Search

The assessment of a pilot-scale hybrid system coupling powdered activated carbon (PAC) adsorption with membrane ultrafiltration (UF), in respect of activated carbon regeneration and organic micropollutant removal, was investigated in this study. Field tests with two adsorbents (i.e. a commercial PAC and a PAC-Fe(II) composite), conducted in the premises of Thessaloniki Water Treatment Plant, demonstrated the high efficiency of the combined PAC/UF process. Regeneration efficiencies varying between approximately 95% and 110%, complete diclofenac (DCF) degradation and rather moderate mineralization (TOC removal) rates of up to 47%, can be achieved by UVC/H2O2 or photo-Fenton oxidation after 4 hours of treatment; this performance is attributed to the in situ generation of reactive oxidant species by photolysis of H2O2, which seems to enhance the process effectiveness. Among the two adsorbent materials tested, composite PAC-Fe(II) exhibited a higher DCF adsorption capacity than the original PAC, probably due to the improved chemisorption and/or the electrostatic attractive interactions between the negatively charged DCF molecules and the positively charged iron species, at neutral pH. Furthermore, a rather insignificant effect of PAC-Fe(II) loading on the regeneration efficiency was observed. The advantages of totally controlled H2O2 dosages and short operating times render the hybrid PAC/UF system a promising alternative to conventional and advanced drinking water purification methods.

Diminishing clean water resources and their pollution (due to human activities and climatic change) are of great concern on a global basis. Under such conditions, the adequacy of drinking water (DW) standards and their meticulous implementation are issues of utmost importance in safeguarding human health. Unfortunately, the significant number of disease outbreaks (and of other suspected/potential health effects) related to DW, even in developed countries, attests to the fact that these issues require vigilance and continuous re-appraisal, particularly considering the assorted emerging contaminants and the ever-improving technological tools to cope with them. Therefore, the present comprehensive assessment addresses the main issues and concerns regarding DW standards and implementation thereof. Emphasis is placed on identifying the inherent deficiencies of standards (due to neglect of potential toxic contaminants and to debatable specifications/limit values) and regulations for their implementation and the monitoring of DW quality (due to weaknesses of available analytical/measurement techniques and inadequacies of tools and protocols). Rather serious deficiencies on these aspects are evident regarding the main categories of contaminants, i.e., synthetic chemicals and biological species. In regard to addressing these issues, progress made in recent years at the scientific/technical level and the main challenges are outlined toward the goal of an improvement in standards and their implementation.

The oxidation of pharmaceutical diclofenac (DCF) was investigated in a catalytic membrane reactor (CMR) comprising a tubular porous alumina (α-Al2O3) membrane, as support material, with embedded iron oxide nanoparticles (nFe) in situ generated in its pores. The performance of the heterogeneous Fenton-type oxidation reaction was examined in a laboratory-scale membrane pilot, with the addition of hydrogen peroxide. Aiming to optimize this CMR, Central Composite Design with Response Surface Methodology was employed to assess the effects of three key process parameters and their interaction on DCF degradation. Encouraging results were obtained, showing a significant DCF oxidation/mineralization by the α-Al2O3/nFe CMR. Feed solution pH was found to be the most significant variable, with acidic values leading to improved performance. Under near optimum operating conditions (pH 3 and 42.9 mg/L H2O2), the removal and mineralization of DCF was ∼65% and ∼48%, respectively. Negligible iron leaching was observed in the tests, which confirms the effective embedding and stability of catalytic nanoparticles in the ceramic porous matrix, thus allowing multiple membrane utilization. Future research directions are proposed towards method optimization, with emphasis given to membrane modification for improving the adsorption of organic pollutants and their oxidation within the catalytic membrane matrix.

Net-type spacers are essential components of the spiral-wound membrane (SWM) modules employed in reverse osmosis (RO) desalination plants. Despite progress made in designing SWM modules in recent years, key issues related to feed-spacer characteristics (particularly spacer material properties in conjunction with their geometry), which affect module performance and membrane integrity, have been neglected by the research community. This paper focuses on the effect of compressive stresses, imposed on membrane spacers and envelopes during SWM-module fabrication to achieve module compactness and resistance to deformation during operation. A comprehensive method is presented (involving experimental and computational tools) which allows to quantify compressive stress effects, on the characteristics of spacer-filled membrane channels, and further to assess their impact on the key SWM-module operating parameters and on overall RO process performance. The method is demonstrated by assessing the performance of 26-mil commercial feed-spacer in an appropriate range of compressive stresses; it is interesting that modest nominal applied pressure, at the level of 1–2 bar, tends to cause significant feed-spacer compression, channel-gap reduction and membrane indentations impairing the active layer. The novel method is considered most helpful for selecting appropriate feed-spacer material and geometric characteristics, for a membrane of particular mechanical properties.

The topic of chemical-free spiral wound membrane (SWM) cleaning, dealt with in this paper, is certainly of great interest to both industry and academia. However, in this Letter, the reasons are outlined for the serious limitations, imposed on implementation of the proposed deformation-induced cleaning (DIC) method to SWM modules, which have been apparently disregarded in this paper. These limitations are due to the dominant structural characteristics of SWM module-design; in particular, to the imposed (during fabrication) interlayer pressure on the membrane leaves and to the overall rigidity of the currently employed modules. These issues/imitations, and the implications of neglecting them are outlined herein.

The reported pilot-scale development of hybrid process alternatives, combines in-situ regenerated powdered activated carbon (PAC), H 2 O 2 -based advanced oxidation techniques and membrane ultrafiltration (UF). Commercial PAC and special iron-impregnated PAC (PAC-Fe) were used as adsorbents, whereas diclofenac (DCF), a commonly detected pharmaceutical in water-sources, was employed as representative micro-pollutant. For PAC regeneration, modest on-line H 2 O 2 dosage was adopted, alone or in combination with UV-C irradiation. Large DCF uptake was measured for both PAC (154.5 mg/g) and PAC-Fe (116.7 mg/g), while the regeneration efficiencies achieved with the photo-assisted H 2 O 2 processes were much greater (122.2% and 131.6% for H 2 O 2 +UVC and photo-Fenton, respectively) than those obtained with H 2 O 2 alone (70% and 77.6% for H 2 O 2 and Fenton, respectively). For PAC-Fe, enhanced recovery of adsorption capacity suggests occurrence of heterogeneous Fenton-like reactions on iron-oxide particle surfaces leading to generation of OH and other reactive species by H 2 O 2 -decomposition. The catalytic adsorbent (PAC-Fe) retained its stability and activity after three operating cycles, which would significantly reduce operating cost in practical applications. Considering the rather short regeneration-time of present tests (180 min) and the nearly basic pH of feed solutions, the novel processes investigated herein, involving PAC regeneration in the same equipment employed for water purification, hold distinct advantages over conventional methods.

The synthesis, characterization, and performance of composite photocatalytic adsorbents are investigated in this work using the dip-coating and the electrophoretic coating methods for the deposition of titanium dioxide (TiO2) on porous activated carbon fiber (ACF) substrates. The adsorption and photocatalytic efficiency of the synthesized catalytic adsorbents were compared using phenol as the model pollutant. Both immobilization techniques resulted in composite ACF/TiO2 adsorbents characterized by large surface area (844.67 &plusmn, 45.58 m2 g&minus, 1), uniform distribution of TiO2 nanoparticles on the activated carbon fibers, and high phenol adsorption. The method and the treatment time affected the phenol adsorption, while the highest sorption was determined in the case of the ACF/TiO2 sample prepared by the electrophoretic coating method (at 20 V) for an electrolysis time of 120 s (7.93 mgphenol g&minus, 1ACF/TiO2). The UV-A irradiation of most ACF/TiO2 samples led to a faster removal of phenol from water as a result of the combined sorption and heterogeneous photocatalysis. The stability and the effective regeneration of the most promising composite photocatalytic adsorbent was proved by multiple filtration and UV-A irradiation cycles.

Enhanced mineralization of phenol in brines with high chloride content was investigated by employing an electrochemical advanced oxidation treatment that couples anodic oxidation, electrochlorination and electro-Fenton in a single process. Experimental work was carried out in a pilot scale unit with an undivided plate-and-frame cell equipped with a boron-doped diamond anode and a carbon-PTFE gas diffusion electrode as cathode, in batch recirculation mode. The effects of operating conditions on phenol degradation, including current density, air flow rate, water feed flow rate, Fe2+ dosage and pH as well as of the water matrix, were evaluated. Applied current exhibited the greatest effect on phenol degradation/mineralization efficiency. Complete degradation of phenol (of initial concentration 50 mg L−1) was achieved under the near-optimum operating conditions (40 mA cm−2, pH 7, 0.4 m3 h−1 water circulation rate) within 30 min. Both air flow rate and Fe2+ dosage did not show a measurable impact on phenol removal. However, increasing the chloride content of water significantly improved the efficiency of treatment due to the enhanced indirect oxidation by the electrogenerated chlorine. Several trihalomethane intermediates (chloroform, bromodichloromethane) and chlorinated/brominated phenol byproducts forming during treatment, were eliminated after 60 min of processing time.

Proper fluid-mechanical characterization of haemo-catharsis (HC) modules is indispensable for optimizing their design and operation. A recently developed methodology allows determination of (heretofore ignored) significant pressure losses in the headers of HC-modules, taking into account the main fluid-mechanical phenomena through appropriate/representative parameters. The latter comprise friction coefficients ff and fs (for fiber-lumen and shell-side flow), membrane permeance K and four parameters ζi1-ζi4 representative of pressure drop in the module-headers. A mechanistic model and pressure/flow-rate measurements, taken with two specially-selected operating modes, are used for such parameter estimation. Implementation of this methodology is reported herein, to the practically important case of counter-current flow, that involves theoretical-model extension/adaptation and experiments with liquids of viscosity 1 cp and 2 cp. These data are employed to investigate and optimize the data-fitting procedure (considering computational accuracy/stability and convenience) and to validate the counter-current flow model. Assessment of experimental and theoretical-model results suggests that the procedure involving data from the two special operating-modes is essential for facile implementation of the novel methodology, including the accurate determination of membrane permeance K. The methodology developed, allows reliable prediction of the key spatially-evolving process-variables, and is recommended for practical applications and advanced overall HC-module performance modeling. Suggested future work is outlined.

A novel, fully automated photocatalytic membrane reactor (PMR) pilot unit is presented and evaluated (at different operating process parameters and real treatment conditions) for degradation of the pharmaceutical diclofenac (DCF), a typical recalcitrant micropollutant. The PMR-pilot has a maximum system capacity of 1.2 m 3 /d of treated water, 52 W of UV-C power, and combines two modern processes; i.e. heterogeneous photocatalysis and membrane separation, employing dispersed TiO 2 nanoparticles with UV-C irradiation, and submerged ultrafiltration hollow fibers, respectively. Pilot tests with two types of natural (tap and surface) water under stable, continuous operation, permitted the assessment of the PMR unit performance (for DCF degradation, TOC mineralization) in terms of hydraulic residence time (HRT), TiO 2 loading, UV-C irradiation dose, and DCF feed concentration. Under the specific conditions tested, the catalyst loading and UV-C radiant power significantly affected DCF removal; however, HRT had a rather insignificant effect. Additionally, several transformation products (TPs) were identified by means of liquid chromatography–mass spectrometry using electrospray interfacing technique, thus, shedding light into possible reaction pathways leading to DCF mineralization.

Reliable simulation of membrane-module dynamic operation is essential in optimizing its detailed geometric characteristics and operation, as well as that of membrane plants, for various types of fluid treatment applications. This paper is part of systematic efforts toward development of such a comprehensive model, considering temporal variability caused by organic membrane fouling. To render the mathematical problem tractable, justified simplifications (retaining the physical parameter interdependencies) lead to a system of basic equations in two spatial planar coordinates, enabling to obtain a realistic temporal evolution of all process parameters. The flexible model structure allows integration of sub-models, for phenomena occurring (and researched) at small spatial scales, which account for retentate spacer effects on friction losses and mass transfer, and constitutive expressions for fouling rates during desalination. The robust numerical algorithm developed to solve the system of differential and algebraic equations exhibits satisfactory convergence, appropriate for applications. The results presented herein demonstrate the versatility of the numerical code and its potential to analyze the interaction of mechanisms involved in fouling evolution, which is impossible by the much simpler one dimensional models. Directions for future developments are indicated.

A common type of ultrafiltration (UF) membrane fouling is due to organic gel, encountered in the treatment of various effluents and feed waters to desalination plants; polysaccharides are a major component of such fouling layers. In standard methods, dead-end UF tests under constant pressure are used to determine membrane-fouling propensity, but the results are applied with questionable success to the case of constant flux filtration, favored in practice. Observed differences between fouling resistances from the two modes of UF filtration have not been adequately explained yet. The temporal evolution of specific resistance α (considered as the most representative fouling parameter) is investigated herein, aiming to elucidate differences and possible similarities in the different filtration modes. During alginate-solution ultrafiltration, typical gel-type membrane fouling layers are formed with time-varying deposited mass density m (g/cm 2 ) depending on permeate flux; thus, the temporal variation of specific resistance α , of pressure difference across the fouling layer/cake ΔP c , and of pressure drop per unit depth of cake (the latter proportional to ΔP c /m ) are studied for both constant pressure P and constant flux J operation, within P and J ranges of practical interest. These detailed data suggest that permeation flux is the process variable controlling the development of resistance α . Specifically, for constant-flux filtration, the resistance α data are well correlated with flux J (as also shown recently), whereas in the constant-pressure mode the commonly observed constant resistance α , over a rather broad range of declining flux, is due to the high flux level prevailing in the early stage of filtration. A successful approach is demonstrated to compare and correlate the specific resistance α data obtained from the two different filtration modes. The usefulness of the new results in practice is also discussed.

Membrane desalination is associated with significantly increased ion concentration of rejected sparingly soluble salts (commonly of calcium, magnesium, barium, etc) right at the membrane surface, leading to undesirable membrane scaling. Optimization of desalination membrane-plant design and its operation strategy requires quantification of scaling evolution phenomena, through appropriate mathematical models involving experimentally determined system parameters. In recent years, a modeling approach based on heterogeneous nucleation-growth theory was developed for this purpose, and proven to be successful in the case of non-stirred dead-end desalination experiments but not in the case of cross-flow and stirred dead-end desalination. To resolve this inconsistency, new results from numerous sets of improved dead-end non-stirred experiments (some with controlled changes of imposed pressure difference during testing) are reported that provide useful quantitative information and insights regarding incipient scaling. The previous modeling approach is generalized considering non-classical nucleation. Moreover, the likely scatter of local membrane-surface properties is considered in a parameter estimation approach. The new modeling approach leads to estimation of a range of scaling-model parameters consistent with the experimental data. Such parameters can be employed in comprehensive modeling tools of membrane-modules, to simulate their operation for the particular type of membrane for which parameters are extracted.

Haemo-catharsis (HC) modules comprise the cylindrical fiber-bundle section and two flow inlet/outlet headers. Available models ignore the substantial pressure-drop (varying non-linearly with flow-rate) in the headers, while employing linear pressure-drop expressions for the entire module, which lead to erroneous assessment of HC-module performance/characteristics. The novel method reported herein to address this issue, enables reliable determination of key HC-module fluid-mechanical parameters. In specially designed facility, accurate pressure measurements are made (for typical flow-rate ranges) under carefully selected inlet/outlet flow-modes. Fitting a mechanistic model to appropriate pressure-difference versus flow-rate data, allows determination of four parameters ζi associated with pressure-drop in headers (for inlet-/outlet-fluid flow), two friction-coefficients for laminar flow in fiber-lumen and shell-side, plus the hydraulic membrane-permeability. Applying this method to a typical commercial high-flux HC-module, realistic values are obtained of membrane-permeability and friction-coefficients, whereas the magnitude of coefficients ζi manifests the significance of pressure-drop in the headers. Friction-coefficient predictions by a theoretical model applicable to the main HC-section, compared to experimentally-determined values, show weaknesses of the former whereas the latter can reliably account for particular geometric/morphological features of real HC-modules. Further testing of the method is recommended, while its adaptation to a comprehensive model for HC-module performance simulation is in progress.

Rigorous peer-review is the corner-stone of high-quality academic publishing [...]

This manuscript summarizes the successful start-up and operation of a hybrid eco-engineered water treatment system, at pilot scale. The pilot unit, with 100 L capacity, has been devised for the efficient electrocatalytic production of H2O2 at an air-diffusion cathode, triggering the formation of OH from Fenton’s reaction with added Fe2+ catalyst. These radicals, in combination with those formed at a powerful boron-doped diamond (BDD) anode in an undivided cell, are used to degrade a mixture of model pesticides. The capability of the plant to produce H2O2 on site was initially optimized using an experimental design based on central composite design (CCD) coupled with response surface methodology (RSM). This aimed to evaluate the effect of key process parameters like current density (j) and solution pH. The influence of electrolyte concentration as well as liquid and air flow rates on H2O2 electrogeneration and current efficiency at optimized j and pH was also assessed. The best operation conditions resulted in H2O2 mass production rate of 64.9 mg min−1, 89.3% of current efficiency and 0.4 kWh m-3 of energy consumption at short electrolysis time. Performance tests at optimum conditions were carried out with 75 L of a mixture of pesticides (pyrimethanil and methomyl) as a first step towards the elimination of organic contaminants by solar photoelectro-Fenton (SPEF) process. The combined action of homogeneous ( OH) and heterogeneous (BDD( OH)) catalysis along with photocatalysis (UV photons collected at a solar CPC photoreactor) allowed the removal of more than 50% of both pesticides in 5 min, confirming the fast regeneration of Fe2+ catalyst through cathodic reduction and photo-Fenton reaction.

Vacuum membrane distillation (VMD) is an attractive variant of the novel membrane distillation process, which is promising for various separations, including water desalination and bioethanol recovery through fermentation of agro-industrial by-products. This publication is part of an effort to develop a capillary membrane module for various applications, as well as a model that would facilitate VMD process design. Experiments were conducted in a laboratory pilot VMD unit, comprising polypropylene capillary-membrane modules. Performance data, collected at modest temperatures (37 °C to 65 °C) with deionized and brackish water, confirmed the improved system productivity with increasing feed-water temperature, excellent salt rejection was obtained. The recovery of ethanol from ethanol-water mixtures and from fermented winery by-products was also studied, in continuous, semi-continuous, and batch operating modes. At low-feed-solution temperature (27–47 °C), ethanol-solution was concentrated 4 to 6.5 times in continuous operation and 2 to 3 times in the semi-continuous mode. Taking advantage of the small property variation in the module axial-flow direction, a simple VMD process model was developed, satisfactorily describing the experimental data. This VMD model appears to be promising for practical applications, and warrants further R&amp, D work.

The investigated novel retentate-spacer design comprises spherical nodes symmetrically connected with cylindrical filaments, thus forming a net-type structure with parallelogram unit-cells. Key feature of this configuration is the contact points (or small contact-regions) of nodes with bounding membranes, instead of filament-membrane contact lines characterizing the conventional unwoven-filament geometries; undesirable contact lines create “dead-flow” zones, that promote reduced mass-transfer (i.e. increased concentration polarization-CP) and fouling. Direct numerical simulations of the flow field and mass transfer, complemented by pressure drop and mass-transfer coefficient measurements, reveal the advantages of novel spacers. The flow field, and local shear-stress and mass-transfer-coefficient distributions on membranes, are rather uniform. The novel spacers, considering also the extensive flow constrictions between filaments and membranes, exhibit broader distributions of these time-averaged transport parameters, which extend to higher values (compared to conventional spacers), with insignificant percentage of zero values; a fairly uniform shear-stress distribution is also identified on spacer filaments. Thus, significant benefits are expected in mitigating CP and fouling phenomena in membrane operation. Friction losses show increased dependence on cross-flow velocity (compared to conventional designs), as they are affected more by flow-shearing at the membrane surfaces, than by form-drag. Useful pressure drop and mass-transfer correlations are obtained.

The rheological properties of membrane biofouling layers developing in Membrane Bioreactors (MBR) have not been measured directly so far, although they are essential for understanding biofouling and for determining conditions to mitigate it. The development of a novel method is reported herein, which permits accurate rheological measurements on membrane biofouling layers, using specimens withdrawn from an operating laboratory-scale MBR. A specially designed test-section (with a pair of flat membrane specimens) allows implementation of realistic operating modes, including automatic periodic back-washing. The membrane specimens, with a biofouling layer representative of a well-controlled MBR operating history, are examined in an advanced parallel-plate rheometer, where special care is taken to prevent ex-situ biofouling layer changes. Measurements performed in oscillatory and steady shearing modes provide valuable information on biofouling layer rheological characteristics, confirming that these layers are typical viscoelastic materials with yield stress exhibiting thixotropic, shear-thinning behavior. Rheological properties of the activated sludge liquor are also assessed. The rheological properties of activated sludge and of biofouling layers display qualitative similarities, although the elastic character of the latter is much more pronounced. The new technique is most useful for studying biofouling layers as a function of the main MBR process parameters. Improvements of the technique for such applications are discussed.

This study is a step toward integrated modeling of the evolving (in space and time) flow and concentration fields during water desalination in membrane spacer-filled channels. Such detailed model predictions are of practical significance, leading to improved understanding of spiral wound membrane (SWM) module operation. Submerged and zigzag spacer filament 2-dimentional configurations are employed, as they create geometric flow-channel features (contact lines, flow constrictions, etc) encountered in SWM modules. The numerical study, performed in a Reynolds number range typical for desalination modules, is focused on the evolution of local mass transfer coefficient k and concentration polarization, which significantly affect SWM module performance. A detailed quantitative prediction is obtained of these parameters, which are linked to the variation of wall shear stress, static pressure and permeation flux. By focusing on a typical region of the membrane channel (adequately capturing the evolving flow characteristics), it is predicted that increasing average k values (with the concomitant reduction of concentration polarization) are associated with the undesirable increase of pressure drop; these conflicting requirements can be balanced through the selection of appropriate spacer geometry. The negative effect of membrane-filament contact lines on concentration polarization is well documented. An assessment is made of the two spacer filament configurations, from this perspective, and directions for future research are outlined.

Development of polysaccharide fouling layers during dead-end membrane ultrafiltration is an inherent problem in water treatment and similar separation processes. Despite its practical significance, ultrafiltration membrane fouling under constant permeate flux J has received inadequate attention. Therefore, constant-flux experiments, with typical alginate model-solutions covering the flux range of practical interest (i.e. 10–100 L/m 2 h), are employed to obtain new insights into fouling layer characteristics. The specific resistance α , as the most representative fouling-layer property, is used for data interpretation. The behavior of resistance α , with increasing permeate volume, confirms that these gel-type layers are strongly affected by flux J and generally compressible. Layer compressibility is evident beyond an initial phase of membrane coverage by alginates. The initial resistance α i, corresponding to thin developing layers and relatively small pressure-drop across the layer/“cake” Δ P c , is independent of Δ P c , approximately linearly increasing with J . For all data-sets, variation of resistance α with Δ P c is well correlated by a generalized expression involving (in addition to α i ) two parameters, n and P a , considered to represent layer/cake compressibility index and reference pressure, respectively; these parameters also increase (and are fairly well correlated) with J . The usefulness of these data is demonstrated in elucidating issues such as "critical flux" for this filtration mode.

The development of reliable criteria for determining the onset of scaling on desalination membranes is of high practical value. A novel approach to systematically investigate incipient membrane crystallization and scale-particle evolution was recently reported by this laboratory, involving modeling-relevant phenomena during membrane desalination and experiments, in well-controlled pressure cells operating in dead-end mode with no agitation. However, parallel studies have shown that flow conditions equivalent to those prevailing in real desalination membrane modules are obtained in pressure cells with agitation. Therefore, this work aims to employ the aforementioned approach to the more realistic case of membrane CaCO3 scaling in dead-end desalination with agitation. Detailed data regarding incipient CaCO3 scaling of desalination membranes were obtained, by analyzing scanning electron microscopy (SEM) images, for small bulk supersaturation ratios S = 1 to 4 and short filtration times (~30–90 min). ...

A laboratory pilot photocatalytic membrane reactor (PMR), employing a hybrid TiO 2 /UV-A catalysis–ultrafiltration process, is evaluated for degradation of diclofenac (DCF), a typical micro-pollutant frequently encountered in source waters. The combination of membrane ultrafiltration with photocatalysis allows TiO 2 separation and reuse whereas the automatic periodic membrane backwashing, combined with moderate permeate flux, effectively control membrane fouling, thus permitting stable continuous operation with no wastewater stream. The operation under steady state conditions permitted the assessment of the PMR system performance (for DCF degradation), in terms of water pH and catalyst dosage, in relatively long-term tests. The results demonstrate excellent system performance. Experiments with groundwater, at UV-A radiant power per unit volume 6.57 W/L, show an optimum DCF removal at pH ∼ 6 and TiO 2 loading near 0.5 g/L with maxima of DCF molecular degradation and mineralization ∼99.5% and ∼69%, respectively. Significant differences of photo-catalytic oxidation performance are identified with different water matrices (i.e. ultrapure water, tap water, groundwater), confirming the important role of feed-water characteristics (i.e. presence of organic and inorganic scavengers) on process effectiveness. The new results highlight, among other factors, the need for good knowledge of feed water properties, to successfully design a PMR treatment process for effectively removing organic micro-pollutant.

Creation of a reliable and comprehensive emission inventory of the pesticides used in Europe is a key step towards quantitatively assessing the link between actual pesticide exposure and adverse health effects. An inventory of pesticide emissions was generated at a 1 × 1 km grid, for the year 2000. The emission model comprises three components: estimates of active substance (AS) wind drift taking into account crop type, volatilization during pesticide application and volatilization from the crop canopy. Results show that atmospheric emission of pesticides varies significantly across Europe. Different pesticide families are emitted from different parts of Europe as a function of the main crop(s) cultivated, agro-climatic conditions and production intensity. The pesticide emission inventory methodology developed herein is a valuable tool for assessing air quality in rural and peri-urban Europe, furnishing the necessary input for atmospheric modelling at different scales. Its estimates have been tested using global sensitivity and Monte Carlo analysis for uncertainty assessment and they have been validated against national and local surveys in four European countries; the results demonstrate the robustness and reliability of the inventory. The latter may therefore be readily used for exposure and health risk assessment studies targeting farmers, applicators, but also bystanders and the general population in Europe.

Data are reported on photocatalytic mineralization rates of sodium alginate (SA), which is representative of polysaccharides commonly encountered in surface waters and effluents requiring purification. An agitated vessel-type photocatalytic reactor was employed, with suspended TiO 2 nanoparticles, to study the effect of UV-A radiant power per unit volume P R (W per L reactor fluid) for which there are insufficient data. Using initial SA concentration ∼30 mg/L, mineralization rates were determined, by monitoring Total Organic Carbon (TOC) reduction, in the range of UV-A P R = 2.15 to 8.57 W L −1 and TiO 2 concentrations 0.5 to 1.5 g/L. For this well-mixed heterogeneous photocatalytic environment, the measured fairly high polysaccharide mineralization rates were fitted to a first order kinetic expression. In the parameter range studied, the kinetic constants k (of order 10 −2 min −1 ) were linearly dependent on P R with a slope in the range ∼0.6 × 10 −2 to ∼1.0 × 10 −2 L/(W min). An optimum catalyst concentration near 1 g/L TiO 2 was observed. The experimental techniques developed and the type of kinetic data obtained are considered useful for designing photocatalytic reactors for practical applications. Moreover, it is encouraging that the measured mineralization rates are in accord with those from a continuously operated hybrid photocatalytic membrane reactor of different geometry.

Spiral-wound membrane (SWM) modules are the most important components of reverse osmosis and nanofiltration desalination and water treatment plants; their optimum design and operation is crucial for achieving satisfactory membrane plant efficiency associated with reduced product cost and environmental impact. This paper is focused on: (a) the main SWM module design parameters, in need of optimization, which include the geometrical characteristics of retentate-side spacers and the membrane sheet dimensions (for a module of fixed total active surface area); (b) the major operating variables, i.e. the permeate flux and the cross-flow velocity at the retentate side. Membrane element problems, usually encountered in operating plants and affected by the above SWM element parameters, include membrane fouling by various species (commonly organic matter, inorganic colloids, and bio-foulants), membrane scaling due to sparingly soluble salts, increased friction losses and uneven flow distribution in the SWM ...

Rheological property data (practically and theoretically significant) of typical desalination-membrane fouling layers are reported for the first time in this study, involving two stages. First, desalination of brackish water (with 30 mg/L sodium alginate) was carried out under constant flux; Ca 2+ -ion concentration 1 mM–8 mM was used. Linear temporal trans-membrane pressure increase, associated with constant specific fouling resistance α , indicated uniformity of growing alginate layers. Fouling resistance α increased with calcium concentration. Subsequently, fouled membrane segments were tested in a parallel-plate rheometer, operating in oscillatory and steady shearing modes. Key rheological parameters of fouling layers, including storage ( G ′) and loss ( G ″) moduli, complex and shear viscosities, were assessed in relation to their permeability and other characteristics. Alginate fouling layers possess viscoelastic properties characteristic of rather stiff gels; moreover, they belong in the class of thixotropic materials with yield stress. Layers from feed-water of small Ca +2 -ion concentration exhibit strong elastic character, high viscosities and yield stresses; the elastic character and rheological parameter values are reduced with increasing Ca 2+ -ion concentration. The high Na-ion concentration in the fouling layers appears to significantly influence their structure; however, considering the complicated dynamics of developing layers, the relative significance of the abundant Na-ions compared to Ca 2+ -ions is unclear.

Significant uncertainty exists regarding incipient membrane scaling (a criterion for determining water recovery and antiscalant use) with negative impact on optimization of RO plant design and operation. Experimental data are reported on the initial phase of RO-membrane scaling by weakly supersaturated brackish water (calcite supersaturation ratio S∼1.9 to 3.1) in a narrow spacer-filled channel, simulating local (constant flux) conditions in spiral-wound membrane (SWM) modules. Scanning electron microscopy image analyses of numerous membrane samples, combined with complete physico-chemical characterization of feedwater and retentate, provide new insights into early-phase membrane scaling. The unavoidable presence of small particles in feedwater (despite the use of tight cartridge filter) apparently promotes heterogeneous nucleation. Additionally, there is a clear evidence that, under these conditions, induction period for membrane scaling is practically non-existent. The initial deposit flux [mgCaCO 3 /(m 2 min)] on the membrane, obtained through measured crystal size distributions, exhibits strong dependence on supersaturation ratio; moreover, measured deposit fluxes are substantial and (if not mitigated) would be eventually detrimental to SWM performance. These data were obtained, while none of the frequently used parameters (bulk fluid turbidity, pH, Ca concentration, permeate flux) indicated particle/scale deposition on membranes. Implications of the new results, on practical and theoretical aspects of scaling, are discussed.

The multi-criteria analysis gives the opportunity to researchers, designers and decision-makers to examine decision options in a multi-dimensional fashion. On this basis, four tertiary wastewater treatment (WWT) technologies were assessed regarding their sustainability performance in producing recycled wastewater, considering a ‘triple bottom line’ approach (i.e. economic, environmental, and social). These are powdered activated carbon adsorption coupled with ultrafiltration membrane separation (PAC-UF), reverse osmosis, ozone/ultraviolet-light oxidation and heterogeneous photo-catalysis coupled with low-pressure membrane separation (photocatalytic membrane reactor, PMR). The participatory method called simple multi-attribute rating technique exploiting ranks was employed for assigning weights to selected sustainability indicators. This sustainability assessment approach resulted in the development of a composite index as a final metric, for each WWT technology evaluated. The PAC-UF technology appears to be the most appropriate technology, attaining the highest composite value regarding the sustainability performance. A scenario analysis confirmed the results of the original scenario in five out of seven cases. In parallel, the PMR was highlighted as the technology with the least variability in its performance. Nevertheless, additional actions and approaches are proposed to strengthen the objectivity of the final results.

By employing fluids of the same composition (regarding foulants and ionic strength) specific fouling resistances are measured in the constant flux filtration mode, in dead-end UF and cross-flow RO tests, and compared with the respective resistances determined under constant pressure. Sodium alginate (SA) alone and mixtures with humic acids (HA), in presence of calcium, were used as typical organic foulants in feedwaters of two ionic strengths (TDS = 500 and 2000 mg/L). Plotting resistances α , from UF (or RO) constant pressure and constant flux tests, versus pressure drop across the cake Δ P c , estimated at the same deposited mass density, very good agreement was obtained; increasing resistance α with Δ P c was observed due to cake compressibility, as expected. Regarding cake structure, the dominant role of alginates, in proportion greater than 25% in mixtures with HA, is reflected in the resistances, which exhibit approximately the same compressibility index. Moreover, at the higher alginate proportions and for the greater salinity (TDS 2000 mg/L), the resistances from both UF and RO tests are correlated satisfactorily, versus Δ P c , by a single expression. These results significantly contribute towards developing reliable tools for assessing water fouling potential and predicting RO membrane fouling evolution in desalination plants.

This study aims to adequately characterize shear stresses and mass transfer coefficients at the membrane surface of a high pressure stirred filtration cell, and to determine equivalent conditions for flow in narrow channels with spacers. Employing electro-chemical techniques, measurements were made of local wall stresses and transfer coefficients, and of space-average mass transfer rates, at various Schmidt numbers and Reynolds numbers 6500–39,000. The local measurements revealed salient features of momentum and mass transfer at the cell base, and the average quantities were successfully correlated with the respective Reynolds numbers. Furthermore, from detailed numerical and experimental studies of cross-flow in narrow spacer-filled channels (simulating conditions in spiral-wound membrane modules), relevant data on wall shear stress and mass transfer were selected for comparison with the new results from the stirred-cell. This comparative data analysis allows estimation of the stirrer rotational speed ω which corresponds to equal area-average shear stress (or mass transfer coefficient) at the membrane surface of stirred cell and of spacer-filled channel (for specific cross flow velocity U). The range of ω values, corresponding to a certain U range, is different for the cases of matching mean shear stress or mass transfer rates in the two flow systems.

Local experimental data on key process parameters, throughout membrane sheets of spiral-wound membrane (SWM) elements, are unavailable and very difficult to obtain; therefore, theoretical models are essential to better understand the evolution of fouling, which is helpful for improving design and operation of these modules. The mathematical model, formulated in a fundamental manner herein, takes into account the spatial flow field non-uniformities inside narrow channels with spacers (simulating SWM elements) and their interaction with the evolving deposits. A realistic meso-scale transient model, typical of the early stages of organic membrane fouling, is employed. Appropriate non-dimensionalization of model equations allows to treat separately the two (practically significant) cases of constant inlet pressure and constant average permeate flux. Several simplified cases are examined, for specific dimensionless parameter values, and asymptotic solutions are derived. The results of this analysis have guided the development of very efficient numerical techniques, most useful for future extensions towards the development of a comprehensive modeling tool, appropriate for module design calculations and for assessing detailed experimental data.

Modeling incipient crystallization (“scaling”) in desalination membrane modules is a very difficult task due to several complications arising from the interplay of physico-chemical solution conditions (leading to supersaturation) with the flow field and related transport processes, including solid phase generation phenomena and membrane surface geometrical changes caused by the developing discrete particles. Although eventually all these aspects must be included in a comprehensive process model, it is fruitful to isolate and tackle them separately, thereby improving our understanding and developing techniques which will facilitate the ensuing synthesis of an integrated modeling framework. The focus in this work is on solid phase generation phenomena accounting for the membrane surface geometrical changes. A mean field model is developed that includes bulk and surface particle nucleation and growth processes. The relative importance of the two types of processes is analyzed. It is shown that, if thick concentration boundary layers exist around surface particles, the mean field theory—although not strictly valid—can be approximately used to estimate the transport coefficients, in conjunction with a unit cell problem for transport processes around a single surface particle. The unit cell problem is formulated and typical results for the flow and concentration field therein are presented as well as the corresponding mass transfer coefficients.

The scope of this paper is to systematically examine the long-term filtration performance of a laboratory scale MBR pilot, under nine different operating states, for elucidating the role of specific organic fractions of the mixed liquor. It is shown that the Solids Retention Time (SRT) is a major parameter influencing the concentration of both Dissolved Organic Matter (DOM) – mainly comprising Soluble Microbial Products (SMP) – and bound Extracellular Polymeric Substances (bEPS). Increasing SRT results in a noticeable reduction of both total and biodegradable fraction of DOM; at high SRT, the total bEPS concentration slightly increases whereas the bEPS biodegradable fraction decreases. The results underpin a close interrelation between DOM and bEPS, which are apparently under a dynamic equilibrium; biosorption phenomena are suggested as a mechanism significantly influencing this equilibrium. Membrane filtration performance has been examined over a wide range of SRT and HRT values. Temporal profiles of reversible and irreversible filtration resistance have been evaluated in connection with DOM and bEPS concentration data. It appears that low DOM concentration and bEPS content inhibit the self-accelerating phenomena leading to a Trans-membrane Pressure (TMP) jump, thus permitting a longer sustainable filtration operation; possible mechanisms are also suggested. This study provides useful data and ample evidence on the critical effect of DOM and bEPS on membrane fouling, helpful for further investigations and eventually for MBR operation.

Significant gaps exist in our knowledge of ultrafiltration (UF) membrane fouling, due to mixtures of poly-saccharides and proteins, despite a fair amount of related research. To get new insights into fouling layer characteristics, experiments were performed under constant-flux, within the range of practical interest (15⁻90 L/m2h), with typical polysaccharides (sodium alginate, SA), proteins (bovine serum albumin, BSA) as well as their mixtures in various proportions (1:3, 1:1, 3:1), and total organic matter concentration of 30 mg/L. The feed-water salinity and calcium ion concentration were 2000 mg/L NaCl and 2 mM, respectively. The temporal evolution of such fouling layers on flat-sheet membranes was monitored by recording the trans-membrane pressure variation. The results show that the specific fouling resistance α is strongly affected by flux, and the fouling propensity of polysaccharide-protein mixtures is significantly enhanced compared to single foulants, i.e., when BSA and SA are alone. The fouling layers are compressible and their resistance α tends to increase with the mass ratio of alginate in the mixture, particularly at high fluxes. To quantify these effects, correlations are presented of the initial fouling resistance αi with permeate flux J and of the evolution of α. R&D priorities are suggested on this topic of mixed foulants.

Membrane bioreactor technology is an attractive approach for combined wastewater treatment and water reclamation. Although considerable practical experience and experimental data are available, regarding the operation of MBR, mathematical models that would be valuable for improved design and optimization of MBR systems are still at an unsatisfactory state. This paper presents a critical review of existing activated sludge mathematical models and variations thereof, with emphasis on the special requirements which arise from the strong interaction of the concurrently occurring biological process and membrane filtration in MBR systems. The desirable attributes of an activated sludge model, focused on MBR technology, emerge by assessing the influence of this interrelation on both processes. Special attention is paid to identifying key variables that can help establish a direct link between bioprocess modeling and membrane fouling models. Various activated sludge models, originally developed for, and commonly used in the conventional activated sludge systems, as well as some modified versions employed in MBR systems, are reviewed. It is pointed out that these models, by design, have inherent deficiencies for MBR simulation mainly because they do not provide sufficient data to comprehensively simulate the fate of different microbial products (like EPS and SMP) that play an important role in membrane fouling. The necessity of an alternative bioprocess modeling framework, de novo focused on the specific needs of the MBR technology, as well as guidelines concerning the development of such a model are suggested in the paper.

Key parameters in the design of spiral wound elements are the geometrical characteristics of net-type spacers at the feed-flow channels, which greatly influence the transport phenomena therein and the overall operation of these modules. Recent results from a comprehensive theoretical and experimental study, on the fl ow characteristics and mass transfer in narrow channels with spacers, summarized in this paper, provide basic information that enables optimization of spacer characteristics. The new results include correlations of pressure drop and of average mass transfer rates to the wall as well as distributions of local shear stress and mass transfer, as a function of Reynolds and Schmidt number and of spacer geometry. An assessment of spacers, based on the above results, indicate trends, regarding geometric parameter values, leading to improved membrane element performance. Numerical results are also presented of an example, involving a single element in a RO seawater desalination plant, which demonstra...

Techniques for assessment of the colloidal fouling propensity of RO membrane feed waters are usually based on filtration tests with less tight membranes. In an effort towards improving such techniques, prototype experiments are performed under constant pressure, employing the same feed waters, with both RO and UF membranes. Fouling species representative of two typical categories of RO membrane foulants are employed, i.e. humic acids and polysaccharides (organic) as well as colloidal iron oxide (inorganic), in the concentration range 0.75–20 mg/L, at various fluid salinities. The RO fouling experiments confirm that there exists a linear dependence of fouling rates on foulant concentration, which permits extrapolation of the new results to lower concentrations, usually encountered in practice. The fouling behavior is analyzed in an effort to relate RO and UF test results. It is found that the specific cake resistances of deposited colloidal iron are much smaller in dead-end UF tests compared to those obtained in RO cross-flow experiments. This difference is smaller in the case of humic acids, whereas UF and RO specific cake resistances are in close agreement for alginate foulants, when corrected for incomplete species rejection by UF and deposition on RO membranes. Four different factors to account for the observed differences, namely pressure , UF membrane rejection , fluid shear , and the effect of cake-enhanced concentration polarization , are examined using additional targeted experiments and theoretical tools. It is concluded that, in the range of conditions tested, the first three factors are important and need to be carefully considered in a systematic way in order to obtain reliable predictions of RO fouling on the basis of UF filtration data.

The very significant impact of European legislation (Directive 91/414/EEC) on the authorization of plant protection products is reviewed herein, which has resulted in withdrawal of 704 active substances (AS) out of 889 assessed so far. The list of currently approved 276 AS includes 194 AS “existing” in the market before 1993 and 82 “new” AS introduced during the last 15 years. Results of toxicity characterization of the approved AS are also summarized, utilizing several well-known databases. Although significant data gaps exist for a rather large part of the approved AS, it is found that 84 AS are positive for at least one health effect (after chronic and/or acute exposure) including carcinogenicity, reproductive and neuro-developmental disorders, as well as endocrine disruption. The toxicity characterization results of this study are compared to those of recent assessments by other organizations (KemI, the Swedish Chemicals Agency, and the Pesticide Safety Directorate of the UK), where interpretation and use is made of AS “cut-off” criteria foreseen in new EU legislation. These studies report a comparatively smaller AS number with positive toxicity characterization. The possibility of some additional AS withdrawal in the near future, combined with the rather small rate of new AS introduction (approx. 5 per year) suggest that the list of approved AS over the next 10–15 years may not change very drastically. Consideration of the above trends is necessary and instructive in evaluating results of existing health impact assessment (HIA) studies, as well as in planning new ones. Due to the very drastic change in the number and type of marketed AS, that took place within the past 8–9 years, it is suggested that new HIA studies (based on epidemiological data after year 2000) should focus on a rather short time frame and, therefore, on appropriate cohort groups, e.g. young children. For the same reason, results of epidemiological studies of the past (involving banned AS) should be carefully interpreted and used with caution. Keywords: Pesticides, EU legislation, Health impact assessment, Directive 91/414/EEC

The influence of different types of humic substances (HS) is studied, in the presence or absence of calcium ions, on the retention of two triazines, atrazine and prometryn. To correlate the retention of triazines with the characteristic chemical parameters of the humic substances, three standard humic materials from the International Humic Substances Society (IHSS) as well as a NOM surrogate, tannic acid, are employed. Differences are observed, in the triazine retention behaviour, between organic-free triazine solutions and those of mixed HS–triazines. The results suggest that physicochemical interactions alone, between triazines and humic substances, cannot fully account for the observed triazine retention by nanofiltration. The HS acidity (total, carboxylic and phenolic) exhibits a weak correlation with triazine rejection. However, the conformation of humic species appears to play a more significant role. The presence of calcium ions improves both HS–triazine association and membrane rejection of the HS–triazine complexes. In support of the membrane filtration results, gel permeation chromatography shows that the hydrophobic interactions can explain the different triazine absorbing capacities of the HS used. In general, the results illustrate the complexity of the adsorption and retention mechanisms of triazines in the presence of humic substances, and confirm the important role of membrane fouling on process efficiency.