Search

ispartof: Metaforum visietekst 9 pages:1-89 status: published

This research investigates the impact of tannery effluent on the self purification capacity of natural streams by biodegradation and on sustaining macroinvertebrates. Water samples were collected from upstream (US) and downstream (DS) of the discharge point of four tanneries along the course of the Tinishu Akaki river (Ethiopia) for physicochemical parameters and macroinvertebrate analysis. As the concentration of chromium and sulfide increases from US to DS of the river, the removal of SS (Suspended Solids), and reduction of BOD5 (Biological Oxygen Demand) and COD (Chemical Oxygen Demand) by the self purification process decreases from 71%, 59% and 50%, respectively, to 10%, 30% and 28%, respectively, while also the nitrification capacity becomes severely hampered. The macroinvertebrate %EPT (ratio of total number of individuals classified as Ephemeroptera, Plecoptera, and Trichoptera to the total number of individuals), which is a measure of the presence of pollution sensitive macroinvertebrates decreased from 24% to zero, starting downstream from the third tannery, indicating severe pollution. The diversity index and correlation values between the concentration of physicochemical parameters, BOD5 reduction, the macroinvertebrate count and the chromium concentration further support these findings. The conclusion from this study is that, although the physicochemical parameters might indicate that, given enough distance between the tanneries, self recovery of the river can be gained for a while, the biodiversity and general health of the river is severely compromised such that aquatic life can hardly be sustained ispartof: location:Mekelle, Ethiopia status: accepted

Transfected Chinese hamster ovary cells were cultured on bare uncoated chemical vapor deposited thin nano- and microcrystalline diamond surfaces, hydrophobic hydrogenand hydrophilic oxygen-terminated. Optical and biochemical analyses show that compared to glass controls, growth and viability were not significantly affected (one-way-analysis of variance, ANOVA). Based on two-way ANOVA analyses, neither grain size nor surface termination had a significant influence until 5 days post-seeding. The authors thank Mrs. R. Beenaerts, Mr. J. Janssen, Mr. P. Pirotte, Mr. J. Soogen, Mr. R. van Werde for expert laboratory assistance, Prof. Dr. I. Lambrichts and Mr.M. Jans for logistics and fixation procedures, Mr. K. Vanstreels and Mr. B. Ruttens for assistance with the scanning electron microscope, Prof. Dr. R. Valcke for the use of his spectrophotometers, Mr. J. Duchateau, Mr. F. Horemans, Mrs. V. Vrindts for cleaning the diamond substrates, Ir. L. Naelaerts (KHLIM) for lending stereomicroscopes. S. W. is a post-doctoral research fellow of the Research Foundation – Flanders (FWO-Vlaanderen). This work was funded by the Research Council of the UHasselt, tUL, the K.U. Leuven (GOA/2006/02) and by a PhD grant of the Institute for the Promotion of Innovation through Science and Technology in Flanders (IWT-Vlaanderen). Support by IAP P6/27 Functional Supramolecular Systems (BELSPO) and by the FWOonderzoeksgemeenschap “Scanning andWide Field Microscopy of (Bio)-organic Systems” is gratefully acknowledged.

In view of the high resistance to ischemic injury of distal tubular cells (see section 1.5.1), a distal tubular cell line is used in this study as an in vitro model to examine some of the cellular defence processes occurring in surviving renal cells during energy depletion. To allow a detailed study of some cellular events and to investigate the relationships between them, a cell line with a low rate metabolism was chosen: the A6 cell line. Moreover, this cell line is a well-characterised model tight epithelium, which resembles the principal cells in mammalian collecting tubules, and that allows electrophysiological measurements to follow transepithelial transport characteristics. To enable us to understand the survival of distal tubular cells in ischemic conditions with threatening high apical sodium levels in the distal tubular lumen, we first characterised the dependency of Na· transport characteristics on the apical Na· concentration (Chapter 3) in normal conditions. Moreover, since there is convincing evidence in literature that intracellular acidification in ischemic conditions might be cytoprotective (see section 1.5.2.2), we determined first the effects of intracellular pH-shifts on Na· transport in control conditions (Chapter 4). A second objective was to investigate the time course of changes in morphology and to determine when irreversible cell death features, such as plasma membrane damage occur in A6 cells facing metabolic inhibition (MI) (Chapter 5). This allowed us to establish a 45 min protocol of MI, since in this time period no detectable cell death occurred and hence the epithelial cells were expected to be sub lethally injured at most. Since A6 cells, like mammalian distal tubular epithelia, are active Na· transporting epithelia, we then focused on the suppression of Na· transport in energy-depleted conditions. The third aim of this study was to look for possible cellular defence strategies related with mechanisms underlying Na· transport suppression during MI. Moreover, it was examined whether transcellular Na· transport and cellular energy levels recovered after MI (Chapter 6). One of the nefast consequences of renal ischemia is excessive tubular epithelial cell swelling, which contributes to renal dysfunction by inducing obstruction of the tubular lumen and vascular congestion in the outer medulla (Mason et al., 1989; Lang et al., 1995). Therefore, the last aim of this thesis was to examine in our experimental model the extent of cell swelling during MI, and to investigate whether metabolically challenged cells still possess the capacity to regulate their cell volume in iso- as well as in anisotonic conditions (Chapter 7).

This paper is the second in a series of two dealing with the spatial optimization of the heat exchanger temperature profiles of exothermic tubular reactors under the assumption of steady state and plug flow characteristics. The minimum principle of Pontryagin (optimal control theory) is applied in a straightforward, analytical sense. To enable a trade-off between process performance and global heat loss a weighted cost criterion is defined. In Part I of this series, it is demonstrated that (weighted combinations of) terminal costs give rise to control inputs of bang-bang type. In this Part II, the terminal cost criterion is extended with an integral term which accounts for the global heat loss during the process. Analysis of the extremal control sequence reveals that integral cost terms generically induce control of bang-singular-bang type. The singular arc is carefully analyzed. Furthermore it is illustrated that a practically more feasible bang-bang profile exhibits near optimal performance.

This series of two papers deals with the spatial optimization of the heat exchanger temperature profile of an exothermic tubular reactor under the assumption of steady state and plug flow characteristics. The minimum principle of Pontryagin (optimal control theory) is applied in a straightforward, analytical sense. To enable a trade-off between process performance and heat loss a combined cost criterion is defined. In Part I of this series terminal costs –related to outlet reactant concentration and to outlet reactor temperature– are analyzed. It is generically proven that the optimal control input is of bang-bang type with no singular arc possible. This control keeps the heat exchanger temperature constant at its maximum or minimum value. In Part II of this series the terminal cost criterion is extended with an integral term that accounts for the global heat loss during the process.

Climate change and associated co-occurrences of increasing water demands and diminishing water supplies have elevated water scarcity concerns from the local to the global level. Advanced water treatment needs to be expanded to allow for water reuse and combat water scarcity. However, the water treatment sector alone already accounts for 3% of total energy consumed in the U.S. and, as such, substantially contributes to greenhouse gas emissions and thus climate change. This dissertation project aims to support the development of novel domestic wastewater treatment systems capable of producing energy instead of consuming it, while achieving excellent effluent quality to allow for water reuse for irrigation or even drinking water production (with additional treatment). By integrating operational, microbial, and life cycle information to guide engineering design, this dissertation research will support the development and subsequent implementation of creative and sustainable approaches to wastewater treatment. status: published

As an efficient technology for wastewater treatment, membrane bioreactors (MBRs) have gained a lot of popularity over the recent years. In an MBR, the secondary biological treatment step is combined with a pressure driven membrane filtration, which results in an effluent of consistently high quality, reliable operation and a reduced footprint in comparison with conventional activated sludge systems that rely on sedimentation. However, fouling, which is the undesired accumulation of retained material on the membrane, remains the main drawback of MBRs. Fouling results in a performance loss through the decline of flux in constant pressure systems, or an increase in required transmembrane pressure in constant flux systems. In this work, the primary aim is to uncover the activated sludge characteristics that are responsible for the membrane fouling and to uncover the mechanisms behind this phenomenon. Three goals were put forward. The first goal was the creation of a framework for bioflocculation and filterability monitoring. Once this framework was set up, the secondary goal was the experimental determination of the relationship between bioflocculation and activated sludge separation. Finally, in the third goal, the specific role of cations for membrane fouling was studied to unravel insights that may validate or disprove existingtheories. The first part of this work tackles the first goal. The activated sludge image analysis program that was previously developed within the division, was modified to accomodate images from the new, more advanced, Olympus IX83 microscope. Improvements include: better segmentation and recognition routines, no more color-image requirements and the ability to analyse images of variable size. Twenty 3x3 images are deemed enough to provide statistically significant results. In the next chapter, the use of quantitative fluorescent image analysis for bioflocculation monitoring was explored. To this end, two image analysis routines were programmed as well. Alexa405-NHS ester was found to provide superior results over fluorescein isothiocyanate for labeling proteins and the results correlate well with the protein content from the heat extracted extracellular polymeric substances (eEPS). The use of the concanavalin A-Alexa488 conjugate, which labels α-polysaccharides, did not yield any eEPS correlation, but showed distinct localisation. Nile Red staining was tested against the microbial adhesion to hydrocarbons measurement, but could only differentiate between inorganic and organic fractions in terms of hydrophobicity. Finally, a fluorescence based method for Gram staining was semi-quantitatively compared to the conventional method and yielded good similarities. In the second part of this work, the focus was on finding correlations between activated sludge characteristics and sludge separation efficiencies. To this end, a measuring campaign was performed on sludge samples from sixteen different wastewater treatment systems. On the data, a multivariate statistical analysis, using partial least squares, was performed to unravel the most influential sludge characteristics for cross-flow and dead-end filtration and settling. In the next chapter, two longterm labscale experiments were performed in which synthetic glucose-based influent was differentiated in terms of cation composition. In accordance with the divalent cation bridging (DCB) theory, changing the monovalent-over-polyvalent (M/P) cation ratio to a high value (17) resulted in deflocculation, which significantly worsened filtration. Changing the ratio back to a low value (2.04) remediated the filterability issues and restored the floc structures over a few weeks. During these experiments, conventional systems were found to be more resilient and recovered faster from such disturbance than membrane systems. The data from both experiments were analysed together with the data from the measuring campaign in the final chapter of the second part. From the results, it is concluded that colloidal and small particulate matter are the most detrimental constituents of activated sludge, together with the polysaccharide fraction of the soluble microbial products (SMP). The morphology of the sludge is found to be of influence as fractal dimension and fouling rates appeared to be correlated. A sludge with a high relative hydrophobicity and a low surface charge proves to be beneficial. For filtration, a reasonable number of filaments was shown to provide a protective barrier against fouling. From the results it became clear that dead-end filtration tests cannot be used for MBR-filterability assessment. Moreover, dead-end filtration appeared to be more influenced by the deflocculation and disturbances in the sludge’s propertiessince the total amount of formed cake was higher. An abundance of filaments appeared to be the main source for settling problems, but sludge density and surface roughness played a significant role as well. No significant effect was found for eEPS on neither separation performance indicator. For the third goal, the effect of cations on bioflocculation and filtration performance was studied. The DCB theory could explain the long term phenomena. A unified theory, combining results from this and previous works was proposed on the basis of the DCB theory: a period of shel l shedding preceeds the more profound deflocculation phase. For the shortterm effects, the osmotic pressure effect was brought forward and experimentally tested. The osmotic pressure effect comes from the accumulation of charged groups in front of the membrane as a consequence of the total sludge retention. Functional groups on flocs and SMP require counterions to be present to preserve electro-neutrality, which leads to a locally increased ionic strength. This ionic strength difference across the membrane leads to a difference in osmotic pressure, which requires an extra driving force during filtration. Different ionic strengths and M/P ratios were tested but the osmotic pressure effect could not be validated or disproved. At high ionic strengths (I>350 mM), filterability worsened, regardless of the M/P ratio. For the high M/P this was attributed to the shortterm deflocculation effect, in accordance with the DCB theory. For the low M/P, precipitation or sweep flocculation,creating small (colloidal) particles, probably played a role. 1 Introduction 1 1.1 Why treat wastewater? . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Biological wastewater treatment . . . . . . . . . . . . . . . . . . 2 1.2.1 MBR compared to CAS . . . . . . . . . . . . . . . . . . 3 1.3 Objectives of this dissertation . . . . . . . . . . . . . . . . . . . 4 1.4 Chapter-by-chapter overview . . . . . . . . . . . . . . . . . . . 6 2 Membrane bioreactors (MBRs) and fouling 9 2.1 MBR fundamentals . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.1.1 What is an MBR? . . . . . . . . . . . . . . . . . . . . . 9 2.1.2 Cross-flow versus dead-end . . . . . . . . . . . . . . . . 12 2.1.3 Membranes and configurations used in MBRs . . . . . . 13 2.1.4 Constant flux or constant TMP mode of operation . . . 17 2.2 Membrane fouling . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.2.1 What is membrane fouling? . . . . . . . . . . . . . . . . 18 2.2.2 Critical, sustainable and threshold fluxes for fouling . . 22 2.3 Relating biomass properties to sMBR fouling . . . . . . . . . . 25 2.3.1 Membrane and module characteristics . . . . . . . . . . 25 2.3.2 Mixed liquor and biomass (bulk) characteristics . . . . . 28 2.3.3 Floc characteristics . . . . . . . . . . . . . . . . . . . . . 35 2.3.4 Compressibility . . . . . . . . . . . . . . . . . . . . . . . 39 2.3.5 Filaments . . . . . . . . . . . . . . . . . . . . . . . . . . 39 2.3.6 Dynamic secondary membrane hypothesis . . . . . . . . 40 2.3.7 Osmotic pressure effect . . . . . . . . . . . . . . . . . . 41 3 Bioflocculation 47 3.1 Bioflocculation fundamentals . . . . . . . . . . . . . . . . . . . 47 3.1.1 Structure of activated sludge flocs . . . . . . . . . . . . 47 3.1.2 Hydrophobic interactions . . . . . . . . . . . . . . . . . 49 3.1.3 Interactions between cations and bioflocculation . . . . 50 3.2 Monitoring bioflocculation . . . . . . . . . . . . . . . . . . . . . 52 3.2.1 Microscopic imaging and image analysis . . . . . . . . . 52 3.2.2 Laser-based particle size detection methods . . . . . . . 60 3.2.3 Floc strength . . . . . . . . . . . . . . . . . . . . . . . . 63 3.2.4 Chemometrics: principal component analysis and partial least squares . . . . . . . . . . . . . . . . . . . . . . . . 64 4 Materials and methods 67 4.1 Activated sludge characteristics . . . . . . . . . . . . . . . . . . 67 4.1.1 Sludge and effluent solids concentration . . . . . . . . . 67 4.1.2 Chemical oxygen demand (COD) . . . . . . . . . . . . . 68 4.1.3 Relative hydrophobicity . . . . . . . . . . . . . . . . . . 68 4.1.4 Surface charge . . . . . . . . . . . . . . . . . . . . . . . 69 4.1.5 Extracellular polymeric substances . . . . . . . . . . . . 69 4.1.6 Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . 72 4.2 Separation performance indicators . . . . . . . . . . . . . . . . 76 4.2.1 Sludge volume index (SVI) . . . . . . . . . . . . . . . . 76 4.2.2 Dead-end filtration characteristics . . . . . . . . . . . . 77 4.2.3 Small-scale cross-flow filterability assessment . . . . . . 77 4.3 Partial least squares (PLS) . . . . . . . . . . . . . . . . . . . . 78 4.3.1 Definitions and principles . . . . . . . . . . . . . . . . . 79 4.3.2 Loading-plots and predicted/observed plots . . . . . . . 80 4.3.3 β PLS regression coefficients . . . . . . . . . . . . . . . 81 4.3.4 Variable influence of projection (VIP) . . . . . . . . . . 82 4.3.5 Selecting the number of PLS-components . . . . . . . . 83 I Monitoring bioflocculation through microscopy and image analysis 86 5 Improvements on the activated sludge image analysis program (ASIAP) 87 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 5.2 A brief introduction to the former ACTIAS program . . . . . . 88 5.2.1 Segmentation in ACTIAS . . . . . . . . . . . . . . . . . 88 5.2.2 Recognition in ACTIAS . . . . . . . . . . . . . . . . . . 95 5.2.3 Data extraction in ACTIAS . . . . . . . . . . . . . . . . 97 5.3 Improvements and adaptations to ASIAP . . . . . . . . . . . . 98 5.3.1 Format compatibility . . . . . . . . . . . . . . . . . . . . 98 5.3.2 Grayscale imaging . . . . . . . . . . . . . . . . . . . . . 98 5.3.3 Improvements on segmentation . . . . . . . . . . . . . . 99 5.3.4 Improvements on the recognition of filaments . . . . . . 103 5.3.5 Improvements on the recognition of fragments: absolute size criterion . . . . . . . . . . . . . . . . . . . 107 5.3.6 Improvements on data extraction . . . . . . . . . . . . . 109 5.4 Statistical analysis to determine the required number of images 112 5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 6 Explorative study on the use of epifluorescence microscopy for bioflocculation monitoring 117 6.1 Selected dyes and staining procedures . . . . . . . . . . . . . . 117 6.1.1 Dye description . . . . . . . . . . . . . . . . . . . . . . . 117 6.1.2 Staining and visualization procedures . . . . . . . . . . 122 6.2 Quantitative fluorescent image analysis . . . . . . . . . . . . . . 123 6.2.1 ASIAP-Fluo: a fluorescence add-on for ASIAP . . . . . 123 6.2.2 FLASIAP: a standalone fluorescent image analysis program126 6.3 A partial least squares regression analysis between quantitative fluorescence methods and volumetric measurements . . . . . . . 127 6.3.1 Selected methods . . . . . . . . . . . . . . . . . . . . . . 127 6.3.2 General fluorescence observations . . . . . . . . . . . . . 127 6.3.3 Comparing quantitative fluorescence measurements with chemo-colorimetric techniques . . . . . . . . . . . . . . . 129 6.4 A novel method for Gram staining: evaluation and semi- quantitative comparison . . . . . . . . . . . . . . . . . . . . . . 134 6.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 II Bioflocculation versus activated sludge separation 139 7 A partial least squares study unravelling the relationship between bioflocculation and activated sludge separation 140 7.1 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . 141 7.1.1 Sample origins . . . . . . . . . . . . . . . . . . . . . . . 141 7.1.2 Sludge characterisation . . . . . . . . . . . . . . . . . . 141 7.1.3 Performance indicators . . . . . . . . . . . . . . . . . . . 141 7.1.4 Data analysis . . . . . . . . . . . . . . . . . . . . . . . . 142 7.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . 143 7.2.1 Reversible fouling - Average fouling rate . . . . . . . . . 144 7.2.2 Irreversible fouling - Global fouling rate . . . . . . . . . 147 7.2.3 Settleability - Sludge volume index . . . . . . . . . . . . 149 7.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 8 A study on the longterm interactions between cations and activated sludge separation 154 8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 8.2 Experiment I: simultaneous M/P differentiation in conventional activated sludge systems . . . . . . . . . . . . . . . . . . . . . . 155 8.2.1 Materials and methods . . . . . . . . . . . . . . . . . . . 155 8.2.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 8.2.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 169 8.3 Experiment II: temporal M/P differentiation in a conventional activated sludge system and a membrane bioreactor . . . . . . 172 8.3.1 Materials and methods . . . . . . . . . . . . . . . . . . . 172 8.3.2 MBR filtration - trends explained . . . . . . . . . . . . . 175 8.3.3 Comparison MBR and CAS . . . . . . . . . . . . . . . . 184 8.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 9 Bioflocculation versus activated sludge separation: what have we learned? 188 9.1 Cross-flow filterability . . . . . . . . . . . . . . . . . . . . . . . 189 9.1.1 Extracellular polymeric substances . . . . . . . . . . . . 189 9.1.2 Hydrophobicity and surface charge . . . . . . . . . . . . 191 9.1.3 Particle sizes and distribution . . . . . . . . . . . . . . . 192 9.1.4 Floc morphology . . . . . . . . . . . . . . . . . . . . . . 193 9.1.5 Filaments . . . . . . . . . . . . . . . . . . . . . . . . . . 194 9.2 Dead-end filterability . . . . . . . . . . . . . . . . . . . . . . . . 194 9.2.1 Particle sizes and distribution . . . . . . . . . . . . . . . 194 9.2.2 Relative hydrophobicity and surface charge . . . . . . . 194 9.2.3 Filaments . . . . . . . . . . . . . . . . . . . . . . . . . . 195 9.3 Settleability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 9.3.1 Extracellular polymeric substances . . . . . . . . . . . . 195 9.3.2 Particle sizes and distribution . . . . . . . . . . . . . . . 196 9.3.3 Filaments . . . . . . . . . . . . . . . . . . . . . . . . . . 196 9.3.4 Relative hydrophobicity and surface charge . . . . . . . 197 9.4 A unified theory on the longterm effects of cations on flocculation197 III Shortterm effects of cations on filtration 200 10 What is the impact of osmotic pressure on membrane fouling? 201 10.1 Introduction and rationale . . . . . . . . . . . . . . . . . . . . . 201 10.2 Preliminary experiments . . . . . . . . . . . . . . . . . . . . . . 202 10.2.1 Materials and methods . . . . . . . . . . . . . . . . . . . 202 10.2.2 Results - MBR sludge . . . . . . . . . . . . . . . . . . . 203 10.2.3 Results - CAS sludge . . . . . . . . . . . . . . . . . . . . 203 10.2.4 Results - municipal . . . . . . . . . . . . . . . . . . . . . 203 10.2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 205 10.3 Supernatant composition . . . . . . . . . . . . . . . . . . . . . 205 10.3.1 Ionic strength and composition of supernatant . . . . . 205 10.3.2 Theoretical calculation of osmotic pressure effect . . . . 207 10.4 Impact of salt solution composition . . . . . . . . . . . . . . . . 208 10.4.1 Effect of ionic strength . . . . . . . . . . . . . . . . . . . 209 10.4.2 Effect of M/P ratio . . . . . . . . . . . . . . . . . . . . . 212 10.4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 214 10.5 Gradual salt addition . . . . . . . . . . . . . . . . . . . . . . . . 215 10.5.1 Experiment 1 - addition of NaCl . . . . . . . . . . . . . 215 10.5.2 Experiment 2 - addition of salt solution with M/P= 0.2 217 10.5.3 Experiment 3 - addition of salt solution with M/P= 0.2 versus reference (inc. supernatant) . . . . . . . . . . . . 218 10.5.4 Experiment 4 - addition of salt solution with M/P= 0.2 versus reference (excl. supernatant) . . . . . . . . . . . 219 10.6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222 10.6.1 No single theory . . . . . . . . . . . . . . . . . . . . . . 222 10.6.2 Low / intermediate ionic strength . . . . . . . . . . . . 222 10.6.3 High ionic strength . . . . . . . . . . . . . . . . . . . . . 223 10.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 11 General conclusion and outlook 226 11.1 General conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 226 11.1.1 Framework for bioflocculation and filterability monitoring 226 11.1.2 Impact of bioflocculation on activated sludge separation 228 11.1.3 The influence of salts on MBR performance and biofloc- culation . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 11.2 Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 11.2.1 Understanding fouling . . . . . . . . . . . . . . . . . . . 232 11.2.2 Efficient fouling control and mitigation . . . . . . . . . . 233 11.2.3 A paradigm shift towards biorefining . . . . . . . . . . . 235 Bibliography 237 nrpages: 255 status: published

Dynamic process models can be used for operating, controlling and optimising important bioprocesses, e.g., pharmaceuticals production, enzyme production and brewing. After selecting an appropriate process model structure, parameter estimates have to be obtained based on real-life experiments. To reduce the amount of labour and often cost intensive experiments optimal experiment design (OED) is an indispensable tool. In optimal experiment design, dynamic input profiles have to be determined in order to obtain informative experiments. In the particular case of optimal experiment design for parameter estimation, a scalar measure of the Fisher Information Matrix is used as an objective function. Over the years, different criteria have been developed. However, the important question that remains is which criterion to choose. In this work, an approach to tackle the criterion selection is presented. In addition, a multi-objective optimisation approach is implemented, which enables to combine two, often competing optimisation criteria. The developed approach is illustrated with two case studies. The first case study is a fed-batch bioreactor model and the second case study is a Lotka Volterra fishing model. © 2012 Elsevier Ltd. ispartof: Chemical Engineering Science vol:78 pages:82-97 ispartof: location:Lommel, Belgium status: published

This research is aimed at solving two major operational issues the wastewater treatment plant of Monsanto in Antwerp was coping with at the end of 2004. At first, inexplicable periodic deteriorations of the activated sludge settling resulted frequently in sludge losses from the final clarifier. And secondly, the Centridry® installation, used to reduce the waste sludge volume by the combination of mechanical dewatering and thermal drying, was performing significantly below its designed capacity. The latter was more specifically caused by two bottlenecks, that is (i) the high vibrations of the decanter centrifuge and (ii) the unexplainable phenomenon of dryer fouling by build-up of drying solids on the dryer wall, inducing substantial operational downtime. Moreover, the investigation of this peculiar fouling phenomenon is seriously impeded by the design of the Centridry® - the design which makes it, at the same time, also a unique technology. By combining both unit operations, dewatering and drying of the sludge, in one enclosed machine, the Centridry® inherently rules out sampling and, hence, analysis of the dewatered cake leaving the centrifuge and thus entering the drying stage. The lack of information on this intermediate product forms a serious drawback from a research and optimization point of view. A 3-year study of the varying Ca++ and Na+ concentrations in the wastewater has been conducted, in relation to the changing sludge settling characteristics. Two mechanisms are demonstrated to exist by which Ca++ ions influence the bioflocculation and so the sludge settleability: (i) in accordance with the Divalent Cation Bridging Theory (DCBT), the divalent Ca++ ions exchange the monovalent Na+ ions from the sludge flocs, bridging the individual flocs into larger and stronger flocs that settle better; (ii) the low soluble CaCO3, formed in the biodegradation basin, precipitates from the water, and is enmeshed within the floc structure, making the flocs heavier and so settling better. Hence, in the case of a Ca++ depletion from the wastewater (when the only Ca++ supplying plant on site goes in shutdown), the sludge settleability starts to worsen over time as a result of (i) the floc disintegration, and (ii) the continuous wasting of the (old) heavy (CaCO3) flocs, being replaced by new grown sludge flocs without the heavy solids incorporated in the floc structure. The first objective of this research, avoiding the sludge settling to deteriorate to such an extent that solids losses manifest in the clarifier overflow, is being accomplished by adding CaCl2 to the wastewater in order to avoid the complete floc disintegration to happen when the influent is for a longer period deficient in Ca++. Conclusively, Ca++ ions are of upmost importance for the bioflocculation, and, hence, for the sludge settleability.In the second part of this research, two lab tests are introduced to (i) simulate the sludge dewatering in the decanter centrifuge (that is, to simulate the important intermediate product of the Centridry®) and (ii) to map the sticky behavior of drying sludge (that is, to make the fouling issue in the Centridry® tangible). A 4-year follow-up of the centrifugal dewaterability of the Centridry® sludge feed has been conducted at lab scale. It is emphatically demonstrated that the dryness of the dewatered sludge cake is mainly governed by the sludge CaCO3 fraction, with a high sludge inorganic fraction yielding a higher cake dryness. Up to a sludge inorganic fraction of 55%, the cake can be made wet again by reducing the amount of inorganic clay conditioner to the Centridry® feed. Furthermore, the lab protocol to map the stickiness of the sludge provides information of equal importance to better understand the fouling issue in the dryer. The drying of sludge is, worldwide, cumbersome because of its sticky phase. When partially dried it behaves as a sticky, gluey substance that likes to cling on a dryer wall. An explanation for this physical consistency is proposed to be found in the 3-dimensional structure of extracellular polymeric substances (EPS) surrounding the micro-organisms, which achieves its strongest (plastic, pasty) configuration when the sludge is partially dried. The sticky phase, i.e., the range of dryness wherein it behaves most sticky, depends on the sludge inorganic fraction as well, with a high sludge inorganic fraction (and thus a low EPS content) resulting in a higher sludge dryness needed to come in the sticky phase. Further, it is shown that the intrinsic stickiness of sludge can be reduced by adding pure polyaluminiumchloride (PACl) to waste sludge. It is postulated that this salient feature of PACl conditioning is due to the large Al13-clusters and their associated hydrated water. When present on the exterior of the drying flocs, these (artificially added) clusters of bound water (water shields) act as some kind of lubricant for the intrinsically sticky solids. This stickiness mitigating effect of pure PACl addition to waste sludge is a totally new application of PACl. In addition to both lab tests, a statistical model is obtained for the fouling propensity, based on a 5-year data base with operational experience. This empirical model clearly demonstrates that the sludge characteristics, resulting from the cations in the incoming wastewater, govern the fouling probability, with sludge characterized by a high inorganic fraction (excellent dewaterability) corresponding with a higher chance of dryer fouling. The second objective of this research, increasing the Centridry® capacity, is then accomplished as follows. The vibrational bottleneck is eliminated by substantially reducing the centrifuge speed of rotation, from 3165 rpm to 2500 rpm. Next, the fouling of the sludge dryer is managed by a two-fold control strategy: (i) for sludge with an inorganic fraction below a threshold of approximately 55%, a wet cake after the centrifuge is aimed for, by lowering the inorganic clay addition to the sludge feed (making wet cake ); (ii) for sludge that dewaters to such a high extent that making wet cake is not an option anymore, pure PACl is added to the sludge feed. The latter strategy is dealing with reducing the sludge s intrinsic stickiness, whereas the former is post-poning the sludge s sticky phase to occur only after the drying solids have passed the most sensitive places at the entrance of the dryer before behaving most sticky. By doing so, the Centridry® capacity could be increased with 107%.In conclusion, this PhD study made the optimization of the industrial-scale Centridry® installation possible by unraveling the relation between the micro-scale activated sludge floc structure and the changing cation concentrations in the wastewater. nrpages: 280 status: published

[**] ispartof: edition:1st edition ispartof: Belgian-French-German Conference on Optimization edition:14 location:Leuven date:14-18 September 2009 edition: 1st edition nrpages: 267 status: published

The optimal operation of a solid oxide fuel cell stack is addressed in this paper. Real-time optimization, performed at a slow time scale via constraint adaptation, is used to account for uncertainty and degradation effects, while model-predictive control is performed at a faster time scale to reject process disturbances and to safely adapt the system to the specified output constraints following changes in cell power demand. To ensure that these constraints are strictly honored, a novel adaptation algorithm that uses the built-in constraint handling of quadratic programming is implemented within the model-predictive controller. An additional feature of this algorithm - its ability to adapt the feasibility region in view of uncertainty - is shown as well. Simulation results illustrate the efficacy of this approach in the solid oxide fuel cell system.