Thesis (MIng)--University of Stellenbosch, 2000. / ENGLISH ABSTRACT: Fouling is readily acknowledged as one of the most critical problems limiting the wider application of membranes in liquid separation processes. A better understanding of fouling layer formation and its monitoring is needed in order to improve on existing cleaning techniques. Plant operation can be optimised if fouling can be monitored by noninvasion means either on the plant itself or on an attached monitoring device. The overall scope of this research was to develop a non-destructive, real-time, in situ visualisation technique or device for concentration polarisation and fouling layer monitoring. Ultrasonic-time-domain-reflectometry (UTDR) was employed as a visualisation technique to provide real-time characterisation of the fouling layer. A 24 cm-long rectangular flat sheet aluminium cell was designed and used as separation device for a desalination system. The experimental results obtained using this module confirmed that there are an excellent correspondence between the flux decline behaviour and the UTDR response from the membrane. The ultrasonic technique could effectively detect fouling layer initiation and growth on the membrane in real-time. In addition to the measurement of fouling, the ultrasonic technique was also successfully employed for monitoring membrane cleaning. Since no real-time permeation data is available during cleaning operations in industrial applications, a UTDR monitoring device may prove to be a very valuable technique in optimising cleaning strategies. The technique was further tested on an 8-inch diameter spiral wrap industrial module and good results were obtained. Stagnant zones, as well as flux flow behaviour inside the module could be determined. However, more research IS needed to fully understand the complex phenomena inside a spiral wrap module. Overall, the UTDR technique and its use in monitoring devices have a major impact in the membrane industry due to its extremely powerful capabilities. / AFRIKAANSE OPSOMMING: Membraan-bevuiling of -verstopping is die grootste struikelblok wat die algemene aanwending van membrane vir verskillende watersuiweringsprosesse negatief beinvloed. 'n Beter begrip van membraan-bevuiling, asook beter metingsmetodes daarvan is nodig om op bestaande skoonmaaktegnieke te verbeter. Die hoofdoel van hierdie studie was die ontwikkeling van 'n nie-destruktiewe-in-lyn visuele tegniek vir die meting van konsentrasie polarisasie en membraan-bevuiling. Deur gebruik te maak van ultrasoniese klank golwe, is 'n tegniek ontwikkel wat 'n direkte visuele aanduiding kon gee van die toestand van membraan-bevuiling binnein die module. 'n Reghoekige aluminium-module, 24 cm lank, is ontwerp en gebou waarbinne die membraan geplaas is vir die skeidingsproses. Resultate dui daarop dat daar 'n uitstekende verband bestaan tussen die afname in permeaatvloei en die ultrasoniese eggo vanaf die membraan. Die ultrasoniese tegniek kon die vorming van en toename in membraan-bevuiling doeltreffend karakteriseer. In teenstelling hiermee, is die tegniek ook suksesvol aangewend om die skoonmaak-proses van membrane te ondersoek. Met min of geen data beskikbaar vir die skoonmaak-proses van membrane in die industriële sektor, het die tegniek enorme potensiaal in die optimisering van bestaande skoonmaak-tegnieke. Die tegniek is verder aangewend op 'n industriële 8-duim deursnee spiraal-module en goeie resultate is verkry. Stagnante sones asook vloed-vloei-patrone binne-in die module kon suksesvol bepaal word. Baie navorsing is egter nog nodig om die ingewikkelde data wat gegenereer word tydens die ondersoek van 'n spiraal-module ten volle te verstaan. Die enorme potensiaal en moontlikhede van die ultrasoniese tegniek kan die begin wees van 'n revolusie in die membraan-industrie.

<p>Pages that are darker in colour were scanned one at a time to produce a high quality image.</p> / <p>Following a review of the gas nitriding process, of the Fe-N system and of the formation of the compound layer, this study investigates the interaction between the diffusing nitrogen and the material to be nitrided.</p> <p>The formation of the compound layer and diffusion case is discussed for pure iron. AISI 1070 low alloy steel is investigated using different microstructures and the resulting effect on the nitrided microstructure is presented. The interaction of a nitride forming element is studied with a Fe/Fe-5Cr diffusion couple specimen. The strengthening mechanisms resulting from the nitriding process for all these materials are discussed using standard models from the literature. The scale of the precipitates produced in pure iron as well as in the 1070 steel pearlitic and spheroidized microstructure is found to provide marginal hardening. The interaction between N and carbides obtained during the nitriding of the 1070 martensitic microstructure gives rise to a significant hardness increase up to 130HV. The most significant hardening effect (up to 900HV) is obtained with the formation of a high density of fine CrN precipitates in the Fe-Cr specimens.</p> / Master of Applied Science (MASc)

Although more than a hundred papers dealing with the agglomeration problem in combustion and gasification of biomass can be found in the literature, very few studies focusing on the bed particle layer formation process in fluidized bed combustion (FBC) and fluidized bed gasification (FBG) can be found. With increased knowledge of the bed particle layer formation process — i.e. the main route behind bed agglomeration and bed material deposition in wood combustion/gasification — suitable combinations of fuel/bed material and/or bed material management measures can be suggested. This would not only aim to reduce the risk of ash related operational problems but also to enhance the catalytic activity of the bed material (e.g. for tar removal in gasification). The present investigation was therefore undertaken to determine the layer formation process on and within typical bed materials (i.e. quartz and olivine) and for a potentially interesting new bed material, K-feldspar. Bed material samples were collected from four different combustion and two different gasification appliances: two bubbling fluidized beds (BFB) (5 kWth/30 MWth), two full-scale circulating fluidized beds (CFB) (90/122 MWth), and two dual fluidized bed gasifiers (DFB) (8/15 MWth). Scanning electron microscopy/energy-dispersive spectroscopy (SEM/EDS) and X-ray diffraction (XRD) were used to explore layer morphology and elemental composition and to gain information about crystalline phases of the layers. Phase diagrams and thermodynamic equilibrium calculations (TECs) were used to interpret the melting behavior of the layers and the melt fragments in deposits. In addition, a diffusion model was used to interpret the layer growth process. For quartz bed particles taken from BFB, the younger particles (&lt; around 1 day) had only one thin layer, but for particles older than 3 days, the layer consisted of inner and outer layers. In addition to the inner and outer layers, a K-rich inner-inner layer was found for bed particles taken from CFB and DFB. No outer layers were found for quartz bed particles taken from DFB. The thin/absence of an outer layer could have resulted from the more significant attrition between particles in CFB and DFB. Reduced availability of Ca and a risk of layer breakage from the particle lead to the formation of the inner-inner layer. Similar elemental compositions of the layers upon the quartz bed particles taken from different fluidized bed techniques were found. The inner-inner layers are dominated by Si, K and Ca (excluding O), and the outer layers are rich in Ca, Si and Mg, which seem to resemble more closely the fuel ash composition. The inner layers, mainly consisted of Si and Ca, were found to have higher concentrations of Ca for older particles. The layer thickness increases with particle age, but the growth rate decreases. Melt was estimated to exist in the inner layer for younger particles (&lt; around 1 day) and in the inner-inner layer. The existence of partially melted inner-inner layers, in particles from CFB and DFB, points towards higher risk of bed agglomeration in these techniques compared to BFB. Based on the experimental results, thermodynamic equilibrium calculations, and diffusion model analyses, a layer formation process on quartz bed particle was suggested: the layer formation is initiated by reaction of gaseous K compounds with quartz to form K-rich silicate melt, which prompts the diffusion of Ca2+. The gradual incorporation of Ca into the melt followed by the precipitation of Ca-silicates, e.g. Ca2SiO4, will result in the continuous inner layer growth. However, because of increasing concentration of Ca and release of K from the inner layer, the melt disappears in the inner layer and the layer formation process gradually becomes Ca diffusion controlled. The diffusion resistance increases with increasing thickness of a more Ca-rich layer, resulting in a decreasing layer growth rate. Crack layers with similar compositions dominated by Si, K and Ca were observed in relatively old quartz bed particles. A melt was predicted to exist in the crack layer according to thermodynamic equilibrium calculations. The crack layers found in quartz particles from BFB and CFB connect with the cracks in the inner layer, whereas for bed samples collected from DFB, the crack layers were found along existing cracks in the quartz particle. The different morphologies may indicate different routes of formation for crack layers in bed particles from different fluidized bed technologies. For quartz particles from BFB and CFB, crack formation through the inner layer down to the interface between the inner layer and the core of quartz bed particle initiates the cracks in the quartz bed particle. This allows for diffusion of gaseous alkali compounds to react with quartz in the bed particle core, thereby forming crack layers. The reaction is accelerated with bridge formation between crack layers. This may later lead to the breakdown of the bed particle into smaller alkali-silicate-rich fragments. For K-feldspar bed particles from BFB and CFB, only one layer was found for particles with an age of 1 day. For bed particles with ages older than 3 days, two layers including a homogenous inner layer containing cracks and a more particle-rich outer layer can be distinguished. Compared to bed particles from BFB with similar ages, the outer layer is thinner for bed particles from CFB. The inner layer is dominated by Ca, Si and Al (excluding O), whereas the outer layer is dominated by Ca, Si and Mg. The average concentration of Ca in the inner layer increases with bed particle age. Increasing layer thickness with decreasing growth rate was found, similar to that on quartz particles. For particles from DFB, the inner layer is also mainly consisted of Ca and Si, but cracks in the inner layer were not found. For all the particles, the Ca/Si molar ratio in the layer decreases towards the bed particle core and the change of concentration is more significant at the bed particle core/layer interface. The overall inner layer growth is resultant from the gradual incorporation of Ca into the layer. For olivine bed particles from DFB, the younger bed particles (&lt; around 24 h) have only one layer, but after 24 h, an inner layer and an outer layer appear. Furthermore, for bed particles older than 180 h, the inner layer is separated into a distinguishable Ca-rich and Mg-rich zone. Two kinds of cracks in the inner layer either perpendicular or parallel to the particle surface were observed. Compared to the younger bed particles, the Ca concentration in the layer of older particles is much higher. A detailed mechanism for layer formation on olivine particles in fluidized bed gasification (most likely also applicable to combustion) based on the interaction between woody biomass ash and olivine has been proposed. The proposed mechanism is based on a solid-solid substitution reaction. However, a possible enabling step in the form of a Ca2+ transport via melts may occur. Ca2+ is incorporated into the crystal structure of olivine by replacing either Fe2+ or Mg2+. This substitution occurs via intermediate states where Ca-Mg silicates, such as CaMgSiO4, are formed. Mg2+ released from the crystal structure most likely forms MgO, which can be found in a distinguishable zone between the main particle layers. Due to a difference in the bond lengths between Mg/Fe and incorporated Ca2+ with their respective neighboring oxygen atoms, the crystal structure shifts, resulting in formation of cracks. The dominating elements in the inner layers are similar for each kind of bed material from BFB, CFB, and DFB, indicating limited effects of atmosphere on the inner layer formation. The initiation of layer formation differs depending on the bed material, but increasing Ca concentration in the inner layer with time for all bed materials indicates that the layer growth resulted from the incorporation of Ca into the layer. Compared to quartz, K-feldspar and olivine are more promising bed materials in wood combustion/gasification, especially in CFB and DFB techniques, from the perspective of mitigating bed agglomeration and bed material deposit build-up.

Die Querstrom-Mikrofiltration mit Kapillarmembranen kommt zunehmend in den Bereichen Lebensmittel-, Pharma-, Chemieindustrie sowie in der Umwelttechnik zum Einsatz. Eine vollständige Beschreibung der Deckschichtbildung innerhalb der Membrananlage ist jedoch noch nicht gelungen. Der erste Teil dieser Arbeit widmet sich der Kuchenbildung in einer Kapillarmembran. Dabei werden sowohl Änderungen der Strömungsverhältnisse entlang der Kapillarlänge als auch Klassiereffekte bei der Querstrom-Mikrofiltration von polydispersen Stoffsystemen berücksichtigt. Die Modellberechnungen wurden mit Experimenten validiert. Weiterhin erfolgte die Untersuchung des Einflusses von polymeren Flockungsmitteln (PFM) auf die Mikrofiltration und der Möglichkeiten einer Filtratstromerhöhung durch Flockung mittels PFM. Es wird gezeigt, dass eine optimale Zudosierung von PFM existiert und diese mit einer maximalen Filtratstromerhöhung verbunden ist. Eine Überdosierung von PFM soll jedoch verhindert werden.

Die Querstrom-Mikrofiltration mit Kapillarmembranen kommt zunehmend in den Bereichen Lebensmittel-, Pharma-, Chemieindustrie sowie in der Umwelttechnik zum Einsatz. Eine vollständige Beschreibung der Deckschichtbildung innerhalb der Membrananlage ist jedoch noch nicht gelungen. Der erste Teil dieser Arbeit widmet sich der Kuchenbildung in einer Kapillarmembran. Dabei werden sowohl Änderungen der Strömungsverhältnisse entlang der Kapillarlänge als auch Klassiereffekte bei der Querstrom-Mikrofiltration von polydispersen Stoffsystemen berücksichtigt. Die Modellberechnungen wurden mit Experimenten validiert. Weiterhin erfolgte die Untersuchung des Einflusses von polymeren Flockungsmitteln (PFM) auf die Mikrofiltration und der Möglichkeiten einer Filtratstromerhöhung durch Flockung mittels PFM. Es wird gezeigt, dass eine optimale Zudosierung von PFM existiert und diese mit einer maximalen Filtratstromerhöhung verbunden ist. Eine Überdosierung von PFM soll jedoch verhindert werden.

A non-reactive model system comprising a highly concentrated and unstable oil-in-water emulsion was used to investigate the retention of oil by the membrane in producing biodiesel with a membrane reactor. Critical flux was identified using the relationship between the permeate flux and transmembrane pressure along with the separation efficiency of the membrane. It was shown that separation efficiencies above 99.5% could be obtained at all operating conditions up to the critical flux. It was observed that the concentration of oil in all collected permeate samples using the oil-water system was below 0.2 wt% when operating at a flux below the critical flux. Studies to date have been limited to the characterization of low concentrated emulsions below 15 vol.%. The average oil droplet size in highly concentrated emulsions was measured as 3200 nm employing direct light scattering (DLS) measurement methods. It was observed that the estimated cake layer thickness of 20 to 80 mm was larger than external diameter of the membrane tube i.e. 6 mm based on a large particle size. Settling of the concentrated emulsion permitted the detection of a smaller particle size distribution (30-100 nm) within the larger particles averaging 3200 nm. It was identified that DLS methods could not efficiently give the droplet size distribution of the oil in the emulsion since large particles interfered with the detection of smaller particles. The content of the smaller particles represented 1% of the total weight of oil at 30°C and 5% at 70°C. This was too low to be detected using DLS measurements but was sufficient to affect ultrafiltration. In order to study the critical flux in the presence of transesterification reaction and the effect of cross flow velocity on separation, various oils were transesterified in another membrane reactor providing higher cross flow velocity. higher cross flow velocity provides better separation by reducing materials deposition on the surface of the membrane due to higher shearing. The oils tested were canola, corn, sunflower and unrefined soy oils (Free Fatty Acids (FFA< 1%)), and waste cooking oil (FFA= 9%). The quality of all biodiesel samples was studied in terms of glycerine, mono-glyceride, di-glyceride and tri-glyceride concentrations. The composition of all biodiesel samples were in the range required by ASTM D6751 and EN 14214 standards. A critical flux based on operating pressure in the reactor was reached for waste cooking and pre-treated corn oils. It was identified that the reaction residence time in the reactor was an extremely important design parameter affecting the operating pressure in the reactor. / Natural Sciences and Engineering Research Council of Canada (NSERC)

A non-reactive model system comprising a highly concentrated and unstable oil-in-water emulsion was used to investigate the retention of oil by the membrane in producing biodiesel with a membrane reactor. Critical flux was identified using the relationship between the permeate flux and transmembrane pressure along with the separation efficiency of the membrane. It was shown that separation efficiencies above 99.5% could be obtained at all operating conditions up to the critical flux. It was observed that the concentration of oil in all collected permeate samples using the oil-water system was below 0.2 wt% when operating at a flux below the critical flux. Studies to date have been limited to the characterization of low concentrated emulsions below 15 vol.%. The average oil droplet size in highly concentrated emulsions was measured as 3200 nm employing direct light scattering (DLS) measurement methods. It was observed that the estimated cake layer thickness of 20 to 80 mm was larger than external diameter of the membrane tube i.e. 6 mm based on a large particle size. Settling of the concentrated emulsion permitted the detection of a smaller particle size distribution (30-100 nm) within the larger particles averaging 3200 nm. It was identified that DLS methods could not efficiently give the droplet size distribution of the oil in the emulsion since large particles interfered with the detection of smaller particles. The content of the smaller particles represented 1% of the total weight of oil at 30°C and 5% at 70°C. This was too low to be detected using DLS measurements but was sufficient to affect ultrafiltration. In order to study the critical flux in the presence of transesterification reaction and the effect of cross flow velocity on separation, various oils were transesterified in another membrane reactor providing higher cross flow velocity. higher cross flow velocity provides better separation by reducing materials deposition on the surface of the membrane due to higher shearing. The oils tested were canola, corn, sunflower and unrefined soy oils (Free Fatty Acids (FFA< 1%)), and waste cooking oil (FFA= 9%). The quality of all biodiesel samples was studied in terms of glycerine, mono-glyceride, di-glyceride and tri-glyceride concentrations. The composition of all biodiesel samples were in the range required by ASTM D6751 and EN 14214 standards. A critical flux based on operating pressure in the reactor was reached for waste cooking and pre-treated corn oils. It was identified that the reaction residence time in the reactor was an extremely important design parameter affecting the operating pressure in the reactor. / Natural Sciences and Engineering Research Council of Canada (NSERC)

A non-reactive model system comprising a highly concentrated and unstable oil-in-water emulsion was used to investigate the retention of oil by the membrane in producing biodiesel with a membrane reactor. Critical flux was identified using the relationship between the permeate flux and transmembrane pressure along with the separation efficiency of the membrane. It was shown that separation efficiencies above 99.5% could be obtained at all operating conditions up to the critical flux. It was observed that the concentration of oil in all collected permeate samples using the oil-water system was below 0.2 wt% when operating at a flux below the critical flux. Studies to date have been limited to the characterization of low concentrated emulsions below 15 vol.%. The average oil droplet size in highly concentrated emulsions was measured as 3200 nm employing direct light scattering (DLS) measurement methods. It was observed that the estimated cake layer thickness of 20 to 80 mm was larger than external diameter of the membrane tube i.e. 6 mm based on a large particle size. Settling of the concentrated emulsion permitted the detection of a smaller particle size distribution (30-100 nm) within the larger particles averaging 3200 nm. It was identified that DLS methods could not efficiently give the droplet size distribution of the oil in the emulsion since large particles interfered with the detection of smaller particles. The content of the smaller particles represented 1% of the total weight of oil at 30°C and 5% at 70°C. This was too low to be detected using DLS measurements but was sufficient to affect ultrafiltration. In order to study the critical flux in the presence of transesterification reaction and the effect of cross flow velocity on separation, various oils were transesterified in another membrane reactor providing higher cross flow velocity. higher cross flow velocity provides better separation by reducing materials deposition on the surface of the membrane due to higher shearing. The oils tested were canola, corn, sunflower and unrefined soy oils (Free Fatty Acids (FFA< 1%)), and waste cooking oil (FFA= 9%). The quality of all biodiesel samples was studied in terms of glycerine, mono-glyceride, di-glyceride and tri-glyceride concentrations. The composition of all biodiesel samples were in the range required by ASTM D6751 and EN 14214 standards. A critical flux based on operating pressure in the reactor was reached for waste cooking and pre-treated corn oils. It was identified that the reaction residence time in the reactor was an extremely important design parameter affecting the operating pressure in the reactor. / Natural Sciences and Engineering Research Council of Canada (NSERC)

A non-reactive model system comprising a highly concentrated and unstable oil-in-water emulsion was used to investigate the retention of oil by the membrane in producing biodiesel with a membrane reactor. Critical flux was identified using the relationship between the permeate flux and transmembrane pressure along with the separation efficiency of the membrane. It was shown that separation efficiencies above 99.5% could be obtained at all operating conditions up to the critical flux. It was observed that the concentration of oil in all collected permeate samples using the oil-water system was below 0.2 wt% when operating at a flux below the critical flux. Studies to date have been limited to the characterization of low concentrated emulsions below 15 vol.%. The average oil droplet size in highly concentrated emulsions was measured as 3200 nm employing direct light scattering (DLS) measurement methods. It was observed that the estimated cake layer thickness of 20 to 80 mm was larger than external diameter of the membrane tube i.e. 6 mm based on a large particle size. Settling of the concentrated emulsion permitted the detection of a smaller particle size distribution (30-100 nm) within the larger particles averaging 3200 nm. It was identified that DLS methods could not efficiently give the droplet size distribution of the oil in the emulsion since large particles interfered with the detection of smaller particles. The content of the smaller particles represented 1% of the total weight of oil at 30°C and 5% at 70°C. This was too low to be detected using DLS measurements but was sufficient to affect ultrafiltration. In order to study the critical flux in the presence of transesterification reaction and the effect of cross flow velocity on separation, various oils were transesterified in another membrane reactor providing higher cross flow velocity. higher cross flow velocity provides better separation by reducing materials deposition on the surface of the membrane due to higher shearing. The oils tested were canola, corn, sunflower and unrefined soy oils (Free Fatty Acids (FFA< 1%)), and waste cooking oil (FFA= 9%). The quality of all biodiesel samples was studied in terms of glycerine, mono-glyceride, di-glyceride and tri-glyceride concentrations. The composition of all biodiesel samples were in the range required by ASTM D6751 and EN 14214 standards. A critical flux based on operating pressure in the reactor was reached for waste cooking and pre-treated corn oils. It was identified that the reaction residence time in the reactor was an extremely important design parameter affecting the operating pressure in the reactor. / Natural Sciences and Engineering Research Council of Canada (NSERC)