J.S. (Hans) Vrouwenvelder

Spiral-wound membrane modules used in water treatment for water reuse and desalination make use of spacer meshes for keeping the membrane leaves apart and for enhancing the mass transfer. Computational fluid dynamics (CFD) has gained importance in the design of new spacers with optimized hy-drodynamic characteristics, but this requires a precise description of the spacer geometry. This study developed a method to obtain accurate three-dimensional (3-D) geometry representations for any given spacer design from X-ray computed tomography (CT) scans. The method revealed that the filaments of industrial spacers have a highly variable cross-section size and shape, which impact the flow characteristics in the feed channel. The pressure drop and friction factors were calculated from numerical simulations on five commercially available feed spacers used in practice. Model solutions compared well to experimental data measured using a flow cell for average velocities up to 0.2 m/s, as used in industrial reverse osmosis and nanofiltration membrane operations. A newly-proposed spacer geometry with alternating strand thickness was tested, which was found to yield a lower pressure drop while being highly efficient in converting the pumping power into membrane shear. Numerical model solutions using CFD with geometries from CT scans were closer to measurements than those obtained using the traditional circular cross-section strand simplification, indicating that CT scans are very well suitable to approximate real feed spacer geometries. By providing detailed insight on the spacer filament shape, CT scans allow better quantification of local distribution of velocity and shear, possibly leading to more accurate estimations of fouling and concentration polarization.

Biocides may be used to control biofouling in spiral-wound reverse osmosis (RO) and nanofiltration (NF) systems. The objective of this study was to investigate the effect of biocide 2,2-dibromo-3-ni-trilopropionamide (DBNPA) dosage on biofouling control. Preventive biofouling control was studied applying a continuous dosage of substrate (0.5 mg/L) and DBNPA (1 mg/L). Curative biofouling control was studied on pre-grown biofilms, once again applying a continuous dosage of substrate (0.5 mg acetate C/L) and DBNPA (1 and 20 mg/L). Biofouling studies were performed in membrane fouling simulators (MFSs) supplied with biodegradable substrate and DBNPA. The pressure drop was monitored in time and at the end of the study, the accumulated biomass in MFS was quantified by adenosine triphosphate (ATP) and total organic carbon (TOC) analysis. Continuous dosage of DBNPA (1 mg/L) prevented pressure drop increase and biofilm accumulation in the MFSs during a run time of 7 d, showing that biofouling can be managed by preventive DBNPA dosage. For biofouled systems , continuous dosage of DBNPA (1 and 20 mg/L) inactivated the accumulated biomass but did not restore the original pressure drop and did not remove the accumulated inactive cells and extracellular polymeric substances (EPS), indicating DBNPA dosage is not suitable for curative biofouling control.

Surface coating of membranes may be a promising option to control biofilm development and biofoul-ing impact on membrane performance of spiral-wound reverse osmosis (RO) systems. The objective of this study was to investigate the impact of an amphiphilic copolymer coating on biofilm formation and biofouling control. The coating was composed of both hydrophilic and hydrophobic monomers hydroxyethyl methacrylate (HEMA) and perfluorodecyl acrylate (PFA), respectively. Commercial RO membranes were coated with HEMA-PFA copolymer film. Long and short term biofouling studies with coated and uncoated membranes and feed spacer were performed using membrane fouling simulators (MFSs) operated in parallel, fed with water containing nutrients. For the long-term studies pressure drop development in time was monitored and after eight days the MFSs were opened and the accumulated biofilm on the membrane and spacer sheets was quantified and characterized. The presence of the membrane coating was determined using X-ray photoelectron spectroscopy (XPS) and Fourier transform infrared spectroscopy (FTIR). Results showed that the amphiphilic coating (i) delayed biofouling (a lower pressure drop increase by a factor of 3 and a lower accumulated active biomass amount by a factor of 6), (ii) influenced the biofilm composition (23% lower polysaccharides and 132% higher protein content) and (iii) was still completely present on the membrane at the end of the biofouling study, showing that the coating was strongly attached to the membrane surface. Using coated membranes and feed spacers in combination with advanced cleaning strategies may be a suitable way to control biofouling.

The porosity of spacer-filled feed channels influences the hydrodynamics of spiral-wound membrane systems and impacts the overall performance of the system. Therefore, an exact measurement and a detailed understanding of the impact of the feed channel porosity is required to understand and improve the hydrodynamics of spiral-wound membrane systems applied for desalination and wastewater reuse. The objectives of this study were to assess the accuracy of porosity measurement techniques for feed spacers differing in geometry and thickness and the consequences of using an inaccurate method on hydrodynamic predictions, which may affect permeate production. Six techniques were applied to measure the porosity namely, three volumetric techniques based on spacer strand count together with a cuboidal (SC), cylindrical (VCC) and ellipsoidal volume calculation (VCE) and three independent techniques based on volume displacement (VD), weight and density (WD) and computed tomography (CT) scanning. The CT method was introduced as an alternative for the other five already existing and applied methods in practice. Six feed spacers used for the porosity measurement differed in filament thickness, angle between the filaments and mesh-size. The results of the studies showed differences between the porosities, measured by the six methods. The results of the microscopic techniques SC, VCC and VCE deviated significantly from measurements by VD, WD and CT, which showed similar porosity values for all spacer types. Depending on the maximum deviation of the porosity measurement techniques from À6% to þ6%, (i) the linear velocity deviations were À5.6% and þ6.4% respectively and (ii) the pressure drop deviations were À31% and þ43% respectively, illustrating the importance of an accurate porosity measurement. Because of the accuracy and standard deviation, the VD and WD method should be applied for the porosity determination of spacer-filled channels, while the CT method is recommended for numerical modelling purposes. The porosity has a linear relationship with the flow velocity and a superlinear effect on the pressure drop. Accurate porosity data are essential to evaluate feed spacer performance in spiral-wound membrane systems. Porosity of spacer-filled feed channels has a strong impact on membrane performance and biofouling impact.

Spiral-wound membrane modules used in water treatment for water reuse and desalination make use of spacer meshes for keeping the membrane leaves apart and for enhancing the mass transfer. Computational fluid dynamics (CFD) has gained importance in the design of new spacers with optimized hy-drodynamic characteristics, but this requires a precise description of the spacer geometry. This study developed a method to obtain accurate three-dimensional (3-D) geometry representations for any given spacer design from X-ray computed tomography (CT) scans. The method revealed that the filaments of industrial spacers have a highly variable cross-section size and shape, which impact the flow characteristics in the feed channel. The pressure drop and friction factors were calculated from numerical simulations on five commercially available feed spacers used in practice. Model solutions compared well to experimental data measured using a flow cell for average velocities up to 0.2 m/s, as used in industrial reverse osmosis and nanofiltration membrane operations. A newly-proposed spacer geometry with alternating strand thickness was tested, which was found to yield a lower pressure drop while being highly efficient in converting the pumping power into membrane shear. Numerical model solutions using CFD with geometries from CT scans were closer to measurements than those obtained using the traditional circular cross-section strand simplification, indicating that CT scans are very well suitable to approximate real feed spacer geometries. By providing detailed insight on the spacer filament shape, CT scans allow better quantification of local distribution of velocity and shear, possibly leading to more accurate estimations of fouling and concentration polarization.

Understanding the factors that determine the spatial and temporal biofilm development is a key to formulate effective control strategies in reverse osmosis membrane systems for desalination and wastewater reuse. In this study, biofilm development was investigated at different water temperatures (10, 20, and 30 C) inside a membrane fouling simulator (MFS) flow cell. The MFS studies were done at the same crossflow velocity with the same type of membrane and spacer materials, and the same feed water type and nutrient concentration, differing only in water temperature. Spatially resolved biofilm parameters such as oxygen decrease rate, biovolume, biofilm spatial distribution, thickness and composition were measured using in-situ imaging techniques. Pressure drop (PD) increase in time was used as a benchmark as to when to stop the experiments. Biofilm measurements were performed daily, and experiments were stopped once the average PD increased to 40 mbar/cm. The results of the biofouling study showed that with increasing feed water temperature (i) the biofilm activity developed faster, (ii) the pressure drop increased faster, while (iii) the biofilm thickness decreased. At an average pressure drop increase of 40 mbar/cm over the MFS for the different feed water temperatures, different biofilm activities, structures, and quantities were found, indicating that diagnosis of biofouling of membranes operated at different or varying (seasonal) feed water temperatures may be challenging. Membrane installations with a high temperature feed water are more susceptible to biofouling than installations fed with low temperature feed water.

Membrane surface hydrophilic modification has always been considered to mitigating biofouling in membrane bioreactors (MBRs). Four hollow-fiber ultrafiltration membranes (pore sizes ~0.1 mm) differing only in hydrophobic or hydrophilic surface characteristics were operated at a permeate flux of 10 L/m 2 h in the same lab-scale MBR fed with synthetic wastewater. In addition, identical membrane modules without permeate production (0 L/m 2 h) were operated in the same lab-scale MBR. Membrane modules were autopsied after 1, 10, 20 and 30 days of MBR operation, and total extracellular polymeric substances (EPS) accumulated on the membranes were extracted and characterized in detail using several analytical tools, including conventional colorimetric tests (Lowry and Dubois), liquid chromatography with organic carbon detection (LC-OCD), fluorescence excitation-emission matrices (FEEM), fourier transform infrared (FTIR) and confocal laser scanning microscope (CLSM). The transmembrane pressure (TMP) quickly stabilized with higher values for the hydrophobic membranes than hydrophilic ones. The sulfonated polysulfone (SPSU) membrane had the highest negatively charged membrane surface, accumulated the least amount of foulants and displayed the lowest TMP. The same type of organic foulants developed with time on the four membranes and the composition of biopolymers shifted from protein dominance at early stages of filtration (day 1) towards polysaccharides dominance during later stages of MBR filtration. Nonmetric multidimensional scaling of LC-OCD data showed that biofilm samples clustered according to the sampling event (time) regardless of the membrane surface chemistry (hydrophobic or hydrophilic) or operating mode (with or without permeate flux). These results suggest that EPS composition may not be the dominant parameter for evaluating membrane performance and possibly other parameters such as biofilm thickness, porosity, compactness and structure should be considered in future studies for evaluating the development and impact of biofouling on membrane performance.

Feed spacers are important for the impact of biofouling on the performance of spiral-wound reverse osmosis (RO) and nanofiltration (NF) membrane systems. The objective of this study was to propose a strategy for developing, characterizing, and testing of feed spacers by numerical modeling, three-dimensional (3D) printing of feed spacers and experimental membrane fouling simulator (MFS) studies. The results of numerical modeling on the hydrodynamic behavior of various feed spacer geometries suggested that the impact of spacers on hydrodynamics and biofouling can be improved. A good agreement was found for the modeled and measured relationship between linear flow velocity and pressure drop for feed spacers with the same geometry, indicating that modeling can serve as the first step in spacer characterization. An experimental comparison study of a feed spacer currently applied in practice and a 3D printed feed spacer with the same geometry showed (i) similar hydrodynamic behavior, (ii) similar pressure drop development with time and (iii) similar biomass accumulation during MFS biofouling studies, indicating that 3D printing technology is an alternative strategy for development of thin feed spacers with a complex geometry. Based on the numerical modeling results, a modified feed spacer with low pressure drop was selected for 3D printing. The comparison study of the feed spacer from practice and the modified geometry 3D printed feed spacer established that the 3D printed spacer had (i) a lower pressure drop during hydrodynamic testing, (ii) a lower pressure drop increase in time with the same accumulated biomass amount, indicating that modifying feed spacer geometries can reduce the impact of accumulated biomass on membrane performance. The combination of numerical modeling of feed spacers and experimental testing of 3D printed feed spacers is a promising strategy (rapid, low cost and representative) to develop advanced feed spacers aiming to reduce the impact of biofilm formation on membrane performance and to improve the cleanability of spiral-wound NF and RO membrane systems. The proposed strategy may also be suitable to develop spacers in e.g. forward osmosis (FO), reverse electrodialysis (RED), membrane distillation (MD), and electrodeionisation (EDI) membrane systems.

Microbial processes inevitably play a role in membrane-based desalination plants, mainly recognized as membrane biofouling. We assessed the bacterial community structure and diversity during different treatment steps in a full-scale seawater desalination plant producing 40,000 m 3 /d of drinking water. Water samples were taken over the full treatment train consisting of chlorination, spruce media and cartridge filters, de-chlorination, first and second pass reverse osmosis (RO) membranes and final chlorine dosage for drinking water distribution. The water samples were analyzed for water quality parameters (total bacterial cell number, total organic carbon, conductivity, pH, etc.) and microbial community composition by 16S rRNA gene pyrosequencing. The planktonic microbial community was dominated by Proteobacteria (48.6%) followed by Bacteroidetes (15%), Firmicutes (9.3%) and Cyanobacteria (4.9%). During the pretreatment step, the spruce media filter did not impact the bacterial community composition dominated by Proteobacteria. In contrast, the RO and final chlorination treatment steps reduced the Proteobacterial relative abundance in the produced water where Firmicutes constituted the most dominant bacterial group. Shannon and Chao1 diversity indices showed that bacterial species richness and diversity decreased during the seawater desalination process. The two-stage RO filtration strongly reduced the water conductivity (>99%), TOC concentration (98.5%) and total bacterial cell number (>99%), albeit some bacterial DNA was found in the water after RO filtration. About 0.25% of the total bacterial operational taxonomic units (OTUs) were present in all stages of the desalination plant: the seawater, the RO permeates and the chlorinated drinking water, suggesting that these bacterial strains can survive in different environments such as high/low salt concentration and with/without residual disinfectant. These bacterial strains were not caused by contamination during water sample filtration or from DNA extraction protocols. Control measurements for sample contamination are important for clean water studies.

Biofouling is a serious problem in reverse osmosis/nanofiltration (RO/NF) applications, reducing membrane performance. Early detection of biofouling plays an essential role in an adequate anti-biofouling strategy. Presently, fouling of membrane filtration systems is mainly determined by measuring changes in pressure drop, which is not exclusively linked to biofouling. Non-destructive imaging of oxygen concentrations (i) is specific for biological activity of biofilms and (ii) may enable earlier detection of biofilm accumulation than pressure drop. The objective of this study was to test whether transparent luminescent planar O 2 optodes, in combination with a simple imaging system, can be used for early non-destructive biofouling detection. This biofouling detection is done by mapping the two-dimensional distribution of O 2 concentrations and O 2 decrease rates inside a membrane fouling simulator (MFS). Results show that at an early stage, biofouling development was detected by the oxygen sensing optodes while no significant increase in pressure drop was yet observed. Additionally, optodes could detect spatial heterogeneities in biofouling distribution at a micro scale. Biofilm development started mainly at the feed spacer crossings. The spatial and quantitative information on biological activity will lead to better understanding of the biofouling processes, contributing to the development of more effective biofouling control strategies.

Oxo-anion binding properties of the thermostable enzyme ferritin from Pyrococcus furiosus were characterized with radiography. Radioisotopes 32 P and 76 As present as oxoanions were used to measure the extent and the rate of their absorption by the ferritin. Ther-mostable ferritin proved to be an excellent system for rapid phosphate and arsenate removal from aqueous solutions down to residual concentrations at the picomolar level. These very low concentrations make thermostable ferritin a potential tool to considerably mitigate industrial biofouling by phosphate limitation or to remove arsenate from drinking water.

Long-term performance and fouling behavior of four full-scale nanofiltration (NF) plants, treating anoxic groundwater at 80% recovery for drinking water production, were characterized and compared with oxic NF and reverse osmosis systems. Plant operating times varied between 6 and 10 years and pretreatment was limited to 10 mm pore size cartridge filtration and antiscalant dosage (2–2.5 mg L À 1) only. Membrane performance parameters normalized pressure drop (NPD), normalized specific water permeability (K w) and salt retention generally were found stable over extended periods of operation (46 months). Standard acid–base cleanings (once per year or less) were found to be sufficient to maintain satisfying operation during direct NF of the described iron rich (r 8.4 mg L À 1) anoxic groundwaters. Extensive autopsies of eight NF membrane elements, which had been in service since the plant startup (6–10 years), were performed to characterize and quantify the material accumulated in the membrane elements. Investigations using scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS), total organic carbon (TOC) and adenosine triphosphate (ATP) measurements revealed a complex mixture of organic, biological and inorganic materials. The fouling layers that developed during half to one year of operation without chemical cleaning were very thin (o2 mm). Most bio(organic) accumulates were found in the lead elements of the installations while inorganic precipitates/deposits (aluminosilicates and iron(II)sulfides) were found in all autopsied membrane elements. The high solubility of reduced metal ions and the very slow biofilm development under anoxic conditions prevented rapid fouling during direct NF of the studied groundwaters. When compared to oxic NF and RO systems in general (e.g. aerated ground waters or surface waters), the operation and performance of the described anoxic installations (with minimal pretreatment) can be described as very stable.

Biofilm formation causes performance loss in spiral-wound membrane systems. In this study a microfiltration membrane was used in experiments to simulate fouling in spiral-wound reverse osmosis (RO) and nanofiltration (NF) membrane modules without the influence of concentration polarization. The resistance of a microfiltration membrane is much lower than the intrinsic biofilm resistance, enabling the detection of biofilm accumulation in an early stage. The impact of biofilm accumulation on the transmembrane (biofilm) resistance and feed channel pressure drop as a function of the crossflow velocity (0.05 and 0.20 m s À1) and feed spacer presence was studied in transparent membrane biofouling monitors operated at a permeate flux of 20 L m À2 h À1. As biodegradable nutrient, acetate was dosed to the feed water (1.0 and 0.25 mg L À1 carbon) to enhance biofilm accumulation in the monitors. The studies showed that biofilm formation caused an increased transmembrane resistance and feed channel pressure drop. The effect was strongest at the highest crossflow velocity (0.2 m s À1) and in the presence of a feed spacer. Simulating conditions as currently applied in nanofiltration and reverse osmosis installations (crossflow velocity 0.2 m s À1 and standard feed spacer) showed that the impact of biofilm formation on performance, in terms of transmembrane and feed channel pressure drop, was strong. This emphasized the importance of hydrodynamics and feed spacer design. Biomass accumulation was related to the nutrient load (nutrient concentration and linear flow velocity). Reducing the nutrient concentration of the feed water enabled the application of higher crossflow velocities. Pretreatment to remove biodegradable nutrient and removal of biomass from the membrane elements played an important part to prevent or restrict biofouling. w a t e r r e s e a r c h 5 0 (2 0 1 4) 2 0 0 e2 1 1 0043-1354/$ e see front matter ª

Biofouling in membrane bioreactors (MBRs) remains a primary challenge for their wider application, despite the growing acceptance of MBRs worldwide. Research studies on membrane fouling are extensive in the literature, with more than 200 publications on MBR fouling in the last 3 years; yet, improvements in practice on biofouling control and management have been remarkably slow. Commonly applied cleaning methods are only partially effective and membrane replacement often becomes frequent. The reason for the slow advancement in successful control of biofouling is largely attributed to the complex interactions of involved biological compounds and the lack of representative-for-practice experimental approaches to evaluate potential effective control strategies. Biofouling is driven by microorganisms and their associated extra-cellular polymeric substances (EPS) and microbial products. Microorganisms and their products convene together to form matrices that are commonly treated as a black box in conventional control approaches. Biological-based antifouling strategies seem to be a promising constituent of an effective integrated control approach since they target the essence of biofouling problems. However, journal h ome page: www.elsevier.com/loca te/watres w a t e r r e s e a r c h 4 7 (2 0 1 3) 5 4 4 7 e5 4 6 3

The influence of polydopamine-and polydopamine-graft-poly(ethylene glycol)-coated feed spacers and membranes, copper-coated feed spacers, and commercially-available biostatic feed spacers on biofouling has been studied in membrane fouling simulators. Feed spacers and membranes applied in practical membrane filtration systems were used; biofouling development was monitored by feed channel pressure drop increase and biomass accumulation. Polydopamine and polydopamine-g-PEG are hydrophilic surface modification agents expected to resist protein and bacterial adhesion, while copper feed spacer coatings and biocides infused in feed spacers are expected to restrict biological growth. Our studies showed that polydo-pamine and polydopamine-g-PEG coatings on feed spacers and membranes, copper coatings on feed spacers, and a commercial biostatic feed spacer did not have a significant impact on feed channel pressure drop increase and biofilm accumulation as measured by ATP and TOC content. The studied spacer and membrane modifications were not effective for biofouling control; it is doubtful that feed spacer and membrane modification , in general, may be effective for biofouling control regardless of the type of applied coating.

The impact of feed spacers on initial feed channel pressure (FCP) drop, FCP increase and biomass accumulation has been studied in membrane fouling simulators using feed spacers applied in commercially available nanofiltration and reverse osmosis spiral wound membrane modules. All spacers had a similar geometry. Our studies showed that biofouling was not prevented by (i) variation of spacer thickness, (ii) feed spacer orientation, (iii) feed spacer coating with silver, copper or gold and (iv) using a biostatic feed spacer. At constant feed flow, a lower FCP and FCP increase were observed for a thicker feed spacer. At constant linear flow velocity, roughly the same FCP development and biomass accumulation were found irrespective of the feed spacer thickness: hydrodynamics and substrate load were more important for development and impact of biofouling than the thickness of currently applied spacers. Use of biostatic and metal coated spacers were not effective for biofouling control. The same small reduction of biofouling rate was observed with copper and silver coated spacers as well as uncoated 45 • rotated spacers. The studied modified spacers were not effective for biofouling prevention and control. The impact of biofouling on FCP increase was reduced significantly by a lower linear flow velocity, while spacer orientation and spacer thickness in membrane modules had a smaller but still significant effect.

There is a strong need for techniques enabling direct assessment of biological activity of biofouling in membrane filtration systems. Here we present a new quantitative and non-destructive method for mapping O 2 dynamics in biofilms during biofouling studies in membrane fouling simulators (MFS). Transparent planar O 2 optodes in combination with a luminescence lifetime imaging system were used to map the two-dimensional distribution of O 2 concentrations and consumption rates inside the MFS. The O 2 distribution was indicative for biofilm development. Biofilm activity was characterized by imaging of O 2 consumption rates, where low and high activity areas could be clearly distinguished. The spatial development of O 2 consumption rates, flow channels and stagnant areas could be determined. This can be used for studies on concentration polarization, i.e. salt accumulation at the membrane surface resulting in increased salt passage and reduced water flux. The new optode-based O 2 imaging technique applied to MFS allows non-destructive and spatially resolved quantitative biological activity measurements (BAM) for on-site biofouling diagnosis and laboratory studies. The following set of complementary tools is now available to study development and control of biofouling in membrane systems: (i) MFS, (ii) sensitive pressure drop measurement, (iii) magnetic resonance imaging, (iv) numerical modelling, and (v) biological activity measurement based on O 2 imaging methodology.

Feed spacer channel pressure drop increase Reverse osmosis a b s t r a c t Phosphate limitation as a method to control biofouling of spiral wound reverse osmosis (RO) membranes was studied at a full-scale installation fed with extensively pretreated water. The RO installation is characterized by (i) a low feed channel pressure drop increase and (ii) low biomass concentrations in membrane elements at the installation feed side. This installation contrasted sharply with installations fed with less extensively pretreated feed water (and therefore higher phosphate concentrations) experiencing a high-pressure drop increase and high biomass concentrations in lead elements. Membrane fouling simulator (MFS) studies showed that low phosphate concentrations (w0.3 mg P L À1) in the feed water restricted the pressure drop increase and biomass accumulation, even at high substrate (organic carbon) concentrations. In the MFS under ortho-phosphate limiting conditions, dosing phosphonate based antiscalants caused biofouling while no biofouling was observed when acids or phosphonate-free antiscalants were used. Antiscalant dosage could increase both phosphate and substrate concentrations of the water. Therefore, antiscalant selection may be critical for biofouling control. Since no biofouling was observed at low phosphate concentrations, restricting biomass growth by phosphate limitation may be a feasible approach to control biofouling, even in the presence of high organic carbon levels.

A three-dimensional (3D) computational model describing fluid dynamics and biofouling of feed channels of spiral wound reverse osmosis and nanofiltration membrane systems was developed based on results from practice and experimental studies. In the model simulations the same feed spacer geometry as applied in practice and the experimental studies was used. The 3D mathematical model showed the same trends for (i) feed channel pressure drop, (ii) biomass accumulation, (iii) velocity distribution profile, resulting in regions of low and high liquid flow velocity also named channeling. The numerical model predicted a dominant biomass growth on the feed spacer, consistent with direct in situ observations on biofouling of spiral wound membrane modules and monitors using Magnetic Resonance Imaging (MRI). The model confirms experimental results that feed spacer fouling is more important than membrane fouling. The paper shows that mathematical modeling techniques have evolved to a stage that they can be used hand-in-hand with experiments to understand the processes involved in membrane fouling.

Water treatment in The Netherlands has developed to an outstanding level. Nevertheless, new challenges, such as protozoa (e.g. Cryptosporidium), Legionella, endocrine disrupting compounds and pharmaceuticals, have to be faced. These challenges can have a negative effect on customer confidence and the overall reputation of water supply. The Dutch water supply companies therefore have set a new ambition for water quality. To realize this ambition an integral approach to water treatment involving new technologies as membrane filtration and UV is necessary. The paper will present the possibilities of the current level of technology, how it relates to the vision and ambition of the Dutch drinking water sector and the role of research – especially in UV technology and membrane filtration – in realizing long-term goals. A panel of Dutch R&D directors of water supply companies have agreed on a new ambition for water quality. The focus of this new ambition is to deliver an impeccable water quality at the tap (Van Dijk and Van der Kooij, 2004). This goal can only be reached by an integral approach, starting at the source, followed by a multi-barrier treatment, and finally a flawless distribution system where water quality does not deteriorate. In order to produce this impeccable water sustainable technology should be used. Typical characteristics of the future water quality are: hygienically safe, free of particles and regrowth potential (Legio-nella), free of trace pollutants and good taste and odor.

Biofouling is a frequently occurring fouling mechanism in membrane applications. Controlling this phenomenon is a challenge due to the difficulty of cleaning biofouling in spiral wound membrane elements. It is assumed that cleaning can be more efficient when biofouling is in an early stage of colonisation. Therefore a sensitive method has to be available for an early identification of biofouling. The present method, the measurement of the normalised pressure drop (NPD) is not specific for biofouling and is not very sensitive. In this research the feasibility of the specific oxygen consumption rate for detection of the activity of biofilms was investigated in membrane systems. The method has the advantages to be specific for active biomass, applicable in situ, non-destructive and more sensitive than NPD. Three experiments demonstrated that the measurement of the rate of oxygen consumption is potentially a simple, reliable method for the measurement of the active biomass in membrane systems. During one of the experiments the method illustrated the effect of cleaning and the regrowth of bacteria afterwards. The method will be further evaluated and standardized. The relation between the specific oxygen consumption rate and the condition of the biofilm, measured by autopsy of membrane elements, will be further explored. The ongoing research will result in an apparatus and procedure, to be used for biofouling identification during the operation of full-scale and pilot plants.