Thesis (MScEng (Process Engineering))--University of Stellenbosch, 2005. / Although selective oxidation catalysts are widely used and extensively studied for their industrial and academic value, their complex mechanisms are, to a large extent, still unclear. The field of so-called allylic (amm)oxidations reactions was chosen for further investigation, in particular the simplistic selective oxidation of propene to acrolein over an α-Bi2Mo3O12 catalyst. One of the most important approaches in selective oxidation is to try to correlate the physicochemical properties of catalysts with their catalytic performance (activity and selectivity). The most interesting, and seemingly most widely invoked parameter, is lattice oxygen mobility. The problem, however, is the difficulty encountered in measuring oxygen mobility. It is hypothesised that the depth of oxygen utilisation and lattice oxygen mobility of bismuth molybdate during the partial oxidation of propene to acrolein may be determined by measuring the rate of acrolein formation and lattice oxygen usage over a range of discrete particle sizes that could be synthesised using reverse micelle technology. Catalyst Preparation A preliminary investigation into the reverse micelle technique showed that discrete nanosized particles could be synthesised, but that there was no size control over the outcome and that, in most cases there were some degree of particle agglomeration. It was also found that nanorod formation occurred due to adsorbtion of surfactant. More in-depth investigation had to be done in order to achieve particle size control and the liberation of the calcined α-Bi2Mo3O12 catalyst particles required for kinetic experiments. Simple precipitation methods, the catalyst calcination step, and the formation and stability of reverse micelles were investigated. A simple precipitation method to prepare α-Bi2Mo3O12, suitable to be integrated into the reverse micelle technique was found by buffering the mixture of bismuth nitrate and ammonium molybdate solutions with an excess of molybdate. This prevented the pH from decreasing below a critical value of 1.3 (at which β-Bi2Mo2O9 forms as an impurity). The excess molybdenum caused the formation of MoO3 in the calcined product, which was selectively and successfully removed using a warm ammonium wash followed by a water rinse and a recalcination step. XRD of a temperature range calcination shows that the calcination starts at temperatures as low as 200°C and almost complete calcination of the catalyst at 280°C. DSC analyses show a 47.15 J/g crystal formation peak only at 351°C. The Mo18O56(H2O)8 4- anion or its double, Mo36O112(H2O)16 8-, is responsible for the formation of α-Bi2Mo3O12 in the precipitation calcination reaction. Reverse micelles were investigated using a Malvern Zetasizer and showed a complex dynamic system in which the reverse micelle sizes and size distributions change over time as a function of surfactant and aqueous concentrations, the salt used and aqueous phase salinity. Although much was accomplished in this study, more investigations into the constituent steps of the reverse micelle technique are needed to develop a method to synthesise the range of discrete catalyst particle sizes required for kinetic studies. Kinetic Studies For the purpose of kinetic experiments a metal reactor was found to be superior to that of a glass reactor. The reactor rig was adequate for these kinetic studies but do not meet the requirements for detailed reaction order experiments. The analysing apparatus could not measure CO2 formation accurately and it had to be calculated using a carbon balance. Only the model proposed by Keulks and Krenzke [1980a] was able to describe the kinetic result, but the model parameter describing the oxidative state of the catalyst surface could not be calculated due to the lack compatibility between published data. Values were awarded to this parameter so to give an Arrhenius plot which corresponded to published data. The parameter describing the oxidative state vs. temperature took on a function that was consistant with the reasoning of Keulks and Krenzke [1980a]. Comprehensive preliminary kinetic studies are needed, both in catalyst reduction and reoxidation, in order to determine the reaction conditions, explore more advanced kinetic models and investigate model parameters that are theoretically and/or empirically obtainable and quantifiable.

Thesis (MEng.)--University of Stellenbosch, 2000. / ENGLISH ABSTRACT: The aim of the thesis was to investigate the use of flow destabilization methods, combined with permeate backflushing (BIF) or on their own, on flux recovery and maintenance in capillary UF membrane systems under cross-flow (XF) and dead-end (DE) operating conditions. Various hydraulic and mechanical methods have been used to remove the accumulated cake layer and improve steady state process flux. Permeate backflushing (B/F) is the most widely used but the drawbacks are loss of product and extensive down-time. In a pilot plant study for ultrafiltration of surface waters containing high NOM, turbidity and cation loads, the use of flow destabilization, or feed flow reversal (FFR) combined with cross-flow B/F was able to improve the normalised flux by 10.7 ± 3.4 %, compared with 3.2 ± 1.6 % improvement for BIF without FFR. When a second B/F included FFR, the flux improvement was 7.0 ± 2.0 % compared with 4.3 ± 2.5 % for a B/F without FFR. The hypothesis was proposed that the flow destabilization caused slight lifting of the oriented cake layer, while the cross-flow B/F was able to sweep the lifted cake out of the lumen. If the flow destabilization may be effected by a simple but effective and low-cost method, and if this flow destabilization may be combined with reverse flow for short durations, the "lift-and-sweep" approach will be the ideal method of maintaining process flux and increasing membrane life. Such a flow destabilization method, now named "reversepressure pulsing" (RIP), was developed. The method involves circulation of feed water in a recycle loop for 2 s to gain momentum, followed by closure of a fast-action valve upstream of the modules. The momentum of the water in the concentrate loop carries it into an air-filled feed accumulator, while concentrate and reverse-flow permeate (which also lifts the fouling layer) are discharged to the atmosphere using the recycle pump for 15 s. When the valve opens again, the air in the accumulator forces the water under pressure through the membrane lumens, causing a pressure pulse and flow perturbations that lift, shift and break up the fouling layer. During 3 such "lift-and-sweep" events, the cake is lifted and the debris is swept out of the lumen. Experimental results for uninterrupted dead-end filtration at a UF pilot plant using RIP only on a severely fouled membrane, indicated that the RIP increased the flux by 18.4 % and decreased the dP by 8.2 % over a 7.2 h period. The method is effective in removing the cake layer intermittently and no long-term flux decline occurred for a period of 555 h since the previous chemical cleaning. / AFRIKAANSE OPSOMMING: Die doel van die tesis was om die gebruik van vloei-destabiliserings metodes, alleen of gekombineer met permeaat-terugwas, op vloed-herwinning en instandhouding in kapillêre UF membraan-stelsels tydens kruisvloei en doodloop bedryf, te ondersoek. Verskeie meganiese en hidrouliese metodes word gebruik in membraan stelsels om die koeklaag op die membraan se oppervlak te verwyder en die gestadigde-toestand vloed te verbeter. Vanhierdie metodes word permeaat-terugwas die meeste gebruik, maar het sy nadele insluitend verlies van produk en produksietyd. In 'n loodsstudie vir die ultrafiltrasie van oppervlakwaters wat hoë beladings NOM, turbiditeit en katione bevat, is die waarneming gemaak dat kruisvloei terugwas met vloeidestabilisering (voerrigting-verandering) die genormaliseerde vloed met 10.7 ± 3.4 % kon verbeter, vergeleke met 'n 3.2 ± l.6 % verbetering sonder voerrigting-verandering. Vir 'n tweede terugwas was die verbetering 7.0 ± 2.0 % vergeleke met 4.3 ± 2.5 % sonder voerrigtingverandering. Die hipotese was voorgestel dat die vloei-destabilisering die geoiënteerde koeklaag van die oppervlak gelig het, en die kruisvloei terugwas die geligde koeklaag uit die lumen kon vee. Indien hierdie vloei-destabilisering bewerk kan word deur 'n eenvoudige maar effektiewe manier, en indien dit gekombineer kan word met terugvloei van produk vir kort tydperke, sal hierdie "lig-en-vee" benadering die ideale metode wees om die membrane se vloed te verbeter en leeftyd te verleng. So 'n vloei-destabiliseringsmetode, nou genoem "terugdruk-pulsering", is ontwikkel. Die metode behels die sirkuiering van voer-water vir 2 s in 'n hersirkulasielus om momentum op te bou, gevolg deur die toemaak van 'n snel-aksie klep stroom-op van die modules. Die water in die konsentraat-lus se momentum dra dit vorentoe tot in In lug-gevulde voer-akkumulator, terwyl konsentraat en terug-vloei permeaat (wat ook tot 'n mate die koeklaag lig) ook na die atmosfeer gewend word vir 15 s deur die hersirkulasiepomp. As die klep weer oopgaan, ontspan die lug in die akkumulator, en forseer die water daarin onder druk deur die membraan-lumens. Die druk-puls en vloei-perturbasies lig, skuif en breek die koeklaag op. Tydens 3 agtereenvolgende "lig-en-vee" aksies word die koeklaag effektief opgebreek en uit die lumen gevee. Eksperimentele uitslae vir ononderbroke doodloop bedryf op uitermate aangevuilde membrane van 'n ultrafiltrasie loodsaanleg toegerus met terugdruk-pulsering, het getoon dat die vloed met 18.4 % verbeter kon word en die dP met 8.2 % verminder kon word in slegs 7.2 h. Die metode breek die koeklaag effektief op, en geen langtermyn vloed-afname is waargeneem vir meer as 555 h sedert die vorige chemiese was-prosedure nie.

Thesis (MScEng)--University of Stellenbosch, 2000. / ENGLISH ABSTRACT: Supercritical extraction is being investigated as a possible alternative to the processes currently used in the fractionation of paraffinic waxes. By removing the lighter carbon fractions from the wax, the wax hardness will be improved and its melting temperature range reduced, hence improving the performance of the wax product in certain applications. In order to evaluate and operate such an extraction process optimally, it is necessary to have a thermodynamic model that accurately represents the process system. There are, however, currently no predictive models available for these systems. In order to fit present models to the systems, accurate phase equilibrium data of the supercritical solvent - n-alkane systems are needed. Unfortunately, the amount of reliable published data on these systems in the required operating range is very limited. A view cell was designed and developed with which these high pressure equilibria could be studied. The binary phase equilibria of supercritical CO2 with n-CI2, n-CI6, n-C20, n-C24, n-C28 and n-C36 and of supercritical ethane with n-CI6, n-C24 and n-C28 were measured in the temperature range 313 - 367 K. It was found that the systems with these two solvents have very different types of phase behaviour. The n-alkane solubility is much higher in ethane, but supercritical CO2 will provide a much better degree of control over the selectivity achieved in an extraction process. Of the various equations of state investigated, it was found that the Patel Teja equation of state provided the best fit of the CO2 - n-alkane systems and that the Soave-Redlich- Kwong equation fitted the ethane - n-alkane systems the best. The interaction parameters of both these equations of state display a functional relationship with temperature and nalkane acentric factor, making it possible to determine parameter values for application at other operating temperatures and with other n-alkane systems. It was found that the current equations of state were not able to represent the phase equilibria accurately over the entire range of operating conditions. The poor performance of the equations of state can be attributed to inherent flaws in the existing equations of state. / AFRIKAANSE OPSOMMING: Superkritiese ekstraksie word tans ondersoek as 'n moontlike altematief vir die prosesse wat huidiglik gebruik word om paraffiese wasse te fraksioneer. Die Iigter koolstofwasse word verwyder om die washardheid te verhoog en die temperatuurgebied waaroor die was smelt te verklein. Dit verbeter dan die was se kwaliteit en werkverrigting. Modelle wat die superkritiese ekstraksie proses akkuraat kan voorstel word egter benodig om die ekstraksie proses te kan evalueer en optimaal te bedryf. Daar is tans geen modelle beskikbaar wat die proses direk kan voorstel nie. Akkurate fase-ewewigsdata word benodig om bestaande modelle aan te pas vir gebruik in hierdie sisteme. Daar is egter baie min betroubare faseewewigsdata vir die superkritiese oplosmiddel - n-alkaan sisteme beskikbaar in die literatuur. 'n Sig-sel, waarrnee hierdie hoe druk data gemeet kan word, is ontwerp en ontwikkel. Die volgende binere fase ewewigte is in die temperatuur gebied 313 - 367 K gemeet: superkritiese CO2 met n-CI2, n-CI6, n-C20, n-C24, n-C28 en n-C36, en superkritiese Etaan met n-CI6, n-C24 en n-C28. Daar is gevind dat hierdie twee superkritiese oplosmiddelsisteme verskillende tipes fase-ewewigsgedragte openbaar. Die n-alkane het 'n baie boer oplosbaarheid in Etaan, maar deur superkritiese C02 in 'n ekstraksie kolom te gebruik, sal tot beheer oor die selektiwiteit van die ekstraksieproses lei. Uit die verskillende toestandsvergelykings wat ondersoek is, is daar gevind dat die Patel- Teja vergelyking die CO2 sisteme die beste kon beskryf en dat die Soave-Redlich-Kwong vergelyking die beste vergelyking was om die Etaan sisteme mee te modelleer. Beide die toestandsvergelykings se interaksie parameters het 'n funksionele verband met temperatuur en die n-alkaan asentrise faktor getoon. Dit is dus moontlik om waardes vir die parameters vir sisteme by ander temperature en met ander n-alkaan tipes te bepaal. Daar was gevind dat die bestaande toestandsvergelykings nie die die fase-ewewigte oor die hele eksperimenele gebied akkuraat kon voorstel nie. Dit kan toegeskryf word aan foute wat inherent is aan die vergelykings.

Thesis (MEng)--University of Stellenbosch, 2000. / ENGLISH ABSTRACT: The growth in intelligent control has primarily been a reaction to the realisation that nonlinear control theory has been unable to provide practical solutions to present day control challenges. Consequently the chemical industry may be cited for numerous instances of overdesign, which result as an attempt to avoiding operation near or within complex (often more economically viable) operating regimes. Within these complex operating regimes robust control system performance may prove difficult to achieve using conventional (algorithmic) control methodologies. Biological neuronal control mechanisms demonstrate a remarkable ability to make accurate generalisations from sparse environmental information. Neural networks, with their ability to learn and their inherent massive parallel processing ability, introduce numerous opportunities for developing superior control structures for complex nonlinear systems. To facilitate neural network learning, reinforcement learning techniques provide a framework which allows for learning from direct interactions with a dynamic environment. lts promise as a means of automating the knowledge acquisition process is beguiling, as it provides a means of developing control strategies from cause and effect (reward and punishment) interaction information, without needing to specify how the goal is to be achieved. This study aims to establish evolutionary reinforcement learning as a powerful tool for developing robust neurocontrollers for application in highly nonlinear process systems. A novel evolutionary algorithm; Symbiotic, Adaptive Neuro-Evolution (SANE), is utilised to facilitate neurocontroller development. This study also aims to introduce SANE as a means of integrating the process design and process control development functions, to obtain a single comprehensive calculation step for maximum economic benefit. This approach thus provides a tool with which to limit the occurrence of overdesign in the process industry. To investigate the feasibility of evolutionary reinforcement learning in achieving these aims, the SANE algorithm is implemented in an event-driven software environment (developed in Delphi 4.0), which may be applied for both simulation and real world control problems. Four highly nonlinear reactor arrangements are considered in simulation studies. As a real world application, a novel batch distillation pilot plant, a Multi-Effect Batch Distillation (MEBAD) column, was constructed and commissioned. The neurocontrollers developed using SANE in the complex simulation studies, were found to exhibit excellent robustness and generalisation capabilities. In comparison with model predictive control implementations, the neurocontrollers proved far less sensitive to model parameter uncertainties, removing the need for model mismatch compensation to eliminate steady state off-set. The SANE algorithm also proved highly effective in discovering the operating region of greatest economic return, while simultaneously developing a neurocontroller for this optimal operating point. SANE, however, demonstrated limited success in learning an effective control policy for the MEBAD pilot plant (poor generalisation), possibly due to limiting the algorithm's search to a too small region of the state space and the disruptive effects of sensor noise on the evaluation process. For industrial applications, starting the evolutionary process from a random initial genetic algorithm population may prove too costly in terms of time and financial considerations. Pretraining the genetic algorithm population on approximate simulation models of the real process, may result in an acceptable search duration for the optimal control policy. The application of this neurocontrol development approach from a plantwide perspective should also have significant benefits, as individual controller interactions are so doing implicitly eliminated. / AFRIKAANSE OPSOMMING: The huidige groei in intelligente beheerstelsels is primêr 'n reaksie op die besef dat nie-liniêre beheerstelsel teorie nie instaat is daartoe om praktiese oplossings te bied vir huidige beheer kwelkwessies nie. Gevolglik kan talle insidente van oorontwerp in die chemiese nywerhede aangevoer word, wat voortvloei uit 'n poging om bedryf in of naby komplekse bedryfsgebiede (dikwels meer ekonomies vatbaar) te vermy. Die ontwikkeling van robuuste beheerstelsels, met konvensionele (algoritmiese ) beheertegnieke, in die komplekse bedryfsgebiede mag problematies wees. Biologiese neurobeheer megamsmes vertoon 'n merkwaardige vermoë om te veralgemeen vanaf yl omgewingsdata. Neurale netwerke, met hulle vermoë om te leer en hulle inherente paralleie verwerkingsvermoë, bied talle geleenthede vir die ontwikkeling van meer doeltreffende beheerstelsels vir gebruik in komplekse nieliniêre sisteme. Versterkingsleer bied a raamwerk waarbinne 'n neurale netwerk leer deur direkte interaksie met 'n dinamiese omgewing. Versterkingsleer hou belofte in vir die inwin van kennis, deur die ontwikkeling van beheerstrategieë vanaf aksie en reaksie (loon en straf) interaksies - sonder om te spesifiseer hoe die taak voltooi moet word. Hierdie studie beaam om evolutionêre versterkingsleer as 'n kragtige strategie vir die ontwikkeling van robuuste neurobeheerders in nie-liniêre prosesomgewings, te vestig. 'n Nuwe evolutionêre algoritme; Simbiotiese, Aanpasbare, Neuro-Evolusie (SANE), word aangewend vir die onwikkeling van die neurobeheerders. Hierdie studie beoog ook die daarstelling van SANE as 'n weg om prosesontwerp en prosesbeheer ontwikkeling vir maksimale ekonomiese uitkering, te integreer. Hierdie benadering bied dus 'n strategie waardeur die insidente van oorontwerp beperk kan word. Om die haalbaarheid van hierdie doelwitte, deur die gebruik van evolusionêre versterkingsleer te ondersoek, is die SANE algoritme aangewend in 'n Windows omgewing (ontwikkel in Delphi 4.0). Die Delphi programmatuur geniet toepassing in beide die simulasie en werklike beheer probleme. Vier nie-liniêre reaktore ontwerpe is oorweeg in die simulasie studies. As 'n werklike beheer toepassing, is 'n nuwe enkelladingsdistillasie kolom, 'n Multi-Effek Enkelladingskolom (MEBAD) gebou en in bedryf gestel. Die neurobeheerders vir die komplekse simulasie studies, wat deur SANE ontwikkel is, het uitstekende robuustheid en veralgemeningsvermoë ten toon gestel. In vergelyking met model voorspellingsbeheer implementasies, is gevind dat die neurobeheerders heelwat minder sensitief is vir model parameter onsekerheid. Die noodsaak na modelonsekerheid kompensasie om gestadigde toestand afset te elimineer, word gevolglik verwyder. The SANE algoritme is ook hoogs effektief vir die soek na die mees ekonomies bedryfstoestand, terwyl 'n effektiewe neurobeheerder gelyktydig vir hierdie ekonomies optimumgebied ontwikkel word. SANE het egter beperkte sukses in die leer van 'n effektiewe beheerstrategie vanaf die MEBAD toetsaanleg getoon (swak veralgemening). Die swak veralgemening kan toegeskryf word aan 'n te klein bedryfsgebied waarin die algoritme moes soek en die negatiewe effek van sensor geraas op die evaluasie proses. Vir industriële applikasies blyk dit dat die uitvoer van die evolutionêre proses vanaf 'n wisselkeurige begintoestand nie koste effektief is in terme van tyd en finansies nie. Deur die genetiese algoritme populasie vooraf op 'n benaderde modelop te lei, kan die soek tydperk na 'n optimale beheerstrategie aansienlik verkort word. Die aanwending van die neurobeheer ontwikkelingstrategie vanuit 'n aanlegwye oogpunt mag aanleiding gee tot aansienlike voordele, aaangesien individuele beheerder interaksies sodoende implisiet uitgeskakel word.

Thesis (MEng)--University of Stellenbosch, 2000. / ENGLISH ABSTRACT: Peach and apricot shells are agricultural waste products. These waste products accumulate around canneries and food-processing plants in South Africa. No effort is being made to utilise these waste products. This study is the first part of the product development from these products i.e. peach shell activated carbon and apricot shell activated carbon. By producing activated carbon from peach and apricot shells the solid waste problem is addressed, but most of all a profit can be made. But why activated carbon? Activated carbons are unique and versatile adsorbent with a vast amount of adsorption applications. It can be produced via a simple oxidation reaction with steam and the nature of peach and apricot shells are such that it is expected that activated carbons with good adsorption properties can be produced from it. The single largest consumer of activated carbon in South Africa is the gold mining industry that uses imported coconut shell activated carbon for gold adsorption in the gold recovery process. Activated carbon is also used as water purification adsorbents. During this study activated carbons were produced in a fluidized bed reactor at various activation conditions: 700 - 900°C, 0.0425 - 0.0629 g steamlg char.min and 30 - 60 min. This was done in order to find the optimum activation conditions within the activation parameter range. The optimal activated carbons were defined as peach and apricot shell activated carbons that showed good microporous as well as mesoporous character. The optimal activated carbons produced are: peach shell activated at 875°C, 0.0533 g stearnlg char. min, 60 min and apricot shell activated carbon at 850°C, 0.0533 g steamlg char.min, 60min. The possible use of these optimal activated carbons and two other activated carbons produced (Peach shell activated carbon 900°C, 0.0425 g steamlg char. min, 60 min and apricot shell activated carbon 900°C, 0.0425 g steamlg char.min, 60min) were tested in gold recovery and water purification. The gold adsorption properties of peach and apricot shell activated carbons were found to be better than two commercial coconut shell activated carbons (Chemquest 650 and GRC 22). No definite conclusions could, however, be drawn about the replacement of coconut shell activated carbon with peach or apricot shell activated carbon, because abrasion test work and thermal regeneration of the experimental carbons still have to be performed. The experimental activated carbons displayed good phenol adsorption characteristic, although further test work is required. / AFRIKAANSE OPSOMMING: Perske- en appelkoospitte is landbouafvalprodukte. Hierdie afvalprodukte versamel rondom inmaakfabrieke en voedselververkingsaanlegte. Tans word daar geen poging in Suid-Afrika aangewend om hierdie afvalprodukte te benut nie. Hierdie studie is die eerste deel van die ontwikkeling van die produkte: Perskepitdop-geaktiveerde koolstof en appelkoospitdop-geaktiveerde koolstof. Deur geaktiveerde koolstof van die perske- en appelkoospitdoppe te maak, word nie net 'n antwoord op die vastestofafvalsprobleem gevind nie, maar daar kan ook geld gemaak word. Hoekom geaktiveerde koolstof? Aktiveerde koolstowwe is veelsydige en unieke adsorbente met 'n groot verskeidenheid adsorpsie toepassings. Dit kan vervaardig word via 'n eenvoudige oksidasie reaksie met stoom en die aard van die perske- en appelkoospitdoppe is sodanig, dat verwag kan word om geaktiveerde koolstowwe met goeie adsorpsie eienskappe daarvan te kry. Die grootste enkelverbruiker van geaktiveerde koolstof in Suid-Afrika is die goudmynbedryf, wat kokosneutdop geaktiveerde koolstof invoer om goud te herwin. Geaktiveerde koolstof word ook gebruik vir watersuiwering. Tydens hierdie studie IS geaktiveerde koolstowwe by verskillende aktiveeringskondisies in 'n gevloeïdiseerde bed vervaardig: 700 - 900oe, 0.0425 - 0.0629g stoornlg gepiroliseerde pitdoppe.min en 30 - 60 mm. Die aktiveringskondisies is gevarieer om sodoende die optimale aktiveringskondisies binne die aktiveringsparameterreeks te kry. 'n Geaktiveerde koolstof is as optimaal geklassifiseer as dit 'n goeie mikro- sowel as mesostruktuur getoon het. Die optimaal geaktiveerde koolstowwe is: geaktiveerde koolstof vervaardig van perskepitdoppe by 875°e, 0.0533 g stoornlg gepiroliseerde pitdoppe.min, 60 mm en geaktiveerde koolstof vervaardig van appelkoospitdoppe by 850oe, 0.0533 g stoornlg gepiroliseerde pitdoppe.min, 60min. Die gebruik van die twee optimale geaktiveerde koolstowwe sowel as twee ander geaktiveerde koolstowwe (perskepitdop-geaktiveerde koolstof, 900oe, 0.0425 g stoornlg gepiroliseerde pitdoppe.min, 60 min en appelkoospitdop-geaktiveerde koolstof, 850°C, 0.0533 g stoom/g gepiroliseerde pitdoppe.min, 60min) is VIr goudadsorpsie en watersuiwering ondersoek. Die goudadsorpsie eienskappe van die perske-en appelkoospitdop-geaktiveerde koolstowwe was beter as die van twee kommersiële kokosneutdop-geaktiveerde koolstowwe (Chemquest 650 and GRC 22). Daar kan egter geen definitiewe gevolgtrekkings gemaak word oor die vervanging van kokosneutdop geaktiveerde koolstowwe met dié van perske of appelkoospitdoppe nie, aangesien daar nog toetsresultate oor die slytweerstand en reaktiverings eienskappe van die eksperimentele geaktiveerde koolstowwe uitstaande is. Die eksperimentele geaktiveerde koolstowwe toon goeie adsorpie ten opsigte van fenol, maar verdere toetswerk is egter nodig.

Phase equilibrium data is vital for designing chemical separation equipment. Traditionally, such data is obtained through laboratory experiments by sampling and analysing each phase of an equilibrated chemical mixture. An alternative means of generating such data is via molecular simulations, which also gives insight into the microscopic structure of the phases. This project was undertaken due to the lack of work on molecular simulations in predicting vapour-liquid-liquid equilibrium (VLLE). Gibbs Ensemble Monte Carlo molecular simulations were performed in the isochoricisothermal (NVT) and isobaric-isothermal (NVT) ensembles to determine the ability and limitations of the Transferable Potentials for Phase Equilibria (United-Atom) and Extended Simple Point Charge (SPC-E) force fields in predicting three-phase fluid equilibrium for two binary and three ternary industrially relevant mixtures: n-hexane/water (1), ethane/ethanol (2), methane/n-heptane/water (3), n-butane/1-butene/water (4) and nhexane/ ethanol/water (5). The NPT ensemble proved inadequate for predicting VLLE for binary mixtures, as for both binary mixtures (1 and 2), the simulations reverted to two phases. This was due in part to the unlike-pair interactions between pseudoatoms in different molecules not being accurately predicted at the specified simulation conditions to reproduce experimental mixture densities and vapour pressures. It was also due to the sensitivity of the NPT ensemble to perturbations which probably removed the system from its three-phase trajectory in Gibbs phase space, since specifying even the correct pressure corresponding to the potential models was unsuccessful in obtaining stable VLLE. Furthermore, ternary VLLE could not be obtained for a mixture exhibiting an extremely narrow three-phase region (4) and simulations for a miscible, non-ideal mixture (5) gave mole fractions that were in poor agreement with experiment. Good results were obtained for mixture 3 which exhibits limited mutual solubilities and a large three phase region. The NVT ensemble overcame the shortcomings of the NPT ensemble by producing three stable phases for the binary mixtures, revealing that the three-phase pressures were shifted by as much as 12%. Also, the narrow three-phase region of mixture 4 was overcome by adjusting the total system volume, producing three stable phases. These were also the first successful binary VLLE simulations involving complex polyatomic molecules. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, Durban, 2008.

Pinch Analysis, in the broadest sense, is concerned with the optimal use of resources (material or energy) in a multi-process system. Pinch Analysis based techniques have emerged for water systems over the past decade. A major assumption that has been made in applying these techniques is that a process system can be segregated into a set of process streams and a set of water streams. With this distinction in place, only the water streams are considered in the Pinch Analysis with the process streams represented implicitly. This approach has obvious limitations in situations where a clear distinction between process streams and water streams cannot be made. The chlor-alkali process is an example of a system in which the clear distinction between process streams and water streams cannot be made. Water is intrinsically involved in the process, serving as a carrier medium for raw materials and eventually becoming part of the products produced by the complex. Hydrochloric acid and caustic soda are reagents which are both used within and produced by the complex. These reagents are required by the process at a range of concentrations and the concentrated reagent is diluted to the required concentrations using demineralised water. Within the chlor-alkali complex, a number of effluents containing the reagent species are available and are typically sent to drain. It is conceivable that these effluents might be recovered and used for dilution purposes instead of demineralised water. This would bring about a reduction in the amount of water and concentrated reagent used and the amount of effluent produced by the complex. Given the economic value of these reagents relative to water, their recovery, if feasible, is likely to dominate the optimal water-use and effluent generation strategy. Current Water Pinch Analysis theory relies on the distinction being made between process streams and water streams and does not consider the recovery of reagents or the presence of desirable species within the system. In addition, the assumption is made that species are non-reactive; reactive species such as hydrogen chloride and sodium hydroxide, fall outside the scope of the current theory. The objectives of this study have included the development of an approach which is able to address these limitations of the existing theory. This approach, termed Combined Water and Materials Pinch Analysis seeks to identify optimal use strategies for raw materials and reagents, in addition to water-use and effluent generation. The approach combines mathematical programming with conceptual insights from Water Pinch Analysis. The approach is based on the optimisation of a superstructure which represents the set of all possible flow configurations for water, reagents and raw materials between the various operations within the process system; this problem is solved as a nonlinear programming (NLP) problem using standard optimisation tools. The application of the developed approach to the Sasol Polymers chlor-alkali complex at Umbogintwini, south of Durban, has been a further objective of this study. Given the variety of process operations present within the complex, which differ both in terms of their physical structure and function, individual process models for these operations were required. These models were described in terms of four basic functional elements, namely, mixing, flow separation, component separation and reaction, and incorporated into the superstructure. Given the complexity of the problem, the process system was divided into three subsystems which were optimised in isolation from each other. These results were subsequently integrated to reflect the performance of the subsystems in combination with each other. The results showed a potential reduction of 14% in water-use and 42% in effluent production by the complex, relative to the existing operating configuration. Amongst other savings in material use, the results indicated a 0.2% reduction in the use of salt, a 1.6% reduction caustic soda use and an 8.3% reduction in the use of hydrochloric acid. Economically, the potential saving identified was R 945 727 per annum, based on operating costs in the year 2000. The final objective of this study was the interpretation of the pinch as it relates to the Combined Water and Materials Pinch Analysis problem. A general definition of the pinch was proposed; according to this definition, the pinch corresponds to that constraint or set of constraints which limits the performance of the system, that is, prevents it from further improvement. For the Combined Water and Materials Pinch Analysis problem, this performance is measured in terms of the operating cost. This definition is thus a departure from its usual thermodynamic interpretation of the pinch; in addition, the pinch is defined in terms of a constraint or a set of constraints instead of a point. These constraints are identified by an analysis of the marginal values provided by the optimisation algorithm. Marginal values are also used as a means of identifying process interventions which may be effected such that the performance of the system may be improved further. / Thesis (M.Sc.Eng.)-University of Natal, 1989.

Due to the opportunity for afforestation in South Africa being severely limited, extensive research is being carried out on obtaining more wood per given area, improving the quality and value of the wood and on gaining a better understanding of how wood properties influence the characteristics of the pulp it produces. The last mentioned is the main focus of this study. If the variations in pulp, due to variations in wood properties, are better understood, then the existing fibre resource could be more efficiently managed and utilised to maximise its value. The main objective of this study was to determine how variation in physical and chemical properties, from the existing P.patula resource drives variation in pulp strength properties. It is well known that differences in tree age and site index lead to major sources of variation of various wood properties. These two easily measured variables were used in this study to capture a significant amount of variation in the wood of the aforementioned genus that enters a mill. Samples were obtained from two extremes in site quality, as measured by site index, (viz. good and poor sites) and three age ranges (viz. 9-10 years, 13-14 years and 20-21 years) from the KwaZulu-Natal Midlands. Wood chips from each of the six sites were pulped for various lengths of time, using the Kraft pulping method and under the same pulping conditions. A wide range of anatomical and chemical properties of wood and pulp were measured to characterise these samples as extensively as possible. All pulp samples were beaten in a PFI mill, at four different levels. The freeness values of the beaten samples were measured. The physical properties measured on handsheets made from the pulp included sheet density, burst index, tear index, tensile index, tensile energy absorbed, stretch and zero-span tensile strength. Regression models were developed to quantify the impact of the measured variables on each of the strength properties. Principal component analysis was performed on the wood characteristics measured and indicated that tree age is a more critical source of variation in wood properties than site index. The predictions of whole tree wood properties from the properties measured at breast height were excellent. Pulping studies showed that pulp yield, at kappa 20-30, increased with tree age. Wood cellulose proved to be a very good predictor of pulp yield. Xylose and mannose appeared to be most resistant to degradation during pulping. The low yield pulps were easier to refine than those with high yield. This has important implications when considering high yield pulping processes. With regard to pulp strength, the younger material could not achieve the high tear strengths obtained by the older material. However, the tear strengths achieved by these younger sites were comparable with, if not higher than, that obtained by hardwood species. The implications of this is that younger P. patula trees could be used for grades of paper where very high tear strength is not essential, but tensile is (e.g. tissue paper) and that older material can be better used for the purpose of providing the high tear strength needed by certain paper grades (e.g. linerboard and sackkraft). The maximum tensile strength achieved by the younger material was higher than that of the older material. When compared at constant freeness or sheet density, longer cooking times had a deleterious effect on strength properties. Strong predictions of pulp strength from basic wood properties were obtained when strength results were compared at constant freeness and sheet density. Because of the great influence of fibre morphology and chemistry on refining rates and on the resultant strength properties, the results of the study suggest that fibres of greatly different chemical and anatomical characteristics should not be refined together, if beating energy and pulp strength are to be optimised. However, further work is required to evaluate if the separation of fibre resources, to improve pulp quality, would be economically viable. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, Durban, 2005.

Biosorption is a property of certain type of inactive, microbial biomass to bind and concentrate heavy metals from even very dilute aqueous solutions. Biomass exhibits this property, acting just as a chemical substance, as an ion exchanger of biological origin. It is particularly the cell wall structure of certain algae that is found responsible for this phenomenon. In these experiments, the rate and extent for removal of copper is subjected to parameters such as pH, initial metal concentration, biosorbent size, contact time, temperature and the ability of the biomass to be regenerated in sorption-desorption experiments. The metal adsorption was found to be rapid within 25 minutes. The maximum copper uptake of 30 mg of copper / g of biomass has been observed, in the following conditions: 100 mg / L, 0.1 g of biomass, pH 4 and at temperature of 25°C. From this study, it was found that copper uptake is increasing with increase in pH, with optimum being pH 4. Copper uptake increases substantially from 0 to 25 minutes. Metal biosorption behaviour of raw seaweed Sargassum in six consecutive sorptiondesorption cycles were also investigated in a packed-bed column, during a continuous removal of copper from a 35 mg/l aqueous solution at pH 4. The sorption and desorption was carried out for an average of 85 and 15 hours, respectively, representing more than 40 days of continuous use of the biosorbent. The weight loss ofbiomass after this time was 13.5%. The column service time decreased from 25 hrs in the first cycle to 10 hrs for the last cycle. / Thesis (M.Sc.)-University of KwaZulu-Natal, 2005.

At SAPPI-Tugela a continuous Kraft wood chip digester operates in EMCC mode (extended modified continuous cooking). Chips are initially exposed to a NaOH / Na2S liquor at high temperature in the top section. The chips move downward in plug flow passing circumferential screens used to draw liquor for various circulations. About midway down the spent black liquor is removed and the chips enter the cooler bottom section where some further reaction and washing occurs. Liquor level and chip level are maintained close to each other near the top. Chips require 8-12 hours to pass through the digester, depending on the chip feed rate. The key parameter of interest at the digester exit is the Kappa number, which is a measure of the extent of delignification which has occurred. Different board and paper products require different Kappa number pulp feed. (Final properties such as tensile, tear and bursting strengths will also depend on the way fibres have been modified in the digestion). The objective of this investigation is to predict the Kappa number of the product pulp in real-time, thus facilitating quicker reaction than the present dependence on laboratory analysis permits, possibly even allowing closed-loop control. The extent of delignification depends on liquor strength, temperature and exposure time, with final Kappa number also depending on the properties of the chip feed (wood type and moisture content). Compensation to maintain a steady Kappa number is made difficult by the long and varying residence time, and the fact that any changes apply to the whole profile held up in the digester. A number of static models for Kappa number prediction have been developed by previous workers, but these do not compare well with plant measurements. The collection of data from the Sappi-Tugela reactor, and the pulp quality reports, have been used to determine an efficient model. This step required a considerable data collection exercise, and similar results to the quality reports have been obtained using a simple linear model based on this data. The problem of model error is being reduced by arrangement as a Smith Predictor, in which the model is intermittently corrected by available laboratory analyses. At the same time, an interface was created, in order to synchronise measurement data for the chips presently leaving the reactor. In order to deal with the dead time, each parcel of chips entering the reactor is effectively tracked, and the changes in Kappa number integrated for reaction time under the varying conditions in transit. Knowing the present inventory of the reactor, this model can also be run forward in time as a predictive controller, to determine optimal control actions for maintenance of the target Kappa number. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, 2005.