal. Die effekte wat bedryfstoestande soos temperatuur, druk, roerder tempo, tipe reagens, en hoeveelheid reagens op die metaal presipitasiegedrag en presipitaat eienskappe het, is eksperimenteel bepaal. Tio-ureum en swaweligsuur is evalueer as presipitasie reagense vir temperature van 80°C en 160°C, roerder tempo’s van 250 rpm en 500 rpm, en drukke gelyk aan omgewingsdruk en 7 bar. 200 % en 320 % oormaat tio-ureum en 720 % en 960 % oormaat swaweldioksied is gebruik. Die presipitasie van AEMe met swaweldioksied was swak in die algemeen; die maksimum persentasie Rh, Ru, en Ir presipitasie wat behaal is, is 35 %, 18 %, en 20 %, onderskeidelik. Daar is egter gevind dat die AEM presipitasie toeneem indien die hoeveelheid reagens toeneem. ʼn Toename in die temperatuur het verder tot ʼn toename in Rh en Ir presipitasie gelei, maar dit het Ru presipitasie negatief affekteer. Tio-ureum het basies al die Rh in oplossing laat presipiteer, ongeag die waardes van die ander prosesveranderlikes wat ondersoek is. Soos wat die geval vir swaweldioksied was, het ʼn toename in die hoeveelheid tio-ureum ʼn toename in die Ru en Ir presipitasie tot gevolg gehad, terwyl hoër temperature Ir presipitasie bevoordeel en Ru presipitasie benadeel het. Die maksimum persentasie Ru en Ir presipitasie wat met tio-ureum behaal is, is 87 % en 60 %, onderskeidelik. Volledige Se presipitasie is by alle proses toestande waargeneem, terwyl Te presipitasie toegeneem het soos wat die temperatuur en die hoeveelheid reagens toegeneem het. Maksimum Te presipitasie van 98 % en 90 % is behaal toe 320 % oormaat tio-ureum en 960 % oormaat swaweligsuur, onderskeidelik, by 160°C en ʼn roerder tempo van 250 rpm gebruik is. ʼn Toename in die hoeveelheid reagens en die temperatuur het egter ook meer koper en nikkel presipitasie tot gevolg gehad. Die statistiese analise van die resultate het dit moontlik gemaak om regressie modelle te pas om die persentasie metaal presipitasie as ʼn funksie van die ondersoekte veranderlikes te voorspel. Hierdie modelle is gebruik om ʼn doelfunksie te definieer ten einde die optimale bedryfstoestande te bepaal. ʼn Temperatuur van 80°C, ʼn druk van 7 bar, en 200 % oormaat tioureum is voorgestel as die optimale bedryfstoestande wat 98 % Rh, 75 % Ru, en 48 % Ir presipitasie met minder as 5 % Cu en Ni kopresipitasie tot gevolg sal hê. Eksperimentele geldigheidsbepalingtoetse het die waardes wat deur die modelle voorspel is bevestig en die herhaalbaarheid van die eksperimentele data bewys.

Thesis (MScIng)--University of Stellenbosch, 2003. / ENGLISH ABSTRACT: The analysis of process data obtained from chemical and metallurgical engineering systems is a crucial aspect of the operating of any process, as information extracted from the data is used for control purposes, decision making and forecasting. Singular spectrum analysis (SSA) is a relatively new technique that can be used to decompose time series into their constituent components, after which a variety of further analyses can be applied to the data. The objectives of this study were to investigate the abilities of SSA regarding the filtering of data and the subsequent modelling of the filtered data, to explore the methods available to perform nonlinear SSA and finally to explore the possibilities of Monte Carlo SSA to characterize and identify process systems from observed time series data. Although the literature indicated the widespread application of SSA in other research fields, no previous application of singular spectrum analysis to time series obtained from chemical engineering processes could be found. SSA appeared to have a multitude of applications that could be of great benefit in the analysis of data from process systems. The first indication of this was in the filtering or noise-removal abilities of SSA. A number of case studies were filtered by various techniques related to SSA, after which a number of neural network modelling strategies were applied to the data. It was consistently found that the models built on data that have been prefiltered with SSA outperformed the other models. The effectiveness of localized SSA and auto-associative neural networks in performing nonlinear SSA were compared. Both techniques succeeded in extracting a number of nonlinear components from the data that could not be identified from linear SSA. However, it was found that localized SSA was a more reliable approach, as the auto-associative neural networks would not train for some of the data or extracted nonsensical components for other series. Lastly a number of time series were analysed using Monte Carlo SSA. It was found that, as is the case with all other characterization techniques, Monte Carlo SSA could not succeed in correctly classifying all the series investigated. For this reason several tests were used for the classification of the real process data. In the light of these findings, it was concluded that singular spectrum analysis could be a valuable tool in the analysis of chemical and metallurgical process data. / AFRIKAANSE OPSOMMING: Die analise van chemise en metallurgiese prosesdata wat verkry is vanaf chemiese of metallurgiese ingenieursstelsels is ‘n baie belangrike aspek in die bedryf van enige proses, aangesien die inligting wat van die data onttrek word vir prosesbeheer, besluitneming of die bou van prosesmodelle gebruik kan word. Singuliere spektrale analise is ‘n relatief nuwe tegniek wat gebruik kan word om tydreekse in hul onderliggende komponente te ontbind. Die doelwitte van hierdie studie was om ‘n omvattende literatuuroorsig oor die ontwikkeling van die tegniek en die toepassing daarvan te doen, beide in die ingenieursindustrie en in ander navorsingsvelde, die navors van die moontlikhede van SSA aangaande die verwydering van geraas uit die data en die gevolglike modellering van die skoon data te ondersoek, ‘n ondersoek te doen na sommige van die beskikbare tegnieke vir nie-lineêre SSA en laastens ‘n studie te maak van die potensiaal van Monte Carlo SSA vir die karakterisering en identifikasie van data verkry vanaf prosesstelsels. Ten spyte van aanduidings in die literatuur dat SSA wydverspreid toegepas word in ander navorsingsvelde, kon geen vorige toepassings gevind word van SSA op chemiese prosesse nie. Dit wil voorkom asof die chemiese nywerhede groot baat kan vind by SSA van prosesdata. Die eerste aanduiding van hierdie voordele was in die vermoë van SSA om geraas te verwyder uit tydreekse. ‘n Aantal tipiese gevalle is ondersoek deur van verskeie benaderings tot SSA gebruik te maak. Nadat die geraas uit die tydreekse van die toetsgevalle verwyder is, is neurale netwerke gebruik om die prosesse te modelleer. Daar is herhaaldelik gevind dat die modelle wat gebou is op data wat eers deur SSA skoongemaak is, beter presteer as die wat slegs op die onverwerkte data gepas is. Die effektiwiteit van lokale SSA en auto-assosiatiewe neurale netwerke om nie- lineêre SSA toe te pas is ook vergelyk. Albei tegnieke het daarin geslaag om nie- lineêre hoofkomponente van die data te onttrek wat nie geïdentifiseer kon word deur die lineêre benadering nie. Daar is egter gevind dat lokale SSA ‘n meer betroubare tegniek is, aangesien die autoassosiatiewe neurale netwerke nie op sommige van die datastelle wou leer nie en vir ander tydreekse sinnelose hoofkomponente onttrek het. Laastens is ‘n aantal tydreekse geanaliseer met behulp van Monte Carlo SSA. Soos met alle ander karakteriseringstegnieke, kon Monte Carlo SSA nie daarin slaag om al die tydreekse wat ondersoek is korrek te identifiseer nie. Om hierdie rede is ‘n kombinasie van toetse gebruik om die onbekende tydreekse te klassifiseer. In die lig van al hierdie bevindinge, is die gevolgtrekking gemaak dat singuliere spektrale analise ‘n waardevolle hulpmiddel kan wees in die analise van chemiese en metallurgiese prosesdata.

Thesis (MScIng)--University of Stellenbosch, 2003. / ENGLISH ABSTRACT: It was shown in an earlier study that it is possible to predict the spectral radiance of rocket combustion plumes directly from the propellant composition and motor parameters. Little is published in the open literature on this subject, but the current trend is to use determinative methods like computational fluid dynamics and statistical techniques to simulate wide band radiance based on blackbody temperature assumptions. A limitation of these methods is the fact that they are computationally expensive and rather complex to implement. An alternative modeling approach was used which did not rely on solving all the nonlinearities and complex relationships applicable to a fundamental model. A multilayer perceptron based Neural Network was used to develop a parametric functional mapping between the propellant chemical composition and the motor design and the resulting spectral irradiance measured in a section of the plume. This functional mapping effectively models the relationship between the rocket design and the plume spectral radiance. Two datasets were available for use in this study: Emission spectra from solid propellant rockets and flare emission spectra. In the case of the solid rocket propellants, the input to the network consisted of the chemical composition of the fuels and four motor parameters, with the output of the network consisting of 146 scaled emission spectra points in the waveband from 2-5 microns. The four motor parameters were derived from equations describing the mass flow characteristics of rocket motors. The mass flow through the rocket motor does have an effect on the shape of the plume of combustion gases, which in turn has an effect on the infrared signature of the plume. The characteristics of the mass flow through the nozzle of the rocket motor determine the thermodynamic properties of the combustion process. This then influences the kind of chemical species found in the plume and also at what temperature these species are radiating energy.The resultant function describing the plume signature is: Plume signature f {p T A fuel composition} t , , , , 1 1 = ε It was demonstrated that this approach yielded very useful results. Using only 18 basic variables, the spectra were predicted properly for variations in all these parameters. The model also predicted spectra that agree with the underlying physical situation when changing the composition as a whole. By decreasing the Potassium content for example, the model demonstrated the effect of a flame suppressant on the radiance in this wavelength band by increasing the predicted output. Lowering the temperature, which drives the process of molecular vibration and translation, resulted in the expected lower output across the spectral band. In general, it was shown that only a small section of the large space of 2 propellant classes had to be measured in order to successfully generate a model that could predict emission spectra for other designs in those classes. The same principal was then applied to predicting the infrared spectral emission of a burning flare. The brick type flare considered in this study will ignite and the solid fuel will burn on all surfaces. Since there are no physical parameters influencing the plume as in the case of the rocket nozzles it was required to search for parameters that could influence the flare plume. It was possible to calculate thermodynamic properties for the flare combustion process. These parameters were then reduced to 4 parameters, namely: the oxidant-fuel ratio, equilibrium temperature, the molar mass and the maximum combustion temperature. The input variables for the flares thus consisted of the chemical composition and 4 thermodynamic parameters described above. The network proposed previously was improved and optimised for a minimum number of variables in the system. The optimised network marginally improved on the pevious results (with the same data), but the training time involved was cut substantially. The same approach to the optimization of the network was again followed to determine the optimal network structure for predicting the flare emission spectra. The optimisation involved starting out with the simplest possible network construction and continuouslyincreasing the variables in the system until the solution predicted by the network was satisfactory. Once the structure of the network was determined it was possible to optimise the training algorithms to further improve the solution. In the case of the solid rocket propellant emission data it was felt that it would be important to be able to predict the chemical composition of the fuel and the motor parameters using the infrared emission spectra as input. This was done by simply reversing the optimised network and exchanging the inputs with the outputs. The results obtained from the reversed network accurately predicted the chemical composition and motor parameters on two different test sets. The predicted spectra of some of the solid propellant rocket test sets and flare test sets did not compare well with the expected values. This was due to the fact that these test sets were in a sparsely populated area of the variable space. These outliers are normally removed from training data, but in this case there wasn’t enough data to remove outliers. To obtain an indication of the strength of the correlation between the predicted and measured line spectra two parameters were used to test the correlation between two line spectra. The first parameter is the Pearson product moment of coefficient of correlation and gives an indication of how good the predicted line spectra followed the trend of the measured spectral lines. The second parameter measures the relative distance between a target and predicted spectral point. For both the solid propellants and the flares the correlation values was very close to 1, indicating a very good solution. Values for the two correlation parameters of a test set of the flares were 0.998 and 0.992. In order to verify the model it was necessary to prove that the solution yielded by the model is better than the average of the variable space. Three statistical tests were done consisting of the mean-squared-error test, T-test and Wilcoxon ranksum test. In all three cases the average of the variable space (static model) and the predicted values (Neural Network model) were compared to the measured values. For both the T-test and the Wilcoxon ranksum test the null hypothesis is rejected when t < -tα = 1.645 and then thealternative hypothesis is accepted, which states that the error of the NN model will be smaller than that of the static model. The mean squared error for the static model was 0.102 compared to the 0.0167 of the neural net, for a solid propellant rocket test set. A ttest was done on the same test set, yielding a value of –2.71, which is smaller than – 1.645, indicating that the NN model outperforms the static model. The Z value for this test set is Z = -11.9886, which is a much smaller than –1.645. The results from these statistical tests confirm that neural network is a valid conceptual model and the solutions yielded are unique. / AFRIKAANSE OPSOMMING: In ‘n vroeër studie is bewys hoe dit moontlik is om die spektrale irradiansie van ‘n vuurpyl se verbrandingspluim te voorspel vanaf slegs die dryfmiddelsamestelling en vuurpylmotoreienskappe. In die literatuur is daar min gepubliseer oor hierdie onderwerp. Dit wil voorkom asof meer deterministiese metodes gebruik word om die probleem op te los. Metodes soos CFD simulasies en statistiese analises word tans verkies om wyeband radiansie te voorspel gebaseer op perfekte swart ligaam teorie. ‘n Groot beperking van hierdie metodes is die feit dat die berekeninge kompleks is en baie lank neem om te voltooi. ‘n Alternatiewe benadering is gebruik, wat nie poog om al die nie-liniêre en komplekse verbande uit eerste beginsels op te los nie. ‘n Neurale netwerk is gebruik om ‘n funksionele verband te skep tussen die chemiese samestelling van die dryfmiddel, vuurpylmotor ontwerp en die spektrale irradiansie van die vuurpyl se pluim. Die funksionele verband kan nou effektief die afhanklikheid van die dryfmiddelsamestelling, vuurpylmotor ontwerp en die spektrale uitset modelleer. Twee datastelle was beskikbaar vir analise: Emissie spektra van vaste dryfmiddel vuurpyle en ook van vaste dryfmiddel fakkels. Die invoer tot die neurale netwerk van die vuurpyle het bestaan uit die chemiese samestelling van die dryfmiddel en 4 vuurpylmotor eienskappe. Die uitvoer van die netwerk het weer bestaan uit 146 spektrale irradiansie waardes in die golflengte band van 2-5μm. Die 4 vuurpylmotor eienskappe is afgelei uit massavloei teorie vir vuurpyl motors, aangesien die uitvloei van die produkgasse ‘n invloed op die pluim van die motor sal hê. Die massavloei het weer ‘n effek op die spektrale handtekening van die pluim. Die eienskappe van die massavloei deur die mondstuk van die vuurpylmotor bepaal die termodinamiese eienskappe van die verbrandingsproses. Die invloed op die verbrandingsproses bepaal weer watter tipe produkte gevorm word en by watter temperatuur hulle energie uitstraal. Die gevolg is dat ‘n funksie gedefinieer kan word wat die pluim beskryf.Pluim handtekening = f{, temperatuur, mondstuk keël grootte, vernouings verhouding van mondstuk, dryfmiddelsamestelling} Deur net 18 invoer nodes te gebruik kon die netwerk die irradiansie suksesvol voorspel met ‘n variansie in al die invoer waardes. Deur byvoorbeeld die Kalium inhoud van die dryfmiddel samestelling te verminder het die model die vermindering van ‘n vlam onderdrukker suksesvol nageboots deurdat die irradiansie ‘n hoër uitset gehad het. Die sensitiwiteit van die model is verder getoets deur die temperatuur in die verbrandingskamer te verlaag, met ‘n korrekte laer irradiansie uitset, as gevolg van die feit dat die temperatuur die molekulêre vibrasie en translasie beweging beheer. Dieselfde benadering is gebruik om die model te bou vir die voorspelling van die fakkels se infrarooi irradiansie. Anders as die vuurpylmotors vind die verbranding in die geval van die fakkels in die atmosfeer plaas. Dit was dus ook nodig om na die termodinamiese eienskappe van die fakkel verbranding te kyk. Verskeie parameters is bereken, maar 4 parameters, naamlik die brandstof-suurstof verhouding, temperatuur, molêre massa en die maksimum verbrandingstemperatuur, tesame met die dryfmiddel samestelling kon die irradiansie van die fakkels suskesvol voorspel. Die bestaande netwerk struktuur vir die vuurpylmotors is verbeter en geoptimiseer vir ‘n minimum hoeveelheid veranderlikes in die stelsel. Die geoptimiseerde netwerk het ‘n klein verbetering in die voorspellings getoon, maar die oplei het drasties afgeneem. Dieselfde benadering is gebruik om die optimale netwerk vir die fakkels te bepaal. Optimisering van die netwerk struktuur is bereik deur met die eenvoudigste struktuur te begin en die hoeveelheid veranderlikes te vermeerder totdat ‘n bevredigende oplossing gevind is. Na die struktuur van die netwerk bevestig is, kon die oordragfunksies op die nodes verder geoptimiseer word om die model verder te verbeter. Dit het verder geblyk dat dit moonlik is om die netwerk vir die vuurpylmotors om te draai sodat die irradiansie gebruik word om die dryfmiddel samestelling en motor eienskappe te voorspel. Die netwerk is eenvoudig omgedraai en die insette het die uitsette geword.Die resultate van die omgekeerde netwerk het bevestig dat dit wel moontlik is om die dryfmiddel samestelling en motor eienskappe te voorspel vanaf die irradiansie. Die voorspelde spektra van beide die vuurpylmotors en die fakkels het nie altyd goed gekorreleer met die gemete data nie. Van die spektra kom voor in ‘n lae digtheidsdeel van die veranderlike ruimte. Dit het tot gevolg gehad dat daar nie genoeg data vir opleiding van die netwerk in die omgewing van die toetsdata was nie. Hierdie data is eintlik uitlopers en moet verwyder word van die opleidingsdata, maar daar is alreeds nie genoeg data beskikbaar om die uitlopers te verwyder nie. Dit is nodig om te bepaal hoe goed die voorspelde data vergelyk met die gemete data. Twee parameters is gebruik om te bepaal hoe goed die data korreleer. Die eerste is die “Pearson product moment of coefficient of correlation”, wat ‘n goeie aanduiding gee van hoe goed die voorspelde waardes die gemete waardes se profiel volg. Die tweede parameter meet die relatiewe afstand tussen die teiken en die voorspelde waardes. Vir beide die vuurpylmotors en die fakkels het die toetsstelle ‘n korrelasiewaarde van baie na aan 1 gegee, wat ‘n goeie korrelasie is. Die waardes van die twee parameters vir een van die fakkel toetstelle was onderskeidelik 0.998 en 0.992. Die model is geverifieer deur te bepaal of die model ‘n beter oplossing bied as die gemiddeld van die veranderlike ruimte. Drie statistiese toetse is gedoen: “Mean-squarederror” toets, T-toets en ‘n “Wilcoxon ranksum” toets. In al drie gevalle word die gemiddelde van die veranderlike ruimte (statiese model) en die voorspelde waardes (Neurale netwerk model) teen die gemete waardes getoets. Vir beide die T-toets en die “Wilcoxon ranksum” toets word die nul hipotese verwerp indien t < ta = 1.645 en dan word die alternatiewe hipotese aanvaar, wat bepaal dat die fout van die neurale netwerk model kleiner is as die van die statiese model. Die “mean-squared-error” van die statiese model was 0.102, in vergelyking met 0.0167 van die neurale netwerk model vir ‘n vuurpylmotor toetsstel. ‘n T-toets is gedoen vir dieselfde toetsstel, met ‘n resultaat van-2.71, wat kleiner is as –1.645 en aandui dat die neurale netwerk model weereens beter presteer as die statiese model. Die Z waarde uit die “Wilcoxon ranksum” toets is Z=- 11.9886, wat baie kleiner is as –1.645. Die resultate van die statitiese toetse toon dat die neurale netwerk ‘n geldige model is en die oplossings van die model ook uniek is.

Thesis (MScIng)--University of Stellenbosch, 2004. / ENGLISH ABSTRACT: Recent years have seen the development of various treatment methods for the purification of industrial waste waters due to the increased demand for reduced pollutant effluents. Aqueous waste streams containing toxic organic compounds are of special interest, since conventional treatment methods such as biological waste treatment can not always be used. Other popular treatment methods are often ineffective. Catalytic oxidation of organic wastes has been investigated since the 1960s with varying degrees of success. A major problem associated with this method is the high temperatures and pressures required to improve the activation energies involved. Electrochemical oxidation has become a popular method in the literature of treating these wastes, since the applied voltage determines the activation energy, and therefore the process can often be performed at ambient conditions. This thesis investigates the capability of a unique reactor system in the treatment of these wastes. The reactor utilises proton-exchange membrane technology to eliminate the requirement of conductivity in treated waste streams; thus the membrane serves as a solid electrolyte. The reactor system has therefore been referred to as a solid-polymer-electrolyte reactor. Novel metal oxide anodes are responsible for the oxidation of the organic molecules. These metal oxide catalysts show promise in the treatment of a wide variety of organic wastes. A SnO2 catalyst doped with ZrO2 is used as anode in this study. Dopants are added to the catalyst to improve properties such as catalytic activity and conductivity. Kinetic data was obtained on a wide range of values for the chosen experimental parameters (current density and flow rate). Phenol, an organic molecule often referred to in the literature as model contaminant due to its resistance to oxidation,was also used as contaminant in this study. The use of the reactor system in the disinfection of water containing selected pathogens, were included in the experimental work. This kinetic data served in the development of a simple model of the process, and provided the basis for a full analysis regarding potential scale-up and economic feasibility. A requirement of the study was the accurate determination of the various oxidation breakdown products of phenol. This led to the refinement of an HPLC analytical method in order to quantitatively determine these products. The full analysis showed that the current reactor system would not be economically viable — mainly due to very long reactor lengths required for the complete removal of all organic material. Both mass transfer and charge transfer at the chosen experimental conditions influenced the electrochemical oxidation of phenol. High pressure drops, causing low flow rates in the reactor, accounted for this because of the narrow flow channels required in the reactor. Some catalyst deactivation was also suspected to affect the overall reaction, but the full extent of the deactivation was not investigated thoroughly. There is still room for improvement in the electrochemical oxidation of organic wastes. The design of the flow channels, a factor that was not investigated, can significantly improve efficiency. Another aspect that was not investigated was the catalyst type. The catalyst has been identified in the literature as the main contributing factor to the success of the oxidation reaction. A wide variety of metal oxide catalysts are currently being researched and may improve the kinetics of the process even further. Further improvement needs to be made on the membrane/electrode assembly to improve current density distribution. Every improvement of the process in terms of the reactor design and catalyst will impact on the economics of the process, thus making the process more competitive with current treatment technologies. / AFRIKAANSE OPSOMMING: In die afgelope paar dekades, is daar ’n wye verskeidenheid metodes ontwikkel wat gebruik kan word om industri¨ele afvoer strome te behandel. Hierdie ontwikkeling het plaasgevind as gevolg van die verhoogde eis aan skoner afvoerstrome. Wateragtige afvoerstrome wat organiese verbindings bevat, is van besonderse belang omdat hierdie tipe strome soms besonders moeilik kan wees om te behandel. Gebruiklike metodes is in die meeste gevalle ongeskik vir behandelings-doeleindes. Katalitiese oksidasie is sedert die 1960’s gebruik, maar hierdie prosesse benodig dikwels ho¨e drukke en temperature om suksesvol te wees. Elektrochemiese oksidasie het intussen ’n populˆere behandelingsmetode geword, aangesien die aktiveringsenergie vir die oksidasieproses hoofsaaklik afhanklik is van die aangewende potensiaal en dus kan die proses by atmosferiese toestande gebruik word. In hierdie tesis word die geskiktheid van ’n unieke reaktorstelsel vir water-suiwering ondersoek. Die reaktor gebruik ’n proton-uitruilings-membraan om die behoefte vir konduktiwiteit in die water uit te skakel. Die membraan dien dus as ’n tipe soliede elektroliet en as gevolg hiervan word na die reaktorstelsel verwys as ’n soliede-polimeer-elektroliet reaktor. Nuwe metaal-oksied anodes word in die reaktor gebruik aangesien hulle belowende resultate toon in die oksidasie van organiese verbindings. In die navorsing, is ’n SnO2 katalis wat klein hoeveelhede ZrO2 bevat gebruik. Oksiede soos ZrO2 word dikwels gebruik om die aktiwiteit en konduktiwiteit van hierdie kataliste te bevorder. Kinetiese data is oor ’n wye bereik van parameter waardes ingesamel. Die hoof parameters in die eksperimentele werk was stroom digtheid en vloeitempo. Fenol, ‘n komponent wat volgens die literatuur in hierdie tipe van werk gebruik word, isas die besoedelende komponent gekies. Die doeltreffendheid van die reaktor in die ontsmetting van water, wat met ’n verskeidenheid skadelike mikro-organismes besmet is, is ook getoets. ‘n Eenvoudinge model is opgestel m.b.v. die kinetiese data, waarna ’n volledige analise met betrekking tot grootskaalse bedryf en ekonomiese uitvoerbaarheid gedoen is. ‘n Vereiste van die studie was om die konsentrasie van die afbreek-produkte van die oksidasie akkuraat vas te stel. As gevolg hiervan is ‘n ho¨e-druk-vloeistofchromatografie analitiese metode verfyn. Die analise het getoon dat die reaktorstelsel nie ekonomies sou wees nie. Een van die hoofredes hiervoor is die onrealistiese reaktorlengtes wat benodig sou word. Resultate het getoon dat die reaksie deur beide massa-oordrag en lading-oordrag be¨ınvloed word. Ho¨e drukvalle in die reaktor wat gelei het tot lae vloeitempo’s was hiervoor verantwoordelik. Die deaktivering van die katalis be¨ınvloed waarskynlik die reaksie, maar die deaktiveringsverskynsel is nie ten volle ondersoek nie. Die reaktorstelsel kan verder verbeter word deur verskeie elemente van die reaktor te ondersoek. Die ontwerp van die vloeikanale in die reaktor is nie ondersoek nie en kan die werksverrigting van die reaktor verhoog. Uit die literatuur is gevind dat die tipe metaaloksied wat as katalis gebruik word, die reaksie direk be¨ınvloed. Dus kan navorsing wat tans op die kataliste gedoen word nuwe kataliste na vore bring wat meer doeltreffend sal wees. Laastens, is die huidige membraan/elektrode samestelling nog oneffektief en kan die reaktor-opstelling dus nog verbeter word. Elke verbetering wat op die bogenoemde faktore van die reaktor ontwerp verkry word, sal die ekomoniese uitvoerbaarheid van die proses be¨ınvloed. So, sal die proses al meer kompeterend met huidige behandelingsmetodes word.

Thesis (MScIng)--University of Stellenbosch, 2003. / ENGLISH ABSTRACT: The emission spectrum measured in the middle infrared (IR) band from the plume of a rocket can be used to identify rockets and track inbound missiles. It is useful to test the stealth properties of the IR fingerprint of a rocket during its design phase without needing to spend excessive amounts of money on field trials. The modelled predictions of the IR spectra from selected rocket motor design parameters therefore bear significant benefits in reducing the development costs. In a recent doctorate study it was found that a fundamental approach including quantum-mechanical and computational fluid dynamics (CFD) models was not feasible. This is first of all due to the complexity of the systems and secondly due to the inadequate calculation speeds of even the most sophisticated modern computers. A solution was subsequently investigated by use of the ‘black-box’ model of a multi-layer perceptron feed-forward neural network with a single hidden layer consisting of 146 nodes. The input layer of the neural network consists of 18 rocket motor design parameters and the output layer consists of 146 IR absorbance variables in the range from 2 to 5 μm wavelengths. The results appeared promising for future investigations. The available data consist of only 18 different types of rocket motors due to the high costs of generating the data. The 18 rocket motor types fall into two different design classes, the double base (DB) and composite (C) propellant types. The sparseness of the data is a constraint in building adequate models of such a multivariate nature. The IR irradiance spectra data set consists of numerous repeat measurements made per rocket motor type. The repeat measurements form the pure error component of the data, which adds stability to training and provides lack-of-fit ANOVA capabilities. The emphasis in this dissertation is on comparing the feed-forward neural network model to the linear and neural network partial least squares (PLS) modelling techniques. The objective is to find a possibly more intuitive and more accurate model that effectively generalises the input-output relationships of the data. PLS models are known to be robust due to the exclusion of redundant information from projections made to primary latent variables, similarly to principal components (PCA) regression. The neural network PLS techniques include feed-forward sigmoidal neural network PLS (NNPLS) and radial-basis functions PLS (RBFPLS). The NNPLS and RBFPLS algorithms make use of neural networks to find non-linear functional relationships for the inner PLS models of the NIPALS algorithm. Error-based neural network PLS (EBNNPLS) and radial-basis function network PLS (EBRBFPLS) are also briefly investigated, as these techniques make use of non-linear projections to latent variables. A modification to the orthogonal least squares (OLS) training algorithm of radial-basis functions is developed and applied. The adaptive spread OLS algorithm (ASOLS) allows for the iterative adaptation of the Gaussian spread parameters found in the radial-basis transfer functions. Over-fitting from over-parameterisation is controlled by making use of leaveone- out cross-validation and the calculation of pseudo-degrees of freedom. After cross-validation the overall model is built by training on the entire data set. This is done by making use of the optimum parameterisation obtained from cross-validation. Cross-validation also gives an indication of how well a model can predict data unseen during training. The reverse problem of modelling the rocket propellant chemical compositions and the rocket physical design parameters from the IR irradiance spectra is also investigated. This problem bears familiarity to the field of spectral multivariate calibration. The applications in this field readily make use of PLS and neural network modelling. The reverse problem is investigated with the same modelling techniques applied to the forward modelling problem. The forward modelling results (IR spectrum predictions) show that the feedforward neural network complexity can be reduced to two hidden nodes in a single hidden layer. The NNPLS model with eleven latent dimensions outperforms all the other models with a maximum average R2-value of 0.75 across all output variables for unseen data from cross-validation. The explained variance for the output data of the overall model is 94.34%. The corresponding explained variance of the input data is 99.8%. The RBFPLS models built using the ASOLS training algorithm for the training of the radialbasis function inner models outperforms those using K-means and OLS training algorithms. The lack-of-fit ANOVA tests show that there is reason to doubt the adequacy of the NNPLS model. The modelling results however show promise for future development on larger, more representative data sets. The reverse modelling results show that the feed-forward neural network model, NNPLS and RBFPLS models produce similar results superior to the linear PLS model. The RBFPLS model with ASOLS inner model training and 5 latent dimensions stands out slightly as the best model. It is found that it is feasible to separately find the optimum model complexity (number of latent dimensions) for each output variable. The average R2-value across all output variables for unseen data is 0.43. The average R2-value for the overall model is 0.68. There are output variables with R2-values of over 0.8. The forward and reverse modelling results further show that dimensional reduction in the case of PLS does produce the best models. It is found that the input-output relationships are not highly non-linear. The non-linearities are largely responsible for the compensation of both the DB- and C-class rocket motor designs predictions within the overall model predictions. For this reason it is suggested that future models can be developed by making use of a simpler, more linear model for each rocket class after a class identification step. This approach however requires additional data that must be acquired. / AFRIKAANSE OPSOMMING: Die emissiespektra van die uitlaatpluime van vuurpyle in die middel-infrarooi (IR) band kan gebruik word om die vuurpyle te herken en om inkomende vuurpyle op te spoor. Dit is nuttig om die uitstralingseienskappe van ‘n vuurpyl se IR afdruk te toets, sonder om groot bedrae geld op veldtoetse te spandeer. Die gemodelleerde IR spektrale voorspellings vir ‘n bepaalde stel vuurpylmotor ontwerpsparameters kan dus grootliks bydra om motorontwikkelingskostes te bemoei. In ‘n onlangse doktorale studie is gevind dat ‘n fundamentele benadering van kwantum-meganiese en vloeidinamika-modelle nie lewensvatbaar is nie. Dit is hoofsaaklik as gevolg van die onvoldoende vermoë van selfs die mees gesofistikeerde moderne rekenaars. ‘n Moontlike oplossing tot die probleem is ondersoek deur gebruik te maak van ‘n multilaag perseptron voorwaartse neurale netwerk met 146 nodes in ‘n enkele versteekte laag. Die laag van invoer veranderlikes bestaan uit agtien vuurpylmotor ontwerpsparameters en die uitvoerlaag bestaan uit 146 IR-absorbansie veranderlikes in die reeks golflengtes vanaf 2 tot 5 μm. Dit het voorgekom dat die resultate belowend lyk vir toekomstige ondersoeke. Weens die hoë kostes om die data te genereer bestaan die beskikbare data uit slegs agtien verskillende tipes vuurpylmotors. Die agtien vuurpyl tipes val verder binne twee ontwerpsklasse, naamlik die dubbelbasis (DB) en saamgestelde (C) dryfmiddeltipes. Die yl data bemoeilik die bou van doeltreffende multiveranderlike modelle. Die datastel van IR uitstralingspektra bestaan uit herhaalde metings per vuurpyltipe. Die herhaalde metings vorm die suiwer fout komponent van die data. Dit verskaf stabilitieit tot die opleiding op die data en verder die vermoë om ‘n analise van variansie (ANOVA) op die data uit te voer. In hierdie tesis lê die klem op die vergelyking tussen die voorwaartse neurale netwerk en die lineêre en neurale netwerk parsiële kleinste kwadrate (PLS) modelleringstegnieke. Die doel is om ‘n moontlik meer insiggewende en akkurate model te vind wat effektief die in- en uitvoer verhoudings kan veralgemeen. Dit is bekend dat PLS modelle meer robuus kan wees weens die weglating van oortollige inligting deur projeksies op hoof latente veranderlikes. Dit is analoog aan hoofkomponente (PCA) regressie. Die neurale netwerk PLS-tegnieke sluit in voorwaartse sigmoïdale neurale netwerk PLS (NNPLS) en radiale-basis funksies PLS (RBFPLS). Die NNPLS en RBFPLS algoritmes maak gebruik van die neurale netwerke om nie-lineêre funksionele verbande te kry vir die binne PLS-modelle van die nie-lineêre iteratiewe parsiële kleinste kwadrate (NIPALS) algoritme. Die fout-gebaseerde neurale netwerk PLS (EBNNPLS) en radiale-basis funksies PLS (EBRBFPLS) is ook weens hulle nie-lineêre projeksies na latente veranderlikes kortiliks ondersoek. ‘n Aanpassing tot die ortogonale kleinste kwadrate (OLS) opleidingsalgoritme vir radiale-basis funksies is ontwikkel en toegepas. Die aangepaste algoritme (ASOLS) behels die iteratiewe aanpassing van die verspreidingsparameters binne die Gauss-funksies van die radiale-basis transformasie funksies. Die oormatige parameterisering van ‘n model word beheer deur kruisvalidering met enkele weglatings en die berekening van pseudo-vryheidsgrade. Na kruisvalidering word die algehele model gebou deur opleiding op die volledige datastel. Dit word gedoen deur van die optimale parameterisering gebruik te maak wat deur kruisvalidering bepaal is. Kruisvalidering gee ook ‘n goeie aanduiding van hoe goed ‘n model ongesiende data kan voorspel. Die modellering van die vuurpyle se chemiese en fisiese ontwerpsparameters (omgekeerde probleem) is ook ondersoek. Hierdie probleem is verwant aan die veld van spektrale multiveranderlike kalibrasie. Die toepassings in die veld maak gebruik van PLS en neurale netwerk modelle. Die omgekeerde probleem word dus ondersoek met dieselfde modelleringstegnieke wat gebruik is vir die voorwaartse probleem. Die voorwaartse modelleringsresultate (IR voorspellings) toon dat die kompleksiteit van die voorwaartse neurale netwerk tot twee versteekte nodes in ‘n enkele versteekte laag gereduseer kan word. Die NNPLS model met elf latente dimensies vaar die beste van alle modelle, met ‘n maksimum R2-waarde van 0.75 oor alle uitvoer veranderlikes vir die ongesiende data (kruisvalidering). Die verklaarde variansie vir die uitvoer data vanaf die algehele model is 94.34%. Die verklaarde variansie van die ooreenstemmende invoer data is 99.8%. Die RBFPLS modelle wat gebou is deur van die ASOLS algoritme gebruik te maak om die PLS binne modelle op te lei, vaar beter in vergelyking met die K-gemiddeldes en OLS opleidingsalgoritmes. Die toetse wat ‘n ‘tekort-aan-passing’ ANOVA behels, toon dat daar rede is om die geskiktheid van die NNPLS model te wantrou. Die modelleringsresultate lyk egter belowend vir die toekomstige ontwikkeling van modelle op groter, meer verteenwoordigde datastelle. Die omgekeerde modellering toon dat die voorwaartse neurale netwerk, NNPLS en RBFPLS modelle soortgelyke resultate produseer wat die lineêre PLS model s’n oortref. Die RBFPLS model met ASOLS opleiding van die PLS binne modelle word beskou as die beste model. Dit is lewensvatbaar om die optimale modelkompleksiteite van elke uitvoerveranderlike individueel te bepaal. Die gemiddelde R2-waarde oor alle uitvoerveranderlikes vir ongesiende data is 0.43. Die gemiddelde R2-waarde vir die algehele model is 0.68. Daar is van die uitvoer veranderlikes wat R2-waardes van 0.8 oortref. Die voor- en terugwaartse modelleringsresultate toon verder dat dimensionele reduksie in die geval van PLS die beste modelle lewer. Daar is ook gevind dat die nie-lineêriteite grootliks vergoed vir die voorspellings van beide DB- en Ctipe vuurpylmotors binne die algehele model. Om die rede word voorgestel dat toekomstige modelle ontwikkel kan word deur gebruik te maak van eenvoudiger, meer lineêre modelle vir elke vuurpylklas nadat ‘n klasidentifikasiestap uitgevoer is. Die benadering benodig egter addisionele praktiese data wat verkry moet word.

Thesis (MScEng) -- University of Stellenbosch, 2004. / ENGLISH ABSTRACT: The grape juice concentrate (GJC) plant of the KWV at Robertson spent significant amounts of money on the decolourisation of grape juice concentrate. A chemically activated powdered activated carbon (PAC) purchased from Norit, namely CA1, was used as decolourisation product. Apart from the expenses involved, it contributed largely to the solid waste produced at this plant. A way was sought to minimise or prevent GJC discolourisation (and possibly solid waste) without increasing operating expenses. Browning reactions in GJC are as old as the product itself. Numerous researchers have studied the origins of these reactions, the reactants and products involved, as well as the reaction kinetics of these reactions. From the work of these researchers four possible browning reaction pathways were identified, namely: • enzymatic oxidative browning, • non-enzymatic oxidative browning, • non-enzymatic browning (the Maillard reaction), and • caramelisation. It was also identified that 5-hydroxymethylfurfural (HMF) are indicative of the browning potential of GJC. A method to analyse for HMF (quantitative and qualitative) was develop for the purposes of this study, namely positive electron-spray ionisation preceded by high-pressure liquid chromatography (HPLC) and followed by dual mass spectrometry. This method showed good repeatability and was used to analyse all samples generated during this study. It was confirmed that the manufacturing process at this plant favours nonenzymatic browning reactions, since mild heat treatment deactivates enzymes. Further investigation indicated that the overruling browning reaction on this plant was non-enzymatic oxidative browning. It was shown that neither the presence, nor the absence of protein had any effect on the rate of formation of HMF. It was, however, confirmed that HMF formation could be attributed to high temperatures and prolonged exposure to these temperatures. Other adsorption products were evaluated against the then current PAC (CA1), namely a steam activated PAC supplied by Norit, SA4, and a polymeric adsorbent, Polyclar V (polyvinylpolypyrrolidone/PVPP). Both SA4 and PVPP indicated superior HMF adsorption capacities. Replacing CA1 with SA4 could result in operating expenses savings and possible solid waste reduction. However, PVPP were too expensive to be considered an economically viable replacement for CA1. Improved concentration technologies such as reverse osmosis (RO) membrane concentration followed by centrifugal evaporation (CE) or twostage CE should be considered as possible replacement for the existing concentration technology (multi-stage falling film evaporator). This should decrease heat treatment/exposure by more than 90% and thus reduce browning significantly. An added advantage could be the reduction of solid waste, since less (if not no) decolourisation will be required. Alternatively, juice should be stored with added sulphur dioxide (SO2), since it was shown that this juice contained much lower HMF concentrations than diluted concentrate (stored for the same time). This should reduce heat exposure by up to 50% and thus minimise browning reactions. / AFRIKAANSE OPSOMMING: Die druiwesapkonsentraat (DSK) aanleg van die KWV in Robertson het jaarliks aansienlike bedrae geld spandeer tydens die ontkleuringsproses van DSK. ‘n Chemies geaktiveerde verpoeierde koolstof (GVK) verkrygbaar van Norit, naamlik CA1, is gebruik as ontkleuringsproduk. Buiten die kostes verbonde aan hierdie produk het dit ook grootliks bygedra tot soliede afval by hierdie aanleg. Oplossings is gesoek om die verbruining/ontkleuring van DSK (en dalk ook soliede afval) te verminder (of selfs te voorkom) sonder om bedryfskostes te verhoog. Verbruiningsreaksies in DSK bestaan al so lank soos DSK self. Verskeie navorsers het die oorsake, reaktante, produkte en reaksiekinetika van hierdie reaksies oor die jare heen bestudeer. Uit die werk van sommige van hierdie navorsers kon vier moontlike verbruiningsreaksieroetes geïdentifiseer word, naamlik: • ensiematiese oksidatiewe verbruining, • nie-ensiematiese oksidatiewe verbruining, • nie-ensiematiese verbruining (die Maillard-reaksie), en • karamelisering. Daar was verder geïdentifiseer dat 5-hidroksiemetielfurfuraal (HMF) aanduidend is van die verbruiningspotensiaal van DSK. ‘n Analitiese metode (kwalitatief en kwantitatief) om vir HMF te analiseer is vir die doel van hierdie studie ontwikkel, naamlik positiewe elektronsproei ionisasie, voorafgegaan deur hoëdruk vloeistof chromatografie en gevolg deur dubbele massa spektrometrie. Hierdie analitiese metode het goeie herhaalbaarheid getoon en was deurgaans gebruik om monsters te analiseer gedurende hierdie studie. Dit was bevestig dat die vervaardigingsproses by hierdie aanleg nieensiematiese verbruiningsreaksies begunstig, aangesien geredelike hittebehandeling ensieme deaktiveer. Verdere navorsing het getoon dat die oorheersende verbruiningsreaksies by hierdie aanleg nie-ensiematiese oksidatief van aard is. Resultate het getoon dat proteinstabiliteit geen invloed op die vormingstempo van HMF het nie. Dit was egter bevestig dat vorming van HMF direk verband hou met hoë temperature en lang blootstellingsperiodes aan hierdie temperature. Ander adsorpsieprodukte was vergelyk met die huidige GVK (CA1), naamlik ‘n stoom geaktiveerde verpoeierde koolstof (Norit se SA4) en ‘n polimeriese adsorbant, Polyclar V (polivinielpolipirrolidoon/PVPP). Beide SA4 en PVPP het CA1 oortref wat betref HMF adsorpsie. Moontlike bedryfskostebesparings (en soliede afval verminderings) potensiaal bestaan indien CA1 vervang word met SA4. Die teenoorgestelde is egter waar vir PVPP wat bedryfskoste aangaan. Instede van die huidige verdampinstegnologie, naamlik vallendefilmverdamping, hoort verbeterde konsentrasietegnologieë soos tru-osmose membraankonsentrasie gevolg deur sentrifugale verdamping, of, alternatiewelik, twee-stadium sentrifugale verdamping, orrweeg te word. Op hierdie wyse behoort hittebehandeling (en dus verbruining) met sowat 90% verminder te word. ‘n Moonlike addisionele voordeel is die vermindering van soliede afval aangesien minder ontkleuring nodig sal wees. Indien die verbeterde tegnologieë te duur is moet daar gekyk word daarna om die ongekonsentreerde sap met addisionele swaweldioksied (SO2) te stoor, aangesien veel laer HMF konsentrasies in sulke sap waargeneem is as in verdunde direkte konsentraat wat vir dieselfde typerk gestoor is. Hittebehandeling sal op hierdie wyse met tot 50% verminder word (en dus verbruining ook).